Skip to main content

Full text of "Protein split products in relation to immunity and disease"

See other formats 


MR 


V 


PROTEIN  SPLIT  PRODUCTS 


IN 


BY 

VICTOR  C.  VAUGHAN,  M.D.,  LL.D. 

DEAN    OF    THE    DEPARTMENT    OF    MEDICINE    AND    SURGERY    OF    THE     UNIVERSITY    OF 

MICHIGAN 


VICTOR  C.  VAUGHAN,  JR.,  M.D,  A.B. 

IN  CHARGE  OF  THE  TUBERCULOSIS  WORK  OF  THE  DETROIT  BOARD  OF  HEALTH 
JUNIOR  ATTENDING  PHYSICIAN  TO  HARPER  HOSPITAL,  DETROIT 


J.  WALTER  VAUGHAN,  M.D.,  A.B. 

JUNIOR  ATTENDING  SURGEON  TO  HARPER    HOSPITAL,    DETROIT 

ILLUSTRATED 


LEA  &  FEBIGER 

PHILADELPHIA    AND    NEW    YORK 
1913 


Entered  according  to  the  Act  of  Congress,  in  the  year  1913,  by 

LEA   &  FEBIGER, 
in  the  Office  of  the  Librarian  of  Congress.      All  rights  reserved. 


PREFACE 


THE  investigations  recorded  in  this  volume,  begun 
nearly  fifteen  years  ago,  were  inaugurated  in  consequence 
of  certain  fundamental  ideas  or  theories  held  by  the  writer, 
and  these  have  directed  and  dominated  all  our  labors  along 
this  line.  The  purpose  of  this  work  has  been  to  solve 
scientific  problems,  rather  than  to  discover  practical 
applications.  The  latter,  so  far  as  they  have  in  any  way 
influenced  our  studies  or  even  received  our  attention,  have 
been  only  incidental.  Quite  naturally  our  theories  have 
been  more  or  less  modified,  and  have  developed  as  the  work 
has  progressed,  but  the  essentials  and  fundamentals  have 
not  been  materially  altered.  From  time  to  time  these 
theories  have  been  given  in  more  or  less  detail,  notably 
in  an  address  at  the  opening  of  the  Medical  Department  of 
the  University  of  Toronto  in  1905  (Canadian  Jour,  of 
Med.  and  Surg.,  xviii,  283)  and  in  the  Shattuck  lecture  for 
1906  (Boston  Med.  and  Surg.  Jour.,  civ,  215).  However, 
it  may  be  well  to  restate  briefly  the  original  conceptions 
which  impelled  us  to  begin  and  continue  these  studies. 

The  only  essential  and  constant  difference  between 
living  and  non-living  matter  is  that  within  the  molecules 
of  the  former  there  is  constant  metabolism,  while  in  the 
latter  no  such  process  operates.  We  are  to  conceive  of 
the  living  molecules  as  made  up  of  numerous  atoms,  and 
each  atom  surrounded  by  its  electrons;  atoms  and  elec- 


IV  PREFACE 

irons  in  ceaseless  motion,  and  groups  of  atoms  being  con- 
stantly cast  out  of  the  molecule  and  replaced  by  new  groups 
split  off  from  outside  matter.  As  soon  as  a  molecule  becomes 
the  seat  of  assimilation  and  excretion,  it  is  no  longer  dead; 
it  lives.  As  a  result  of  assimilation  it  acquires  the  property 
of  building  up  its  own  structure;  then  polymerization 
follows  and  reproduction  in  its  simplest  form  begins.  The 
one  phenomenon  always  manifested  by  living  matter,  and 
never  exhibited  by  non-living  matter,  is  metabolism. 

When  matter  becomes  endowed  with  life  it  does  not 
cease  to  be  matter;  it  does  not  lose  its  inherent  properties; 
it  is  not  released  from  the  laws  that  govern  its  structure, 
its  attractions,  and  its  motions.  In  studying  living  things 
it  should  be  borne  in  mind  that  they  are  material  in  com- 
position and  subject  to  the  fundamental  laws  that  govern 
matter,  and  possessed  of  those  properties  essential  to 
matter. 

Matter  is  alive  when  it  feeds  and  excretes.  The  living 
molecule  not  only  absorbs;  it  assimilates.  It  chemically 
alters  what  it  absorbs,  and  within  limits,  it  may  be  altered 
by  what  it  absorbs.  Atomic  groups  taken  into  living 
molecules  enter  into  new  combinations.  The  living  mole- 
cule is  not  stabile,  but  is  highly  labile.  Its  composition  is 
never  constant,  and  it  is  never  in  a  condition  of  equilibrium. 
It  is  in  constant  chemical  reaction  with  outside  matter. 
Apart  from  other  matter  it  could  not  exist.  There  is  a 
constant  interchange  of  atoms  between  it  and  external 
matter.  A  condition,  best  designated  as  latent  life,  may 
exist  without  interchange  of  atoms.  This  is  exemplified 
in  spores,  seeds,  and  ova.  Matter  existing  in  these  forms 
may  be  awakened  into  activity  by  proper  stimuli;  active 
life  begins  with  the  interchange  of  atoms. 


PREFACE  V 

Why  is  there  this  constant  change  of  atomic  groups 
between  the  living  molecule  and  outside  matter?  It  is 
for  the  purpose  of  supplying  the  living  molecule  with 
energy.  It  is  probable  that  in  the  absorption  of  energy 
by  the  living  molecule,  oxygen  is  released  from  its  combina- 
tion with  carbon  or  hydrogen,  and  is  attached  to  nitrogen, 
while  in  the  liberation  of  energy  the  reverse  takes  place. 
Nitrogen  seems  to  be  the  master  element  within  the  living 
molecule.  It  is  by  virtue  of  its  chemism  that  groups  are 
torn  from  non-living  matter,  taken  into  the  living  molecule, 
and  assimilated  by  atomic  rearrangement;  and  furthermore, 
it  is  on  account  of  the  lability  of  the  compound  thus  formed 
that  potential  energy  is  converted  into  kinetic  and  work 
is  accomplished.  A  nitrogen  side-chain  serves  as  a  receptor 
and  transmitter  of  oxygen,  and  thus  the  traffic  in  energy 
within  the  living  molecule  goes  on  rhythmically.  It  is  not 
to  be  supposed  that  the  nitrogen  side-chain,  which  serves 
as  the  receptor  and  transmitter  of  oxygen,  consists  of  so 
simple  a  body  as  nitrogen  or  nitrogen  oxide,  but  it  is  probably 
a  highly  complex  nitrogenous  body  in  which  the  location 
of  the  nitrogen  is  central,  as  suggested  by  Allen.  Nor  is  it 
probable  that  only  oxygen  is  broken  off  from  the  pabulum, 
but  substances  containing  this  element.  This  is  the  way 
in  which  the  living  molecule  keeps  up  its  constant,  rhythmic 
traffic  in  energy,  absorbing  heat  by  assimilation,  and  giving 
it  off  by  dissociation.  Each  living  molecule  has  not  only 
one,  but  many  of  these  nitrogenous  groups  that  act  as 
receptors.  Moreover,  metabolism  within  the  molecule  is 
not  confined  to  the  absorption  of  oxygen,  and  the  casting 
out  of  non-nitrogenous  products  of  combustion.  The 
whole  molecule  is  labile,  and  there  is  probably  in  every 
living  molecule  a  nitrogenous,  as  well  as  a  non-nitrogenous, 


vi  PREFACE 

metabolism.  Nitrogen  absorbed  with  the  oxygen  is,  in 
part  at  least,  utilized  in  replacing  the  waste  in  this  element, 
and  the  carbon  brought  into  the  molecule  at  the  same 
time  is  in  part  detached  by  the  free  valences  in  the  carbo- 
hydrate groups,  and  used  to  repair  loss  in  this  part  of  the 
molecular  structure.  In  the  living  molecule  it  is  probable 
that  nitrogenous  metabolism  proceeds  much  more  slowly 
than  the  carbon  and  hydrogen  metabolism,  but  both  move 
rhythmically,  and  the  tempo  depends  upon  the  swing  of 
the  atomic  groups  that  constitute  the  molecule,  and  this 
rate  can  be  changed,  hastened  or  retarded,  by  alterations, 
either  physical  or  chemical,  in  the  medium  in  which  life 
resides.  When  the  molecule  is  in  active  life  its  food  is 
prepared  for  it  by  ferments,  and  it  is  quite  certain  that 
these  ferments  have  their  origin  in  the  nitrogenous  metab- 
olism of  the  living  molecule. 

The  keystone  or  archon  of  the  protein  molecule  is  our 
poison.  It  is  common  to  all  protein  molecules.  It  is  the 
primary  group.  One  protein  differs  from  another  in  the 
secondary  and  tertiary  groups.  Ordinary  proteins  are  not 
poisonous,  because  in  them  the  chemism  of  the  primary 
group  is  satisfied  by  combination  with  secondary  groups. 
Strip  off  the  secondary  groups  and  the  primary  becomes 
poisonous  on  account  of  the  avidity  writh  which  it  combines 
with  the  secondary  groups  of  other  molecules. 

The  specificity  of  proteins  resides  in  the  secondary  groups 
of  their  molecules,  and  all  specific  protein  reactions  are 
due  to  these  groups.  This  is  true  of  agglutination,  precipi- 
tin,  and  lytic  reactions.  Biological  relationship  between 
proteins  is  dependent  upon  the  chemical  structure  of  their 
molecules.  The  poisonous  part  of  a  protein  is  its  primary 
group;  the  sensitizing  part  is  found  among  the  secondary 


PREFACE  vi  i 

groups.  The  former  ^  physiologically  the  same  in  all 
proteins.  There  probably  are  chemical  differences  in  the 
primary  groups  of  varied  proteins,  and  it  is  possible  that 
fine  physiological  differences  may  be  detected  by  more 
careful  study,  but  the  primary  group  is  the  ring  about 
which  all  proteins  are  built,  or  at  least,  it  contains  this 
ring;  and  just  as  innumerable  compounds  may  be  built 
with  the  benzol  ring  as  a  nucleus,  so  all  proteins  are  con- 
structed about  a  common  centre.  The  secondary  groups 
are  not  identical  in  any  two  kinds  of  proteins.  There  may 
be  one  or  more  common  to  the  two,  but  in  some  respects 
there  are  differences. 

The  cell  is  not  the  unit  of  life;  life  is  molecular.  The  cell 
is  not  only  made  up  of  protein  molecules,  but  its  form  and 
function  are  determined  by  the  chemical  structure  of  its 
constituent  molecules.  The  lines  along  which  the  spore, 
seed,  or  ovum  develops  are  determined  by  the  chemical 
structure  of  its  ^proteins.  Growth  in  other  directions  is 
impossible,  and  this  accounts  for  stability  in  reproduction. 
However,  gradual  changes  in  the  chemical  structure  of 
living  proteins  occur,  and  in  these  lies  the  basis  of  organic 
evolution. 

The  basic  points  of  our  theory,  as  stated  above,  will  be 
in  evidence  throughout  this  volume.  The  experimental 
wyork  devoted  to  the  development  of  this  theory  could  not 
have  been  done  without  the  aid  of  able  assistants  who 
have  devoted  much  time  to  it,  and  all  without  adequate 
reward.  Besides  those  associated  with  me  in  the  prepa- 
ration of  this  volume,  special  mention  is  due  Drs.  Sybil 
May  Wheeler  and  Mary  Leach.  The  former  gave  eight 
years  and  the  latter  two  years  of  most  devoted  and  skilful 
service  to  the  elaboration  of  the  problems  discussed  here. 


vill  PREFACE 

I  regard  the  studies  recorded  in  this  volume  as  the  mere 
beginnings  of  work  which  should  be  developed.  I  dare  say 
that  our  record  contains  many  imperfections  and  possibly 
some  errors.  Future  studies  will  perfect  the  former  and  elimi- 
nate the  latter.  Attempts  to  solve  the  problems  stated  in 
this  volume  have  occupied  many  years  and  filled  them  with 
the  interest  and  pleasure  that  always  come  to  those  who 
seek  to  widen  the  fields  of  the  known. 

THE  SENIOR  AUTHOR. 

ANN  ARBOR,  1913. 


CONTENTS 


CHAPTER  I 

INTRODUCTION 

Bacteria  are  participate  proteins;  all  true  proteins  contain  a  poi- 
sonous group;  the  chemical  nucleus  contains  the  poison;  when  proteins 
are  disrupted  the  poisonous  group  may  be  set  free;  the  patho- 
genicity  of  a  bacterium  is  not  determined  by  its  capability  of  forming 
a  poison,  but  is  dsterminsd  by  its  ability  to  grow  and  multiply 
in  the  animal  body;  any  foreign  protein  which  can  grow  and  mul- 
tiply in  the  body  of  a  given  animal  is  pathogenic  to  that  animal; 
the  infectious  diseases  result  from  the  parenteral  digestion  of  proteins; 
natural  bacterial  immunity,  that  which  follows  an  infectious  disease, 
and  that  induced  by  vaccination,  result  from  inability  of  the  organ- 
ism to  grow  and  multiply  in  the  animal  body;  protein  sensitization 
and  bacterial  immunity,  apparently  antipodal,  are  in  reality  identical; 
protein  sensitization  consists  in  the  development  of  a  new  function 
in  certain  body  cells — that  of  elaborating  a  specific,  proteolytic 
ferment;  a  foreign  protein  introduced  into  the  blood  is  distributed 
through  the  tissues;  vaccines  are  protein  sensitizers;  toxin  and 
bacterial  immunities  are  different;  the  protein  poison  is  not  a  toxin; 
it  is  not  specific;  it  elaborates  no  antibody;  it  develops  a  specific* 
ferment;  different  proteins  tend  to  accumulate  in  predilection  places; 
the  symptoms  of  the  infectious  diseases  are  largely  determined  by  the  . 
organ  or  tissues  in  which  the  foreign  protein  accumulates;  the  poison 
elaborated  in  all  the  infectious  diseases  is  the  same;  when  a  cell  in 
the  animal  body  is  permeated  by  a  foreign  protein,  the  former  strives 
to  elaborate  a  ferment  by  which  the  latter  is  destroyed;  this  we 
believe  to  be  a  biological  law 18 

CHAPTER  II 

THE  GROWTH  OF  MASSIVE   CULTURES  OF  BACTERIA 
The  large  tanks  and  the  preparation  of  bacterial  cellular  substances       29 


x  CONTENTS 

CHAPTER  III 

PRELIMINARY  EXPERIMENT  OF  BACTERIAL  CELLULAR  SUBSTANCES 
They  consist  essentially  of  complex  proteins,  each  of  which  contains 
a  poisonous  group 37 

CHAPTER  IV 

THE  CHEMISTRY  OF  BACTERIAL  CELLULAR  SUBSTANCES 
Their  protein,  nuclein,  carbohydrate,  fatty  and  amino  constituents       52 

CHAPTER  V 

THE  CLEAVAGE  OF  PROTEINS  WITH  DILUTE  ALKALI  IN  SOLUTION 

IN  ABSOLUTE  ALCOHOL 

Bacterial,  vegetable,  and  animal  proteins  can  be  split  into  poison- 
ous and  non-poisonous  parts;  the  former  are  non-specific,  the  latter 
are  specific;  the  exact  nature  of  neither  of  these  portions  is  yet  known  95 

CHAPTER  VI 

ACTION  OF  ANIMALS 

The  action  of  the  living  bacillus,  of  the  dead  bacillus,  and  of  the 
poisonous  split  product 119 

CHAPTER  VII 

THE  PRODUCTION  OF  ACTIVE  IMMUNITY  WITH  THE  SPLIT 

PRODUCTS  OF  THE  COLON  BACILLUS 

The  establishment  of  a  certain  degree  of  tolerance  with  the  poi- 
sonous product;  this  is  non-specific;  the  production  of  a  mild  degree 
of  immunity  with  the  non-poisonous  product;  this  is  specific  .  .  137 

CHAPTER  VIII 

THE  SPLIT  PRODUCTS  OF  THE  TUBERCLE    BACILLUS  AND  THEIR 

EFFECTS  ON  ANIMALS 

The  cellular  substance ;  the  cell  poison ;  the  cell  residue ;  the  precipi- 
tate poison;  the  precipitate  residue;  the  final  filtrate;  the  action  of 
these  on  animals;  the  effects  of  the  tuberculo-poison;  sensitization 
with  tuberculo-protein;  the  relation  of  tuberculo-sensitization  to 
immunity 164 


CONTENTS  Xl 

CHAPTER  IX 

THE  ANTHRAX  PROTEIN 

Literature,  investigations;  the  anthrax  cellular  substance,  like 
other  proteins,  contains  a  poison;  sensitization  with  anthrax  protein  189 

CHAPTER  X 

THE  CELLULAR  SUBSTANCE  OF  THE  PNEUMOCOCCUS 
Difference  in  virulence  in  strains;  properties  and  effects  on  animals 
of  the  cellular  substance;  the  action  of  the  poisonous  portion;  auto- 
lysis  of  the  pneumococcus;  sensitization  with  pneumococcus  protein     205 

CHAPTER   XI 

PROTEIN  SENSITIZATION 

Introduction;  definition;  the  sensitizer;  all  true  proteins  sensitize; 
volatile  sensitizers;  the  sensitizing  group  in  the  protein  molecule; 
the  effects  on  different  animals;  period  of  incubation;  the  anaphyl- 
actic  state;  the  reinjection;  symptoms;  the  mechanism  of  anaphyl- 
axis;  passive  anaphylaxis;  anti-anaphylaxis;  the  Arthus  phenomenon; 
anaphylaxis  and  toxic  sera;  the  toxogens;  anaphylaxis  in  vitro;  the 
poison;  j8-iminazolylethylamin;  the  kyrins;  anaphylatoxin ;  physio- 
logical action  of  the  protein  poison;  general  physiological  action  of 
proteins;  sensitization  is  cellular;  theories;  theory  of  Friedberger; 
1  theory  of  Vaughan  and  Wheeler;  theory  of  Nolf 214 

CHAPTER  XII 

PARENTERAL  DIGESTION 

The  disposition  of  peptones;  the  fate  of  proteins  introduced 
directly  into  the  circulation;  the'  poisonous  action  of  proteins;  egg- 
white  injected  into  the  stomach  of  a  rabbit  may  be  in  part  absorbed 
unchanged;  egg-white  injected  into  the  rectum  of  a  rabbit  may  be, 
in  part  at  least,  absorbed  unchanged;  egg-white  injected  into  the 
peritoneal  cavity  of  a  rabbit  may  be  absorbed  unchanged;  egg- 
white  injected  intravenously  in  a  rabbit  quickly  disappears  from  the 
circulating  blood;  egg-white  injected  intravenously  in  a  rabbit  may 
be  detected  in  the  peritoneal  cavity,  in  the  bile,  and  in  certain 
organs  after  it  has  disappeared  from  the  circulating  blood;  the  'njec- 
tion  of  a  large  amount  of  egg-white  intravenously  in  a  rabbit  may 
prove  fatal;  the  blood  is  a  digestive  fluid;  proteolytic  digestion  in 
the  blood  is  regulated  by  the  accumulation  of  digestive  products  .  342 


xii  CONTENTS 

CHAPTER   XIII 

PROTEIN  FEVER 

The  production  of  acute,  intermittent,  remittent,  and  continued 
fevers  by  the  injection  of  foreign  proteins;  fever  results  from  the 
parenteral  digestion  of  proteins;  the  sources  of  fever  in  the  paren- 
teral  digestion  of  proteins;  fever  per  se  is  a  beneficent  process  .  .  373 

CHAPTER   XIV 

SPECIFIC  FERMENTS  OF  THE  CANCER  CELL 

Extra-  and  intracellular  ferments;  a  ferment  developed  in  animals 
by  injections  of  cancer  protein;  the  nature  and  action  of  this  ferment  416 

CHAPTER   XV 

THE  PHENOMENA  OF  INFECTION 

How  bacteria  grow;  how  bacteria  cause  disease;  the  phenomena 
of  the  period  of  incubation;  the  phenomena  of  the  active  period  of 
the  infectious  diseases;  the  germicidal  properties  of  the  blood;  the 
phenomena  of  tubercular  infection;  the  tuberculin  test;  vaccines  and 
sensitization;  sensitization  and  idiosyncrasies  to  food  and  medicine  .  436 


PROTEIN    POISONS 


CHAPTER   I 
INTRODUCTION 

MANY  years  ago  the  senior  contributor  to  this  volume 
began  a  research  on  the  chemistry  of  bacterial  cellular 
substance.  This  work  has  grown  and  the  progress  made 
has  been  reported  from  time  to  time  in  current  scientific 
and  medical  literature.  Able  assistants  have  rendered 
valuable  service,  and  as  the  research  has  developed  it  has 
been  correlated  with  that  done  along  similar  lines  in  other 
laboratories.  We  feel  that  the  time  has  come  when  the 
more  important  facts  ascertained  along  this  and  related 
lines,  by  all  investigators,  should  be  classified  and  proper 
deductions  drawn  from  them.  We  are  the  more  inclined 
to  do  this  because  we  believe  that  the  proper  interpretation 
of  the  results  obtained  opens  up  a  view  of  the  etiology  and 
development  of  both  immunity  and  disease,  which  has 
hitherto  not  been  appreciated.  We  have  thought  it  best 
to  state  briefly  in  this  introduction  some  of  the  most  impor- 
tant points  dwelt  upon  in  the  volume.  We  have  done  this 
somewhat  dogmatically,  hoping  that  they  will  impress  the 
reader  and  hold  his  attention  while  they  are  more  fully 
detailed  in  subsequent  chapters. 

1.  Bacteria  are  essentially  particulate,  specific  proteins. 
Bacteria  are  usually  classified  as  microscopic  plants,  but  we 
have  sought  diligently  for  the  presence  of  cellulose  in  their 
structure,  with  uniformly  negative  results.  We  have 
shown  that  some  bacteria,  at  least,  contain  two  carbohy- 
2 


18  PROTEIN  POISONS 

drates,  but  neither  of  these  gives  the  reactions  characteristic 
of  cellulose.  One  of  these  is  certainly  a  constituent  of  the 
nucleic  acid  group,  having  the  same  relation,  or  at  least  a 
similar  relation,  to  the  other  members  of  this  group  as 
exists  in  the  nucleins  and  nucleoproteins  found  in  the 
vegetable  and  animal  world.  Our  studies  together  with 
those  of  other  investigators  render  it  quite  certain  that 
bacterial  cellular  substance  yields  the  nuclein  bases  on 
hydrolysis.  The  position  of  the  second  carbohydrate  in 
the  molecular  structure  has  not  been  determined  with 
certainty.  It  is  thought  possible  that  it  is  simply  stored 
in  the  cell  as  a  reserve  food  supply;  but  that  this  is  not 
true  is  indicated  by  the  fact  that  it  cannot  be  removed  by 
simple  solvents,  and  that  its  separation  is  secured  only 
after  disruption  of  the  molecular  structure.  We  are  inclined 
to  the  opinion,  subject  to  change  as  the  result  of  more 
exact  knoAvledge,  that  the  second  carbohydrate  group  found 
in  at  least  some  bacteria  is  an  essential  constituent  of  the 
protein  structure.  Our  work  on  the  amino-acids,  both 
mono-  and  di-amino,  makes  it  certain  that  the  greater  part 
of  the  bacterial  cell  is  made  up  of  true  proteins.  We  have 
not  only  isolated  and  identified  many  of  the  amino-acids, 
but  we  have  shown  that  they  exist  in  widely  different 
proportions  in  different  species  of  bacteria,  just  as  they  do 
in  different  proteins  obtained  from  plants  and  animals. 
While  fats  and  waxes  are  found  in  relatively  large  amount 
in  certain  bacteria,  notably  in  bacillus  tuberculosis,  we 
see  no  reason  for  concluding  that  they  are  essential  con- 
stituents of  the  living  molecule.  That  they  are  specific 
products  of  the  life  activities  of  certain  bacteria  we  are 
convinced,  but  we  have  seen  no  reason  for  believing  that 
they  are  essential  constituents  of  the  bacterial  molecules. 
We  conclude  that,  chemically,  bacteria,  at  least  those 
with  which  we  have  worked,  are  nucleoproteins  or  glyco- 
nucleoproteins.  While  bacteria  are  morphologically  simple 
in  structure  and  without  differentiation  in  parts,  chemically 
they  are  complicated  in  structure,  quite  as  much  so  as 
many  of  the  tissues  of  the  higher  plants  and  animals.  The 


INTRODUCTION  19 

demonstration  that  bacteria  are  not  only  proteins,  but 
relatively  complex  proteins,  is  a  matter  of  marked  impor- 
tance. It  shows  that  in  many  of  their  life  processes  they 
must  bear  a  close  resemblance  to  the  cells  of  the  higher 
animals;  that  they  require  the  same  kind  of  food,  which 
they  select,  assimilate,  and  excrete  in  much  the  same  way; 
that  the  conditions  of  life  are  much  the  same;  what  is  favor- 
able to  one  bearing  a  like  relation  to  the  other,  and  what 
proves  injurious  to  one  having  a  like  effect  upon  the  other. 
2.  All  true  proteins  contain  a  poisonous  group.  At  first 
we  found  that  the  cellular  substance  of  certain  pathogenic 
bacteria  could  be  split  up  with  the  liberation  of  a  poisonous 
substance,  then  we  tested  non-pathogenic  bacteria,  then 
animal  and  vegetable  proteins,  and  all  with  the  same  result. 
Not  only  do  all  these  contain  a  poison,  but  so  far  as  its 
gross  effects  on  the  higher  animals  have  been  studied,  the 
same  poison.  We  have  held  that  when  we  know  more 
about  these  poisonous  oodies  obtained  from  all  proteins, 
it  will  be  found  that  chemically  they  are  not  identical,  but 
physiologically  they  are  so  closely  similar  that  up  to  the 
present  time  we  have  not  been  able  to  distinguish  one 
from  the  other  by  the  symptoms  induced.  The  poison 
obtained  from  the  typhoid  bacillus,  that  from  egg-white, 
and  that  from  edestin  of  hemp-seed  kill  animals  in  the  same 
doses,  with  the  same  symptoms  and  with  the  same  lesions. 
This  is  striking  evidence  of  the  similarity  in  the  structure 
of  the  protein  molecule,  whether  it  be  of  bacterial,  animal, 
or  vegetable  origin.  One  cannot  resist  the  temptation  to 
formulate  a  theory  to  fit  these  facts.  Indeed,  the  theory 
Unfolds  itself  and  may  be  briefly  expressed  as  follows:  All 
proteins  are  constructed  on  the  same  model  and  contain  a 
chemical  nucleus,  archon,  or  key-stone.  This  is  the  poison- 
ous group  and  is  practically  the  same  in  all  proteins.  One 
protein  differs  from  all  others  in  its  secondary  and  possibly 
its  tertiary  groups.  In  these  lies  the  specificity  of  proteins. 
Living  proteins  function  through  their  secondary  and 
tertiary  groups.  When  the  primary  group  is  detached 
from  its  own  subsidiary  and  specific  groups  it  manifests 


20  PROTEIN  POISONS 

its  poisonous  action  through  the  avidity  which  it  has  for 
the  secondary  groups  of  other  proteins.  These  are  thus 
detached  from  their  normal  positions  and  consequently 
the  living  protein  is  deprived  of  its  capability  of  functioning 
normally.  This  is  only  a  theory,  but  it  is  one  which  naturally 
suggests  itself. 

3.  The  chemical  nucleus  does  not  become  a  poison  until 
stripped  in  part  at  least  of  its  secondary  groups,  and  the 
intensity  of  its  poisonous  action  is  determined  by  the  thorough- 
ness with  which  the  secondary  groups   have  been  removed. 
The  protein  molecule  may  be  regarded  as  a  highly  complex 
neutral   salt,   made  up   of  many   basic  and   acid   groups. 
One  of  these  components,  it  may  be  either  a  basic  or  an 
acid  group  (or  it  may  have  within  itself  both  a  basic  and 
an  acid  group),  is  the  chemical  nucleus  of  the  molecule.    In 
its   natural   condition   its   chemism   is   satisfied   by   nicely 
adjusted  combination.    When  this  combination  is  disrupted, 
which  may  be  accomplished  either  by  chemical  agents  or 
by  enzymes,  the  chemical  nucleus  is  set  free,  more  or  less 
completely,  and  to  the  extent  that  it  is  released  from  com- 
bination, it  becomes,  in  the  presence  of  living  proteins,  a 
poison  because  it  disrupts  the  same.     We  have  shown  by 
direct  experiment  that  the  protein  poison  may  be  at  least 
partly  neutralized   by  being  kept  for  some  days   in   the 
presence  of  an  alkaline  carbonate  at  37°  C. 

4.  When  proteins  are  submitted  to  the  action  of  disrupting 
agents  there  is  the  possibility  of  the  chemical  nucleus  being 
set  free  more  or  less  completely,  and  to  the  extent  that  it  is 
detached  it  becomes  a  poison.     We  have  found  that  this 
occurs   when   proteins   are   carefully   disrupted   by   either 
dilute  acid  or  dilute  alkali.     So  far  as  our  work  has  gone 
the  best  agent  with  which  to  disrupt  the  protein  molecule 
and  obtain  the  largest  yield  of  poison  is  a  2  per  cent,  solu- 
tion of  caustic  soda  in  absolute  alcohol.     This  is  a  crude 
procedure  and   much   of  the  poison   is  destroyed   in   the 
process.     The  disruption  easily  extends  beyond  the  point 
where  the  poison  is  set  free  and  much  of  the  product  sought 
is  destroyed.     In  peptic  digestion  the  poison  becomes  active 


INTRODUCTION  21 

at  about  the  stage  of  the  formation  of  peptone,  and  it  has 
long  been  known  that  peptones  are  quite  highly  poisonous 
when  administered  parenterally.  This  also  is'  a  crude 
method  of  obtaining  the  poison,  and  with  all  the  work  that 
has  been  done  along  this  line,  we  do  not  know  whether  the 
peptone  is  itself  poisonous  or  whether  its  poisonous  action 
is  due  to  admixture  with  some  other  digestive  product. 
We  do  know  that  as  alimentary  digestion  proceeds  the 
protein  poison  itself  is  destroyed.  Indeed,  we  had  no 
conception  of  the  small  amount  of  protein  necessary  to 
furnish  a  lethal  dose  of  the  poison,  until  we  submitted 
proteins  to  the  blood  sera  and  organ  extracts  of  sensitized 
animals.  Then  we  found  that  1  mg.  of  protein  may  supply 
enough  poison  to  kill  a  guinea-pig  when  injected  intra- 
venously. But  to  produce  the  poison  in  this  way  necessi- 
tates a  delicate  adjustment  between  substrate  and  ferment 
which  is  imperfectly  understood,  and  consequently  inade- 
quately controlled,  and  we  can  know  that  we  have  produced 
the  poison  in  any  given  experiment  only  by  its  effect  on 
an  animal.  Thus  it  happens  that  after  years  of  study  we 
are  still  quite  ignorant  of  the  true  nature  and  chemical 
composition  and  structure  of  the  protein  poison. 

5.  The  pathogenicity  of  a  bacterium  is  not  determined  by 
its  capability  of  forming  a  poison.    Non-pathogenic  bacteria 
yield  just  as  much  of  the  protein  poison  as  the  most  highly 
pathogenic,  and  the  proteins  of  our  food  contain  the  same 
poison  that  is  found  in  pathogenic  bacteria. 

6.  The  pathogenicity  of  a  bacterium  is  dependent  upon  its 
ability  to  grow  and  multiply  in  the  animal  body.    Any  micro- 
organism which  can  grow  and  multiply  in  an  animal  body 
is  pathogenic  to  that  animal.     Growing  and  multiplying 
in  the  animal  body  means  that  the  invader  converts  the 
proteins  of  the  animal  into  its  own  proteins,  transforms 
native  into  foreign  proteins,  and  the  accumulation  of  foreign 
proteins  can  result  only  from  the  destruction  of  the  native. 
There  are  two  conditions  which  determine  whether  or  not  a 
foreign  protein  can  grow  and  multiply  in  the  animal  body: 
One  is  the  capability  of  the  invader  of  digesting  and  utilizing 


22  PROTEIN  POISONS 

the  proteins  of  the  body.  All  living  cells  grow  by  means 
of  their  own  digestive  ferments,  and  these  must  act  upon 
the  pabulum  within  their  reach.  If  the  ferment  of  the 
bacterial  cell  cannot  digest  and  prepare  food  for  the  bac- 
terium from  the  body  proteins,  then  the  invading  bacterial 
cell  dies.  The  second  factor  in  determining  whether  a 
given  bacterial  cell  will  grow  in  the  animal  body  is  the 
effect  of  the  ferments  of  the  body  cells  on  the  invader.  If 
these  are  rapidly  and  thoroughly  destructive  there  is  no 
bacterial  development,  and  the  organism  is  innocuous. 
The  prodigiosus  is  not  pathogenic,  but  the  cellular  sub- 
stance of  this  bacillus  obtained  by  growth  on  artificial  culture 
media  is  highly  poisonous  to  animals.  This  is  true,  with 
modification  as  to  degree,  of  the  cellular  substance  of  all 
non-pathogenic  bacteria.  It  is  not  the  lack  of  poison  in 
the  substance  when  placed  under  conditions  favorable  to 
its  growth,  but  it  is  its  inability  to  grow  under  unfavorable 
conditions.  The  smallpox  virus  is  pathogenic  to  the  unvac- 
cinated  but  non-pathogenic  to  the  vaccinated,  because  by 
vaccination  there  has  been  developed  in  the  body  a  ferment 
which  destroys  the  smallpox  virus  before  it  can  develop. 

7.  Any  foreign  protein  which  can  grow  and  multiply  in  the 
body  of  a  given  animal  may  prove  pathogenic  to  that  animal. 
Our  idea  of  the  development  of  an  infectious  disease  may 
be  stated  as  follows:  An  infective  agent  is  any  protein 
which  possesses  the  capability  of  growth  in  the  animal 
body.  What  these  essentials  are  have  been  stated  under  6. 
We  will  take  as  illustration  typhoid  fever.  The  infective 
agent  is  the  typhoid  bacillus,  a  specific,  particulate  protein. 
It  is  infective  because  by  means  of  its  digestive  ferment  it 
can  feed  upon  the  proteins  of  man's  body.  This  means 
that  it  can  convert  man's  proteins  into  typhoid  proteins 
and  thus  multiply  its  kind.  Moreover,  it  is  not,  imme- 
diately on  its  entrance  in  man's  body,  destroyed  by  the 
ferments  of  the  body  cells.  Having  found  admission  to  the 
body  it  proceeds  to  grow  and  multiply.  This  continues 
through  the  period  of  incubation,  which  in  this  disease  is 
somewhere  about  ten  days.  During  this  period  of  incuba- 


INTRODUCTION  23 

tion  there  is  no  effective  resistance  on  the  part  of  the  body 
cells  to  the  growth  and  multiplication  of  the  foreign  pro- 
tein. During  this  time  the  man  is  not  sick,  and  we  conclude 
therefore  that  it  is  not  the  growth  of  the  foreign  protein 
which  per  se  gives  rise  to  the  symptoms  of  typhoid  fever. 
However,  during  this  time  the  body  cells  are  being  pre- 
pared for  their  combat  with  the  foreign  protein.  This 
preparation  consists  in  the  development  in  certain  of  the 
body  cells  of  a  new  function,  that  of  elaborating  a  new  and 
specific  ferment  which  will  digest  and  destroy  the  foreign 
protein.  When  this  new  ferment  begins  its  action  the 
first  symptoms  of  the  disease  appear.  The  active  stage  of 
the  disease,  with  its  symptoms  and  the  lesions  induced, 
marks  the  period  over  which  the  parenteral  digestion  of 
the  foreign  protein  extends.  Death  may  come  from  the 
too  rapid  Jbreaking  up  of  the  foreign  protein  and  the  conse- 
quent liberation  of  a  fatal  dose  of  the  protein  poison,  which 
is  always  formed  on  the  disruption  of  the  protein  molecule, 
or  it  may  result  from  some  lesion  induced  by  the  products 
of  this  disruption,  such  as  perforation  and  hemorrhage,  or 
it  may  follow  from  chronic  intoxication  and  consequent 
exhaustion.  In  case  of  recovery  the  individual  is  for  a 
time  at  least  immune  to  the  typhoid  bacillus  because  his 
body  cells  are  now  able  to  elaborate  and  make  immediately 
effective  the  specific  ferment  which  destroys  the  typhoid 
protein. 

8.  The  infectious  diseases  result  from  parenteral  protein 
digestion.  Parenteral  digestion,  like  all  fermentative  pro- 
cesses, is  influenced  in  its  rate  of  progress  by  many  condi- 
tions, among  which  may  be  mentioned  the  relation  between 
amount  of  ferment  and  substrate,  the  physical  condition 
of  the  substrate,  and  the  presence  of  the  fermentation 
products.  These  influences  upon  parenteral  digestion  are 
not  easily  ascertained,  and  consequently  not  as  yet  measur- 
able or  controllable.  The  liberation  of  heat  as  measured  by 
body  temperature  has  recently  received  attention,  and  we 
can  say  in  a  general  way  that  fever  is  one  of  the  most  easily 
recognizable  effects  of  the  process.  While  natural  infec- 


24  PROTEIN  POISONS 

tion  is  due  to  living  proteins,  we  have  recently  learned 
that  experimental  fever  can  be  induced  by  repeated  injec- 
tions of  foreign  proteins  and  by  changes  in  size  of  dose  and 
in  intervals  between  doses,  fever  of  any  desired  type  can 
be  induced. 

9.  Natural  immunity  to  any  infection  is  due  to  inability 
of  the  infecting  agent  to  grow  in  the  animal  body.    This,  of 
course,  does  not  include  toxin  immunity,  which  is  due  to 
the  presence  in  the  body  of  an  antitoxin  or  of  something 
which  destroys  or  neutralizes  the  toxin.     The  inability  of 
an   infection   to   multiply   in   the   animal   body  has   been 
explained  under  6. 

10.  The  immunity  which  is  due  to  recovery  from  an  infec- 
tion is  the  result  of  the  development  in  the  body  during  the 
course  of  the  infection  of  a  specific  ferment  which   imme- 
diately destroys  the  infection  on  renewed  exposuret     As  has 
been  stated,  the   cells  of  the  body   acquire  the  function 
of  developing  the   specific   ferment,  and  this  function   is 
awakened  and  made  immediately  effective  on  subsequent 
exposure.    This  new  function   developed  in  the  body  cells 
by  disease  may  continue  throughout  life,  or  it  may  be  lost 
after  a  period  which  is  variable  in  different  diseases.    The 
immunity  induced  by  one  attack  of  yellow  fever  is  believed 
to   continue   through   life;   that   from   smallpox   generally 
holds  through  life,  and  but  few  have  typhoid  fever  more 
than  once.     Some  infectious  diseases,  such  as  pneumonia, 
are  apparently  not  followed  by  immunity.    In  most  instances 
it  seems  that  the  immunity  induced  by  one  attack  of  an 
infectious  disease  is  not  absolute,  but  only  relative,  and 
may  be  overcome  by  severe  or  prolonged  exposure  to  a 
virulent  form  of  the  infection. 

11.  Immunity    established    by    vaccination    is    similar    to 
that  induced  by  an  attack  of  the  disease.    The  vaccine  is  the 
same  protein  that  causes  the  disease.    It  must  be  so  modified 
that  it  will  not  induce  the  disease,  but  yet  so  little  altered 
that  it  will  stimulate  the  body  cells  to  form  a  specific  ferment 
which   will   promptly   and   quickly   destroy   the   infecting 
agent  on  exposure.     The  smallpox  virus  is  modified  by 


INTRODUCTION  25 

passage  through  the  cow.  The  anthrax  bacillus  is  con- 
verted into  a  vaccine  by  growth  at  high  temperature. 
The  typhoid  bacillus  is  killed  by  heat.  In  all  these  instances 
the  protein  is  so  little  changed  in  being  converted  from  an 
active,  infecting  agent  into  a  vaccine  that  it  still  sensitizes 
the  animal  to  itself  in  the  unmodified  state.  It  seems 
reasonable  to  conclude  that  the  protein  retains  its  capa- 
bility of  sensitizing  so  long  as  there  is  no  radical  alteration 
in  its  chemical  structure.  The  results  secured  by  vaccina- 
tion with  killed  typhoid  bacilli  prove  that  in  this  instance 
at  least  a  vaccine  is  not  necessarily  a  living  organism.  The 
possibility  of  obtaining  vaccines  from  the  split  products  of 
pathogenic  proteins  has  led  to  some  of  the  investigations 
detailed  in  this  volume,  and  while  we  do  not  claim  success 
in  this  particular,  we  think  that  continued  efforts  in  this 
direction  are  justified. 

12.  Protein  sensitization  and  bacterial  immunity,  appar- 
ently antipodal,  are  in  reality  identical.    This  statement  first 
made  by  us  in  1907  has  since  met  with  wide  acceptance, 
and  will  be  discussed  in  detail  in  later  chapters. 

13.  Protein    sensitization    consists    in    developing    in    the 
animal  body  a  specific  proteolytic  ferment  which  digests  the 
same  protein  on  reinjection.     Protein   sensitizers   may   be 
living  or  dead,  particulate  or  in  solution.    Soluble  proteins 
sensitize  more  readily  and  more  fully  than  those  not  in 
solution.      The    development    of    the    specific    proteolytic 
ferment  in  sensitization  is  due  to  the  action  of  the  foreign 
protein  upon  the  body  cells.    There  is  developed  in  certain 
body  cells  a  new  function,  that  of  elaborating  this  new 
ferment.    In  order  for  a  given  body  cell  to  be  thus  influenced 
by  a  foreign  protein,  the  latter  must  come  in  contact  with 
the   former.     Cell   permeation   by  the   foreign   protein   is 
probably  essential  to  the  perfect  elaboration  of  this  process. 

14.  When  a  foreign  protein  is  introduced  into  the  blood  of 
an  animal  it  soon  leaves  the  circulating  fluid  and  is  distributed 
throughout  the  tissues.    The  truth  of  this  has  been  demon- 
strated by  the  researches  of  independent  investigators,  and 
will  be  detailed  later.    This  is  true  of  both  particulate  and 


26  PROTEIN  POISONS 

soluble  proteins,  but  the  distribution  is  more  prompt  and 
effective  with  soluble  than  with  particulate  proteins.  This 
explains  why  the  former  are  more  efficient  sensitizers  than 
the  latter.  It  will  be  understood  that  a  protein  relatively 
insoluble  in  vitro  may  become  more  readily  soluble  in  vivo. 
15.  Vaccines  are  protein  sensitizers.  One  of  the  most 
important  problems  in  scientific  medicine  now  awaiting 
solution  is  that  of  the  preparation  and  employment  of 
vaccines.  The  term  vaccine — from  vacca,  a  cow — was 
first  used  when  Jenner  employed  the  infection  of  cowpox 
to  induce  immunity  to  smallpox.  Since  that  time  the  use 
of  the  word  "vaccine"  has  been  extended  to  include  every' 
form  of  preventive  inoculation.  Through  the  researches 
of  Wright  vaccine  therapy  has  been  developed,  and  now 
vaccines  are  employed  not  only  in  the  prevention  but  in 
the  treatment  of  disease.  Under  11  we  have  spoken  of 
the  employment  of  vaccines  in  inducing  immunity,  and 
now  we  wish  to  speak  briefly  of  vaccine  therapy.  As  we 
understand  it,  these  two  uses  of  vaccines  depend  upon  the 
same  principle.  The  action  of  the  vaccine  is  the  same  in 
both  instances.  The  protein  of  the  organism  responsible 
for  the  diseased  condition  and  that  of  the  vaccine  must  be 
identical  or  closely  related  bodies.  Both  must  be  protein 
sensitizers.  In  most,  if  not  in  all,  of  the  systemic  infectious 
diseases  the  infecting  organism  sensitizes  the  body  either 
throughout  or  over  large  areas.  It  seems  to  us  to  treat 
such  diseases  with  vaccines  is  irrational,  and  we  believe 
that  much  harm  has  been  done  by  such  attempts.  There 
are,  however,  local  infections  in  which  the  area  of  sensitiza- 
tion  is  limited  and  circumscribed.  Such  diseases  may  be 
treated  rationally  with  vaccines,  provided  such  agents 
can  be  obtained  in  such  forms  that  they  will  act  both 
effectively  and  harmlessly.  The  future  of  vaccine  therapy, 
in  our  opinion,  depends  upon  our  ability  to  secure  such 
vaccines.  That  we  have  not  yet  fully  established  our 
ability  to  obtain  vaccines  that  are  both  harmless  and 
effective  we  are  ready  to  admit.  This  does  not  mean, 
however,  that  all  efforts  to  accomplish  this  should  be  dis- 


INTRODUCTION  27 

continued.  In  our  own  attempts  in  this  direction  we  have 
met  with  enough  encouragement  to  lead  us  to  be  hopeful 
of  ultimate  success,  while  admitting  present  failure  In 
our  opinion,  it  is  not  only  unwise,  but  unjustifiable  to  treat 
advanced  cases  of  tuberculosis  with  tuberculin  or  other 
tuberculo-sensitizer.  So  long  as  the  disease  is  strictly  localized 
and  the  body  in  general  is  not  sensitized,  such  treatment 
can  find  reasonable  justification.  Our  theory  is  that  in 
strictly  localized  infections  the  proper  use  of  a  specific 
sensitizer  may  cause  the  more  general  and  abundant  forma- 
tion of  a  specific  proteolytic  ferment  which  may  aid  in  the 
destruction  of  the  infecting  organism.  In  our  opinion 
sensitization  consists  in  the  development  of  a  new  function 
in  the  body  cells.  In  strictly  local  infections  this  new 
function  has  been  developed  only  in  the  infected  area,  and 
to  establish  a  like  function  in  more  distant  cells  may  be 
beneficial.  The  present  tendency  on  the  part  of  the  pro- 
fession to  employ  all  kinds  of  bacterial  proteins  as  vaccines 
is,  in  our  opinion,  not  only  unscientific,  but  wholly  without 
justification.  It  should  be  clearly  understood  that  with 
every  protein  injected  into  the  body  a  most  potent  poison 
is  introduced,  and  caution  in  the  use  of  vaccines  is  not 
out  of  place. 

16.  Toxin  immunity  and  bacterial  immunity  are  radically 
different.     This  is  a  point  upon  which  we  shall  frequently 
touch,  and  we  hold  that  attempts  to  describe  one  of  these 
forms  of  immunity  in  terms  of  the  other  have  not  only 
been  unwarranted  by  the  facts,  but  have  led  to  unnecessary 
confusion. 

17.  The  protein  poison  is  not  a  toxin.     It  elaborates  no 
antibody,  and  while  its  repeated  use  in  non-fatal  doses  may 
establish  a  certain  tolerance,  it  gives  no  immunity  com- 
parable in  either  nature  or  degree  with  that  obtained  by 
like  employment  of  toxins. 

18.  The  protein  poison  is  not  specific. 

19.  The  tolerance  which  may  be  secured  by  the  protein 
poison  is  not  specific. 


28  PROTEIN  POISONS 

20.  The  sensitization  developed  by  a  protein  is  specific, 
but  is  not  due  to  the  poisonous  group  in  the  protein.    As  we 
have  stated,  the  specificity  of  a  protein  is  not  due  to  its 
poisonous  group,  which  is  much  the  same  in  all  proteins, 
but  to  its  secondary  groups,  for  it  is  in  these  that  one  protein 
differs  from  all  others. 

21.  Different  proteins  find  in  the  body  certain  predilection 
places  in  which  they  are  most  prone  to  accumulate.     The 
pneumococcus    accumulates    in    the    lungs,    the    smallpox 
virus  in  the  skin,  the  typhoid  bacillus  in  the  spleen,  and 
mesenteric   glands;    the    tubercle    bacillus   finds   its    most 
frequent  location  in  the  lungs,  but  it  has  been  a  parasite 
so  long  that  it  may  grow  on  any  human  tissue. 

22.  The  symptoms  of  a  given  disease  are  largely  determined 
by  the  location  of  the  foreign  protein.     The  most  skilful 
physician  may  not  be  able  to  tell  what  organism  is  respon- 
sible for  a  meningitis.     The  symptoms  of  acute  miliary 
tuberculosis  and  those  of  typhoid  fever  are  much  alike. 
It  is  the  location  of  the  infection  rather  than  the  exact 
nature  of  the  infecting  agent  which  gives  rise  to  the  more 
or  less  characteristic  symptoms  of  the  several  infectious 
diseases. 

23.  The  poison  elaborated  in  all  the  infectious  diseases  is 
the  same.    It  is  the  protein  poison,  and  it  is  physiologically 
the  same  whatever  its  source,  whether  it  comes  from  coccus, 
bacterium,  spirillum,  or  protozoan.     The  specificity  which 
characterizes  the  infectious  diseases  is  not  due  to  the  poison 
formed,  but  to  the  protein  cause  and  the  specific  ferment 
produced. 

24.  When  a  cell  in  the  animal  body  is  permeated  by  a 
foreign  protein,  the  former  strives  to  elaborate  a  ferment  by 
which  the  latter  is  destroyed.    We  believe  this  to  be  a  biological 
law,  and  we  think  that  it  lies  at  the  foundation  of  a  correct 
understanding  of  many  of  the  problems  of  immunity  and 
disease. 


CHAPTER  II 

THE  GROWTH  OF  MASSIVE  CULTURES  OF 
BACTERIA 

HAVING  decided  to  study  the  chemistry  of  the  bacterial 
cell,  it  soon  became  evident  that  we  must  devise  some  way 
of  obtaining  this  substance  in  large  quantity  and  fairly  free 
from  admixture  with  foreign  material.  Bacteria  had  been 
grown  only  in  test-tubes,  Petri  dishes,  and  Roux  flasks, 
and  none  of  these  methods  of  growth  gave  the  amount  of 
material  necessary  to  promise  any  satisfactory  investiga- 
tion. As  a  medium,  agar  suits  the  purpose  admirably, 
because  the  bacterial  growth  can  be  detached  and  washed 
from  the  surface  of  this  medium  quite  free  from  admixture, 
but  it  remained  to  devise  some  means  of  obtaining  a  large 
surface  so  protected  as  to  be  guarded  against  contamina- 
tion. At  first  we  tried  the  Roux  flasks,  and  by  inoculating 
one  hundred  of  them  with  the  colon  bacillus,  allowing  the 
cultures  to  grow  for  from  two  to  three  weeks  at  room  tem- 
perature, or  for  a  shorter  time  in  the  incubating  room,  and 
then  washing  off  the  growth  with  alcohol,  we  secured  a 
somewhat  bulky  and  promising  volume  of  bacterial,  cellular 
substance;  but  when  this  had  been  thoroughly  washed 
with  sterile  salt  solution,  extracted  with  alcohol  and  ether, 
dried  and  weighed,  we  found  the  total  yield,  under  the 
most  favorable  circumstances;  was  not  more  than  three 
grams.  This  enabled  us  to  make  some  preliminary  experi- 
ments and  to  demonstrate  that  the  dead  bacterial  cells, 
thus  prepared,  gave  all  the  general  color  reactions  for  pro- 
teins and  were  highly  poisonous  to  animals,  but  the  possi- 
bility of  making  any  satisfactory  chemical  study  was  not 
promising.  Moreover,  the  labor  and  care  necessary  to 


30 


PROTEIN  POISONS 
FIG.  1 


Tank  with  raised  lids. 
FIG.  2 


Tank  with  lids  lowered. 


GROWTH  OF  MASSIVE  CULTURES  OF  BACTERIA     31 

remove  with  anything  like  completeness  the  bacterial 
growth  from  one  hundred  Roux  flasks  were  greater  than 
we  could  afford  to  exert  more  than  a  few  times;  therefore, 
attempts  to  secure  the  desired  quantity  of  bacterial  cellular 
substance  by  this  method  were  abandoned.  We  then  tried 

FIG.  3 


The  incubating  room,  lids  lowered. 

growing  the  colon  bacillus  in  ordinary  moist  chambers,  used 
as  Petri  dishes,  but  with  the  greatest  care  many  of  these 
cultures  became  contaminated.  Finally,  we  devised  the 
large  copper  double  tanks  which  have  proved  wholly  satis- 
factory and  have  supplied  abundant  growths,  easily  obtain- 


32 


PROTEIN  POISONS 


able  and  free  from  contamination.  This  tank,  photographs 
of  which  are  here  given,  was  first  put  into  operation  in  1900, 
and  was  described  in  the  following  year1  (Figs.  1,  2,  3,  4). 
A  copper  tank  ten  feet  long,  two  feet  wide,  and  four 


FIG.  4 


The  incubating  room,  lids  raised. 

inches  deep,  with  a  trough  around  the  edge  one  inch  deep, 
has  a  cover  which,  when  lowered  into  place,  rests  in  the 
trough.  This  tank  is  supported  by  an  iron  frame  of  gas 
piping,  the  legs  of  which  rest  on  rollers,  so  that  the  whole 

]Trans.  Assoc.  Amer.  Phys.,  1901,  xvi,  217. 


GROWTH  OF  MASSIVE  CULTURES  OF  BACTERIA     33 

may  be  easily  moved  about  the  room.  An  inner  tank,  two 
inches  shorter  and  two  inches  narrower,  also  provided  with 
a  trough  that  runs  around  the  edge,  sits  in  the  large  one, 
and  is  supported  two  inches  from  the  bottom  of  the  larger 
one  by  iron  cross-bars.  The  bottom  of  the  outer  tank  and 
the  seal  trough  on  its  edges  are  filled  with  water.  The  seal 
trough  of  the  inner  tank  is  filled  with  glycerin.  Both  lids 
are  raised  and  lowered  by  wire  ropes  passed  through  pulleys 
fixed  in  the  ceiling.  The  iron  frame  supporting  the  tanks 
may  be  of  any  desired  height.  In  our  incubating  room  we 
have  a  nest  of  six  tanks,  three  of  which  are  on  frames  four 
feet  high  and  three  on  frames  two  feet  high.  This  economizes 
space,  as  the  lower  ones  can  be  rolled  under  the  higher  ones. 
Both  lids  are  supplied  with  vent  tubes  which  are  plugged 
with  cotton  in  sterilization.  Twenty  liters  of  3  per  cent, 
agar  is  placed  in  the  inner  tank;  both  lids  are  lowered  into 
their  respective  troughs,  and  with  large  gas  burners  at  full 
blast  underneath  the  apparatus  is  a  sterilizer.  After  three 
sterilizations  on  successive  days  the  medium  is  inoculated 
by  pouring  a  liquid  culture  through  the  vent  tubes  in  the 
lid  of  the  inner  tank.  Then  with  upper  lid  lowered  into 
the  water  trough  and  gentle  heat,  which  may  be  controlled 
by  a  thermoregulator,  it  becomes  an  incubator.  With  a 
number  of  tanks  in  a  small  room  it  is  better  to  heat  the  room 
to  the  desired  temperature,  thus  regulating  the  heat,  than 
it  is  to  heat  each  tank  separately. 

When  the  growth  has  reached  its  maximum,  the  time 
necessary  for  this  varying  with  the  organism  grown  and  the 
temperature  maintained,  both  lids  are  raised,  the  growth 
is  detached  from  the  subjacent  agar  with  sterilized  bent 
glass  rods,  sterile  salt  solution  added  if  necessary,  and  the 
bacterial  mass  is  drawn  by  means  of  a  water  pump  into  a 
sterilized  receiver. 

The  tanks  are  inoculated  from  special  glass  bulbs  in  which 
the  organism  has  been  grown  for  some  days.  With  the 
colon  bacillus  we  have  usually  employed  Uschinsky's 
solution,  or  some  modification  of  it,  in  the  inoculating  bulbs, 
in  order  that  there  may  be  no  trace  of  foreign  protein  in 
3 


34  PROTEIN  POISONS 

the  bacterial  growth.  While  it  was  highly  desirable  that 
this  should  be  done  at  least  once  in  order  to  demonstrate 
that  the  protein  reaction  given  by  the  bacterial  substance 
was  not  due  to  some  constituent  of  the  culture  medium, 
ordinarily  beef-tea  cultures  may  be  employed.  As  will  be 
seen  later,  we  did  grow  the  colon  bacillus  once  in  liquid 
Uschinsky  medium  for  the  purpose  of  fully  satisfying 
ourselves  that  the  protein  material  obtained  did  not  come 
from  the  culture  medium. 

After  removal  from  the  tanks  the  bacterial  cellular  sub- 
stance may  be  washed  with  various  fluids.  As  a  rule,  we 
have  washed  once  or  twice  with  sterile  salt  solution  by 
decantation  and  then  repeatedly  with  alcohol,  beginning 
with  50  per  cent,  and  increasing  to  95  per  cent.  The  sub- 
stance is  then  placed  in  large  soxhlets  and  extracted  first 
for  one  or  two  days  with  absolute  alcohol,  and  then  for 
three  or  four  days  with  ether.  These  extractions  with 
alcohol  and  ether  should  be  thorough  in  order  to  remove 
all  traces  of  fats  and  waxes. 

After  extraction,  the  cellular  substance  is  ground,  first 
in  porcelain,  then  in  agate  mortars,  and  passed  through 
the  finest  meshed  sieves.  If  there  be  bits  of  agar  in  the 
bacterial  cellular  substance,  which  is  seldom  the  case,  it  is 
separated  by  the  sieve  and  discarded.  The  one  who  grinds 
the  cellular  substance  should  wear  a  mask  in  order  to  pro- 
tect himself;  notwithstanding  this  precaution,  several 
workers  have  been  acutely  poisoned,  especially  with  the 
typhoid  bacillus.  Of  course,  there  is  no  danger  of  infection, 
as  the  material,  after  the  treatment  already  described, 
contains  no  living  bacilli.  The  finely  ground  cellular  sub- 
stance in  the  form  of  an  impalpable  powder  may  be  kept 
in  wide-mouthed  bottles  in  a  dark  place,  and  if  so  kept 
it  retains  its  toxicity  for  years,  but  when  long  exposed 
to  the  light,  even  if  kept  perfectly  dry,  it  becomes  less 
poisonous. 

The  yield  from  the  tanks  varies  with  the  organism,  but 
generally  amounts  to  from  60  to  80  grams  of  the  purified 
cellular  substance  for  each  tank,  and  with  six  tanks  in 


GROWTH  OF  MASSIVE  CULTURES  OF  BACTERIA     35 

operation,  and  with  a  crop  every  three  weeks,  one  may 
obtain  several  kilograms  within  a  few  months. 

Three  successive  crops  of  the  colon  bacillus  have  been 
grown  on  the  same  agar,  resterilizing  and  reinoculating 
after  each  harvest,  but  the  third  crop  is  not  abundant.  It 
has  been  found  to  be  well  to  follow  the  example  of  the 
scientific  agriculturalist  and  rotate  the  crops.  Colon 
grows  well  after  typhoid,  but  typhoid  does  not  grow  well 
after  colon.  Five  crops  have  been  obtained  in  the  following 
order:  (1)  pneumococcus,  (2)  typhoid,  and  (3)  three  suc- 
cessive colon  growths,  or  better,  non-pathogenic  bacteria 
following  one  colon  growth. 

Many  non-pathogenic  organisms,  the  colon,  typhoid, 
pneumococcus,  and  diphtheria  organisms,  have  been  grown 
on  the  tanks,  and  so  simple  is  this  method  of  obtaining 
bacterial  cellular  substance  in  large  amount  that  any  intel- 
ligent person,  after  some  experience,  may  repeatedly  go 
through  the  whole  manipulation,  producing  growth  after 
growth,  without  contamination.  In  the  laboratory  it  is 
best  to  have  one  man  make  a  specialty  of  producing  these 
growths. 

The  anthrax  and  tuberculosis  cellular  substances  with 
which  we  have  worked  have  not  been  produced  in  the 
tanks.  The  former  has  been  grown  in  Roux  flasks  and  the 
latter  in  glycerin  beef-tea  cultures.  After  these  growths 
have  been  obtained,  their  further  preparation  has  been  the 
same  as  that  already  outlined. 

Prepared,  as  described,  the  bacterial  cellular  substances 
form  fine,  white,  or  yellowish-white  powders.  This  is  true 
even  of  the  chromogenic  bacteria,  such  as  b.  violaceus  and 
b.  prodigiosus,  the  pigment  being  removed  from  the  cells 
by  its  solubility  in  alcohol.  One  of  our  former  students, 
Detweiler,1  studied  some  of  these  pigments,  but  these  do 
not  concern  us  at  present,  because  they  constitute  no  part 
of  the  cellular  protein.  The  same  is  true  of  the  other  bodies 
soluble  in  alcohol  and  ether.  The  extracts  made  with  these 

i  Trans.  Assoc.  Amer.  Phys.,  1902,  xvii,  246. 


36  PROTEIN  POISONS 

solvents  contain  fats,  waxes,  pigments,  and  possibly  other 
substances,  but  in  this  review  we  are  interested  solely  in 
the  cellular  proteins. 

Microscopic  examination  of  the  powdered  bacterial 
substances  show  the  bacilli,  mostly  intact,  though  many 
are  more  or  less  broken  by  the  attrition  to  which  they  have 
been  subjected.  The  cells  still  take  the  ordinary  stains, 
but  the  tubercle  bacillus  and  others  of  this  group  are  no 
longer  acid-fast,  showing  that  the  property  of  retaining 
the  stain  when  washed  with  mineral  acid  is  due  to  some 
constituent  removed  by  the  alcohol  and  ether. 


CHAPTER  III 


BACILLUS  coli  communis  was  selected  in  the  earlier 
experiments  for  the  following  reasons:  (1)  It  is  easily 
obtained  at  any  time  from  the  normal  feces  of  man.  (2)  It 
is  quite  stable  in  artificial  cultures,  varying  but  little,  if 
transplanted  from  day  to  day,  in  its  effects  upon  experi- 
mental animals.  (3)  It  elaborates  no  extracellular  poison,  at 
least  under  ordinary  conditions  and  in  beef-tea  cultures. 

Our  early  findings  were  reported  in  1901, *  and  these, 
confirmed  and  enlarged  by  subsequent  work,  will  be  briefly 
reported  as  follows: 

1.  The  poison  is  contained  within  the  bacterial  cell 
from  which  it  does  not,  at  least  under  ordinary  conditions, 
diffuse  into  the  culture  medium. 

This  was  demonstrated  by  the  following  experiment, 
which  was  repeatedly  made,  and  always  with  the  same  result : 
Beef-tea  cultures  of  the  colon  bacillus,  grown  for  three 
weeks  or  longer,  in  the  incubator,  were  filtered  through 
porcelain.  From  8  to  10  c.c.  of  the  clear,  sterile  filtrate 
was  injected  intra-abdominally  in  guinea-pigs.  The  animals 
thus  treated  were  restless  and  evidently  in  pain  for  some 
minutes  after  the  injection,  probably  due  to  the  volume  of 
the  fluid  and  its  slightly  irritating  character,  but  gave  no 
other  evidence  of  any  effect  of  the  injection. 

As  controls  to  the  above,  other  guinea-pigs  received 
intra-abdominally  0.25  c.c.  of  the  same  culture  unfiltered, 
and  all  died  within  twelve  hours.  It  will  be  understood 

1  Trans.  Assoc.  Amer.  Phys.,  xvi,  201. 


38  PROTEIN  POISONS 

that  these  animals  died  from  infection,  and  the  object  in 
inoculating  them  was  to  show  that  the  culture  contained 
living,  virulent  bacteria,  while  its  filtrate  was  without 
effect  on  animals.  However,  it  might  be  claimed  that  the 
poison,  although  in  solution  in  the  beef-tea,  will  not  pass 
through  porcelain.  This  suggestion  is  reasonable,  and 
calls  for  further  experimentation ;  consequently  an  un weighed 
portion  of  the  dead  cellular  substance  of  the  colon  bacillus 
was  suspended  in  water,  heated  in  the  autoclave  at  154° 
under  2  kilos  of  pressure,  and  filtered  through  porcelain. 
Four  cubic  centimeters  of  this  clear,  sterile  filtrate  injected 
intra-abdominally  into  a  guinea-pig  caused  death  within 
thirty-six  hours,  and  section  showed  the  same  lesions  that 
are  found  after  death  from  either  the  living  bacillus  or  the 
dead  cellular  substance.  This  demonstrates  that  when 
the  bacterial  cells  have  been  disrupted  by  superheated 
steam,  their  poisonous  constituent  becomes  to  some  extent 
soluble  in  water,  and  may  be  passed  through  porcelain. 
Furthermore,  experiment  showed  that  colon  cultures  when 
boiled  in  open  dishes  and  filtered  through  porcelain  supplied 
inert  filtrates.  This  indicates  that  the  disrupting  effect 
of  a  high  temperature  is  necessary  to  the  extraction  of  the 
poison  from  the  cell. 

Filtrates  from  living  cultures  of  the  diphtheria  bacillus 
contain  a  toxin  which  is  a  secretion  of  the  living  micro- 
organism. The  colon  bacillus  produces  no  such  active 
toxin.  Old,  dead  cultures  of  the  colon  or  typhoid  bacillus 
may  contain  soluble  poisons,  but  these  are  not  secretions 
of  the  living  cells.  They  come  from  the  autolysis  of  the 
dead  cells,  and,  as  we  shall  see  later,  they  are  not  properly 
toxins,  capable  of  producing  antibodies,  but  are  chemical 
poisons.  Moreover,  the  toxin  of  the  diphtheria  bacillus  is 
specific,  while  the  cellular  poison  is  not. 

2.  The  poison  is  not  extracted  from  the  bacterial  cell 
by  dilute  saline  solution,  alcohol,  or  ether,  either  at  ordi- 
nary temperature  or  at  the  boiling-point  of  these  fluids. 
Extracts  of  the  cellular  substance  of  the  colon  bacillus  with 
these  agents  were  repeatedly  made,  filtered,  evaporated 


BACTERIAL  CELLULAR  SUBSTANCE  39 

in  vacua,  taken  up  in  a  small  volume  of  water,  and  injected 
into  animals  without  effect.  That  the  extraction  of  the 
cellular  substance  of  the  colon  bacillus  with  alcohol  and 
ether  has  no  destructive  action  on  the  intracellular  poison 
is  shown  by  the  fact  that  this  material  after  extraction  with 
these  agents  does  not  lose  any  of  its  toxicity.  Furthermore, 
it  may  be  stated  that  prolonged  boiling  of  the  cellular  sub- 
stance of  the  colon  bacillus  with  alcohol  or  ether,  or  with 
the  two  successively,  neither  sets  free  nor  destroys  the 
intracellular  poison.  These  agents  dissolve  the  fats,  waxes, 
and  coloring  matter  from  bacterial  cells,  but  do  not  remove 
or  destroy  the  intracellular  poison.  As  will  be  seen  later, 
this  is  also  true  of  the  intracellular  poison  of  those  bacteria 
that  produce  a  soluble  toxin  in  their  cultures.  Moreover, 
it  will  later  appear  that  the  poisonous  group  not  only  in 
bacterial  but  in  vegetable  and  animal  proteins  is  soluble 
in  absolute  alcohol  after  it  has  been  well  detached  from 
the  other  groups  in  the  protein  molecule,  but  it  is  not 
removed  from  its  place  in  the  complex  molecule  by  either 
alcohol  or  ether.  Indeed,  the  poisonous  group  in  the  protein 
molecule  is  not  removed  from  its  attachment  to  other 
groups  by  any  purely  physical  solvents,  and  the  molecule 
must  be  disrupted  by  high  temperature,  chemical  agents, 
or  enzymes  before  its  poisonous  constituent  can  be  extracted 
by  physical  solvents. 

3.  The  cellular  substance  of  the  colon  bacillus  may  be 
heated  with  water  without  destruction  of  its  poisonous  group. 

Two  hundred  milligrams  of  the  cellular  substance  of  the 
colon  bacillus  was  suspended  in  10  c.c.  of  water  in  a  tube, 
which  was  then  sealed  and  heated  at  184°  for  thirty  minutes. 
On  opening  the  tube,  the  milky  content  was  found  on 
microscopic  examination  to  contain  granular  debris  with  a 
few  unbroken  cells.  Portions  of  this  heated  substance 
injected  into  guinea-pigs  caused  death,  and  autopsy  revealed 
the  same  lesions  that  are  seen  after  death  from  either  the 
living  bacillus  or  the  unbroken  cellular  substance. 

Another  portion  of  the  content  of  this  heated  tube  was 
placed  in  a  centrifuge  and  separated  into  a  deposit  and  a 


40  PROTEIN  POISONS 

supernatant  fluid,  the  latter  being  somewhat  opalescent. 
Guinea-pigs  were  treated  with  both  portions,  and  all  died. 
This  demonstrates  that  superheated  steam  disrupts  the 
bacterial  cells,  but  does  not  destroy  the  intracellular  poison. 
Heating  the  cellular  substance  of  the  colon  bacillus  in 
physiological  salt  solution  to  140°  in  the  autoclave  does 
not  destroy  or  even  weaken  the  poison.  The  cell  substance 
used  in  this  experiment  had  been  prepared  six  years  pre- 
viously. One  gram  was  thoroughly  mixed  with  100  c.c.  of 
salt  solution.  Two  cubic  centimeters  of  this  mixture,  con- 
taining 20  mg.  of  the  cell  substance,  killed  guinea-pigs  of 
300  grams'  weight  when  injected  intra-abdominally,  and 
0.5  c.c.  or  5  mg.  made  the  animals  very  sick.  The  emulsion 
was  then  heated  in  the  autoclave  to  140°  and  held  at  this 
temperature  for  ten  minutes.  One  cubic  centimeter  of 
this  heated  emulsion,  containing  10  mg.  of  the  cell  sub- 
stance, killed  guinea-pigs  of  300  grams'  weight,  and  0.1  c.c. 
or  1  mg.  made  the  animals  sick.  Animals  treated  with  the 
heated  emulsion  died  more  promptly  and  from  smaller 
doses  than  those  treated  with  the  unheated  preparation. 
Evidently  the  heating  prepared  the  cell  substance  so  that 
it  was  more  promptly  split  up  by  the  ferments  of  the  body. 
When  heated  to  this  temperature  .a  part  of  the  poison 
passes  into  solution.  This  was  shown  by  filtering  the 
heated  emulsion  through  hard  paper.  The  filtrate  was 
clear,  slightly  acid  to  litmus,  gave  the  biuret,  Millon,  and 
«-naphthol  tests,  and  killed  guinea-pigs  of  300  grams' 
weight.  The  animals  that  had  the  unheated  suspension 
showed  marked  peritoneal  inflammation,  with  bloody  exu- 
date.  In  those  having  the  larger  doses  (40  mg.  of  the  cell 
substance)  the  inflammatory  condition  extended  to  the 
muscular  walls  of  the  abdomen.  In  those  that  had  the 
heated  suspension  the  inflammatory  condition  was  much 
less  marked,  there  being  only  a  slight  serous  exudate,  less 
and  less  stained  with  blood  as  the  amount  of  the  cell  sub- 
stance injected  decreased.  In  those  killed  with  the  filtrate 
there  was  no  evidence  of  peritoneal  inflammation.  This 
furnishes  a  beautiful  illustration  of  the  nature  of  inflam- 


BACTERIAL  CELLULAR  SUBSTANCE       41 

mation  as  caused  by  bacterial  cells.  When  such  cells  are 
disrupted  wholly  by  the  body  cells,  and  in  a  restricted 
locality,  there  is  marked  destruction  of  the  local  body 
cells,  and  this  is  the  condition  which  we  designate  as  "  local 
inflammation."  On  the  other  hand,  when  the  bacterial 
cells  are  disrupted  and  the  cellular  poison  made  soluble 
before  being  introduced  into  the  body  of  the  animal,  there 
is  no  local  reaction,  or  no  special  local  reaction,  and  conse- 
quently no  recognizable  inflammatory  conditions. 

While  heating  suspensions  of  the  cellular  substance  in 
water  and  salt  solution  does  not  lessen  its  toxicity,  the 
poison  passes  into  solution  only  partially.  The  heated 
suspension  is  more  than  twice  as  poisonous  as  the  filtrate 
from  the  same.  Some  of  the  poison  actually  goes  into 
solution  and  the  filtrate  from  hardened  paper  may  be 
perfectly  clear,  provided  the  first  portion  be  returned  to 
the  filter,  but  the  greater  part  of  the  poison  is  removed  by 
filtration  through  paper.  It  is  worthy  of  note  that  heated 
suspensions  of  the  colon  cellular  substance,  filtered  or 
unfiltered,  easily  become  contaminated,  and  apparently 
furnish  acceptable  culture  media. 

4.  Dilute  (0.5  per  cent.)  solutions  of  the  caustic  alkalies 
disrupt  the  cellular  substance  of  the  colon  bacillus  slowly 
and  imperfectly. 

This  is  shown  by  the  following:  100  mg.  of  the  cellular 
substance  was  boiled  for  five  minutes  in  an  open  dish  with  a 
0.5  per  cent,  solution  of  sodium  hydroxide.  The  fluid 
was  centrifuged  and  the  deposit  found  to  be  still  poisonous, 
while  the  supernatant  fluid  was  without  effect.  However, 
stronger  solutions  (2  per  cent.)  of  alkali  completely  disrupt 
the  bacterial  cell  and  dissolve  the  poison  after  prolonged 
heating.  As  will  appear  later,  the  method  finally  selected 
for  splitting  off  the  poisonous  group  consists  in  heating 
the  cellular  substance  with  a  2  per  cent,  solution  of  sodium 
hydroxide  in  absolute  alcohol. 

5.  Boiling  with  a  0.2  per  cent,  dilution  of  hydrochloric 
acid  has  but  little   effect   upon  the   bacterial   cell  or  its 
contained  poison. 


42  PROTEIN  POISONS 

One  hundred  milligrams  of  the  cellular  substance  was 
boiled  in  an  open  test-tube  with  10  c.c.  of  a  0.2  per  cent, 
dilution  of  hydrochloric  acid.  This  induced  no  visible 
alteration  in  the  bacterial  cells,  as  seen  under  the  micro- 
scope, and  the  injection  of  this  material  into  animals  caused 
death  in  the  usual  time  and  with  the  usual  findings. 

6.  Heating   the   cellular   substance   for   an   hour   in   an 
open  dish  on  the  water-bath  (about  80°),  with  from  1  to  5 
per  cent,  solutions  of  hydrochloric  acid,  breaks  up  the  cells 
but  does  not  wholly  destroy  the  toxicity  of  the  cell  content; 
however,  prolonged  boiling  with   1   per  cent,  or  stronger 
dilutions  of  hydrochloric  acid  does  destroy  the  poison. 

Five  hundred  milligrams  of  the  colon  cellular  substance 
was  heated  on  the  water-bath  for  one  hour  with  500  c.c. 
of  a  5  per  cent,  solution  of  hydrochloric  acid  and  then 
decanted  through  a  hard  filter.  The  filtrate  was  clear 
and  colorless  and  gave  no  appreciable  precipitate  when 
dropped  into  absolute  alcohol,  but  that  the  acid  had  dis- 
solved some  part  of  the  cellular  substance  was  shown  by 
the  response  of  the  filtrate  to  the  biuret  test. 

The  undissolved  material  was  suspended  in  a  dilute 
solution  of  sodium  bicarbonate,  sufficient  to  neutralize  the 
acid,  and  injected  into  guinea-pigs  which  died  in  the  charac- 
teristic way,  and  showed  the  usual  lesions. 

7.  The  bacterial  cellular  proteins  are,   so  far  as  their 
toxicity  is  concerned,  quite  resistant  to  the  action  of  pepsin 
and  trypsin. 

A  given  sample  of  the  cellular  substance  of  the  colon 
bacillus  was  tested  upon  a  large  number  of  guinea-pigs,  in 
order  to  determine  the  minimum  lethal  dose,  which  was 
found  to  be  for  half-grown  animals  0.5  mg.  given  intra- 
abdominally,  and  1  mg.  subcutaneously.  This  material 
was  then  subjected  for  three  days  to  an  artificial  gastric 
juice,  the  efficiency  of  which  was  demonstrated  simul- 
taneously by  its  action  on  coagulated  egg-white.  The 
soluble  and  insoluble  parts  were  separated  and  their  toxicity 
tested.  One-half  milligram  of  the  undigested  part  given 
intra-abdominally  did  not  kill  but  1  mg.  did;  while  1  mg. 


BACTERIAL  CELLULAR  SUBSTANCE       43 

given  subcutaneously  no  longer  killed  but  2  mg.  did.  Of 
the  part  that  was  dissolved  in  the  acid  pepsin  solution, 
doses  up  to  100  mg.  had  no  effect.  A  like  result  was  obtained 
with  the  cellular  substance  of  the  typhoid  bacillus.  The 
amount  of  cellular  substance  left  undigested  after  three 
days'  exposure  to  the  acid-pepsin  was  about  10  per  cent,  of 
that  originally  taken.  The  conclusion  is  that  the  gastric 
juice  slowly  digests  the  bacterial  cellular  proteins,  and  in 
so  doing  destroys  the  poison. 

With  trypsin  the  effect  is  somewhat  different.  The 
cellular  protein  goes  into  solution  more  rapidly,  and  at 
least  a  part  of  the  poison  goes  into  solution  without  complete 
loss  of  its  properties.  The  parts  of  both  the  colon  and 
typhoid  cellular  substance  that  passed  into  solution  after 
three  exposures  to  trypsin  killed  in  doses  of  from  35  to  40 
mg.  given  intra-abdominally,  while  the  undigested  portion 
killed  in  doses  of  from  4  to  7.5  mg. 

One  gram  of  the  cellular  substance  given  to  a  rabbit 
through  a  stomach-tube  had  no  recognizable  effect  on  the 
animal. 

At  one  time  early  in  these  investigations  we  had  an  idea 
that  the  poison  in  the  colon  cell  resisted  peptic  digestion, 
and  we  therefore  quite  naturally  suspected  that  it  might 
be  a  nuclein.  This  belief  was  founded  upon  the  following: 
The  growth  on  fifty  Roux  flasks  was  removed,  extracted 
with  96  per  cent,  alcohol  so  long  as  the  alcohol  took  up 
coloring  matter,  then  dried,  placed  in  a  beaker,  and  stirred 
with  1  liter  of  0.2  per  cent,  hydrochloric  acid  dilution  in 
which  0.5  grams  of  active  pepsin  had  been  dissolved.  The 
beaker  with  content  was  kept  in  the  incubator  for  two 
days,  with  occasional  stirring.  The  undigested  portion 
on  microscopic  examination  was  found  to  be  amorphous, 
but  still  easily  stained  with  methylene  blue.  It  was  col- 
lected on  a  filter,  washed  thoroughly  with  96  per  cent, 
alcohol,  dried  at  100°,  and  pulverized.  One  hundred  milli- 
grams of  this  powder  was  shaken  with  50  c.c.  of  water, 
forming  an  acid,  colloidal  mixture.  On  adding  sodium 
bicarbonate  to  a  faintly  alkaline  reaction,  the  substance 


44  PROTEIN  POISONS 

dissolved  to  an  opalescent  fluid.  This  was  heated  in  order 
to  insure  sterilization,  and  injected  into  guinea-pigs,  which 
it  killed  within  from  six  to  twenty-four  hours.  One  milli- 
gram and  even  less  of  this  undigested  portion  killed  the 
animals  thus  treated,  but  subsequent  investigation  showed 
that  the  poisonous  portion,  when  fully  detached  from  the 
other  constituents  of  the  protein  molecule,  kills  in  a  few 
minutes,  and  we  concluded  that  the  undigested  part  con- 
sisted of  several  groups  still  attached,  or,  in  other  words,  of 
a  larger  and  more  complex  group  of  which  the  poison  is 
only  a  part.  The  action  of  the  proteolytic  enzymes  on 
bacterial  cells  deserves  a  more  thorough  study  than  we 
have  given  it. 

These  preliminary  studies  quite  convinced  us  so  long  ago 
as  1901  that  a  typical  colon  bacillus,  obtained  from  normal 
human  feces,  does  not  elaborate  in  its  cultures  a  soluble 
poison,  but  that  its  cells  do  contain  a  highly  active  body. 
Moreover,  these  studies  indicate  that  the  poison  of  the  colon 
bacillus  exists  in  the  essential  proteins  of  the  bacterial 
cell,  and  that  it  cannot  be  isolated  until  these  proteins  are 
broken  up  into  their  constituent  parts.  In  other  words, 
the  poison  consists  of  one  or  more  groups  in  the  protein 
molecule.  Since  the  colon  bacillus  may  grow  in  a  medium 
consisting  solely  of  inorganic  matter  and  a  small  amount 
of  some  organic  compound,  as  asparagin,  its  protein  must 
be  formed  synthetically,  and  its  poison,  as  a  constituent 
of  its  protein,  must  be  developed  in  the  same  way.  One 
is  forced  to  the  conclusion  that  the  poison  of  this  bacillus, 
at  least,  does  not  result  from  the  cleavage  action  of  the 
bacterial  cell  or  its  soluble  ferments  on  the  constituents 
of  the  medium  in  which  it  grows,  but  that  it  is  built  up 
synthetically,  and  is  set  free  only  when  the  cellular  protein 
is  disrupted.  In  other  words,  the  harmful  action  of  bacillus 
coli  communis  upon  animals  is  not  due  directly  to  the 
growth  and  multiplication  of  the  organism  in  the  animal 
body,  but  to  the  breaking  up  of  the  bacterial  protein  and 
the  consequent  liberation  of  its  poisonous  group. 

Subsequent  and  more  extended  research  has  shown  that 


BACTERIAL  CELLULAR  SUBSTANCE       45 

the  stability  of  the  protein  molecule  varies  within  wide 
limits,  and  that  the  facts  learned  in  the  study  of  the  special 
strain  of  the  colon  bacillus  do  not  hold  strictly  true  in 
every  particular  with  all  proteins.  This  was  expected,  and 
we  hoped  to  obtain  from  these  preliminary  studies  nothing 
more  than  certain  standards  by  which  our  findings  in  more 
extended  studies  might  be  measured.  With  some  strains 
of  the  colon,  and  in  many  more  of  the  typhoid  bacillus, 
we  have  obtained  evidence  of  the  presence  of  soluble  poisons 
in  old  cultures.  For  the  most  part  at  least  these  come  from 
autolysis  of  the  bacterial  cell.  This  is  a  subject  to  which 
we  shall  return  in  recording  the  development  of  these 
researches. 

The  findings  in  our  early  studies  quite  naturally  developed 
several  inquiries,  some  of  which  may  be  formulated  as 
follows:  If  the  protein  of  the  colon  bacillus  contains  a 
poisonous  group,  may  not  the  proteins  of  other  pathogenic 
bacteria  contain  similar  groups,  and  if  the  proteins  of  patho- 
genic bacteria  contain  poisonous  groups,  why  should  not  the 
proteins  of  non-pathogenic  bacteria  possess  like  constituents. 
If  bacterial  proteins  contain  poisonous  groups,  why  should 
not  other  proteins,  such  as  those  of  vegetable  and  animal 
origin,  contain  like  groups,  and  if  all  proteins  possess  in 
their  structure,  poisonous  bodies,  how  is  it  that  the  animal 
world,  including  man,  lives  so  largely  on  proteins  ?  Attempts 
to  solve  these  questions  have  taken  our  time  and  energy, 
and  given  us  much  pleasure.  The  results  of  these  labors 
constitute  the  principal  record  of  this  volume. 

Marshall  and  Gelston1  made  an  exhaustive  study  of  the 
toxicity  of  the  cellular  substance  of  the  colon  bacillus.  At 
first  they  employed  material  coarsely  ground  in  a  porcelain 
mortar.  This  was  suspended  in  water,  boiled  to  insure 
complete  sterilization,  and  then  injected  intra-abdominally 
in  guinea-pigs.  Up  to  1  part  of  poison  to  40,000  parts  of 
body  weight  all  animals  treated  in  this  way  died.  When 
the  proportion  was  reduced  to  1  to  50,000  and  less,  none  of 

1  Trans.  Assoc.  Amer.  Phys.,  1902,  xvii,  298. 


46  PROTEIN  POISONS 

the  animals  died.  The  same  powder  more  finely  ground  in 
an  agate  mortar  killed  15  out  of  16  animals  up  to  1  to 
75,000;  out  of  28  pigs  it  killed  9  at  1  to  100,000;  out  of  8 
it  killed  5  at  1  to  200,000;  out  of  34  it  killed  4  at  1  to 
2,000,000.  Heating  the  cellular  protein  for  fifteen  minutes 
under  2  kilos  of  pressure  at  134°  did  not  appreciably  lessen 
its  toxicity.  The  finely  ground  powdered  substance  killed 
rabbits  when  injected  intra-abdominally :  2  out  of  4  at  1 
to  75,000;  4  out  of  7  at  1  to  100,000;  4  out  of  12  at  1  to 
200,000;  and  2  out  of  18  at  1  to  2,000,000. 

One  gram  of  the  cellular  protein  was  incinerated  and  the 
whole  ash  injected  intraperitoneally  in  a  guinea-pig  without 
effect.  The  toxic  effect  when  subcutaneously  injected  is 
practically  the  same  as  when  employed  intraperitoneally, 
but  death  is  longer  delayed. 

The  intracellular  poison  of  the  diphtheria  bacillus  was 
studied  by  Gelston,1  using  different  strains,  among  which 
was  the  well-known  extracellular  toxin  producer  desig- 
nated as  Park  No.  8.  These  were  grown  in  Roux  flasks, 
as  it  was  found  that  this  organism  does  not  grow  well  in 
the  tanks,  supposedly  on  account  of  limited  aeration. 
The  growth  scraped  from  the  surface  of  the  agar  was  placed 
in  physiological  salt  solution  and  heated  for  two  and  one- 
half  hours  at  50°  to  secure  sterilization.  This  suspension 
was  then  poured  onto  hard  filters  and  placed  in  an  ice-box 
until  filtration  was  complete.  The  mass  on  the  filter  was 
washed  with  sterile  physiological  salt  solution,  dried  on 
porous  plates  over  sulphuric  acid  in  vacua,  and  reduced 
to  a  fine  powder  in  agate  mortars.  It  is  worthy  of  note 
that  the  physiological  salt  solution  contained  small  amounts 
of  the  extracellular  toxin,  and  that  still  larger  quantities 
could  be  obtained  by  macerating  the  agar,  on  which  the 
bacillus  had  grown,  with  salt  solution.  It  will  be  observed 
that  this  cellular  material  was  not  extracted  with  alcohol 
and  ether.  It  killed  guinea-pigs  when  injected  subcu- 
taneously or  intraperitoneally  up  to  1  to  33,000.  On  post- 

i  Trans.  Assoc.  Amer.  Phys.,  1902,  xvii,  308. 


BACTERIAL  CELLULAR  SUBSTANCE 


47 


mortem  examination  all  animals  dying  from  intraperitoneal 
injection  showed  marked  congestion  of  the  mesentery, 
omentum,  peritoneum,  and  adrenals;  also  numerous  ecchy- 
moses  and  hemorrhagic  effusions  in  the  serous  coat  of  the 
stomach,  intestine,  and  peritoneum.  The  spleen,  liver, 
and  kidneys  were  slightly  enlarged  and  dark.  In  the 
kidneys  a  dark  line  sharply  divided  the  cortex  from  the 
medulla.  The  heart  was  in  diastole.  When  mixtures  of 
diphtheria  antitoxin  and  suspensions  of  the  cell  substance 
in  salt  solution'  were  injected  into  guinea-pigs,  death  followed 
as  promptly  as  when  the  dead  germ  only  was  given.  One 
thousand  immunity  units  failed  to  protect  against  10 
minimum  fatal  doses  of  the  cell  substance  as  shown  by  the 
following : 

TABLE  I 


Amount 

Immunity 

No. 

Weight. 

injected. 

Antitoxin.1 

Antitoxin.2 

units. 

Result. 

1 

155 

47  .  1  mg. 

0 

.0005 

c.c. 

0 

.00031 

c.c. 

0 

.118 

+ 

2 

135 

40.9  mg. 

0 

.0050 

c.c. 

0 

,  00290 

c.c. 

1 

175 

+ 

3 

200 

60.6  mg. 

0 

.0242 

c.c. 

0 

.01940 

c.c. 

5 

,700 

+ 

4 

150 

46.0  mg. 

0 

.0500 

c.c. 

0 

.03000 

c.c. 

11 

750 

+ 

5 

135 

40.9  mg. 

0 

,2000 

c.c. 

0 

10900 

c.c. 

47 

000 

+ 

6 

175 

53.0  mg. 

0 

.5000 

c.c. 

0 

35000 

c.c. 

117 

500 

+ 

7 

120 

37.0  mg. 

1 

.0000 

c.c. 

0 

48000 

c.c. 

235 

000 

+ 

8 

120 

37.0  mg. 

2 

0000 

c.c. 

0 

96000 

c.c. 

470, 

000 

+ 

In  these  experiments  the  minimum  lethal  dose  of  the 
cellular  substance  was  1  to  33,000,  and  1  c.c.  of  antitoxin 
was  equivalent  to  235  immunity  units. 

With  another  sample  of  cell  substance  and  of  antitoxin, 
the  following  figures  were  obtained: 


No.  Weight. 

1  257 

2  280 

3  245 


Amount 
injected. 

55.39  mg. 
0.05  c.c. 

52.80  mg. 


TABLE  II 

Antitoxin. 
2.5000  c.c. 
0.0025  c.c. 
control 


Toxin         Anti-     Immunity 
cor.      toxin,  cor.  units.  Result. 

2.5680       1000       + 
0.056       0.0028  1        — 


In  this  series  No.  2  received  10  minimum  lethal  doses  of 
a  filtered  bouillon  culture  of  the  same  bacillus  from  which 


1  Antitoxin  for  250  grams  body  weight. 

2  Antitoxin  for  actual  body  weight. 


48        •  PROTEIN  POISONS 

the  cellular  substance  had  been  obtained,  and  1  immunity 
unit  of  antitoxin.  These  experiments  show  that  while 
diphtheria  antitoxin  protects  against  the  extracellular 
toxin  it  fails  to  protect  against  the  intracellular  poison. 

Suspensions  of  the  diphtheria  cellular  substance  when 
exposed  to  a  temperature  of  50°  or  higher,  for  fifteen  minutes 
or  longer,  gradually  decrease  in  toxicity,  but  this  is  not 
wholly  lost  after  exposure  to  122°  in  the  autoclave  for 
thirty  minutes.  The  minimum  lethal  dose  of  our  prepara- 
tion having  been  found  to  be  1  to  33,000,  it  "was  heated  to 
122°  for  20  minutes,  and  then  proved  to  be  1  to  6400.  It 
is  worthy  of  note  that  the  several  strains  of  the  diphtheria 
bacillus  employed  yielded  cellular  substances  that  varied 
widely  in  toxicity.  As  a  rule,  one  that  supplied  a  potent 
extracellular  poison  yielded  relatively  an  indifferent  intra- 
cellular poison.  Possibly  this  is  due  to  the  greater  lability 
of  the  molecular  structure  of  the  former,  which  leads  to 
the  partial  breaking  down  of  the  protein  molecule  in  the 
heating  resorted  to  in  order  to  sterilize  the  growth. 

The  cellular  substance  of  the  anthrax  bacillus  was  pre- 
pared and  studied  by  J.  Walter  Vaughan.1  This  kills 
guinea-pigs  in  only  relatively  large  doses,  and  this  fact 
indicates  that  the  intensity  of  the  infectious  properties 
of  a  microorganism  is  not,  always  at  least,  measured  by 
the  potency  of  its  intracellular  poison.  The  bacillus  pro- 
digiosus  is  non-pathogenic  to  the  higher  animals,  not  from 
its  inability  to  furnish  a  poison,  but  because  it  cannot 
grow  and  multiply  in  the  animal  body;  while,  on  the  other 
hand,  the  anthrax  bacillus  is  highly  infectious  to  some  of 
the  higher  animals,  not  from  the  intensity  of  the  poison 
which  it  elaborates,  but  rather  from  the  fact  that  in  these 
animals  this  bacillus  finds  conditions  favorable  to  its  growth 
and  multiplication. 

Detweiler2  prepared  and  demonstrated  the  poisonous 
action  of  the  cellular  substances  of  b.  prodigiosus,  b. 
violaceus,  sarcina  aurantiaca,  and  s.  lutea. 

1  Trans.  Assoc.  Amer.  Phys.,  1902,  xvii,  313. 

2  Ibid.,  257. 


BACTERIAL  CELLULAR  SUBSTANCE       49 

The  relative  toxicity  of  the  finely  powdered  cellular  sub- 
stances of  certain  bacteria  in  proportion  to  body  weight  of 
animal,  is  shown  in  the  following  figures: 

Bacillus  anthracis 1  to     1,700 

Sarcina  lutea 1  to    2,050 

Micrococcus  pneumonias 1  to  10,000 

Sarcina  aurantiaca 1  to  25,500 

Bacillus  violaceus 1  to  26,500 

Bacillus  diphtherias 1  to  33,000 

Bacillus  typhosus 1  to  40.000 

Bacillus  pyocyaneus 1  to  50,000 

Bacillus  coli 1  to  75,000 

Bacillus  prodigiosus 1  to  90,000 

Numerous  and  varied  attempts  to  immunize  animals 
with  the  bacterial  substances  were  made.  Guinea-pigs, 
rabbits,  and  goats  were  used  in  these  experiments.  The 
following  quotation  is  taken  from  the  report  Marshall  and 
Gelston1  made  in  1902: 

"1.  Although  guinea-pigs  and  rabbits  acquire  immunity 
to  the  germ  substance  but  slowly,  yet  if  sufficient  time  be 
given,  and  the  animal  be  allowed  to  recover  before  a  second 
dose  is  administered,  a  fair  degree  of  immunity  may  be 
obtained  after  many  months  of  treatment.  Unless  the 
animal  be  allowed  to  recover  completely  before  a  second 
injection  is  given,  death  generally  results.  In  these  ex- 
periments for  the  purpose  of  inducing  immunity,  we 
employed  the  finely  divided  germ  substance  which  had 
been  used  in  determining  the  toxicity.  Our  previous 
experiments  had  shown  that  one  part  of  the  finely  divided 
powder  to  75,000  parts  of  body  weight  in  a  guinea-pig 
was  surely  fatal.  Guinea-pig  No.  93  received  in  the  fifteenth 
injection  a  dose  of  28  mg.  (1  to  17,321),  and  recovered  from 
the  same.  Guinea-pig  No.  109  received  on  the  ninth  injec- 
tion 25.5  mg.  (1  to  22,745),  and  recovered.  Rabbit  No.  17 
received  on  the  fifteenth  injection  372.3  mg.  (1  to  4780), 
and  recovered.  Rabbit  No.  29  received  on  the  thirteenth 
injection  293.3  mg.  (1  to  6409),  and  recovered. 

1  Loc.  cit. 


50  PROTEIN  POISONS 

"2.  Study  of  the  blood  serum  from  animals  partially 
immunized  to  the  germ  substance  led  to  the  following 
results : 

"  (a)  The  blood  serum  of  rabbit  No.  2,  which  had  received 
on  the  twelfth  injection  562.3  mg.  (1  to  3503),  obtained 
twenty  days  after  the  last  injection,  had  no  bacteriolytic 
action  on  the  organisms  contained  in  a  beef-tea  culture 
of  the  colon  bacillus  of  the  same  strain  as  that  which  had 
been  used  in  immunizing  the  animals. 

"  (6)  Tubes  of  immune  serum  and  others  of  normal 
serum  were  inoculated  with  virulent  cultures  of  the  colon 
bacillus,  and  placed  in  the  incubator  at  37.5°.  Both  tubes 
showed  good  and  equal  growth  within  seven  hours. 

"  (c)  Tubes  of  immune  and  normal  serum  were  inoculated 
writh  virulent  cultures  of  colon  and  typhoid  bacillus,  and, 
from  these  gelatin  plates  were  made  at  the  end  of  one 
minute,  thirty  minutes,  one  hour,  and  forty-eight  hours. 
The  number  of  colonies  on  the  colon  plates  did  not  diminish 
while  those  on  the  typhoid  plates  did  with  both  sera,  thus 
showing  that  the  immune  serum  had  no  specific  action. 

"  (d)  Immune  serum,  when  mixed  with  filtered  cultures 
of  the  colon  bacillus,  gives  a  slight  precipitate,  which, 
however,  is  also  given  by  normal  serum. 

"  (e)  The  immune  serum  gives  very  positive  agglutinating 
reactions  with  suspensions  of  the  dead  colon  germ  used  in 
immunizing  the  animals.  Suspensions  of  from  1  to  50  mg. 
of  the  germ  substance  in  1  c.c.  of  physiological  salt  solution 
are  completely  precipitated  in  from  three  to  five  minutes 
on  the  addition  of  an  equal  quantity  of  immune  serum. 
Control  tubes  to  which  normal  serum  had  been  added  gave 
negative  results,  inasmuch  as  complete  subsidence  did  not 
occur  in  these  writhin  fourteen  days. 

"  (/)  The  immune  serum  exerts,  when  mixed  with  suspen- 
sions of  the  germ  substance  and  injected  into  the  abdominal 
cavity  of  rabbits,  a  slight  protective  action,  as  is  shown  by 
the  following: 

Rabbit  No.  1,  850  grams,  had  17.0  mg.  (1  to  50,000),  2  c.c.,  I.  S.,  — . 
Rabbit  No.  2,  700  grams,  had  14.0  mg.  (1  to  50,000),  2  c.c.,  N.  S.,  +. 
Rabbit  No.  3,  840  grams,  had  16.8  gm.  (1  to  50,000),  2  c.c.,  beef-tea,  +. 


BACTERIAL  CELLULAR  SUBSTANCE       51 

"  (g)  The  immune  serum  of  the  rabbit  exerts  no  protec- 
tive action  for  the  guinea-pig  against  the  germ  substance. 

"3.  The  immunity  obtained  by  treating  animals  with 
the  germ  substance  is  apparently  of  short  duration,  and 
if  the  interval  between  the  administration  of  doses  be 
prolonged,  death  is  likely  to  follow  even  when  there  is  no 
increase  in  the  dose.  It  will  thus  be  seen  that  the  attempt 
to  immunize  animals  to  the  germ  substance  of  the  colon 
bacillus  is  beset  with  difficulties.  If  the  intervals  be  too 
short,  and  the  animal  has  not  fully  recovered,  death  is 
likely  to  result,  and  the  same  will  also  probably  happen 
when  the  interval  is  unduly  long.  Even  after  a  marked 
degree  of  immunity  has  been  obtained,  this  is  apparently 
lost  within  a  few  weeks,  and  a  repetition  of  a  dose  of  the 
same  size  causes  death.  The  following  table,  showing  next 
to  the  last,  and  the  last  injections  given  to  certain  animals, 
will  illustrate  our  meaning: 

January  19,  1902,  rabbit  No.  17,  1780  grams,  had  372.3  mg.  (1  to  4780), 
and  recovered. 

March  29,  1902,  rabbit  No.  17,  1880  grams,  had  273.5  mg.  (1  to  5048),  . 
and  died  March  31. 

January  18,  1902,  rabbit  No.  29,  2000  grams,  had  293.3  mg.  (1  to  6819), 
and  recovered. 

March  29,  1902,  rabbit  No.  29,  1880  grams,  had  293.3  mg.  (1  to  6409), 
and  died  March  31. 

January  6,  1902,  pig  No.  93,  485  grams,  had  28  mg.  (1  to  17,329),  and 
recovered. 

March  29,  1902,  pig  No.  93,  645  grams,  had  28  mg.  (1  to  19,464),  and 
died  March  31. " 

Later  these  experiments  were  repeated  and  extended  to 
goats  by  V.  C.  Vaughan,  Jr.,  and  Gumming,  with  prac- 
tically the  same  results.  We  concluded  that  the  capability  of 
the  animal  to  bear  increased  doses  of  the  cellular  substance 
was  not  sufficiently  marked  to  be  designated  by  the  term 
immunity,  and  it  was  decided  to  recognize  it  as  increased 
tolerance.  This  opinion  we  still  hold.1 

1  It  will  be  seen  from  this  work  as  here  recorded  that  we  met  with  the 
phenomena  of  protein  sensitization  or  so-called  anaphylaxis  as  early  as 
1902  in  our  studies  on  the  bacterial  cellular  substances,  but  that  we  failed 
to  follow  it  up  and  indeed  did  not  attach  much  importance  to  it. 


CHAPTER   IV 

CHEMICAL  STUDIES  OF  BACTERIAL  CELLULAR 
SUBSTANCE 

Proteins. — Xencki  and  Schaffer1  obtained  the  cellular 
substance  from  a  mixed  culture  of  putrefying  bacteria, 
dried  it  to  a  constant  weight,  first  on  the  water-bath  and 
then  at  110°,  pulverized,  and  extracted  with  alcohol  and 
ether.  The  residue  thus  obtained  was  extracted  on  the 
water-bath  with  0.5  per  cent,  potassium  hydroxide.  From 
the  alkaline  extract  a  protein,  designated  as  mykroprotein, 
was  precipitated  by  neutralization  and  saturation  with 
sodium  chloride.  Mykroprotein  when  freshly  precipitated 
was  found  to  consist  of  amorphous  flakes  soluble  in  water, 
but  losing  in  solubility  when  dried  at  110°.  It  contains 
52.32  per  cent,  of  C,  7.55  per  cent,  of  H,  14.75  per  cent, 
of  N,  and  neither  sulphur  nor  phosphorus.  In  aqueous 
solution  it  gives  an  acid  reaction,  and  is  not  precipitated 
by  alcohol,  but  is  precipitated  by  picric  acid  and  other 
alkaloidal  reagents.  It  gives  the  biuret  and  Millon  reac- 
tions, but  not  the  xanthoproteic.  On  being  fused  with 
potash  it  furnishes  ammonia,  amylamin,  phenol,  valerianic 
acid,  leucin,  and  traces  of  indol  and  skatol.2  Later,  Xencki3 
attempted  to  prepare  mykroprotein  from  anthrax.  He 
obtained  from  anthrax  spores  a  substance  which  he  desig- 
nated as  anthrax  protein,  closely  related  to  plant  casein 
and  animal  mucin,  soluble  in  alkalies,  but  insoluble  in 
water,  acetic,  and  dilute  mineral  acids.  Like  mykroprotein, 
it  contains  no  sulphur.  Dyrmont4  made  an  analysis  of 

1  Jour.  f.  prakt,  Chem.,  1879,  xx,  443. 

2  Ibid.,  1881,  xxiii,  302.  3  Berichte,  1884,  xxvii,  2605. 
4  Arch.  f.  exper.  Path.  u.  Pharm.,  1886,  xxi,  309. 


BACTERIAL  CELLULAR  SUBSTANCE       53 

anthrax  protein,  with  the  following  results:  C,  52.1  per 
cent.;  H,  6.82  per  cent.;  N,  16.2  per  cent.,  and  a  trace  of 
ash.  We  now  know  that  both  mykroprotein  and  anthrax 
protein  are  cleavage  products  obtained  from  bacterial 
cellular  substances  through  the  action  of  alkali. 

Brieger1  found  in  Friedlander's  pneumococcus  a  protein, 
partially  soluble  in  water  and  precipitated  on  boiling, 
containing  less  nitrogen  than  is  found  in  mykroprotein. 
Lewith,2  reporting  work  done  by  Hellmich  on  a  hay  bacillus 
grown  on  synthetic  medium,  stated  that  a  globulin  was 
extracted  from  the  cellular  substance  by  neutral  salts  at 
ordinary  temperature.  This  probably  was  in  fact  no  part 
of  the  cellular  protein.  Subsequent  extraction  with  dilute 
alkali  gave  a  protein  body  described  as  an  albuminate  and 
said  to  resemble  casein. 

Buchner3  demonstrated  that  certain  bacterial  cells  con- 
tain pyogenetic  bodies.  These  are  extracted  from  the 
cellular  substances  tjiith  dilute  alkalies  from  which  they 
are  precipitated  with  dilute  acids.  The  amount  of  protein 
obtained  in  this  way  varied  greatly  with  the  species  of 
bacteria.  Bacillus  pyocyaneus  gave  the  most  abundant 
yield,  supplying  19.3  per  cent,  of  the  dried  cells.  By 
heating  on  the  sand-bath  under  a  reflux  condenser,  or 
in  an  autoclave  at  120°,  filtering  through  sand,  and  precipi- 
tating with  absolute  alcohol,  he  obtained  a  more  soluble 
protein. 

Brieger  and  Frankel,4  Proskauer  and  Wassermann,5  and 
Dzierzgowski  and  Rekowski6  prepared  so-called  toxal- 
bumins  from  diphtheria  cultures,  and  made  ultimate 
analyses  of  the  same,  but,  as  we  now  know,  these  were 
mixtures  and  gave  us  no  information  concerning  the  com- 
position of  the  bacterial  cell. 

1  Zeitsch.  f.  physiol.  Chem.,  1885,  ix,  1. 

2  Arch.  f.  exper.  Path.  u.  Pharm.,  1890,  xxvi,  341. 

3Berl.  klin.  Woch.,  1890,  xxvii,  673;  ibid,  1084;  Munch,  med.  Woch., 
1891,  xxxviii,  841. 

<  Berl.  klin.  Woch.,  1890,  xxvii,  241,  268. 
6  Deutsch.  med.  Woch.,  1891,  xvii,  585. 
6  Arch.  d.  Sci.  biol.,  1892,  i,  167. 


54  PROTEIN  POISONS 

Hammerschlag1  extracted  tubercle  bacilli  with  dilute 
alkali  and  precipitated  a  protein  with  ammonium  sulphate. 
Buchner2  extracted  tubercle  bacilli  with  from  40  to  50  per 
cent,  glycerin,  and  obtained  what  he  believed  to  be  the 
active  principle  of  tuberculin,  but  which  in  reality  consisted 
of  a  mixture  of  the  autolytic  products  of  this  bacillus. 
Hoffman3  reported  the  isolation  of  six  proteins  from  the 
tubercle  bacillus,  but  these  were  mixtures.  Weyl4  believed 
that  he  had  succeeded  in  separating  the  membrane  from 
the  protoplasmic  content  of  the  bacterial  cell,  and  from  the 
latter  he  obtained  a  body  which  he  designated  as  a  toxo- 
mucin,  but  there  is  no  proof  that  the  bacterial  cell  has 
any  such  structure  as  he  supposed.  Ruppel5  prepared 
from  the  tubercle  bacillus  a  body  that  he  named  tuber- 
culosamin. 

Vandervelde6  reported  the  presence  of  nuclein  in  bacillus 
subtillis,  and  Dreyfuss,7  basing  his  opinion  on  the  behavior 
of  bacteria  toward  the  basic  aniline  dyes,  concluded  that 
nuclein  is  a  constituent  of  all  bacteria.  Gottstein,8  finding 
that  various  bacteria  decompose  hydrogen  peroxide,  both 
during  life  and  after  death,  concludes  from  this,  from  the 
presence  of  phosphorus,  and  from  the  affinity  of  bacteria 
for  basic  aniline  dyes,  that  they  contain  nuclein.  Nishi- 
mura9  reported  the  finding  of  nuclein  in  a  water  bacillus 
grown  on  potato.  The  bacterial  cells  were  removed  from 
the  potatoes,  extracted  with  alcohol  and  ether,  heated  under 
a  reflux  condenser  with  0.15  per  cent,  sulphuric  acid,  and 
then  heated  in  an  autoclave  at  105°.  From  this  acid  extract 
he  obtained  0.17  per  cent,  xanthin,  0.08  per  cent,  adenin, 
and  0.14  per  cent,  guanin.  Lustig  and  Galeotti10  prepared 

1  Monats.  f.  Chem.,  1899,  x,  9;  Centralbl.  f.  klin.  Med.,  1891,  xii,  9. 

2  Munch,  med.  Woch.,  1891,  xxxviii,  45. 

3  Wien.  klin.  Woch.,  1894,  712. 

4  Deutsch.  med.  Woch.,  1891,  xvii,  256. 

s  Zeitsch.  f.  Physiol.  Chem.,  1898,  xxvi,  218. 

6  Ibid.,  1884,  viii,  367. 

7  Ibid.,  1893,  xviii,  358. 

8  Virchow's  Archiv,  1893,  cxxxiii,  302. 

9  Arch.  f.  Hygiene,  1893,  xviii,  318. 

10  Deutsch.  med.  Woch.,  1897,  xxiii,  228. 


BACTERIAL  CELLULAR  SUBSTANCE       55 

a  nucleoprotein  from  the  pest  bacillus.  From  the  effect 
of  methylene  blue  on  this  bacillus  and  its  microscopic 
appearance  after  extraction,  they  conclude  that  the  alkali 
removes  the  nuclein  without  destruction  of  the  cell  mem- 
brane. Our  studies  do  not  indicate  the  existence  of  a  mem- 
brane in  any  bacterial  cells.  Galeotti1  extracted  an  organ- 
ism similar  to  bacillus  ranicidus  with  1  per  cent,  potassium 
hydroxide,  and  obtained  a  protein  body  containing  from 
11.99  to  12.21  per  cent,  of  nitrogen  and  from  0.94  to  1.16 
per  cent,  of  phosphorus,  and  which  he  believed  to  be  a 
nucleoprotein.  The  percentage  of  phosphorus  increased 
after  several  reprecipitations. 

Aronson2  extracted  the  diphtheria  bacillus  with  from 
one-tenth  to  one-fifth  normal  alkali  in  the  cold,  at  100° 
and  at  130°,  precipitated  these  extracts,  first  with  acetic 
acid,  then  with  acidulated  alcohol,  sometimes  with  alcohol 
to  which  ether  and  a  little  acetic  acid  had  been  added. 
The  precipitate  with  acid  furnished  a  white  powder,  giving 
the  biuret,  xanthoproteic,  and  Adamkiewicz  reactions. 
Xanthin  bases,  a  pentose,  and  an  albumin  were  found 
among  its  decomposition  products,  thus  proving  the  presence 
of  a  nucleoprotein.  On  the  addition  of  alcohol  to  this 
acid  filtrate,  a  new  precipitate  was  formed,  and  this  on 
purification  yielded  nucleic  acid,  from  which  xanthin 
bases,  a  pentose,  and  a  phosphate  were  obtained.  Blandin3 
obtained  a  nuclein  and  a  nucleo-albumin  from  typhoid 
cultures,  but  these  may  have  come  from  constituents  of 
the  medium  or  from  the  bacterial  cells.  Klebs4  concluded 
that  nuclein  makes  up  a  large  part  of  the  tubercle  bacillus. 
He  extracted  the  bacilli  with  ether  and  benzol,  digested 
with  hydrochloric  acid  and  pepsin  and  dissolved  the  residue 
in  alkali.  By  precipitating  the  alkaline  extract  with  alcohol, 
he  obtained  a  nuclein  containing  from  8  to  9  per  cent,  of 
phosphorus.  Hahn5  rubbed  moist  tubercle  bacilli  with 

1  Zeitsch.  f.  physiol.  Chem.,  1898,  xxv,  48. 

2  Arch.  f.  Kinderheilkunde,  1900,  xxx,  23. 

3  La  Riforma  Medica,  April  17,  1901. 

4  Centralbl.   f.   Bakteriol.,    1896,   xx,  488. 
6  Munch,  med.  Woch.,  1897,  xliv,  1344. 


56  PROTEIN  POISONS 

sand,  mixed  with  water,  20  per  cent,  glycerin,  or  physio- 
logical salt  solution,  to  a  dough  consistency,  and  subjected 
the  mass  to  a  gradually  increased  pressure  of  from  400  to 
500  atmospheres.  The  clear,  expressed  fluid  contained  a 
large  quantity  of  coagulable  protein,  which  decomposed 
hydrogen  peroxide,  but  lost  this  property  on  being  heated. 
It  gave  the  protein  tests  and  behaved  like  a  nucleoprotein. 
Ruppel1  obtained  a  nuclein  containing  9.42  per  cent,  of 
phosphorus  from  the  residue  left  after  the  preparation  of 
his  tuberculosamin.  This  he  called  tuberculinic  acid,  and 
he  believed  that  it  exists  in  the  cell  partly  combined  with 
the  tuberculosamin  and  partly  free.  Levene2  prepared 
three  proteins  from  tubercle  bacilli  grown  on  protein-free 
medium.  The  dried  bacilli  were  ground  for  two  or  three 
days  in  a  porcelain  mill,  then  extracted  repeatedly  for  two 
days  with  an  8  per  cent,  solution  of  ammonium  chloride. 
These  proteins  coagulated  at  from  50°  to  64°,  72°  to  75°, 
and  94°  to  95°  respectively.  Ammonium  sulphate  pre- 
cipitated all  of  them;  sodium  chloride  only  the  first;  50 
per  cent,  magnesium  sulphate  the  first;  magnesium  sul- 
phate to  saturation  the  second,  but  not  the  third.  It 
required  less  acid  to  precipitate  the  first,  but  0.2  per  cent, 
hydrochloric  acid  precipitated  all  three.  The  third  was 
richer  in  phosphorus  than  the  others,  and  Levene  con- 
cluded that  the  tubercle  bacillus  consists  principally  of 
nucleoproteins,  one  of  which  differs  from  the  others  in 
that  it  is  not  precipitated  by  magnesium  sulphate  and  does 
not  give  the  biuret  reaction.  He  called  attention  to  the 
coincidence  between  the  coagulation  temperature  of  the 
first  protein  and  that  necessary  for  the  sterilization  of  the 
bacillus.  He  believed  that  tuberculin  is  a  specific  substance 
having  the  constitution  of  a  nucleoprotein.  He  also  made 
a  study  of  tuberculinic  acid,  finding  but  little  of  this  free 
in  mannite  synthetic  cultures,  but  considerable  in  beef- 
broth  cultures.  Samples  differ  in  composition,  and  experi- 
ments suggest  that  tuberculinic  acid  is  less  stabile  than 

1  Loc.  cit. 

2  Jour.  Med.  Research,  July,  1901,  135;  Medical  Record,  1898,  liv,  873. 


BACTERIAL  CELLULAR  SUBSTANCE       57 

any  other  known  nucleic  acid.  De  Schweinitz1  thought 
that  a  nucleo-albumin  is  the  fever-producing  agent  in  the 
tubercle  bacillus.  Maragliano2  made  an  aqueous  extract 
of  the  tubercle  bacillus  by  digesting  it  on  the  water-bath 
and  obtained  a  poisonous  substance.  This  comes  from  the 
autolytic  cleavage  of  the  bacillus. 

Carbohydrates. — One  of  the  earliest  studies  of  the  chemistry 
of  bacteria  was  made  by  Scheibler3  upon  leuconostoc  mesen- 
teroides.  The  viscous  growth  of  this  germ  in  beet  juice, 
after  extraction  with  alcohol  was  boiled  with  milk  of  lime. 
The  filtrate  furnished  a  gum  which  was  regarded  as  an 
anhydride  of  dextrose,  since  by  slow  hydrolysis  it  is  con- 
verted into  the  latter.  This  substance,  to  which  the  name 
of  dextran  was  given,  is  a  white,  amorphous  powder,  soluble 
in  water  and  dextrorotatory,  having  three  times  the  rotary 
power  of  cane  sugar.  Scheibler  stated  that  this  germ  con- 
tains ash,  fat,  water,  dextra,  and  a  substance  containing 
nitrogen,  believed  to  be  protagon  or  some  closely  related 
body.  Kramer4  separated  from  the  slime  of  bacillus  viscosus 
sacchari  two  modifications  of  a  carbohydrate  of  the  formula 
C6Hi0O5.  Both  were  optically  active,  and  on  being  boiled 
with  acid  reduced  Fehling's  solution.  Ward  and  Green5 
found  that  a  species  of  bacterium  from  Madagascar  sugar- 
cane secretes  invertose.  In  sugar  solutions  it  produces 
a  viscous  growth  that  gives  an  opalescent  solution  in  water, 
which,  when  treated  with  alcohol,  yields  a  bulky  flocculent 
precipitate,  found  to  contain  two  carbohydrates,  one  of 
which  gives  an  osazone,  and  is  optically  active,  while  the 
other  is  inactive.  They  regard  these  bodies  as  related  to, 
but  not  identical  with,  Scheibler's  dextran,  and  are  not 
certain  whether  they  are  products  of  the  vital  processes 
of  the  organism  or  are  cleavage  products.  Vincenzi6  could 

1  Bulletin  No.  7,  Bureau  of  Animal  Industry;  Jour.  Amer.  Chem.  Society, 
1897,  xix,  782. 

2  Berl.  klin.  Woch.,  1899,  xxxvi,  385. 

3  Zeitsch.  f.  Rubenzuckerindustrie,  1874,  xxiv,  309. 

4  Monats.  f.  Chem.,  1889,  x,  467. 
6  Proc.  Roy.  Soc.,  1899,  Ixv,  65. 

6  Zeitsch.  f.  physiol.  Chem.,  1887,  ix,  181. 


58  PROTEIN  POISONS 

find  no  evidence  of  cellulose  in  bacillus  subtillis,  but  Drey- 
fuss1  heated  masses  of  this  organism  to  180°  with  concen- 
trated alkali  and  from  this  extract  obtained  a  substance 
that  reduced  Fehling's  solution  and  from  which  crystals 
of  glucosazone  were  prepared.  From  this  Dreyfuss  con- 
cluded that  cellulose  is  present.  However,  this  conclusion 
is  hardly  justifiable.  Like  results  were  obtained  from 
pyogenic  bacilli.  Hammerschlag2  concluded  that  the 
tubercle  bacillus  contains  cellulose.  The  cell  substance, 
previously  extracted  with  alcohol,  ether,  and  1  per  cent, 
potassium  hydroxide,  was  dissolved  in  concentrated  sul- 
phuric acid,  diluted,  and  boiled,  after  which  it  reduced 
Fehling's  solution.  A  second  portion  was  treated  with 
potassium  chlorate  and  nitric  acid,  but  most  of  the  substance 
remained  undissolved.  A  third  portion  was  partly  dissolved 
in  ammoniacal  copper  solution.  Hammerschlag  stated 
that  if  one  assumes  that  the  nitrogenous  material  in  the 
tubercle  bacillus  is  all  protein  and  that  the  protein  contains 
16  per  cent,  of  nitrogen,  this  bacillus  contains  36.9  per 
cent,  of  protein,  28.1  per  cent,  of  cellulose,  27  per  cent,  of 
substance  soluble  in  alcohol,  and  8  per  cent,  of  ash.  Nishi- 
mura3  thought  that  he  found  hemicellulose  in  a  water 
bacillus,  in  prodigiosus,  and  in  staphylococcus  pyogenes. 
De  Schweinitz  and  Dorset4  extracted  dried  tubercle  bacilli 
with  alcohol,  digested  the  residue  with  1.25  per  cent,  sodium 
hydroxide  for  from  forty  to  sixty  minutes,  washed  the 
residue,  then  digested  with  1.25  per  cent,  sulphuric  acid, 
washed,  dried,  and  ignited.  The  loss  by  ignition  they 
calculated  should  give  the  cellulose.  Accordingly,  they 
reported  6.95  per  cent,  cellulose  in  the  tubercle  bacillus. 
However,  this  conclusion  is  hardly  accepted  by  these 
authors  themselves,  since  in  the  same  paper  they  state  that 
cellulose  is  probably  present  in  small  amount  in  the  tubercle 
bacillus,  and  not  present  in  the  bacillus  of  glanders.  Brown5 
boiled  "the  membrane"  of  bacterium  xylinum  twenty 

1  Loc.  cit.  2  Loc.  cit.  3  Loc.  cit. 

4  Jour.  Amer.  Chem.  Soc.,  1895,  xvii,  605;  ibid.,  1896,  xviii,  449;  ibid., 
1897,  xix,  782;  ibid.,  1898,  xx,  618. 
6  Jour.  Chem.  Soc.,  1886,  xlix,  432;  ibid.,  1887,  li,  643. 


BACTERIAL  CELLULAR  SUBSTANCE  59 

minutes  with  10  per  cent,  potassium  hydroxide  and  found 
that  this  digested  the  bacteria  but  left  "the  film"  appar- 
ently unchanged.  The  residue  was  washed  with  dilute 
hydrochloric  acid,  then  with  water,  and  treated  with 
bromine  according  to  Miiller's  method  for  obtaining  cellu- 
lose. The  product  seemed  to  be  identical  with  that  from 
cotton,  dissolving  in  ammoniacal  copper  solution  and  in 
strong  sulphuric  acid.  It  gave  a  reducing  sugar,  dextro- 
rotatory, even  when  grown  on  media  containing  only  levoro- 
tatory  substances.  Analysis  showed  a  close  agreement 
with  cellulose,  and  Brown  regarded  this  as  a  cellulose  proper, 
differing  from  the  metacellulose  usually  found  in  yeast  and 
the  fungi.  No  trace  of  dextran  was  found.  Brown  regarded 
the  formation  of  cellulose  as  a  process  of  assimilation  and 
not  of  fermentation.  Bendix1  extracted  dried  bacilli  with 
5  per  cent,  hydrochloric  acid  over  the  free  flame,  cooled, 
made  alkaline,  then  acidulated  with  acetic  acid  in  order  to 
precipitate  the  protein.  The  filtrate  gave  with  phenyl- 
hydrazin  an  osazone  which  when  purified  melted  at  153° 
to  155°,  thus  showing  it  to  be  pentosazone.  It  gave  the 
orcin  and  optical  tests  for  pentose.  He  obtained  pentose 
from  diphtheria  and  tubercle  bacilli,  also  from  mixed  fecal 
bacteria,  but  not  from  the  typhoid  bacillus.  The  pentose 
exists  in  the  nucleoprotein.  Aronson2  found  a  nucleo- 
protein  containing  pentose  in  alkaline  extracts  of  diphtheria 
bacilli.  This  author  stated  that  the  residue  after  complete 
extraction  with  alkali  contains  carbohydrate  which  is 
dextrorotatory  and  yields  an  osazone;  it  is  neither  cellulose 
nor  chitin. 

Meyer3  came  to  the  conclusion  that  in  some  species  of 
bacteria  fat  is  stored  up,  while  in  others  complex  carbo- 
hydrates take  their  place.  One  species  grown  on  barley 
gave  a  substance,  colored  blue  by  iodine,  and  easily  soluble 
in  malt  diastase  and  in  saliva.  Bacillus  subtilis  gave 
a  body  that  is  colored  red  by  iodine  and  is  dissolved  by 
saliva,  and  on  boiling  with  dilute  sulphuric  acid.  Meyer 

1  Deutsch.  med.  Woch.,  1901,  xxvii,  18. 

2  Loc.  cit.  3  Flora,  1889,  432. 


60  PROTEIN  POISONS 

thought' that  this  might  be  either  glycogen  or  amylodextrin. 
He  also  obtained  a  substance  that  he  considered  a  mixture 
of  much  amylodextrin  and  a  little  /3-amylose.  It  is  easily 
extracted  from  the  cell  with  water.  It  may  be  remarked 
here  that  our  researches  have  shown  that  substances 
removable  from  the  cell  by  physical  solvents  constitute  no 
part  of  the  actual  cell  protein.  The  carbohydrate  in  the 
essential  part  of  the  cell  is  a  constituent  of  the  protein 
molecule.  All  carbohydrates,  fats,  waxes,  and  inorganic 
salts  that  may  be  washed  out  of  the  cell  substance  are 
either  no  part  of  the  cell  proper  or  result  from  autolytic 
changes  in  the  cell  molecules. 

Levene1  obtained  a  glycogen-like  body  from  the  tubercle 
bacillus.  The  cell  substance  was  extracted  with  salts,  or 
better,  with  alkali,  the  albumins  removed  from  the  extract 
with  picric  and  acetic  acids,  the  nuclein  and  carbohydrate 
carried  down  together  with  alcohol,  and  then  separated  by 
means  of  copper  chloride.  The  glycogen  thus  obtained  is 
soluble  in  water,  gives  the  iodine  color  test,  and  reduces 
Fehling's  solution  after  being  boiled  with  dilute  mineral 
acid.  Emmerling2  prepared  chitin  or  a  closely  related  body 
from  the  zooglea  of  bacterium  xylinum.  From  110  grams 
of  moist,  impure  material  he  secured  0.2  gram  of  crystal- 
line glucosamine  hydrochloride.  Helbing3  concluded  that 
chitin  makes  up  a  large  part  of  the  tubercle  bacillus,  and 
to  this  constituent  he  attributed  the  peculiar  staining 
properties  of  this  organism.  He  was  clearly  wrong  in  this 
inference.  All  the  early  work  on  the  carbohydrate  con- 
stituent of  the  bacterial  cell,  when  the  material  was  grown 
on  media  containing  carbohydrate,  must  be  regarded  as  not 
possessed  of  practical  value. 

Fat,  Wax,  etc. — In  the  earlier  studies  of  the  chemistry 
of  bacterial  cells  it  was  assumed  that  the  alcoholic  and 
ethereal  extracts  consisted  of  fats  exclusively.  Kramer4 

1  Loc.  cit.  2  Berichte,  1899,  xxxii,  541. 

3  Deutsch.  med.  Woch.,  xxvi,  Vereinsbeilage,  1900,  133. 

4  Arch.  f.  Hygiene,  1891,  xiii,  71;  ibid.,  1893,  xvi,  151;  ibid.,  1895,  xxii, 
167;  ibid.,  1897,  xxviii,  1. 


BACTERIAL  CELLULAR  SUBSTANCE  61 

noted  that  such  an  extract  had  the  appearance  of  fat  and 
melted  not  much  over  40°.  Hammerschlag1  obtained  from 
the  tubercle  bacillus  free  fatty  acids  that  melt  at  63°  and 
concluded  that  the  fat  of  this  organism  consists  mainly  of 
tripalmatin  and  tristearin,  and  that  it  contains  little  or 
no  triolein.  Nishimura2  obtained  from  the  alcoholic  and 
ethereal  extracts  of  his  water  bacillus  a  putty-like  mass 
with  the  properties  of  lecithin.  Meyer3  found  that  the 
fat  in  bacillus  tumescens  gradually  increases  until  spore 
formation  occurs,  when  it  disappears,  the  spores  also  being 
free  from  fat.  Klebs4  found  in  the  tubercle  bacillus  20.5 
per  cent,  of  a  red  fat,  melting  at  42°,  and  1.14  per  cent,  of  a 
white  fat  melting  above  50°;  the  latter  being  insoluble  in 
ether,  but  soluble  in  benzol.  De  Schweinitz  and  Dorset"' 
saponified  fats  from  the  tubercle  bacillus  and  from  the 
melting-points  of  the  acids  concluded  that  the  fat  of  this 
organism  contains  palmitic  and  arachidic  acids,  while 
that  of  the  glanders  bacillus  contains  oleic  and  palmitic. 
They  also  found  a  crystalline  acid,  for  which  they  suggested 
the  name  tuberculinic  acid,  though  this  is  quite  different 
from  Ruppel's  nucleic  acid.  This  new  fatty  acid  was 
obtained  mainly  from  the  culture  medium,  only  in  small 
amounts  from  the  bacilli.  The  crystals  are  prismatic  or 
needles,  melting  at  161°  to  164°,  readily  soluble  in  water, 
alcohol,  and  ether,  and  not  responsive  to  the  biuret  test. 
Analysis  showed  close  correspondence  to  the  formula, 
C7H10O4.  The  authors  called  attention  to  the  similarity  in 
composition  and  properties  of  this  body  to  teraconic  acid, 
and  suggested  that  this  may  be  the  substance  which  is 
responsible  for  the  coagulation  necrosis  and  that  it  is  the 
temperature  reducing  substance.  In  a  later  paper  they 
described  a  crude  fat  extracted  from  the  tubercle  bacillus 
and  from  which  they  obtained  an  acid  melting  at  62°, 
unchanged  by  recrystallization.  In  concluding  they  decided 
that  the  fat  of  the  tubercle  bacillus  consists  principally  of 
a  glyceride  of  palmitic  acid  with  a  minute  amount  of  the 

1  Loc.  cit.  2  Loc.  cit.  3  Loc.  cit. 

4  Loc.  cit.  5  Loc.  cit. 


62  PROTEIN  POISONS 

glyceride  of  a  volatile  fatty  acid  to  which  cultures  of  this 
bacillus  owe  their  characteristic  odor,  also  a  very  small 
amount  of  an  acid  (probably  lauric)  melting  at  42°  to  43°, 
and  an  unusually  high  melting  acid,  one  apparently  with  a 
larger  carbon  content  than  any  before  noted  in  plants. 
Ruppel1  obtained  three  extracts  from  the  tubercle  bacillus 
by  using  successively  cold  alcohol,  hot  alcohol,  and  ether. 
The  first  contains  free  fatty  acids  and  a  fat  melting  between 
60°  and  65°,  easily  saponified  and  decomposed  into  a  free 
acid  and  a  higher  alcohol.  The  second  contained  a  waxy 
mass,  saponified  with  difficulty,  and  which  seemed  to  be 
the  ester  of  a  fatty  acid  and  a  high  alcohol.  The  third 
melted  at  65°  to  67°,  and  had  an  odor  resembling  that  of 
beeswax.  Aronson2  obtained  from  tubercle  bacilli,  by 
means  of  a  mixture  of  five  parts  of  ether  and  one  of  abso- 
lute alcohol,  a  yellowish-brown  tenacious  mass,  constituting 
from  20  to  25  per  cent,  of  the  dried  bacilhis.  From  the 
growth  of  several  hundred  liters  of  culture  70  grams  were 
secured.  This  contained  17  per  cent,  of  free  fatty  acids. 
The  remainder  was  wax,  not  acid  and  glycerin,  but  esters 
of  acid  and  alcohol  insoluble  in  water.  Most  of  this  wax  is 
not  in  the  cells,  but  lies  around  and  between  them.  Levene3 
found  almost  30  per  cent,  of  fat  or  wax  in  tubercle  bacilli. 
Kresslig4  extracted  tubercle  bacilli  successively  with  ether, 
chloroform,  benzol,  and  alcohol,  and  obtained  38.95  per 
cent,  of  fatty  and  waxy  substances.  Repeated  extraction 
with  chloroform  gave  a  dark  brown  mass  of  the  consistency 
and  color  of  beeswax  and  melting  at  46°.  He  found  14.38 
per  cent,  of  free  fatty  acid,  77.25  per  cent,  of  neutral  fat 
and  esters  of  fatty  acids,  and  some  volatile  fatty  acid, 
probably  butyric.  He  decided  that  the  fat  of  the  tubercle 
bacillus  is  quite  different  from  that  obtained  from  any 
other  source. 

Reducing    Action    of    Bacteria. — Although    the    general 
subject  of  the  reducing  action  of  bacteria  scarcely  falls 


1  Loc.  cit.  2  Berl.  klin.  Woch.,  1896,  xxxv,  484. 

3  Loc.  cit.  4  Centralbl.  f.  Bakteriol.,  1901,  xxx,  897. 


BACTERIAL  CELLULAR  SUBSTANCE       63 

within  the  domain  of  this  work,  it  may  be  well  to  mention 
the  results  of  a  few  investigations.  W.  Smith1  found  that 
many  bacilli,  including  the  colon,  decolorize  methylene 
blue,  sodium  indigo  sulphate,  litmus,  etc.  He  concluded 
that  this  reducing  action  is  common  to  all  bacteria,  both 
aerobic  and  anaerobic;  that  the  velocity  of  reduction 
depends  upon  the  number  of  bacteria  and  the  temperature; 
that  it  is  a  function  of  the  bacterial  plasma,  and  that  the 
reducing  substance  does  not  diffuse  into  the  culture  medium, 
but  that  the  cell  retains  this  property  for  a  time  after 
death.  Klett,2  testing  this  reducing  action  of  bacteria  on 
sodium  silicate,  tellurate,  and  some  other  salts,  also  con- 
cluded that  the  reducing  agent  exists  in  the  cell,  and  is  not 
found  among  the  cleavage  products.  Jegunow3  showed 
that  hydrogen  sulphide  is  formed  by  the  reducing  action 
of  bacteria  on  sulphates  and  on  organic  bodies  contain- 
ing sulphur.  Sulphur  bacteria  oxidize  hydrogen  sulphide 
and  store  sulphur  in  the  form  of  oily  spheres,  which  may 
constitute  as  much  as  90  per  cent,  of  the  cell  substance. 
This  sulphur  is  oxidized  to  sulphuric  acid,  thus  serving  as 
a  source  of  energy  in  the  vital  processes  of  the  bacterium. 
The  sulphuric  acid  is  neutralized  by  carbonates  and  sepa- 
rated as  a  sulphate;  then  by  bacterial  activity  the  sulphate 
is  reduced,  thus  forming  a  complete  cycle.  If  the  bacteria 
can  obtain  no  sulphur  they  use  that  stored  up  in  their 
cells  and  die  in  from  one  to  two  days. 

The  above  is  a  resume  of  the  work  done  on  the  chemistry 
of  bacterial  cells  up  to  the  time  when  our  work  was  begun. 
It  should  be  clearly  understood  that  we  are  not  now  con- 
cerned with  the  cleavage  products  of  bacteria  produced  in 
the  media  in  which  they  groAv.  This  subject  is  discussed  in 
Cellular  Toxins  by  Vaughan  and  Novy  (fourth  edition,  1902). 

Moisture,  Ash,  and  Nitrogen. — Leach,4  in  studying  the 
chemistry  of  the  cellular  substance  of  the  colon  bacillus 

1  Centralbl.  f.  Bakteriol.,  1896,  xix,  181. 

2  Zeitsch.  f.  Hygiene,  1900,  xxxiii,  137. 

3  Annuaire  g6ologique  et  mineralogique  de  la  Russie,  1900,  ii,  157. 
*  Jour.  Biol.  Chem.,  1906,  i,  463. 


64  PROTEIN  POISONS 

prepared  in  our  laboratory,  makes  substantially  the  follow- 
ing statement.  The  cell  substance,  prepared  by  the  method 
already  described,  takes  up  moisture  readily  and  holds  it 
tenaciously,  but  may  be  dried  to  constant  weight  by  heating 
small  amounts  in  a  steam-drying  oven  for  many  days  at  from 
85°  to  95°.  If  the  temperature  falls  to  60°,  it  may  absorb 
moisture  even  in  the  oven.  One  sample  was  heated  to  105° 
during  working  hours  for  three  days  and  kept  in  a  desiccator 
during  the  intervals;  it  increased  in  weight.  Drying  in 
vacua  over  sulphuric  acid  is,  on  the  whole,  the  most  satis- 
factory method,  although  it  may  require  days  and  even 
weeks. 

The  dried  cellular  substance  burns  with  a  flame,  forming 
volatile  and  liquid  products,  giving  off  odors  characteristic 
of  nitrogen  compounds,  and  finally  leaving  a  greenish  ash. 
Two  determinations  gave  the  following  results: 

0.346  gram  gave  0.0296  gram  of  ash,  or  8.55  per  cent. 
0.496  gram  gave  0.0431  gram  of  ash,  or  8.68  per  cent. 

Values  reported  for  other  bacteria  vary  from  3  per  cent, 
in  putrefactive  organisms  to  13  per  cent,  in  prodigiosus,  or, 
by  using  special  media,  to  nearly  30  per  cent,  in  the  cholera 
bacillus;  while  in  the  tubercle  bacillus  the  ash  has  been 
found  to  vary  from  1.77  to  5.92  per  cent,  according  to 
conditions.  The  ash  from  the  colon  bacillus,  as  we  have 
prepared  it,  contains  sodium,  potassium,  small  amounts  of 
calcium,  aluminum,  copper,  and  phosphates.  A  slight 
residue  insoluble  in  acid  is  probably  silica.  Sulphate  is 
present  in  so  small  an  amount  that  it  may  escape  detection, 
and  chloride  has  not  been  found.  In  comparison  with  the 
data  obtained  with  other  bacteria,  these  findings  are  note- 
worthy only  in  the  absence  of  magnesium,  and  in  the 
presence  of  copper  and  aluminum.  Presumably  the  former 
comes  from  the  tanks  and  the  latter  from  the  agar. 

Phosphorus  was  the  only  constituent  of  the  ash  quanti- 
tatively determined.  The  ash  was  dissolved  in  nitric  acid, 
the  phosphate  precipitated  with  ammonium  molybdate, 
dissolved  in  ammonia,  precipitated  with  magnesia  mixture, 


BACTERIAL  CELLULAR  SUBSTANCE       65 

and    weighed    as    pyrophosphate.      The    following    results 
were  obtained: 

Weight  of  Weight  of 

sample.  phosphate.  Weight  of  P.  Per  cent,  of  P. 

0 . 496  gram  0 . 0475  gram  0 . 01323  gram  2 . 68 

0 . 346  gram  0 . 0380  gram  0 . 0 1 059  gram  3 . 06 

The  mean  of  these  determinations,  2.87  per  cent.,  agrees 
quite  closely  with  Levene's  finding,  2.67  per  cent,  of  phos- 
phorus in  the  tubercle  bacillus  from  mannite  cultures. 
Most  observers  report  smaller  results,  but  it  should  be 
noted  that  our  samples  are  free  from  fat  and  wax,  and 
therefore  the  percentage  is  higher  than  if  calculated  for 
the  cells  not  previously  extracted  with  alcohol  and  ether. 

In  view  of  the  fact  that  the  cellular  substance,  notwith- 
standing the  washings  to  which  it  has  been  subjected, 
cannot  be  regarded  as  chemically  pure,  we  have  not 
wasted  time  in  making  elementary  analyses.  Wheeler 
has  collected  the  nitrogen  and  ash  determinations  made 
in  bacterial  cellular  substance  in  this  laboratory  and  has 
arranged  them  in  the  following  table: 

Substance.  Per  cent,  of  nitrogen.         Per  cent,  of  ash. 

Typhoid 11.55  5.70 

Colon 10.65  8.615 

Colon 8.38 

7.20   (air-dried) 

Tuberculosis 10.55  11.47 

9.27  (air-dried)  9.98  (air-dried) 

Anthrax 10.285  7.76 

Subtilis 5.964  10.83 

Proteus  vulgaris 6.791  10.88 

Ruber  of  Kiel 10 . 655  6.71 

Megaterium 8.349  10.18 

Pyocyaneus 10.843  9.04 

Violaceus 11.765  6.90 

Sarcina  aurantiaca       ....      11.460  6.40 

As  will  be  seen,  the  nitrogen  varies  from  5.964  per  cent, 
in  subtilis  to  11.765  per  cent,  in  violaceus,  and  the  ash 
from  5.7  per  cent,  in  the  typhoid  bacillus  to  11.47  per  cent, 
in  the  bacillus  tuberculosis. 
5 


66  PROTEIN  POISONS 

Nicolle  and  Alilaire1  give  the  following  table,  showing  the 
percentage  of  water,  nitrogen,  substance  soluble  in  acetone, 
and  phosphorus  in  the  bacterial  cells  named: 


Organism.               *g  ^               *3  g  co_;  E 

.+3  X                                    +3     0)  -*J     CO     O 

fl  do  "       fe          "-1      «            ^  *-• 

S  g«  g.nP         s.a.2          go 


B 

ll 

PH 

._ 

PH 

n  55o 
• 
PH 

t,  <a  o 

£ 

& 

(-  3 

& 
PH 

B.  glanders  . 

76.49 

10.47 

11.69 

8.59 

3.10 

2.530 

8.0 

B.  chicken  cholera 

79.35 

10.79 

7.54 

6.30 

1.24 

2.370 

7.5 

B.  cholera 

73.38 

9.79 

8.70 

6.77 

1.93 

2  370 

7.5 

B.  of  Shiga  .      .      . 

78.21 

8.89 

12.80 

10.57 

2.23 

1.570 

5.0 

B.  proteus    . 

79.99 

10.73 

10.87 

7.10 

3.77 

1.580 

5.0 

B.  typhoid   . 

78.93 

8.28 

15.44 

10.64 

4.80 

1.160 

3.5 

B.  anthrax   . 

81.74 

9.22 

6.31 

1.48 

4.83 

0.948. 

3.0 

B.     pseudotubercu- 

losis     .... 

78.83 

10.36 

15.63 

10.31 

5.32 

0.793 

2.5 

B.  pneumonia    . 

85.55 

8.33 

15.45 

7.36 

8.06 

0.790 

2.5 

B.  coli     .... 

73.35 

10.32 

15.25 

11.77 

3.48 

0.790 

2.5 

B.  prodigiosus   . 

78.00 

10.55 

9.00 

6.60 

2.40 

0.474 

1.5 

B.  psittacosis     . 

78.05 

9.55 

11.08 

•  7.03 

4.05 

0.474 

1.5 

B.  diphtheria     . 

84.50 

7.04 

5.23 

1.81 

0.158 

0.5 

B.  pyocyaneus  . 

74.99 

9.79 

15.77 

10.67 

5.10 

0.157 

0.5 

B.  lymphangitis    .  . 

77.90 

9.17 

6.83 

2.53 

4.30 

0.157 

0.5 

Froberg's  yeast 

69.25 

10.00 

4.22 

2.92 

1.30 

0.000 

0.0 

Chlorella  vulgaris   . 

63.06 

3.96 

21.10 

12.81 

8.29 

0.000 

0.0 

Carbohydrates. — In  no  case  have  we  been  able  to  detect 
cellulose  in  the  bacterial  cell  substance.  Wheeler  made 
special  search  for  it  in  sarcina  lutea.  Twenty  grams  of 
substance  was  autoclaved  with  25  parts  (500  c.c.)  of  10 
per  cent,  potassium  hydroxide  at  120°,  first  for  thirty 
minutes  and  then  for  an  hour.  There  remained  a  consider- 
able residue  which  no  longer  gave  the  protein  reactions, 
but  did  respond  to  the  carbohydrate  test  with  a-naphthol, 
although  it  did  not  reduce  Fehling's  solution  even  after 
prolonged  boiling  with  dilute  hydrochloric  acid.  Cellulose 
could  not  be  detected  by  any  of  the  tests  employed. 
Schweitzer's  reagent  failed  to  dissolve  it,  and  it  gave  no 

1  Annales  de  1'Institut  Pasteur,  1909,  xxiii,  547. 


BACTERIAL  CELLULAR  SUBSTANCE       67 

color  with  iodine  even  after  treatment  with  sulphuric  acid. 
A  portion  was  dried,  and  heated  with  soda  lime,  when  it 
evolved  a  gas  which  turned  red  litmus  paper  blue,  thus 
indicating  nitrogen  which  had  been  reduced  to  ammonia. 
The  odor  of  burning  feathers  also  indicated  the  presence 
of  nitrogen.  From  these  results  it  was  concluded  that  the 
residue  left  after  extraction  of  the  cellular  substance  with 
10  per  cent,  potassium  hydroxide  at  120°  contains  a  car- 
bohydrate, but  there  is  nothing  to  indicate  that  it  is  cellu- 
lose. Leach  made  a  search  for  cellulose  in  the  cells  of  the 
colon  bacillus,  with  like  negative  results. 

There  are  two  carbohydrate  bodies  in  bacterial  cellular 
substances.  One  of  these  furnishes  a  reducing  sugar  after 
being  boiled  with  dilute  mineral  acid,  while  the  other  does 
not.  The  former  may  be  extracted  from  the  cells  with 
either  alkali  or  acid,  better  with  the  former.  In  Wheeler's 
studies  of  sarcina  lutea  the  portion  soluble  in  10  per  cent, 
potassium  hydroxide  was  filtered  through  paper,  acidified 
with  hydrochloric  acid,  and  treated  with  three  volumes  of 
95  per  cent,  alcohol,  which  produced  an  abundant  white, 
curdy,  sticky  precipitate.  Precipitation  by  means  of 
alcohol  with  acetic  and  picric  acids  was  also  tried,  but  did 
not  prove  satisfactory.  Purification  was  attempted  by 
repeated  solution  and  precipitation  with  acidified  alcohol, 
but  the  quantity  was  diminished  each  time  on  account  of 
its  relative  solubility  in  dilute  alcohol.  After  filtering, 
washing,  and  drying  in  an  atmosphere  of  carbon  dioxide, 
the  powder  obtained  weighed  only  0.7872  grams.  It  con- 
tained phosphorus,  responded  to  the  carbohydrate  test, 
and  reduced  Fehling's  solution  after  prolonged  boiling  with 
dilute  mineral  acid.  Its  phosphorus  content  was  deter- 
mined and  found  to  be  0.861  per  cent.,  which  is  too  low  to 
indicate  the  presence  of  nucleic  acid  or  nuclein  in  anything 
like  a  pure  condition.  The  same  investigator  at  one  time 
took  300  grams  of  the  colon  germ  substance  and  heated 
on  the  water-bath  with  six  liters  of  2  per  cent,  potassium 
hydroxide.  The  extract  was  filtered  through  paper  and 
acidified  with  acetic  acid.  The  precipitate  produced, 


68  PROTEIN  POISONS 

presumably  protein,  was  filtered  out  after  standing  twenty- 
four  hours,  and  was  so  small  that  it  was  lost  on  the  filter 
paper. 

The  filtrate  was  then  poured  into  three  volumes  of  95 
per  cent,  alcohol  acidified  to  the  extent  of  0.5  per  cent,  with 
hydrochloric  acid.  This  formed  a  heavy,  white,  curdy, 
fibrous  precipitate,  which  was  filtered,  washed  acid-free 
with  alcohol  and  then  with  ether.  It  was  purified  by 
repeated  solution  in  0.5  per  cent,  alkali  and  precipitation 
with  alcohol.  Finally  there  was  obtained  a  fine  white 
powder  amounting  to  something  less  than  10  per  cent,  of 
the  original  cellular  substance,  but  much  had  been  lost  by 
its  partial  solubility  in  dilute  alcohol.  This  powder  consists 
almost  wholly  of  a  carbohydrate  which  is  converted  into 
a  reducing  sugar  after  prolonged  boiling  with  dilute  mineral 
acid.  However,  it  contained  5.9  per  cent,  of  ash  and  0.194 
per  cent,  of  phosphorus.  Solutions  of  this  powder  give 
none  of  the  protein  reactions,  with  the-  exception  of  the 
xanthoproteic,  to  which  they  responded  imperfectly. 

Leach  prepared  the  same  body,  but  with  a  higher  phos- 
phorus content,  from  the  colon  bacillus.  The  cellular 
substance,  after  repeated  extraction  with  dilute  (1  to  5 
per  cent.)  sulphuric  acid,  was  extracted  upon  the  water- 
bath  or  over  a  free  flame  with  from  2  to  4  per  cent,  of 
sodium  hydroxide.  *The  alkaline  extract,  after  filtration, 
was  neutralized  with  hydrochloric  acid  and  poured  into  95 
per  cent,  alcohol.  A  light  colored,  flocculent  precipitate 
was  obtained.  This  turned  dark  on  the  exposure  to  air 
incident  to  filtration.  It  was  twice  dissolved  in  0.5 
per  cent,  potassium  hydroxide  and  reprecipitated  with 
acidified  alcohol.  Each  time  the  fresh  precipitate  was 
white  or  nearly  so,  but  the  utmost  care  in  filtering,  even 
in  an  atmosphere  of  carbon  dioxide,  did  not  prevent  its 
turning  dark.  The  solution  in  alkali  gave  the  xantho- 
proteic and  furfurol  tests,  but  neither  the  biuret  nor  Millon 
test.  Copper  chloride  gave  a  precipitate,  but  picric  acid 
and  platinum  chloride  did  not.  The  solution  was  accord- 
ingly acidified  with  picric  and  acetic  acids,  copper  chloride 


BACTERIAL  CELLULAR  SUBSTANCE  69 

added,  and  the  mixture  poured  into  three  volumes  of 
alcohol.  An  aqueous  solution  of  the  precipitate  thus 
obtained  did  not  reduce  Fehling's  solution,  but  after  boiling 
with  hydrochloric  acid  it  reduced  both  Fehling's  and 
Nylander's  solutions,  and  also  gave  the  furfurol  test,  thus 
showing  the  presence  of  a  carbohydrate.  The  original 
powder  burned  readily,  puffing  up  and  glowing  as  does 
nucleic  acid,  then  fusing  and  leaving  a  dark  ash.  Two 
determinations  of  phosphorus  gave  the  following: 

Weight  of  Weight  of 

sample.  pyrophosphate.  Weight  of  P.  Per  cent,  of  P. 

0 . 4723  gram            0 . 0524  gram  0 . 01450  gram           3 . 09 

0.6469  gram            0.0685  gram  0.01895  gram           2.93 

In  our  attempts  to  extract  the  poisonous  groups  from 
bacterial  proteins  this  carbohydrate  gave  us  great  trouble. 
It  is  readily  soluble  in  water,  whether  acid  or  alkaline,  and 
more  or  less  soluble  in  alcohol,  the  degree  of  solubility 
depending  upon  the  strength  of  the  alcohol.  In  absolute 
alcohol  it  is  insoluble,  but  it  cannot  be  precipitated  com- 
pletely from  aqueous  solution  by  the  addition  of  alcohol. 
Concentrated  solutions  and  residues  obtained  by  evapora- 
tion in  vacuo  are  sticky  and  unsatisfactory  in  all  attempts 
at  purification.  As  we  ascertained  after  much  experimenta- 
tion, the  poisonous  group  in  the  protein  molecule  is  freely 
soluble  in  absolute  alcohol,  and  finally  when  we  disrupted 
the  protein  molecule  with  a  dilute  solution  of  alkali  in 
absolute  alcohol  we  secured  a  complete  separation  of  the 
poisonous  group  from  both  carbohydrates.  Therefore  the 
best  material  in  which  the  bacterial  or  other  protein  carbo- 
hydrates can  be  studied  is  the  non-poisonous  portion  after 
complete  removal  of  the  poisonous  group  by  heating  the 
protein  repeatedly  with  a  2  per  cent,  solution  of  sodium 
hydroxide  in  absolute  alcohol.  This  method  will  be  dis- 
cussed in  detail  later,  but  it  needs  to  be  stated  here  that 
when  this  is  done  the  poisonous  group,  free  from  any  trace 
of  either  carbohydrate,  goes  into  solution  in  the  alkaline 
alcohol,  while  the  non-poisonous  part,  or  the  haptophor,  as 


70  PROTEIN  POISONS 

we  have  designated  it,  containing  all  the  carbohydrates, 
remains  insoluble  in  this  menstruum. 

Leach1  has  studied  the  carbohydrate  in  the  haptophor 
portion  of  the  cellular  substance  of  the  colon  bacillus. 
Gram  samples  of  the  haptophor  portion  were  dissolved  in 
water  containing  a  little  alkali,  neutralized  with  hydro- 
chloric acid  to  definite  strength  and  heated  on  a  water-bath 
in  a  flask  with  a  reflux  condenser.  The  hydrolyzed  solution 
was  neutralized  and  titrated  with  Fehling's  solution.  Al- 
though there  is  undoubtedly  some  pentose  present,  there 
is  no  proof  that  the  reducing  substance  is  all  carbohydrate. 
However,  for  purposes  of  comparison  the  reducing  sub- 
stance was  calculated  as  xylose.  In  order  to  find  conditions 
giving  the  maximum  yield,  amount  and  strength  of  acid 
as  well  as  time  of  boiling  were  varied  as  shown  in  the 
following  table: 

REDUCING  POWER  OF  COLON  HAPTOPHOR 

No.  of  Amount  of  Per  cent,  of  Hours  Per  cent,  calculated 

sample.  HC1.  HC1.  boiled.  as  xylose. 

1  26.0  c.c.  1.0  1  7.05 

2  38.8  c.c.  2.5  1  16.45 

3  38.5  c.c.  2.5  2  21.56 

4  .   38.5   c.c.  2.5  4  23.12 

5  72.0  c.c.  2.5  3  23.93 

6  72.0  c.c.  2.5  9  23.53 

As  shown  by  these  figures  the  maximum  amount  of 
reducing  substance  was  obtained  by  using  2.5  per  cent, 
acid,  and  boiling  for  three  hours.  Longer  heating  changes 
the  result  very  little. 

Attempts  were  made  to  separate  this  carbohydrate  from 
the  other  constituents  of  the  haptophor  of  the  colon  bacillus. 
A  5  per  cent,  aqueous  solution  of  the  haptophor  was  poured 
into  four  volumes  of  absolute  alcohol,  containing  10  c.c.  of 
hydrochloric  acid  and  100  c.c.  of  ether  per  liter.  After 
settling,  the  supernatant  liquid  was  siphoned  off  and  the 
precipitate  (known  as  G)  collected  with  suction,  washed 

i  Jour.  Biol.  Chem.,  1907,  iii,  443. 


BACTERIAL  CELLULAR  SUBSTANCE       71 

with  alcohol  containing  ether,  then  with  ether,  dried,  and 
pulverized.  The  yield  from  50  grams  of  haptophor  was  19 
grams.  This  was  twice  dissolved  in  water  made  faintly 
alkaline  with  sodium  acid  carbonate,  and  reprecipitated  by 
alcohol  containing  hydrochloric  acid  and  ether.  The  final 
precipitate  G  weighed  16  grams.  With  water,  G  forms  an 
emulsion,  acid  in  reaction  and  cleared  by  the  addition  of 
alkali.  The  biuret  test  is  negative,  Millon  doubtful,  xantho- 
proteic,  Adamkiewicz,  a-naphthol,  and  orcin  tests  are  all 
positive,  the  carbohydrate  tests  being  very  marked.  After 
boiling  with  acid  there  is  copious  reduction  of  Fehling's 
solution.  G  was  tested  for  glycogen,  with  negative  results. 
One  gram  of  G  boiled  two  and  one-half  hours  with  72  c.c. 
of  2.5  per  cent,  hydrochloric  acid,  gave  38.63  per  cent,  of 
reducing  substance  calculated  as  xylose.  A  second  gram 
boiled  for  five  hours  yielded  43.77  per  cent. 

The  second  carbohydrate,  or,  more  properly,  the  second 
substance  giving  the  a-naphthol  test  in  bacterial  proteins, 
is  not  converted  into  a  reducing  sugar  on  being  boiled  with 
dilute  mineral  acid.  In  Wheeler's  work  with  sarcina  lutea 
it  remained  as  a  residue  after  extracting  the  cell  material 
with  10  per  cent,  potassium  hydroxide  at  120°.  This 
residue  responded  to  the  a-naphthol  test,  did  not  give  the 
protein  reactions,  and  did  contain  nitrogen.  The  same 
investigator  also  found  this  body  in  alkaline  extracts  of 
the  residue  left  after  extraction  with  dilute  acid.  In  her 
work  with  the  cell  substance  of  the  colon  bacillus,  Leach 
makes  the  following  statement  touching  this  body:  "The 
residue,  after  repeated  extraction  with  dilute  sulphuric 
acid  (from  1  to  5  per  cent.),  was  treated  with  2  to  4  per 
cent,  sodium  hydrate  either  upon  the  water-bath  or  over 
a  free  flame.  In  every  case  the  substance  went  into  solution 
readily,  leaving  only  a  slight  coating  on  the  filter.  The 
slight  residue  gave  no  protein  test,  contained  no  nitrogen, 
but  gave  test  for  carbohydrate.  In  one  case  it  was  removed 
from  the  filter,  and  the  organic  matter  approximately 
determined.  The  total  residue  was  about  0.4  gram,  equiva- 
lent to  0.8  per  cent,  of  the  cell  substance  used.  The  organic 


72  PROTEIN  POISONS 

matter  .was  only  0.15  gram,  equivalent  to  0.3  per  cent,  of 
the  original."  This  body  is  also  found  in  the  dilute  acid 
extracts  of  cellular  proteins,  as  is  shown  by  the  following 
additional  quotation  from  Leach:  "Some  earlier  investi- 
gations in  this  laboratory  upon  the  toxicity  of  the  colon 
germ  showed  the  desirability  of  studying  the  action  of 
dilute  acid  upon  the  cell  substance.  Accordingly,  samples 
were  treated  with  1  per  cent,  sulphuric  acid  under  varying 
conditions.  On  filtering,  a  light  brown  or  straw-colored 
fluid  was  obtained.  This  readily  reduced  nitric  acid  and 
gave  the  typical  xanthoproteic  color  on  the  addition  of 
ammonia.  In  no  case  was  there  more  than  a  slight  biuret 
test,  and  there  was  too  much  sulphate  present  for  a  satis- 
factory Millon  test.  The  a-naphthol  test  for  furfurol  was 
positive.  Alcohol  gave  a  voluminous  precipitate,  A,  which 
will  be  described  more  fully  under  another  heading.  The 
alcoholic  filtrate,  B,  was  neutralized  with  sodium  hydroxide, 
the  sodium  sulphate  filtered  out,  together  with  some  organic 
matter  mechanically  carried  down,  and  the  liquid  distilled 
under  diminished  pressure  at  30°  to  38°.  The  liquid  residue, 
C,  left  after  distillation,  turned  yellow  on  heating  with 
potassium  hydroxide,  but  gave  neither  the  biuret  nor 
Millon  test.  Again,  the  xanthoproteic  and  a-naphthol 
tests  were  positive,  but  it  failed  to  reduce  either  Fehling's 
or  Nylander's  solution  (after  boiling  with  dilute  mineral 
acid).  It  yielded  precipitates  with  ammonium  molybdate, 
phosphomolybdic  acid,  ammoniacal  silver  nitrate,  and 
picric  acid.  A  guinea-pig  was  injected  with  5  c.c.  of  C 
with  no  apparent  effect." 

We  are  inclined  to  attribute  the  sticky,  mucilaginous 
properties  of  both  acid  and  alkaline  extracts  of  bacterial 
cellular  substances  to  these  bodies  giving  the  furfurol 
reaction,  and  here  regarded  as  carbohydrates.  Further- 
more, we  are  of  the  opinion,  though  this  cannot  be  con- 
sidered as  conclusive,  that  the  one  yielding  a  reducing  sugar 
after  boiling  with  dilute  mineral  acid  exists  in  the  cellular 
molecule  as  a  constituent  of  the  nucleic  acid  group,  while 
the  other  is  a  part  of  the  protein  component. 


BACTERIAL  CELLULAR  SUBSTANCE       73 

The  reducing  carbohydrate  is  present  in  the  bacterial 
cellular  substance  examined  in  this  laboratory  in  a  minimum 
of  something  over  10  per  cent.;  in  the  colon  haptophor  in 
about  24  per  cent. ;  and  in  precipitate  G  from  the  haptophor 
in  about  44  per  cent.  So  far,  we  have  not  obtained  it  free 
from  phosphorus.  The  percentage  of  the  other  furfurol 
giving  body  we  have  no  means  of  determining  even  approxi- 
mately, although  we  can  safely  say  that  the  amount  is 
much  smaller.  The  one  yielding  a  reducing  sugar  probably 
exists  in  the  nucleic  acid  group  as  a  pentose. 

Nuclein  Bodies. — In  her  work  on  sarcina  lutea,  Wheeler1 
makes  the  following  statement:  "So  far  as  the  xanthin 
bases  are  concerned,  Nishimura2  found  0.17  per  cent,  of 
xanthin,  0.08  per  cent,  of  adenin,  and  0.14  per  cent,  of 
guanin  in  his  water  bacillus.  It  has  been  suggested  that  in 
Nishimura's  experiments  these  bases  might  have  come 
from  the  potato  upon  which  his  organism  was  grown,  but 
inasmuch  as  the  potato  contains  only  a  very  small  per- 
centage of  protein,  this  is  not  likely.  Lustig  and  Galleotti3 
report  guanin  obtained  from  the  pest  bacillus,  and  Galleotti4 
says  that  a  nucleoprotein  separated  from  the  bacillus 
ranicidus  yielded  xanthin  bases,  although  the  percentage 
of  nitrogen  was  low. 

"I  have  gone  through  the  process  of  testing  for  xanthin 
bases  four  times.  Three  times  the  acid  extracts  were 
carefully  precipitated  with  powdered  silver  nitrate  crystals 
until  a  drop  of  the  solution  gave  a  yellow  instead  of  a  white 
precipitate  with  barium  hydrate.  The  precipitate  was 
filtered  out,  washed,  dried,  and  then  worked  up  for  xanthin 
bases.  The  fourth  time  the  process  was  almost  the  same, 
the  difference  being  that  33^  per  cent,  acid  extract  had 
been  made.  This  was  first  almost  neutralized  with  barium 
hydrate,  the  barium  sulphate  filtered  out,  carefully  washed 
out  and  boiled  with  water,  and  then  the  slightly  acid 
extract  was  precipitated  with  silver  sulphate  instead  of 
silver  nitrate.  The  first  silver  nitrate  precipitate  was 

1  Trans.  Assoc.  Amer.  Phys.,  1902,  xxvii,  265. 

2  Loc.  cit.  3  Loc.  cit.  4  Loc.  cit. 


74  PROTEIN  POISONS 

investigated  according  to  the  method  given  by  Kruger  and 
Solomon;1  as  no  satisfactory  separation  was  thereby  ob- 
tained, the  last  three  precipitates  were  separated  according 
to  the  method  of  Kossel  and  his  pupils,  as  outlined  by 
Hammersten,2  for  the  separation  of  the  four  bases,  xanthin, 
hypoxanthin,  guanin,  and  adenin.  The  precipitate  was 
dissolved  as  completely  as  possible  in  boiling  nitric  acid 
(specific  gravity,  1.1),  a  little  urea  having  been  added 
to  prevent  nitrification,  filtered  hot,  concentrated  some- 
what, and  allowed  to  cool.  On  cooling,  only  a  slight 
residue  of  the  guanin-hypoxanthin-adenin  portion  separ- 
ated out.  On  decomposing  this  precipitate,  treating  with 
ammonia  and  evaporating,  the  amount  obtained  was  so 
small  that  it  was  not  possible  to  make  separation  of  the 
bases,  but  ammoniacal  solution  produced  a  comparatively 
heavy  flocculent  organic  precipitate.  The  nitric  acid  filtrate 
containing  the  xanthin  portion  was  precipitated  with 
ammonia.  A  heavy,  reddish-brown,  mucilaginous  precipi- 
tate came  down,  but  was  not  sufficient  in  quantity  or 
sufficiently  free  from  impurities  to  justify  an  ultimate 
analysis." 

Leach3  obtained  from  1  per  cent,  sulphuric  acid  extracts 
of  the  colon  cellular  substance  a  body  containing  7.33  per 
cent,  of  phosphorus.  It  gave  none  of  the  protein  color 
reactions  except  the  ubiquitous  xanthoproteic.  It  could 
hardly  be  anything  else  than  a  nucleic  acid.  However,  the 
percentage  of  nitrogen  was  only  8.98,  and  no  known  nucleic 
acid  contains  so  small  an  amount  of  nitrogen.  In  the  same 
extracts  she  obtained  indications  of  the  presence  of  two 
xanthin  bases,  xanthin  and  guanin.  The  evidence  of  the 
existence  of  nuclein  bodies  in  the  haptophor  of  the  colon 
cellular  substance  will  be  given  later. 

Diamino-acids. — Wheeler  failed  to  obtain  any  evidence 
of  the  presence  of  arginin  or  histidin  in  sarcina  lutea,  but 
in  each  of  five  attempts  she  secured  quite  convincing  proof 

1  Zeitsch.  f.  Physiol.  Chem.,  1898,  xxvi,  373. 

2  Physiol.  Chem.,  p.  120,  as  translated  by  Mendel. 

3  Jour.  Biol.  Chem.,  1906,  i,  463. 


BACTERIAL  CELLULAR  SUBSTANCE  75 

of  the  existence  of  lysin.  She  says:  "As  to  lysin,  I  have 
obtained  in  every  instance  a  yellow,  thick,  oily  body,  where 
lysin  picrate  should  be  formed,  which,  however,  could  not 
be  crystallized.  This  oily  body  was  shaken  up  with  ether 
to  remove  excess  of  picric  acid,  but  when  I  attempted  to 
purify  it  by  redissolving  it  in  alcohol,  it  was  no  longer 
completely  soluble,  inasmuch  as  a  part  of  it  hardened  into 
a  solid  which  seemed  somewhat  crystalline,  and  the  remainder 
was  precipitated  by  alcohol.  However,  it  was  found  to  be 
readily  soluble  in  water,  especially  in  hot  water,  and  although 
no  crystals  were  obtained,  on  concentration  of  the  aqueous 
solution,  the  same  heavy  oil  separated.  In  order  to  obtain 
the  hydrochloride,  if  possible,  the  oily  substance  was  treated 
with  hydrochloric  acid  in  a  little  water,  and  the  picric 
acid  shaken  out  with  ether.  From  the  solution  on  concen- 
tration an  imperfectly  crystalline  mass  was  obtained.  It 
may  be  that  I  have  not  lysin,  but  there  is  undoubtedly 
present  some  organic  body  which,  in  its  chemical  behavior 
at  least,  is  very  similar  to  lysin." 

In  searching  for  the  hexon  bases  in  the  cellular  substance, 
Leach  proceeded  as  follows :  "  The  cell  substance  was  stirred 
with  nine  times  its  weight  of  33.33  per  cent,  sulphuric 
acid,  allowed  to  stand  overnight,  then  heated  in  an  evapor- 
ating dish  on  the  water-bath.  When  all  danger  of  frothing 
was  over,  the  mixture  was  transferred  to  a  flask  fitted  with  a 
reflux  condenser,  and  boiled  on  a  sand-bath  for  eight  hours 
one  day  and  six  the  next.  After  cooling  and  filtering,  some 
water  was  added  to  the  filtrate,  and  it  was  neutralized  by 
the  addition  of  the  calculated  amount  of  barium  hydrate. 
When  the  barium  sulphate  had  completely  settled,  the 
supernatant  liquid  was  siphoned  off,  the  precipitate  stirred 
up  with  boiling  water,  heated  to  boiling,  settled  overnight, 
and  again  siphoned.  This  was  repeated  until  the  wash 
water  was  nearly  colorless.  The  extract  and  wash  water 
were  united,  acidified  with  acetic  acid,1  concentrated  on 


1  If  there  is  a  large  excess  of  barium  present,  it  is  well  to  remove  it  by 
carbon  dioxide. 


76  PROTEIN  POISONS 

the  water-bath,  cooled,  and  filtered  to  remove  any  tyrosin 
and  leucin  that  may  crystallize  out.  The  filtrate  was 
diluted  to  about  one  and  one-half  liters  for  each  100  grams 
of  cell  substance,  made  decidedly  acid  with  nitric  acid, 
and  20  per  cent,  silver  nitrate  added  as  long  as  it  gave  a 
precipitate.  It  was  left  overnight  to  settle,  and  the  silver 
precipitate  of  xanthin  bases  filtered  out.  To  this  filtrate 
excess  of  silver  nitrate  and  barium  hydrate  were  added  to 
remove  arginin  and  histidin.  After  their  removal,  silver 
and  barium  were  precipitated  by  hydrochloric  and  sulphuric 
acids,  these  inorganic  precipitates  boiled  out  with  water 
several  times,  the  filtrate  and  wash  water  united,  and  con- 
centrated. The  solution,  which  should  contain  some  5 
per  cent,  of  acid,  was  treated  with  a  50  per  cent,  solution 
of  phosphotungstic  acid  as  long  as  it  gave  an  immediate 
precipitate.  The  precipitate  was  rubbed  up  with  5  per 
cent,  sulphuric  acid,  carefully  washed  with  the  same 
solution,  and  filtered  with  suction.  The  heavy  white 
precipitate  was  again  rubbed  up  with  water,  hot  saturated 
solutions  of  barium  hydrate  added,  until  the  mixture  was 
no  longer  acid,  settled  overnight,  and  the  supernatant 
liquid  siphoned  off.  The  precipitate,  consisting  of  barium 
phosphate,  tungstate,  etc.,  was  washed  several  times  with 
hot  barium  hydrate  solution,  decanted,  and  finally  filtered 
by  suction.  The  filtrate  and  wash  water  were  united,  and 
barium  was  removed  as  carefully  as  possible,  first  by  running 
in  carbon  dioxide,  and  then  by  adding  ammonium  carbonate 
to  the  solution.  This  precipitate,  like  all  the  other  inorganic 
ones,  was  boiled  out  several  times  with  water,  and  the 
washings  added  to  the  original  filtrate.  The  resulting 
liquid  was  concentrated  nearly  to  dryness  on  the  water- 
bath,  the  residue  taken  up  with  water,  filtered  to  remove 
barium  carbonate,  and  again  concentrated  to  a  thick 
syrup. 

"The  alkaline  syrup  was  vigorously  stirred  with  alcohol, 
and  then  with  an  alcoholic  solution  of  picric  acid.  Some- 
times a  crystalline  precipitate  came  down  at  once,  sometimes 
there  was  a  viscous  mass  like  molasses  candy,  which  became 


BACTERIAL  CELLULAR  SUBSTANCE  77 

granular  or  crystalline  after  long  kneading  and  stirring,  as 
Fischer  and  Weigert  suggest.  When  picric  acid  would  no 
longer  give  a  precipitate  even  on  standing,  the  crystals 
were  filtered  out  by  suction,  washed  with  alcohol,  and 
dried  on  a  porous  plate.  On  concentration  the  alcoholic 
mother  liquid  became  gummy  and  viscous,  but  no  more 
crystals  were  obtained.  The  crude  picrate  was  recrystallized 
from  hot  water  several  times.  On  dissolving  there  was 
much  sediment  which  mainly  filtered  out,  but  on  concen- 
tration more  appeared  upon  the  sides  of  the  vessel.  The 
loss  of  substance  by  the  first  crystallization  was  very  large; 
as  it  became  pure,  however,  it  crystallized  like  an  inorganic 
salt.  All  mother  liquors  were  treated  with  hydrochloric 
acid  to  remove  picric  acid,  reprecipitated  with  phospho- 
tungstic,  the  precipitate  worked  up  as  before,  and  a  further 
crop  of  crystals  obtained.  The  crystals  are  slender,  yellow, 
silky,  felted  needles  or  prisms.  On  heating  in  a  melting- 
point  tube  the  substance  begins  to  change  color  at  216°, 
and  is  very  dark  at  230°.  Heated  side  by  side  with  lysin 
picrate  from  fibrin  and  from  gelatin,  they  agree  within  a 
degree.  The  authorities  all  agree  that  lysin  picrate  turns 
black  at  230°  to  232°,  while  Kutscher  and  Lohmann  also 
say  that  it  begins  to  change  color  at  215°. 

"To  change  the  picrate  into  the  chloride,  2  grams  were 
dissolved  in  33  c.c.  of  hot  water,  5  c.c.  of  concentrated 
hydrochloric  acid  added,  cooled,  the  picric  acid  filtered  out 
and  washed  with  water  containing  hydrochloric  acid.  The 
filtrates  were  shaken  out  with  ether  as  long  as  there  was 
any  yellow  color.  The  solution  should  be  colorless  or 
nearly  so;  if  it  is  not,  it  can  be  decolorized  by  treatment 
with  animal  charcoal.  The  solution  was  evaporated  nearly 
to  dryness,  first  on  the  water-bath,  and  finally  in  a  desic- 
cator. When  down  to  a  thick  syrup,  stirring  gave  crystals. 
These  were  recrystallized  out  of  hot  water  containing 
hydrochloric  acid,  giving  long  prisms,  which  melt  at  192°, 
again  agreeing  with  the  corresponding  salt  from  gelatin 
and  fibrin.  Henze  says  that  lysin  chloride  becomes  soft 
at  193°  and  melts  at  195°;  Lawrow  says  that  it  has  no  sharp 


78  PROTEIN  POISONS 

melting-point,  but  begins  to  melt  at  194°  to  195°.  Hender- 
son collected  samples  melting  from  190°  to  200°,  prepared 
by  different  individuals  from  widely  different  sources. 
By  careful  purification  he  obtained  from  each  sample  a 
product  melting  at  192°  or  193°.  Thus  it  would  appear 
that  the  apparent  discrepancy  in  the  melting-point  is  due 
to  impurities.  Reactions,  crystalline  form,  properties,  and 
melting  (or  decomposing)  show  that  the  picrate  and  chloride 
from  the  germ  are  identical  with  lysin  picrate  and  chloride 
from  gelatin  and  from  fibrin.  Thus  the  presence  of  one  of 
the  hexon  bases  in  the  bacterial  cell  has  been  demonstrated, 
and  another  point  of  resemblance  between  bacterial  and 
other  proteins  has  been  established." 

Mono-amino-acids. — The  phosphotungstic  acid  filtrate  ob- 
tained by  Leach  in  her  work  on  the  hexon  bases  was  turned 
over  to  Wheeler,  who  has  made  the  following  report  on  it: 
"From  the  solution  phosphotungstic  acid  was  removed 
with  barium  hydrate  and  carbon  dioxide  used  to  remove 
excess  of  barium.  By  concentration  and  crystallization 
bodies  were  obtained  resembling  tyrosin  and  leucin  under 
the  microscope.  These  were  purified  by  repeated  recrystal- 
lization  from  water  or  from  ammoniacal  water,  the  tyrosin 
being  so  much  less  soluble  than  the  leucin  that  they  could 
be  separated  by  difference  of  solubility.  It  was  necessary 
to  boil  the  leucin  fraction  with  animal  charcoal  to  remove 
the  coloring  matter.  The  tyrosin  formed  the  characteristic, 
colorless,  silky  needles,  many  grouped  in  the  characteristic 
sheaves.  As  it  became  more  pure  the  needles  were  longer 
and  longer,  and  grouped  in  the  sheaves  less  positively. 
After  many  purifications  the  crystals  melted  at  a  constant 
temperature,  though  this  was  difficult  to  determine,  since 
tyrosin  melts  with  decomposition.  The  melting-point 
maintained  after  each  of  two  or  three  recrystallizations 
was  288°,  unconnected.  The  correction  was  8.13°,  which 
made  the  corrected  point  296.13°.  A  Kahlbaum  prepara- 
tion of  tyrosin  in  the  laboratory  melted  within  a  degree  of 
the  same  point,  and  agreed  in  the  chemical  tests,  to  be 
mentioned  presently.  Richter  gives  the  melting-point  of 


BACTERIAL  CELLULAR  SUBSTANCE  79 

tyrosin  as  235°;  Cohn  as  295°;  while  Fischer  says  that  with 
rapid  heating  the  corrected  point  is  314°  to  318°. 

"The  tyrosin  obtained  gives  a  Hofmann  test  with 
Millon's  reagent.  It  gives  Scherer's  test  with  nitric  acid 
and  sodium  hydrate  on  platinum  foil,  and  also  a  beautiful 
Piria  test  with  sulphuric  acid,  then  barium  carbonate  and 
ferric  chloride,  which  test  is  characteristic  for  tyrosin. 

"The  leucin  crystallized  in  the  characteristic  knobs  or 
balls.  As  it  became  purer  it  crystallized  more  and  more  in 
shining,  white,  very  thin  plates,  sometimes  in  radial  groups, 
sometimes  not.  The  crystals  were  finally  obtained  with  a 
practically  constant  melting-point,  262°  to  263°  or  corrected, 
268.5°  to  269.6°.  The  pure  laboratory  leucin  (Kahlbaum) 
melted  at  the  same  point.  Schwanert,  Hammersten,  and 
others  give  the  melting-point  for  active  leucin  as  170°; 
that  for  the  inactive  form  is  given  as  270°.  Fischer  says 
the  melting-point  is  293°  to  295°  (corrected)  if  heated 
quickly  in  a  closed  tube.  Cohn  gives  275°  to  276°.  The 
leucin  obtained  melted  with  darkening  and  decomposition. 
With  careful  heating  in  an  open  tube  it  sublimed  with 
the  characteristic  white,  woolly  deposit.  It  also  responded 
to  Scherer's  test  on  platinum  foil  with  nitric  acid  and 
sodium  hydrate,  which  test  Hammersten  says  is  charac- 
teristic for  leucin." 

Agnew1  has  made  the  following  contribution  to  the 
mono-amino-acid  content  of  the  cellular  proteins  of  the 
colon  and  tubercle  bacilli. 

The  material  used  in  this  research  consisted  of  the  cellular 
substance  of  the  bacteria.  Growths  from  massive  cultures 
were  placed  in  large  Soxhlets  and  extracted  for  three  days 
with  absolute  alcohol  and  for  the  same  time  with  ether. 
The  protein  bacterial  substance  thus  freed  from  everything 
soluble  in  alcohol  and  ether  was  ground  into  a  powder  and 
passed  through  a  fine-meshed  sieve.  For  the  preparation 
of  the  amino-acids  Fischer's  method  slightly  modified  was 
employed.  The  cellular  protein  was  boiled  with  three  and 

1  Unpublished  research. 


80  PROTEIN  POISONS 

one-half  times  its  weight  of  strong  hydrochloric  acid  under  a 
reflux  condenser  until  it  failed  to  respond  to  the  biuret 
test.  The  humus  substance  was  filtered  out,  repeatedly 
extracted  with  hydrochloric  acid,  and  the  concentrated 
extracts  added  to  the  filtrate  from  the  humus.  From  this 
filtrate  glutamic  acid  was  isolated  as  a  hydrochloride  by 
saturation  with  hydrochloric  acid  gas  and  standing  for  some 
days  in  the  ice-box.  The  deposited  glutamic  acid  was 
collected  on  a  filter  with  the  aid  of  a  pump.  The  filtrate 
was  concentrated  in  vacua  to  a  syrup,  diluted  with  an  equal 
volume  of  alcohol,  and  esterified  by  saturating  with  hydro- 
chloric acid  gas,  the  solution  being  warmed  to  complete 
the  esterification.  The  alcohol  was  distilled  off  at  a  tem- 
perature under  40°,  more  alcohol  added,  and  removed  by 
distillation,  this  being  repeated  three  times.  At  this  point 
glycocoll,  as  the  hydrochloride  of  the  ethyl  ester,  crystal- 
lized out  on  account  of  difficult  solubility  in  alcohol.  For 
convenience,  the  thick  syrup  containing  the  esters  of  the 
hydrochlorides  of  the  amino-acids  was  divided  into  portions 
and  each  treated  separately  but  in  the  same  way.  The 
syrup  was  diluted  with  half  its  volume  of  water,  cooled  in  a 
freezing  mixture,  and  made  slightly  alkaline  with  sodium 
hydrate.  This  alkaline  solution  was  extracted  with  ether, 
the  liquid  being  kept  alkaline  by  the  addition  of  a  few  drops 
of  a  saturated  solution  of  sodium  hydrate.  The  liquid  was 
then  converted  into  a  paste  by  the  addition  of  solid  potas- 
sium carbonate  and  repeatedly  shaken  with  ether.  The 
combined  ethereal  extract  was  dried  by  shaking  with  solid 
potassium  carbonate  for  a  short  time,  and  by  being  left  for 
twelve  hours  over  fused  sodium  sulphate.  The  ether  was  then 
distilled  off  and  the  free  esters  subjected  to  fractional  dis- 
tillation in  vacua.  The  highest  vacuum  that  I  was  able  to 
obtain  varied  from  20  to  30  mm.  The  pasty  mass  left  after 
the  distillation  of  the  esters,  according  to  Abderhalden's 
recommendation,  was  made  acid  with  hydrochloric  acid, 
concentrated,  and  the  inorganic  salts  allowed  to  crystallize. 
The  filtrate  freed  from  these  crystals  was  again  esterified. 
furnishing  a  further  but  smaller  amount  of  esters. 


BACTERIAL  CELLULAR  SUBSTANCE       81 

Hydrolysis  of  the  Cellular  Protein  of  Bacillus  Coli  Com- 
munis. — Three  hundred  and  fifty  grams  of  this  cellular 
protein  was  thoroughly  mixed  with  1100  c.c.  of  hydrochloric 
acid,  of  specific  gravity  1.19,  and  allowed  to  stand  over- 
night. The  next  morning  it  was  found  to  be  a  frothy  mix- 
ture of  deep  purple  color  (Liebermann's  reaction).  It  was 
placed  in  a  2-liter  flask  connected  with  a  long  reflux  con- 
denser, and  boiled  on  the  sand-bath  for  six  hours,  when  it 
no  longer  gave  the  biuret  reaction.  It  was  then  filtered 
through  heavy  paper,  giving  a  clear  brown  filtrate  and 
leaving  a  heavy  deposit  of  humus  on  the  paper.  The  latter 
was  repeatedly  extracted  with  dilute  hydrochloric  acid, 
the  extracts  concentrated,  and  added  to  the  filtrate  The 
humus,  air-dried,  weighed  77  grams,  making  22  per  cent, 
of  the  cellular  protein. 

The  filtrate  was  concentrated  in  vacua  to  half  its  volume, 
placed  in  a  freezing  mixture,  saturated  with  hydrochloric 
acid  gas  and  left  in  the  ice-box  for  two  days.  By  this  time 
the  hydrochloride  of  glutamic  acid  had  crystallized  out. 
An  equal  volume  of  ice-cold  alcohol,  previously  saturated 
with  hydrochloric  acid  gas,  was  added,  and  the  glutamic 
acid  collected,  dried  over  solid  sodium  hydrate,  and 
weighed.  It  yielded  10.5  grams  of  3  per  cent,  of  the  cellular 
protein. 

The  filtrate  from  the  glutamic  acid  was  esterified  with 
hydrochloric  acid  gas,  followed  by  gentle  heat.  The  alcohol 
was  distilled  off  in  vacuo,  an  equal  volume  added,  again 
saturated  with  the  gaseous  acid,  heated,  and  removed  by 
distillation,  this  being  repeated  several  times.  The  fluid 
was  left  in  the  ice-box  for  forty-eight  hours,  when  crystals 
were  deposited.  These  were  purified  by  reprecipitating 
them  from  alcoholic  solution  with  hydrochloric  acid  gas. 
The  melting-point  was  144°  and  the  crystals  were  identified 
as  the  ethyl  ester  of  glycocoll.  The  yield  of  glycocoll  was 
1  gram. 

The  filtrate  from  the  glycocoll   containing  the  hydro- 
chlorides  of  the  esters  was  treated  as  already  stated  and 
fractioned  with  the  following  yields: 
6 


82  PROTEIN  POISONS 

Bath  temperature.  Yield. 

Fraction  I 40°  to    60°  16.5  grams 

Fraction  II 60°  to    80°  8.5  grams 

Fraction  III 80°  to  100°  22.5  grams 

Fraction  IV 100°  to  130°  20.0  grams 

Fraction  V 130°  to  160°  16.5  grams 

The  dark  red  residue  left  in  the  distillation  flask  was 
dissolved  in  hot  alcohol,  decolorized  with  animal  charcoal, 
filtered,  and  from  the  filtrate  0.5  gram  of  leucinimid  was 
obtained. 

Fraction  I. — This  was  saponified  by  boiling  with  five 
times  its  weight  of  water  for  five  hours  with  a  reflux  con- 
denser. On  concentration  two  crops  of  crystals  were 
obtained.  The  first  was  dissolved  in  the  smallest  possible 
amount  of  hot  water,  then  treated  with  an  equal  volume 
of  alcohol,  and  left  in  the  ice-box  for  twenty-four  hours. 
The  white  crystalline  mass,  tasting  sweet  and  closely 
resembling  alanin,  weighed  1.75  grams. 

The  second  crop  was  obtained  by  evaporation  to  dryness, 
and  weighed  1  gram.  It  was  dissolved  in  3  c.c.  of  hot 
water;  this  was  poured  into  an  equal  volume  of  hot  absolute 
alcohol,  and  set  in  the  ice-box  for  twenty-four  hours.  White 
needle-like  crystals,  sweet  to  the  taste  and  arranged  in 
bundles,  formed.  These  were  recrystallized,  washed  with 
absolute  alcohol,  dried,  and  weighed.  The  yield  was  0.5 
gram.  These  crystals  sublimed  on  heating  and  were  prob- 
ably valin,  although  it  will  be  seen  that  neither  of  these 
bodies  has  been  positively  identified. 

Fraction  II. — This  on  being  saponified  by  boiling  with 
five  times  its  weight  of  water  for  five  hours  under  a  reflux 
condenser  and  concentrated,  yielded  a  crystalline  mass 
that  contained  no  prolin,  since  no  part  of  it  was  soluble  in 
hot  absolute  alcohol.  The  crystalline  mass. was  dissolved 
in  700  c.c.  of  water  and  boiled  for  an  hour  with  an  excess  of 
freshly  prepared  copper  oxide.  On  concentration  of  the 
blue  filtrate,  a  difficultly  soluble  salt  was  obtained.  This 
was  crystallized  and  the  percentage  of  copper  in  it  found 
to  be  20.35,  thus  identifying  it  as  the  double  salt  of  copper 
with  leucin  and  valin. 


BACTERIAL  CELLULAR  SUBSTANCE  83 

Calculated  per  cent,  of  Cu  in  Cu-leucin  .  .  .  .  19 . 66 
Calculated  per  cent,  of  Cu  in  Cu-valin  .  .  .  .25.52 
Calculated  per  cent,  of  Cu  in  Cu-leucin-valin  .  .  20.48 

The  more  soluble  copper  salt  was  freed  from  its  copper 
with  hydrogen  sulphide.  The  filtrate  furnished,  on  treating 
it  with  80  per  cent,  alcohol  and  allowing  it  to  stand  in  the 
ice-box,  0.5  gram  of  alanin. 

Fraction  III. — After  saponification  by  boiling  for  five 
hours  under  reflux  condenser  with  five  times  its  weight  of 
water,  this  fraction  yielded  7.3  grams  of  solid.  The  absence 
of  prolin  was  demonstrated  by  the  fact  that  hot  absolute 
alcohol  dissolved  nothing.  The  solution  was  boiled  with 
copper  oxide  which  took  up  all  the  substance.  The  per- 
centage of  copper  was  found  to  be  20.38,  thus  showing  the 
compound  to  be  the  double  salt  of  copper  with  leucin  and 
valin. 

Fractions  IV  and  V. — When  these  were  mixed  with 
water  a  brownish  oil  separated.  This  was  filtered  off  and 
saponified  by  heating  for  several  hours  on  the  water-bath 
with  an  excess  of  baryta.  When  the  barium  had  been 
removed  there  was  obtained  0.5  gram  of  phenylalanin.  It 
was  found  difficult  to  purify  this,  but  finally  we  did  so,  and 
found  the  melting-point  to  be  262.5°.  The  filtrate  from  the 
oily  ester  of  phenylalanin  was  also  saponified  by  heating 
on  the  water-bath  with  an  excess  of  baryta.  The  absence 
of  aspartic  acid  was  shown  by  failure  to  obtain  an  insoluble 
barium  asparaginate.  When  the  barium  was  removed  a 
small  amount  of  glutamic  acid  was  found  in  the  filtrate. 

The  pasty  mass  left  after  the  extraction  of  the  esters 
was  neutralized  with  hydrochloric  acid  and  prepared  for  a 
second  esterification,  but  owing  to  an  accident  this  was 
not  completed.  No  ultimate  analyses  were  made  in  this 
or  the  subsequent  hydrolyses,  and  we  have  relied  for  the 
recognition  of  the  amino-acids  on  (1)  the  boiling-point  of 
the  esters,  (2)  crystalline  form,  (3)  melting-point,  and  (4)  the 
percentage  of  copper  in  the  copper  compounds. 

In  this  hydrolysis  we  have  accounted  for  only  a  little 
more  than  10  per  cent,  of  the  nitrogen,  distributed  as 


84  PROTEIN  POISONS 

follows:  glutamic  acid,  3  per  cent.;  glycocoll,  0.33  per  cent.; 
alanin,  1  per  cent.;  valin,  1.6  per  cent.;  leucin,  2  per  cent.; 
and  phenylalanin,  0.2  per  cent. 

The  cellular  protein  of  the  bacillus  coli  communis  with 
which  we  did  this  work  contained  13.74  per  cent,  of  moisture 
and  7.2  per  cent,  of  ash,  or  8.38  per  cent,  of  ash  in  the 
moisture-free  substance. 

Hydrolysis  of  the  Cellular  Protein  of  the  Bacillus  Tuber- 
culosis.— Five  hundred  grams  of  this  substance  was  hydro- 
lyzed  after  the  manner  already  given.  The  air-dried  humus 
from  this  substance  weighed  120  grams  or  24  per  cent,  of 
the  cellular  substance,  or  27  per  cent,  of  the  moisture-free 
substance.  The  humus  was  found  to  contain  14.34  per  cent, 
of  moisture,  0.15  per  cent,  of  ash,  and  1.69  per  cent,  of 
nitrogen. 

Before  hydrolyzing  this  substance  samples  were  taken, 
and  the  following  determinations  made. 

Percentage  of  moisture 12.07 

Percentage  of  ash .       7 . 08 

Percentage  of  ash  in  moisture-free  substance 8.05 

Nitrogen  was  found  distributed  as  follows: 

Percentage  of  nitrogen  in  cell  substance 9 . 27 

Percentage  of  nitrogen  in  moisture-free  cell  substance      .      .      .      .  10.55 

Percentage  of  nitrogen  in  ash-free  cell  substance 9.98 

Percentage  of  nitrogen  in  ash-  and  moisture-free  cell  substance       .  11 .47 

Total  amount  of  nitrogen  in  500  grams  of  cell  substance  .  46 . 3500  grams 

Total  amount  of  nitrogen  in  the  hydrolyzed  fluid    .      .      .  36.3075  grams 

Total  amount  of  nitrogen  in  the  extract  of  humus        .      .  4 . 0530  grams 

Total  amount  of  nitrogen  in  the  extracted  humus  ...  2 . 0388  grams 

Total  amount  of  nitrogen  in  500  grams  of  cell  substance   .  46.3500  grams 

Total  amount  of  nitrogen  in  the  hydrolytic  product     .      .  42 . 3993  grams 

Total  amount  of  nitrogen  lost  in  hydrolysis        ....  3 . 9507  grams 

Amount  of  nitrogen  in  hydrolyzed  fluid 36 . 3075  grams 

Amount  of  nitrogen  in  extract  of  humus 4 . 0530  grams 

Amount  of  nitrogen  in  fluid  to  be  esterified        ....  40.3605  grams 

It  is  thus  seen  that  there  is  in  the  fluid  to  be  esterified, 
40.3605  grams  of  nitrogen,  and  it  is  supposed  that  this 
exists  in  the  form  of  mono-  and  diamino-acids. 


BACTERIAL  CELLULAR  SUBSTANCE       85 

Glutamic  acid  was  separated  as  the  hydrochloride. 
Chlorides  of  ammonium  and  sodium  were  present  in  large 
amount,  but  were  easily  separated  on  account  of  their 
greater  solubility  in  water.  We  obtained  1  gram  of  glutamic 
acid,  equivalent  to  0.2  per  cent,  of  the  cellular  substance 
or  0.23  per  cent,  of  the  moisture-free  substance. 

The  filtrate  from  the  glutamic  acid  on  being  esterified,  the 
esters  extracted,  dried,  and  distilled,  yielded  the  following: 

Temperature          Temperature          Weight  of 


of  bath. 

of  vapor. 

distillate. 

Fraction  I 

50°  to    85° 

35°  to    60° 

14  grams 

Fraction  II    . 

85°  to  100° 

60°  to    85° 

33  grams 

Fraction  III 

100°  to  140° 

85°  to  105° 

34  grams 

Fraction  IV  . 

140°  to  180° 

105°  to  130° 

12  grams 

Fraction  V    . 

180°  to  210° 

130°  to  160° 

12  grams 

The  pasty  mass  left  after  extraction  of  the  esters  was 
acidified  with  hydrochloric  acid,  evaporated,  the  salt 
filtered  out  from  time  to  time,  and  when  brought  to  a  thick 
syrup  it  was  diluted  with  an  equal  volume  of  absolute 
alcohol  and  the  esterification  repeated.  However,  the  yield 
from  this  esterification  was  exceedingly  small,  not  more 
than  a  few  drops  for  each  fraction. 

From  the  residue  left  after  distillation,  we  obtained  2 
grams  of  leucinimide,  equivalent  to  0.4  per  cent,  of  the 
cellular  protein  or  0.45  per  cent,  of  the  moisture-free  sub- 
stance. 

Fraction  I. — This  was  saponified  by  being  boiled  with 
five  times  its  weight  of  water  for  five  hours  under  a  reflux 
condenser.  It  was  evaporated  to  dryness  and  the  white 
crystalline  mass  was  dissolved  in  25  c.c.  of  hot  water, 
treated  with  an  equal  volume  of  hot  absolute  alcohol  and 
left  in  the  ice-box  for  two  days  when  a  white  crystalline 
mass  separated.  This  was  purified  and  yielded  7  grams  of 
alanin,  equivalent  to  1.4  per  cent,  of  the  cellular  substance 
or  1.57  per  cent,  of  the  moisture-free  substance. 

Fraction  II. — From  this  there  was  obtained  by  the  method 
already  described  a  copper-leucin-valin  compound  repre- 


86  PROTEIN  POISONS 

senting  3.7  grams  of  leucin  and  12.4  grams  of  valin.  Prolin 
was  not  present. 

Fraction  III. — From  this  there  was  secured  by  the 
formation  of  the  copper  salt  5.4  grams  of  leucin  and  10.6 
grams  of  valin. 

Fractions  II  and  III  gave  a  combined  yield  of  9.1  grams  of 
leucin  and  23  grams  of  valin,  equivalent  to  1.82  per  cent, 
of  leucin  and  4.6  per  cent,  of  valin  in  the  cellular  substance 
or  2.04  per  cent,  and  5.17  per  cent,  respectively  in  the 
moisture-free  substance. 

Fractions  IV  and  V.— Each  was  shaken  with  three  times 
its  weight  of  cold  water  and  filtered  through  a  damp  paper, 
leaving  a  brown  oil.  The  oil,  after  being  washed  twice 
with  cold  water,  was  saponified  with  excess  of  baryta. 
From  this  there  was  obtained  2.5  grams  of  crude  phenyl- 
alanin.  This  was  purified  and  the  melting-point  found  to 
be  263°.  The  yield  of  phenylalanin  amounted  to  0.5  per 
cent,  of  the  cellular  substance  or  0.56  per  cent,  of  the 
moisture-free  substance.  From  the  filtrate  from  the  oil  a 
few  crystals  which  were  probably  glutamic  acid,  were 
obtained,  but  the  amount  was  too  small  for  identification. 

From  500  grams  of  the  cellular  protein  of  the  bacillus 
tuberculosis  the  following  substances  in  the  amounts  and 
percentages  given  were  obtained. 

Per  cent,  of  Per  cent,  of  dry 

Substance.                Amount.  cell  substance,  cell  substance. 

Humus  .      .      .      .  120.0  grams  24.00  27.00 

Glutamic  acid  .      .          1.0  gram  0.20  0.225 

Alanin    .      .      .      .          7.0  grams  1 . 40  1 . 57 

Leucin    .      .      .      .          9.1  grams  1.82  2.04 

Valin      ....  23.0  grams  4.60  5.17 

Phenylalanin     .      .          2.5  grams  0.50  0.56 

Leucinimide      .      .          2.0  grams  0.40  0.45 

It  will  be  seen  that  of  the  total  of  40.3605  grams  of  nitrogen 
in  the  fluid  esterified  there  has  been  recovered  in  the  form 
of  mono-ammo-acids  only  5.2  grams  of  nitrogen  or  12.88 
per  cent.  However,  under  the  best  conditions  one  cannot 
hope  to  obtain  more  than  a  part  of  the  mono-amino-acids 
present  and  the  diamino-acids  probably  take  up  a  consider- 
able part  of  the  nitrogen. 


BACTERIAL  CELLULAR  SUBSTANCE       87 

Hydrolysis  of  the  Non-poisonous  Portion  of  the  Cellular 
Protein  of  the  Bacillus  Tuberculosis. — Five  hundred  grams 
of  this  non-poisonous  bacterial  split  product,  known  in  this 
laboratory  as  "residue"  or  haptophor,  was  hydrolyzed. 
The  humus  was  found  to  constitute  29  per  cent,  of  the  air- 
dried,  or  40  per  cent,  of  the  moisture-free  substance,  the 
percentage  of  moisture  being  27.5. 

The  distribution  of  nitrogen  was  studied  with  the  fol- 
lowing results: 

Percentage  of  nitrogen  in  residue 4 . 59 

Percentage  of  nitrogen  in  moisture-free  residue 6 . 34 

Percentage  of  nitrogen  in  ash-free  residue 5 . 54 

Percentage  of  nitrogen  in  ash-  and  moisture-free  residue  .  .  .  .  8 . 29 
Amount  of  nitrogen  in  the  hydrolyzed  fluid  ....  18 . 980  grams 
Amount  of  nitrogen  in  the  extract  of  humus  ....  1 . 888  grams 
Amount  of  nitrogen  in  the  extracted  humus  ....  2 . 030  grams 


Total 22.898  grams 

Total  amount  of  nitrogen  in  the  500  grams  of  residue        .        22.990  grams 
Total  amount  of  nitrogen  in  the  products  of  hydrolysis     .        22 . 898  grams 


Loss  during  hydrolysis 0 . 092  gram 

Amount  of  nitrogen  in  the  hydrolyzed  fluid        ....        18 . 980  grams 
Amount  of  nitrogen  in  the  extract  of  humus      ....          1 . 888  grams 


Amount  of  nitrogen  in  the  fluid  to  be  esterified       ...        20 . 868  grams 

No  glutamic  acid  could  be  obtained  from  the  residue. 
The  result  of  the  fractional  distillation  is  shown  as  follows : 

Temperature  of  bath.  Yield. 

Fraction  I 40°  to    60°  21. 5  grams 

Fraction  II 60°  to    80°  5.0  grams 

Fraction  III 80°  to  100°  8.0  grams 

Fraction  IV 100°  to  130°  9.0  grams 

Fraction  V 130°  to  170°  10.0  grams 

By  the  hydrolysis  of  500  grams  of  tubercle  residue  the 
following  substances  were  obtained  in  the  amounts  and 
percentages  given: 

Per  cent,  of  Per  cent,  of 

Substance.  Amount.  residue.  dry  residue. 

Humus     ....  145. 00  grams  29.00  40.00 

Alanin      ....  4.00  grams                    0.80  1.10 

Leucin      .      .      .      .  6 . 40  grams                    1 . 28  1 . 76 

Valin        ....  2. 30  grams                   0.46  0.63 

Phenylalanin       .      .  1.70  grams                   0.34  0.46 


88  PROTEIN  POISONS 

It  has  been  possible  in  this  hydrolysis  to  recover  only 
8.19  per  cent,  of  the  nitrogen  in  the  fluid  esterified.  Doubt- 
less we  should  have  obtained  a  greater  percentage  had 
we  been  able  to  secure  a  higher  vacuum.  Fischer  does  his 
work  with  a  vacuum  of  1  to  2  mm.  of  mercury,  while  the 
best  wre  could  obtain  with  the  facilities  at  our  command 
varied  from  20  to  30  mm. 

While  the  results  of  this  work  are  not  so  satisfactory  as 
one  might  wish  it  does  indicate  that  the  proteins  of  the 
two  bacilli  studied  are  different  in  their  chemical  compo- 
sition. This  is  shown  by  the  distribution  of  the  amino- 
acids  as  is  indicated  by  the  following  figures: 

Colon.  Tuberculosis. 

Glutamic  acid       .      .      .  3.00  per  cent.  0.20  per  cent. 

Glycocoll 0.33  per  cent.  0.00  per  cent. 

Alanin 1.00  per  cent.  1.40  per  cent. 

Valin 1 . 60  per  cent.  4 . 60  per  cent. 

Leucin 2 . 00  per  cent.  1 . 82  per  cent. 

Phenylalanin  .      .      .      .  0.20  per  cent.  0.50  per  cent. 

Wheeler  has  reported  as  follows  upon  the  mono-amino 
acids  of  the  toxophor  group: 

The  poisons  selected  for  this  work  were  those  from 
tuberculosis,  typhoid,  and  colon  germ  substances,  an,d  for 
comparison  that  from  egg  albumin.  For  the  tuberculosis, 
the  poison  from  900  grams  germ  substance  was  used — 296 
grams;  for  the  typhoid,  100  grams  of  the  poison;  for  the 
colon,  that  from  300  grams  of  germ  substance,  estimated 
as  61.5  grams;  while  for  the  albumin,  that  from  200  grams 
of  the  protein  was  employed,  yielding,  it  was  estimated, 
93  grams  of  crude  poison.  These  poisons  were  hydrolyzed 
by  boiling  under  a  reflux  condenser  for  fourteen  hours 
with  concentrated  hydrochloric  acid.  The  nitrogen  of 
each  extract  was  determined  by  the  Kjeldahl  method, 
using  an  aliquot  part  of  each  as  a  sample,  and  giving  the 
following  results : 

NITROGEN  OF  ACID  EXTRACT  OF  POISONS 
Source  of  poison  Per  cent,  of  N. 

Tuberculosis 9.895 

Typhoid 10.380 

Colon       .      .      . 10.185 

Egg  albumen 11.477 


BACTERIAL  CELLULAR  SUBSTANCE 


89 


From  this  point  Fischer's  ester  method  for  obtaining  the 
individual  mono-amino-acids  was  carried  out. 

This  method  is  so  well  known  that  it  is  not  necessary  to 
outline  it  here  other  than  to  say  that  after  the  amino-acids 
have  been  produced  by  cleavage  of  the  protein  with  concen- 
trated acid,  the  hydrochloride  of  their  ethyl  esters  is  formed, 
and  later  the  free  esters  are  separated  by  distillation  in  the 
highest  possible  vacuum.  These  are  then  saponified  and 
the  amino-acids  crystallized  and  purified. 

The  efficiency  of  this  method  depending  in  large  measure 
upon  the  vacuum  secured,  the  yields  here  presented  for 
cleavage  of  the  poisons  might  have  been  materially  increased 
with  a  better  vacuum,  the  highest  one  possible  with  the 
apparatus  at  hand  varying  from  20  to  30  mm.  The  following 
table  shows  the  results  of  the  distillations  of  the  free  esters, 
both  the  bath  and  vapor  temperatures  being  given,  the 
amount  of  the  distillates,  and  the  yield  of  crude  crystals 
after  saponification. 


DISTILLATION  OF  THE  ESTERS  OF  THE   MONO-AMINO-ACIDS  FROM  PROTEIN 

POISONS 

Tuberculosis  Poison 


Temperature 

Temperature 

Amount  of            Weight  of 

Fraction. 

of  oil  bath. 

of  vapor. 

distillate.         crude  crystals. 

1 

40°  to    60° 

20°  to    40° 

5  .  0   c  c             2.0  grams 

2 

60°  to    80° 

40°  to    60° 

5.0  c  c 

2.0  grams 

3 

80°  to  100° 

60°  to    80° 

25  .  0  c  c 

16.0  grams 

4 

100°  to  130° 

80°  to  100° 

25  .  0  c  c 

16.0  grams 

5 

130°  to  160° 

25  .  0   c  c 

7.0  grams 

Typhoid  Poison 

Fraction. 

Temperature 
of  bath. 

Temperature 
of  vapor. 

Amount  of 
distillate. 

Weight  of 
crude  crystals. 

1 

25°  to    60° 

10°  to    20° 

12.0  c.c. 

0.3158  gram 

2 

60°  to    80° 

20° 

8.0  c.c. 

1  .  2558  grams 

3 

80°  to  110° 

20° 

2.0  c.c. 

4 

110°  to  130° 

95°  to  108° 

7.0  c.c. 

6  .  0000  grams 

5 

130°  to  145° 

108°  to  110° 

4.0  c.c. 

6 

145°  to  200° 

138°  to  185° 

4.0  c.c. 

No  distillate  passed  over  between  20' 
between  110°  and  138°. 


and  95°,  inside  temperature,  or 


Temperature 

Temperature 

Amount  of 

Weight  of 

of  bath. 

of  vapor. 

distillate. 

crude  crystals. 

40°  to    60° 

28°  to    41° 

1.5   c.c. 

0.07  gram 

60°  to    80° 

41°  to    56° 

1.0   c.c. 

0.06  gram 

80°  to  104° 

56°  to    84° 

4.0    c.c. 

2.00  grams 

104°  to  120° 

84°  to    88° 

2.0    c.c. 

0.63  gram 

120°  to  160° 

88°  to  139 

5.5  c.c. 

90  PROTEIN  POISONS 

Colon  Poison 

Temperature         1 
Fraction. 

1 
2 
3 
4 
5 

The  yield  from  the  colon  poison  was  exceedingly  small,  due  to  the  fact 
that  at  one  stage  of  the  process  part  of  the  solution  was  lost. 

Albumin  Poison. 

Temperature  Amount  of  Weight  of 

Fraction.  of  bath.  distillate.  crude  crystals. 

1  40°  to    60°  5   c.c.  0.2638  gram 

2  60°  to    80°  3   c.c.  0.3338  gram 

3  80°  to  100°  8  c.c.  4. 0000  grams 

4  100°  to  130°  10   c.c.  6.0000  grams 

5  130°  to  160°  7  c.c.  7. 0000  grams 

After  repeated  recrystallizations  these  crude  products 
were  obtained  in  a  state  of  chemical  purity.  From  the 
tuberculosis  poison,  fractions  1  and  2  yielded  needle-shaped 
crystals,  soluble  in  water  and  alcohol,  sweet  to  the  taste, 
and  containing  15.773  per  cent,  of  nitrogen,  the  average 
of  four  determinations  by  the  Kjeldahl  method.  Alanin, 
C3H7NO2,  has  all  these  properties  and  contains  15.73  per 
cent,  of  nitrogen,  thus  identifying  the  crystals  as  alanin. 
Fisher,  Fraenkel,  and  others  do  not  give  a  melting  point  for 
d-alanin,  saying  that  it  is  not  sharp,  due  to  the  presence  of 
a  mixture  of  the  optically  active  and  racemic  forms.  The 
melting-point  of  the  crystals  from  the  tubercle  poison 
varied  from  268°  to  280°  corrected,  showing  no  constant 
temperature.  In  fractions  3  and  4  the  crystals  were  beau- 
tiful, shiny,  satiny  plates,  sweet  to  the  taste,  soluble  in 
water,  but  almost  insoluble  in  alcohol.  These  sublimed 
readily,  melted  with  decomposition,  and  contained  11.976 
per  cent,  of  nitrogen  (average  of  eight  determinations). 
These  properties  and  the  nitrogen  correspond  with  valin, 
a-aminoisovalerianic  acid,  C5HiiNO2,  which  contains  11.965 
per  cent,  nitrogen.  Frankel  gives  the  melting-point  of 


BACTERIAL  CELLULAR  SUBSTANCE       91 

valin  as  298°,  corrected,  when  heated  in  a  closed  tube, 
decomposition  taking  place  at  the  same  time.  The  valin 
from  the  tubercle  poison  melted  as  high  as  296.28°,  cor- 
rected, but  after  continued  recrystallizations  the  melting- 
point  dropped  as  low  as  285°,  and  was  never  reliable. 
Whether  this  was  due  to  a  partial  racemization  on  repeated 
heating  is  not  known.  Heated  in  a  closed  tube  the  melting- 
point  of  the  final  product  was  275.8°  to  278.2°.  As  is  well 
known,  valin  closely  resembles  leucin  in  its  properties,  so 
that  it  is  very  difficult  to  demonstrate  the  existence  of  one  in 
the  presence  of  the  other.  On  page  78,  by  another  method, 
the  presence  of  leucin  in  the  poison  has  been  shown,  but 
by  the  Fischer  method  of  ester  distillation  valin  seems  to  be 
the  one  obtained.  The  presence  of  leucin  was  further 
demonstrated  by  the  fact  that  from  the  final  residue  left 
after  the  esters  had  been  distilled,  crystals  of  its  decomposi- 
tion product,  leucinimide  were  obtained.  This  crystallized 
from  dilute  alcohol  in  the  form  of  needles  and  melted  at 
295.4°.  Cohn  gives  the  melting-point  of  leucinimide  as  295°, 
Frankel  as  262°.  From  fraction  5  of  the  tubercle  poison  a 
qualitative  test  only  was  obtained  for  phenylalanin,  the 
quantity  obtained  being  too  small  for  complete  purification. 
After  evaporation  of  the  ethereal  solution  of  the  thick, 
oily  ester,  according  to  the  method,  the  ester  is  saponified 
by  twice  evaporating  with  hydrochloric  acid.  It  is  then 
evaporated  with  ammonia,  dissolved  in  a  small  amount  of 
water,  and  poured  into  a  large  volume  of  absolute  alcohol, 
which  precipitates  the  phenylalanin.  From  this  precipitate 
the  qualitative  test  was  obtained,  according  to  Frankel, 
by  dissolving  in  dilute  sulphuric  acid  and  adding  an  excess 
of  potassium  dichromate,  producing  the  characteristic 
odor  of  phenylacetaldehyde  and  showing  thus  the  presence 
of  phenylalanin.  From  fraction  5,  after  removal  of  the 
phenylalanin,  upon  saponification  with  barium  hydrate, 
there  was  obtained,  after  the  barium  had  been  removed, 
rhombic  hemihedral  crystals  which  had  a  distinctly  sour 
taste.  These,  after  purification,  showed  9.54  per  cent,  of 
nitrogen,  the  average  of  two  Kjeldahl  determinations, 


92  PROTEIN  POISONS 

identifying  them  as  glutamic  acid,  C5H9NO4,  which  has  9.52 
per  cent,  of  nitrogen.  Frankel  gives  the  melting-point  of 
glutamic  acid  as  202°  to  202.5°,  or  quickly  heated,  213°,  with 
decomposition.  The  product  above  obtained  melted  in  an 
open  tube  at  242°  to  245°,  in  a  closed  tube  at  236°  to  238°. 
From  fraction  1  of  the  typhoid  poison  was  obtained  alanin, 
with  characteristic  properties,  as  described  above.  These 
crystals  showed  15.633  per  cent,  of  nitrogen  and  melted 
at  267°  to  271°.  Fractions  2,  3,  4,  and  5  contained  only 
valin,  which  showed  11.932  per  cent,  of  nitrogen,  the  average 
of  four  determinations.  This  melted  at  287°  to  290.6°  in 
an  open  tube,  278°  to  280°  in  a  closed  one.  From  fraction  6 
the  qualitative  test  for  phenylalanin  was  obtained  as  from 
the  tubercle  poison. 

Owing  to  the  small  yield  of  esters  and  crystals,  fractions 
1  and  2  from  the  colon  poison  could  only  be  determined 
qualitatively.  Both  fractions,  however,  showed  needle- 
shaped  crystals  and  a  sweet  taste,  which  in  conjunction  writh 
the  temperature  at  which  their  esters  distilled  indicated 
alanin.  Fractions  3  and  4  gave  characteristic  valin  crystals 
containing  11.942  per  cent,  of  nitrogen  and  melting  at 
283.4°  to  285°  in  an  open  tube,  or  274°  to  277°  in  a  closed 
tube.  Phenylalanin  was  obtained  qualitatively  from  this 
as  from  the  two  preceding  poisons.  After  its  extraction 
from  fraction  5,  and  after  saponification  with  barium 
hydrate  and  its  removal,  crystals  in  the  form  of  rhombic 
plates  and  prisms,  insoluble  in  alcohol,  were  obtained. 
These  corresponded  with  those  of  aspartic  acid,  and  as  the 
quantity  was  not  sufficient  for  purification  by  recrystalli- 
zation,  the  copper  salt  was  formed  with  copper  acetate. 
This  was  obtained  in  the  form  of  needles,  very  difficultly 
soluble  in  cold  water,  difficultly  in  hot,  which  again 
corresponded  with  the  properties  of  aspartic  acid. 

When  the  crystals  from  the  fractions  obtained  from 
albumin  poison  were  examined  the  result  was  not  different. 
Fractions  1  and  2  produced  characteristic  alanin  crystals 
with  11.75  per  cent,  of  nitrogen,  the  average  of  four  deter- 
minations. The  melting-point  was  277°  to  279.6°.  Valin, 


BACTERIAL  CELLULAR  SUBSTANCE       93 

with  form  and  properties  as  already  given,  was  obtained 
from  both  fraction  3  and  4,  containing  11.935  per  cent,  of 
nitrogen,  and  showing  a  melting-point  of  282°  to  286.4°  in 
an  open  tube,  and  of  279°  to  283  in  a  closed  tube.  Like- 
wise from  fraction  5  the  heavy  oil  of  phenylalanin  ethyl 
ester  was  obtained,  and  from  this  as  in  the  other  cases  the 
qualitative  test  for  phenylalanin  by  the  production  of 
phenylacetaldehyde.  The  remaining  portion  of  fraction  5 
yielded  the  same  rhombic  plates  and  prisms  as  described 
under  the  colon  fractions,  and  which  are  like  those  of  aspartic 
acid  properly  obtained  at  this  point  if  present.  The  copper 
salt  was  again  formed,  the  same  needles,  very  difficultly 
soluble  in  cold  water,  difficultly  in  hot,  being  obtained. 
The  amount  of  crystals  was  too  small  for  further  identi- 
fication. 

From  this  it  will  be  seen  that  mono-amino-acids  are 
obtained  from  the  protein  poisons  after  hydrolysis  with 
strong  acid.  It  is  not  claimed  that  these  are  the  only 
mono-amino-acids  present,  or  that  all  of  these  have  been 
sufficiently  identified,  but  in  consideration  of  the  fact  that 
those  discussed  were  found  in  the  proper  fraction  according 
to  Fischer's  separation  and  according  to  the  boiling-points 
of  their  esters,  that  the  crystalline  form  and  qualitative 
properties  corresponded,  and  that,  when  it  could  be  deter- 
mined, the  percentage  of  nitrogen  was  close  to  the  theo- 
retical, it  seems  fair  to  conclude  that  the  following  tabulation 
is  not  far  from  correct: 

MONO-AMINO-ACIDS  OF  THE  PROTEIN  POISONS 


Tuberculosis 
poison. 
Alanin 
Valin 
Phenylalanin 

.  Typhoid 
poison, 
alanin 
valin 
phenylalanin 

Colon 
poison, 
alanin 
valin 
phenylalanin 

Albumin 
poison, 
alanin 
valin 
phenylalanin 

Leucinimide 


Aspartic  acid         aspartic  acid 


This  is  sufficient  to  establish  the  point  for  the  proof  of 
which  the  method  was  employed,  that  is,  the  protein  nature 


94  PROTEIN  POISONS 

of  the  poisonous  group  of  the  protein  molecule.  Attention 
is  called  also  to  the  comparative  simplicity  of  the  group  and 
to  the  great  similarity  of  acids  obtained  from  the  different 
poisons.  This  accords  well  with  the  great  similarity  and 
non-specificity  of  their  physiological  action. 

It  is  interesting  that  the  final  residue  left  after  distilla- 
tion of  the  esters  gives  still  a  very  intense  Millon  reaction, 
which  cannot  be  ascribed  to  the  presence  of  tyrosin. 

It  will  be  clearly  understood  that  this  work  does  not 
show  that  the  active  agent  or  agents  in  the  "crude  soluble 
poison"  is  or  are  protein  in  nature. 


CHAPTER  V 

THE  CLEAVAGE  OF  PROTEINS  WITH  DILUTE 
ALKALI  IN  SOLUTION  IN  ABSOLUTE 
ALCOHOL 

THE  researches  detailed  in  the  preceding  pages  seem  to 
establish  the  following  propositions: 

1.  The  cellular  substances  of  bacteria  consist  largely  of 
proteins    that    yield    split    products    identical    with    those 
obtained  by  the  hydrolysis  of  vegetable  and  animal  proteins. 

It  has  been  shown  that  the  bacterial  cellular  substances, 
when  broken  up  with  mineral  acids  or  alkalies,  furnish 
ammonia,  mono-amino  and  diamino  nitrogen,  one  or  more 
carbohydrate  groups,  and  humin  substances.  It  seemed 
therefore  logical  to  conclude  that  the  bacterial  cell  consists 
largely  of  proteins. 

2.  The  proteins  of  the  bacterial  cell   contain   at  least 
one  group  which  when  injected  intra-abdominally,  subcu- 
taneously,   or  intravenously  in  anmials,   has  a  markedly 
poisonous  effect. 

3.  This  poisonous  group  may  be  detached  from  the  cell 
protein  by  hydrolysis  with  either  dilute  acids  or  alkalies. 

4.  The  dilute  alkali   furnishes  the  better  means  of  ex- 
tracting the  poisonous  group. 

5.  When  the  bacterial  protein  is  broken  up  with  alkali 
in  dilute  aqueous  solution,  at  least  two  groups  are  split 
off  and  pass  into  solution.     These  are  the  carbohydrate 
and  the  poisonous  groups.     Both  are  soluble  in  water  and 
in  dilute  alcohol,  and  their  separation,  when  the  cell  protein 
is  disrupted  by  alkali  in  aqueous  solution,  is  difficult  and 
unsatisfactory. 

6.  Since  the  carbohydrate  group  is  insoluble  in  absolute 
alcohol,  while  the  poisonous  group  is  more  readily  soluble 


96  PROTEIN  POISONS 

in  this  menstruum  than  in  water,  it  was  decided  to  attempt 
to  disrupt  the  cell  protein  with  a  solution  of  alkali  in  absolute 
alcohol.  Another  idea  also  acted  as  a  determining  factor 
in  attempting  this  method  of  hydrolysis,  and  in  fact  it 
was  at  that  time  the  dominating  factor.  The  effect  of  the 
poisonous  group  on  animals  so  closely  resembles  that  of 
neurin  that  it  was  thought  that  the  two  might  be  identical, 
or  at  least  that  the  poisonous  body  might  contain  neurin. 
Knowing  that  neurin  can  be  heated  without  decomposition 
in  alkaline  alcohol  was,  therefore,  a  reason  for  trying  this 
method. 

7.  Previous  experiments  had  demonstrated  the  advantage 
of  extracting  the  cell  substance  thoroughly  with  alcohol  and 
ether  before  submitting  it  to  hydrolysis.  This  frees  the 
material  from  fat,  wax,  and  other  substances  soluble  in 
alcohol  or  ether,  and  since  it  had  been  shown  that  these 
are  no  part  of  the  cell  protein  it  is  beneficial  to  get  rid  of 
them  in  toto  before  hydrolysis  is  attempted. 

The  following  preliminary  trials  were  made  by  Vaughan 
and  Wheeler  (in  the  fall  of  1903)  in  order  to  compare 
hydrolysis  with  aqueous  and  alcoholic  solutions  of  alkali. 

Two  samples,  of  10  grams  each,  of  the  cellular  substance 
of  the  colon  bacillus  were  taken.  This  material  had  pre- 
viously been  thoroughly  extracted  with  alcohol  and  ether. 
One  sample  was  mixed  with  250  c.c.  of  a  1  per  cent,  aqueous 
solution  of  sodium  hydroxide  and  the  other  with  the  same 
volume  of  an  absolute  alcohol  solution  of  the  same  substance 
in  the  same  strength.  These  mixtures  were  heated  in 
flasks,  fitted  with  reflux  condensers,  for  one  hour  on  the 
water-bath.  Ten  cubic  centimeters  of  the  clear  filtrate 
from  each  was  evaporated,  the  aqueous  solution  to  5  c.c. 
and  the  alcoholic  to  dryness,  and  then  taken  up  in  5  c.c. 
of  water.  Each  was  carefully  neutralized  with  dilute 
hydrochloric  acid  and  injected  into  the  abdominal  cavity 
of  a  guinea-pig.  Both  animals  developed  in  a  characteristic 
manner  the  first  and  second  stages  of  poisoning  with  the 
split  product,  but  neither  died.  This  experiment  showed 
that  the  poison  was  present  in  both  extracts,  and,  so  far 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI       97 

as  we  could  judge  by  the  development  and  intensity  of 
the  symptoms,  in  similar  amounts.  That  the  poison  could 
be  extracted  by  alkaline  alcohol  was  proved.  '  However, 
the  yield  was  not  satisfactory,  and  a  second  test  was  made, 
and  in  this  the  strength  of  the  alkali  was  doubled.  These 
were  treated  as  before,  and  the  pigs  that  received  the 
injections  developed  the  characteristic  symptoms  and 
died.  The  one  that  had  the  alcoholic  extract  died  within 
six,  and  the  other  within  eight  minutes.  This  confirmed 
the  hope  that  the  alcoholic  alkali  was  quite  as  efficient  as 
the  aqueous  in  the  extraction  of  the  poisonous  group. 
While  the  aqueous  extract  contained  a  large  amount  of 
the  carbohydrate  group,  it  was  found  that  the  alcoholic 
extract,  after  evaporation  to  dryness  and  solution  in  water, 
gave  the  biuret,  Millon,  and  xanthoproteic  tests,  but  failed 
wholly  to  give  the  Molisch  reaction.  The  carbohydrate 
group  had  been  split  off  in  both  samples,  but  being  insoluble 
in  absolute  alcohol,  it  remained  with  the  insoluble  portion 
of  the  cellular  substance. 

The  above  and  many  other  experiments  have  demon- 
strated that  the  best  method,  so  far  devised,  for  extracting 
the  poisonous  group  from  the  cell  protein,  or,  as  subsequent 
work  has  shown,  from  any  protein,  is  by  means  of  a  2  per 
cent,  solution  of  sodium'  hydroxide  in  absolute  alcohol. 
If  satisfactory  results  are  obtained,  the  alcohol  used  in  the 
extraction  must  be  absolute.  If  it  is  not,  more  or  less  of 
the  carbohydrate  will  be  mixed  with  the  poison;  a  sticky 
mass  will  be  obtained,  and  the  patience  of  the  experimenter 
will  be  taxed  severely.  Previous  thorough  extraction  of 
the  protein  with  alcohol  and  ether  for  the  removal  of  fats, 
waxes,  and  other  substances  soluble  in  these  agents,  is 
also  essential  to  satisfactory  work. 

The  method  for  preparing  the  bacterial  cellular  substance 
has  been  given,  but  it  may  be  well  to  give  here  some  details 
for  the  preparation  of  egg-white  before  splitting  it  up  into 
poisonous  and  non-poisonous  proteins. 

Fresh  eggs  (we  have  usually  taken  twenty  dozen  at  a 
time)  are  broken  and  the  whites  dropped  into  a  beaker  or 
7 


98  PROTEIN  POISONS 

precipitating  jar,  then  poured  with  constant  stirring  into 
four  volumes  of  95  per  cent,  alcohol.  This  stands  with 
frequent  stirring  for  two  days,  then  the  alcohol  is  decanted, 
and  replaced  with  the  same  volume  of  absolute  alcohol. 
This  is  allowed  to  stand  for  from  one  to  two  days,  when 
the  coagulated  albumin  is  collected  on  a  filter,  allowed  to 
drain,  then  placed  in  large  Soxhlets  and  extracted  with 
ether  for  from  one  to  two  days.  It  is  then  ground  in  porce- 
lain mortars  and  passed  through  fine  meshed  sieves.  This 
gives  a  beautifully  white  powder  which  may  be  kept  in 
bottles  in  stock  from  which  portions  are  taken  for  the 
purpose  of  hydrolyzing  it. 

Twenty  dozen  eggs  yield  about  735  grams  of  this  powder, 
a  little  more  than  3  grams  per  egg. 

A  weighed  portion  of  the  protein,  prepared  as  above,  is 
placed  in  a  flask,  covered  with  from  fifteen  to  twenty-five 
times  its  weight  of  absolute  alcohol  in  which  2  per  cent,  of 
sodium  hydroxide  has  been  dissolved.  The  flask,  fitted 
with  a  reflux  condenser,  is  heated  on  the  water-bath  for 
one  hour,  when  it  is  allowed  to  cool  and  the  insoluble  portion 
collected  on  a  filter.  After  thorough  draining  the  insoluble 
part  is  returned  to  the  flask  and  the  extraction  repeated. 
It  has  been  found  that  three  extractions  are  necessary  in 
order  to  split  off  all  the  poisonous  group.  The  temperature 
of  these  extractions  is  78°,  the  temperature  of  boiling 
absolute  alcohol.  By  this  method  the  protein  is  split  into 
two  portions,  one  of  which  is  soluble  in  absolute  alcohol 
and  is  poisonous,  while  the  other  is  insoluble  in  absolute 
alcohol  and  is  not  poisonous. 

A  large  number  of  protein  bodies,  bacterial,  vegetable, 
and  animal,  have  been  split  up  in  this  way  and  no  true 
protein  has  failed  to  yield  a  poisonous  portion.  Among 
the  proteins  with  which  we  have  worked  the  following  may 
be  mentioned:  egg-white,  casein,  serum  albumin,  edestin, 
zein,  Witte's  peptone,  Macquaire's  peptone,  de  Chapoteaut's 
peptone,  the  tissue  of  cancers,  and  the  cellular  substance 
of  bacillus  coli  communis,  b.  typhosus,  b.  anthracis,  b. 
tuberculosis,  b.  Moelleri  (timothy),  sarcina  lutea,  b.  ruber 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI      99 

of  Kiel,  b.  proteus,  b.  subtilis,  b.  megaterium,  b.  pyo- 
cyaneus,  b.  pneumonise,  and  b.  diphtheriae.  Gelatin  con- 
tains no  poison,  but  gelatin  is  an  albuminoid  and  gives  the 
Millon  test  imperfectly,  if  at  all.  Nicolle  and  Abt1  found 
that  Defresne's  peptone  does  not  yield  a  poison  when 
treated  by  our  method,  and  we  have  confirmed  this  finding. 
It  would  be  interesting  to  know  whether  this  peptone  is 
made  from  gelatin  or  from  a  true  protein.  The  probabilities 
are  that  in  peptic  digestion  a  point  is  reached  when  the 
poisonous  group  in  proteins  is  disrupted.  In  fact,  as  has 
been  stated  (page  42),  we  have  shown  that  the  poison  in 
the  cellular  substance  of  the  colon  bacillus  is  slowly  digested 
and  destroyed  by  digestion  with  pepsin-hydrochloric  acid. 
Therefore,  it  is  not  strange  that  certain  peptones  fail  to 
yield  a  poisonous  body  when  disrupted  with  dilute  alkali 
in  absolute  alcohol. '  Witte's  peptone,  so-called,  as  is  well 
known,  is  not  a  peptone,  but  an  albumose. 

This  poison,  like  the  whole  protein  of  which  it  is  a  part, 
is  formed  synthetically  by  the  living  cell.  In  case  of  the 
colon  poison  we  demonstrated  this  by  growing  the  bacillus 
in  Fraenkel's  modification  of  Uschinsky's  medium,  which 
has  the  following  composition: 

Water 10,000  parts 

Sodium  chloride 50  parts 

Asparagin 34  parts 

Ammonium  lactate 63  parts 

Di-sodium  hydrogen  phosphate 20  parts 

After  a  week's  development  the  contents  of  these  flasks 
were  poured  into  from  two  to  three  volumes  of  95  per  cent, 
alcohol.  The  precipitate  was  filtered  out  and  put  into 
absolute  alcohol;  next  it  was  extracted  in  Soxhlets  with 
ether,  dried,  and  powdered.  This  powdered  cellular  sub- 
stance, when  split  up  with  2  per  cent,  sodium  hydroxide 
in  absolute  alcohol,  furnished  the  poison,  the  action  of 
which  was  demonstrated  on  guinea-pigs.  Moreover,  the 
poison  obtained  in  this  way  gave  all  the  protein  reactions 

1  Annales  de  1'Institut  Pasteur,  February,  1908. 


100  PROTEIN  POISONS 

hereafter  described  as  being  obtained  from  the  poison  from 
agar-grown  cultures.  This  demonstrates  that  the  poison  is. 
an  integral  part  of  the  cellular  substance,  and  it  is  evident 
that  the  bacterial  cell  must  synthetically  produce  this 
protein  body  during  its  growth  from  the  chemical  con- 
stituents of  the  medium. 

When  the  protein  is  split  up  by  dilute  alkali  in  absolute 
alcohol  according  to  the  method  described,  the  poison  is 
in  solution  in  the  alkaline  alcohol.  The  preparation  is 
filtered  and  the  filtrate  neutralized  with  hydrochloric  acid, 
avoiding  an  excess  of  acid.  This  throws  down  the  greater 
part  of  both  base  and  acid  as  sodium  chloride,  which  is 
removed  by  filtration.  In  this  way  a  solution  of  the  poison 
in  absolute  alcohol  is  obtained.  This  is  evaported  in 
vacuo  at  40°,  redissolved  in  absolute  alcohol  to  remove 
traces  of  sodium  chloride,  and  again  evaporated  in  vacuo 
at  40°  or  less.  Evaporation  may  be  done  in  an  open  dish, 
but  the  toxicity  of  the  substances  is  somewhat  decreased 
when  this  is  done.  The  poisonous  part  of  the  protein 
molecule  when  obtained  in  this  way  and  powdered,  when 
there  is  no  water  present,  forms  a  dark  brown  scale  which 
pulverizes  into  a  lighter  brown  powder. 

It  should  be  clearly  understood  that  we  regard  this 
method  of  extracting  the  poisonous  group  from  the  protein 
molecule  as  by  no  means  ideal.  We  know  that  it  is  crude 
and  that  much  of  the  poison  is  destroyed  in  the  process. 
In  disrupting  a  protein  by  our  method  with  dilute  alkali 
in  absolute  alcohol,  ammonia  is  given  off  and  the  odor  of 
this  gas  is  apparent  even  at  the  end  of  the  third  extraction. 
An  effort  was  made  to  discover  how  much  nitrogen  was 
converted  into  ammonia  in  the  process.  A  device  was 
arranged  for  conducting  the  ammonia  into  standard  acid, 
and  four  10-gram  samples  of  Witte's  peptone  were  extracted 
with  2  per  cent,  sodium  hydrate  in  absolute  alcohol,  one 
for  three  hours  in  a  current  of  air,  the  others  in  a  current 
of  hydrogen  for  two  and  one-half,  eight  and  one-half,  and 
nineteen  and  one-half  hours  respectively.  At  the  end  of 
each  operation  the  excess  of  acid  was  titrated  with  deci- 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     101 


normal  sodium  hydrate,  and  the  percentage  of  nitrogen 
calculated.  The  relative  toxicity  of  the  split  products  was 
determined.  In  every  case  ammonia  was  still  being  pro- 
duced when  the  process  was  interrupted.  Again,  a  10-gram 
sample  of  the  poison  from  egg  ablumen  was  boiled  for 
fifty-four  and  one-half  hours  with  2  per  cent,  alcoholic 
alkali  to  ascertain  if  ammonia  could  be  split  from  the 
poison  itself.  The  results  of  this  work  are  shown  in  the 
following  table: 

AMMONIA  PRODUCED  BY  CLEAVAGE  OF  PKOTEIN  WITH  DILUTE  ALKALI  IN 
ABSOLUTE  ALCOHOL. 
Per  cent,  of 
Time  in   Atmos-       N  given         Rate  per 


Sample.  hours. 

Witte  peptone         3.0 
Witte  peptone         2,5 


Witte  peptone 
Witte  peptone 
Poison 


8.5 
19.5 
54.5 


phere. 
air 
H 


H 
H 
II 


off. 

hour. 

Toxicity. 

0.4305 

0.1435 

Diminished. 

0.3956 

0.1582 

Greater   than 

that  in  air. 

0.7383 

0.0868 

Diminished. 

1.0517 

0.0539 

Diminished. 

1.4800 

0.0270 

Diminished 

by  half. 

The  albumin  poison  as  ordinarily  obtained  contains 
13.74  per  cent,  of  nitrogen.  By  the  fifty-four  and  one-half 
hours'  heating  with  alcoholic  alkali,  10.77  per  cent,  of  its 
nitrogen  was  converted  into  ammonia.  After  this  treatment 
the  poison  still  gave  a  good  Millon  test,  but  no  longer  the 
biuret. 

It  is  probable  that  by  continued  heating  in  the  same 
manner  quite  all  of  the  nitrogen  could  be  separated,  though 
it  is  noticeable  that  the  rate  was  greatly  diminished  as  the 
time  lengthened.  The  decrease  in  toxicity  with  the  evolu- 
tion of  ammonia  suggests  that  this  group  is  essential  to  the 
toxicity  of  the  poison.  This  seems  to  be  highly  probable. 

Properties  of  the  Crude  Soluble  Poison.  The  poison  split 
off  from  the  protein  molecule  by  the  method  above  given 
is  designated  as  "the  crude  soluble  poison;"  "crude" 
because  it  is  undoubtedly  a  mixture  of  chemical  bodies, 
and  "soluble"  in  contradistinction  to  the  bacterial  cellular 
substance,  from  which  it  was  first  prepared,  and  which  is 
poisonous,  but  not  soluble. 


102  PROTEIN  POISONS 

The  brownish  toxic  powder,  varying  in  shade  of  color 
somewhat  with  the  protein  from  which  it  has  been  obtained, 
has  a  peculiar  odor.  It  is  highly  hydroscopic,  and  the 
poisonous  .portion  is  freely  soluble  in  water.  The  solubility 
of  the  whole  powder,  however,  varies  with  the  protein 
from  which  it  is  obtained,  and  possibly  with  the  length 
of  time  that  it  has  been  exposed  to  the  alkali  in  the  alcohol. 
Any  portion  insoluble  in  water  should  be  removed  by 
filtration,  and  in  some  instances  we  have  found  filtration 
through  porcelain  necessary.  Generally  the  powder  dis- 
solves in  water  with  a  slight  opalescence  easily  removed  by 
filtration  through  paper.  In  all  cases  we  have  found  the 
portion  insoluble  in  water  free  from  toxic  effect.  Aqueous 
solutions  of  the  poison  are  decidedly  acid  to  litmus,  the 
acidity  being  due  to  some  organic  body  and  probably 
not  to  the  poison  itself.  On  neutralization  with  sodium 
bicarbonate  a  brownish,  non-toxic  precipitate  is  formed. 
Prolonged  contact  with  alkali,  as  we  shall  see  later,  lessens 
the  activity  of  the  poison,  and  even  neutralization  has 
some  effect,  which  is  more  marked  the  longer  the  prepara- 
tion stands.  We  are  inclined  to  attribute  this  to  the  forma- 
tion of  a  salt  with  the  acid  poison  and  the  alkali.  The 
poison  is  freely  soluble  in  alcohol,  more  readily  than  in 
water.  Alcoholic  solutions  on  long  standing  deposit  small 
brownish  sediments  which  we  have  always  found  to  be 
inert.  When  an  alcoholic  solution  is  evaporated,  there  is 
a  part  of  the  residue  that  is  insoluble  in  absolute  alcohol. 
These  portions  also  are  devoid  of  toxic  effect.  Alcoholic 
solutions  have  been  kept  for  five  years  without  recog- 
nizable loss  in  toxicity,  and  even  aqueous  solutions  decom- 
pose very  slowly.  The  poison  is  soluble  in  methyl  as  well 
as  in  ethyl  alcohol.  It  is  insoluble  in  ether,  chloroform, 
and  petroleum  ether.  Each  of  these  removes  a  small 
amount  of  fatty  substance,  which  is  non-toxic,  but  they 
do  not  dissolve  an  appreciable  quantity  of  the  poison. 
From  its  alcoholic  solution  the  poison  is  precipitated  by 
ether,  but  contact  with  ether  decreases  its  toxicity  to  such 
an  extent  that  this  method  is  not  applicable  in  attempts 
at  purification. 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     103 

The  "crude  soluble  poison"  is  soluble  in  strong  mineral 
acids,  and  such  solutions  remain  clear  on  being  boiled  and 
on  dilution  with  water.  However,  a  few  drops  of  mineral 
acid  added  to  an  aqueous  solution  cause  a  precipitate, 
which  seems  to  indicate  that  the  acidity  of  the  aqueous 
solution  is  caused  by  the  presence  of  some  organic  acid. 

The  poison  diffuses  slowly  through  collodion  sacs  both 
within  the  animal  body  and  when  suspended  in  distilled 
water.  The  following  experiments  bear  on  this  point: 
Two  hundred  milligrams  of  the  crude  soluble  poison  from 
the  cellular  substance  of  the  typhoid  bacillus  dissolved  in 
20  c.c.  of  water  was  placed  in  each  of  two  collodion  sacs 
which  were  then  suspended  in  distilled  water.  At  the 
end  of  twenty-four  hours,  the  Millon  reaction  was  given 
by  the  dialysate.  This  was  replaced  every  twenty-four 
hours  by  fresh  distilled  water,  and  the  dialysis  continued 
for  ninety-six  hours.  At  the  end  of  this  time  the  combined 
dialysates  were  concentrated  to  dryness,  the  residues  dis- 
solved in  absolute  alcohol,  filtered,  and  again  evaporated. 
The  brown,  sticky  residue,  thus  obtained,  dissolved  in 
water,  was  acid  in  reaction,  had  the  characteristic  odor, 
and  when  injected  into  a  guinea-pig,  killed  in  twenty  minutes 
with  typical  symptoms,  thus  showing  that  the  poison  does 
diffuse  through  a  collodion  sac.  So  slowly,  however,  does 
it  diffuse  that  at  the  end  of  ninety-six  hours  it  was  not 
wholly  removed  from  the  sac.  In  another  experiment 
one  gram  of  the  same  poison  in  8  c.c.  of  water  was  put  into 
a  collodion  sac  which  was  introduced  into  the  abdominal 
cavity  of  a  medium-sized  rabbit.  After  twelve  days,  the 
animal  not  being  visibly  affected,  the  sac  was  removed 
and  found  to  contain  6  c.c.  of  a  clear  fluid  which  looked 
more  like  blood  serum  than  anything  else.  Five  cubic 
centimeters  of  this  injected  into  the  abdomen  of  a  guinea- 
pig  had  no  effect.  We  conclude  from  this  that  the  poison 
had  diffused  from  the  sac,  but  so  slowly  that  it  was  disposed 
of  by  the  animal's  body  without  recognizable  discomfort. 

Notwithstanding  the  ready  solubility  of  the  crude  soluble 
poison  in  absolute  alcohol,  we  must  regard  it  as  either 


104  PROTEIN  POISONS 

being  a  protein  itself  or  as  being  mixed  with  one  or  more 
proteins.  Its  aqueous  solutions  give  all  the  protein  color 
reactions  with  the  important  exception  of  that  of  Molisch. 
It  is  worthy  of  note  that  the  part  that  separates  from 
alcoholic  solution  on  long  standing  is  inert  and  does  not 
give  the  protein  reactions,  while  the  solution  does  not 
decrease  in  toxicity.  This  indicates  that  the  protein  is 
permanently  soluble  in  absolute  alcohol.  The  Millon 
reaction  shows  most  perfectly  and  persistently  whenever 
the  poison  is  found.  It  is  generally  believed  by  physio- 
logical chemists  that  this  reaction  is  given  by  all  benzene 
derivatives  in  which  one  hydrogen  atom  has  been  replaced 
by  a  hydroxyl  group,  and  it  is  also  generally  supposed  that 
tyrosin  is  the  only  oxyphenyl  compound  in  the  protein 
molecule,  therefore  this  reaction  is  presumed  to  show  the 
presence  of  tyrosin.  This  is  interesting  in  view  of  the 
fact  already  stated  that  gelatin,  which  contains  no  tyrosin, 
or  but  little,  yields  no  poison.  The  fact  that  the  poison 
contains  no  carbohydrate,  as  shown  by  its  failure  to  respond 
to  the  Molisch  test,  an  exceedingly  delicate  test,  is,  in  our 
opinion,  strong  evidence  that  the  cleavage  in  the  protein 
molecule  induced  by  dilute  alkali  in  absolute  alcohol  at 
the  temperature  of  78°  follows  along  structural  lines.  If 
the  change  were  one  of  simple  degradation  without  chemical 
cleavage  it  wrould  be  difficult  to  explain  the  absolute  failure 
of  the  carbohydrate  test  in  the  crude  soluble  poison.  It 
seems  quite  evident  from  our  work  that  in  the  process  the 
complex  protein  molecule  is  split  into  several  groups,  one 
of  which  is  the  poison  and  another  is  a  carbohydrate,  the 
former  being  freely  soluble  in  absolute  alcohol,  while  the 
latter  is  insoluble  in  this  reagent.  It  should  be  stated  that 
the  crude,  soluble  poison  not  only  fails  to  respond  to  the 
Molisch  test,  but  it  also  fails  to  reduce  Fehling's  solution 
after  prolonged  boiling  with  dilute  mineral  acid. 

The  crude  soluble  poison  gives  the  biuret  test  beauti- 
fully, therefore  we  must  say  that  the  poison  either  is  itself 
a  biuret  body  or  is  mixed  with  such  a  body.  As  is  well 
known,  the  biuret  test  is  regarded  as  the  landmark  between 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     105 

proteins  and  their  simpler  non-protein  disruption  products, 
and,  so  long  as  a  disrupted  protein  continues  to  give  the 
biuret  test  it  must  still  be  classed  among  the  proteins.  It 
will  certainly  be  understood  that  the  pure  poison  may 
not  be  a  protein,  but  until  it  is  purified  sufficiently  to  fail 
to  give  the  biuret  test  it  must  be  regarded  as  a  protein. 

The  poison  responds  nicely  to  the  Adamkiewicz  or  gly- 
oxylic  acid  test.  Hopkins  and  Cole  have  shown  quite 
convincingly  that  this  color  test  depends  upon  the  presence 
of  tryptophan  or  indol-amino-propionic  acid;  therefore,  while 
we  have  made  no  direct  search  for  tryptophan  in  our  poison, 
we  assume  its  presence  on  account  of  the  unequivocal 
response  to  this  test. 

When  the  poison  is  boiled  with  concentrated  hydrochloric 
acid  to  which  a  drop  of  concentrated  sulphuric  acid  has 
been  added,  the  powder  passes  into  solution  and  a  violet 
color  results,  thus  giving  Liebermann's  test.  At  one  time 
Hofmeister  believed  this  to  be  a  carbohydrate  reaction  in 
which  furfurol  and  the  aromatic  oxyphenyl  radicals  take 
part,  but  Cole  has  shown  that  this,  like  the  Adamkiewicz 
test,  also  once  regarded  as  a  carbohydrate  test,  is  due  to 
the  tryptophan  group.  We  are  quite  convinced  that  our 
soluble  poison  contains  no  carbohydrate,  and  we  regard 
the  fact  that  it  does  respond  to  the  Liebermann  test  as  a 
strong  confirmation  of  the  error  of  Hofmeister's  explanation 
of  this  test,  and  in  favor  of  the  explanation  given  by  Cole. 

When  heated  with  strong  nitric  acid  the  powdered  poison 
goes  into  solution,  more  or  less  yellow  according  to  the 
amount  used,  and  this  becomes  orange  on  the  addition  of 
ammonia,  thus  giving  the  xanthroproteic  test  and  indicating 
the  existence  of  aromatic  radicals. 

The  ordinary  test  for  sulphur  in  proteins,  that  of  heating 
with  excess  of  sodium  hydrate  in  the  presence  of  a  small 
amount  of  acetate  of  lead,  is  not  given  by  the  portion  of 
the  protein  split  off  "by  alkali  in  absolute  alcohol.  If, 
however,  a  portion  of  the  substance  in  a  test-tube  is  fused 
with  metallic  sodium  and  the  cooled  mass  treated  with 
water,  a  few  drops  of  a  freshly  prepared  solution  of  sodium 


106  PROTEIN  POISONS 

nitroprussiate  added  to  a  part  of  the  clear  filtrate,  a  beauti- 
ful violet  color  is  produced,  indicating  the  presence  of  sulphur. 
Also,  if  the  other  part  of  the  clear  filtrate  be  treated  with  a 
lead  acetate  solution,  lead  sulphide  is  precipitated.  If  the 
solution  be  acidified  before  lead  acetate  is  added  a  faint 
but  unmistakable  odor  of  hydrogen  sulphide  is  detected. 
It  is  known  that  sulphur  may  exist  in  the  protein  molecule 
in  at  least  two  forms,  one  part  being  readily  split  off  with 
dilute  alkali  as  a  sulphide,  the  other  being  obtained  only 
when  the  disruption  of  the  protein  molecule  is  carried  much 
farther.  It  is  still  a  question  whether  or  not  both  of  these 
sulphur  groups  come  from  cystin.  Since  the  nitroprussiate 
reaction  is  very  delicate,  no  conclusion  as  to  the  amount 
of  sulphur  can  be  drawn  from  this  test,  and  although  a 
good  precipitate  of  lead  sulphide  is  formed,  the  amount  of 
sulphur  in  the  poison  is  probably  not  large,  since  Leach 
failed  entirely  to  find  sulphur  in  the  ash  of  the  colon  bacillus, 
though  both  the  cellular  substance  and  the  non-poisonous 
portion,  as  well  as  the  poison,  respond  to  the  nitroprussiate 
test  for  sulphur  and  also  give  the  lead  sulphide  precipitate 
in  the  clear  acidified  filtrate  from  the  fused  mass. 

A  solution  of  this  toxic  substance  is  not  coagulated  by 
heat  in  acid,  neutral,  or  alkaline  solution,  though,  as  already 
stated,  a  few  drops  of  a  mineral  acid  added  to  an  aqueous 
solution  causes  the  appearance  of  a  considerable  precipitate, 
which  is  not  soluble  on  heating  or  on  the  further  addition 
of  acid.  This  precipitate  is  produced  regardless  of  the 
previous  removal  of  the  opalescence  from  the  aqueous 
solution. 

Among  the  metallic  salts,  copper  sulphate  produces  no 
precipitate  and  ferric  chloride  only  on  heating.  Silver 
nitrate  naturally  precipitates  any  trace  of  chlorides  present, 
but  after  the  addition  of  an  excess  of  ammonia  there  still 
remains  a  small  precipitate.  Potassium  ferrocyanide  gives 
a  precipitate,  also  potassium  bismuth  iodide  in  acid  solution. 
Lead  acetate,  mercuric  chloride,  and  platinum  chloride 
all  produce  heavy  precipitates.  With  lead  acetate  and 
mercuric  chloride,  however,  after  removal  of  lead  and 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     107 

mercury  with  hydrogen  sulphide  from  their  respective 
precipitates  and  filtrates,  the  protein  reactions  are  given 
by  the  filtrates,  and  here  also  is  found  the  poison  in  each 
case.  From  10  to  15  per  cent,  of  the  crude  poison  can  be 
precipitated  by  the  use  of  platinum  chloride  in  either  water 
or  alcoholic  solution.  All  attempts  to  crystallize  this 
precipitate  failed,  as  only  a  small  part  of  it  is  dissolved  by 
hot  water,  and  the  insoluble  part  is  unaffected  by  any  of 
the  ordinary  solvents.  The  protein  reactions  are  given 
by  the  platinum  precipitate,  by  both  soluble  and  insoluble 
parts,  but  not  by  the  filtrate.  The  poison  is  found  in  the 
insoluble  part  of  the  precipitate  after  removal  of  the  plati- 
num by  hydrogen  sulphide,  its  toxicity  being  markedly 
increased.  The  other  parts,  after  removal  of  the  platinum, 
are  inert. 

The  most  active  products  have  been  obtained  by  precipi- 
tation from  solution  in  absolute  alcohol  with  alcoholic  solu- 
tions of  the  chlorides  of  platinum,  mercury,  and  copper  and 
removal  of  the  base  from  the  precipitate  with  hydrogen 
sulphide.  By  this  method  we  have  obtained  a  body  which 
kills  guinea-pigs  of  from  200  to  300  grams'  weight  in  doses 
of  0.5  mg.  given  intravenously. 

From  a  water  solution  of  the  poison,  bodies  giving  protein 
reactions  may  be  salted  out  by  the  addition  of  ammonium 
sulphate  or  sodium  chloride  to  saturation,  but  in  neither 
case  is  the  separation  complete,  the  filtrates  still  responding 
to  the  protein  color  tests  after  removal  of  the  neutral 
salts.  In  case  of  salting  out  with  ammonium  sulphate,  the 
solubility  of  both  parts  is  thereby  lessened  and  the  toxicity 
diminished,  possibly  on  account  of  decreased  solubility, 
though  both  parts  exhibit  some  poisonous  action,  and  like- 
wise both  show  the  protein  color  tests. 

Phosphotungstic,  phosphomolybdic,  and  picric  acids  all 
give  abundant  precipitates.  Since  these  reagents  are  also 
used  in  the  precipitation  of  alkaloidal  bodies,  the  precipitates 
.with  phosphomolybdic  and  phosphotungstic  acids  were 
further  examined,  the  possibility  suggesting  itself  that  the 
toxic  body  might  be  alkaloidal  in  nature,  and  that  the 


108  PROTEIN  POISONS 

protein  part  might  be  entirely  separate  from  the  poison. 
A  sample  was  precipitated  with  phosphomolybdic  acid 
in  acid  solution,  the  precipitate  removed,  washed,  and 
dissolved  in  ammoniacal  water.  This  solution  was  then 
shaken  with  amyl  alcohol,  but  the  alcohol  was  not  colored 
and  the  residue  obtained  on  concentration  was  so  slight  as 
to  be  practically  nothing.  Another  sample  was  precipitated 
with  phosphotungstic  acid,  the  solution  being  acid  in 
reaction.  The  precipitate  was  allowed  to  settle,  removed 
by  filtration,  washed  with  acidulated  water,  decomposed 
with  a  saturated  solution  of  barium  hydrate,  and  the 
remaining  insoluble  part  filtered  out.  So  far  as  possible, 
the  barium  was  removed  from  the  filtrate  with  carbon 
dioxide,  alternating  with  concentration,  and  further  addi- 
tion of  carbon  dioxide.  The  solution  was  then  allowed  to 
concentrate  to  dryness,  when  the  residue  was  dissolved  in 
absolute  alcohol,  leaving  barium  salts  behind.  On  con- 
centrating the  slightly  opalescent  solution,  more  barium 
salts  came  down  during  the  process  and  were  filtered  out. 
The  dry  residue  was  taken  up  in  water  and  ammonium 
carbonate  used  to  precipitate  the  barium  that  still  remained. 
After  removing  the  barium  carbonate  by  evaporating  on 
the  water-bath,  both  carbon  dioxide  and  ammonia  were 
expelled,  the  solution  again  becoming  acid.  Dryness  being 
reached,  absolute  alcohol  was  once  more  used,  leaving 
undissolved  a  small  amount  of  inorganic  material.  In  this 
way  the  final  residue  after  evaporation  of  the  alcohol  was 
practically  freed  from  inorganic  impurities.  Sulphuric 
acid  no  longer  gave  a  barium  precipitate  in  water  solution. 
The  amount  obtained  by  this  method  was  very  small  and 
an  exceedingly  small  part  of  the  original  toxic  powder. 
Since  the  substance  obtained  in  this  way  still  gave  good 
Millon's,  biuret  and  xanthoproteic  reactions,  it  is  fair  to 
say  that  it  was  not  alkaloidal.  The  very  small  amount 
obtained  by  this  method  given  to  a  guinea-pig  intra- 
abdominally  made  the  animal  sick,  but  did  not  kill. 
Either  phosphotungstic  acid  does  not  precipitate  the  toxic 
body  or  else  the  amount  obtained  was  less  than  a  fatal 
dose. 


Should  the  poison  consist  of  an  alkaloidal  body  existing 
as  a  salt  in  the  acid  solution,  the  possibility  of  extracting 
the  base  with  ether  or  chloroform,  after  the  solution  had 
been  made  alkaline  with  ammonia,  is  apparent.  This 
was  tried  with  negative  results.  To  a  water  solution  of 
colon  poison,  acid  in  reaction,  ammonia  was  added,  drop 
by  drop,  to  a  slightly  alkaline  reaction,  the  mixture  shaken 
with  ether,  the  ether  separated  and  evaporated.  The  residue 
remaining  was  non-toxic.  The  ammoniacal  water  solution 
was  next  shaken  with  chloroform,  the  slightly  colored 
chloroform  drawn  off  and  evaporated  at  low  temperature, 
leaving  a  small  amount  of  a  dark,  thick,  semiliquid,  which 
was  not  poisonous  either  as  it  was  or  after  faintly  acidify- 
ing with  hydrochloric  acid.  The  water  solution  remaining 
being  still  poisonous,  it  is  evident  that  the  toxic  part  is  not 
an  alkaloidal  body  capable  of  being  extracted  directly. 

Potassium  bismuth  iodide  in  acid  solution  of  the  crude 
soluble  poison  produces  an  abundant  precipitate,  apparently 
more  or  less  soluble  in  excess,  and  soluble  in  ammoniacal 
water. 

Kowalewsky  has  shown  that  uranyl  acetate  will  com- 
pletely remove  from  various  albuminous  fluids  every  trace 
of  protein  giving  a  biuret  reaction,  while  Jacoby  and  others 
have  used  this  reagent  for  the  removal  of  proteins  from 
faintly  alkaline  solutions.  Abel  and  Ford  used  it  to  remove 
protein  from  an  extract  of  poisonous  fungi.  In  a  slightly 
alkaline  solution  of  albumin  poison,  uranium  acetate  gave 
an  abundant  precipitate,  but  not  a  complete  separation, 
as  both  precipitate  and  filtrate  still  gave  the  Millon  and 
biuret  tests,  and  the  filtrate,  after  removal  of  excess  of 
uranium  with  a  solution  of  di-sodium  hydrogen  phosphate, 
filtration,  evaporation,  solution  in  alcohol,  and  reevapora- 
tion,  was  still  poisonous.  In  acid  solution,  the  precipitation 
was  complete,  the  filtrate  no  longer  giving  the  protein 
reactions. 

Freshly  prepared  metaphosphoric  acid  also  produced 
an  abundant  precipitate,  but  not  a  complete  separation, 
the  filtrate  showing  both  Millon  and  biuret  reactions. 


110  PROTEIN  POISONS 

Likewise  a  heavy  precipitate  is  produced  by  the  use  of  a 
saturated  solution  of  picric  acid,  but  the  poison  is  not  in  the 
precipitate,  which  gives  only  a  very  poor  Millon  test  after 
removal  of  the  picric  acid,  and  no  biuret.  Hofmeister  has 
given  a  method  for  introducing  iodine  into  the  molecule 
of  egg  albumen.  This  was  tried  with  the  poison  split  from 
egg  albumen.  The  iodized  compound  no  longer  gave  either 
the  Millon  or  biuret  reactions,  and  while  it  affected  animals 
more  or  less,  they  did  not  die,  and  the  symptoms  were  not 
those  induced  by  toxin  poisoning.  The  iodine  seemed  to 
have  entered  into  chemical  combination  in  the  poison 
molecule,  and  to  have  thus  changed  its  characteristics. 
The  iodized  body  was  freely  soluble  in  absolute  alcohol, 
and  in  alkaline  water,  not  in  water  alone,  and  was  precipi- 
tated by  acid  water  from  alcoholic  solution,  also  on  acidi- 
fying an  alkaline  water  solution.  Though  it  no  longer 
responded  to  the  Millon  and  biuret  reactions,  a  good  test 
for  nitrogen  was  obtained  after  fusing  with  metallic  sodium. 
An  attempt  was  made  to  benzoylate  the  poison  by  the 
Schotten-Baumann  method,  using  albumin  poison.  Prac- 
tically no  precipitate  was  obtained.  From  the  filtrate  in  a 
part  soluble  in  hot  alcohol  there  were  obtained  shiny,  glis- 
tening plates  or  flat  needles  which  matted  together  under 
suction,  and  had  much  the  appearance  of  some  of  the  fatty 
acids.  These  were  insoluble  in  water  or  very  difficultly 
so,  if  at  all,  difficultly  soluble  in  cold  alcohol,  readily  in  hot. 
They  gave  no  Millon  test,  no  biuret,  no  Molisch,  and  con- 
tained no  nitrogen.  After  recrystallization  from  alcohol 
they  melted  constantly  at  62°.  Palmitic  acid  melts  at  62° 
and  boils  at  339°  to  350°  (Mulliken).  A  Merck  preparation 
of  palmitic  acid  melted  at  62°  and  boiled  at  about  345°  to 
350°.  Our  crystals  had  not  yet  boiled  at  360°,  though 
above  300°  there  was  some  decomposition.  From  the 
remainder  of  the  filtrate  there  was  obtained  from  the  part 
soluble  in  cold  alcohol  a  non-crystallizable  body,  giving 
both  Millon  and  biuret  tests  and  containing  9.335  per  cent, 
of  nitrogen,  and  from  the  part  soluble  only  in  water,  likewise 
a  non-crystalline  compound,  with  9.66  per  cent,  nitrogen, 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     111 

and  showing  both  Millon  and  biuret  tests,  but  not  seriously 
affecting  animals  in  usual  doses. 

The  nitrogen  in  a  number  of  the  crude  poisons  has  been 
determined  by  Gidley  in  this  laboratory  as  follows: 

PERCENTAGE  OF  NITROGEN  IN  PROTEIN  POISONS. 

Per  cent,  of  N. 

Source  of  poison.                                                         in  crude  poison. 

Colon  bacillus 13.49 

Typhoid  bacillus 11.52 

Tubercle  bacillus 11.00 

Pyocyaneus 10.50 

Ruber  of  Kiel 10.495 

Subtilis 8.12 

Megaterium 8.595 

Proteus  vulgaris 10.17 

Yellow  sarcine 6.145 

Egg  albumen  (Leach) 13 . 74 

Serum  albumin 10.48 

Edestin 12.78 

Zein 10.69 

Witte  peptone 11.14 

De  Chapoteaut  peptone        ........  12.735 

To  study  the  distribution  of  the  nitrogen,  determinations 
were  made  in  both  the  colon  and  albumin  poisons,  of  the 
ammonia  nitrogen,  the  mono-amino,  and  diamino  nitrogen, 
by  the  method  already  described  under  cleavage  with 
dilute  mineral  acids.  The  following  are  the  results: 

DISTRIBUTION  OF  NITROGEN  IN  PROTEIN  POISONS. 

Total  N  Mono- 
Source  of               Total           of  acid  Ammonia  amino  Diamino 

poison.                 poison.          extract.              N.                  N.  N. 

Colon  bacillus  .      13.49%       10.185%  1.525%  6.472%  1.753% 

Egg  albumen    .      13.74%       11.477%  0.745%  7.999%  1.400% 

It  will  be  seen  that  the  greater  part  of  the  nitrogen  is  to 
be  found  in  mono-amino  combination.  From  the  phospho- 
tungstic  filtrates,  from  both  the  albumin  and  colon  poisons, 
containing  the  mono-amino  acids,  crystalline  bodies  were 
obtained.  Judged  by  the  strong  Millon  test,  tyrosin  was 


112  PROTEIN  POISONS 

undoubtedly  present,  but  the  crystalline  masses  were 
largely  leucin,  and  no  tyrosin  was  obtained  in  purified 
form.  From  the  crude  crystals,  after  many  and  repeated 
crystallizations,  what  was  thought  to  be  leucin  was  obtained 
pure,  melting  at  264°  to  265°  uncorrected,  or  269.42°  to 
270.46°  corrected.  The  crystals  were  thin  plates  charac- 
teristically grouped,  and  sublimed  readily.  From  another 
5  per  cent,  sulphuric  acid  extract  of  albumin  poison  was 
obtained  a  large  mass  of  crystals  in  characteristic  tyrosin- 
like  sheaves,  and  giving  a  deep  Millon  reaction.  These 
were  undoubtedly  tyrosin,  though  at  the  time  no  melting- 
point  was  taken. 

Properties  of  the  Haptophor  or  Non-poisonous  Group. — Leach1 
has  investigated  this  split  product  with  the  following 
general  results:  After  cleavage  of  the  protein  with  alkaline 
alcohol,  the  haptophor  remains  undissolved.  It  is  collected 
on  a  filter,  then  transferred  to  Soxhlets,  and  for  some  days 
extracted  with  95  per  cent,  alcohol.  This  is  for  the  purpose 
of  removing  as  thoroughly  as  possible  the  alkali  which  it 
has  absorbed  from  the  alkaline  alcohol.  This  cannot, 
however,  be  wholly  washed  out  by  this  method,  and  it  is 
possible  that  in  part  it  is  held  chemically.  After  this 
extraction  the  substance  is  easily  reduced  to  a  fine  brownish 
powder.  On  burning  it  puffs  up,  gives  off  the  characteristic 
odor  of  nitrogenous  compounds,  and  leaves  a  copious  ash 
containing  phosphate.  The  solubility  of  the  haptophors 
from  different  proteins  differs  widely;  that  from  egg-white 
is  wholly  soluble  in  water,  while  that  from  the  cellular 
substance  of  the  tubercle  bacillus  is  only  sparingly  soluble. 
However,  it  is  only  the  part  soluble  in  water  from  any  of 
these  haptophors  that  is  of  special  interest.  The  studies  of 
Leach,  referred  to,  were  made  with  the  non-poisonous 
portion  of  the  colon  bacillus.  This  is  mainly  soluble  in 
water,  giving  an  opalescent  solution  from  which  a  light- 
colored  sediment  is  deposited  on  standing,  leaving  a  clear, 
golden  brown  solution.  The  sediment  is  not  soluble  in 

1  Jour.  Biolog.  Chem.,  1907,  iii,  443. 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI      113 

either  dilute  alkali  or  acid  in  the  cold,  but  is  soluble  in 
alkali  on  boiling.  The  clear,  aqueous  solution  of  the  hapto- 
phor  is  alkaline  from  sodium  hydrate  held  either  mechanic- 
ally or  chemically;  it  is  precipitated  by  mineral  acids  and 
by  alcohol.  It  responds  to  the  biuret,  xanthoproteic, 
Millon,  and  Adamkiewicz  tests.  Millon's  test  is  not  very 
satisfactory,  and  in  some  samples  has  failed  altogether, 
even  after  care  has  been  exercised  in  neutralizing  the  alkali. 
It  is  quite  evident  that  the  substance  or  substances  in  the 
protein  molecule  to  which  the  Millon  test  is  due  are  for 
the  most  part  found  in  the  toxophor  group.  However, 
the  readiness  of  response  to  this  test  varies  greatly  in  the 
different  haptophors.  The  haptophor  substance  does  not 
reduce  Fehling's  solution  directly,  but  does  so  readily  and 
abundantly  after  prolonged  boiling  with  dilute  hydro- 
chloric acid.  The  presence  of  carbohydrate  in  the  hapto- 
phor has  already  been  discussed  (page  70).  Tests  with 
a-naphthol,  phloroglucin,  and  orcin  give  positive  results. 
Ammonium  molybdate  gives  an  organic  precipitate,  but 
no  evidence  of  free  phosphoric  acid.  The  preliminary  tests 
show  the  presence  of  protein,  nucleic,  and  carbohydrate 
groups.  Comparing  these  results  with  those  obtained  in 
the  study  of  the  toxophor,  the  following  statements  may 
be  formulated:  (1)  The  toxophor  is  freely  soluble  in  abso- 
lute alcohol,  the  haptophor  is  insoluble  in  this  menstruum. 
(2)  The  toxophor  contains  no  carbohydrate,  all  of  which  is 
found  in  the  haptophor.  (3)  The  toxophor  freely  responds 
to  the  Millon  test,  while  the  haptophor  does  so  slightly  and 
in  some  instances  not  at  all.  (4)  The  toxophor  contains  no 
phosphorus,  or  but  little  of  this  element,  while  the  hapto- 
phor is  rich  in  phosphorus.  (5)  The  toxophor  from  different 
proteins  seems  to  be  the  same,  possibly  with  unrecognizable 
differences  in  chemical  structure,  while  the  haptophor  of 
each  protein  differs  from  that  from  all  other  proteins. 

Leach1  gives  the  following  table  showing  the  percentages 
of  ash,  nitrogen,  and  phosphorus  in  the  haptophor  of  the 
colon  bacillus: 

1  Loc.  cit. 


ash. 

ash. 

N. 

P. 

free. 

free. 

N:P. 

8.61 

10.65 

2.87 

26.08 

5.56 

2.34 

7.52 

3.99 

2.38 

20.36 

15.66 

6.76 

3.61 

8.02 

4.28 

1.87 

30.74 

4.87 

1.50 

7.03 

2.16 

3.25 

15.05 

8.48 

4.95 

2.25 

5.41 

2.46 

2.20 

1.66 

4.65 

1.74 

4.73 

1.77 

2.67 

5.50 

1.36 

3.43 

1.35 

3.48 

1.37 

2.53 

14.00 

27.67 

5.98 

2.68 

8.27 

3.71 

2.23 

2.08 

5.50 

1.79 

5.62 

1.83 

3.07 

11.71 

3.74 

3.16 

2.47 

3.28 

2.70 

1.28 

3.47 

5.35 

1.58 

5.55 

1.64 

3.39 

114  PROTEIN  POISONS 

Fixed     Inorganic  N  ash      P  ash     Ratio 

Ash. 

Cell  substance  . 

Haptophor     .  .  33.25 

Prep.  A.  .      .  .  26.76 

Prep.  B.   .      .  .  35.34 

Prep.  D.  .      .  .  15.38 

Prep.  G.  .      .  .  6.99 

Prep.  G.  pur.  .  5.50 

Prep.  H.  .      .  .  35.91 

Prep.  K2  .      .  .  7.57 

Prep.  M.        .  .  11.71 

Prep.  Ma        .  .  8.30 

Explanation  of  the  Table. — Ash,  residue  heating  at  low 
redness.  Fixed  ash,  residue  after  heating  to  full  heat  of 
powerful  burner.  Inorganic  ash,  ash  less  calculated  amount 
of  PO4.  N,  nitrogen  by  Kjeldahl-Groening  method.  P, 
phosphorus  by  the  Neumann  method.  N  and  P  ash  free, 
reckoned  free  from  "inorganic  ash."  N:P,  quotient  of 
column  4  divided  by  column  5.  A,  portion  of  haptophor 
dissolved  by  acid  alcohol.  B,  portion  of  haptophor  not 
dissolved  by  acid  alcohol.  D,  substance  precipitated  by 
acid  alcohol  from  solution  of  B  in  aqueous  alkali.  G, 
substance  precipitated  by  acid  alcohol  from  aqueous  solu- 
tion of  haptophor.  H,  obtained  by  concentration  of  the 
alcoholic  filtrate  from  G.  K,  substance  precipitated  by 
dilute  acetic  acid  from  aqueous  solution  of  haptophor.  K2, 
same  as  K,  except  that  strong  acid  was  used.  M,  precipi- 
tated by  alcohol  from  filtrate  from  K.  M2,  precipitated 
from  filtrate  from  K2. 

Leach  states:  "As  these  preparations  are  all  mixtures, 
the  absolute  values  are  worth  nothing  taken  singly,  but 
the  comparative  values,  especially  the  ratio  of  N  to  P,  as 
given  in  the  last  column,  are  of  interest.  The  determina- 
tions were  made  for  the  sake  of  tracing  the  nucleo  com- 
pounds. There  are  many  indications  of  nucleic  acid,  but 
the  amount  of  both  nitrogen  and  phosphorus  is  much  too 
small.  The  ratio  between  them  is,  however,  quite  within 
the  range  for  nucleic  acids  from  other  sources,  as  may  be 
seen  by  comparison  in  the  following  table.  Moreover,  the 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     115 


nucleic  acid  and  the  nucleates  are  the  only  nucleo  com- 
pounds in  which  the  ratios  are  at  all  comparable  with 
those  given  in  the  preceding  table.  Nuclein  contains  a 
little  less  phosphorus  than  any  of  these  preparations  from 
the  germ,  while  other  nucleo  compounds  are  much  richer 
in  nitrogen  and  poorer  in  phosphorus.  It  is  perhaps  worthy 
of  mention  that  contact  with  mineral  acid  apparently 
breaks  up  the  nucleic  acid,  the  phosphoric  acid  going  into 
solution;  thus,  preparation  A  gives  evidence  of  phosphorus 
in  inorganic  combination,  while  G  does  not." 


Substance. 

Source. 

Observer. 

N. 

P. 

N:P. 

Nucleic  acid 

Salmon  sperm 

Miescher 

15.24 

9.62 

1.58 

Nucleic  acid 

Sea  urchin  sperm 

Mathews 

15.34 

9.59 

1.60 

Nucleic  acid 

Yeast 

Miescher 

16.03 

9.04 

1.77 

Nucleic  acid 

Pancreas 

Bang 

18.20 

7.67 

2.37 

Nucleic  acid 

Thymus 

Kostytschew 

15.55 

9.25 

1.69 

Nucleic  acid 

Thymus 

Kostytschew 

15.26 

9.30 

1.65 

Nucleic  acid 

Wheat  embryo 

Osborne  and 

15.88 

8.70 

1.83 

Harris 

Inosinic  acid 

Muscle 

Haiser 

16.00 

8.60 

1.86 

Clupein  nucleate 



Mathews 

21.06 

6.07 

3.48 

Nucleohiston 

Thymus 

Huiskamp 

18.37 

3.70 

4.97 

Nucleoprotein 

Thymus 

Huiskamp 

16.42 

0.95 

17.30 

Nucleoprotein 

Pancreas 

Umber 

17.82 

1.67 

10.65 

Nuclein 

Pancreas 

Umber 

17.39 

4.48 

3.88 

Ba  a-nucleate 

Thymus 

Kostytschew 

12.83 

7.63 

1.68 

Ba  /8-nucleate 

Thymus 

Kostytschew 

10.16 

8.48 

1.38 

In  a  later  paper,  Leach1  has  made  a  study  of  the  hapto- 
phor  of  egg-white.  The  percentages  of  ash,  nitrogen,  phos- 
phorus, and  sulphur  in  egg-white  and  its  split  products 
are  given  and  compared  with  the  cellular  substance  of  the 
colon  bacillus  in  the  following  table: 


Egg-white    . 
Toxophor 
Haptophor   . 

Cell  substance 
Toxophor 
Haptophor  . 


Ash. 

2.48 

1.14 

13.57 

8.61 

2.33 

33.25 


Inorganic 
ash. 
2.066 


12.80 


26.08 


N. 

14.48 
13.74 
12.67 


P. 

0,135 
Trace 
0.253 


10.65      2.870 

11.15       Trace 

5.56       2.340 


N  ash  P  ash  S  ash 

S.           free.  free.  free. 

2.66       14.70  0.138  2.73 

2.19       13.90           ..  2.22 

2.79       14.53  0.290  3.20 


7.52       3.990 


1  Jour.  Biol.  Chem.,  1908,  v,  253. 


116  PROTEIN  POISONS 

Leach  split  up  edestin,  casein,  egg-white,  and  colon 
cellular  substance  with  alkaline  alcohol.  The  insoluble 
part  of  each  gave  the  various  protein  color  tests,  Millon's 
reaction  less  satisfactorily  than  the  others.  On  stirring 
with  water,  the  edestin  preparation  was  entirely  soluble, 
there  was  a  slight  flocculence  with  the  casein  preparation, 
the  others  were  mainly  but  not  wholly  soluble.  Addition  of 
a  little  sodium  hydroxide  increases  the  solubility.  Mineral 
acids  give  precipitates  with  the  casein  and  egg  preparations. 

The  most  marked  difference  was  found  on  testing  for 
carbohydrates.  As  edestin  contains  no  carbohydrate,  its 
preparation  showed  no  evidence  of  such  a  group.  Although 
casein  is  said  to  contain  no  carbohydrate,  it  has  been  found 
to  respond  to  the  Molisch  test,  and  so  does  its  haptophor. 
As  was  to  be  expected,  the  egg  preparation  gives  evidence 
of  hexose  and  not  pentose.  The  lead  sulphide  reaction 
shows  the  presence  of  loosely  combined  sulphur  in  the 
preparations  from  egg  and  edestin,  not  in  the  ones  from 
casein  and  the  colon  bacillus. 

Samples  of  the  haptophor  of  egg-white  were  stirred  with 
water,  filtered,  and  attempts  made  to  separate  protein  and 
carbohydrate  in  the  filtrate  by  means  of  uranium  acetate. 
The  acetate  was  added  both  with  and  without  sufficient 
alkali  to  keep  the  solution  alkaline.  A  copious  precipitate 
resulted  in  both  cases  and  this  was  filtered  out  with  some 
difficulty.  The  slight  excess  of  uranium  was  removed 
from  the  filtrate  by  the  addition  of  sodium  phosphate. 
The  filtrate  gave  evidence  of  carbohydrate,  but  the  separa- 
tion was  not  sufficiently  sharp,  and  that  method  was 
abandoned.  Acidifying  until  there  was  a  slight  permanent 
precipitate,  the  addition  of  either  ethyl  or  methyl  alcohol 
cleared  the  solution.  Phosphotungstic  acid  precipitated 
both  protein  and  carbohydrate.  In  short,  no  method  was 
found  that  would  remove  the  protein  from  the  solution  and 
leave  the  carbohydrate.  It  is  perhaps  a  legitimate  infer- 
ence that  the  combination  of  the  two  is  a  chemical  one. 

Samples  were  subjected  to  hydrolysis  and  titrated  with 
Fehling's  solution.  The  proteins  and  possibly  other  bodies 


THE  CLEAVAGE  OF  PROTEINS  WITH  ALKALI     117 

present  interfered  with  the  reaction,  but  by  adding  the 
solution  all  or  nearly  all  at  once  it  was  possible  to  obtain 
comparative  results.  Experiments  with  the  haptophor  of 
the  colon  bacillus  had  shown  that  the  maximum  reduction 
was  obtained  by  boiling  for  two  and  one-half  hours  with 
2.5  per  cent,  hydrochloric  acid  (see  p.  70). 

Three  grams  of  the  haptophor  of  egg-white  was  mixed 
with  200  c.c.  of  water,  and  20  c.c.  of  25  per  cent,  hydro- 
chloric acid.  A  second  sample  was  prepared  in  the  same 
way  except  that  it  was  filtered  before  adding  the  acid. 
Both  were  boiled  with  reflux  condenser.  After  boiling  half 
an  hour  and  then  at  intervals  of  three  hours,  aliquot  parts 
were  removed,  neutralized,  titrated  with  Fehling's  solution, 
and  the  amount  of  reducing  substance  calculated.  Other 
samples  were  hydrolyzed  with  sulphuric  acid,  with  less 
satisfactory  results.  These  preliminary  experiments  indi- 
cated that  the  reducing  substance  is  all  present  in  the 
portion  soluble  in  water,  and  that  the  maximum  yield, 
which  if  calculated  as  dextrose,  is  about  9  per  cent.,  is 
obtained  by  boiling  from  ten  to  twelve  hours,  and  until  the 
mixture  no  longer  gives  the  biuret  test. 

Accordingly,  25  grams  of  the  egg-white  haptophor  was 
shaken  for  two  hours  on  a  shaker  with  ten  times  its  weight 
of  water,  filtered,  200  c.c.  more  of  water  added,  the  solution 
neutralized  with  hydrochloric  acid,  then  50  c.c.  of  25  per 
cent,  hydrochloric  acid  added,  thus  making  approximately 
a  5  per  cent,  solution  of  material  in  2.5  per  cent.  acid.  This 
was  boiled  with  a  reflux  condenser  for  ten  or  twelve  hours, 
until  the  solution  no  longer  gave  the  biuret  test.  It  was 
then  filtered,  leaving  very  little  on  the  filter.  The  clear, 
red-brown  filtrate  was  cooled,  neutralized  with  sodium 
hydroxide,  and  benzolated  by  the  Schotten-Baumann 
method.  The  mixture  became  very  warm,  but  was  cooled 
by  surrounding  the  flasks  with  pounded  ice  and  salt.  When 
the  reaction  ceased,  the  compound  settled  nicely,  and  was 
filtered  by  suction  after  standing  two  or  three  hours.  The 
precipitate  was  washed  with  water  containing  a  little 
ammonia,  and  treated  with  boiling  water,  in  which  a  large 


118  PROTEIN  POISONS 

portion  was  freely  soluble.  On  cooling  and  concentrating 
the  alcoholic  solution,  a  fine  yield  of  crystals  was  obtained. 
The  crystals  from  several  samples  were  united  and  recrys- 
tallized  from  hot  absolute  alcohol  until  the  solution  was 
clear  and  colorless.  Macroscopic  bundles  of  needles  were 
thus  obtained,  showing  very  characteristic  grouping.  They 
were  washed  in  alcohol  and  in  ether,  dried  upon  porous 
plates,  the  operations  being  repeated  until  samples  from 
two  recrystallizations  melted  side  by  side  within  1°  or  1.5°. 
The  crystals  are  pure  white,  readily  soluble  in  benzol, 
chloroform,  and  in  glacial  acetic  acid  as  well  as  in  alcohol, 
and  melt  at  203°.  When  boiled  with  sodium  hydroxide, 
ammonia  is  given  off;  after  removing  benzoic  acid  by  boil- 
ing with  hydrochloric  acid,  the  resulting  product  reduces 
Fehling's  solution. 

0.4150  gram  gave  0.00891  gram  N,  corresponding  to  2.14  per  cent.  N. 
0.4220  gram  gave  0.00962  gram  N,  corresponding  to  2.279  per  cent.  N. 
Average  is  2.213  per  cent.  N. 

These  characteristics  suffice  to  identify  the  compound 
as  glucosamin  benzoate  which  Pumm  reports  as  melting 
at  203°.  Kueny  prepared  different  benzoates  of  glucosamin 
by  varying  the  conditions  of  the  experiment.  The  one  most 
readily  formed  was  the  tetrabenzoate,  melting  at  199° 
when  recrystallized  from  alcohol,  and  at  207°  when  re- 
crystallized  from  glacial  acetic  acid.  He  tried  by  various 
methods  to  prepare  a  pentabenzoate,  but  without  success. 
Langstein  prepared  glucosamin  benzoate  from  egg-white, 
which,  after  once  recrystallizing  from  hot  alcohol,  melted 
at  201°  to  202°,  and  gave  1.95  per  cent,  of  nitrogen.  The 
theoretical  amount  of  nitrogen  in  the  tetrabenzoate  is 
2.35  per  cent.  Thus,  the  benzoate  prepared  from  the 
haptophor  of  egg-white  agrees  with  glucosamin  benzoate 
prepared  from  glucosamin  and  from  egg-white,  at  least  as 
well  as  those  preparations  do  with  each  other.  Numerous 
observers  have  found  glucosamin  in  egg-white,  and  this 
work  shows  that  it  remains  in  the  haptophor  when  egg-white 
is  disrupted  by  alkaline  alcohol. 


CHAPTER  VI 
ACTION  ON  ANIMALS1 

IT  will  be  interesting  and  instructive  to  compare  the 
effects  of  the  living  bacillus,  the  dead  cellular  substance, 
and  the  soluble  poison  on  animals.  • 

The  Action  of  the  Living  Bacillus. — When  a  guinea-pig  is 
inoculated  with  a  fatal  dose  of  the  living  colon  germ,  prac- 
tically no  symptoms  whatever  are  noticeable  for  a  period 
varying  from  five  to  twelve  hours,  according  to  the  size 
of  the  dose  given.  This  may  be  considered  as  the  period 
of  incubation  and  is  roughly  proportional  to  the  amount  of 
living  germ  injected.  We  have  always  worked  with  a 
bacillus  1  c.c.  of  a  twelve-hour  or  older,  bouillon  culture  of 
which  has  invariably  proved  fatal  to  guinea-pigs  within 
twenty-four  hours.  If  1  c.c.  of  such  a  culture  is  given,  no 
effects  will  be  seen  for  a  period  of  from  ten  to  twelve  hours. 
If,  on  the  other  hand,  2  c.c.  of  the  same  culture  be  injected, 
the  animal  will  begin  to  manifest  symptoms  of  illness  in 
from  eight  to  ten  hours,  and  if  larger  doses  are  given  the 
symptoms  will  become  apparent  in  a  shorter  time.  This 
period  of  incubation  undoubtedly  represents  the  time 
taken  for  the  bacillus  to  multiply  and  to  be  destroyed  to 
such  an  extent  that  sufficient  poison  may  be  liberated 
through  its  disintegration  to  produce  noticeable  toxic 
effects  in  the  animal.  This  period  of  incubation  is,  therefore, 
in  reality  the  crisis  of  the  disease  and  the  outcome  depends 
solely  on  whether  all  bacteria  have  been  destroyed  before 
a  lethal  dose  of  the  poison  has  been  set  free  or  not.  It  is 

1  This  chapter  is  a  reproduction,  without  material  change,  of  an  article 
by  Victor  C.  Vaughan,  Jr.,  published  in  the  Jour.  Amer.  Med.  Assoc.  in 
1905. 


120 


PROTEIN  POISONS 


during  this  period  that  individual  resistance  and  acquired 
immunity  are  important  factors  acting  by  causing  increased 
bacteriolysis  and  the  destruction  of  all  bacilli  before  a 
fatal  dose  of  poison  has  been  set  free.  During  this  time 
the  temperature  of  the  animal  may  rise  to  a  greater  or  less 
extent  or  may  remain  stationary;  the  animal  remains 
active,  eats;  its  coat  is  not  roughened  and  it  appears  in 


FIG.  5 


102 
101 

iooc 


\ 


Temperature  curve  of  guinea-pig  after  inoculation  with  1  c.c.,  sixteen- 
hour  bouillon  culture  of  the  colon  bacillus.  Death  occurred  twenty  hours 
after  inoculation. 

all  respects  as  well  as  a  normal  animal.  At  the  end  of  this 
period,  however,  the  appearance  changes.  The  animal 
becomes  less  active.  It  remains  in  one  corner  of  its  cage; 
its  coat  becomes  roughened;  it  hangs  its  head  and  apparently 
enters  into  a  state  of  stupor.  At  the  same  time  the  rectal 
temperature  begins  to  fall  abruptly,  as  can  be  seen  from 
a  study  of  Fig.  5. 


ACTION  ON  ANIMALS  121 

Indeed,  this  fall  of  body  temperature  is  often  the  first 
marked  symptom  and,  when  occurring  to  a  marked  degree, 
it  is  invariably  a  bad  omen.  The  body  temperature  will 
often  fall  from  101°  to  94°  F.  or  even  lower  within  from  two 
to  four  hours,  and  this  fall  is  progressive  and  continuous 
until  the  animal's  death,  immediately  preceding  which  a 
temperature  as  low  as  87°  or  86°  F.  is  not  uncommon.  At 
the  same  time  the  animal  shows  signs  of  the  most  marked 
peritoneal  inflammation,  as  is  evidenced  by  rigidity  and 
spasm  of  the  abdominal  muscles  on  pressure.  At  autopsy, 
the  only  gross  lesion  present  is  a  marked  hemorrhagic 
peritonitis  with  a  large  amount  of  bloody  fluid  containing 
intact  red  corpuscles  and  leukocytes  in  the  peritoneal 
cavity.  The  parietal  and  visceral  peritoneum  are  studded 
with  minute  punctiform  hemorrhages.  Hemorrhage  is  an 
especially  prominent  feature  in  the  great  omentum  and  is 
present  to  a  less  marked  degree  in  the  mesentery. 

The  Action  of  the  Cellular  Substance. — The  dead  bacterial 
substance  used  in  the  following  work  was  obtained  by 
growing  a  large  amount  of  the  colon  germ  on  tanks  filled 
with  agar  for  a  period  of  two  weeks  at  room  temperature. 
At  the  end  of  this  time  the  growth  was  removed  from  the 
tanks  and  extracted  with  absolute  alcohol  and  ether.  The 
crude  bacterial  substance  thus  obtained  was  reduced  to  a 
fine  powder  by  pulverization  in  an  agate  mortar,  and  was 
then  ready  for  use. 

It  is  interesting  to  note  that  the  person  who  did  the 
pulverizing  was  often  quite  seriously  poisoned  during  the 
process  unless  he  took  the  precaution  of  wearing  a  mask 
which  hindered  the  inhalation  of  the  powder.  The  symp- 
toms of  such  poisoning  were  exceedingly  interesting.  The 
first  thing  noticed  was  a  marked  irritation  of  the  nasal 
mucous  membrane  and  a  huskiness  of  the  voice,  due  no 
doubt  to  the  mechanical  irritation  of  the  inhaled  powder. 
This  was  followed  by  a  feeling  of  depression  and  malaise, 
and  chilly  sensations.  Occasionally  a  decided  chill  would 
be  experienced.  It  is  unfortunate  that  no  accurate  obser- 
vations of  temperature  were  taken  in  these  cases.  Nausea 


122  PROTEIN  POISONS 

and  even  vomiting  were  occasionally  noted.  After  a  period 
of  discomfort  varying  from  six  to  ten  hours,  during  which 
the  patient  often  complained  of  dull  pain  in  the  various 
joints,  recovery  would  rapidly  and  completely  take  place. 

On  examining  the  powder  obtained  in  this  manner 
microscopically  we  found  that  it  consisted  of  colon  bacilli 
which  still  retained  their  morphological  characteristics  and 
could  still  be  stained  by  aniline  dyes.  On  the  other  hand 
cultures  made  from  this  powder  have,  of  course,  never 
given  a  growth.  In  other  words,  the  bacillus  has  not  been 
broken  up  by  this  treatment,  but  simply  has  been  deprived 
of  life  and  of  the  power  of  reproduction.  It  is  worthy  of 
note  that  neither  by  the  action  of  alcohol,  ether,  physiological 
salt  solution,  distilled  water,  nor  any  simple  solvent  have 
we  been  able  to  extract  a  poison  from  the  colon  bacillus. 
Nor,  again,  can  a  poison  be  split  off  by  the  action  of  heat 
even  when  the  germ  substance  is  heated  to  184°  C.  in  a 
sealed  tube  for  thirty  minutes.  It  is  only  when  we  make 
use  of  agents  which  will  chemically  break  up  the  colon 
bacillus  that  we  are  enabled  to  obtain  a  poison  apart  from 
the  rest  of  the  cellular  substance.  The  powdered  bacterial 
substance  is  not  soluble,  but  can  be  held  in  suspension  in 
normal  salt  solution  and,  since  it  can  be  boiled  without 
appreciably  affecting  its  toxicity,  suspensions  were  always 
heated  to  100°  for  fifteen  minutes  before  injection  in  order 
to  insure  sterility. 

This  coarsely  powdered  cellular  substance  killed  guinea- 
pigs  when  injected  intraperitoneally  in  doses  of  1  to  40,000, 
body  weight,  and  invariably  proved  fatal  within  twelve 
hours,  usually  causing  death  at  the  end  of  from  six  to  eight 
hours.  On  the  injection  of  a  fatal  dose  of  the  cellular 
substance  intraperitoneally,  we  noticed  that  the  most 
marked  change  was  in  the  length  of  the  period  of  incu- 
bation. Thus,  whereas  in  the  case  of  the  living  germ  from 
eight  to  twelve  hours  passed  before  noticeable  symptoms 
appeared,  in  that  of  the  dead  germ  substance  the  animal 
almost  invariably  showed  symptoms  of  illness  at  the  end 
of  four  hours.  In  regard  to  the  character  of  these  symptoms 


ACTION  ON  ANIMALS  123 

it  may  be  stated  that  they  are  similar  in  all  respects  to 
those  induced  by  the  living  bacillus.  The  temperature 
remains  the  same  or  may  rise  slightly  during  the  first  two 
hours.  At  the  end  of  the  four  hours  it  has  begun  to  fall, 
and  there  is  a  decided  drop  from  then  on  until  the  time  of 
death,  provided  the  dose  given  is  a  fatal  one.  If  a  non-fatal 
dose  has  been  injected  intraperitoneally  the  temperature, 
as  will  be  seen  from  Fig.  6,  has  reached  a  minimum  at  the 
end  of  from  six  to  eight  hours  and  has  returned  to  normal 
again  in  from  twelve  to  twenty  hours. 

Moreover,  as  a  general  rule,  it  may  be  stated  that  the 
fall  in  non-fatal  cases  seems  to  be  directly  proportional  to 
the  amount  of  bacterial  substance  injected.  That  this 
.should  be  the  case  seems  to  be  olily  natural  when  we  con- 
sider the  fact  that  in  this  instance  we  have  largely  done 
away  with  that  factor  which  is  known  as  the  individual 
resistance  of  the  animal.  As  has  been  previously  mentioned 
in  the  case  of  the  living  bacillus,  the  individual  resistance 
plays  an  important  part  in  determining  the  amount  of 
poison  which  will  ultimately  be  set  free  in  the  body.  For 
example,  whereas  1  c.c.  of  a  twelve-hour  culture  of  our 
colon  bacillus  invariably  proved  fatal,  0.25  c.c.  never  did. 
The  explanation  of  this  is  to  be  found  in  the  fact  that  with 
the  smaller  dose  all  animals  were  able  to  cause  disintegra- 
tion of  all  bacilli  injected  before  a  fatal  dose  of  poison  was 
set  free.  If  now  0.5  c.c.  be  given  some  would  recover,  while 
others  would  die.  In  this  case  we  would  speak  of  the  former 
as  possessing  a  greater  individual  resistance  than  the  latter. 
This  simply  means  that,  in  the  first  instance  the  animal 
has  possessed  a  sufficient  quantity  of  bactericidal  substance 
directly  available  to  cause  disintegration  of  all  bacilli  before 
the  latter  have  multiplied  to  a  sufficient  extent  to  furnish 
enough  poison  to  kill  the  animal  on  its  liberation.  On  the 
other  hand,  those  animals  which  succumbed  did  not  possess 
quite  enough  of  the  bactericidal  substance,  or  at  least  did 
not  possess  it  in  a  form  available  for  immediate  use.  When, 
however,  the  dead  bacterial  substance  is  given  the  dose 
of  poison  which  the  animal  receives  is  a  certain  definite 
amount  and  is  not  capable  of  subsequent  increase. 


124 


PROTEIN  POISONS 


Accompanying  the  fall  in  temperature  there  is  apparent 
lassitude,  stupor,  and  roughening  of  the  coat.  In  cases  in 
which  many  times  the  fatal  dose  has  been  given,  the  animals 
occasionally  die  within  from  four  to  six  hours  with  convul- 
sions, a  feature  which  can  now  and  then  be  observed  after 
the  injection  of  large  quantities  of  the  living  bacillus.  At 
autopsy  we  find  a  picture  similar  in  all  respects  to  that 
following  inoculation  with  the  living  colon  bacillus.  There 
is  a  marked  hemorrhagic  peritonitis,  the  peritoneal  cavity 

FIG.  6 


101 
100° 

99 
98 
97° 
96 
95 
94 


Temperature  curve  of  guinea-pig  after  intraperitoneal  injection  of 
non-fatal  dose  of  crude  bacterial  cell  substance. 

containing  bloody  fluid,  together  with  unabsorbed  bacterial 
cell  substance,  and  the  omentum  and  mesentery  showing 
numerous  punctiform  hemorrhages.  It  is  needless  to 
state  that  in  all  cases  cultures  were  made  from  the  peritoneal 
cavity  and  heart's  blood  immediately  after  death,  and 
these  proved  to  be  sterile.  From  this  we  see  that  practically 
the  sole  difference  between  the  effects  following  inoculation 
with  the  living  bacillus  and  the  injection  of  the  dead  bac- 
terial substance  is  a  shortening  of  the  period  of  incubation 


ACTION  ON  ANIMALS  125 

due,  no  doubt,  to  the  fact  that  the  intracellular  poison  is 
liberated  much  more  rapidly  and  in  greater  concentration 
in  the  second  case.  As  will  be  seen  later,  it  is  not  so  much 
the  absolute  quantity  of  the  poison  which  is  injected  that 
determines  the  result,  as  the  amount  which  is  active  at  a 
given  time. 

The  Action  of  the  Soluble  Poison. — When  doses  of  this 
powder  are  given  intraperitoneally  in  amounts  varying 
from  8  to  60  milligrams,  according  as  to  whether  we  have 
been  careful  to  remove  most  of  the  common  salt  or  not,  a 
fatal  result  follows  in  guinea-pigs  in  from  thirty  to  sixty 
minutes.  Within  fifteen  minutes  after  injection  the  temper- 
ature begins  to  fall  and  sometimes  within  half  an  hour  has 
reached  94°  F.  or  even  lower.  At  first,  after  an  interval  of 
from  five  to  ten  minutes  immediately  following  the  injec- 
tion, the  animal  appears  restless,  runs  about  the  cage,  and 
shows  a  great  tendency  to  scratch  itself,  this  undoubtedly 
being  due  to  itching  sensations  in  the  skin  caused  by  irrita- 
tion of  the  peripheral  nerves.  The  animal  then  begins  to 
show  evidence  of  lack  of  coordination,  which  is  rapidly 
followed  by  partial  paralysis,  which  is  especially  marked 
in  the  hind  extremities.  This  stage  lasts  for  from  five  to 
ten  minutes,  during  the  later  part  of  which  the  animal 
usually  lies  quietly  on  one  side.  From  this  state  the  animal 
passes  into  what  one  might  term  the  convulsive  stage. 
These  convulsions  are  usually  clonic  in  nature  and,  as  a 
rule,  at  first  involve  only  the  neck  muscles,  the  head  being 
momentarily  drawn  backward  on  the  back.  At  first  these 
convulsions  are  but  slight  in  degree  and  are  separated  by 
considerable  intervals  of  time.  Soon,  however,  they  become 
much  more  frequent  and  of  much  greater  severity.  Gradu- 
ually  they  become  more  and  more  general  in  their  extent, 
until  all  the  muscles  of  the  body  become  involved  in  violent 
clonic  convulsions.  This  stage  when  present  presages  a 
fatal  outcome;  rarely  an  animal  recovers  after  reaching 
the  convulsive  stage.  During  a  convulsion,  or  occasionally 
in  the  interval  of  calm,  respiration  ceases.  The  heart, 
however,  continues  to  beat,  at  first  with  perfect  regularity 


126  PROTEIN  POISONS 

and  no  acceleration;  indeed,  the  rate  seems  to  be  somewhat 
slower  than  normal.  Gradually  the  beat  becomes  more 
and  more  feeble,  the  rate  and  regularity  being  preserved 
to  the  end.  It  is  usually  only  after  an  interval  of  from 
three  to  four  minutes  after  the  cessation  of  respiration  that 
the  heart  ceases  to  beat.  As  has  been  previously  stated,  a 
fatal  issue,  if  it  occurs  at  all,  always  results  within  one 
hour  after  injection  and  usually  within  from  thirty  to  forty 
minutes.  This  is  to  a  large  extent  independent  of  whether 
the  dose  is  the  minimum  lethal  one  or  two  or  three  times 
that  amount.  It  is  certainly  entirely  independent  of  the 
size  of  the  pig.  Death,  of  course,  results  at  slightly  different 
times  with  different  batches  of  the  poison,  but  even  in  this 
case  the  interval  of  time  between  injection  and  a  fatal 
issue  does  not  vary  to  any  great  extent.  A  dose  wrhich  has 
proved  to  be  the  minimum  fatal  dose  for  one  pig  will  almost 
surely  prove  to  be  the  same  for  another.  In  other  words, 
we  have  done  away  practically  entirely  with  the  period  of 
incubation,  and  the  poison  acts  so  rapidly  that  individual 
resistance  plays  no  part;  hence,  the  animal  acts  almost 
with  the  exactitude  of  a  chemical  compound  into  which 
for  all  practical  purposes  it  has  been  converted.  The  period 
of  incubation  has  ceased  to  exist  since  the  poison  is  no  longer 
contained  within  either  the  dead  or  the  living  bacillus,  but 
is  present  in  a  free  and  uncombined  form,  capable  of  uniting 
immediately  with  those  body  cells  for  which  it  may  possess 
a  special  affinity. 

At  autopsy  no  special  gross  lesions  can  be  made  out. 
The  peritoneum  is  smooth  and  shiny  throughout,  and  there 
is  not  the  slightest  evidence  of  either  hemorrhage  or  even 
marked  congestion  in  the  omentum  or  mesentery.  This  is 
very  important  and  in  marked  contrast  to  the  hemorrhagic 
peritonitis  found  after  injection  of  either  the  living  or  the 
dead  colon  bacillus.  We  are  inclined  to  believe  that  it  is 
the  distinguishing  feature  between  the  injection  of  the 
poison  in  a  comparatively  free  and  in  a  combined  state. 
At  one  time  we  attempted  to  obtain  the  poison  by  a  simpler 
method,  omitting  the  extraction  of  the  crude  substance 


ACTION  ON  ANIMALS  127 

with  ether.  The  result  was  that  on  evaporation  of  the 
alcoholic  filtrate  we  obtained  a  sticky  residue  which  it 
was  utterly  impossible  to  pulverize  or  to  weigh.  We  were 
compelled,  therefore,  to  content  ourselves  with  evaporating 
it  to  a  sticky  mass,  which  was  then  immediately  dissolved 
in  water.  The  solution  of  the  substance  thus  prepared 
was .  very  poisonous,  but,  as  a  rule,  took  from  one  to  two 
hours  or  even  longer  to  bring  about  a  fatal  result.  The 
animals  showed  the  roughening  of  the  coat  and  the  stupor 
characteristic  of  the  living  and  dead  bacillus,  but  not  as  a 
rule  seen  in  the  case 'of  the  soluble  poison.  ..Furthermore, 
the  majority  of  the  animals  showed  during  life  unmistakable 
signs  of  peritoneal  inflammation.  They  died  in  convulsions. 
At  autopsy  an  intense  hemorrhagic  peritonitis  was  present, 
which  was  particularly  prominent  in  the  omentum  and 
mesentery,  and  hemorrhage  was  often  present  in  the  cap- 
sules of  the  liver  and  the  spleen.  From  the  fact  that  death 
was  slower  in  these  cases  and  that  the  symptoms  were 
more  like  those  seen  after  inoculation  with  the  living  bacillus, 
we  are  inclined  to  believe  that  in  this  instance  the  poison, 
although  split  off  from  the  bacillus  itself,  still  exists  in 
combination  with  some  other  cell  group,  and  that  it  is 
essential  that  this  combination  be  broken  up  before  the 
poison  can  be  set  free  and  can  act  on  the  body  cell. 

Another  interesting  fact  in  this  connection  is  furnished 
by  the  action  of  the  poison  in  solutions  which  have  been 
rendered  strongly  alkaline  by  the  addition  of  sodium  bicar- 
bonate. As  has  been  previously  stated,  the  aqueous  solu- 
tions of  the  poison  are  slightly  acid  in  reaction,  and  in  order 
to  avoid  the  irritative  effects  which  might  follow  their 
injection  into  the  peritoneal  cavity,  they  were  neutralized 
or  rendered  slightly  alkaline  by  the  addition  of  sodium 
bicarbonate.1  At  first  no  attempt  was  made  to  secure 
perfect  neutralization,  with  the  result  that  sometimes 
we  were  making  use  of  neutral,  while  again  slightly  or 


1  The    precaution    of   neutralizing    the    soluble    poison,    when    properly 
prepared,  is  unnecessary  as  it  has  no  appreciable  irritative  action. 


128 


PROTEIN  POISONS 


decidedly  alkaline  solutions  were  employed.  It  was  soon 
noticed,  however,  that  the  results  obtained  in  the  three 
cases  were  very  different.  Thus,  whereas  in  the  neutral 
or  faintly  alkaline  solution  the  injection  of  60  mg.  of  the 
powder  invariably  killed,  in  the  case  of  a  stronger  alkaline 
solution  the  same  amount  did  not  cause  a  fatal  result, 
although  the  animals  were  very  ill.  From  this  fact  it 
became  evident  that  some  change  had  taken  place  in  the 
poison  on  standing  in  alkaline  solution.  In  order  to  study 
this  change  more  in  detail,  experiments  were  conducted 
with  solutions  of  different  degrees  of  alkalinity,  with  the 
results  found  in  the  following  tables: 


TABLE  III. — RESULTS  WITH  SOLUTION  OF  POISON  BARELY  NEUTRALIZED 
WITH  SODIUM  BICARBONATE  AND  PLACED  IN  INCUBATOR 


Solution 

Result  of             Time  of 

No.  of 

Dose  of 

kept  in 

Weight 

injec-              death  after 

animal. 

poison. 

incubator. 

of  pig. 

tion.                injection. 

1 

60  mg. 

Fresh 

325  gm. 

+                30  minutes 

2 

60  mg. 

2  hours 

330  gm. 

+               20  minutes 

3 

60  mg. 

20   hours 

370  gm. 

+                20  minutes 

4 

60  mg. 

2  days 

320  gm. 

+                15   minutes 

5 

60  mg. 

4  days 

350  gm. 

+               20  minutes 

6 

60  mg. 

6  days 

350  gm. 

+                45   minutes 

7 

60  mg. 

8  days 

320  gm. 

+                20  minutes 

TABLE  IV. — RESULTS  WITH  SOLUTION  OF  POISON  RENDERED  DECIDEDLY 
ALKALINE  WITH  SODIUM  BICARBONATE  AND  PLACED  IN  INCUBATOR 


Solution 

No  of 

Dose  of 

kept  in 

Weight 

animal. 

poison. 

incubator. 

of  pig. 

1 

60  mg. 

Fresh. 

350  gm. 

(barely 

neut.) 

2 

60  mg. 

Fresh. 

370  gm. 

decided- 

ly alk. 

3 

60  mg. 

2   hours 

325  gm. 

4 

80  mg. 

4   hours 

310  gm. 

5 

120  mg. 

24   hours 

280  gm. 

160  mg.  3   days  350  gm. 


Result  of 
injection. 


Very    sick 
for  2  hours 


Sick 

+ 

+ 


Time  of 

death  after 

injection. 

30  minutes 
Recovered 


Recovered 
Recovered 
More  than 

5  hours. 
7  hours. 


ACTION  ON  ANIMALS 


129 


TABLE    V. — RESULTS    WITH  SOLUTION  OF  POISON  RENDERED    DECIDEDLY 

ALKALINE  wiTri  SODIUM  BICARBONATE  AND  KEPT  AT  ROOM 

TEMPERATURE 


Time  at 

No.  of 

Dose  of 

room  tem- 

Weight 

animal. 

poison. 

perature. 

of  pig. 

1 

60  mg. 

Fresh 

350  gm. 

2 

60  mg. 

12  hours 

265  gm. 

3 

90  mg. 

2  days 

280  gm. 

4 

120  mg. 

2  days 

460  gm. 

5 

120  mg. 

7  days 

405  gm. 

6         160  mg.         7  days 


440  gm. 


Result  of 
injection. 

Sick 
Sick 

Sick  for  5 

hours 

Sick  for 

several  hours 


Time  of 
death  after 
injection. 
35  minutes 
Recovered 
Recovered 
20  minutes 
Recovered 

Recovered 


From  the  above  tables  it  will  be  seen  that  the  degree  of 
alkalinity  of  the  solution,  and  especially  the  length  of  time 
that  the  poison  has  stood  in  alkaline  solution  are  very 
important  factors  in  determining  its  toxicity.  Thus  in 
Table  III,  in  which  the  solution  was  barely  neutralized,  the 
poison  seems  to  have  retained  its  full  potency  after  eight 
days  in  the  incubator,  whereas,  in  the  case  of  the  strongly 
alkaline  solution,  the  potency  has  decreased  markedly 
within  from  twenty-four  to  forty-eight  hours.  Again, 
there  are  great  differences  to  be  seen  depending  on  whether 
the  strongly  alkaline  solution  has  been  kept  at  room  tem- 
perature or  at  that  of  the  incubator,  the  decrease  in  toxicity 
being  much  less  rapid  in  the  first  instance. 

A  more  detailed  report  of  the  effects  on  animals  than 
it  was  possible  to  give  in  the  above  tables  is  not  without 
interest.  For  example,  in  Table  IV,  No.  2,  which  received 
60  milligrams  immediately  after  the  solution  had  been 
rendered  decidedly  alkaline,  was  very  sick  indeed,  whereas 
No.  3,  which  received  the  same  amount  after  two  hours 
in  the  incubator,  was  only  slightly  affected.  In  the  case 
of  Nos.  5  and  6  the  effects  observed  corresponded  more 
closely  to  those  obtained  with  the  crude  bacterial  cell 
substance.  It  is  unfortunate  that  the  time  of  death  was 
not  ascertained  in  the  case  of  No.  5.  No.  6  did  not  suc- 
cumb until  seven  hours  after  the  injection.  On  autopsy 
9 


130  PROTEIN  POISONS 

there  was  considerable  fluid  in  the  peritoneal  cavity,  and 
the  vessels  of  the  mesentery  were  markedly  congested. 
The  omentum  was  particularly  injected  and  a  few  minute 
hemorrhages  could  be  made  out.  The  most  plausible  explan- 
ation of  the  above  facts  is  found  in  the  theory  that  the 
poison  has  not  been  destroyed  in  the  alkaline  solution,  but 
rather  has  entered  into  chemical  combination  with  the 
alkali  and  that  we  are  again  dealing  with  it  in  a  combined 
instead  of  in  a  free  state.  The  fact  that  the  same  amount 
will  not  cause  a  fatal  result  is  thus  readily  explained,  since 
the  outcome  depends  largely  on  the  rapidity  with  which 
the  poison  acts.  If  it  is  present  in  a  state  of  combination 
which  must  be  broken  up  before  it  can  exert  its  deleterious 
action  on  the  body,  and  if  this  combination  is  only  slowly 
decomposed,  the  nerve  cells,  for  which  it  apparently  has  a 
special  affinity,  are  not  subjected  to  an  overwhelming  dose 
at  one  time,  as  in  the  case  of  the  intraperitoneal  injection 
of  the  free  poison. 

The*  results  obtained  in  animals  Nos.  5  and  6,  Table  V, 
are  very  interesting.  In  these  instances  there  were  two 
distinct  illnesses,  the  first  becoming  manifest  within  from 
twenty  to  thirty  minutes  after  the  injection  and  corre- 
sponding in  all  respects  to  that  following  a  non-fatal  dose 
of  what  we  have  for  convenience  termed  the  free  poison. 
The  animals  were  decidedly  in  better  condition  at  the  end 
of  an  hour;  however,  they  then  began  to  show  symptoms 
similar  to  those  noticed  after  the  injection  of  the  crude 
cell  substance,  i.  e.,  roughening  of  the  coat,  stupor,  and 
slight  convulsive  movements.  Recovery  from  this  state 
did  not  occur  until  after  the  lapse  of  from  five  to  six  hours. 
It  is  evident  that  here  the  first  signs  of  illness  were  due 
to  some  of  the  poison  which  had  not  as  yet  combined  with 
the  alkali,  and  hence  still  existed  in  the  free  state,  whereas 
the  later  symptoms  were  due  to  the  effects  of  the  slow 
liberation  of  the  same  poison  from  its  combination.  In  this 
connection  it  is  interesting  to  note  that  the  combination 
between  the  poison  and  the  alkali  which  apparently  takes 
place  in  decidedly  alkaline  solutions  is  not  an  immediate 


ACTION  ON  ANIMALS  131 

one,  but  occurs  gradually  and  reaches  a  maximum  only 
after  the  lapse  of  a  considerable  interval  of  time.  That 
the  rapidity  with  which  this  combination  is  effected  depends 
largely  on  temperature  is  shown  by  the  fact  that  it  occurs 
much  more  rapidly  in  a  solution  kept  in  the  incubator 
than  in  one  which  is  allowed  to  stand  at  room  temperature. 

The  results  which  follow  the  injection  of  a  fatal  dose  of 
the  soluble  poison  intraperitoneally  have  already  been 
described.  When  a  non-fatal  dose  has  been  injected  the 
symptoms  first  noticed  are  similar  in  all  respects  to  those 
following  a  fatal  dose.  The  animal  becomes  restless,  shows 
signs  of  irritation  of  the  peripheral  nerves,  incoordination, 
and  partial  paralysis.  The  convulsive  stage  is  not  present, 
as  a  rule,  and  when  ,it  is  noticed  is  evidenced  solely  by 
slight  movements  separated  by  considerable  intervals  of 
time.  We  have  never  seen  a  case  showing  marked  general- 
ized convulsions  which  recovered.1  Recovery  is  apparently 
rapid  and  complete,  and  within  two  hours  after  injection 
the  animal  which  has  been  desperately  ill  appears  as  well 
as  any  untreated  animal.  Tne  maximum  effect  is  obtained 
within  from  forty-five  to  sixty  minutes  in  every  instance. 
The  study  of  the  changes  in  temperature  in  these  animals 
is  particularly  interesting.  Within  fifteen  minutes  the 
rectal  temperature  has  begun  to  fall  and  has  reached  a 
minimum  within  one  hour. 

It  remains  stationary  for  a  short  time  and  then  begins 
to  rise  again,  and  at  the  end  of  three  hours  after  the  injection 
has  usually  returned  to  normal  or  above. 

On  injecting  the  soluble  poison  subcutaneously,  we  find 
that  animals  are  able  to  withstand  a  much  larger  dose 
than  when  the  poison  is  given  intraperitoneally.  Thus,  in 
the  case  of  a  poison,  60  mg.  of  which  invariably  killed 
when  given  intraperitoneally,  it  was  found  that  120  mg. 
could  be  given  subcutaneously  without  causing  a  fatal 
result.  However,  the  injection  of  a  solution  containing 
180  mg.  invariably  caused  death,  the  fatal  issue  occurring 

«  l  This  does  rarely  occur. 


132 


PROTEIN  POISONS 


in  about  the  same  length  of  time  as  in  the  case  of  animals 
treated  intraperitoneally.  Thus  a  dose  of  180  mgs.  always 
proved  fatal  in  from  one-half  to  three-quarters  of  an  hour. 
The  symptoms  are  practically  identical  with  those  following 
the  intraperitoneal  injection  with  the  exception  of  the  fact 
that  the  various  stages  are  much  more  sharply  defined. 
For  example,  the  stage  of  peripheral  irritation  is  much  more 
marked.  The  animal  soon  after  injection  becomes  very 
restless,  runs  around  his  cage,  and  scratche's  his  body. 
This  itching  seems,  however,  to  be  general  from  the  outset, 
and  is  not,  apparently,  more  pronounced  in  the  immediate 

FIG.  7 


1U:> 

/ 

- 

k 

/ 

'& 

V 

z 

\ 

z 

\ 

/ 

<*<;' 

_ 

1  —  . 

—  — 

—  — 

—  -" 

Temperature  curve  of  guinea-pig  treated  with  45  mgs.  of  the  soluble 
poison  intraperitoneally. 

neighborhood  of  the  site  of  injection.  If  the  animal  has 
been  injected  under  the  skin  of  the  abdomen,  its  attention 
is  not  necessarily  first  attracted  to  this  spot,  but  it  may 
begin  by  scratching  its  nose  or  one  of  the  extremities. 
Another  peculiar  symptom,  which  is  probably  due  to 
peripheral  irritation,  and  which  is  seldom  seen  in  cases  of 
intraperitoneal  injection,  is  the  tendency  which  the  animals 
show  to  dig  furiously  in  the  shavings  in  the  bottom  of  their 
cages.  This  feature  is  quite  characteristic,  and  is  seldom 
absent  in  pigs  which  have  been  treated  subcutaneously. 
The  later  stages  are  similar  in  all  respects  to  those  seen 
following  the  intraperitoneal  injection.  The  animal  shows 


ACTION  ON  ANIMALS  133 

symptoms  of  incoordination,  lies  on  one  side,  and  finally 
develops  convulsions,  with  failure  of  respiration,  the  heart 
continuing  to  beat  regularly  for  some  time  after  the  com- 
plete cessation  of  respiration.  Here  also  the  symptoms  are 
accompanied  by  a  decided  fall  in  the  body  temperature. 

The  results  following  the  intravenous  injection  of  the 
soluble  poison  are  given  in  the  following  table: 

TABLE  VI 

Time  of  cessa-  Time  of  cessa- 

No.  of  animal.        Amount   of  tion  of  respira-  tion  of  heart- 
poison  injected  tion  after  beat  after 
intravenously.  injection.  injection. 

1  ...      10  mg.  4  minutes  7  minutes 

2  ...      10  mg.  Recovered 

3  ...      10  mg.  3  minutes  6  minutes 

4  ...      10  mg.  Recovered 

5  ...      15  mg.  4  minutes  6  minutes 

6  ...      15  mg.  3  minutes  5  minutes 

7  ...      15  mg.  4  minutes  7  minutes 

8  ...      15  mg.  4  minutes  6  minutes 

9  ...     20  mg.  3  minutes  5  minutes 

10  ...     20  mg.  4  minutes  6  minutes 

11  ...     20  mg.  3  minutes  6  minutes 

12  ...     20  mg.  3  minutes  7  minutes 

From  the  above  table  it  will  be  seen  that  in  all  cases 
respiration  ceased  within  four  minutes  after  injection. 
Indeed,  the  respiratory  embarrassment  becomes  pronounced 
immediately  following  the  injection.  The  animal  struggles 
for  breath  and  there  is  violent  retraction  of  the  sternum. 
No  convulsions  are  seen  following  the  intravenous  injection, 
this  being  probably  due  to  the  inhibitory  influence  of  the 
anesthetic  which  has  been  used  during  the  preparation  of 
the  animal  for  the  operation.  The  failure  of  respiration  in 
the  absence  of  convulsions  would  seem  to  be  conclusive 
evidence  that  the  cessation  of  this  function  is  due  not  to 
mechanical  interference  during  a  convulsive  attack,  but  to 
a  direct  paralysis  of  the  respiratory  centre  itself.  Further- 
more, the  fact  that  the  heart  continues  to  beat  in  a  perfectly 
normal  manner  for  from  two  to  four  minutes  after  respira- 
tion has  entirely  ceased,  would  tend  to  show  that  the 


134  PROTEIN  POISONS 

immediate  cause  of  death  is  asphyxia  brought  about  by 
paralysis  of  respiration  through  the  action  of  the  poison.1 

This  action  of  the  heart  after  the  cessation  of  respiration 
is  exceedingly  interesting,  and  is  entirely  analogous  to  that 
mentioned  as  following  the  intraperitoneal  injection.  The 
rate  of  the  beat  is  decidedly  lessened  and  at  first  the  indi- 
vidual beats  are  stronger.  They  gradually  become  more 
and  more  feeble,  however,  until  finally  the  heart  stops  in 
diastole,  the  rate  after  the  preliminary  slowing  remaining 
unchanged  until  the  end.  It  is  worthy  of  note  that  in  the 
case  of  intravenous  injections  the  fall  of  temperature,  which 
is  so  marked  a  feature  after  the  intraperitoneal  and 
subcutaneous  injections  of  the  poison  does  not  occur. 
The  explanation  of  this  fact  is  doubtless  to  be  found  in  the 
very  short  interval  of  time  which  elapses  between  the 
injection  and  a  fatal  outcome.  As  regards  the  size  of  the 
lethal  dose  when  given  intravenously,  we  see  that  10  mg. 
often,  and  15  mg.  invariably,  proved  fatal.2 

For  purposes  of  comparison,  we  have  always  made  use 
of  the  poison  obtained  from  the  same  extraction  in  our 
intravenous,  subcutaneous,  and  intraperitoneal  injections, 
and  have  therefore  been  able  to  ascertain  with  a  fair  degree 
of  accuracy  the  differences  in  dose  required  to  bring  about 
a  fatal  result  in  the  three  cases.  Thus,  whereas  60  mg. 
represents  the  fatal  dose  when  given  intraperitoneally,  it 
requires  between  120  and  180  mgs.  subcutaneously,  and 
only  from  10  to  15  mgs.  of  the  same  poison  to  cause  death 
when  given  intravenously.  These  differences  are  un- 
doubtedly due  to  the  rapidity  of  absorption  in  the  various 
cases,  and  a  fatal  issue  depends  entirely  on  whether  suffi- 
cient poison  reaches  the  sensitive  area  at  one  time  to  cause 
cessation  of  respiration  OP  not. 

It  has  now  been  shown  that  a  very  powerful  intracellular 
poison  can  be  obtained  from  the  colon  bacillus.  As  has 
been  previously  stated,  the  results  given  in  the  foregoing 
experiments  are  those  obtained  with  the  poison  from  one 

1  The  physiological  action  of  this  protein  poison  is  discussed  on  page  315. 

2  The   fatal   dose   of  the  purest   form    of    the    poison  which    we    have 
obtained  is  0.5  mg.  intracardially. 


ACTION  ON  ANIMALS  135 

extraction  only.  It  must  be  understood  that  the  poison 
is  not  in  a  pure  state  and  when  it  is  stated  that  60  mgs. 
causes  death  when  injected  intraperitoneally  we  refer  simply 
to  the  powder  obtained  from  a  given  extraction.  We  have 
been  able  to  procure  powders  which  kill  in  doses  of  15  and 
even  as  low  as  8  mgs.  when  given  intraperitoneally.  This 
difference  is  in  large  amount  due  to  the  presence  of  sodium 
chloride,  since  no  attempt  has  been  made  to  remove  this 
salt  in  the  case  of  the  less  toxic  powders  by  redissolving 
in  absolute  alcohol. 

There  are  several  facts  which  lead  us  to  believe  that 
this  poison  is  the  one  which  causes  the  symptoms  of  illness 
and  death  in  animals  infected  with  the  colon  germ.  Most 
of  these  facts  have  already  been  brought  out,  but  it  may 
not  be  out  of  place  to  briefly  recapitulate  at  this  point. 
As  has  been  previously  seen,  the  results  obtained  with  the 
living  germ,  the  dead  bacterial  substance,  and  the  soluble 
poison  can  best  be  explained  on  the  ground  that  the  poi- 
sonous body  in  each  case  is  the  same.  The  differences  in 
action  are  not  differences  in  symptoms,  but  simply  in  the 
rapidity  with  which  these  symptoms  become  manifest. 
While  it  is  undoubtedly  true  that  in  animals  dying  with 
the  minimum  fatal  dose  of  the  living  germ,  the  convulsive 
stage  is  not  present  or  is  only  slightly  marked,  it  is  rarely 
absent  in  cases  where  from  three  to  four  times  the  fatal 
amount  has  been  given.  The  sole  difference  between  the 
living  germ  and  the  soluble  poison  which  would  appear  to 
demand  an  explanation  is  the  lack  of  evidence  of  a  perito- 
nitis in  the  latter  case.  This,  we  think,  is  best  explained  by 
the  fact  that  in  the  case  of  the  soluble  poison  the  poisonous 
substance  exists  in  an  uncombined  form,  which,  of  course,  is 
not  true  in  the  case  of  either  the  living  or  the  dead  bacte- 
rial cell.  The  uncombined  poison  is  rapidly  absorbed  from 
the  peritoneal  cavity,  and  hence  the  irritative  effects  which 
would  result  from  its  retention  in  this  place  are  absent. 

As  has  been  stated,  one  of  the  first  signs  of  the  action  of 
the  poison  is  a  lowering  of  the  body  temperature.  This 
hypothermia  is  usually  present  to  a  marked  degree,  and 
is  noticeable  before  any  visible  symptoms  occur.  It  is, 


136  PROTEIN  POISONS 

therefore,  the  best  index  which  we  have  as  to  the  exact 
time  at  which  the  poison  begins  to  exert  its  effect.  In  the 
same  manner  the  rise  of  temperature  after  the  development 
of  hypothermia  is  the  first  indication  of  recovery.  More- 
over, if  in  an  animal  with  a  subnormal  temperature  a  rise 
occurs,  it  is  an  infallible  sign  of  ultimate  recovery,  no  matter 
how  grave  the  general  condition  may  appear  to  be  at  the 
time.  We  have  laid  great  stress,  therefore,  on  the  changes 
in  body  temperature  as  furnishing  the  most  delicate  test 
of  the  action  of  the  poison.  It  may  be  here  stated  that 
the  body  temperature  of  guinea-pigs  is  ordinarily  fairly 
constant  within  certain  narrow  limits,  and  they  are  much 
more  satisfactory  animals  to  work  with  in  this  respect 
than  are  rabbits.  Moreover,  their  temperature  does  not 
seem  to  be  materially  altered  by  the  injections  of  sterile 
salt  solution  or  such  inert  substances  as  a  suspension  of 
pumice  stone  into  the  peritoneal  cavity.  As  has  been  seen 
in  the  case  of  the  living  germ,  it  is  only  after  the  lapse  of 
several  hours  that  a  fall  in  temperature  occurs.  This 
would  indicate  that  it  is  not  until  this  time  that  sufficient 
poison  is  liberated  to  cause  notable  toxic  effects  in  the 
animal.  That  it  takes  an  appreciable  time  for  the  poison 
to  be  liberated  from  the  bodies  of  the  bacilli  is  well  illus- 
trated in  the  instance  of  the  dead  bacterial  substance. 
Here  it  is  only  a  question  of  dissolution  of  the  bacilli  and 
the  setting  free  of  the  contained  poison,  and  yet  it  will 
be  noticed  that  an  interval  of  at  least  two  hours  and  usually 
longer  elapses  before  there  is  any  noticeable  fall  in  temper- 
ature. The  maximum  effect  in  this  case  is  reached  between 
four  and  six  hours,  and  if  the  dose  has  been  a  non-fatal 
one,  recovery  begins  at  the  end  of  from  eight  to  ten  hours, 
as  is  indicated  by  the  upward  trend  of  the  temperature 
curve  at  this  point.  In  the  case  of  the  soluble  poison,  the 
toxic  effect  begins  at  once.  Within  fifteen  minutes  the 
temperature  has  begun  to  drop  and  within  an  hour  has 
reached  a  minimum.  Recovery  then  begins  and  within 
three  hours  the  effect  of  the  poison  has  worn  off,  as  is  best 
evidenced  by  the  return  of  the  body  temperature  to  normal 
or  above  bv  this  time. 


CHAPTER  VII 

THE  PRODUCTION  OF  ACTIVE  IMMUNITY 

WITH  THE  SPLIT  PRODUCTS  OF  THE 

COLON  BACILLUS1 

IT  may  be  stated  that  the  work  which  we  have  done 
with  the  colon  bacillus  up  to  the  present  time  has  in  every 
instance  upheld  the  belief  that  the  substances  which  give 
rise  to  the  phenomena  occurring  in  animals  infected  with 
the  living  colon  germ  exist  as  essential  groups  within 
the  bacterial  cell  and  can  be  liberated  from  the  latter 
only  by  its  disruption.  Moreover,  until  these  substances 
have  been  separated  from  the  other  constituents  of  the 
bacterial  cell  with  which  they  are  normally  combined  they 
are  unable  to  exert  any  deleterious  action  upon  the  body 
cells.  If  the  belief  that  the  phenomena  which  result  from 
infection  with  the  colon  bacillus  are  due  to  the  action  of 
the  intracellular  constituents  of  this  organism  is  correct, 
we  would  expect  that  it  might  be  possible  by  chemical 
means  to  split  up  this  bacillus  into  different  groups,  the 
injection  of  some  of  which  into  animals  would  be  followed 
by  some  of  the  results  which  are  seen  after  inoculation  with 
the  living  germ.  In  a  previous  chapter  we  have  shown  that 
it  is  possible  to  split  off  a  toxic  group  which  causes  death 
in  animals  with  symptoms  similar  to  those  observed  after 
the  injection  of  a  fatal  dose  of  living  bacilli.  However, 
death  is  by  no  means  the  sole  phenomenon  which  results 
from  the  inoculation  of  animals  with  the  colon  bacillus. 
The  results  which  follow  the  injection  of  non-fatal  doses  of 
the  living  germ  are  of  equal  if  not  of  greater  importance. 

1  This  chapter  is  taken,  with  but  few  changes,  from  an  article  by  Victor 
C.  Vaughan,  Jr.,  in  the  Journal  of  Medical  Research,  1905,  xiv,  67. 


138  PROTEIN  POISONS 

It  is  a  well-known  fact  that  animals  which  have  been 
treated  with  non-fatal  doses  of  either  the  living  or  dead 
colon  germ  acquire  a  certain  degree  of  immunity  toward 
subsequent  infection  with  this  bacillus.  If,  now,  our  theory 
as  regards  the  action  of  this  bacillus  is  correct,  one  would 
suppose  that  among  the  groups  which  we  have  been  able 
to  split  off  there  exist  certain  ones  which  possess  the  power 
of  producing  immunity  when  injected  into  susceptible 
animals.  To  ascertain  whether  such  an  active  immunity 
can  be  established  in  animals  through  treatment  with  the 
split  products  of  the  colon  bacillus  is  the  aim  of  this  chapter, 
and  we  shall  find  it  convenient  to  take  up  (1)  the  active 
immunity  obtained  with  the  toxic  portion,  and  (2)  the 
immunity  obtained  with  the  residue  which  remains  after 
the  separation  of  the  poisonous  portion  from  the  cellular 
substance. 

1.  Immunization  with  the  Poisonous  Portion  of  the  Cellular 
Substance  of  the  Colon  Bacillus. — From  the  description  of 
the  action  of  the  "crude  soluble  poison"  of  the  colon  bacillus 
given  in  the  preceding  chapter,  it  can  be  readily  seen  that 
the  poison  with  which  we  are  working  is  one  which  exerts 
its  action  with  great  rapidity.  The  difficulties  of  immunizing 
animals  with  such  a  poison  can  be  readily  appreciated,  and 
it  is  inevitable  that  during  the  course  of  treatment  a  large 
number  should  be  lost.  Up  to  the  present  time  our  attempts 
to  produce  immunity  with  the  toxic  portion  have  been 
largely  confined  to  intraperitoneal  and  subcutaneous  injec- 
tions with  what  we  have  termed  the  free  or  uncombined 
poison.  However,  as  has  been  shown  in  a  previous  chapter, 
it  is  possible  to  make  use  of  this  poison  in  a  combined  state 
by  rendering  the  solution  of  the  toxic  part  decidedly  alka- 
line with  sodium  bicarbonate  and  allowing  it  to  stand  for 
some  time,  preferably  at  incubator  temperature.  With 
this  combined  poison,  one  is  able  to  give  much  larger  doses 
without  producing  a  fatal  result,  and,  moreover,  the  effect 
of  the  poison  is  in  this  instance  manifested  over  a  much 
longer  period  of  time.  These  two  factors  are,  of  course,  of 
primary  importance  in  the  production  of  immunity  and  it 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        139 

is  quite  possible  that  with  the  employment  of  the  combined 
poison  a  higher  degree  of  immunization  may  be  obtained. 

That  it  is  possible  by  means  of  repeated  doses  to  induce  a 
certain  amount  of  tolerance  in  animals  for  the  poisonous 
portion  is  illustrated  in  the  following  tables: 

TABLE  VII 

20  MG.  OF  THIS  POISON  INVARIABLY  CAUSED  DEATH  IN  UNTREATED  PIGS 
WITHIN  ONE  HOUR 


Total 

Weight 

Dose  of  poison  in  mg. 

and  when 

given. 

amount 

Guinea- 
pig  No. 

in 
grams. 

3/2   3/7 

3/15 

3/28 

3/28 

4/1 

4/7 

4/14 

4/22 

poison 
in  mg. 

1 

815 

15.0   20 

25.0 

30.0 

30 

30 

35 

0 

40.0 

45 

270 

2 

705 

12.5   .. 

15.0 

17.5 

20 

25 

30 

.0 

35.0 

40 

195 

3 

575 

10.0   .. 

15.0 

20.0 

20 

25  1 

30 

,0 

35.0 

40 

195 

4 

700 

10.0   .. 

15.0 

20.0 

20 

25 

30 

0 

35.0 

40 

195 

5 

580 

9.0   .. 

12.5 

15.0 

20 

25 

27 

5 

30.0 

35 

174 

6 

790 

10.0   .. 

15.0 

20.0 

25 

25 

30 

0 

35,0 

35 

195 

7 

545 

7,5   .. 

15.0 

20.0 

20 

25 

30, 

0 

32.5 

35 

185 

8 

655 

10.0   .. 

15.0 

20.0 

20 

25 

30 

0 

35.0 

40 

195 

9 

640 

15.0   .. 

20.0 

25.0 

25 

30 

35 

0 

40.0 

45 

235 

The  following  table  furnishes  an  index  to  the  degree  of 
tolerance  established  in  rabbits  through  the  administration 
of  gradually  increasing  doses  of  the  poison: 

TABLE  VIII 

350  MG.  OF  THIS  POISON  CAUSED  DEATH  IN  UNTREATED  ANIMALS 
WITHIN  ONE  HOUR 

Weight         Dose  of  poison  in  mgs.  and  when  given. 
Rabbit 


No.      grams.  12/12  l2/^    ^2/2»     1/9      1/14      1/19       1/26 

1  1850       200       200       400       500       700       1000       1400       Died  20  min.  after 

last  injection. 

2  2800       250       300       400       500       700       900         1000 

3  1950       300       250       400       600       700       ........       Died  30  min.  after 

last  injection. 

4  2100       200       250       400       500       700       1000       1200       Died  35  min.  after 

last  injection. 

From  a  study  of  the  above  tables  it  can  be  seen  that  in 
the  case  of  both  guinea-pigs  and  rabbits  after  the  adminis- 
tration of  several  doses  of  gradually  increasing  strength, 
a  point  is  reached  at  which  the  animal  is  able  to  withstand 


140  PROTEIN  POISONS 

the  injection  of  from  two  to  three  times  the  amount  which 
would  surely  have  proved  fatal  for  an  untreated  control. 
This  would  indicate  that  during  the  course  of  the  treatment 
the  animal  had  developed  either  a  slight  degree  of  immunity, 
or  had  established  a  certain  amount  of  tolerance  for  the 
poison.  Which  of  these  explanations  is  the  correct  one 
can  only  be  determined  after  a  careful  study  of  the  subject 
of  the  possible  production  of  passive  immunity  and  the 
demonstration  of  a  possible  antibody  in  the  blood  of  treated 
animals.  At  present,  owing  to  the  slight  amount  of  increased 
resistance  which  the  animals  exhibit  to  the  action  of  the 
poison,  we  are  inclined  to  believe  that  the  question  is  one 
of  tolerance.  Although  the  degree  or  tolerance  thus  far 
secured  has  been  limited,  we  do  not  feel  justified  in  con- 
cluding that  greater  resistance  to  the  poison  may  not  be 
obtained.  There  are  many  factors  of  primary  importance 
in  this  work,  all  of  which  must  be  carefully  studied  before 
definite  conclusions  can  be  drawn.  For  example,  the 
interval  of  time  which  is  allowed  to  elapse  between  the 
injections  is  a  matter  of  first  importance.  Since  the  length 
of  time  over  which  the  poison  acts  is  apparently  so  short, 
it  seemed  quite  probable  that  any  reaction  which  might 
occur  on  the  part  of  the  body  would  develop  in  a  compara- 
tively short  time  after  the  injection.  With  the  object  of 
ascertaining  whether  this  was  true  or  not,  animals  were 
treated  daily  with  gradually  increasing  doses  with  the 
following  results: 

TABLE  IX 

60  MG.  OF  THIS  POISON  INVARIABLY  CAUSED  DEATH  IN  UNTREATED 
ANIMALS  WITHIN  ONE  HOUR 

Pig  No.  1.  Pig  No.  2.  Pig  No.  3.  Pig  No.  4.  Pig  No.  5. 

Day.     Wt.,      Dose,      Wt.,      Dose,      Wt.,      Dose,      Wt.,      Dose.      Wt.,      Dose, 
gm.  mg.      gm.         mg.         gin.         mg.         gm.         mg.         gm.         mg. 


1 

425 

45 

385 

45 

460 

45 

390 

45 

405 

45 

2 

380 

50 

360 

50 

415 

50 

355 

50 

385 

50 

3 

385 

60 

375 

60 

420 

60 

365 

60 

385 

60 

4 

385 

80 

375 

80 

450 

80 

370 

80 

380 

80 

5 

405 

100 

370 

100 

450 

100 

375 

100 

390 

100 

6 

405 

112 

385 

112 

460 

112 

385 

112 

7 

420 

125 

400 

125 

470 

100 

410 

112 

Died  in  30 

Died  in  30 

Died  in  30 

minutes. 

minutes. 

minutes. 

THE  PRODUCTION  OF  ACTIVE  IMMUNITY        141 

From  this  we  see  that  it  is  possible  to  establish  a  certain 
amount  of  tolerance  by  means  of  daily  injections  of  the 
poisonous  portion.  Here  again  we  find  that  it  is  compara- 
tively easy  to  reach  a  dose  which  corresponds  to  about 
twice  the  fatal  amount,  but  above  this  the  animal  cannot 
be  carried.  When  death  does  result  from  a  dose  of  the 
poison  which  is  too  large  to  be  borne  by  the  treated  pig,  the 
symptoms  are  identical  in  all  respects  with  those  which 
occur  in  the  case  of  an  untreated  animal,  and  a  fatal  result 
follows  in  the  same  length  of  time. 

The  question  now  arose  as  to  whether  these  animals 
which  had  acquired  a  tolerance  for  the  poisonous  portion  of 
the  colon  bacillus  were  more  resistant  to  inoculation  with 
the  living  germ  than  were  untreated  animals.  In  order  to 
ascertain  this  point,  guinea-pigs  which  had  received  from 
174  to  235  mg.  of  the  toxic  portion  were  inoculated  intra- 
peritoneally  with  doses  of  the  living  germ  with  the  following 
results  : 

TABLE  X 

1  c.c.  OF  A  16-HOUR  CULTURE  OF  THE  COLON  BACILLUS  USED  IN  THESE 

EXPERIMENTS  INVARIABLY  KILLED  A  CONTROL  WITHIN 

TWENTY-FOUR  HOURS. 


No.  of  in- 

Total 
amount 

Interval  be- 
tween last 
injection  and 

Guinea-     jections 

of  poison 

inoculation 

Amount  and  age  of 

pig  No. 

of  poison. 

received. 

with  germ. 

culture. 

Result. 

1 

9 

270  mg. 

11  days 

1  c.c. 

24-hour  culture 

Recovery. 

2 

8 

195  mg. 

11  days 

1  c.c. 

24-hour  culture 

Recovery. 

3 

8 

195  mg. 

11  days 

1  c.c. 

4-day  culture 

Died  in  22  hrs. 

4 

8 

195  mg. 

11  days 

2  c.c. 

4-day  culture 

Dead  in  24  hrs. 

5 

8 

174  mg. 

15  days 

2  c.c. 

24-hour  culture 

Recovery 

6 

8 

195  mg. 

25  days 

2  c.c. 

24-hour  culture 

Recovery 

7 

8 

185  mg. 

33  days 

2  c.c. 

24-hour  culture 

Recovery 

8 

8 

195  mg. 

25  days 

2  c.c. 

24-hour  culture 

Recovery 

9 

8 

235  mg. 

8  days 

2  c.c 

24-hour  culture 

Recovery 

That  an  active  immunity  to  the  living  colon  bacillus  is 
also  developed  in  rabbits  which  have  been  treated  with 
repeated  injections  of  the  toxic  portion  is  illustrated  by  the 
following  experiments: 


142  PROTEIN  POISONS       • 

Rabbit  No.  1  received  between  May  25  and  July  5  eight 
injections  of  the  toxic  part,  the  total  amount  of  poison 
injected  being  855  mg.  On  July  8  this  animal  received 
5  c.c.  of  a  twenty-four-hour  culture  of  the  living  germ 
without  apparent  effect.  The  control  inoculated  at  the 
same  time  was  found  dead  in  eight  hours. 

Rabbit  No.  2  received  between  May  28  and  July  18 
eleven  injections  of  the  toxic  portion,  the  total  amount  of 
poison  given  being  2475  mg.  On  July  27  this  animal  was 
inoculated  with  5  c.c.  of  a  sixteen-hour  culture  of  the  colon 
bacillus  without  effect.  The  control  was  found  dead  in 
ten  hours. 

Rabbit  No.  3  received  between  June  27  and  July  18  seven 
injections  of  the  poison,  the  total  amount  given  being  2100 
mg.  On  July  28  this  animal  was  inoculated  with  6  c.c.  of 
a  twenty-four-hour  colon  culture.  Recovered. 

Rabbit  No.  4  received  the  same  amount  of  poison  as  the 
preceding  one.  Twelve  days  after  his  last  treatment  this 
animal  was  given  6  c.c.  of  a  forty-hour  colon  culture  and 
recovered.  The  control  was  found  dead  in  eight  hours. 

From  the  foregoing  experiments  it  becomes  evident  that 
animals  which  have  been  treated  with  the  toxic  portion  of 
the  colon  bacillus  acquire  a  certain  degree  of  immunity  to 
the  living  germ.  We  are  as  yet  unable  to  state  whether  it 
is  possible  to  obtain  a  high  degree  of  immunity  with  the 
poisonous  portion  or  not.  Thus  far  we  have  had  animals 
which  have  withstood  inoculation  with  from  two  to  four 
times  the  fatal  dose  of  the  living  germ.  It  is  worthy  of 
note  in  this  connection  that  animals  which  have  received 
one  injection  of  a  non-fatal  dose  of  the  poison  are  able  to 
withstand  inoculation  with  twice  the  lethal  dose  of  the 
living  germ  on  the  following  day.  The  immunity  which 
follows  a  single  injection  is,  however,  exceedingly  transitory, 
and  has  usually  disappeared  on  the  second  day  following 
the  treatment.  This  would  seem  to  be  the  most  marked 
difference  between  the  immunity  which  results  from  a 
single  injection  of  the  toxic  part  and  that  which  follows  a 
series  of  injections  extending  over  a  considerable  interval 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        143 

of  time.  In  the  first  instance  the  protection  afforded  is 
very  temporary,  while  in  the  second  it  is  still  present  even 
after  the  lapse  of  from  twenty-five  to  thirty  days.  It  may 
be  possible  that  in  the  case  of  the  injections  extending  over 
a  long  period  of  time  the  immunity  obtained  is  of  a  higher 
degree.  This  is  a  point  which  will  require  further  study. 

In  the  case  of  an  animal  which  has  been  treated  with- 
the  poisonous  portion  and  has  subsequently  received  a  dose 
of  the  living  germ  which  would  surely  have  proved  fatal 
for  a  normal  animal,  the  symptoms  noticed  are  identical 


—  Curve  of  normal  animal  inoculated  with  non-fatal  dose  of  living  bacillus. 
•  —  Curve  of  immune  animal  inoculated  with  living  bacillus. 

in  every  respect  with  those  which  follow  the  injection  of  a 
non-lethal  dose  in  an  untreated  animal.  This  is  a  very 
important  fact  and  one  on  which  we  have  laid  much  stress. 
Moreover,  as  can  be  seen  from  Fig.  8,  the  temperature 
curve  corresponds  very  closely  with  that  obtained  in  the 
case  of  a  normal  animal  inoculated  with  a  non-fatal  dose 
of  the  living  colon  bacillus. 

We  see  that  in  both  instances  there  is  no  appreciable  fall 
in  body  temperature  until  from  six  to  eight  hours  after 
inoculation.  At  this  time  the  minimum  temperature  has 


144  PROTEIN  POISONS 

been  reached  in  each  case,  and  within  from  ten  to  twelve 
hours  it  has  again  returned  to  normal.  The  similarity  of 
the  symptoms  in  the  two  instances  leads  us  to  believe  that 
in  all  probability  we  are  here  dealing  with  an  immunity 
which  is  identical  in  character  with  that  which  is  usually 
spoken  of  as  natural  immunity.  This  idea  has  been  further 
upheld  by  the  fact  that  we  have  been  able  to  obtain  from 
egg  albumen  and  peptone  poisonous  substances  which 
resemble  the  toxic  portion  of  the  colon  bacillus  in  their 
action,  and  by  the  injection  of  single  non-fatal  doses  of 
which  it  is  possible  to  obtain  the  same  transitory  immunity 
to  the  living  colon  germ  as  is  observed  after  the  injection 
of  the  colon  poison.  That  this  toxic  group  is  common  to 
certain  bacteria  and  other  protein  bodies  is  not  improbable, 
and  this  would  furnish  an  explanation  not  only  of  the 
increased  resistance  to  certain  bacterial  infections  occurring 
in  animals  treated  with  albumin  and  peptone,  but  of  some 
phases  of  natural  immunity  as  well.  However,  this  subject 
will  be  more  fully  considered  in  a  future  paper  on  a  compari- 
son of  these  various  poisons.  It  may  be  well  to  reiterate 
at  this  point  that  we  have  conclusively  shown  that  the 
poison  which  we  have  been  able  to  obtain  from  the  colon 
bacillus,  and  to  which  death  is  due  in  colon  infection, 
does  not  come  directly  from  the  peptone  or  albumen  in 
the  culture  medium,  since  we  have  obtained  the  same 
poison  from  the  bacillus  when  grown  upon  a  protein-free 
medium. 

2.  Immunization  with  the  Residue  Remaining  after  the 
Extraction  of  the  Poison  from  the  Colon  Bacillus. — The  residue 
remaining  after  the  extraction  of  the  poison  from  the 
colon  bacillus,  which  is  insoluble  in  alkaline  absolute  alcohol, 
is  soluble  in  water.  The  resulting  solution  is,  however, 
quite  decidedly  alkaline  in  reaction,  owing  to  the  presence 
of  free  alkali.  Since  it  is  essential  that  we  should  avoid 
the  irritative  effects  which  would  follow  the  injection  of 
this  free  alkali  into  the  peritoneal  cavity,  the  solution  is 
first  rendered  slightly  acid  with  hydrochloric  acid,  and 
then  neutralized  with  sodium  bicarbonate  before  injection. 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        145 

The  solution  of  the  residue  thus  obtained  after  sufficient 
extraction  with  alkaline  alcohol  is  non-toxic  in  the  ordinary 
sense  of  the  term.  However,  the  toxicity  of  a  substance 
for  the  body  as  a  whole  depends  largely  upon  whether  the 
cells  which  it  attacks  are  of  fundamental  importance  in 
maintaining  the  life  of  the  animal  or  not.  Thus  a  poison 
which  possesses  a  special  affinity  for  the  cells  of  the  respira- 
tory centre  will  inevitably  lead  to  the  production  of  marked 
symptoms  of  poisoning  on  the  part  of  the  animal,  while 
one  which  exerts  its  effect  upon  the  blood  or  connective- 
tissue  cells  would  not  necessarily  do  so.  Of  course,  in  the 
latter  case,  treatment  over  a  prolonged  period  of  time  would 
undoubtedly  result  in  symptoms  of  chronic  poisoning. 
The  residue  is  as  potent  a  cell  poison  as  is  the  toxic  portion, 
but  the  cells  which  it  poisons  are  not  directly  concerned  in 
the  carrying  on  of  a  function,  the  cessation  of  which  would 
prove  immediately  fatal  to  the  organism  as  a  whole.  That 
the  residue  is  possessed  of  but  slight  toxicity  is  seen  from 
the  fact  that  the  injection  of  from  300  to  400  mg.  into  the 
peritoneal  cavity  of  guinea-pigs  at  a  single  dose  has  appar- 
ently no  effect  upon  the  animal.  There  is  no  fall  of  tem- 
perature such  as  is  observed  after  the  injection  of  the 
poisonous  portion,  nor,  on  the  other  hand,  is  there  any 
appreciable  rise.  It  may  be  well  to  emphasize  at  this  point 
that  in  order  to  study  this  portion  of  the  colon  bacillus  and 
its  action  it  is  absolutely  essential  that  the  toxic  portion 
should  have  been  completely  removed.  In  order  to  accom- 
plish this  it  is  necessary  to  extract  the  bacterial  cell  sub- 
stance at  least  three  times  with  the  alkaline  alcohol,  and 
frequently  a  fourth  extraction  is  required.  If  all  of  the 
poisonous  portion  has  not  been  removed,  the  treated  animal 
begins  to  show  evidences  of  poisoning,  as  lowering  of  tem- 
perature, stupor,  and,  provided  the  extraction  has  been 
very  imperfect,  death.  These  symptoms  do  not,  however, 
become  manifest  to  a  marked  degree  until  from  two  to 
four  hours  after  the  injection.  This  is  in  marked  contrast 
to  the  rapidity  with  which  the  free  poison  acts,  and  would 
indicate  that  the  poison  in  the  imperfectly  extracted  residue 
10 


146 


PROTEIN  POISONS 


still  exists  in  combination  with  other  constituents  of  -the 
bacterial  cell. 

The  question  now  arose  as  to  whether  the  animals  treated 
with  increasing  doses  of  the  residue  had  acquired  any 
immunity  to  infection  with  the  living  colon  bacillus.  To 
ascertain  this  point,  guinea-pigs  were  treated  with  this 
portion  and  subsequently  inoculated  with  the  living  germ 
with  the  following  results: 


TABLE  XI 

Nos.  1,  2,  3,  AND  4  RECEIVED  A  CULTURE,  1  c.c.  OF  A  12-nouR  CULTURE 
OF  WHICH   CAUSED   DEATH   IN  UNTREATED   PIGS  WITHIN   TWENTY- 
FOUR  HOURS.    THE  REST  RECEIVED  A  CULTURE,  \  c.c.  OF  WHICH 
INVARIABLY  PROVED  FATAL  TO  UNTREATED  ANIMALS. 


No.  of 

Total 

Time  between 

Guinea- 

injections 

amount. 

last  injection 

pig 

of 

received 

and  inocula- 

Amount of  culture 

No. 

residue. 

in  gm. 

tion. 

injection. 

Result. 

1 

.     9 

0.29 

16  days 

1  c.c.  16-hour  culture 

Recovery 

2 

.     9 

0.26 

15  days 

1  c.c.  24-hour  culture 

Recovery 

3 

.     9 

0.3 

16  days 

2  c.c.  16-hour  culture 

Recovery 

4 

.     2 

0.3 

4  days 

2  c.c.  16-hour  culture 

Death 

5 

.     4 

0.9 

3  days 

2  c.c.  18-hour  culture 

Recovery 

6 

.     3 

1.0 

5  days 

2  c.c.  24-hour  cultu  e 

Recovery 

7 

.      3 

1.0 

5  days 

3  c.c.  24-hour  cultu  e 

Recovery 

8 

.     4 

0.8 

7  days 

3  c.c.  20-hour  cultu  e 

Recovery 

9 

.     4 

0.9 

14  days 

3  c.c.  20-hour  cultu  e 

Recovery 

10 

.      5 

1.1 

17  days 

3  c.c.  16-hour  cultu  e 

Death 

11 

.     4 

0.8 

30  days 

3  c.c.  16-hour  culture 

Death 

12 

.     4 

0.9 

4  days 

4  c.c.  16-hour  culture 

Recovery 

13 

.     4 

0.8 

7  days 

4  c.c.  16-hour  culture 

Recovery 

14 

.     4 

0.9 

14  days 

4  c.c.  18-hour  culture 

Death 

15 

.     4 

0.9 

5  days 

5  c.c.  16-hour  culture 

Death 

16 

.     4 

0.8 

7  days 

6  c.c.  16-hour  culture 

Death 

From  the  above  table  it  will  be  seen  that  guinea-pigs 
which  have  been  treated  with  that  portion  of  the  colon 
bacillus  which  is  represented  by  the  residue  have  acquired 
an  active  immunity  to  at  least  eight  times  the  fatal  dose  of 
the  living  germ.  The  degree  of  immunity  produced  does 
not  depend  so  much  upon  the  amount  of  residue  which 
has  been  injected  as  upon  the  number  of  treatments  and 
the  interval  of  time  over  which  they  have  been  continued. 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        147 

For  example,  No.  3,  which  received  a  total  amount  of  0.3 
gram  in  nine  doses,  was  able  to  withstand  2  c.c.  of  a  sixteen- 
hour  culture  after  an  interval  of  sixteen  days,  while  No.  4, 
which  received  the  same  total  amount  in  two  doses,  suc- 
cumbed to  the  injection  of  2  c.c.  of  a  sixteen-hour  culture 
given  at  an  interval  of  four  days  after  the  last  dose  of 
residue.  Again  we  notice  that  the  length  of  time  over  which 
the  immunity  lasts  is  rather  short  in  the  case  of  animals 
which  have  received  a  large  amount  of  the  substance  in  a 
few  doses  continued  over  a  short  period.  Thus  the  immunity 
to  3  c.c.  of  a  culture,  0.5  c.c.  of  which  proved  fatal  for 
untreated  pigs,  was  lost  between  the  fourteenth  and  the 
seventeenth  day  following  the  injection  of  the  last  dose  of 
residue,  and  that  to  4  c.c.  of  the  same  culture  disappeared 
between  the  seventh  and  the  fourteenth  day. 

That  it  is  possible  to  secure  active  immunity  to  the  living 
colon  bacillus  in  rabbits  by  the  injection  of  the  colon  residue 
is  shown  in  the  following  experiments: 

Rabbit  No.  1  received  on  August  13  a  solution  which 
contained  1  gram  of  the  residue.  On  August  20  a  second 
injection  of  2  grams  was  given.  Eighteen  days  later  this 
animal  was  inoculated  with  5  c.c.  of  an  eighteen-hour  culture 
of  the  colon  bacillus.  From  two  to  four  hours  after  injec- 
tion this  rabbit  was  apparently  very  sick,  the  symptoms 
resembling  those  which  are  seen  following  the  injection 
of  the  toxic  portion.  However,  it  then  began  to  improve, 
and  eventually  completely  recovered.  A  control  inoculated 
at  the  same  time  with  an  equal  amount  of  the  same  culture 
died  in  five  hours. 

Rabbit  No.  2  received  doses  of  1  gram  of  the  residue  on 
October  5,  10,  and  20.  Four  days  after  the  last  injection 
this  animal  was  inoculated  with  5  c.c.  of  an  eighteen-hour 
culture  from  which  he  recovered.  A  control  given  the 
same  dose  at  the  same  time  died  within  five  hours. 

Rabbit  No.  3  had  the  same  treatment  as  did  No.  2,  and 
six  days  after  the  last  dose  of  residue  withstood  5  c.c.  of 
an  eighteen-hour  culture.  Eight  days  later  this  animal 
received  10  c.c.  of  a  sixteen-hour  culture  and  did  not  die. 


148  PROTEIN  POISONS 

Rabbits  No.  4  and  5  had  the  same  treatment  as  did  Nos. 
2  and  3,  and  seven  days  after  the  last  dose  each  received 
5  c.c.  of  a  twenty-four-hour  culture  of  the  living  colon 
bacillus  without  a  fatal  result.  A  control  which  received 
the  same  amount  of  the  same  culture  died  within  five  hours. 

When  we  turn  our  attention  to  the  symptoms  which 
follow  the  injections  of  living  cultures  of  the  colon  bacillus 
into  animals  which  have  been  actively  immunized  with  the 
split  products,  we  find  that  the  clinical  picture  differs 
materially  according  as  to  whether  they  have  been  treated 
with  the  toxic  portion  or  with  the  residue.  As  has  been 
previously  mentioned,  the  symptoms  which  are  observed 
in  a  pig  immunized  with  the  toxic  part  are  apparently 
identical  with  those  which  one  sees  in  a  normal  animal 
after  inoculation  with  a  non-lethal  dose  of  the  living  bacillus. 
The  picture  which  is  obtained  on  inoculation  of  animals 
rendered  immune  by  previous  treatment  with  the  residue 
is,  however,  quite  different.  In  this  case  the  animals 
become  apparently  very  ill  within  an  hour  after  inoculation 
with  the  living  germ.  Indeed,  so  noticeable  was  this  fact 
and  the  treated  pigs  appeared  so  much  sicker  than  did  the 
controls  that  our  first  thought  was  that  we  had  in  some 
manner  increased  their  susceptibility  to  subsequent  infec- 
tion by  treatment  with  the  residue.  However,  after  from 
six  to  eight  hours  the  treated  animals  appeared  in  much 
better  condition  and  eventually  recovered,  whereas  the 
controls  invariably  died.  The  temperature  of  the  treated 
animals  runs  a  course  entirely  in  accord  with  the  symptoms. 
Thus  in  a  pig  which  had  received  290  mg.  of  the  residue, 
and  subsequently  was  inoculated  with  1  c.c.  of  a  twenty-four- 
hour  culture,  the  temperature  fell  from  100°  to  95°  F.  within 
four  hours,  and  by  six  hours  had  once  more  begun  to  rise, 
as  is  illustrated  in  Fig.  9. 

The  difference  between  the  behavior  of  animals  treated 
with  the  toxic  part  and  those  which  have  been  treated  with 
the  residue  toward  cultures  of  the  living  germ  is  easily 
explained  if  we  consider  the  fact  that  in  the  first  case  we 
are  dealing  with  an  animal  which  has  acquired  a  certain 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        149 

amount  of  tolerance  for  the  intracellular  poison  of  the 
colon  bacillus  as  represented  by  the  toxic  part.  In  the 
case  of  animals  treated  with  the  residue,  however,  no 
tolerance  for  the  poison  contained  within  the  colon  bacillus 
has  been  developed.  If  now  the  process  which  takes  place 
in  both  instances  is  a  bacteriolytic  one,  it  results  that  in  the 
case  of  the  animal  immunized  with  the  toxic  group  the 
effects  of  the  poison  contained  within  the  bacterial  cell  and 
liberated  upon  its  disintegration  will  not  become  manifest 

FIG.   9 


103 
102° 
101 
100° 
99° 
98 
97° 
96° 
95° 


The  temperature  curve  of  an  animal  immunized  with  the  colon  residue  and 
afterward  inoculated  with  twice  the  lethal  dose  of  the  living  culture. 

until  a  sufficient  amount  of  poison  has  been  set  free  to 
overcome  the  tolerance  which  the  animal  has  attained 
during  the  process  of  immunization.  In  the  case  of  the 
animal  immunized  with  the  residue  there  is  no  tolerance 
to  be  overcome  other  than  that  which  is  present  in  all 
animals,  and  the  effects  of  the  poison  liberated  through 
bacteriolysis  become  apparent  sooner  and  to  a  more  marked 
extent.  Again,  the  fact  that  bacteriolysis  may  occur  more 
rapidly  in  the  case  of  residue  pigs  than  in  those  immunized 
with  the  toxic  group  might  explain  in  part  the  difference 


150  PROTEIN  POISONS 

in  behavior  in  the  two  cases.    This  is  a  point  on  which  we 
are  as  yet  unable  to  give  any  definite  results. 

In  order  to  study  the  differences  in  reaction  to  the  living 
germ  in  animals  treated  with  the  toxic  part  and  those 
immunized  with  the  residue  it  is  not  only  essential  that 
they  should  receive  the  same  amount  of  the  same  culture, 
but  the  dose  given  should  not  exceed  twice  that  which  would 
prove  fatal  for  a  control.  When  a  larger  amount  of  the 
living  culture  is  given  the  differences  are  by  no  means  so 
clearly  defined,  although  even  in  this  case  the  animal  which 
has  been  treated  with  the  residue  shows  symptoms  of 
severity  at  a  much  earlier  time.  As  can  be  seen  from  Fig. 
9,  the  temperature  of  a  residue  pig  which  had  been  inocu- 
lated with  twice  the  fatal  dose  of  a  living  colon  culture  had 
begun  to  rise  at  an  interval  of  six  hours  after  injection. 
However,  if  an  animal  which  has  been  rendered  immune 
by  treatment  with  the  residue  is  inoculated  with  six  to 
eight  times  the  fatal  dose  of  the  living  culture  we  find  that 
the  temperature  curve  obtained  is  somewhat  different  in 
character.  The  temperature  falls  with  the  same  initial 
rapidity,  but  instead  of  showing  an  early  rise  it  continues 
for  some  time  at  a  low  point  and  it  is  only  at  the  end  of 
from  eight  to  ten  hours  that  any  appreciable  rise  is  mani- 
fest. This  we  think  is  due  to  the  fact  that  there  has  not 
been  enough  of  the  bacteriolytic  substance  directly  avail- 
able to  destroy  all  the  bacilli  contained  in  the  large 
amount  of  culture  injected.  The  remainder  of  the  germs 
are  destroyed  by  the  same  factors  which  are  operative  in 
normal  animals  after  the  injection  of  a  non-fatal  dose  of 
the  living  bacillus.  As  can  be  seen  from  Fig.  8,  it  is  only 
after  an  interval  of  six  to  eight  hours  that  there  is  any 
appreciable  fall  in  temperature  in  the  case  of  a  normal 
animal  inoculated  with  a  non-fatal  dose  of  the  living  culture. 
This,  we  think,  indicates  that  it  is  not  until  this  time  that 
any  appreciable  amount  of  poison  is  liberated  by  bac- 
teriolysis since,  as  we  have  seen  in  a  previous  chapter, 
one  of  the  first  signs  of  the  action  of  the  intracellular  poison 
is  a  fall  in  body  temperature.  In  a  pig  which  has  been 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        151 

immunized  with  the  residue  and  subsequently  inoculated 
with  a  large  amount  of  the  living  germ,  we  obtain  evidence 
of  hypothermia  at  a  much  earlier  period,  owing  to  the 
fact  that  bacteriolysis  takes  place  very  rapidly  since  the 
bacteriolytic  substance  is  present  in  a  form  available 
for  immediate  use.  If,  however,  the  amount  of  this  sub- 
stance directly  available  is  not  sufficient  to  cause  death 
and  bacteriolysis  of  all  germs  present,  those  bacilli  which 
remain  are  still  capable  of  further  reproduction.  The  same 
mechanism  which  causes  destruction  of  the  bacteria  in 
normal  animals,  and  which  is  probably  connected  with 
the  phenomenon  of  phagocytosis  is,  however,  still  operative 
in  the  immune  animal.  Thus  we  shall  have  two  influences 
at  work  in  the  immune  animal  to  cause  bacteriolysis,  one 
acting  rapidly,  and  the  other  manifesting  its  action  only 
after  a  considerable  interval  of  time.  We  should  there- 
fore expect  theoretically  that  we  would  find  in  the  im- 
munized animal  a  marked  fall  in  temperature  at  an  early 
time,  due  to  the  setting  free  of  the  poison  from  the  bodies 
of  the  bacteria  disintegrated  by  the  directly  available 
bacteriolytic  substance  followed  by  a  secondary  rise,  and 
a  succeeding  fall  due  to  the  liberation  of  the  poison  by 
means  of  the  factors  present  in  the  normal  animal.  How- 
ever, this  is  not  actually  the  case,  since  the  effect  of  the 
poison  liberated  at  first  has  not  worn  off  before  the  second 
period  of  bacteriolysis  becomes  well  established.  Conse- 
quently, the  intermediate  rise  of  temperature  is  absent. 

The  results  which  follow  the  injection  of  the  dead  bac- 
terial substance  into  animals  immunized  with  the  residue 
are  very  interesting.  As  has  been  previously  mentioned, 
whereas  animals  treated  with  the  residue  develop  an  active 
immunity  to  colon  infection,  they  do  not  possess  any  greater 
degree  of  tolerance  for  the  colon  poison  than  do  untreated 
animals.  This  is  shown  by  the  fact  that  the  fatal  dose  of 
the  soluble  poison  is  the  same  for  the  treated  pig  as  for  the 
untreated  control.  This  would  lead  to  the  belief  that  the 
immunity  obtained  to  the  living  colon  bacillus  is,  in  the 
case  of  the  residue  animals,  purely  a  bacteriolytic  one.  If 


152  PROTEIN  POISONS 

this  is  true,  one  would  suppose  that  on  the  injection  of  a 
fatal  amount  of  the  dead  bacterial  substance  death  would 
occur  more  rapidly  in  the  immunized  than  in  a  normal 
pig,  provided  the  immune  animal  possesses  a  sufficient 
amount  of  bacteriolytic  substance  directly  available  to 
cause  disintegration  of  all  bacteria  present.  If,  now,  a 
pig  which  has  been  immunized  with  the  residue  receives 
5  c.c.  of  a  twenty-four-hour  culture  of  the  colon  bacillus 
which  has  been  deprived  of  life  by  means  of  heat,  the 
animal  is  very  sick  within  from  fifteen  to  twenty  minutes. 
The  symptoms  noted  are  similar  in  all  respects  to  those 
which  are  observed  after  the  injection  of  the  soluble  poison. 
The  pig  runs  about  the  cage,  scratches  itself,  and  shows 
the  same  evidence  of  lack  of  coordination  and  partial 
paralysis  of  the  hind  extremities.  This  behavior  is  in 
marked  contrast  to  that  seen  in  the  case  of  a  normal  animal 
which  has  received  an  injection  of  5  c.c.  of  a  twenty-four- 
hour  culture  which  has  been  rendered  sterile  by  means 
of  heat.  In  this  instance  the  animal  appears  perfectly 
well  until  after  the  lapse  of  about  an  hour,  when  it  begins 
to  show  signs  of  illness  such  as  roughening  of  the  coat, 
stupor,  and  indications  of  a  beginning  peritonitis.  The 
latter  symptoms  are  those  which  we  have  described  as 
being  due  to  the  slow  liberation  of  the  poison  from  a  com- 
bined state.  The  same  symptoms  are  observed  in  the  case 
of  the  immune  pig,  and  are  noticed  at  the  same  length  of 
time  after  the  injection.  The  difference  in  the  behavior 
of  the  immune  and  the  normal  pig  is  seen  to  consist  in  the 
fact  that  in  the  first  instance  we  have  symptoms  of  the 
action  of  the  free  poison  shortly  after  the  injection  of  the 
dead  culture,  which  are  entirely  lacking  in  the  second  case. 
This  shows  beyond  doubt  that  in  the  immune  pig  there  is 
marked  bacteriolysis  of  the  dead  bacilli  and  a  consequent 
liberation  of  the  contained  poison  shortly  after  the  injec- 
tion of  the  dead  bacterial  cell  into  the  peritoneal  cavity. 
Although  we  have  as  yet  been  unable  to  actually  cause 
death  in  an  immune  pig  at  an  early  period,  the  animals 
are  in  every  instance  very  ill  within  thirty  minutes  after 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        153 

the  injection  of  the  dead  culture.  In  fact,  several  of  them 
have  shown  signs  of  the  commencement  of  the  convulsive 
stage  as  evidenced  by  slight  convulsive  movements  of  the 
head  separated  by  considerable  intervals  of  time.  We  have 
been  unable  to  secure  a  fatal  result  in  these  animals  up  to 
the  present  time  simply  because  we  have  worked  with  pigs 
which  did  not  possess  a  sufficient  amount  of  bacteriolytic 
substance  directly  available  to  cause  disintegration  of 
enough  bacilli  to  liberate  a  fatal  amount  of  poison  at  one 
time.  It  is  worthy  of  note  that  this  behavior  of  animals 
immunized  with  the  residue  toward  the  dead  bacterial  sub- 
stance furnishes  additional  proof  of  the  fact  that  the  poison 
of  the  colon  bacillus  is  an  intracellular  one.  If  the  poison 
existed,  free  in  the  culture  medium  we  should  expect  that 
the  control  would  show  evidences  of  its  action  at  as  early 
a  period  as  does  the  treated  animal.  However,  as  has 
been  stated  above,  this  is  not  the  case.  The  fact  that  the 
treated  animal  shows  symptoms  of  poisoning  to  a  much 
greater  degree  and  at  an  earlier  time  than  does  the  control 
can  be  explained  only  on  the  ground  that  the  poison  with 
which  we  are  dealing  is  an  intracellular  one  and  is  set  free 
only  after  the  disintegration  of  the  bacillus  by  bacteriolysis. 
The  question  now  arose  as  to  whether  the  immunity 
induced  through  the  residue  is  specific  for  the  colon  bacillus 
or  not.  In  order  to  test  this  point,  animals  which  had  been 
treated  with  this  portion  were  inoculated  with  living  cul- 
tures of  the  typhoid  bacillus  with  the  following  results: 

TABLE  XII 

1  c.c.  OF  A  16-HOUR  CULTURE  OF  THIS  TYPHOID  BACILLUS  KILLED 

CONTROLS  WITHIN  TWENTY-FOUR  HOURS.     5  c.c. 

DID  NOT  CAUSE  DEATH 

No.  Total       Time  between 

Guinea-       injections     amount       last  injection 


pig 

of 

received 

and  inocula- 

Amountof typhoid 

No. 

residue. 

in  gm. 

tion. 

culture  injection. 

Result. 

1 

.     4 

0.8 

6  days 

1  c.c.  16-hour  culture 

Death 

2 

.     4 

0.8 

6  days 

2  c  c.  16-hour  culture 

Death 

3 

.     4 

0.8 

2  days 

2  c.c.  16-hour  culture 

Death 

4 

.     4 

0.8 

6  days 

3  c.c.  16-hour  culture 

Death 

5 

.      4 

0.8 

5  days 

4  c.c.  16-hour  culture 

Death 

154  PROTEIN  POISONS 

Although  these  experiments  are  by  no  means  sufficient 
in  extent  to  warrant  the  conclusion  that  the  injection  of 
the  residue  obtained  from  the  colon  bacillus  furnishes  no 
increased  resistance  to  typhoid  infection,  it  can  be  seen 
that  the  degree  of  immunity  established  must  be  very 
slight.  As  far  as  the  typhoid  bacillus  is  concerned,  the 
immunity  produced  by  the  colon  residue  would  appear  to 
be  specific.  If  the  immunity  induced  by  the  colon  bacillus 
is  indeed  specific,  one  would  suppose  that  the  immunizing 
group  is  one  which  is  found  only  in  the  residue  obtained 
from  the  colon  germ. 

As  has  been  previously  mentioned,  we  have  found  it 
possible  by  treatment  similar  to  that  which  we  have  used 
in  splitting  up  the  colon  bacillus  to  secure  toxic  substances 
from  egg  albumen  and  peptone,  which  possess  a  similar 
action  when  injected  into  the  animal  body,  to  that  observed 
after  the  injection  of  the  colon  poison.  We  have  also 
stated  that  the  same  transitory  immunity  to  colon  infec- 
tion followed  the  injection  of  the  albumin  and  peptone 
poison  as  was  obtained  with  the  colon  poison  itself.  The 
albumen  and  peptone  bear  a  further  resemblance  to  the 
bacterial  cell  substance  in  that  the  residue  which  remains 
after  alcoholic  extraction  is  non-toxic.  The  question  now 
arose  as  to  whether  the  injection  of  the  albumen  and  pep- 
tone residue  afford  any  immunity  to  the  living  colon  germ 
or  not.  In  order  to  ascertain  this  point  animals  were 
treated  with  gradually  increasing  doses  of  these  residues, 
and  •  subsequently  inoculated  with  the  colon  bacillus  with 
the  following  results: 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        155 


TABLE  XIII 


1     C.C.    OF    A 


16-HOUR  CULTURE  OF  COLON   BACILLUS  KILLED    CONTROLS 
IN  TWENTY-FOUR  HOURS.     5  c.c.  DID  NOT  KILL 

Peptone  Residue 


Total 

Time  between 

Guinea- 

No.  injec- 

amount 

last  injection 

pig 

tions  of 

received 

and  inocu- 

Amount of  colon 

No. 

residue. 

in  gm. 

lation. 

culture  injection. 

Result. 

1 

.     4 

0.9 

3  days 

2  c.c.  16-hour  culture 

Death 

2 

.     4 

0.9 

4  days 

2  c.c.  16-hour  culture 

Death 

3 

.     4 

0.9 

6  days 

2  c.c.  16-hour  culture 

Death 

4 

.     4 

0.9 

3  days 

2  c.c.  18-hour  culture 

Recovery 

5 

.     4 

0.9 

4  days 

2  c.c.  18-hour  culture 

Recovery 

6 

.     4 

0.9 

5  days 

2  c.c.  18-hour  culture 

Death 

7 

.     4 

0.9 

4  days 

2  c.c.  16-hour  culture 

Death 

8 

.     4 

0.9 

6  days 

3  c.c.  16-hour  culture 

Death 

Guinea- 
pig 

Total 
No.  injec-   amount 
tions  of     received 

TABLE  XIV 
Albumin  Residue 
Time  between 
last  injection 
and  inocu-              Amount  of  colon 

No. 

residue. 

in  gm. 

lation. 

culture  injection. 

Result, 

1 

.     4 

0.9 

3  days 

1  c.c.  16-hou 

culture 

Death 

2 

.     4 

0.9 

5  days 

1  c.c.  16-hou 

culture 

Death 

3 

.     4 

0.9 

3  days 

2  c.c.  16-hou 

culture 

Death 

4 

.     4 

0.9 

4  days 

2  c.c.  18-hou 

culture 

Death 

5 

.     4 

0.9 

5  days 

2  c.c.  16-hou 

culture 

Death 

6 

.     4 

0.9 

5  days 

3  c.c.  16-hou 

culture 

Death 

7 

.     4 

0.9 

4  days 

4  c.c.  16-hou 

culture 

Death 

From  the  above  tables  it  can  be  readily  seen  that  the 
residues  obtained  from  the  peptone  and  albumin  possess 
little  if  any  immunizing  properties  against  infection  with 
the  colon  bacillus.  In  the  case  of  the  animals  treated  with 
the  peptone  residue,  the  first  three  pigs  received  a  different 
residue  from  that  given  to  the  remainder.  This  residue 
had  been  thoroughly  extracted  with  alkaline  alcohol,  and 
was  evidently  possessed  of  no  immunizing  properties  what- 
ever. On  the  other  hand,  the  residue  which  was  received 
by  pigs  No.  4  to  No.  8  inclusive  had  not  been  subjected  to 
so  thorough  an  extraction.  It  is  therefore  highly  probable 
that  the  slight  degree  of  immunity  apparently  obtained 
in  some  of  the  latter  animals  was  due  to  the  presence  of 


156  PROTEIN  POISONS 

some  of  the  toxic  portion  which  had  been  left  in  the  residue 
as  the  result  of  incomplete  extraction.  In  the  animals 
treated  with  the  albumen  residue  we  were  unable  to  obtain 
any  evidence  whatever  of  increased  resistance  to  colon 
infection. 

It  has  now  been  shown  that  active  immunity  to  the 
colon  germ  can  be  produced  in  animals  by  treatment  with 
the  split  products  obtained  from  this  bacillus.  This  would 
seem  to  furnish  conclusive  evidence  that  there  exist  within 
the  colon  bacillus  certain  immunizing  groups  wrhich  are 
capable  of  being  separated  more  or  less  completely  from 
the  other  constituents  of  the  bacterial  cell  by  means  which 
bring  about  a  chemical  cleavage  of  the  latter.  Furthermore, 
it  has  been  seen  that  the  colon  bacillus  contains  at  least 
two  different  groups,  each  of  which  when  injected  into  the 
animal  body  is  capable  of  establishing  a  certain  degree  of 
immunity  toward  subsequent  infection  with  the  living 
germ.  One  of  these  groups  is  contained  within  the  toxic 
portion,  and  probably  represents  a  group  which  is  common 
to  many  protein  bodies,  since,  as  has  been  shown,  it  is 
contained  in  the  poisons  secured  through  the  chemical 
cleavage  of  egg  albumen  and  peptone,  as  well  as  from  the 
colon  bacillus.  The  degree  of  immunity  thus  far  obtained 
through  the  agency  of  this  group  is  not  great.  The  fact 
that  this  group  is  apparently  not  specific  to  the  colon 
bacillus,  but  can  also  be  obtained  from  other  protein  bodies, 
furnishes  an  explanation  of  the  increased  resistance  to 
infection  observed  in  animals  previously  treated  with 
solutions  of  egg  albumen  and  peptone.  Again  it  has  been 
shown  that  the  residue,  which,  as  has  been  previously 
stated,  is  for  all  practical  purposes  non-toxic,  also  contains 
an  immunizing  group.  The  immunizing  group  contained 
in  the  residue  differs  from  that  found  in  the  toxic  portion 
in  one  very  important  respect.  It  represents  a  group 
which  up  to  the  present  time  we  have  been  able  to  find 
only  in  the  colon  bacillus,  and  which  when  injected  into 
animals  affords  protection  against  this  bacillus  alone.  In 
other  wrords,  the  immunity  produced  with  the  residue  is 
strictly  specific  in  character.  Moreover,  the  degree  of 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        157 

immunity  to  the  living  germ  obtained  through  the  employ- 
ment of  the  residue  is  apparently  much  higher  than  that 
which  follows  treatment  with  the  toxic  portion.  It  does 
not  seem  improbable  that  the  specific  immunizing  group 
which  is  contained  in  the  residue  represents  that  group  of 
the  colon  bacillus  which  is  of  primary  importance  in  the 
development  of  specific  acquired  immunity  to  this  germ. 
The  work  just  detailed  may  be  summed  up  as  follows: 

1.  Guinea-pigs    treated    at    intervals    of   from   three    to 
four   days   with   intra-abdominal   injections   of   the   colon 
residue  acquire  an  active  immunity  to  at  least  eight  times 
the  ordinary  fatal  dose  of  the  living  bacterium. 

2.  The  degree  of  immunity  secured  does  not  depend  so 
much  upon  the  amount  of  the  residue  or  non-poisonous 
portion  that  has  been  injected   as   upon  the  number  of 
treatments  and  the  interval  of  time  over  which  they  have 
been  continued. 

3.  The  length  of  time  over  which  the  immunity  continues 
is  rather  short  in  the  case  of  animals  that  have  received  a 
large  amount-  of  the  residue  in  a  few  doses  continued  over  a 
short  period. 

4.  Rabbits  that  received  from  two  to  three  injections  of 
0.5   gram   each   of   the   residue   acquire   an   immunity   to 
quantities  of  the  living  bacillus  that  kill  the  controls  within 
five  hours. 

5.  The  immunity  induced  by  the  colon  residue  or  non- 
poisonous    part    is    specific,    and    previous    treatments    of 
animals  with  the  residues  of  egg-white,  peptone,  and  the 
typhoid  bacillus  give  no  immunity  to  the  colon  bacillus. 

In  continuing  this  work  Vaughan  and  Wheeler1  decided 
in  the  first  place  to  ascertain  whether  or  not  a  single  dose 
of  the  residue  gives  any  immunity;  if  so,  what  degree  of 
immunity  does  it  afford  and  how  long  does  it  continue? 

The  number  of  immunizing  doses  given  in  the  work 
already  reported  ran  from  three  to  nine,  and  the  special 
object  of  the  work  here  reported  is  to  ascertain  the  effects 
of  a  smaller  number  of  immunizing  doses. 

1  New  York  Med.  Jour.  June,  29,  1907. 


158  PROTEIN  POISONS 

TABLE  XV 

These  animals  had  one  dose  of  50  mg.  of  the  residue, 
as  the  non-poisonous  portion  is  designated.  The  protocol 
number,  the  weight  of  the  animal,  the  interval  in  days 
between  the  administration  of  the  residue  and  the  inocula- 
tion with  the  twenty-four  hours  beef-tea  culture  of  the 
bacillus,  the  amount  of  the  culture  and  the  result  are  shown : 

Protocol,  Weight,  Interval,  Amount  given, 

No.                      gm.  days.                     c.c.  Result. 

324  400  1                          5  Recovery 

325  420  3                          5  Recovery 

269  505  4  3  Recovery 
275                         300                          4                          5                          Recovery 

326  315  7  5  Death 

270  600  9  3  Recovery 
328                         345                        11                           2  Recovery 
207                         455                        21                          2                          Recovery 

209  495  27  4  Death 

210  625  27  4  Death 

211  485  28  3  Death 

139  240  36  2  Recovery 

140  250  40  3  Death 

141  250  42  2  Death 

142  255  42  2  Death 

In  studying  these  results  it  will  be  well  to  consider  the 
minimum  fatal  dose  as  the  unit,  which  in  case  of  the  cultures 
used  in  these  experiments  is  1  c.c.  of  the  twenty-four  hour 
beef -tea  growth,  and  we  will  regard  the  animal  that  suc- 
cumbs to  2  c.c.  as  having  practically  lost  its  immunity. 
With  this  measure  it  will  be  seen  that  a  single  dose  of  50 
mg.  of  the  colon  residue  gives  to  the  animal  a  temporary 
immunity  of  at  least  5  units,  which  is  in  force  twenty-four 
hours  after  the  treatment,  and  continues  for  at  least  four 
days,  but  has  begun  to  disappear  by  the  seventh  day. 
However,  some  slight  degree  of  immunity  continues  up  to 
the  thirty-sixth  day,  but  practically  all  is  lost  by  the 
fortieth  day. 

TABLE  XVI 

These  animals  had  a  single  dose  of  25  mg.  of  the  residue. 
The  data  are  the  same  as  given  in  the  preceding  table. 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        159 

Protocol,  Weight,  Interval,       Amount  given, 

No.  gm.  days.  c.c.  Result. 

330  460  1  5.0  Recovery 

331  300  3  5.0  Death 

332  225  7  5.0  Death 

333  290  9  2.5  Recovery 

334  235  11  2.0  Death 

335  360  14  1.0  Recovery 

Comparing  Tables  XV  and  XVI,  it  will  be  seen  that  the 
immunity  given  by  25  mg.  of  the  residue,  although  it  may 
be  as  great  as  that  given  by  50  mg.  at  the  end  of  the  first 
twenty-four  hours,  declines  more  rapidly  and  is  less  at  the 
end  of  three  days,  and  continues  to  be  less  at  eleven  days. 
The  only  element  of  doubt  that  we  can  see  in  these  con- 
clusions lies  in  the  small  size  of  all  the  animals,  save  No.  1 
used  in  Table  XVI.  However,  we  have  not  found  that  size 
or  weight  of  guinea-pigs  are  important  factors,  provided, 
of  course,  that  the  animals  are  in  good  condition,  in  influ- 
encing the  result  after  inoculation  with  the  colon  or  the 
typhoid  bacillus. 

TABLE  XVII 

These  animals  had  from  two  to  three  treatments,  receiving 
each  time  50  mg.  of  the  residue.  These  treatments  were 
at  intervals  of  three  days.  The  protocol  number,  the  weight 
of  the  animals,  the  number  of  treatments,  the  total  amount 
of  residue  given,  the  interval  in  days  between  the  last 
treatment  and  the  inoculation,  the  amount  of  the  culture 
twenty-four  hours  old,  and  the  result  are  given: 

Amount  of  Amount  of 

Protocol,          Weight,         No.  of         residue,  Interval,      culture, 

No.                  gm.        treatments.        mg.  days.             c.c.                 Result. 

270  600                 2                  100  33                 Recovery 

271  600                  2                  100  34                 Recovery 

212  520  2  100  33  Recovery 

213  530  2  100  3  4  Recovery 

272  530  3  150  5  4  Recovery 

273  540  3  150  5  5  Recovery 

274  475  3  150  7  6  Recovery 
278  580  3  150  7  6  Recovery 
280  615  3  150  7  6  Recovery 

214  585  3  150  12  5  Recovery 

215  600  3  150  12  6  Recovery 

216  485  3  150  12  6  Recovery 

217  530  3  150  12  6  Recovery 

218  475  3  150  12  7  Recovery 


160  PROTEIN  POISONS 

Comparing  Table  XVII  with  Tables  XV  and  XVI,  it  is 
plainly  evident  that  this  immunity  induced  by  two  and  three 
doses  at  intervals  of  three  to  four  days  is  greater  in  degree 
and  more  lasting  in  its  effects  than  that  produced  by  a 
single  injection.  This  confirms  the  conclusion  already 
stated,  but  at  the  same  time  this  additional  work  shows 
that  a  single  dose  may  serve  to  furnish  protection  against 
at  least  five  times  the  ordinary  fatal  dose  for  a  few  days 
and  against  twice  the  fatal  dose  for  one  month. 

TABLE  XVIII 

These  animals  received  a  single  dose  of  100  mg.  of  the 
typhoid  residue.  The  protocol  number,  weight  of  animal, 
interval  between  treatment  and  inoculation,  amount  of 
twenty-four-hour  culture  given,  and  the  result  are  shown: 

Amount  of 

Protocol,  Weight,  Interval,  culture, 

No.  ..  .         gm.  days.  c.c.  Result. 

354  250  1  3  Recovery 

355  320  1  4  Death  on  second  day 

356  280  3  3  Death 

357  265  3  4  Death 

358  270  6  2  Death 

359  230  6  1  Recovery 
219  570  28  2  Death 


TABLE  XIX 

These  animals  received  a  single  dose  of  50  mg.  of  the 
typhoid  residue. 

Amount  of 
Protocol,  Weight,          Interval,          culture, 

No.  gm.  days.  c.c.  Result. 

336  360         1         3  Recovery 

337  285         3         3  Death 

338  245         7         3  Death 

339  350         9         2  Death  on  second  day 

340  290        11         1  Recovery 

341  255        14         1  Recovery 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        161 

TABLE  XX 

These  animals  received  two  and  three  immunizing  doses 
of  the  typhoid  residue.  The  protocol  number,  the  weight, 
the  number  of  immunizing  doses,  the  interval  in  days  between 
the  last  treatment  and  the  inoculation,  the  amount  of 
culture  twenty-four  hours  old,  and  the  result  are  given: 


Protocol,  Weight,  No.  of  residue,  Interval,      culture, 

No.  gm.  treatments.  mg.  days.             c.c.  Result. 

220  375  2  100  32  Recovery 

221  495  2  100  33  Death 

222  650  3  150  52  Recovery 

223  480  3  150  5  3  Recovery 

224  435  3  150  53  Recovery 

144  605  3  150  74  Recovery 

145  665  3  150  13  4  Recovery 

The  minimum  fatal  dose  of  the  twenty-four-hour  cul- 
ture of  the  typhoid  bacillus  employed  in  this  case  is  0.5 
c.c.  It  will  be  seen  from  Table  XVIII  that  a  single  dose  of 
100  mg.  of  the  residue  gives  the  animal  an  immunity  of 
six  units  at  the  end  of  twenty-four  hours,  and  that  the 
immunity  was  less  than  eight  units  at  that  time.  On  the 
third  day  the  immunity  was  found  to  be  diminishing,  but 
on  the  sixth  day  the  animal  bore  twice  the  fatal  dose.  The 
animal  which  received  eight  units  at  the  end  of  the  first 
day  evidently  was  close  to  the  borderline,  because  it  did 
not  die  until  the  second  day,  and  the  normal  time  for  an 
untreated  guinea-pig  to  live  after  receiving  the  minimum 
fatal  dose  is  less  than  twelve  hours.  Table  XIX  shows  that 
a  single  dose  of  50  mg.  is  quite  as  efficient  as  one  of  100  mg. 
Table  XX  indicates  that  multiple  immunizing  doses  give 
a  higher  degree  and  a  more  lasting  immunity  than  that 
secured  by  a  single  dose. 

Theoretical  Considerations  and  Conclusions. — We  wish  to 
offer  certain  theories  that  we  have  reached  after  making 
these  experiments.    In  order  to  save  space  we  will  condense 
our  views  as  follows: 
11 


162  PROTEIN  POISONS 

1.  All  the  proteins  with  which  we  have  worked  contain 
a  poisonous  group,  and  the  probabilities  are  that  this  is 
true  of  all  proteins,  be  they  bacterial,  vegetable,  or  animal. 

2.  Proteins  may  be  split  into  poisonous  and  non-poisonous 
groups,  either  artificially  in  the  retort  or  in  the  animal 
body. 

3.  The  splitting  up  of  the  protein  in  the  animal  body  is 
due  to  a  proteolytic  ferment  which  is  the  product  of  certain 
cells. 

4.  This  ferment  is  specific  for  the  protein  which  calls  it 
into  existence. 

5.  Our  conception  of  the  origin  and  nature  of  these  specific 
ferments  is  as  follows:    The  cell  is  made  up  of  molecules; 
the  molecules  consist  of  atoms,  and  the  atoms  of  electrons. 
The  molecule  may  be  likened  to  the  universe,  composed 
of  suns,  planets,  and  satellites.     These  are  in  harmonious 
and  rhythmic  motion.    The  molecule  of  the  foreign  protein 
introduced  into  the  body  has  a  structure  similar  to  that 
of  the  cell  molecule,  and  when  one  is  brought  within  the 
attractive  range  of  the  other,  one  or  the  other,  or  both, 
must    undergo    certain    disturbances.      Suppose    that    an 
atomic  group  is  split  off  from  the  animal  cell  and  enters 
the  attraction  sphere  of  the  molecule  of  the  foreign  protein, 
then  the  harmonious  arrangement  of  the  atoms  and  elec- 
trons of  the  latter  will  be  affected ;  indeed,  the  molecule  may 
be  disrupted  as  completely  as  it  is  in  the  retort  under  the 
influence  of  dilute  alkali. 

Our  residues  evidently  have  the  same  effect  on  the  cells 
of  the  body  that  the  proteins  from  which  they  come  do, 
and  .in  this  way  we  may  explain  the  specific  action  of  the 
residues.  The  special  group  broken  off  from  the  cell  mole- 
cule depends  upon  the  composition  of  the  protein  which 
comes  within  its  attractive  sphere,  and  it  seems  that  our 
residues  contain  that  portion  of  the  molecule  of  the  foreign 
protein  which  possesses  this  property  of  bringing  into 
existence,  or  rather  of  activating,  its  own  specific  ferment. 
The  ferment  is,  according  to  our  conception,  a  portion  of 
the  animal  cell,  an  atomic  group  within  the  cell  molecule, 


THE  PRODUCTION  OF  ACTIVE  IMMUNITY        163 

and  does  not  become  a  real  active  ferment,  or  it  is  not 
activated  until  the  foreign  protein  comes  within  its  sphere 
of  attraction.  It  occurred  to  us  that  if  this  theory  has 
much  of  truth  in  it  we  might  test  it.  We  thought  that  the 
introduction  of  a  small  portion  of  the  residue  might  give 
some  immunity  immediately.  We  therefore  injected  doses 
of  25  mg.  of  the  colon  and  12.5  mg.  of  the  typhoid  residue 
into  the  abdominal  cavity  of  guinea-pigs,  and  thirty  min- 
utes later  inoculated  these  animals  intra-abdominally  with 
living  cultures,  and  found  that  the  colon  animals  had  in 
that  short  time  acquired  an  immunity  of  five  units  and 
the  typhoid  one  of  six  units.  However,  we  found  that 
larger  immunizing  doses  did  not  give  us  results  so  good, 
and  this  is  easily  explained  by  supposing  that  this  ferment 
set  free  or  activated  by  the  residue  is  in  part  used  up 
in  its  reaction  with  the  residue  itself.  The  reason  multiple 
doses  repeated  at  intervals  give  us  a  higher  degree  of  im- 
munity than  single  doses  may  be  due  to  more  cells  being 
acted  upon  or  to  the  accumulation  of  the  ferment  in  the 
blood.  The  theory  of  the  modus  operandi  of  the  residue 
which  we  have  offered  is  tentative,  and  we  hope  to  be  able 
to  investigate  it  further. 

It  will  undoubtedly  occur  to  the  reader,  as  it  has  to  us, 
to  ask  the  question  how  it  is  that  the  residue  sensitizes  or 
activates,  while  the  bacillus  itself,  living  or  dead,  has  no  such 
effect  or  at  least  is  not  nearly  so  effective.  The  only  answer 
that  we  can  suggest  to  this  question  is  that  in  order  to  be 
effective  in  its  action  the  sensitizer  must  be  in  solution, 
and  that  being  in  this  state  it  reaches  every  part  of  the 
circulatory  system  in  a  few  seconds.  Possibly  cell  permea- 
tion may  be  necessary  to  the  most  perfect  sensitization. 


CHAPTER  VIII 

THE  SPLIT  PRODUCTS  OF  THE  TUBERCLE 

BACILLUS  AND  THEIR  EFFECTS 

UPON  ANIMALS1 

The  Organism. — The  tubercle  bacillus  employed  in  the 
experimental  work  herewith  reported  is  one  which,  after 
having  been  grown  for  many  years  on  artificial  culture- 
media,  has  lost  its  virulence  for  rabbits  and  guinea-pigs. 
We  have  repeatedly  demonstrated  this  fact  during  the 
past  six  years,  but  in  order  to  have  renewed  evidence  we 
have,  at  the  beginning  of  this  research,  inoculated  4  rabbits 
and  5  guinea-pigs  intra-abdominally  with  loops  of  the 
glycerin  beef-tea  culture,  and  these  animals  having  been 
killed  from  three  to  six  months  after  inoculation,  have  in 
no  instance  showed  any  evidence  of  infection.  On  artificial 
culture-media  this  bacillus  grows  abundantly,  and  in  this 
respect  it  has  served  our  purpose  in  furnishing  a  large 
amount  of  cellular  substance.  It  has  been  grown  in  the 
ordinary  glycerin  beef-tea  and  has  been  harvested  after 
periods  of  from  one  to  six  months.  Growths  obtained 
after  from  one  to  two  months  at  37°  have  given  us  the 
most  satisfactory  material. 

The  Cellular  Substance. — The  bacterial  substance  is 
collected  on  hard  filters,  dried  between  folded  filters,  and 
thoroughly  extracted,  first  with  alcohol  and  then  with 
ether,  in  large  Soxhlets.  In  this  paper  we  will  say  nothing 
concerning  the  fats  and  waxes  extracted  with  alcohol  and 
ether.  The  cellular  substance  is  next  rubbed  up  in  a  mortar 
and  passed  through  a  fine-meshed  sieve.  As  thus  prepared, 

1  The  first  part  of  this  chapter  is  taken  from  a  paper  read  by  Vaughan 
and  Wheeler  before  the  International  Congress  on  Tuberculosis  in  1908. 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     165 

the  powder  shows  the  bacilli  more  or  less  broken  into  a 
cellular  debris  when  examined  microscopically.  The  indi- 
vidual bacilli  take  the  carbolic  stain,  but  this  is  now 
washed  out  with  dilute  nitric  acid. 

The  Cleavage  of  the  Cell. — The  cellular  substance,  pre- 
pared as  stated  above,  is  placed  in  large  flasks,  fitted  with 
reflux  condensers,  covered  with  from  fifteen  to  twenty 
times  its  weight  of  absolute  alcohol,  in  which  2  per  cent, 
of  sodium  hydroxide  has  been  dissolved,  and  heated  for 
one  hour  at  78°,  the  boiling-point  of  absolute  alcohol. 
Three  successive  extractions  are  made,  using  a  new  portion 
of  alkaline  alcohol  each  time.  This  treatment  splits  the 
cellular  substance  into  two  portions — one  soluble  and  the 
other  insoluble  in  absolute  alcohol.  These  portions  we  will 
designate  as  the  "cell  poison"  and  the  "cell  residue." 

The  Cell  Poison. — This  is  soluble  in  the  alcohol,  which  is 
carefully  neutralized  with  hydrochloric  acid.  The  precipi- 
tated sodium  chloride  is  removed  by  filtration  and  the 
filtrate  containing  the  poison  is  evaporated  in  vacua  at  or 
below  40°.  This  leaves  the  cell  poison  as  a  brownish  mass 
containing  a  small  amount  of  sodium  chloride,  which  can 
be  removed  by  repeated  solutions  in  absolute  alcohol  and 
evaporation.  The  poison  resembles  that  obtained  from  other 
protein  bodies.  It  is  freely  soluble  in  absolute  alcohol ;  less 
freely  in  water.  Its  aqueous  solutions  give  all  the  color  pro- 
tein reactions,  including  that  of  Molisch,  which  is  not  given 
by  the  poisonous  groups  that  we  have  obtained  from  other 
proteins.  In  powder  form  it  is  deliquescent  and  becomes 
darker  as  it  absorbs  water.  The  tubercle  protein  apparently 
contains  much  less  poison  than  the  cellular  proteins  of  the 
colon  and  typhoid  bacilli.  The  latter  are  split  up  by  our 
method  into  about  one-third  poison  and  two-thirds  residue, 
while  25  grams  of  the  cellular  substance  of  the  tubercle 
bacillus  yielded  less  than  3  grams  of  poison. 

The  Cell  Residue. — This  is  the  portion  insoluble  in  the 
alkaline  alcohol.  It  is  placed  in  Soxhlets  and  extracted  for 
many  hours  with  absolute  alcohol  in  order  to  remove 
traces  of  the  cell  poison  and  free  alkali.  After  this  it  is 


166  PROTEIN  POISONS 

dried  and  powdered.  It  is  partially  soluble  in  water,  and 
the  soluble  part  constitutes  that  which  we  have  used  in  our 
experiments. 

The  Bacterial  Filtrate. — After  the  bacilli  have  been 
removed,  the  culture  medium  is  concentrated  on  a  steam- 
bath  to  one-sixth  its  volume.  This  concentrated  fluid  is 
poured  into  five  times  its  volume  of  absolute  alcohol,  which 
throws  down  a  heavy,  sticky  precipitate.  This  precipitate 
is  placed  in  Soxhlets  and  extracted,  first  with  alcohol  and 
then  with  ether.  Next,  it  is  powdered  and  split  up  with 
alkaline  alcohol  after  the  method  used  with  the  cellular 
substance.  This  breaks  it  up  into  poisonous  and  non- 
poisonous  groups,  which  we  distinguish  from  the  corre- 
sponding bodies  obtained  from  the  cellular  substance  by 
designating  them  as  "the  precipitate  poison"  and  "the 
precipitate  residue." 

The  Precipitate  Poison. — This  differs  in  none  of  its  physical 
or  chemical  properties,  so  far  as  we  have  investigated,  from 
the  cell  poison. 

The  Precipitate  Residue. — This  is  freely  and  wholly  soluble 
in  water.  It  gives  all  the  protein  reactions  and  is  precipitated 
by  uranyl  acetate  and  metaphosphoric  acid. 

The  Final  Filtrate.- — In  this  manner  we  have  designated 
that  portion  of  the  culture-medium  that  remains  after  the 
concentrated  medium  has  been  precipitated  by  five  times 
its  volume  of  absolute  alcohol.  The  alcoholic  filtrate  gives 
a  voluminous  precipitate  with  an  alcoholic  solution  of 
mercuric  chloride,  showing  that  all  the  protein  material 
has  not  been  precipitated  by  the  alcohol.  This  filtrate  is 
freed  from  alcohol  by  distillation  and  has  been  used  in 
some  of  the  animal  experiments  described  later. 

It  will  be  seen  that  we  have  split  up  the  tubercle  cell 
into  two  portions:  the  cell  poison  and  the  cell  residue. 
The  culture-medium  has  been  concentrated  and  then  pre- 
cipitated with  five  times  its  volume  of  absolute  alcohol, 
and  this  precipitate  has  been  broken  up  into  two  portions: 
the  precipitate  poison  and  the  precipitate  residue,  and  the 
portion  of  the  culture-medium  left  after  the  removal  of 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS      167 

the  alcoholic  precipitate  we  have  designated  as  the  final 
filtrate. 

The  Effect  of  the  Cellular  Substance  on  Animals. — It  must 
be  borne  in  mind  that  the  cellular  substance  with  which 
we  are  now  dealing  is  that  of  a  tubercle  bacillus  that  is 
avirulent  to  rabbits  and  guinea-pigs  and  that  it  has  been 
thoroughly  extracted  with  alcohol  and  ether.  There  remains, 
as  it  were,  only  the  protein  skeleton  of  the  bacillus. 

We  have  injected  into  the  abdominal  cavities  of  twenty- 
four  guinea-pigs  single  doses,  varying  in  amount  from  5 
to  200  mg.  of  the  cellular  substance,  and  from  these  experi- 
ments we  make  the  following  statements: 

1.  In  no  case  was  death  caused  directly  by  the  injection. 
One  pig  that  received  20  mg.  was  found  dead  six  days 
later.    There  were  several  caseous  nodules  in  the  omentum 
and  one  on  the  under  surface   of   the   liver.     Microscopic 
examination    showed    that    these    consisted    of    masses    of 
leukocytes  and  the  debris  of  the  injected  bacilli.     Another 
that  received  5  mg.  was  found  dead  nine  days  later,  but 
careful  search  failed  to  reveal  any  traces  or  effect  of  the 
injection.     Animals  that  received  from   100  to  200  mg. 
remain  apparently  well  four  months  after  the  injection. 

2.  It  gives  in  guinea-pigs  no  immunity  to  a  subsequent 
inoculation  with  a  virulent  bacillus.     Six  pigs  that  had 
received  single  intra-abdominal  injections  of  the  cellular 
substance  in  amounts  varying  from   15  to  200  mg.  were 
inoculated  one  month  later  with  a  loop  of  a  virulent  culture 
of  bacillus  tuberculosis  and  all  developed  tuberculosis  and 
died  from  it  within  from  nineteen  to  one  hundred  days. 

3.  It  does,  for  a  short  time  at  least,  sensitize  guinea-pigs 
to  the  tuberculosis  bacillus.     This  is  an  interesting  and, 
in  our  opinion,  a  hopeful  point.     The  following  are  illus- 
trations of  this  action:    Pig  No.  159,  weight  530  grams, 
received,  December  18,  25  mg.  of  the  cellular  substance. 
Thirteen  days  later  it  was  given  intra-abdominally  a  large 
loop  of  the  avirulent  culture  suspended  in  salt  solution. 
The  animal  was  sick  within  a  few  minutes.     Within  half 
an  hour  it  developed  the  first  and  second  stages  of  anaphyl- 


168  PROTEIN  POISONS 

axis.  Within  forty-five  minutes  its  rectal  temperature 
had  fallen  to  96°  F.  and  it  was  found  dead  the  next  morning. 
Postmortem  showed  a  hemorrhagic  peritonitis.  Pig  No.  163, 
weight  535  grams,  received  45  mg.  of  the  cellular  substance 
December  18.  Its  subsequent  treatment  and  its  results 
were  the  same  as  recorded  of  the  preceding  animal.  Pig. 
No.  151,  weight  555  grams,  received,  December  18,  125 
mg.  of  the  cellular  substance.  Twenty-three  days  later 
it  had  intra-abdominally  a  large  loop  of  the  avirulent 
culture.  It  died  within  sixteen  hours  and  showed  a 
hemorrhagic  peritonitis. 

If  we  interpret  these  results  correctly  we  infer  that  the 
cellular  substance  had  sensitized  these  animals  and  that 
the  bacilli  of  the  second  dose  were  broken  up  so  rapidly  and 
their  poisonous  constituents  set  free  so  speedily  that  the 
animals  died.  If  this  interpretation  be  correct,  there 
remains  at  least  the  possibility  that  there  may  be  found 
in  the  bacillary  substance  some  constituent  that  may 
stimulate  the  cells  of  the  animal  body  to  split  up  and 
destroy  tubercle  bacilli.  We  will  return  to  this  before 
we  close. 

The  Effect  of  the  Cell  Poison  on  Animals. — This  body, 
obtained  by  splitting  up  the  cellular  substance  with  alkali 
in  absolute  alcohol,  is,  like  all  similar  bodies  that  we  have 
obtained  from  bacterial,  vegetable,  and  animal  proteins, 
a  poison.  It  develops  the  three  stages  of  peripheral 
irritation,  partial  paralysis,  and  terminal  convulsions. 
When  given  in  sufficient  quantity  it  kills  within  an  hour 
both  healthy  and  tuberculous  animals.  When  given  to 
healthy  animals  in  very  small  repeated  doses  it  has  no 
visible  effect.  In  larger  repeated  doses  it  causes  in  healthy 
animals  a  condition  of  chronic  intoxication  characterized 
by  loss  of  flesh  and  general  marasmus.  When  given  even 
in  very  small  repeated  doses  to  tuberculous  animals  it 
intensifies  the  tuberculous  process,  and  in  all  cases  the 
treated  animals  die  before  the  controls.  There  is  no  evi- 
dence that  it  elaborates  any  antitoxin,  and  it  is  harmful,  so 
far  as  our  experiments  show,  and  we  have  made  many  with 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     169 

this  body  upon  both  normal  and  tuberculous  animals;  it 
has  nothing  to  recommend  it.  What  is  true  of  the  cell 
poison  is  equally  true  of  the  precipitate  poison  and  the 
final  filtrate.  The  effects  of  these  poisons  on  animals  are 
harmful — only  harmful. 

The  Effects  of  the  Cell  Residue  on  Animals. — This  is  the 
non-poisonous  group  obtained  by  splitting  up  the  cellular 
substance  with  alkali  in  absolute  alcohol.  On  healthy 
animals  it  has  no  recognizable  ill  effect,  either  in  single  or 
repeated  doses,  either  large  or  small.  In  this  product  we 
see  the  one  small  ray  of  hope  of  finding,  among  the  split 
products,  a  body  that  may  possibly  be  of  service  in  the 
treatment  of  incipient  and  localized  tuberculosis.  Before 
giving  the  basis  of  this  slight  hope  we  will  tell  what  we  have 
done  with  this  product. 

In  the  first  place,  it  sensitizes  guinea-pigs  to  the  tubercle 
bacillus.  The  following  are  illustrations:  Guinea-pig  No. 
199,  weight  530  grams,  received,  December  20,  50  mg.  of 
the  residue.  Thirteen  days  later  it  had  intra-abdominally  a 
large  loop  of  the  avirulent  culture  (the  same  amount  given 
to  pigs  159,  163,  and  151,  see  p.  167).  This  injection  was 
made  at  11.30  A.M.  At  12  M.  the  temperature  had  fallen 
to  96°  F.  At  2.30  P.M.  it  was  97.1°,  and  at  5  P.M.  it  was 
99°,  and  the  animal  was  apparently  well. 

Pig  No.  200,  weight  530  grams,  received  50  mg.  of  the 
residue  intra-abdominally.  Thirteen  days  later  it  had 
intra-abdominally  one  large  loop  of  the  avirulent  culture. 
The  rectal  temperature  before  the  injection  was  101°  F. 
Within  half  an  hour  it  had  fallen  one  degree,  but  went  no 
lower  and  the  animal  seemed  to  be  but  little  disturbed. 
We  have  similar  records  of  other  animals. 

We  compare  these  results  with  those  recorded  of  animals 
159,  163,  and  151,  and  tentatively  conclude  that  the  sensi- 
tizing agent  in  the  cellular  substance  is  the  portion  that 
we  have  designated  as  the  residue.  But  the  avirulent 
bacilli  killed  the  animals  sensitized  with  the  cellular  sub- 
stance because  when  the  bacteriolytic  ferment  was  set 
free  or  activated  by  the  second  injection,  it  split  up  not 


170  PROTEIN  POISONS 

only  the  bacilli  introduced  by  the  second  injection,  but  also 
those  remaining  in  the  body  from  the  first  injection  and 
these  together  supplied  enough  free  poison  to  kill. 

It  will  require  much  experimentation  to  show  what  degree 
of  sensitization  can  be  secured  by  the  residue,  what  size 
doses  should  be  used,  and  how  long  the  condition  of  sensi- 
tization continues.  If  men,  as  well  as  guinea-pigs,  can 
be  sensitized  with  the  residue,  there  is  the  possibility  that 
it  may  be  of  service  in  the  treatment  of  initial  and  localized 
tuberculosis,  because  it  may  be  used  to  bring  into  existence 
and  activate  a  specific  bacteriolytic  ferment  which  will 
split  up  and  destroy  the  few  bacilli  that  are  in  the  body, 
but  we  can  readily  see  that  this  might  be  harmful  rather 
than  beneficial  when  the  number  of  bacilli  in  the  body  is 
large  enough  to  furnish  a  dangerous  amount  of  the  poison 
when  set  free.  In  this  case  the  old  adage  that  it  is  not  wise 
to  disturb  sleeping  dogs  might  be  remembered. 

The  Effect  of  the  Precipitate  Residue  on  Animals. — This 
is  the  most  interesting  of  the  split  products  of  the  tubercle 
bacillus,  and  it  deserves  much  more  study  than  we  have 
as  yet  been  able  to  give  it.  On  healthy  animals  it  has  no 
recognizable  ill  effects  either  in  single  or  repeated  doses, 
large  or  small.  We  took  6  half-grown  pigs  and  injected 
into  the  abdominal  cavity  every  third  or  fourth  day  50 
mg.  of  this  residue.  These  injections  were  begun  April  20 
and  continued  until  June  11.  During  this  time  each  animal 
received  sixteen  injections,  a  total  of  800  mg.  each.  All 
increased  normally  in  weight,  and  five  days  after  the  last 
injection  all  were  killed  and  carefully  examined  and  found 
to  be  perfectly  normal.  We  were  led  to  do  this  because  of 
the  following  experience:  Three  sets  of  pigs  were  inocu- 
lated intra-abdominally  with  the  living  a  virulent  culture. 
These  inoculations  were  made  December  14,  1905.  The 
animals  of  the  first  set  had  no  treatment,  and  when  killed 
April  11,  1906,  were  found  to  be  perfectly  normal.  Those 
of  the  second  set  had,  December  18,  50  mg.,  December  22,  60 
mg.,  and  December  26,  70  mg.  of  the  cell  residue.  All  in 
this  set  were  killed  April  11,  and  were  also  found  to  be 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS      171 

wholly  free  from  infection.  Those  of  the  third  set  had, 
December  18,  30  mg.,  December  22,  75  mg.,  and  December 
26,  100  mg.  of  the  precipitate  residue.  Half  of  this  set  died 
before  April  11  of  tuberculosis,  and  the  other  half  were 
found  to  be  tuberculous  when  killed  on  that  date.  To  us 
this  indicates  that  the  precipitate  residue  has  some  specific 
effect  upon  tuberculous  animals.  We  suspect  that  the  ill 
effect  in  these  instances  was  due  to  the  size  of  the  doses, 
because  in  reality  the  doses  of  the  precipitate  residue  were 
much  larger  than  those  of  the  cell  residue — much  larger, 
indeed,  than  the  figures  indicate,  because,  as  we  have 
stated,  the  cell  residue  is  not  freely  soluble  in  water,  while 
the  precipitate  residue  is  wholly  soluble.  In  making  up  our 
solutions  we  weighed  out  each  residue  and,  in  fact,  the 
animals  received  only  the  soluble  parts  of  the  amounts 
stated  in  the  figures.  We  can  easily  understand  how  exces- 
sive doses  given  soon  after  inoculation  with  the  avirulent 
culture  might  induce  such  a  result.  This  culture  is  aviru- 
lent because  it  makes  only  an  ineffectual  attempt  to  grow 
in  the  animal  body.  The  feeble  effort  is  resisted  and  over- 
come by  the  natural  defences  of  the  healthy  body.  Now, 
if  these  natural  defences  were  wholly  occupied  in  disposing 
of  the  material  injected,  which  should  have  been  only 
sufficient  to  awaken  these  defences,  then  the  bacilli  would 
meet  with  no  resistance  and  would  multiply. 

The  precipitate  residue  sensitizes  guinea-pigs  to  the 
tubercle  bacillus  just  as  the  cell  residue  does.  Evidently 
our  so-called  residues  are  much  alike,  and  it  is  more  than 
probable  that  they  contain  the  same  active  constituent. 
In  cultures  from  three  to  six  months  old  many  of  the  bacilli 
have  undergone  autolytic  changes  and  the  cellular  sub- 
stance has  in  part  passed  into  solution.  This  is  true  of 
both  the  poisonous  and  the  non-poisonous  groups  of  the 
protein  that  makes  up  the  cell  substance.  Of  one  thing 
we  have  satisfied  ourselves  at  least,  and  that  is  that  no 
preparation  from  the  tubercle  bacillus  should  be  used  in 
the  treatment  of  tuberculosis  until  the  poisonous  group  of 
the  tuberculous  protein  and  other  proteins  in  the  culture- 


172  PROTEIN  POISONS 

medium  be  removed.  This  is  too  powerful  a  poison  to  be 
injected  repeatedly  even  in  small  doses  into  the  animal 
body. 

One  of  us  has  for  the  past  two  years  used  solutions  of 
the  cell  residue  in  the  treatment  of  tuberculosis  in  man. 
The  most  suitable  preparation  is  a  1  per  cent,  solution 
filtered  through  porcelain.  The  cell  residue  in  weighed 
quantity  is  placed  in  a  bottle  with  the  proper  volume  of  a 
0.5  per  cent,  solution  of  carbolic  acid,  and  the  bottle  is 
carried  on  a  mechanical  shaker  for  twenty-four  hours, 
after  which  the  content  is  passed  through  a  porcelain  filter. 
Such  a  solution  will  keep  indefinitely.  We  have  used  this 
solution  sufficiently  to  justify  the  following  statements: 
(1)  It  is  of  no  value  in  advanced  cases  of  pulmonary  tuber- 
culosis. (2)  It  may  prove  harmful  even  in  initial  cases  if 
the  dose  be  too  large  or  if  small  doses  be  too  frequently 
repeated.  (3)  When  properly  used  in  initial  cases  or  in 
localized  tuberculosis,  its  action  is  apparently  prompt  and 
specific.  If  the  tubercle  bacilli  wholly  disappear  from  the 
sputum,  as  they  may,  the  injections  should  be  repeated 
at  intervals  of  from  two  to  four  weeks  for  some  months. 
We  .wish  it  clearly  understood  that  in  well-established  cases 
of  pulmonary  tuberculosis  no  benefit  from  this  treatment 
can  be  expected.  We  believe  that  in  initial  cases  this  pre- 
paration is  preferable  to  any  form  of  tuberculin. 

Toxophor  Group. — White  and  Avery1  have  reported  an 
interesting  research  on  the  split  products  of  the  cellular 
substance  of  the  tubercle  bacillus,  especially  of  the  toxo- 
phor  group.  They  used  a  strain  virulent  to  guinea-pigs. 
This  was  grown  on  glycerin  broth  cultures  for  six  weeks, 
and  the  cellular  substance  was  washed  with  alcohol  and 
ether,  ground  in  a  ball  mill  and  split  up  by  our  method. 
The  toxophor  obtained  by  them  agreed  with  that  which 
we  have  prepared.  It  is  a  yellowish-brown  powder  of 
characteristic  pungent  odor,  readily  soluble  in  alcohol.  Its 
aqueous  solutions  are  faintly  turbid,  and  give  the  biuret, 

1  Jour.  Med.  Research,  1912,  xxvi,  317. 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     173 

xanthoproteic,  Adamkiewicz,  Liebermann,  Millon,  and 
Molisch  tests;  the  last  two  faintly.  Bromine  water  pro- 
duces a  white  flocculent  precipitate,  but  no  color,  showing 
the  absence  of  tryptophan.  Injections  were  made  into  the 
right  external  jugular  vein  of  guinea-pigs  of  about  200 
grams  weight.  The  poison,  as  prepared,  killed  in  doses  of 
1  to  15,000  body  weight.  White  and  Avery  give  such  an 
excellent  statement  of  the  symptoms  and  gross  pathology 
that  we  are  induced  to  make  the  following  quotation :  "  When 
a  quantity  approaching  the  minimum  fatal  dose  is  given, 
the  first  symptoms  appear  immediately,  or,  at  most,  within 
thirty  seconds.  The  animal  becomes  restless,  scratches 
its  nose,  and  frequently  utters  a  sharp  hiccough.  The 
movements  become  incoordinate,  the  gait  is  unsteady. 
The  eyes  are  fixed,  and  stare.  Respiratory  embarrassment, 
with  diaphragmatic  spasm  sets  in  and  increases  to  a  degree 
which  causes  the  animal  to  spring  from  its  feet,  to  buck, 
and  finally  to  fall  on  its  side  with  convulsive  twitching  of 
its  legs,  intermittent,  and  both  clonic  and  tonic  in  character. 
Involuntary  micturition  and  defecation  frequently  take 
place.  The  dyspnea  becomes  more  marked,  and  then 
ensue  successive  periods  of  apnea,  lasting  as  long  as  twenty 
to  thirty  seconds.  These  are  followed  by  violent  inspira- 
tory  efforts,  during  which  the  chest  wall  becomes  fixed 
in  maximum  inspiration.  Cyanosis  is  noticeable  in  the 
lips  and  ears,  and  becomes  more  marked.  The  convulsive 
gasps  increase  in  frequency  and  decrease  in  depth,  until 
finally  only  the  lips  move,  the  feeble  and  rapid  dilatations 
of  the  alse  nasi  marking  the  onset  of  death.  This  sequence 
of  symptoms  is  accompanied  by  a  rapid  and  progressive 
fall  of  the  body  temperature.  Death  takes  place  in  from 
one  and  one-half  to  six  or  seven  minutes.  Immediate 
autopsy  reveals  first  a  cyanotic  hue  of  the  subcutaneous 
and  muscular  tissues.  The  blood  is  dark  in  color  and  does 
not  clot  readily.  Beyond  an  exaggerated  peristaltic  move- 
ment of  the  intestines,  the  abdominal  viscera  appear  to 
be  normal.  On  opening  the  chest  the  lungs  are  found  to 
be  in  a  state  of  maximum  inflation,  overlapping  the  peri- 


174  PROTEIN  POISONS 

cardium,  and  forming  a  cast  of  the  thoracic  cavity.  They 
are  pale  and  often  slightly  bluish  in  color,  and  frequently 
exhibit  punctate  hemorrhages  on  the  surface.  The  heart 
still  beats.  Not  infrequently  there  is  definite  heart  block, 
with  an  auriculoventricular  arrhythmia  of  three  to  one. 
Often  there  are  petechial  hemorrhages  in  the  epicardium, 
Greater  extravasations  are  also  seen,  and  in  two  cases 
actual  rupture  of  the  ventricle  had  apparently  taken  place. 
On  section  the  lungs  do  not  collapse,  and  on  pressure  only  a 
little  frothy  serum  exudes.  They  are  not  edematous.  They 
float  on  water.  The  excised  heart  continues  to  beat  for 
several  minutes.  The  gross  appearance  of  the  brain  is 
normal.  A  study  of  the  pathological  changes  in  the  his- 
tology of  the  lungs,  heart,  and  brain  has  been  undertaken, 
but  has  not  yet  progressed  sufficiently  to  warrant  any 
conclusions.  When  the  dose  is  larger  the  acute  symptoms 
appear  instantaneously,  and  their  sequence  is  more  rapid. 
With  a  sublethal  dose  the  onset  is  slower  and  the  manifes- 
tations are  less  violent.  The  animal  shows  evidence  of 
weakness,  drops  its  hind  legs,  and  frequently  lies  on  its 
side  in  collapse.  The  apneic  stage  is  never  reached,  its 
appearance  therefore  signifies  inevitable  death.  Recovery 
from  a  non-fatal  dose  is  comparatively  prompt  even  when 
near  the  lethal  borderline.  Recovered  animals  exhibit  no 
visible  sequela?  of  the  intoxication." 

These  investigators  have  compared  the  acute  intoxi- 
cation produced  in  animals  by  the  tuberculopoison  with 
anaphylactic  shock,  and  conclude  that  there  are  no  appre- 
ciable points  of  difference  in  the  symptomatology  and  gross 
pathology  of  the  two  conditions.  "They  would  therefore 
appear  to  be  identical."  The  tuberculopoison,  like  that 
obtained  from  other  proteins,  is  thermostabile.  It  also 
agrees  with  the  like  poison  obtained  by  the  cleavage  of 
other  proteins  in  the  following  particulars:  (1)  It  lowers 
the  temperature  when  given  in  doses  sufficient  to  produce 
recognizable  effects.  (2)  It  does  not  sensitize  animals 
to  the  unbroken  tuberculoprotein,  while  the  haptophor 
group  is  not  poisonous  and  does  sensitize  to  the  whole 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     175 

protein.  (3)  The  injection  of  non-fatal  doses  of  the  poison 
renders  animals  at  least  temporarily  refractory  to  subsequent 
injections  of  what  would  normally  be  fatal  doses.  We 
have  always  held  that  this  is  due  to  the  establishment  of  a 
tolerance.  This  is  important  and  we  will  refer  to  it  again 
when  we  discuss  the  action  of  tuberculin.  (4)  The  poison 
is  not  absorbed  in  vitro  by  brain,  lung,  or  liver  tissue.  "These 
experiments  seem  to  emphasize  the  absence  of  any  possible 
identity  of  this  protein  fragment  with  the  true  toxins. 
The  results,  however,  are  in  accord  with  the  symptoms 
produced  by  the  cell  poison.  Recovery  from  a  sub-lethal 
dose  is  rapid  and  complete,  and  this  would  imply  that  the 
contact  between  the  body  cells  and  the  poison  is  transi- 
tory, and  non-destructive.  It  appears  to  be  more  like  a 
fulminating  irritation,  and  may  result  in  an  arrest  of  func- 
tion due  to  a  disturbance  in  the  physical  equilibrium  of  the 
cells  affected."  (5)  The  serum  of  normal  guinea-pigs 
incubated  with  the  poison  does  not  materially,  at  least, 
decrease  its  action.  (6)  The  poison  does  not  induce  any 
local  reaction  when  introduced  intradermically  in  guinea- 
pigs  sensitized  to  tuberculoprotein.  Four  animals  sensitized 
nineteen  days  previously  with  the  cell  residue,  and  which 
had  been  found  to  be  sensitive  to  an  extract  emulsion  of 
tubercle  bacilli,  and  four  animals  rendered  and  proved 
sensitive  by  a  watery  extract  of  tubercle  bacilli  received 
intradermal  injections  of  the  poison,  and  under  close  obser- 
vation showed  no  reaction.  This  is  as  should  be  expected. 
The  skin  reaction,  like  other  tuberculin  reactions,  results 
from  the  cleavage  of  the  tuberculoprotein.  The  cleavage 
products  produce  no  such  reactions.  (7)  Auer  and  Lewis1 
showed  that  prophylactic  treatments  with  atropine  save 
a  large  percentage  of  animals  from  death  by  anaphylactic 
shock  on  the  reinjection  of  the  homologous  protein.  We 
claim  that  our  protein  poison  is  the  active  agent  in 
anaphylaxis.  Now,  White  and  Avery  show  that  atropine 
protects  75  per  cent,  of  guinea-pigs  from  death  after  the 

1  Amer.  Jour.  Physiology,  1910,  xxvi,  439. 


176  PROTEIN  POISONS 

administration  of  lethal  and  slightly  supralethal  doses  of 
the  poison.  (8)  Morphine  sulphate  has  been  shown  by 
White  and  Avery  to  antagonize  the  action  of  the  poison." 
Of  the  animals  tested  (19)  only  3  showed  typical  symptoms, 
and  with  two  of  these  death  was  slightly  delayed.  The 
three  autopsies  revealed  typical  inflation  of  the  lungs,  with 
epicardial  hemorrhages  in  two.  Nine  of  the  pigs  had  only 
slight  symptoms,  and  although  the  issue  was  fatal,  death 
was  delayed  from  forty-two  minutes  to  over  six  hours.  On 
section,  however,  six  of  the  animals  showed  inflated  lungs 
with  epicardial  hemorrhages.  Two  animals  recovered.  It 
will  be  noted  that  in  five  cases  a  dose  of  1  to  12,000  failed 
to  produce  typical  immediate  symptoms.  Further  investi- 
gations of  the  effects  of  morphine  might  lead  to  a  better 
knowledge  of  the  factors  concerned  in  the  sequelae  of 
parenteral  administration.  (9)  Banzhaf  and  Steinhardt1 
studied  the  effects  of  chloral  hydrate  upon  the  action  of 
our  poison  prepared  from  egg-white,  and  came  to  the 
following  conclusions:  "Normal  guinea-pigs  under  the 
influence  of  chloral  (by  intracardiac  and  intramuscular 
injections)  were  completely  protected  against  one  and 
one-fourth  fatal  doses  of  the  poison  (given  intracardiacly). 
If  two  or  more  fatal  doses  were  given  death  resulted.  Chloral 
mixed  with  the  poison  and  then  given  caused  irregular  results 
which  were  interpreted  as  meaning  that  there  is  no  chemical 
union  of  the  chloral  and  poison  in  vitro.  We  assume  that 
the  chloral  protected  by  union  with  certain  vital  cells." 
White  and  Avery  used  a  2.5  per  cent,  solution  of  chloral 
in  normal  salt.  The  injections  were  made  intravenously. 
"With  the  exception  of  2  animals  displaying  typical  symp- 
toms, both  of  which  received  amounts  of  the  poison  con- 
siderably in  excess  of  that  required  to  kill,  6  of  the  13 
survived  the  injection  with  slight  or  no  symptoms,  while 
5  succumbed  in  from  two  to  thirteen  hours  without  exhibiting 
the  classic  respiratory  spasms.  Autopsy  showed  the  typical 
findings  in  2,  while  in  2  others  there  was  a  partial  inflation 

1  Jour.  Med.  Research,  1910,  xxiii,  1. 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS      177 

of  the  lungs  with  punctate  hemorrhages  beneath  the  peri- 
cardium. The  results,  although  not  so  strikingly  positive 
as  those  of  Banzhaf  and  Steinhardt,  at  least  tend  to  confirm 
their  conclusion."  (10)  Banzhaf  and  Steinhardt  found 
that  lecithin  given  intraperitoneally  in  doses  of  from  250 
to  500  mg.  or  more  to  serum-sensitized  guinea-pigs  pro- 
tected them  from  a  second  injection  of  5  c.c.  of  horse  serum 
given  twenty-four  hours  later.  When  lecithin  was  emul- 
sified with  the  Vaughan  poison  or  given  twenty-four  hours 
before  the  poison  was  injected,  no  protection  was  afforded. 
From  this  Banzhaf  and  Steinhardt  conclude  that  lecithin 
prevents  the  cleavage  of  the  protein  in  a  sensitized  animal 
on  reinjection  and  that  it  does  not  neutralize  or  modify 
the  action  of  the  preformed  poison.  White  and  Avery, 
from  their  experiments,  come  to  the  following  conclusion: 
"Lecithin  emulsion  injected  simultaneously  with  the 
poison  seems  to  possess  a  slight  and  irregular  prophylactic 
action.  Incubation  of  the  poison  with  lecithin  emulsion 
for  an  hour  at  37.5°  increases  this  neutralizing  property. 
A  dose  of  1  to  12,000  was  not  affected.  The  preliminary 
administration  of  lecithin  protected  some  of  the  animals, 
delayed  death  in  others,  and  was  without  effect  in  the 
remainder.  The  results  were  too  inconstant  to  warrant 
definite  conclusions." 

White  and  Avery  are  inclined  to  the  opinion  that  our 
crude  protein  poison  contains  a  plurality  of  active  substances, 
and  in  this  they  are  probably  right.  They  say:  "The 
effects  provoked  by  the  parenteral  administration  of  the 
artificially  obtained  poisonous  substance  in  non-fatal  doses, 
and  as  modified  by  atropine,  morphine,  chloral,  and  other 
drugs,  seem  to  suggest  the  plurality  of  its  action.  It  is 
conceivable  that  the  poisonous  fraction  obtained  by 
Vaughan's  method  contains  either  an  essential  component 
which  is  several  in  its  physiological  action,  and  which  in 
sufficient  doses  exerts  its  primary  and  dominating  effect  on 
the  respiratory  mechanism,  or  that  it  contains  groups  or 
individual  constituents  of  different  selective  vital  affinities, 
the  most  eminent  of  which  is  for  the  peripheral  or  central 
12 


178  PROTEIN  POISONS 

cells  functioning  in  respiration.  It  is  not  unreasonable  to 
hope  that  a  further  separation  of  the  poisonous  fraction 
into  its  components  and  a  more  intimate  study  of  their 
various  actions  on  the  animal  economy  may  furnish  valuable 
clues  not  only  to  the  relation  of  these  chemical  substances 
to  true  anaphylactic  processes,  but  also  to  the  physiological 
nature  of  the  varied  phenomena  of  hypersensitiveness." 
It  seems  to  us,  theoretically,  that  there  must  be  a  whole 
spectrum  of  poisons  in  the  protein  molecule.  We  have 
shown  that  at  least  one  group  in  this  molecule  is  poisonous. 
The  poisonous  action  of  the  protein  molecule  becomes 
more  marked  as  we  proceed  in  stripping  off  certain  side 
chains.  Peptone  is  more  poisonous  than  the  native  protein 
from  which  it  is  obtained.  Our  product  is'  more  active 
than  peptone.  Between  the  two  there  must  be  a  group  of 
bodies,  each  of  which  is  more  active  than  the  peptone  and 
less  active  than  our  split  product.  Indeed,  we  are  con- 
fident that  we  have  discovered  some  of  these  intermediate 
bodies.  As  has  been  stated,  when  the  alcohol  employed 
in  the  cleavage  of  the  protein  molecule  is  not  absolute  we 
obtain  products  that  are  quite  unlike  our  poison  in  physical, 
chemical,  and  physiological  properties.  They  are  sticky 
and  gummy.  They  contain  some  carbohydrate,  responding 
to  the  Molisch  test,  and  yielding  a  reducing  substance 
after  prolonged  boiling  with  dilute  mineral  acid;  while  our 
final  product  is  not  gummy  and  fails  to  show  any  evidence 
of  carbohydrate  content,  except  in  that  from  tubercle 
bacilli.  These  other  bodies  kill  much  less  promptly.  The 
paralytic  symptoms  are  more  marked,  and  the  convulsive 
stage  is  either  only  slightly  in  evidence  or  wholly  wanting. 

Interesting  experiments  on  sensitization  to  tuberculo- 
protein  have  been  made  by  Baldwin1  and  Krause.2  We 
make  the  following  extracts  from  this  work:  Animals  may 
be  sensitized  by  any  of  the  ordinary  products  of  the  tubercle 
bacillus.  Sensitization  may  be  secured  by  introducing 
the  protein  by  any  parenteral  route,  by  the  peritoneal 

1  Jour.  Med.  Research,  1910,  xxii,  189.         2  Ibid.,  xxii,  275;  xxiv,  361. 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     179 

cavity,  subcutaneously,  subdurally,  intracerebrally,  by  post- 
orbital  injection,  and  probably  by  intravenous  injection, 
though  the  last-mentioned  method  was  not  tried.  Sensi- 
tization  may  be  obtained  by  the  injection  of  only  0.05  mg. 
of  the  protein.  The  best  preparations  for  sensitization 
are  those  in  which  the  protein  is  in  solution.  The  shortest 
period  of  incubation  found  was  six  days.  This  was  when 
the  sensitizing  dose  was  given  postorbitally.  Before  the 
twenty-first  day  sensitization  is  uneven  and  inconstant. 
After  this  period  it  proceeds  with  great  regularity,  and  the 
longest  duration  noted  was  two  hundred  and  eighty-six 
days.  It  is  likely  that  it  continues  in  the  guinea-pig  through- 
out life.  The  size  of  the  sensitizing  dose  bears  no  relation 
to  the  period  of  incubation.  Acute  anaphylactic  shock 
follows  when  the  reinjection  is  given  intravenously  or  post- 
orbitally. The  minimum  toxic  dose  on  reinjection  was  found 
to  be  0.99  mg.  of  the  dry  protein,  and  the  minimum  fatal 
dose  on  reinjection  1.6  mg.  Attempts  to  establish  passive 
anaphylaxis  have  been  uniformly  unsuccessful.  Infected 
animals  become  autosensitized  and  are  killed  by  injections 
of  large  amounts  of  the  tuberculoprotein.  This  protein  does 
not  act  like  a  toxin,  and  when  injected  into  animals  does 
not  lead  to  the  elaboration  of  an  antitoxin.  "If  an  animal 
be  infected  experimentally  it  begins  to  react  to  tuberculin 
about  the  fifteenth  day;  in  like  manner,  the  non-tuber- 
culous but  protein-treated  animal  will  react  to  a  second 
injection  about  two  weeks  after  the  first.  Again,  both 
the  tuberculous  and  the  sensitized  non-tuberculous  animals 
react  to  exceedingly  small  doses  of  the  protein;  indeed,  a 
certain  proportion  of  the  tuberculous  will  undergo  an 
intoxication  that  is  identical  with  acute  anaphylaxis, 
provided  the  toxic  dose  is  applied  postorbitally,  while  if 
the  sensitized  animal  receives  its  toxic  injection  by  a  route 
that  renders  absorption  less  rapid — e.  g.,  an  intraperitoneal 
injection— the  resulting  intoxication  will  tend  to  approxi- 
mate what  is  generally  observed  as  the  tuberculin  reaction 
in  the  infected  guinea-pig  (without,  of  course,  any  focal 
reaction).  Therefore,  while  the  facts  will  not  at  present 


180  PROTEIN  POISONS 

warrant  the  flat  declaration  that  the  two  phenomena  result 
from  the  same  fundamental  causes,  there  are  enough  data 
at  hand  to  justify  the  elaboration  of  a  working  hypothesis 
that  such  is  the  case."  The  important  question  of  the 
relation  between  sensitization  and  immunity  to  infection 
has  been  tested  by  Baldwin  and  Krause.  Series  of  guinea- 
pigs  were  sensitized  to  tuberculoprotein.  The  fact  that 
they  were  in  full  sensitization  was  demonstrated  by  testing 
some  of  each  set.  Those  that  recovered  from  anaphylactic 
shock  and  known  as  "refractory"  were  inoculated  an  hour 
after  the  reinjection.  Lot  A  had  received  a  total  of  25  c.c. 
of  the  watery  extract  in  seventeen  doses  over  a  period  of 
thirty-nine  days.  Lot  B  had  received  a  total  of  13  c.c. 
of  the  watery  extract  in  ten  doses  over  a  period  of  thirty- 
nine  days.  Lot  C  had  received  a  total  of  8  c.c.  of  the  watery 
extract  in  six  doses  over  a  period  of  thirty-nine  days.  Lot 
D  had  not  been  sensitized.  The  last  sensitizing  doses  were 
given  June  14,  1910.  All  of  these  animals  were  inoculated 
with  the  same  amount  of  a  virulent  culture  of  the  tubercle 
bacillus  July  1,  1910.  Sixty-two  days  after  inoculation 
all  the  animals  were  killed  and  examined.  A  summary  of 
the  findings  is  stated  as  follows:  "The  refractory  animals 
suffered  most.  The  disease  was  pretty  well  disseminated 
in  all  of  them,  and  they  exhibited  far  more  tuberculosis 
than  any  of  the  animals  that  had  not  been  intoxicated, 
and  than  any  of  the  controls.  .  .  .  The  animals  that 
were  sensitized  in  various  ways  all  became  diseased.  As 
a  general  thing,  we  may  say  that  the  more  protein  the 
animal  received  during  preliminary  treatment,  the  less  was 
the  resultant  infection.  So  far  as  one  could  tell  from  the 
toxic  symptoms  of  the  test  animals  there  was  very  little 
difference  in  the  average  degree  of  sensitization  in  the 
several  sets  of  guinea-pigs.  The  results  of  inoculation 
were,  however,  different.  It  is  most  likely  that  the  differ- 
ences were  altogether  independent  of  any  degree  of  raised 
or  lowered  resistance  conferred  by  the  sensitive  state,  but 
that  they  were  due  to  the  heightened  immunity  that  followed 
the  protein  injections." 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     181 

Krause  concludes  the  paper  from  which  the  above  was 
taken  as  follows: 

"1.  Sensitization  of  non-tuberculous  guinea-pigs  with 
tuberculoprotein  does  not  alter  their  resistance  to  experi- 
mental tuberculous  infection. 

"2.  Sensitization  to  tuberculoprotein  and  relative 
immunity  (increased  resistance)  to  infection  can  occur 
coincidently  in  the  same  animals. 

"3.  Resistance  to  infection  is  markedly  lowered  during 
the  period  that  a  sensitized  animal  is  suffering  from  symp- 
toms of  anaphy lactic  shock." 

The  third  conclusion  is  certainly  justified  from  the  results 
of  the  research,  and  is  what  might  have  been  predicted 
at  the  start.  Whether  the  first  conclusion  is  in  any  way 
contradictory  to  the  previous  statement  "that  the  more 
protein  the  animal  received  during  preliminary  treatment, 
the  less  was  the  resultant  infection,"  we  leave  the  reader 
to  determine  for  himself.  This  line  of  experimentation 
should  be  continued  with  all  the  tuberculoprotein  prepara- 
tions and  with  variations  in  size  and  frequency  of  dosage. 

Thiele  and  Embleton1  have  reviewed  the  literature  of 
Sensitization  in  tuberculosis,  and  have  experimented  with 
reference  to  both  active  and  passive  hypersensitiveness  to 
tubercle  bacilli,  and  the  relation  to  the  tuberculin  reaction 
in  man.  The  conclusions  reached  are  stated  as  follows: 
(1)  Guinea-pigs  may  be  typically  sensitized  with  pow- 
dered tubercle  bacilli.  (2)  Guinea-pigs  may  be  passively 
sensitized  with  the  blood  or  tissues  of  animals  actively 
sensitized.  (3)  Guinea-pigs  may  be  sensitized  to  tuberculin 
with  the  blood  of  tuberculous  patients  who  are  highly 
sensitive  to  tuberculin.  (4)  Likewise,  guinea-pigs  may 
be  sensitized  with  tuberculous  tissue  from  man,  or  with 
that  from  tuberculous  guinea-pigs.  (5)  By  regulating  the 
dose  one  can  induce  fever  or  cause  the  temperature  to 
fall  below  the  normal  in  actively  sensitized  guinea-pigs 
with  tuberculin.  (6)  The  same  results  can  be  obtained  in* 

1  Zeitsch.  f.  Immunitatsforschung,  1913,  xvi,  411. 


182  PROTEIN  POISONS 

animals  sensitized  with  heterologous  or  homologous  tissue. 
(7)  A  cutaneous  reaction  has  not  been  obtained. 

This  work  is  confirmed  and  supplemented  by  that  of 
Sata,1  who  sensitizes  guinea-pigs  with  a  single  injection  of 
tuberculous  serum,  in  doses  of  0.1,  0.5,  or  1.0  c.c.  subcu- 
taneously,  intraperitoneally,  or  intravenously,  and  uses  a 
reinjection  of  old-tuberculin  intravenously.  When  the 
dose  of  the  reinjection  is  as  much  as  0.5  c.c.  acute  ana- 
phy lactic  death  results.  With  smaller  doses  there  is 
elevation  of  temperature. 

Many  investigators  have  failed  to  sensitize  animals 
with  tuberculin,  while  most  have  succeeded  with  dead 
bacilli  and  with  aqueous  extracts.  This  is  not  surprising; 
indeed  it  is  what  should  have  been  expected.  Tuberculin 
consists  of  digested,  denatured  proteins  of  relatively  simple 
composition.  It  is  well  known  that  peptones  and  poly- 
peptids  do  not  sensitize.  The  protein  poison  when  detached 
from  other  groups  in  the  protein  molecule  sensitizes  neither 
to  itself,  nor  to  the  unbroken  protein.  The  fact  that  tuber- 
culin does  not  sensitize  or  does  so  imperfectly  raises  a 
serious  question  as  to  its  employment  as  a  therapeutic 
agent.  It  is  undoubtedly  an  excellent  diagnostic  agent 
because  its  relatively  simple  structure  may  favor  its  prompt 
cleavage  when  injected  into  an  animal  already  sensitized 
by  the  disease.  But  if  it  is  not  a  sensitizer  its  therapeutic 
good  effect,  if  it  has  any  such  effect,  must  be  confined  to 
the  possible  establishment  of  a  tolerance  to  the  tuberculo- 
poison.  Sensitization  to  tuberculoprotein  can  be  induced 
by  bacillary  emulsions,  with  watery  extracts,  and  with 
the  non-poisonous  residue.  If  the  sensitization  secured  by 
the  last-mentioned  agent  is  as  good  as  that  produced 
by  the  others,  it  has  the  advantage  of  not  containing  any 
poison.  On  the  other  hand,  if  the  therapeutic  effect  desired 
consists  in  the  development  of  a  tolerance  to  the  poison, 
tuberculin  must  be  preferred  unless  wre  should  use  the  more 
•completely  isolated  poison. 

>  Zeitsch.  f.  Immunitatsforschung,  1913,  xvii,  62. 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS     183 

There  are  those  who,  while  admitting  that  animals  can 
be  sensitized  to  tuberculoprotein,  hold  that  the  tuberculin 
reaction  is  not  an  anaphylactic  one.  We  think  that  it  is, 
and  that  the  fact  that  tuberculin  does  not  sensitize  or  does 
so  imperfectly  does  not  contradict  this.  It  is  probable  that 
when  tuberculin  does  sensitize  at  all  it  is  due  to  the  fact 
that  it  contains  traces  of  but  little  altered  or  unaltered 
tuberculoprotein.  The  tuberculin  reaction  should  be 
regarded  as  a  phenomenon  resulting  from  a  reinjection. 
The  animal  is  already  sensitized  by  the  disease. 

Koch  in  his  early  work  pointed  out  two  facts,  which  in 
a  way  seemed  to  be  contradictory,  but  which  have  been 
found  to  be  true.  First,  he  showed  that  a  tuberculous 
animal  behaves  toward  a  second  infection  differently  from 
a  normal  animal,  the  former  resisting  the  second  infection 
by  forming  an  inflammatory  area  about  the  point  of  the 
second  inoculation,  this  leading  to  necrosis,  and  recovery 
without  extension  of  the  infection.  Second,  Koch  stated 
that  the  injection  of  dead  tubercle  bacilli  into  tuberculous 
guinea-pigs  killed  them  within  from  six  to  forty-eight 
hours,  while  like  injections  into  normal  guinea-pigs  had 
no  such  effect.  These  apparently  contradictory  statements 
have  not  only  been  confirmed,  but  have  been  found  not  in 
any  way  in  conflict.  The  studies  of  Romer,  Hamburger, 
and  others  have  shown  that  the  following  conditions  must 
prevail  in  order  to  fully  demonstrate  the  first  statement  of 
Koch:  (a)  The  first  injection  must  be  a  weak  one,  per- 
mitting the  disease  to  run  a  chronic  course.  (6)  The  time 
interval  between  the  first  and  second  inoculations  must 
be  relatively  long,  the  resistance  to  the  second  infection 
increasing  with  time,  (c)  The  dose  of  the  second  injection 
must  not  exceed  a  certain  limit.  What  happens  to  the 
bacilli  of  the  second  inoculation?  Why  should  these  fail 
to  develop,  while  those  of  the  first  inoculation  continue  to 
grow?  Kraus  and  Hofer1  have  found  that  tubercle  bacilli 
injected  into  the  peritoneum  of  a  tuberculous  animal  are 

»  Deutsch.  med.  Woch.,  1912. 


184  PROTEIN  POISONS 

destroyed  by  lysis  within  an  hour.  There  is  some  lytic 
destruction  of  tubercle  bacilli  in  the  peritoneum  of  a 
healthy  guinea-pig,  but  this  does  not  compare  in  rapidity 
and  completeness  with  that  occurring  in  the  tuberculous 
animal.  But  why  is  the  action  of  this  lytic  agent  manifested 
so  effectively  against  the  bacilli  of  the  second  inoculation 
while  those  of  the  first  apparently  proceed  in  uninterrupted 
growth?  One  of  Hamburger's  experiments1  may  throw 
some  light  on  this  question.  He  made  his  reinoculation 
subcutaneously  on  each  side.  On  the  left  he  injected  a 
small  dose,  on  the  right  a  large  one.  On  the  left  there  was 
no  development;  on  the  right  a  tuberculous  nodule  developed. 
We  infer  from  this  and  similar  observations  made  by  others 
that  the  lytic  agent  which  destroys  the  tubercle  bacillus 
and  which  is  produced  in  larger  amount  in  tuberculous 
than  in  normal  animals,  because  the  cells  of  the  former  have 
been  sensitized,  is  stored  in  the  cells  as  a  zymogen,  and 
is  activated  only  when  tuberculoprotein  is  brought  into 
contact  with  the  cell,  and  possibly  is  active  only,  or  is 
most  active,  in  statu  nascendi.  The  ferment  is  capable  of 
destroying  only  a  given  amount  of  bacilli  or  is  wholly 
inactive  in  the  presence  of  a  great  excess  of  substrate, 
or  its  action  is  soon  interrupted  by  the  accumulation  of 
fermentative  products.  However,  it  is  quite  certain  that 
all  the  bacilli  of  the  second  inoculation  are  not  always 
killed  because  it  may  happen  according  to  observations 
of  Hamburger  that  some  months  after  the  second  inoc- 
ulation, during  which  time  there  may  have  been  no 
evidence  of  infection,  tubercular  processes  appear  and 
develop  like  a  primary  infection.  Whatever  the  true 
explanation,  it  is  a  fact  that  the  tuberculous  animal  is  more 
resistant  to  additional  infection  than  the  normal  animal  is 
to  primary  infection.  This  led  Lowenstein2  to  say:  "Only 
the  tuberculous  organism  is  tuberculosis-immune."  Ham- 


1  Beitrage  z.  klin.  d.  Tuberculose,  xii. 

2  Handbuch  d.  path.   Mikroorganismen,  Kolle  u.  Wassermann,  zweite 
Auflage. 


185 

burger  and  Toyosuku1  infected  guinea-pigs  subcutaneously, 
and  after  the  disease  had  become  chronic,  they  submitted 
these  animals  along  with  normal  ones  to  a  dust  rich  in 
tubercle  bacilli.  The  normal  animals  developed  pulmonary 
tuberculosis,  while  the  tuberculous  ones  failed  to  do  so. 
Romer2  developed  chronic  subcutaneous  tuberculosis,  and 
then  inoculated  intracutaneously  and  intravenously,  and 
in  this  way  demonstrated  the  immunity  of  the  tissues  of 
the  tuberculous  animal  to  infection  with  tuberculosis.  The 
submental  and  cervical  glands  of  normal  guinea-pigs  become 
tuberculous  on  feeding  with  as  small  an  amount  as  0.1 
mg.  of  living  bacilli,  but  these  glands  are  not  affected  when 
tuberculous  guinea-pigs  are  fed  with  living  bacilli.  Many 
other  investigators  have  experimented  along  the  same 
line,  with  like  results.  That  the  unaffected  tissues  and 
organs  of  tuberculous  men  are  largely  immune  to  infection 
with  the  tubercle  bacillus  is  a  matter  of  every-day  observa- 
tion. In  pulmonary  tuberculosis  the  sputum  laden  with 
bacilli  passes  through  the  upper  air  passages  without,  as 
a  rule,  infecting  them.  Besides,  there  are  cases  of  healed 
tuberculosis  with  virulent  tubercle  bacilli  in  their  expec- 
toration. There  are  tuberculosis  carriers  just  as  there  are 
typhoid  carriers. 

Koch  tried  various  methods  in  his  attempts  to  immunize 
animals  to  tuberculosis.  Early  in  his  investigations  he 
tried  feeding  animals  with  both  living  and  dead  cultures. 
For  two  months  he  fed  rats  exclusively  on  the  bodies  of 
animals  dead  from  tuberculosis.  From  time  to  time  a 
rat  was  killed,  and  most  of  them  were  found  normal.  In 
a  few,  small  nodules  were  detected  in  the  lungs.  But  these 
animals  after  feeding  for  weeks  upon  tubercular  tissue, 
developed  tuberculosis  promptly  when  inoculated  intra- 
peritoneally.  Later,  Koch  made  the  following  statement: 
"All  attempts  to  cause  absorption  of  living  or  dead  bacilli, 
by  administration  subcutaneously,  intraperitoneally,  or 
intravenously,  have  failed  me  and  also  other  investiga- 
tors. When  dead  bacilli  are  injected  subcutaneously  they 

1  Beitrage  z.  klin.  d.  Tuberkulose,  xviii  2  Ibid.,  xiii. 


186  PROTEIN  POISONS 

uniformly  cause  suppuration,  and  they  can  be  easily  stained 
and  detected  in  large  numbers  in  the  abscesses  thus  formed 
after  months.  When  injected  into  the  peritoneal  cavity 
they  are  better  absorbed,  and  I  have  obtained  some  immunity 
in  this  way,  but  they  generally  cause  local  inflammations, 
which  lead  to  adhesions  with  stenosis  and  occlusion  of  the 
intestine,  so  that  a  large  percentage  of  the  animals  is  lost. 
When  injected  intravenously  into  rabbits,  dead  bacilli 
cause  tubercular  nodules,  similar  to  those  observed  after 
infection,  in  the  lungs,  and  in  these  nodules  the  unaltered 
bacilli  can  be  found  after  a  long  time.  By  this  method 
absorption  does  not  proceed  in  the  desired  way.  Having 
been  convinced  that  the  unaltered  bacilli  could  not  be 
used,  I  attempted  to  render  them  absorbable  through  the 
action  of  chemical  agents  on  them.  The  only  method  of 
this  kind  which  I  have  found  effective  consists  in  boiling 
the  bacilli  with  dilute  mineral  acid  or  with  'strong  alkali. 
In  this  way  tubercle  bacilli  may  be  so  changed  that  they  are 
absorbed  in  toto,  and  in  large  amount,  though  slowly,  when 
administered  subcutaneously.  But  marked  immunity  has 
not  been  reached  in  this  way,  and  it  seems  that  these  chemi- 
cal agents  cause  such  thorough  alteration  in  the  bacillary 
substance  that  its  immunizing  property  is  destroyed." 

This  conclusion  reached  by  Koch  has  been  justified  by 
all  subsequent  investigators.  Levy1  has  tried  to  prepare  a 
vaccine  by  such  chemically  indifferent  substances  as  glycerin, 
25  per  cent,  solution  of  milk  sugar,  and  10  to  25  per  cent, 
solutions  of  urea.  The  object  in  these  experiments  was  to 
kill  the  bacilli  by  the  withdrawal  of  water  and  without 
changing  their  immunizing  properties.  Levy  stated  that 
these  vaccines  contain  no  living  bacilli,  and  with  them  he 
apparently  increased  the  resistance  of  guinea-pigs  to  infec- 
tion with  tubercle  bacilli,  but  Romer  doubts  the  complete 
killing  of  the  bacilli  by  these  agents.  Heating  the  bacilli 
to  70°  or  80°  has  failed  to  furnish  an  effective  vaccine. 
Lowenstein2  tried  to  prepare  a  vaccine  by  exposing  tubercle 

1  Med.  Klinik,  1905,  1906;  Centralbl.  f.  Bak.,  xlii,  xlvi,  and  xlvii. 

2  Zeitsch.  f.  Tuberkulose,  1905,  vii. 


THE  SPLIT  PRODUCTS  OF  TUBERCLE  BACILLUS      187 

cultures  to  daylight  for  a  year,  but  this  failed.  The  same 
investigator  tried  formalin,  with  a  like  result.  Bartel1 
made  a  brie  of  tubercle  bacilli  and  lymph  glands,  and 
Schroder  tried  a  like  experiment  with  spleen  pulp,  but 
neither  of  these  preparations  proved  effective.  Calmette 
and  his  students  have  tried  preparations  obtained  by  the 
action  of  iodine  and  its  salts  upon  tubercle  bacilli,  but 
without  success.  Similar  preparations  with  chlorine  have 
been  tried  by  Mossu  and  Goupil.2  Noguchi3  stated:  "The 
inoculation  of  guinea-pigs  with  tubercle  bacilli,  which  have 
been  killed  by  soaps  (sodium  oleate),  develops  in  these 
animals  a  complete  or  partial  resistance  to  subsequent 
inoculation  with  a  virulent  culture  of  the  same  strains 
of  bacilli.  In  short,  a  condition  of  immunity  to  tuberculosis 
can  be  induced  in  guinea-pigs  by  the  injection  of  an  emulsion 
of  tubercle  bacilli  in  oleic  soaps."  Zeuner,1  following  the 
work  of  Noguchi,  has  used  an  extract  of  tubercle  bacilli  in 
solution  of  sodium  oleate  in  the  treatment  of  tuberculosis. 
Broil5  found  that  guinea-pigs  treated  with  this  preparation 
survived  only  a  few  weeks.  Marxner6  tried  it  on  goats,  and 
found  no  evidence  of  infection  in  two  cases  on  postmortem, 
but  Lowenstein7  states  that  this  is  explained  by  the  fact 
that  the  animals  were  sectioned  too  soon  after  inoculation, 
and  adds  that  he  sectioned  goats  three  years  after  this 
treatment,  followed  by  inoculation,  and  found  tubercular 
cavities,  the  size  of  a  man's  head,  in  the  lungs.  Deycke 
and  Much8  found  that,  "One  part  of  tubercle  bacilli  is 
dissolved  in  two  parts  of  a  25  per  cent,  solution  of  neurin 
when  kept  at  52°  for  twenty-four  hours.  This  forms  a 
perfectly  clear  syrup  which  becomes  cloudy  on  cooling. 
We  attribute  this  phenomenon  to  the  presence  in  the  bacilli 
of  fatty  bodies  with  high  melting-points."  Later  it  was 

1  Wien.  klin.  Wochenschrift,   1905. 

2  Compt.  Rend,  de  1'Acad.,  1907. 

3  Centralbl.  f.  Bak.,  1909,  lii,  85. 

4  Zeitschf.  f.  Tuberkulose,  xv. 

6  Berlin  tierarztliche  Wochenschft.,  No.  47. 

i    •  Zeitsch.  f.  Immunitatsforschung,  x,  xi,  xii.  7  Loc.  cit. 

8  Beitrage  z.  klinik  d.  Tuberkulose,  xv,  Heft  2. 


188  PROTEIN  POISONS 

found  that  solutions  of  cholin  have  a  similar,  but  less 
marked,  solvent  action  on  tubercle  bacilli;  also  that  these 
solutions  injected  into  men  are  without  harmful  effect. 
Attempts  to  immunize  animals  with  these  solutions  have 
been  made,  without  success.  Aronson1  has  extracted  fat 
from  tubercle  bacilli  with  trichlorethylen,  and  has  attempted 
to  immunize  with  the  residue,  but  has  not  reported  any 
success.  Calmette  and  Guerin2  have  tested  the  protective 
action  of  tubercle  bacilli  grown  on  media  containing  bile 
acids.  They  stated  in  1910  that  by  the  tenth  generation, 
these  cultures  of  bovine  tubercle  bacilli  become  so  attenu- 
ated that  they  can  be  used  as  a  vaccine.  Since  these 
investigators  have  made  no  later  report  it  is  fair  to  assume 
that  their  expectations  have  not  been  realized.  Attempts 
have  been  made  to  immunize  animals  to  tuberculosis  with 
the  virulent  bacillus  by  beginning  with  so  small  an  amount 
as  one  bacillus  (Webb  and  Williams)  and  increasing  the 
dose;  with  cultures  attenuated  in  varying  degrees;  with 
living  and  dead  cultures  of  various  varieties  of  the  tubercle 
bacillus,  such  as  human,  bovine,  avian,  chicken,  from  cold- 
blooded animals,  etc.;  with  strains  of  other  acid-fast  bacilli, 
as  those  of  timothy,  butter,  manure,  etc.;  with  normal 
and  specific  sera;  but  up  to  the  present  time  no  satisfactory 
results  have  been  obtained.  This  is  an  interesting  subject, 
and  we  would  like  to. go  into  some  detail,  but  it  lies  outside 
the  scope  of  this  book. 

Lowenstein3  has  shown  that  the  tubercle  bacillus  will 
grow,  though  not  abundantly,  on  a  medium  of  the  following 
simple  composition: 

Ammonium  phosphate 6  parts 

Glycerin 40  parts 

Distilled  water 1000  parts 

Although  the  growth  is  slow  in  developing  and  sparse, 
it  elaborates  an  active  tuberculin.  This  is  additional 
evidence  that  growth  of  bacteria  consists  essentially  of 
synthetical  processes. 

1  Berl.  klin.  Woch.,  1910   No.  35.  *  Comp.  rend,  de  1'Acad.,  cli,  1. 

»  Centralbl.  f.  Bak.,  1913,  Ixviii,  591. 


CHAPTER  IX 
THE  ANTHRAX  PROTEIN1 

Literature. — Since  anthrax  is  the  most  typically  infec- 
tious of  all  diseases,  and  since  so  many  theories  have  been 
evolved  concerning  it,  we  may  be  pardoned  for  briefly 
reviewing  the  literature.  As  early  as  1805  Kausch2  wrote 
a  monograph  on  this  disease  in  which  he  held  that  it  is 
due  to  paralysis  of  the  nerves  of  respiration;  but  he  offered 
no  explanation  of  the  paralysis.  Delafond3  held  that 
anthrax  has  its  origin  in  the  influence  of  the  chemical 
composition  of  the  soil  on  the  food,  thus  inducing  patho- 
logical changes  from  malnutrition.  The  contagious  nature 
of  the  disease  was  clearly  established  in  1845  by  Gerlach.4 
This  was  confirmed  by  the  studies  of  Heuzinger,5  and  was 
endorsed  by  Virchow  in  1855,  since  which  time  it  has  never 
been  questioned.  However,  as  early  as  1849  the  bacilli 
had  been  seen  by  Pollender.6  Pollender  did  not  publish  his 
observations  until  1855,  but  he  states  that  they  were  made 
in  the  fall  of  1849.  First,  he  examined  the  blood  of  five 
cows  dead  from  anthrax,  and  compared  this  with  material 
taken  from  the  spleen  of  a  healthy  animal.  The  examina- 
tions were  not  made  until  from  eighteen  to  twenty-four 
hours  after  death,  and  he  states  that  the  blood  was  stinking, 
thus  indicating  that  it  had  become  contaminated  with 
putrefactive  organisms,  but  the  description  which  he  gives 

1  The  first  part  of  this  chapter  is  abstracted  from  a  paper  by  J.  Walter 
Vaughan,  published  in  the  Trans.  Assoc.  Amer.  Phys.,  1902,  xvii,  313. 

2  Ueber  der  Milzbrand  des  Rindviehes. 

3  Traite  sur  la  Maladie  du  Sang  des  Betes  a  laine,  1843. 

4  Magazin  f.  Thierheilkunde. 

6  Die  Milzbrandkrankheiten  der  Thieren  und  der  Menschen. 
6  Vierteljahresschrift  f.  gerichtliche  Medicin,  1855,  viii,  103. 


190  PROTEIN  POISONS 

shows  that  he  actually  saw  anthrax  bacilli.  He  used  a 
crude  compound  microscope  made  by  Plossl,  and  he  gave 
his  attention  to  the  blood  corpuscles,  chyle  globules,  and 
the  bacilli.  His  description  of  the  microorganisms  may  be 
condensed  as  follows:  The  third  and  most  interesting 
microscopic  bodies  seen  in  anthrax  blood  are  innumerable 
masses  of  rod-like,  solid,  opaque  bodies,  the  length  of  which 
varies  from  -j^-g-  to  2~ro  °f  a  nne>  and  the  breadth  averages 
ToVo  °f  a  nne-  They  resemble  the  "vibrio  bacillus"  or 
"vibrio  ambiguosus."  They  are  non-motile  and  neither 
water  nor  dilute  acids,  nor  strong  alkalies  have  any  effect 
upon  them,  and  for  this  reason  he  concluded  that  they  must 
be  regarded  as  vegetable  organisms.  He  questioned  whether 
they  existed  in  the  blood  of  the  living  animal  or  resulted  from 
putrefaction,  but  was  inclined  to  believe  the  former,  and 
thought  they  might  represent  the  infecting  organism,  or  at 
least  the  bearer  of  the  infection.  It  will  be  seen  that  Pol- 
lender  presented  no  positive  proof  that  these  rod-like  bodies 
had  any  causal  relation  to  the  disease.  In  1856  Brauell1 
inoculated  sheep,  horses,  and  dogs  with  blood  taken  from 
animals  sick  with  anthrax,  and  in  this  way  demonstrated 
that  the  disease  could  be  transmitted  to  sheep  and 
horses,  but  not  to  dogs.  He  found  sheep  highly  sus- 
ceptible, horses  less  so,  and  dogs  quite  immune.  He  also 
demonstrated  the  presence  of  the  bacilli  in  the  blood  of 
sick  animals  before  death.  It  is  interesting  to  note  that 
he  fell  into  an  error  concerning  the  motility  of  the  bacilli. 
He  states  that  when  seen  in  fresh  blood  they  are  non-motile, 
but  later  they  become  highly  motile.  This  was,  of  course, 
due  to  contamination.  It  should  be  noted  that  Brauell 
also  made  examinations  of  the  blood  of  various  domestic 
animals  suffering  from  other  diseases,  and  demonstrated 
the  absence  of  the  bacillus  in  these.  In  1863  Davaine2 
published  three  valuable  papers  on  anthrax.  In  the  first 
he  states  that  in  1850  Rayer  inoculated  sheep  with  the 
blood  of  others  dead  from  anthrax,  and  in  this  way  trans- 

1  Virchow's  Archiv,  1857,  xi,  132. 

»  Compt.  Rend,  de  1'Academie  des  Sciences,  Ivii,  220,  351,  386. 


THE  ANTHRAX  PROTEIN  191 

mitted  the  disease.  It  appears  that  Rayer  published  a  short 
note  of  this  work  in  the  Bull,  de  la  Soc.  de  Biologie  in  1850, 
but  we  have  not  had  access  to  this  publication.  Davaine's 
own  work  was  of  the  greatest  value  and  shows  great  skill 
for  that  time.  Probably  the  most  important  experiments 
that  he  made  were  those  in  which  he  demonstrated  that 
the  blood  of  an  animal  sick  with  anthrax  is  not  capable  of 
transmitting  the  disease  to  others  unless  it  contains  the 
bacillus.  It  may  be  of  interest  to  describe  briefly  the 
experiments  which  led  to  the  establishment  of  this  fact. 
Rabbit  A  was  inoculated  with  anthrax  blood.  Forty-six 
hours  later,  examination  showed  no  bacilli  in  the  blood  of 

A.  At  that  time  twelve  or  fifteen  drops  of  blood  were 
taken  from  the  ear  of  this  animal  and  injected  into  rabbit 

B.  Nine  hours  later  the  blood  of  A  was  reexamined  and 
found  to  contain  a  large  number  of  bacilli.    This  blood  was 
injected   subcutaneously  into  rabbit  C.     One  hour  later 
rabbit  A  died,  and  twenty  hours  later  C  died,  while  B 
remained  free  from  infection.     Space  will  not  permit  us 
to  follow  the  literature  of  anthrax  further,  save  those  parts 
that  bear  on  the  presence  of  a  chemical  poison.     Pasteur, 
De  Barry,  Koch,  and  others  studied  the  morphology,  life 
history,  and  cultural  characteristics  of  the  bacillus,  and  in 
this  way  founded  the  science  of  bacteriology.     The  reader 
is  referred  for  the  theories  of  the  action  of  this  bacillus 
to  the  works  of  Bollinger,1  Szpilman,2  Joffroy,3  Touisant,4 
and  Nencki.5 

Investigations. — In  1877  Pasteur  and  Joubert6  filtered 
anthrax  cultures  and  the  blood  of  animals  sick  of  this 
disease  through  porcelain,  and  injected  the  germ-free 
filtrate  into  animals  without  inducing  the  disease,  and 
concluded,  quite  properly,  that  this  bacillus  does  not 


1  Zur  Path,  des  Milzbrandes,  1872. 

2  Zeitsch.  f.  phys.  Chem.,  1880,  iv,  350. 

3  Compt.  Rend.  Soc.  de  Biol.,  1873  and  1874.  • 

4  Comp.  Rend,  de  1'Acad.,  1879,  xci,  195;  xciii,  163. 

6  Berichte  d.  deutsch.  Chem.  Gesellschaft,  1884,  2605. 
6  Comp.  Rend,  de  1'Acad.,  Ixxxiv,  905. 


192  PROTEIN  POISONS 

produce  a  soluble  poison.  Subsequent  investigations,  in 
our  opinion,  have  established  the  correctness  of  this  con- 
clusion. However,  there  have  been  several  claims  to  the 
discovery  of  soluble  poisons  in  cultures  of  the  anthrax 
bacillus,  and  in  the  bodies  of  animals  dead  with  this  disease, 
and  we  will  now  briefly  review  some  of  these  claims  which 
are  of  historical  interest. 

Hoffa1  obtained  from  pure  cultures  of  the  anthrax  bacillus 
small  quantities  of  a  substance  which  he  believed  to  be  a 
ptomain,  and  the  specific  poison  of  this  disease.  When  in- 
jected under  the  skin  of  certain  animals  it  at  first  increased 
the  respiration  and  the  action  of  the  heart.  After  a  short 
period  the  respirations  became  deep,  slow,  and  irregular. 
Later  the  temperature  fell  below  the  normal,  the  pupils  were 
dilated,  and  a  bloody  diarrhea  set  in.  Autopsy  showed 
the  heart  to  be  in  systole,  the  blood  dark,  and  ecchymoses 
were  found  on  the  pericardium  and  peritoneum.  Subse- 
quently, Hoffa  believed  that  he  had  succeeded  in  isolating 
this  substance  from  the  bodies  of  animals  dead  of  anthrax. 
He  named  it  anthracin,  and  undoubtedly  convinced  himself 
for  the  time  at  least  that  he  had  discovered  the  specific 
poison  of  this  disease.  No  subsequent  investigator — and 
several  have  repeated  the  experiments — has  been  able  to 
confirm  Hoffa's  results,  and  it  is  now  more  than  probable 
that  his  "anthracin"  resulted  from  the  action  of  the  agents 
used  in  its  detection  and  separation  upon  the  constituents 
of  the  fluids  with  which  he  worked. 

In  1889  Hankin2  grew  the  anthrax  bacillus  in  Liebig's 
meat  extract,  to  which  fibrin  had  been  added,  and  from 
this  filtered  culture  he  precipitated  with  ammonium  sul- 
phate an  albumose,  which,  while  not  directly  poisonous  to 
animals  when  injected  simultaneously  with  an  inoculation 
of  the  anthrax  bacillus,  caused  more  speedy  death  than 
when  the  bacillus  only  was  used.  He  concluded  that  the 
albumose  destroys  or  lessens  the  natural  resistance  of  the 


1  Ueber  die  Natur  des  Milzbrandgiftes,  1886. 

^  British  Med.  Jour.,  1890,  ii,  65;  Proc.  Royal  Soc.,  xlviii,  93. 


THE  ANTHRAX  PROTEIN  193 

animal  to  the  disease,  after  which  the  bacillus  is  able  to 
continue  the  elaboration  of  its  poison  in  the  body. 

Petermann1  repeated  Hankin's  experiments,  and  obtained 
an  albumose  which  elevates  the  temperature  from  one  to 
two  degrees,  but  is  otherwise  without  poisonous  effects  and 
without  protective  influence  against  the  anthrax  bacillus. 
Hankin  and  Wesbrook2  repeated  and  modified  the  experi- 
ments of  the  former,  and  reached  the  following  conclusions: 
(1)  The  anthrax  bacillus  elaborates  a  proteolytic  ferment 
by  means  of  which  albumose  may  be  formed  from  proteins, 
but  these  have  no  immunizing  action.  (2)  The  bacillus 
produces  another  albumose  which  is  not  due  to  the  soluble 
ferment,  but  to  an  intracellular  ferment.  (3)  This  albumose 
was  obtained  in  a  relatively  pure  condition.  This  was  done 
by  growing  the  bacillus  in  a  solution  of  pure  peptone.  It 
confers  partial  immunity  against  anthrax,  when  given  in 
small  doses  to  mice.  (4)  To  animals  susceptible  to  anthrax 
this  albumose,  in  ordinary  doses  at  least,  is  not  poisonous. 

(5)  Those  animals  which  are  relatively  immune  to  anthrax, 
such  as  the  rat  and  frog,  are  easily  poisoned  by  this  albumose. 

(6)  On  the  contrary,  young  rats  which  are  susceptible  to 
anthrax  are  not  poisoned  by  this  substance. 

Klemperer3  obtained  from  cultures  of  the  anthrax  bacillus 
a  substance  which  caused  elevation  of  temperature  when 
injected  subcutaneously,  but  which  was  not  submitted  to 
further  investigation.  Brieger  and  Frankel4  endeavored 
to  prepare  a  tox-albumin  from  the  organs  of  animals  dead 
of  anthrax.  They  cut  the  tissue  into  fine  pieces,  rubbed 
up  with  water,  allowed  to  stand  for  twelve  hours  in  an  ice- 
box, and  filtered  through  porcelain.  The  filtrate  was 
concentrated  in  vacuo  at  30°  to  one-third  its  volume,  and 
after  being  acidified  with  a  few  drops  of  acetic  acid,  was 
treated  with  ten  times  its  volume  of  absolute  alcohol.  The 
mixture  was  then  allowed  to  stand  twelve  hours  longer  in 
an  ice-box,  after  which  the  precipitate  was  collected  on  a 

1  Ann.  de  1'Institut  Pasteur,  1892,  vi,  32.  2  Ibid.,  vi,  633. 

3  Zeitsch.  f.  klin.  Med.,  1892,  xx,  165. 
«  Berl.  klin.  Woch.,  1890,  xxvii,  241,  268,  1133. 
13 


194  PROTEIN  POISONS 

filter,  dissolved  in  a  small  volume  of  water,  refiltered, 
reprecipitated  with  alcohol,  this  being  repeated  until  a 
perfectly  clear  aqueous  solution  was  obtained.  The  albu- 
mose  was  further  purified  by  dialysis,  and  as  thus  obtained, 
it  was  found  to  be  freely  soluble  in  water  and  to  give  the 
ordinary  reactions  for  proteins.  These  investigators  failed 
to  make  any  satisfactory  study  of  this  product,  and  the 
repetition  of  their  work  by  others  has  led  to  negative  results. 
Marmier1  attempted  to  isolate  a  poison  from  cultures 
grown  in  a  medium  of  the  following  composition: 

Water 1000.0 

Peptone 40.0 

Sodium  chloride 15.0 

Sodium  phosphate 0.5 

Potassium  phosphate 0.2 

Glycerin 40.0 

Before  inoculation  this  fluid  was  filtered  through  porcelain, 
and  then  sterilized  at  110°.  The  peptone  used  was  obtained 
from  the  commercial  preparation  by  the  precipitation  of 
the  albumoses  with  ammonium  sulphate,  and  the  salts 
were  removed  by  dialysis.  In  this  medium  the  anthrax 
bacillus,  especially  the  sporeless  form,  grew  abundantly. 
In  order  to  obtain  the  poison  the  culture  was  filtered  and 
saturated  at  room  temperature  with  ammonium  sulphate, 
which  produced  a  more  or  less  abundant  precipitate.  This 
was  allowed  to  stand  for  some  hours  and  filtered,  after  which 
the  precipitate  was  washed  with  a  saturated  solution  of 
ammonium  sulphate.  Subsequently  the  precipitate  was 
dissolved  in  water,  freed  from  salts  by  dialysis,  concen- 
trated, feebly  acidified  with  sulphuric  acid,  and  precipitated 
with  alcohol.  The  substance  thus  obtained  was  found  to 
be  soluble  in  water  and  in  a  1  per  cent,  solution  of  phenol. 
It  was  said  not  to  give  any  of  the  reactions  for  albumoses 
or  alkaloids,  but  this  can  scarcely  be  true.  This  work  has 
had  no  confirmation,  and  is  mentioned  here  simply  because 
of  its  historical  interest. 

1  Ann.  de  1'Institut  Pasteur,  1895,  ix,  533. 


THE  ANTHRAX  .PROTEIN  195 

Heim  and  Geiger1  grew  anthrax  bacilli  in  eggs  after  the 
method  of  Hueppe,  extracted  with  alcohol,  precipitated 
the  extract  with  mercuric  chloride,  filtered,  treated  the 
filtrate  with  platinum  chloride,  decomposed  the  precipitate 
thus  formed  with  hydrogen  sulphide,  filtered,  rendered 
alkaline  with  potassium  hydrate,  and  divided  into  two 
portions,  one  of  which  was  extracted  with  ether  and  the 
other  with  benzol.  The  amount  of  material  removed  with 
ether  was  small,  but  that  obtained  in  the  benzol  extract 
was  large.  When  either  of  these  residues  was  taken  up  in 
a  few  cubic  centimeters  of  feebly  acidified  water  and  injected 
intra-abdominally  into  mice,  it  caused  salivation  and 
lacrymation,  followed  by  muscular  convulsions  and  death. 
The  smallest  dose  of  the  benzol  extract  which  proved  to 
be  fatal  was  0.5  c.c.,  while  a  similar  amount  of  the  ether 
extract  caused  only  transient  symptoms.  Apparently  no 
controls  were  employed  by  these  investigators,  and  the 
evidence  that  they  obtained  any  poison  from  the  anthrax 
bacillus  is  too  slight  to  deserve  serious  attention. 

Ivanow2  has  demonstrated  the  presence  of  certain  volatile 
acids,  formic,  acetic,  and  caproic,  in  anthrax  cultures,  but 
there  is  no  proof  that  the  bacilli  had  anything  to  do  with 
the  production  of  these  bodies,  or  that  they  are  concerned 
in  any  way  in  the  symptomatology  or  pathology  of  the 
disease;  certainly  these  same  volatile  acids  are  found  in 
the  cultures  of  many  bacteria,  both  pathogenic  and  non- 
pathogenic. 

Petri  and  Massen3  detected  hydrogen  sulphide  in  anthrax 
cultures,  but  inasmuch  as  at  the  same  time  they  found  this 
gas  in  every  one  of  the  thirty-six  other  bacteria  examined, 
it  cannot  be  said  to  be  of  any  specific  importance.  More- 
over, spectroscopic  examination  of  anthrax  blood  fails  to 
show  the  presence  of  hydrogen  sulphide  or  any  of  its  com- 
pounds, and  there  is  no  evidence  that  this  gas  has  any 
connection  with  the  disease. 

1  Lehrbuch  der  Bakteriolog.  Untersuchungen  u.  Diagnostik,  1894,  229. 

2  Ann.  de  1'Institut  Pasteur,  1892,  vi,  131. 

3  Arbeiten  aus  d.  kaiserlich.  Gesundheitsamte,  1893,  viii,  318. 


196  PROTEIN  POISONS 

Fermi1  has  shown  the  presence  of  both  diastatic  and 
proteolytic  ferments  in  anthrax  cultures,  but  as  all  living 
cells,  including  bacteria,  elaborate  such  ferments,  this 
discovery  fails  to  make  us  acquainted  with  the  poison  of 
the  anthrax  protein.  Maumus2  found  that  by  its  growth 
on  potato  the  anthrax  bacillus  converts  some  starch  into 
sugar,  and  Reyer3  showed  the  presence  in  anthrax  cultures 
of  a  ferment  which  coagulates  casein. 

Klein4  removed  anthrax  bacilli  from  agar  cultures  of 
forty-eight  hours'  growth,  placed  them  in  5  c.c.  of  bouillon, 
and  after  the  tube  had  been  held  for  five  minutes  in  boiling 
water,  injected  the  contents  into  the  peritoneal  cavity  of 
a  guinea-pig,  without  results.  After  a  few  days  the  injec- 
tion was  repeated,  and  four  or  five  days  later  these  animals 
were  inoculated  subcutaneously  and  intra-abdominally, 
with  small  doses  of  a  living  culture.  All  died  within  forty- 
eight  hours,  of  typical  anthrax.  From  these  experiments 
Klein  concluded  that  the  anthrax  bacillus  contains  no 
intracellular  poison,  and  that  treatment  with  the  cellular 
substance  confers  no  immunity  on  guinea-pigs. 

Conradi5  attempted  to  solve  the  question  of  the  existence 
of  an  anthrax  poison  by  the  following  methods: 

1.  Guinea-pigs  were  inoculated  intraperitoneally  with 
anthrax,  and  immediately  after  death  the  peritoneal  fluid, 
which  varied  in  amount  in  different  individuals,  but  aver- 
aged from  10  to  15  c.c.,  and  contained  in  each  field  from  ten 
to  twenty  microorganisms,  was  filtered  through  porcelain. 
In  some  of  the  experiments  the  filter  of  Chamberland  was 
employed  while  in  others  that  of  Kitasato  was  used.  The 
filtered  peritoneal  exudate  was  injected  into  mice,  rats, 
and  guinea-pigs  subcutaneously,  intravenously,  and  intra- 
peritoneally, and  always  without  effect.  The  amounts  of  the 
filtered  exudate  injected  into  mice  varied  from  2  to  4  c.c.; 

1  Arch.  f.  Hygiene,  1890,  x,  1. 

2  Compt.  Rend,  de  la  Soc.  de  Biologie,  1893,  v,  1071. 

3  Ibid.,  309. 

4  Centralbl. -f .  Bakteriologie,  1894,  xv,  598. 

5  Zeitsch.  f.  Hygiene,  1899,  xxxi,  287. . 


THE  ANTHRAX  PROTEIN  197 

that  into  rats  from  5  to  12  c.c.;  in  guinea-pigs  from  4  to  15 
c.c.;  in  rabbits  from  10  to  20  c.c.;  and  in  one  dog,  25  c.c. 
was  injected  subcutaneously.  In  proportion  to  body  weight 
the  mice  received  by  far  the  larger  injections.  The  experi- 
ments indicate  that  in  the  peritoneal  exudate  of  guinea- 
pigs  inoculated  with  anthrax  there  is  no  appreciable  amount 
of  soluble  toxin. 

2.  Many  guinea-pigs  were  inoculated  with  anthrax,  and 
directly  after  death  their  livers  and  spleens  were  removed 
under   aseptic   conditions   and   rubbed   up   in   a   sterilized 
mortar  with  sterile  sand  to  which  a  little  physiological  salt 
solution   had   been   added.     After   thorough   rubbing   the 
mixture  was  diluted  with  physiological  salt  solution  and 
filtered  through  a  Chamberland  tube  under  four  atmos- 
pheres of  pressure.    The  filtrate  was  injected  subcutaneously, 
intravenously,  and  intraperitoneally  into  mice,  rats,  guinea- 
pigs,  and  rabbits,  and  in  every  case  without  effect. 

3.  Conradi,   finding   the   preparation   of   collodion   sacs 
difficult,   substituted  for  them  the  vegetable  membranes 
from    phragmites    communis,    first    used    by    Podbelsky.1 
These  sacs,  after    sterilization,   were   filled  with    bouillon 
cultures   of   the   anthrax   bacillus,    and   after   laparotomy 
under  ether  were  placed  in  the  abdominal  cavities  of  animals. 
The  animals  used  were  guinea-pigs,  rabbits,  and  dogs,  all 
of  which  remained  well,  notwithstanding  the  presence  of 
these  tubes  containing  virulent  cultures  of  the  anthrax 
bacillus  in  their  abdominal  cavities.     These  experiments 
satisfied  Conradi  that  the  anthrax  bacillus  does  not  produce 
any  soluble  toxin,  and  he  next  turned  his  attention  to 
determining  whether  or  not  this  organism  possesses  any 
intracellular  poison. 

4.  The  anthrax  exudates  obtained  as  described  in  the  first 
series,  in  quantities  of  from  5  to  6  c.c.,  were  placed  in  test- 
tubes,  0.5  c.c.  of  toluol  added  to  each,  the  tube  closed  with 
a  sterilized  cork,  thoroughly  shaken,  and  then  allowed  to 

1  To  one  experienced  in  the  preparation  of  collodion  sacs  this  must  be 
regarded  as  a  clumsy  substitution. 


198  PROTEIN  POISONS 

stand  for  ten  days  in  the  dark  at  room  temperature.  At 
the  expiration  of  this  time  the  contents  of  many  tubes  were 
placed  in  a  separator  and  the  toluol  removed.  Having  shown 
by  inoculation  that  the  germ  contained  in  these  tubes  was 
dead,  the  exudate  thus  sterilized  was  injected  into  suscep- 
tible animals  subcutaneously,  without  effect. 

5.  Asporogenous    cultures    were    sterilized    by    exposure 
for  one  hundred  and  ten  hours  to  — 16°.     After  exposure 
to  this  temperature,  the  tubes  were  kept  in  an  incubator 
at  20°,  long  enough  to  see  that  they  remained  sterile,  after 
which  they  were  injected  subcutaneously  into  susceptible 
animals,  and  always  without  effect. 

6.  A   number   of   rabbits   and   guinea-pigs   were   simul- 
taneously infected  with  anthrax,  and  after  death  the  livers 
and  spleens  were  subjected  to  a  hydraulic  pressure  of  500 
atmospheres.     The  fluid  thus  obtained,  which  on  micro- 
scopic examination  showed  the  presence  of  bacteria,  was 
passed  through  a  Chamberland  filter,  and  then  injected 
subcutaneously,  intraperitoneally,   and  intravenously  into 
mice,  rats,  guinea-pigs,  and  rabbits,  and  always  without  effect. 

7.  The  experiment  of  Brieger  and  Frankel  in  which  they 
prepared    their    anthrax    toxalbumin    was    repeated    with 
negative  results. 

From  these  experiments  Conradi  reaches  the  following 
conclusions:  "By  no  method  known  at  present  can  it  be 
shown  that  the  anthrax  bacillus  forms  either  an  extra- 
cellular or  an  intracellular  poison  within  the  bodies  of  either 
susceptible  or  insusceptible  animals.  Indeed,  these  experi- 
ments increase  the  probability  that  the  anthrax  bacillus 
does  not  form  any  poisonous  substance,  therefore  the  solu- 
tion of  the  manner  in  which  anthrax  infection  results  remains 
unknown.  Whether  improved  chemical  methods  will 
lead  to  its  detection  or  not  cannot  be  determined,  but  for 
the  present  the  anthrax  bacillus  at  least  must  be  regarded 
as  a  purely  infectious  microorganism."  If  this  conclusion 
reached  by  Conradi  be  true,  the  mechanical  interference 
theory  is  the  best  that  can  at  present  be  offered  so  far  as 
anthrax  is  concerned. 


THE  ANTHRAX  PROTEIN  199 

In  all  our  work  the  anthrax  bacillus  has  been  grown  in 
lloux  flasks,  as  we  have  not  dared  try  it  in  the  large  tanks, 
consequently  the  amount  of  cellular  substance  obtained 
has  been  small.  This  work  was  begun  in  1900  and  carried 
on  intermittently.  We  have  felt  it  desirable  to  exercise 
great  care  in  handling  quantities  of  the  anthrax  bacillus. 
We  have  always  opened  the  flasks  and  removed  the  growth 
over  a  shallow  tray  containing  some  powerful  germicide. 
We  will  make  a  few  extracts  from  our  protocols.  In  May, 
1900,  the  growth  was  removed  from  twelve  Roux  flasks, 
placed  in  one  liter  of  1  per  cent,  sulphuric  acid,  and  kept 
in  the  incubator  at  37°  for  twenty-four  hours.  At  the 
expiration  of  this  time  cultures  showed  the  bacillus  still 
alive.  The  suspension  was  then  placed  in  the  autoclave 
and  heated  at  100°  for  thirty  minutes.  Cultures  now  showed 
that  the  bacillus  had  been  killed.  The  suspension  was 
passed  through  a  Chamberland  filter  with  the  aid  of  a 
pump.  The  clear  filtrate  was  poured  drop  by  drop  into 
twice  its  volume  of  absolute  alcohol.  A  finely  flocculent, 
white  precipitate  formed,  was  collected  on  a  hard  paper, 
washed  with  alcohol  until  the  filtrate  was  no  longer  acid, 
and  dried  in  vacuo  over  potash.  This  substance  is  not 
colored  with  nitric  acid  and  heat,  but  on  the  addition  of 
ammonia  the  characteristic  orange  of  the  xanthoproteic 
test  is  developed  beautifully.  It  does  not  give  the  biuret 
and  other  protein  tests.  Later  an  unweighed  portion  of 
this  substance  was  dissolved  in  water  and  injected  intra- 
abdominally  in  a  guinea-pig.  Twelve  hours  later  the 
animal  was  found  dead.  The  heart  was  in  partial  diastole 
with  clots  in  both  auricles  and  ventricles.  The  peritoneal 
cavity  contained  a  few  cubic  centimeters  of  a  clear  fluid. 

Two  days  later  a  guinea-pig  received  intra-abdominally 
60  mg.  of  the  alcoholic  precipitate.  When  last  seen  that 
night,  five  hours  after  the  injection,  the  animal  was  dying. 
The  next  morning  it  was  posted  and  the  condition  described 
above  was  found.  The  heart's  blood  was  found  to  be  sterile. 
This  experiment  was  repeated  a  number  of  times,  with 
similar  results. 


200  PROTEIN  POISONS 

This  work  was  not  resumed  until  1902,  and  on  repeating 
the  work  the  alcoholic  precipitate  failed  to  manifest  any 
poisonous  action,  but  in  this  instance  we  heated  the  acid 
extract  for  two  hours.  Inasmuch  as  the  prolonged  heating 
seemed  to  be  the  only  difference  in  the  methods  of  pro- 
cedure, another  trial  was  made  in  which  the  acid  extract 
was  heated  in  the  autoclave  at  110°  for  exactly  ten  minutes. 
The  alcoholic  precipitate  freed  from  acid  as  before  was 
ground  to  a  fine  powder  in  an  agate  mortar.  An  unweighed 
portion  of  this  powder  was  dissolved  in  5  c.c.  of  water  and 
injected  intra-abdominally.  When  last  seen  that  night, 
nine  hours  after  the  injection,  the  breathing  was  difficult 
and  irregular.  The  animal  was  found  dead  the  next  morning. 
Autopsy  showed  extreme  subcutaneous  edema  over  the 
abdomen.  The  peritoneal  cavity  contained  a  few  cubic 
centimeters  of  a  clear  fluid,  and  a  smaller  amount  of  bloody 
exudate  was  found  in  the  pleural  cavity.  The  heart  was 
in  diastole  and  the  most  marked  changes  were  found  in 
the  lungs.  These  were  greatly  congested  and  the  left  upper 
lobe  seemed  to  be  consolidated.  Closer  examination  showed 
portions  of  the  lungs  to  be  completely  hepatized.  Many 
of  the  air  cells  were  filled  with  exudate  and  blood  corpuscles. 
The  kidneys  were  highly  congested  and  the  liver  seemed 
pale  and  friable.  Further  experiments  with  weighed  por- 
tions of  the  powder  showed  the  minimum  fatal  dose  for 
a  guinea-pig  when  given  intra-abdominally  to  be  about 
50  mg.  Smaller  doses  down  to  15  mg.  made  the  animals 
very  sick,  but  failed  to  kill. 

The  poisonous  group  obtained  from  the  anthrax  protein 
by  cleavage  with  1  per  cent,  sulphuric  acid  is  destroyed, 
at  least  greatly  weakened,  by  prolonged  boiling  in  aqueous 
solution. 

The  cellular  substance  of  the  anthrax  bacillus,  prepared 
by  our  method,  is  the  least  toxic  of  all  the  bacterial  proteins 
with  which  we  have  worked.  This  is  true  whether  the  cell 
protein  is  derived  from  a  pathogenic  or  a  non-pathogenic 
organism.  It  requires  not  less  than  JO  mg.  of  the  cell 
substance,  after  extraction  with  ohol  and  ether,  to 


THE  ANTHRAX  PROTEIN  201 

kill  a  guinea-pig  after  intra-abdominal  injection.  Smaller 
doses  make  the  animals  sick,  but  do  not  kill,  and  confer 
no  marked  immunity  to  inoculation  with  living  cultures. 

The  cell  substance  has  been  split  up  by  our  method  into 
poisonous  and  non-poisonous  portions.  The  former  differs 
in  no  recognizable  way  from  the  poisonous  group  obtained 
from  other  proteins,  and  treatment  of  animals  with  the 
latter  fails  to  establish  any  noticeable  immunity  to  inocula- 
tion with  the  bacillus. 

It  has  been  shown  that  the  poisonous  group  in  the  cellular 
protein  of  a  non-pathogenic  bacillus  may  be  more  effective 
than  that  in  the  anthrax  bacillus.  The  minimum  lethal 
dose  of  the  air-dried  cell  of  the  prodigiosus  for  a  guinea-pig 
of  from  200  to  300  grams  body  weight,  when  injected  intra- 
abdominally,  is  less  than  3  mg.,  while  that  of  the  anthrax 
bacillus  for  the  same  animal  is  about  200  mg.  Even  the 
lemon  sarcine,  the  least  toxic  of  the  non-pathogenic  organ- 
isms examined,  surpasses  the  anthrax  bacillus  in  toxic 
action.  These  facts  convince  us  that  the  pathogenicity  of 
a  bacterium  is  not  measured  by  its  capability  of  furnishing 
a  poisonous  group,  but  by  its  ability  to  grow  and  multiply 
in  the  animal  body.  The  high  degree  of  infectivity  shown 
by  the  anthrax  bacillus  in  some  animals  is  due  to  the  fact 
that  it  grows  without  hindrance  on  the  part  of  the  secre- 
tions of  those  animals.  On  the  other  hand,  its  failure  to 
infect  other  species  is  due  to  the  inhibiting  action  of  certain 
secretions  of  these  animals. 

Like  other  proteins,  that  of  the  anthrax  bacillus  contains 
a  poisonous  group.  The  chief  constituent  of  the  anthrax 
bacillus  is  a  glyconucleoprotein,  and  by  this  we  do  not 
mean  a  physical  mixture  of  carbohydrate,  nuclein,  and 
protein,  but  a  molecule  containing  these  constituents  as 
atomic  groups.  The  intracellular  poisons  contained  in 
bacterial  cells  are  not  preformed  toxins,  as  supposed  by 
Pfeiffer,  but  are  atomic  groups  in  a  complex  molecule. 
The  poison  can  be  obtained  from  the  protein  only  by 
processes  which  disrupt  the  molecule.  Mere  solvents, 
such  as  water,  alcohol,  ether,  saline  solution,  and  glycerin, 


202  PROTEIN  POISONS 

do  not  detach  the  poisonous  group.  We  have  obtained 
poisonous  substances  from  the  anthrax  bacillus  by  two 
methods.  The  substances  thus  obtained  differ  physically, 
chemically,  and  physiologically.  The  one  obtained  by  the 
action  of  1  per  cent,  sulphuric  acid  is  insoluble  in  alcohol 
and  doos  not  give  the  distinctive  protein  reactions.  The 
one  obtained  by  cleavage  of  the  bacterial  cell  with  a  2  per 
cent,  solution  of  sodium  hydroxide  in  absolute  alcohol  is 
soluble  in  alcohol  and  does  give  the  biuret  and  Millon 
reactions.  The  former  kills  only  after  some  hours,  and 
leaves  marked  pathological  changes.  The  other  kills  in  a 
few  minutes  and  leaves  no  gross  alterations.  This,  however, 
does  not  prove  that  the  poisonous  group  in  the  two  prepara- 
tions is  not  the  same.  It  may  be  that  in  the  one  the  poisonous 
group  is  still  closely  attached  to  other  groups,  and  energetic 
measures  may  be  necessary  to  tear  it  off,  and  as  a  result 
of  this  the  injury  done  to  the  body  cells  and  recognized 
at  autopsy  may  be  due.  In  the  other  preparation  the 
poisonous  group  is  already  detached,  and  consequently 
its  effects  are  manifest  immediately.  On  the  other  hand, 
our  work  does  not  show  that  the  poisonous  group  in  the 
two  preparations  is  the  same.  It  leaves  this  question  quite 
undetermined.  As  we  have  stated  elsewhere,  there  is 
probably  in  the  protein  molecule  a  whole  spectrum  of 
poisons,  one  derivable  from  the  other,  a  chain  of  poisonous 
groups,  one  differing  from  the  one  next  it  by  having  one 
more  or  one  less  link.  There  is  at  present  no  more  impor- 
tant and  no  more  difficult  subject  than  that  of  the  chemistry 
of  the  protein  molecule.  The  researches  of  Fischer  have 
done  much  to  show  some  features  of  the  structure  of  the 
protein  molecule.  We  know,  as  a  result  of  Fischer's  work, 
that  proteins  are  to  be  regarded  as  polymers  or  condensa- 
tion products  of  the  amino  acids,  but  between  the  native 
protein  and  the  amino  acids  into  which  it  may  be  split, 
there  is  a  long  list  of  intermediary  products  about  which 
we  know  practically  nothing.  The  ordinary,  native  proteins 
are  not  primarily  poisons.  The  amino-acids  which  result 
from  their  ultimate  cleavage  are  not  poisonous,  but  between 


THE  ANTHRAX  PROTEIN  203 

the  two  there  are  many  split  products,  of  varied  sizes, 
which  are  poisonous.  Besides,  some  of  the  amino-acids 
may  be  converted  into  highly  poisonous  substances. 
Whether  a  given  protein  molecule,  on  being  disrupted,  sup- 
plies an  active  poison  or  not  is  determined  by  the  lines  of 
cleavage  and  these  are  dependent  upon  the  cleavage  agent 
and  the  conditions  under  which  it  acts.  The  work  of  Fischer, 
as  valuable  as  it  is,  has  been  and  is  of  little  or  no  service  in 
elucidating  the  processes  of  parenteral  digestion,  which  must 
be  better  understood  before  we  can  read  the  first  line  in  the 
true  history  of  disease,  either  exogenous  or  endogenous. 

Rosenau  and  Anderson1  sensitized  animals  by  subcu- 
taneous injections  of  extracts  of  the  anthrax  bacillus. 
Sobernheim2  was  not  able  to  confirm  this  work,  and  made 
some  statements  that  deserve  attention.  He  said  that  the 
cell  substance  of  the  anthrax  bacillus  is  quite  different 
chemically  and  biologically  from  that  of  other  bacteria, 
and  that  it  is  wholly  devoid  of  poisonous  properties,  what- 
ever the  amount  and  method  of  administration  may  be. 
Our  own  work,  as  already  stated,  shows  that  this  is  not 
true.  Busson3  has  reinvestigated  this  question  of  sensiti- 
zation  with  anthrax  protein.  Preisz4  has  shown  that 
when  the  anthrax  bacillus  is  grown  at  a  high  temperature 
(42.5°  C.)  after  the  manner  used  by  Pasteur  in  preparing 
his  vaccine,  the  membrane  becomes  mucilaginous  and 
more  permeable.  With  bacilli  thus  prepared  Busson  suc- 
ceeded in  inducing  a  mild  form  of  sensitization  by  intra- 
peritoneal  injections.  The  sensitized  state  was  recognized 
by  a  more  marked  elevation  of  temperature  over  the  con- 
trols on  reinjection.  It  is  undoubtedly  true  that  the  anthrax 
bacillus  is  protected  by  its  capsule  against  the  action  of 
ferments  produced  in  the  bodies  of  infected  animals,  but 
that  anthrax  protein  is  so  radically  different  from  other 

1  Hygienic  Lab.  Bull.,  1907,  No.  36. 

2  Kraus  und  Levaditi,  Handbuch  d.  Technik  u.  Methodik  d.  Immuni- 
tatsforschung,  ii. 

3  Zeitsch.  f.  Immunitatsforschung,  1912,  xii,  671 

4  Centralbl.  f.  Bak.,  1911,  Iviii. 


204 


PROTEIN  POISONS 


bacterial  proteins  is  an  unwarranted  assumption,  and  that 
it  is  not  poisonous  in  any  dose  we  have  shown  not  to  be 
true.  When  the  anthrax  protein  is  obtained  in  solution 
without  alteration  of  its  constitution,  and  when  this  solu- 
tion is  properly  administered  we  dare  say  that  it  will  be 
found  to  sensitize  animals  as  well  as  any  other  protein. 
As  we  have  had  occasion  to  point  out  more  than  once, 
it  is  necessary  to  have  a  protein  in  solution  in  order  to 
develop  exquisite  sensitization,  and  it  must  be  in  solution  on 
reinjection  in  order  to  induce  the  most  striking  form  of 
anaphylactic  shock.  Permeation  of  the  body  cells  seems  to 
be  essential  to  the  most  complete  sensitization,  also  to  its 
demonstration  on  reinjection. 

Roos1  has  shown  that  salvarsan  is  an  effecient  germicide 
for  the  anthrax  bacillus,  both  in  vitro  and  in  vivo,  and 
Becker2  and  Bettman3  have  successfully  treated  anthrax 
in  man  with  this  preparation. 

1  Zeitsch.  f.  Immunitatsforschung,  1912.  xv,  487. 
z  Deutsch.  med.  Woch.,  1911. 
3  Ibid.,  1912. 


CHAPTER  X 

THE  CELLULAR  SUBSTANCE  OF  THE 
PNEUMOCOCCUS  * 

The  Strain. — The  strain  of  the  pneumococcus  with  which 
this  work  was  done  was  presented  us  by  Dr.  J.  J.  Kinyoun, 
of  Washington.  When  the  culture  was  received,  a  mouse, 
guinea-pig,  and  rabbit  received  intraperitoneal  inoculations. 
The  mouse  died  in  twenty-three  hours,  the  guinea-pig  in 
twenty-four,  and  the  rabbit  in  twenty-seven.  Cultures 
were  made  from  the  heart  blood  of  each  of  these  animals, 
and  all  found  to  be  pure.  Our  growths  were  made  in  5  per 
cent,  glycerin  bouillon,  and  with  these,  Roux  flasks  and 
the  tanks  containing  a  medium  made  of  3  per  cent,  of  agar 
and  1  per  cent,  of  ox  serum  in  the  5  per  cent,  glycerin 
bouillon  were  inoculated.  The  flasks  and  tanks  were  kept 
at  38°  for  four  days,  when  the  growth  was  removed.  The 
growth  seemed  to  reach  maturity  in  this  time  at  the  tem- 
perature mentioned.  When  kept  longer  it  began  to  dry 
and  contract.  In  some  instances  the  growth  was  not 
harvested  until  the  sixth  day.  The  cellular  substance  thus 
obtained  was  handled  in  the  usual  manner,  i.  e.,  it  was 
thoroughly  extracted  with  alcohol  and  ether. 

The  strain  was  found  to  be  highly  virulent  and  remained 
so  throughout  the  year  of  work  with  it.  Fig.  10  shows  the 
effects  of  the  living  organism  on  guinea-pigs  after  intra- 
abdominal  inoculation  of  0.00001,  0.000001,  and  0.0000001 
c.c.  of  a  twenty-four-hour  bouillon  culture. 

Fig.  11  shows  the  relative  effects  of  the  living  organ- 
ism and  5  mg.  of  the  cellular  substance.  It  will  be  seen 

1  The  first  part  of  this  chapter  is  founded  on  work  done  in  the  Hygienic 
Laboratory  of  the  University  of  Michigan  in  1905-06  by  Dr.  J.  F.  Munson. 


206 


PROTEIN  POISONS 


that  with  the  living  organism  there  was  no  marked  fall 
in  temperature  until  about  the  tenth  hour,  when  the  fall 
became  evident  and  continued  until  death.  This  held 


FIG.  10 


\ 


FIG.  11 


TIME 

TEMP.  I 

TEMPI! 

v        • 

. 

3 

1 

0       1 

1     1 

.'     1 

.,    1 

4      1 

•       1 

6       1 

;     1 

8       1 

j     : 

c    : 

1 

N.T. 



99.9 

/ 

\ 

1  HR. 

101.4 

98.2 

v 

/ 

S 

*" 

•--. 

/ 

2  HR8. 

99.7 

100. 

\ 

/ 

^N 

/ 

\ 

3  MRS. 

99.4 

01.3 

\ 

/ 

\ 

\ 

I 

5  MRS. 

98.2 

00.1 

A 

/ 

\ 

\ 

6  MRS. 

96.8 

00.6 

\ 

\ 

7  HR8. 

96.2 

01. 

—  \ 

'  

— 

s 

VI 

a  H  s. 



00.7 

^ 

j 

k 

\ 

96 

»Jn  s. 

91.8 



\ 

I 

S 

r 

\ 

10  H    S. 

96.4 

01.9 

\ 

—  f 

— 

\ 

as 

1Z  H    S. 

95 

02.2 

\ 

\ 

14  H    S. 

55 

99.5 

\ 

21   H    8 

/ 

93 

HF- 

COVEREC 

\l 

\ 

\ 

\ 

91 

\ 

-90- 

good  with  some  variations  in  the  exact  time  when  the  fall 
began  in  all  our  experiments  with  the  living  organism  when 
twenty-four-hour  cultures  were  employed.  During  the 
time  before  the  fall  begins,  the  organism  is  growing  and 
multiplying.  The  fall  indicates  dissolution  of  the  cell  and 
the  liberation  of  its  poisonous  constituents.  When  older 
cultures  are  used,  especially  when  the  amount  is  large,  the 
fall  in  temperature  may  appear  much  earlier,  and  is  due  to 
the  presence  of  autolyzed  cells  in  the  cultures.  This  is 
true  not  only  of  the  pneumococcus,  but  of  the  cholera, 
typhoid,  and  other  bacteria. 

The  cellular  substance  of  the  pneumococcus,  prepared 
by  our  method,  is  a  white  powder,  in  which  the  individual 
cells  readily  take  the  stains,  and  it  is  found  to  be  quite  free 
from  debris.  We  administered  it  in  suspension  in  sterile 
salt  solution,  and  generally  intraperitoneally.  It  seems 
to  be  more  irritating  than  other  bacterial  cellular  sub- 
stances with  which  we  have  worked,  and  even  5  mg.  sus- 
pended in  5  c.c.  of  salt  solution  and  injected  into  the  peri- 
toneal cavity  seems  to  cause  pain.  The  animal  soon  becomes 
quite  normal  in  appearance,  and  remains  so  for  an  hour 
or  two,  when  the  fur  behind  the  ears  begins  to  roughen  and 
gradually  the  whole  coat  takes  on  this  state.  The  posterior 
extremities  become  weak  and  the  animal  is  unable  to  main- 
tain the  erect  posture.  The  weakness  intensifies  into  a 
paralytic  state,  and  finally  the  animal  lies  stretched  out 
on  its  side,  and  seems  quite  unable  to  make  a  struggle. 
Rarely  there  are  convulsions,  but,  as  a  rule,  respiration 
slowly  and  quietly  fails,  and  it  is  often  difficult  to  tell 
just  when  it  stops.  Following  these  symptoms,  the  tem- 
perature which  at  first  may  be  slightly  elevated  grad- 
ually falls,  and  has  been  frequently  at  85°,  rarely  at 
75°,  before  respiration  wholly  ceases.  When  the  dose 
is  a  non-fatal  one  the  lowest  point  is  usually  reached 
about  the  seventh  hour,  when  the  temperature  rises  as 
gradually  as  it  fell.  The  cellular  substance  of  the  pneu- 
mococcus is  not  highly  poisonous  compared  with  similar 
preparations  from  other  bacteria.  It  is  rather  interesting 


208  PROTEIN  POISONS 

to  make  some  comparisons  here.  As  has  been  stated,  the 
strain  with  which  our  work  was  done  was  highly  virulent, 
killing  half-grown  guinea-pigs  in  doses  of  0.0000001  c.c. 
of  a  twenty-four-hour  culture  given  intraperitoneally.  At 
the  same  time  our  old  stock  culture  of  the  pneumococcus 
did  not  kill  in  doses  of  less  than  1  c.c.,  and  yet  the  cellular 
substances  of  the  two,  measured  by  toxicity,  were  prac- 
tically the  same.  This  and  similar  observations  wTith  other 
bacteria  lead  us  to  conclude  that  virulence  is  measured 
by  rate  of  multiplication  and  not  by  chemical  differences 
in  cellular  poison  content.  Moreover,  when  two  animals 
were  killed  with  the  two  strains  the  cells  seemed  to  be  as 
abundant  in  one  as  in  the  other.  The  more  virulent  strain 
multiplies  the  faster.  Virulence  may  depend  upon  several 
factors,  but  rate  of  multiplication  is  certainly  one  of  them, 
and  on  a  common  medium  as  the  animal  body  this  must 
depend  upon  the  effectiveness  of  the  ferments  whose  func- 
tion it  is  to  prepare  and  utilize  the  pabulum  on  which 
the  organism  feeds.  Our  highly  virulent  strain  furnished 
a  cellular  substance  which  killed  guinea-pigs  in  doses  of 
1  to  10,000.  Occasionally  smaller  doses  killed.  The  smallest 
fatal  dose  on  first  injection  of  which  we  have  a  record  was 
1  to  19,000,  but  the  surely  fatal  minimum  was  1  to  10,000, 
and  as  we  have  stated,  the  less  virulent  strain  of  pneumo- 
coccus furnished  cellular  substance  of  the  same  degree  of 
toxicity.  A  comparison  of  the  virulent  strain  of  the  pneu- 
mococcus with  our  strain  of  colon  is  also  of  interest.  With 
our  colon  bacillus  the  minimum  constantly  fatal  dose  was 
1  c.c.  of  a  twenty-four-hour  bouillon  culture.  Sometimes 
a  dose  of  0.5  c.c.  killed,  and  the  smallest  fatal  dose,  as  we 
found  it,  was  0.25  c.c.  This  organism  yielded  a  cellular 
substance,  which  as  a  coarsely  ground  powder  always  killed 
1  to  50,000;  when  finely  ground  it  killed  1  to  75,000,  and 
sometimes  as  high  as  1  to  2,000,000.  Our  virulent  pneumo- 
coccus killed  in  0.0000001  c.c.  doses,  and  yielded  a  cellular 
substance  which,  when  ground  to  the  finest  possible  powder, 
killed  only  1  to  10,000.  Surely  these  are  strong  arguments 
for  our  belief  that  the  pathogenicity  of  a  microorganism  is 


not  measured  by  its  poisonous  cell  content,  but  by  the  rate 
with  which  it  multiplies  in  the  animal  body  or  the  intensity 
and  rapidity  with  which  it  converts  body  proteins  into  its 
own  proteins. 

It  must  be  borne  in  mind  in  considering  what  we  are 
about  to  say  in  this  paragraph  that  at  the  time  these  experi- 
ments were  conducted  we  knew  but  little  about  protein 
sensitization,  and  they  were  not  conducted  with  the  phe- 
nomena of  sensitization  in  view.  Had  we  known  then  what 
we  now  know  the  lines  of  investigation  would  have  been 
drawn  somewhat  differently.  However,  this  makes  a  review 
of  our  old  protocols  all  the  more  interesting  and  valuable. 
We  tried  to  immunize  animals  with  the  cellular  substance. 
It  will  be  worth  while  to  follow  one  set  of  these  experiments 
through.  We  take  the  three  tables  on  p.  210  just  as  they 
stand  in  the  protocol. 

It  will  be  observed  that  in  the  second  and  third  injections 
made  at  intervals  of  five  and  six  days  we  killed  one-third 
of  our  animals.  Now  we  know  that  this  was  due  to  the 
fact  that  we  partially  sensitized  the  animals. 

Failing  absolutely  to  even  render  our  animals  tolerant 
to  the  dead  germ  substance,  we  tried  to  weaken  it  by  heat, 
but  in  this  we  were  equally  unsuccessful.  However,  we 
did  prove  that  heating  the  cellular  substance  of  the  pneu- 
mococcus  to  144°  for  five  minutes  in  the  autoclave  does 
not  destroy  its  intracellular  poison.  We  also  found  that 
by  heating  the  cells  some  of  the  poison  passes  into  solution, 
and  may  be  filtered  through  porcelain. 

We  split  up  the  cellular  substance  with  a  2  per  cent, 
solution  of  sodium  hydroxide  in  absolute  alcohol,  and 
obtained  a  non-poisonous  and  a  poisonous  portion.  In 
both  small  and  large  doses  the  former  had  no  visible  effect 
on  animals,  but  it  gave  no  immunity  to  subsequent  inocu- 
lations. 

The  poisonous  fraction  kills  animals  in  about  the  same 

doses  as  are  required  by  similar  preparations  from  other 

proteins.     The  symptoms  are  not  wholly  identical  with 

those  induced  by  poisons  obtained  from  other  proteins. 

14 


TABLE  XXI. — MARCH    9,    1905.      CAGE    VI.      PNEUMOCOCCUS    GERM    SUBSTANCE; 
DOSES  SUSPENDED  IN  5  c.c.  SALT  SOLUTION. 


"§ 

o. 
g 

a 

a 

a 

a 

3 

a 

ft 

d 

d 

ft 

a 

| 

a 

a 

> 

si 

41 

I 

§ 

a 

i 

i 

a 

8 

I 

§ 

•i 

» 

| 

9 

e 

M~ 

i* 

1 

S 

1 

1 

s 

i 

8 

•p 

h 

-iS 

I*' 

r^ 

-^> 
L 

00 
O 
P 

3 

S 

'3 
£ 

O 

P 

PQ 

tH 

1 

XJ 

1-1 

jg 

<N 

XI 
CO 

J3 
•* 

xi 

>o 

A 
t- 

O 

<N 

o 
IN 

12.20 

Black 

340 

2.5 

136,000  100.0 

102.9 

104.3 

99.0 

98.9 

98.0 

96.7 

99.0 

99.7 

99.5 

12.18 

Gray 

350    5.0    70,000  100.2 

101.5 

103.4 

100.4 

97.1  + 

96.6 

94.2—  95.0 

94.8 

98.2 

12.17 

Gray  curly 

370  10.0    37,000  100.3 

100.1 

99.9 

97.6 

97.6 

98.8 

97.4     194.2— 

99.2 

100.8 

12.15 

Wh.,  yel., 

black 

390  15.0    26,000  100.8 

101.3 

101.2 

96.4 

98.1 

PS  6 

95.8 

94.2— 

976 

101  4 

12.10 

Yel.,  white 

39520.0    19.750  101.0 

101.7 

102.4 

99.2 

98.9 

P76 

94.6 

96.8 

P66 

1009  + 

12.05 

Yellow 

025 

25.0    25.000  101.2 

102.0 

102.7 

101.0 

98.1 

P73 

97.6 

97.4 

PP3 

10">4 

TABLE  XXII. — MARCH  14,  1905.  THIS  is,  so  FAR  AS  POSSIBLE,  AN  EXACT 
REPETITION  OF  THE  WORK  OF  MARCH  9,  1905,  USING  THE  SAME  PIGS.  DOSES 
GIVEN  AT  11.45  A.M. 


Black 
Gray 

Gray,  curly 
Yel.,  white, 

black 

Yel.,  white 
Yellow 


1 

521 
508 
545 

542 
545 

P 

2.5 

50 

PQ 

-S 

o 
£ 

a 

I 

a 

! 

a 

I 

1 

ft 

a 
1 

XI 
00 

i 
i 

o 

101.0 
100.6 

ft 
i 

XI 
(N 

XI 

XI 

co 

X) 

98.4 
100.2 

XI 

XI 

96.4 
P87 

208,400 
101,600 

100.9 
100.2 

102.6 
105.4 

105.0 
103.0 

103.0 
103.8 

98.8 
100.1 

102.3 
101.2 

101.9 
100.8 

Died 
Died 

10.0 
150 

54,550 
36,133 

101.7 
101.2 

103.3 
102.7 

104.1 
100.5 

102.4 
100.0 

99.3 
99.8 

98.4 
98.1 

95.3 
98.1 

100.2 
90.7 

101.0 
91.4 

101.8 
95.4 

20.0 

27,250 

100.4 

100.6 

101.2 

97.9 

96.6 

95.2 

87.8    86.7 

88.7 

90.3 

601 

25.0 

24,040 

102.0 

103.0 

103.7 

100.1 

100.0 

97.0 

91.4 

92.5 

97.0 

97.3 

TABLE  XXIII. — MARCH  20,  1905.  CAGE  VI.  THE  USUAL  DOSES  WERE  GIVEN 
(THIRD  TIME).  ANIMALS  REACTED  MORE  SEVERELY  THAN  BEFORE.  AN  IMPRESSION 
DRAWN  FROM  THEIR  BEHAVIOR.  Nos.  3  AND  6  DIED.  PLACING  IN  INCUBATOR  DID 

NOT  SAVE  THEM 


Sf 

£ 

4! 

• 

•si 

a 
p 

PQ 

1 
i 

| 

B 

o 

a 

S 

t-I 

a 
c 

CM 

a 
ri 

CO 

ft 

a 
i 

a 

XI 

co 

f 

£ 

H« 

a 

J3 
0? 

PI  4 

S 

S 

o 

1 

M 

>D 

Black 
Gray 
Gray,  curly 
Yellow 

550    2.5 
542    5.0 
552  10.0 
60925.0 

99.6 

103.4 

104.3 

102.6 

P6  4 

90.5 

PI  4 

PO  5 

96  7 

Dead 
Dead 

101.0 
100.6 

103.6 
100.0 

103.2 
100.3 

101.7 
96.0 

97.1 
86  9 

92.392.7 
86.9  84.2 

95.0 
P6  8 

93.2 

98.2 

101.0 

98.2 

97.8 

93.2 

PO  5 

86.086.0 

P5  P 

The  temperatures  of  Nos.  3  and  6  at  nine  and  one-half  hours  rose  because  they  were  placed 
in  the  incubator. 


THE  CELLULAR  SUBSTANCE  OF    PNEUMOCOCCUS      211 

The  difference  lies  in  the  less  marked  convulsive  character 
of  the  third  stage.  When  injected  intraperitoneally  in 
guinea-pigs,  the  coat  soon  roughens,  and  for  some  minutes 
the  animal  seems  quiet.  Weakness  of  the  hind  limbs 
develops  and  the  upper  part  of  the  body  is  shaken  by 
spasms  resembling  severe  hiccough.  The  paralysis  rapidly 
develops  and  spreads,  and  the  animal  lies  on  one  side. 
It  dies  in  most  instances  without  general  convulsions. 
The  respiration  becomes  slower  and  the  heart  continues  to 
pulsate  for  some  minutes  after  respiration  has  stopped. 
Death  occurs  in  from  twenty  minutes  to  an  hour.  The 
temperature  curve  begins  to  fall  soon  after  the  injection, 
and  continues  until  death.  In  cases  of  recovery  the  first 
sign  of  improvement  is  a  rise  in  temperature,  and  it  comes 
up  more  slowly  than  it  went  down. 

It  is  worthy  of  note  that  in  case  of  inoculation  with  the 
living  organism  there  is  an  incubation  period  of  about  ten 
hours.  This  is  followed  by  the  complete  triumph  of  the 
infection,  and  is  shown  by  the  even  and  constant  fall  in 
temperature.  With  the  dead  cellular  substance  the  incu- 
bation period  is  shortened,  but  the  character  of  the  fall  is 
the  same.  With  the  free  poison  there  is  no  period  of  incu- 
bation, and  the  temperature  begins  to  fall  in  a  few  minutes. 
In  all,  the  temperature  curve  is  the  same  in  general  char- 
acter. Certainly,  it  must  be  true  that  the  poison  which 
affects  and  kills  the  animal  must  be  the  same  in  the  living 
and  dead  cell,  and  in  the  split  product. 

We  demonstrated  that  the  poisonous  portion,  like  that 
obtained  from  other  proteins,  establishes  on  repeated 
injections  of  non-fatal  doses  a  certain  degree  of  tolerance, 
but  gives  no  immunity  against  infection. 

The  recent  work  of  Rosenow1  on  the  autolytic  cleavage 
products  of  the  pneumococcus,  and  certain  other  bacteria 
is  of  great  interest  and  value.  The  pneumococcus  readily 
undergoes  autolysis,  and  Rosenow  has  studied  the  products 
resulting  in  this  way.  The  following  are  some  of  the  more 

1  Jour.  Infect.  Dis.,  ix,  190;  x,  113;  xi,  286,  480. 


212  PROTEIN  POISONS 

important  facts  demonstrated  by  this  investigator:  (1) 
Animals  may  be  sensitized  with  dead  pneumococci  or 
with  extracts  from  the  same.  The  sensitizing  dose  may  be 
given  subcutaneously,  intraperitoneally,  intravenously,  or 
intrapleurally.  In  order  to  induce  anaphylactic  shock  the 
reinjection  must  be  made  intravenously  or  intracardiacly. 
In  the  sensitized  animal  both  dead  and  living  pneumococci 
are  dissolved  more  rapidly  than  in  normal  animals.  This 
explains  the  slight  but  definite  immunity  to  virulent  cultures 
manifested  by  sensitized  animals.  (2)  Fresh  pneumococci 
suspended  in  salt  solution  and  kept  at  37°  for  forty-eight 
hours,  under  ether  or  over  chloroform,  undergo  autolysis 
by  which  a  poison  is  liberated.  This  poison  injected 
intravenously  or  intracardiacly  in  normal  animals  causes 
anaphylactic  shock.  In  guinea-pigs  this  poison  induces 
death  by  spasm  of  the  bronchioles  and  consequent  arrest 
of  respiration.  In  dogs  it  causes  marked  fall  in  blood- 
pressure  and  delays  the  coagulation  of  the  blood.  This 
poison  is  split  off  from  the  pneumococcus  protein  not  only 
in  autolysis,  but  also  by  normal  and  immune  sera  and  by 
leukocytic  extracts.  (3)  The  cleavage  of  pneumococcus 
cell  substance  by  autolysis  or  the  other  agents  mentioned, 
is  accomplished  by  proteolytic  ferments,  as  is  shown  by 
increased  production  of  amino  bodies  as  the  poison  is  set 
free.  Finally,  the  digestive  process  reaches  a  point  when 
the  poison  itself  is  digested  and  rendered  inert.  "The  fact 
that  virulent  pneumococci  have  within  themselves  a  proteo- 
lytic enzyme  which  splits  their  protein  into  a  highly  toxic 
substance,  is  strong  indication  that  certain  strains  of  pneu- 
mococci may  cause  infection  forthwith  without  first  rendering 
the  host  allergic.  This  is  quite  in  keeping  with  the  fact 
that  in  pneumococcus  infections  an  incubation  period  is 
not  an  invariable  rule.  On  the  other  hand,  in  certain 
instances,  a  previous  sensitization  before  symptoms  set 
in,  probably  occurs.  This  might  well  be  the  case  in  lobar 
pneumonia  when  the  chill  occurs  a  week  or  ten  days  after 
the  patient  contracted  a  severe  cold  or  bronchitis.  The 
distribution  by  lobes  in  typical  cases  may  be  related  to 


THE  CELLULAR  SUBSTANCE  OF    PNEUMOCOCCUS     213 

the  bronchial  spasm  which  this  toxic  substance  produces. 
That  early  dyspnea  and  increased  respiration  before  con- 
solidation is  demonstrable  is  in  keeping  with  this  idea. 
(4)  Morphine,  ether,  urethane,  atropine,  and  adrenalin, 
protect  normal  guinea-pigs  against  the  toxic  material 
obtained  in  vitro  from  pneumococci,  and  also  sensitized 
guinea-pigs  on  reinjection." 

Recently  (December,  1912)  we  found  a  small  bottle  of 
the  powdered  pneumococcus  cellular  substance  prepared 
by  Munson  nearly  seven  years  before  (March,  1906).  It 
is  a  fine,  yellowish-white  powder,  looking  very  much  like 
wheat  flour.  It  has  stood  during  these  years  in  a  cupboard, 
kept  closed  except  when  momentarily  opened  to  put  some- 
thing in  or  take  something  out.  Microscopic  examination 
showed  the  pneumococci  as  clearly  and  in  as  perfect  form 
as  in  a  fresh  preparation.  It  kills  guinea-pigs  on  intra- 
abdominal  injection  in  the  same  doses  (1  to  10,000  of  body 
weight),  and  just  as  promptly  as  it  did  more  than  six  years 
ago.  Five  hundred  milligrams  of  this  was  weighed,  sus- 
pended in  500  c.c.  of  salt  solution,  10  c.c.  of  chloroform 
added,  and  the  whole  allowed  to  stand  at  37°.  After  twenty- 
four  hours,  10  c.c.  of  the  opalescent  supernatant  fluid  was 
injected  into  the  external  jugular  vein  of  a  guinea-pig. 
Within  two  hours  the  rectal  temperature  had  fallen  below 
94°,  and  the  animal  remained  sick  for  some  hours,  but 
gradually  recovered.  The  same  experiment  repeated  at 
the  end  of  forty-eight  and  seventy-two  hours  killed  the 
guinea-pigs  within  two  hours.  These  animals  died  with 
the  symptoms  of  a  subacute  anaphylactic  shock.  We 
conclude  from  this  that  the  intracellular  autolytic  ferment 
had  remained  intact  during  the  years  that  had  elapsed 
since  the  preparation  of  the  cellular  protein.  Six  days 
after  the  suspension  had  been  prepared  and  placed  in  the 
incubator  a  like  injection  killed  the  guinea-pig  within  three 
minutes.  This  animal  died  with  the  symptoms  of  acute 
anaphylactic  shock,  and  autopsy  showed  the  lungs  distended 
and  minute  petechial  hemorrhages  in  the  pericardium. 


CHAPTER  XI 
PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS 

Introduction. — The  older  medical  literature  occasionally 
records  facts  which  in  the  light  of  more  recent  and  extended 
knowledge  are  known  as  the  phenomena  of  protein  sensi- 
tization.  Such  were  some  of  the  experiences  recorded  in 
the  early  attempts  at  the  transfusion  of  blood.  Many  of  the 
untoward  results  reached  in  this  procedure  and  beyond 
the  ken  of  that  time  are  now  fully  explained.  Behring  and 
Kitashima1  found  on  immunizing  an  animal  to  tetanus 
toxin  that  it  died  in  convulsions  notwithstanding  the  fact 
that  the  blood  serum  was  richly  charged  with  antitoxin. 
They  explained  this  by  assuming  the  existence  of  a  con- 
dition of  "  hypersensitiveness"  to  the  toxin.  With  our 
present  knowledge  we  see  no  reason  for  ascribing  this  to 
the  toxin.  There  is,  so  far  as  we  know,  no  evidence  that 
animals  can  be  rendered  hypersensitive  to  either  toxin  or 
antitoxin.  Neither  has  ever  been  obtained  free  from  pro- 
teins, and  since  all  true  proteins,  so  far  as  we  know,  sensitize, 
there  seems  no  sufficient  justification  in  ascribing  a  sensi- 
tization  induced  by  a  protein  solution  containing  a  toxin 
to  the  latter.  Buchner2  repeatedly  injected  bacterial  pro- 
teins into  men  and  noticed  that  the  cardinal  indications 
of  local  inflammation,  tumor,  rubor,  dolor,  and  calor 
resulted.  Furthermore,  he  noted  that  fever  increased  with 
repeated  injections.  Krehl  and  Matthes3  induced  fever  in 
animals  by  repeated  injections  of  albumose  and  peptone. 
Weichardt4  made  an  advanced  study  in  the  domain  which 

1  Berl.  klin.  Woch.,  1901,  No.  6. 

2  Berl.  klin.  Woch.,  1890,  216;  Munch,  med.  Woch.,  1891,  No.  3. 

3  Arch.  f.  exper.  Path.  u.  Pharm.,  1895,  xxxv,  232;  ibid.,  1896,  xxxvi,  437. 

4  Berl.  klin.  Woch.,  1903,  No.  1. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      215 

we  now  designate  as  anaphylaxis.  He  repeatedly  treated 
rabbits  with  protein  expressed  from  placental  cells,  and 
found  that  some  of  these  died  promptly  on  subsequent 
injections.  Furthermore,  he  mixed  the  serum  of  animals 
thus  treated  with  placental  cells  and  obtained  a  soluble 
poison  which  he  named  synzytiotoxin.  Later,  he  showed 
that  hay  fever  results  from  the  parenteral  digestion  of  the 
proteins  of  pollen.  Both  of  these  points  will  be  discussed 
in  more  detail  later.  Wolff-Eisner1  discussed  the  theory 
of  endotoxins  and  their  application  to  various  diseased 
conditions,  in  a  very  suggestive  manner,  but  added  little 
to  our  exact  knowledge.  Richet2  has  made  many  valuable 
contributions  on  this  subject.  In  his  first  report  made  with 
Portier  in  1902,  he  worked  with  an  extract  from  the  ten- 
tacles of  a  muscle  and  showed  that  an  injection  of  this  made 
the  animal  much  more  susceptible  to  a  second  one.  Unfor- 
tunately, he  coined  the  word  anaphylaxis  as  most  suitable 
to  cover  this  condition  of  increased  susceptibility.  He  used 
this  word  understanding  it  to  mean  "without  protection," 
and  indicating  that  the  first  injection  destroyed  any  natural 
resistance  that  the  animal  might  possess  against  the  poison. 
Now,  we  know  that  the  condition  of  sensitization  is  essential 
to  certain  forms  of  immunity,  as  was  first  indicated  by 
Vaughan  and  Wheeler,3  and  the  inappropriateness  of  the 
term  anaphylaxis  is  self-evident.  However,  the  word  has 
come  into  general  use,  and  with  this  explanation  we  will 
continue  it.  V.  Pirquet4  proposed  and  has  continued  the 
use  of  the  word  "allergic,"  meaning  altered  energy.  This 
is  much  more  suitable,  inasmuch  as  it  simply  expresses  a 
fact  and  binds  no  one  to  any  theory.  However,  "allergic" 
has  not  been  usually  employed,  and  we  will  use  "protein 
sensitization,"  "hypersensitiveness,"  "anaphylaxis,"  and 
"allergic"  as  synonyms. 

1  Zentralbl.  f.  Bakt.,   1904,  xxxvii;  Munch,  med.  Woch.,   1906;  Derm. 
Zentralbl.,  1906;  Berl.  klin.  Woch.,  1907. 

2  Compt.  rend,  de  la  Soc.  biol.,  1902;  Ann.  de  1'Institut  Pasteur,  1907,  xxi, 
497;  ibid.,  "1908,  xxiii;  ibid.,  1909. 

3  Jour.  Infect.  Dis.,  1907. 

4  Munch,  med.  Woch.,  1906. 


216  PROTEIN  POISONS 

The  fact  that  animals  which  have  once  received  an 
injection  of  protein  are  liable  to  sudden  death  after  a  second 
injection  of  the  same  kind  has  been  known  for  many  years. 
Ever  since  the  opening  of  the  Hygienic  Laboratory  of  the 
University  of  Michigan  (1888),  animals  once  used  have 
been  segregated  and  kept  in  cages  marked  "used  animals," 
which  indicated  that  conclusions  could  not  be  safely  drawn 
from  results  obtained  when  these  animals  were  employed  a 
second  time.  In  the  standardization  of  diphtheria  antitoxin 
it  soon  became  evident  that  the  guinea-pigs  that  survived 
one  test  could  not  be  relied  upon  in  a  second  one.  In  the 
late  nineties,  Parke,  Davis  &  Co.,  large  manufacturers  of 
antitoxins,  ascertained  this  fact  and  offered  to  supply  the 
Hygienic  Laboratory  of  the  University  of  Michigan  with 
"  used"  guinea-pigs  at  a  small  price.  The  offer  was  accepted, 
but  the  animals  were  found  dear  at  any  price,  as  they 
suddenly  and  unexplainably  died  when  treated  with  horse 
serum. 

This  condition  evidently  was  observed  by  others,  and 
Theobald  Smith  mentioned  it  to  Ehrlich,  who  set  Otto  to 
work  to  find  the  explanation.  Otto1  published  his  results 
under  the  title  "Das  Theobald  Smithsche  Phanomen  der 
Serumiiberempfindlichkeit."  However,  simultaneously  with 
these  observations  on  animals  used  in  the  standardization 
of  antitoxin,  the  profession  had  occasion  to  observe  the 
effects  of  injections  of  antitoxin  in  human  beings.  As 
early  as  1903,  v.  Pirquet2  wrote  concerning  certain  clinical 
effects  following  antitoxin  treatment,  and  in  1905  he  and 
Shick  published  a  monograph  "the  serum  disease,"  "Die 
Serumkrankheit. ' ' 

Definition. — Friedemann3  offers  the  following  definition: 
"We  speak  of  anaphylaxis  when  the  organism,  in  conse- 
quence of  a  previous  treatment  with  an  antigen,  after  a 
period  of  incubation  becomes  hypersensitive  to  the  same 
or  to  a  closely  related  substance,  and  when  this  condition 

1  V.  Leuthold  Gedenkschrift,  1906. 

a  Wien.  klin.  Woch. 

3  Jahresb.  u.  d.  Ergeb.  d.  Immunitatsforschung,  1910,  vi. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     217 

can  be  passively  transferred  to  fresh  animals  by  the  serum 
or  organ  extracts  of  the  sensitized  animal."  Biedl  and 
Kraus,1  omitting  passive  anaphylaxis,  give  the  following: 
"By  anaphylaxis  we  mean  that  state  of  specific  hypersen- 
sitiveness  induced  in  animals  by  protein  injections,  and  in 
which  symptoms  of  poisoning  follow  subsequent  injections 
of  the  same  protein  in  doses  which  would  have  no  effect 
upon  untreated  animals."  With  some  explanation  to  be 
given  later  we  accept  these  definitions  as  quite  satisfactory. 
In  the  meantime  it  is  desirable  to  have  a  clear  understanding 
of  the  meaning  of  the  terms  employed  in  discussing  this 
subject.  The  substance  which  induces  the  anaphylactic 
state  is  generally  known  as  the  "antigen."  This  implies 
that  it  gives  rise  to  the  production  of  an  antibody,  and  the 
selection  of  this  word  has  been  determined  by  an  attempt 
to  correlate  the  phenomena  of  anaphylaxis  with  the  theory 
evolved  by  Ehrlich  in  explanation  of  the  production  of 
antitoxins  by  treatment  with  toxins.  In  truth  the  "  antigen" 
of  anaphylaxis  is  not  a  toxin,  nor  is  the  new  substance 
generated  in  the  body  of  the  treated  animal  an  antitoxin. 
The  term  "  anaphylactogen"  is  unobjectionable,  since  it  is 
applicable  to  any  substance  which  induces  the  anaphylactic 
state.  Sensitizer  is  a  good  word,  and  commits  one  to  no 
theory.  The  same  is  true  of  the  term  "  sensibilisinogen" 
used  by  our  French  confreres.  The  sensitizer  causes  the 
body  cells  of  the  treated  animal  to  elaborate  a  specific 
proteolytic  ferment  which  digests  or  splits  up  the  sensitizer. 
Again,  following  the  nomenclature  of  Ehrlich,  this  ferment 
elaborated  as  a  consequence  of  the  introduction  of  the 
sensitizer  is  generally  designated  as  the  "antibody."  It 
would  be  equally  rational  to  speak  of  pepsin  as  an  antibody 
to  beefsteak,  because  the  former  digests  the  latter.  The 
theory  evolved  by  Ehrlich  in  his  studies  on  toxin  immunity 
is  the  product  of  a  genius  of  the  highest  order.  It  has 
stimulated  research,  which  has  resulted  in  discoveries  of 
the  greatest  importance,  but  the  attempt  to  explain  all 

1  Kraus  and  Levaditi's  Handbuch  d.  Technik  u.  Methodik  d.  Immunitats- 
forschung.  Erganzungsband. 


218  PROTEIN  POISONS 

physiological  and  pathological  processes  by  this  theory, 
and  to  describe  them  in  the  nomenclature  of  this  theory  is 
unscientific.  To  say  that  anaphylaxis  is  the  result  of 
protein — antiprotein  reaction — is  to  talk  jargon.  When 
foreign  proteins  are  taken  into  the  alimentary  canal  they 
must  be  digested  before  they  are  absorbed.  This  means 
that  their  large  molecules  must  be  split  into  smaller  ones, 
and  this  must  be  continued  until  there  are  no  more  protein 
molecules  left.  Every  protein  molecule  contains  a  poisonous 
group,  and  in  normal,  alimentary  digestion  this  group  is 
rendered  non-poisonous  by  further  cleavage  before  absorp- 
tion takes  place.  When  foreign  proteins  find  their  way  into 
the  blood  and  tissues  they  must  be  digested.  This  is  accom- 
plished, as  it  is  in  the  alimentary  canal,  by  proteolytic 
ferments,  but  the  danger  from  the  poisonous  group  in  the 
protein  molecule  is  evidently  greater  in  parenteral  than  in 
enteral  digestion.  Both  enteral  and  parenteral  digestion 
are  physiological  processes.  Every  living  cell  has  its  own 
proteolytic  ferments,  otherwise  it  could  not  live.  When 
stimulated  it  pours  out  this  ferment,  and  it  does  so  only 
when  stimulated.  The  function  of  a  cell  ferment  depends 
upon  the  kind  of  cell  elaborating  it,  and  to  a  certain  extent 
upon  the  stimulating  substance.  The  proteins  are  the 
normal  stimulants  to  cell  secretion.  When  a  foreign  pro- 
tein is  introduced  into  the  blood  or  tissue  it  stimulates 
certain  body  cells  to  elaborate  that  specific  ferment  which 
will  digest  that  specific  protein.  When  such  a  protein  first 
comes  in  contact  with  the  body  cells  the  latter  are  unpre- 
pared to  digest  the  former,  but  this  function  is  gradually 
acquired.  The  protein  contained  in  the  first  injection  is 
slowly  digested,  and  no  ill  effects  are  observable.  When 
subsequent  injections  of  the  same  protein  are  made,  the 
cells,  prepared  by  the  first  injection,  pour  out  the  specific 
ferment  more  promptly  and  the  effects  are  determined  by 
the  rapidity  with  which  the  digestion  takes  place.  The 
poisonous  group  in  the  protein  molecule  may  be  set  free 
so  rapidly  and  in  amount  sufficient  to  kill  the  animal.  This 
in  brief  is  an  explanation  of  the  phenomena  of  anaphylaxis. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      219 

The  Sensitizer. — The  sensitizing  agent  most  thoroughly 
studied  is  blood  serum.  When  a  small  dose  of  blood  serum 
is  injected  into  a  guinea-pig  intravenously,  subcutaneously, 
intracranially,  or  intra-abdominally,  and,  after  a  period 
of  twelve  days  or  longer  has  elapsed,  a  second  injection  is 
made,  the  animal  develops  the  symptoms  of  anaphylactic 
shock,  which,  in  the  majority  of  instances,  terminate 
fatally.  This  reaction  is  specific.  The  animal  is  sensitized 
to  the  blood  serum  of  the  species  of'  animal  from  which  the 
blood  was  taken  and  not  to  the  sera  of  other  species.  The 
amount  of  serum  necessary  to  sensitize  a  guinea-pig  is 
surprisingly  small.  Rosenau  and  Anderson  found  0.000001 
c.c.  of  horse  serum  sufficient.  Besredka  places  the  minimum 
amount  necessary  to  secure  uniform  results  at  0.001  c.c. 
while  one-tenth  of  this  proved  sufficient  in  a  considerable 
percentage  of  the  animals.  The  sensitizing  dose  of  horse 
serum  ordinarily  employed  in  experiments  upon  guinea-pigs 
is  0.01  c.c.  Large  doses  sensitize,  but  a  longer  time  is 
required.  When  5  c.c.  is  given  the  time  which  elapses 
before  complete  sensitization  results  may  be  as  long  as 
three  months.  The  larger  the  dose  the  longer  the  time 
essential  to  sensitization.  Besredka  is  inclined  to  the 
opinion  that  when  large  doses  are  given  there  is  no  sensitiza- 
tion until  the  greater  part  of  the  injected  protein  is  elimin- 
ated. If  he  means  that  it  is  eliminated  unchanged,  he  is 
certainly  wrong.  The  protein  of  the  first  injection  is  slowly 
digested,  and  the  larger  the  amount  the  longer  the  time 
required  for  the  digestion,  and  complete  sensitization  does 
not  occur  until  all  the  protein  of  the  first  injection  has  been 
disposed  of  and  the  cells  have  had  time  to  accumulate  a 
reserve  of  the  preferment.  At  least  this  is  our  explanation 
of  this  point.  The  second  dose,  in  order  to  produce  a  fatal 
result,  must  be  considerably  larger  than  the  minimum 
sensitizing  dose.  The  proportion  between  the  minimum 
sensitizing  and  minimum  fatal  dose  has  been  placed  by 
Doerr  and  Russ  at  1  to  1000.  The  second  dose,  in  order 
to  kill  the  animal  promptly,  must  contain  at  least  a  fatal 
dose  of  the  protein  poison,  but  it  may  contain  many  times 


220  PROTEIN  POISONS 

this  amount  and  not  kill.  Whether  the  second  dose  kills 
or  not  depends  not  only  upon  the  amount  of  poison  it 
contains  but  upon  the  rapidity  with  which  the  poison  is 
set  free. 

There  has  been  some  difference  of  statement  concerning 
the  effect  of  heat  on  the  sensitizing  properties  of  blood 
serum.  Rosenau  and  Anderson1  found  that  animals  could 
not  be  sensitized  with  serum  which  had  been  heated  at  100°. 
Doerr  and  Russ2  placed  the  point  at  which  loss  of  sensitizing 
properties  occurs  at  80°.  Kraus  and  Volk3  raised  it  to  90°. 
Besredka  has  straightened  out  this  matter  and  has  correctly 
shown  that  the  sensitizing  properties  of  a  protein  are  in  part 
at  least  dependent  upon  its  physical  state,  and  that  diluted 
serum  may  be  heated  even  to  120°  without  losing  its 
capability  of  sensitizing.  It  is  probable  that  no  protein 
completely  sensitizes  the  body  cells  unless  it  be  in  at  least 
partial  solution.  Heating  undiluted  blood  coagulates  the 
protein,  and  in  this  way  leads  to  a  decrease  of  its  stimu- 
lating effects  upon  the  body  cells.  Besredka  has  shown  that 
the  sensitizing  property  of  blood  serum  is  thermostabile. 
Wells4  has  very  properly  pointed  out  that  it  is  the  physical 
change  induced  in  the  protein  by  coagulation  and  not 
chemical  alteration,  which  decreases  its  efficiency  as  a 
sensitizer,  and  he  calls  attention  to  the  fact  first  shown 
by  Besredka  that  proteins  not  coagulated  by  heat,  do  not 
decrease  in  their  sensitizing  effects  when  their  solutions  are 
boiled.  This  is  true  of  casein,  for  instance,  but  when  milk 
sours  and  coagulation  of  the  casein  results  it  is  not  so  ready 
a  sensitizer.  Wells,  furthermore,  shows  that  other  methods 
of  coagulation,  as  by  precipitation  with  alcohol,  lessen  the 
sensitizing  properties.  He  suggests  that  the  finely  coagu- 
lated particles  of  protein  may  be  seized  upon  by  phagocytes 
and  destroyed.  In  confirmation  of  this  we  have  found  that 
proteins  insoluble  in  water,  such  as  edestin,  sensitize  more 
efficiently  when  dissolved  in  salt  solution  than  when  sus- 

1  Hygienic  Laboratory,  Bulletin  No.  45. 

2  Zeitsch.  f.  Immunitatsforschung,  i,  110. 

3  Ibid.,  731.  "  Jour.  Infect.  Dis.,  v. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     221 

pended  in  water.  Furthermore,  we  have  found  that  bacterial 
proteins  suspended  in  normal  salt  solution  and  heated  to 
154°  in  the  autoclave  under  2  kilos  of  pressure  are  more 
efficient  sensitizers  than  the  unheated  suspensions.  All 
these  facts  support  the  theory  that  body  cells  are  best 
sensitized  when  the  protein  comes  in  intimate  contact  with 
them.  Possibly  cell  permeation  is  necessary  for  the  most 
complete  sensitization. 

Besredka  finds  that  when  the  protein  of  the  second 
injection  is  heated  it  is  less  likely  to  kill,  and  he  concludes 
that  proteins  contain  a  thermostabile,  sensitizing,  and  a 
thermolabile  toxic  component.  We  fail  to  see  how  such  a 
conclusion  follows  his  findings.  If  the  physical  condition 
of  a  protein  affects  its  sensitizing  properties,  why  should 
it  not  affect  its  toxic  action  on  reinjection?  The  poison  is 
set  free  by  the  digestive  action  of  the  specific  ferment 
elaborated  as  a  result  of  the  first  injection.  Why  should 
not  the  physical  state  of  the  protein  affect  the  rapidity 
and  thoroughness  with  which  it  is  digested,  and  consequently 
the  amount  of  the  protein  poison  set  free  or  activated  at 
one  time?  Doerr  and  Russ  have  apparently  answered  this 
question  in  a  satisfactory  manner.  By  carefully  conducted 
experiments  they  show  that  heat  affects  the  sensitizing 
and  toxic  properties  of  proteins  in  the  same  ratio. 

It  should  be  understood  that  temperatures  high  enough 
to  disrupt  and  destroy  proteins  are  destructive  to  their 
sensitizing  properties.  According  to  Rosenau  and  Anderson 
a  temperature  of  200°  removes  every  trace  of  the  sensitizing 
property  of  proteins. 

The  influence  of  the  digestive  ferments  of  the  alimentary 
canal  on  the  sensitizing  properties  of  proteins  is  an  interesting 
and  important  subject,  since  it  bears  upon  the  possibility 
of  sensitization  by  administration  through  the  digestive 
tract.  This  point  has  been  especially  studied  by  Wells1 
and  Pick  and  Yamanouchi.2  The  former  submitted  egg 


1  Loc.  cit. 

2  Zeitsch.  f.  Immunitatsforschung,  i,  676;  Wien.  klin.  Woch.,  1908,  1513. 


222  PROTEIN  POISONS 

albumen  to  tryptic  digestion  and  found  that  as  the  digestive 
action  advanced  the  sensitizing  property  receded.  Some 
have  claimed  to  sensitize  animals  with  peptone  and  even 
with  amino-acids,  but  since  the  minutest  quantity  of 
protein  suffices  to  sensitize,  it  is  more  reasonable  to  suppose 
that  the  peptone  and  amino-acid  preparations  were  not 
absolutely  free  from  protein.  Vaughan  and  Wheeler  have 
shown  that  the  poisonous  portion  of  the  protein  molecule 
does  not  sensitize  in  either  small  or  large  doses.  Frances- 
chelli1  found  that  when  tissue  is  autolyzed  for  months,  and 
until  every  trace  of  the  biuret  reaction  is  lost,  the  fluid 
shows  no  diminution  in  its  sensitizing  properties.  This 
agrees  with  the  finding  of  Vaughan  and  Wheeler,  that  their 
non-poisonous  portion  of  the  protein  molecule  sensitizes 
even  when  it  does  not  respond  to  the  biuret  reaction.  All 
this  suggests  that  the  sensitizing  group  in  the  protein 
molecule  is  not  itself  a  protein,  or  at  least  not  a  biuret,  body. 
However,  the  sensitizing  group  is  destroyed  in  normal 
digestion,  and  it  is  only  under  abnormal  conditions  that 
protein  sensitization  results  through  the  alimentary  canal. 
We  will  return  to  this  subject  later. 

Whether  or  not  all  proteins  contain  the  sensitizing  group 
cannot  as  yet  be  answered  with  certainty.  According  to 
Doerr  and  Russ  the  globulin  of  the  blood  serum  is  the  only 
protein  in  that  fluid  which  sensitizes,  while  Wells  concludes 
that  in  egg-white  the  albumen  is  the  only  active  agent. 
Wells  purified  the  albumen  of  egg-white  by  recrystallization 
after  the  method  of  Hopkins  and  Pinkus,  and  he  found 
that  the  purer  his  albumen,  the  smaller  the  amount  neces- 
sary to  sensitize.  Gay  and  Adler2  attempted  by  fractional 
precipitation  of  blood  serum  with  ammonium  sulphate  to 
separate  the  anaphylactogenic  from  the  other  protein 
constituents,  and  they  obtained  an  euglobulin  which  sensi- 
tizes but  does  not  prove  toxic  on  the  second  injection. 
Quite  naturally  it  seemed  to  them  that  they  had  succeeded 
in  isolating  the  sensitizing  constituent  of  blood  serum,  and 

i  Archiv  f.  Hygiene,  1909,  Ixx,  163.  2  Jour.  Med.  Research,  xviii,  433. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     223 

they  proposed  to  call  it  "anaphylactin."  But  ammonium 
sulphate  alters  the  chemical  nature  of  proteins,  and  Armit1 
has  shown  that  after  precipitation  with  this  reagent  the 
poisonous  group  in  it  cannot  be  extracted  by  the  method 
of  Vaughan  and  Wheeler.  It  is  probable  therefore  that 
the  "anaphylactin"  of  Gay  and  Adler  contains  the  poisonous 
group,  but  so  combined  that  it  is  not  set  free  either  in  vitro 
or  in  vivo. 

Besredka  has  pointed  out  that  in  anaphylactic  experiments 
with  milk  the  fluid  should  be  boiled  for  about  twenty 
minutes,  the  injections  should  be  made  into  the  peritoneal 
cavity,  larger  guinea-pigs  of  from  300  to  400  grams'  weight 
should  be  used,  and  a  period  of  from  sixteen  to  twenty 
days  allowed  to  elapse  between  the  sensitizing  and  test 
injections.  When  these  conditions  are  complied  with,  an 
exquisite  sensitization  with  uniform  results  is  secured. 
Besredka  was  not  able  to  sensitize  guinea-pigs  with  milk 
given  by  mouth  or  rectum.  We  have  found  rabbits,  espe- 
cially young  ones,  easily  sensitized  by  either  of  these  avenues. 
In  our  work  we  have  observed  that  hungry  rabbits  will 
eat  milk  when  mixed  with  other  food  and  seldom  are  sensi- 
tized, but  when  the  milk  is  introduced  into  the  stomach  of 
a  fasting  rabbit  through  a  tube,  or  injected  into  the  rectum, 
the  milk  can  soon  be  detected  in  the  heart's  blood,  and  the 
animal  becomes  sensitized.  Evidently  when  the  milk  is 
taken  normally  into  the  stomach,  it  is  digested;  when 
forcibly  fed  through  a  tube,  it  is  in  part,  at  least,  absorbed 
undigested.  Like  the  blood,  milk  as  between  different 
species  of  animals  shows  a  strictly  specific  action.  Animals 
sensitized  to  woman's  milk  do  not  react  when  treated  with 
cows'  milk,  and  vice  versa.  By  this  method  we  have  iden- 
tified the  source  of  milk  stains  deposited  on  wood  for 
months.  The  evidence  concerning  the  difference  between 
the  proteins  of  the  milk  and  those  of  the  blood  of  the  same 
animal  is  somewhat  conflicting.  Besredka2  found  that 


1  Zeitsch.  f.  Immunitatsforschung,  1910,  v.  703. 

2  Compt.  rend.  Soc.  biol.,  Ixiv,  888;  Ann.  de  1'  Institut  Pasteur,  1909,  xxiii. 


224  PROTEIN  POISONS 

animals  sensitized  to  cows'  serum  were  not  affected  on  the 
subsequent  injection  of  cows'  milk,  and  vice  versa.  Wells1 
obtained  results  that  were  not  constant.  Uhlenhuth  and 
Handel2  and  Thomsen3  did  sensitize  to  serum  with  milk  and 
vice  versa;  the  latter  used  woman's  milk  and  human  serum. 
Bauer4  by  fixation  of  the  complement  method  seems  to  have 
shown  that  the  albumin  and  globulin  of  milk  are  closely 
related  to  the  same  constituents  of  the  blood,  while  the 
casein  of  the  milk  is  a  protein  unlike  any  in  the  blood. 
This  is  probably  correct  and  explains  the  inconstancy  in 
the  experiments. 

The  differences  between  the  proteins  of  the  blood  serum 
and  those  of  the  erythrocytes  have  been  demonstrated 
by  the  anaphylactic  reaction.  This  has  been  shown  uni- 
formly by  the  experiments  of  H.  Pfeiffer,5  Pfeiffer  and 
Mita,6  Thomsen,7  Doerr  and  Moldovan,8  and  Uhlenhuth 
and  Handel.9  These  investigators  have  found  it  impossible 
to  sensitize  guinea-pigs  against  blood  serum  with  erythro- 
cytes and  vice  versa.  In  demonstrating  this  fact  it  is  necessary 
to  fully  separate  the  corpuscles  and  serum.  The  corpuscles 
should  be  well  washed  in  order  to  accomplish  this,  and 
provision  must  be  made  against  solution  of  the  corpuscles 
in  the  serum.  The  corpuscles  of  each  species  contain 
specific  proteins  and  therefore  those  of  one  species  do  not 
sensitize  to  those  of  another. 

According  to  Dunbar,10  the  sexual  cells  are  as  specific  as 
blood  sera.  The  proteins  of  organ  extracts  are  specific  as 
between  different  species  of  animals,  with  some  exceptions 
to  be  noted  later,  but  as  between  different  organs  from  the 
same  species,  and  between  the  blood  serum  and  organ 


Loc.  cit. 

Zeitsch.  f.  Immunitatsforschung,  iv,  761.  3  Ibid.,  iii,  539. 

Munch,  med.  Woch.,  1908,  No.  16;  Zeitschr.  f.  exper.  Path.  u.  Ther., 
1909,  vii. 

Zeitsch.  f.  Immunitatsforschung,  1910,  viii. 

Ibid.,  vi;  ibid.,   v.  '  Ibid.,  i;  ibid.,  iii. 

>>  Zeitsch.  f.  Bakt.  Ref.,  v. 
»  Zeitsch.  f.  Immunitatsforschung,  iv,  761. 
10  Ibid.,  740;  vii,  454. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     225 

extracts  they  are  not  strictly  specific.  It  will  be  readily 
understood  that  it  is  difficult  to  obtain  organ  extracts 
wholly  free  from  the  blood  of  the  same  animal. 

That  the  crystalline  lens  contains  proteins  different  from 
those  found  in  any  other  part  of  the  body  was  demonstrated 
some  years  ago  by  Uhlenhuth,  by  the  precipitin  reaction, 
and  it  was  believed  that  the  proteins  of  the  crystalline  lens 
are  identical  in  all  animals.  This  was  apparently  confirmed 
by  anaphylactic  tests,  as  shown  by  the  work  of  Andrejew1 
and  that  of  Kraus  and  Sohma.2  Guinea-pigs  can  be  sensi- 
tized with  the  proteins  of  their  own  lenses,  or  with  those  of 
other  animals,  and  sensitization  with  the  proteins  of  the 
lens  from  any  animal  responds  to  the  same  from  any  animal. 
The  question  whether  the  proteins  of  the  crystalline  lens 
are  identical  in  all  animals  is  of  the  greatest  biological 
interest.  The  precipitin  and  the  first  sensitizing  tests 
seemed  to  establish  this  belief,  but  quantitative  experi- 
ments, such  as  those  of  Kapsenberg,3  indicate  that  there  are 
slight  differences  in  the  proteins  of  the  crystalline  lens  from 
different  species.  Kapsenberg,  after  reviewing  the  literature 
and  detailing  his  own  work,  concludes:  (1)  Guinea-pigs  are 
easily  and  uniformly  sensitized  with  the  lens  substance  of 
other  animals.  The  dose  on  reinjection  necessary  to  induce 
fatal  anaphylactic  shock  is  small.4  (2)  Guinea-pigs  may  be 
sensitized  to  the  protein  of  their  own  lenses,  but  this  is 
done  with  difficulty  and  the  fatal  dose  on  reinjection  is 
large  (40  mg.).  (3)  The  proteins  of  the  crystalline  lens  are 
specific,  but  not  markedly  so.  Dunbar  finds  that  the 
specificity  of  fish  proteins  is  not  so  marked  as  those  from 
mammals. 

Rosenau  and  Anderson5  succeeded  in  sensitizing  guinea- 
pigs  to  placental  tissue  from  the  same  animal,  and  Locke- 
mann  and  Thies6  sensitized  rabbits  to  the  serum  of  the 

1  Arb.  aus  d.  Kaiserl.  Gesundheitsamte,  xxx,  450. 

2  Wien.  klin.  Woch.,  1908,  1084. 

3  Zeitsch.  f.  Immunitatsforschung,  1912,  xv,  518. 

4  In  one  instance  as  low  as  2.5  mg.  of  protein  but  usually  6  mg. 

5  Hygienic  Laboratory  Bulletin,  No.  45. 
e  Biochem.  Zeitsch.,  xxv. 

15 


226  PROTEIN  POISONS 

rabbit  fetus.  Gozony  and  Wiesinger1  passively  sensitized 
rabbits  with  the  blood  serum  to  amniotic  fluid  in  two  cases 
of  eclampsia. 

Some  years  ago  Obermayer  and  Pick2  found  that  the 
serum  of  rabbits  treated  with  proteins  radically  changed 
by  being  iodized,  nitrified,  or  diazoized,  did  not  precipitate 
the  native  protein,  but  did  act  upon  the  altered  protein 
with  which  the  animal  had  been  treated,  and  this  occurred 
without  reference  to  the  original  sources  of  the  protein. 
Wells3  and  Pick  and  Yamanouchi4  were  not  able  to  sensitize 
animals  with  iodized  protein  to  an  iodized  protein  obtained 
from  another  species. 

Sensitization  to  egg-white  has  been  studied  by  Vaughan 
and  Wheeler5  also  by  Wells.6  The  former  used  in  most  of 
their  experiments  egg-white  diluted  with  an  equal  volume 
of  salt  solution.  Guinea-pigs  sensitized  to  egg-white  from 
chickens  responded  to  test  injections  of  egg-white  from 
tame  ducks,  though  less  energetically  and  less  constantly, 
and  still  less  to  egg-white  from  robbins.  Vaughan  and 
Wheeler  by  the  method  already  described  (p.  98)  split 
egg-white  into  a  non-poisonous,  sensitizing  portion  and  a 
poisonous,  non-sensitizing  portion.  They  believe  that  a 
similar  cleavage  occurs  as  the  result  of  ferment  action  on 
the  second  injection  in  sensitized  animals.  This  work 
forms  the  basis  of  their  theory  of  anaphylaxis,  which  will 
be  discussed  later. 

All  bacterial  proteins  are  anaphylactogens.  Indeed,  the 
Koch  reaction  with  tuberculin  is  an  anaphylactic  test,  but 
this  will  be  discussed  later.  Bacterial  proteins  act  as  ana- 
phylactogens, whether  living  or  dead,  formed  or  in  solution. 
On  account  of  the  physical  state  of  the  protein  the  reaction 
is  generally  less  pronounced,  and  strong  than  with  proteins 
in  solution.  Bacterial  anaphylaxis  has  been  studied  by 

»  Orvosi  hetilap,  liii,  418.  2  Wien.  klin.  Woch.,  1906. 

3  Loc.  cit. 

4  Zeitsch.  f.  Immunitatsforschung,  i,  676. 
6  Jour.  Infect.  Dis.,  June,  1907. 

6  Loc.  cit. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     227 

Rosenau  and  Anderson,1  Kraus  and  his  students,2  and 
others.  The  first  mentioned  sensitized  animals  with  sub- 
tilis,  colon,  typhoid,  anthrax,  and  tubercle  bacilli.  The 
second  dose  given  after  eleven  days  or  longer  induced 
anaphylactic  symptoms.  In  some  instances  repeated 
injections  seem  to  be  necessary  in  order  to  induce  a  high 
degree  of  sensitization.  The  evidence  concerning  the 
specificity  of  bacterial  anaphylaxis  is  somewhat  conflicting. 
Kraus  and  Doerr  sensitized  guinea-pigs  with  intraperitoneal 
injections  of  one  loop  or  less  of  cultures  of  typhus,  dysentery, 
cholera,  v.  Nasik  and  v.  El-Tor.  A  second  injection  of  a 
maceration  of  homologous  cultures  given  intravenously 
after  from  twenty  to  twenty-five  days  was  followed  by 
marked  dyspnea,  discharge  of  urine  and  feces,  and  coma. 
Some  recovered,  but  others  died  within  ten  minutes.  These 
investigators  found  this  reaction  strictly  specific.  In 
another  experiment  they  found  that  animals  sensitized  with 
a  maceration  of  the  dysentery  bacillus  did  not  respond  to 
the  toxin  of  this  bacillus,  but  did  to  a  second  treatment 
with  the  maceration.  There  is  no  proof  that  toxins  sensi- 
tize. Delanoe  sensitized  guinea-pigs  to  the  typhoid  bacillus. 
He  secured  the  most  marked  effects  when  sensitization  was 
induced  by  repeated  injections,  and  one  month  or  longer 
elapsed  before  the  test  injection.  He  did  not  find  the 
reaction  markedly  specific.  Vaughan  and  Wheeler  sensi- 
tized guinea-pigs  to  colon,  typhoid,  and  tubercle  proteins, 
and  in  this  way  secured  a  certain  degree  of  immunity  to 
living  cultures.  They  also  sensitized  animals  with  the 
non-poisonous  proteins  of  the  colon  and  typhoid  bacilli,  and 
secured  the  same  degree  of  immunity  to  living  cultures. 
This  subject  will  be  enlarged  when  we  discuss  the  relation 
of  anaphylaxis  to  the  infectious  diseases. 

The  purest  known  proteins  act  as  sensitizers.  Even  the 
crystallized  proteins  as  hemoglobin,  crystalline  egg-white, 
and  such  pure  vegetable  proteins  as  edestin  and  excelsin 


1  Loc.  cit. 

2  Wien.  klin.  Woch.,  1908,  Nos.  18,  28,  and  30. 


228  PROTEIN  POISONS 

act  as  exquisite  sensitizers.  Moreover,  it  has  been  found 
that  the  more  thoroughly  a  protein  is  purified  the  more 
perfectly  it  sensitizes  and  the  smaller  the  dose  necessary 
to  sensitize  or  to  kill  on  reinjection.  Wells  found  that 
purified  casein  acts  more  perfectly  and  in  smaller  doses 
than  a  corresponding  quantity  of  milk,  and  the  sensitizing 
dose  of  crystallized  egg-white  is  less  than  one  one-hundredth 
that  of  native  egg-white,  and  the  killing  dose  on  reinjec- 
tion less  than  one-fifth.  These  facts  have  led  Wells  to 
suggest  that  the  mixed  albumins  may  contain  substances 
which  antagonize  the  anaphylactic  reactions.  Since  pure 
proteins  sensitize  and  kill  on  reinjection,  it  seems  reasonable 
to  conclude  that  the  sensitizing  and  poisonous  groups  are 
constituents  of  the  same  molecule.  Edestin  in  its  most 
highly  purified  form  is  believed  to  be  a  chemical  unit,  and 
not  a  mixture  of  proteins.  This  can  be  split  by  our  method 
into  sensitizing  and  poisonous  portions.  It  is  true  that  the 
amount  of  the  non-poisonous  portion  necessary  to  sensitize 
is  larger  than  that  of  the  unbroken  molecule  necessary  to 
accomplish  the  same  purpose,  and  it  is  possible  that  sensi- 
tization  with  this  product  is  due  to  the  fact  that  it  contains 
a  trace  of  the  unbroken  molecule,  but  the  fact  that  no 
amount  of  this  portion  induces  the  slightest  anaphylactic 
symptoms  on  reinjection  is  not  in  harmony  with  this  view. 
It  seems  more  reasonable  to  assume  that  in  the  process  of 
cleavage,  which  is  crude,  a  large  part  of  the  sensitizing 
group  is  destroyed.  It  is  certain  that  the  poisonous  portion 
does  not  sensitize  to  either  itself  or  the  unbroken  molecule. 
By  our  method  the  molecule  is  disrupted  and  in  so  doing 
both  portions  are  largely  destroyed.  The  final  word  on 
this  matter  cannot  be  spoken  until  we  know  that  we  have 
absolutely  pure  proteins  to  start  with,  and  we  have  more 
perfect  methods  for  the  cleavage  of  the  protein  molecule. 
However,  it  seems  certain  that  the  sensitizing  properties 
of  the  protein  molecule  reside  in  a  group  or  in  groups  which 
are  destroyed  by  digestion  long  before  the  poisonous  group 
is  markedly  impaired.  The  sensitizing  group  seems  more 
labile  than  the  poisonous  one. 


PROTEIN  SEN  SIT  IZ  AT  ION  OR  ANAPHYLAXIS     229 

Through  the  courtesy  of  White  and  Avery  we  have  been 
permitted  to  read  an  unpublished  research  of  theirs  on 
"Some  Immunity  Reactions  of  Edestin."  From  this  we 
excerpt  the  following  findings:  (1)  The  smallest  sensitizing 
dose  of  pure  crystallized  edestin  given  intraperitoneally 
is  0.0001  mg.  Guinea-pigs  sensitized  with  this  dose  react 
fatally  when  the  reinjection  intravenously  is  not  less  than 
50  mg.  When  the  sensitizing  dose  is  from  0.1  to  5  mg., 
0.5  mg.  causes  a  fatal  dose  on  intravenous  reinjection. 
(2)  Guinea-pigs  sensitized  to  edestin  do  not  react  on  intra- 
venous injection  of  gliadin,  or  globulins  from  squash  seed, 
castor  bean,  and  the  hazel-nut.  Two  animals  reacted, 
one  fatally,  to  intravenous  injections  of  flaxseed  globulin. 
The  fatal  dose  of  flaxseed  globulin  was,  however,  from 
forty  to  one  hundred  and  twenty  times  the  minimum  fatal 
intoxicating  dose  of  edestin.  (3)  Guinea-pigs  from  a  sensi- 
tized mother  inherit  sensitization,  though  in  a  lessened 
degree.  (4)  The  intraperitoneal  injection  of  from  0.05  to 
0.1  c.c.  of  edestin  immune  serum  into  a  guinea-pig  sensitizes 
the  latter  to  such  an  extent  that  it  reacts  fatally  to  an 
intravenous  injection  of  edestin  on  the  following  day.  (5) 
"When  edestin  is  hydrolyzed  by  an  alcoholic  solution  of 
sodium  hydrate  by  the  method  of  Vaughan,  a  substance 
is  formed  which  produces  a  fatal  intoxication  in  the  guinea- 
pig  apparently  identical  with  true  anaphylactic  shock.  The 
intravenous  injection  of  one  part  of  this  poison  to  forty 
thousand  parts  of  guinea-pig  by  weight  constitutes  the 
minimum  fatal  dose."  It  should  be  stated  that  this  is  the 
crude  poison.  (6)  "When  suitable  amounts  of  edestin  and 
edestin-immune  serum  are  allowed  to  remain  in  contact 
for  a  given  length  of  time,  a  precipitate  is  formed  which, 
when  washed  with  salt  solution  and  mixed  with  fresh 
guinea-pig  complement  and  incubated  at  body  temperature, 
yields  a  substance  or  substances  which  when  injected  into 
a  guinea-pig  intravenously  produces  a  fatal  intoxication, 
apparently  identical  in  every  way  with  the  anaphylactic 
reaction.  Fresh  complement,  when  allowed  to  act  under 
similar  conditions  with  edestin  alone,  yields  no  poisonous 


230  PROTEIN  POISONS 

substance.  From  edestin,  therefore,  by  the  action  of 
immune  serum  and  complement,  under  the  experimental 
conditions  noted,  a  toxic  product  is  obtained  which  seems 
to  correspond  with  the  anaphylatoxin  of  Friedberger." 

It  seems  most  probable  that  anaphylactogens,  agglu- 
tinogens,  precipitinogens,  and  lysinogens  are  identical. 
In  other  words,  one  group  in  the  protein  molecule  causes 
the  animal  cells  to  develop  a  substance  which  under  certain 
conditions  may  act  as  an  agglutinin,  a  precipitin,  or  a 
lysin.  We  are  inclined  to  the  belief — not  yet  positively 
demonstrated — that  the  same  ferment  may  under  varied 
conditions  act  as  an  agglutinin,  a  precipitin,  a  lysin,  or  it 
may  cause  a  deeper  cleavage  in  the  protein  molecule, 
resulting  in  the  liberation  of  the  protein  poison.  Through 
the  researches  of  Friedberger,  Doerr  and  Russ,  and  others, 
it  has  been  made  quite  certain  that  anaphylactogens  and 
precipitinogens  are  identical,  and  that  these  properties 
reside  in  the  same  intramolecular  group.  As  proteins  are 
altered  by  heat  or  digestion,  their  properties  as  anaphylac- 
togens and  precipitinogens  are  decreased  in  the  same 
ratio.  The  protein  obtained  by  one-third  to  one-half 
saturation  of  serum  with  ammonia  sulphate  is  strongly 
active  both  as  a  precipitinogen  and  as  an  anaphylactogen, 
while  that  obtained  by  full  saturation  is  inactive  in  either 
direction.  Whether  this  is  due  to  physical  or  chemical 
alteration  has  not  been  determined. 

We  may  condense  our  statements  concerning  anaphylac- 
togens as  follows :  They  are  proteins  which  when  introduced 
parenterally  into  animals  stimulate  the  body  cells  to  elaborate 
specific  ferments  for  the  purpose  of  their  digestion.  When 
introduced  into  a  sensitized  animal  they  are  digested  so 
rapidly  that  the  split  products,  some  of  which  are  poisonous, 
produce  certain  more  or  less  violent  and  characteristic 
symptoms  which  may  terminate  in  death.  All  anaphylac- 
togens are  proteins,  and  all  proteins  contain  a  certain 
poisonous  intramolecular  group.  This  group  is  physio- 
logically the  same  in  all  proteins,  hence  the  identity  of 
the  symptoms  of  anaphylactic  shock  whatever  the  protein 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     231 

by  which  it  is  induced.  All  anaphylactogens  contain  a 
sensitizing  intramolecular  group  which  is  not  the  same  in 
any  two  kinds  of  proteins,  hence  the  specificity  of  sensiti- 
zation.  We  have  succeeded  in  splitting  some  proteins  into 
non-poisonous,  sensitizing,  and  into  poisonous,  non-sensi- 
tizing portions.  Whether  all  proteins  contain  a  sensitizing 
group  or  not  has  not  been  determined.  Our  views  con- 
cerning anaphylactogens  differ  from  those  held  by  others. 
They  think  that  in  mixed  proteins,  such  as  blood-serum, 
corpuscles,  organ  cells,  egg-white,  etc.,  there  is  some  one 
protein  which  sensitizes  and  some  other  one  which  is  toxic. 
We  hold  that  the  sensitizing  and  toxic  proteins  are  groups 
in  the  same  molecule.  We  think  that  we  have  demon- 
strated this  by  obtaining  both  groups  from  such  pure  proteins 
as  edestin.  Artificially  crystallized  proteins,  such  as  egg 
albumen  prepared  by  the  method  of  Hopkins  and  Pinkus, 
are  not  suitable  for  this  work  because  they  are  changed 
chemically  by  the  ammonium  sulphate,  and  are  not  split 
up  by  our  method.  From  our  researches  we  conclude  that 
the  sensitizing  group  of  the  protein  molecule  is  much  more 
complicated  in  its  chemical  structure  than  the  toxic  group. 
Further  discussion  along  this  line  will  be  indulged  in  when 
we  take  up  the  poisonous  portion. 

We  are  aware  of  the  claims  made  by  Bogomoletz1  and 
by  Pick  and  Samanouchi,2  that  lipoids  may  act  as  anaphyl- 
actogens, but  they  have  not  convinced  us  that  their  prepara- 
tions were  wholly  free  from  proteins.3  Besides,  it  is  possible 
that  a  non-protein  may  act  indirectly  as  an  anaphylactogen. 
This  may  be  due  to  the  substance  causing  some  cleavage 
in  the  proteins  of  the  body  and  these  products  may  sensitize. 
This  question  will  arise  again  when  we  discuss  hypersensi- 
tiveness  to  certain  medical  agents. 

Volatile  Sensitizers. — Rosenau  and  Amos4  have  demon- 
strated that  the  exhaled  air  contains  a  substance  which 

1  Zeitsch.  f.  Immunitatsforschung,  v  and  vi.  2  Ibid.,  i,  676. 

s  Thiele  and  Embleton,  Zeitschr.  f.  Immunitatsforschung,  1913,  xvi,  160, 
have  investigated  the  claim  of  Bogomoletz  that  lipoids  act  as  anaphylacto- 
gens and  have  been  unable  to  confirm  his  work. 

4  Jour.  Med.  Research,  1911,  xxv,  35. 


232  PROTEIN  POISONS 

sensitizes  animals  to  the  blood  serum.  The  exhaled  breath 
of  men  condensed  and  injected  into  guinea-pigs  sensitizes 
these  animals  to  subsequent  injections  of  man's  serum.  Of 
99  guinea-pigs  submitted  to  this  test,  26  manifested  recog- 
nizable symptoms  of  anaphylactic  shock,  and  4  of  these 
died  on  injection  of  human  serum.  "  The  fact  that  a  number 
of  our  experiments  resulted  negatively  may  mean  either 
that  the  organic  matter  is  present  in  the  expired  air  in 
exceedingly  small  amounts,  or  that  the  guinea-pigs  with 
which  we  worked  did  not  come  from  a  very  sensitive  race. 
There  are  indications  in  our  work  which  suggest  that  the 
expired  breath  from  certain  persons  contains  more  organic 
matter  than  from  other  persons;  also  that  the  amount 
varies  with  conditions.  We  obtained  a  greater  percentage 
of  reactions  in  the  guinea-pigs  injected  with  the  liquid 
condensed  from  the  expired  breath  of  females  than  in  those 
injected  with  the  liquid  condensed  from  the  expired  breath 
of  males.  Whether  this  is  a  mere  coincidence  or  not  may 
be  determined  only  by  collecting  more  extensive  data. 

"  The  logical  conclusion  from  our  results  is  that  protein 
substances  under  certain  circumstances  may  be  volatile. 
It  seems  unlikely  that  such  a  complex  molecule  should 
possess  the  power  of  passing  into  the  air  in  a  gaseous  form. 
The  volatility,  however,  now  in  question,  may  resemble 
that  solubility  which  deals  with  particles  in  suspension  hi 
a  physicochemical  state  (colloidal  suspension).  The  protein 
may  simply  be  carried  over  in  'solution'  in  the  water  vapor. 

"A  comparatively  large  number  of  the  guinea-pigs 
inoculated  subcutaneously  with  the  condensed  liquid  from 
the  expired  breath  developed  sloughs  at  the  site  of  the 
injection.  It  is  not  certain  whether  this  was  due  to  the 
pressure  of  the  relatively  large  amount  of  liquid  injected, 
or  to  some  irritating  principle  contained  in  the  liquid. 
Occasionally  the  local  effects  may  have  been  due  to  the 
fact  that  the  liquid  was  cold  when  injected.  The  injection 
of  the  condensed  liquid  caused  no  other  untoward  symp- 
toms upon  the  animals,  which  is  quite  contrary  to  the 
observation  on  rabbits  of  Brown-Sequard  and  others." 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     233 

Later,  Rosenau  has  announced  that  guinea-pigs  kept  in 
stables  with  horses  become  sensitized  to  horse  serum. 

Wells  and  Osborne1  have  studied  the  anaphylactic  reac- 
tions of  some  pure  vegetable  proteins,  such  as  the  globulins 
from  castor  bean,  flax-seed,  and  squash-seed,  edestin  from 
hemp-seed,  excelsin  from  Brazil  nuts,  legumins  from  peas 
and  vetch,  vignin  from  cow  peas,  glycinin  from  soy  beans, 
gliadin  from  wheat  and  rye  flour,  hordein  from  barley,  and 
zein  from  maize.  "  It  has  been  found  that  all  these  proteins 
cause  typical  anaphylactic  reactions  in  sensitized  animals, 
with  all  features  essentially  the  same  as  when  serum  and 
other  animal  materials  containing  proteins  are  used.  The 
minimum  doses  which  produce  sensitization  and  the  time 
of  incubation  are  about  the  same  as  with  animal  proteins 
but  as  a  rule  the  symptoms  are  of  somewhat  slower  onset 
and  less  stormy  course  than  are  those  obtained  with  foreign 
sera,  and  the  minimum  intoxicating  doses  are  larger.  There 
are  also  considerable  differences  in  the  toxicity  of  the  several 
vegetable  proteins  to  sensitized  animals,  but  the  reasons  for 
these  differences  have  not  yet  been  investigated.  The  most 
toxic  proteins  as  measured  by  the  frequency  of  severe  and 
fatal  reactions,  were  the  globulin  of  squash-seed,  vignin, 
excelsin,  and  castor-bean  globulin,  which  usually  caused 
death  when  given  in  0.1  gram  doses  to  properly  sensitized 
animals.  Edestin  caused  the  least  severe  reactions  of  any 
of  the  proteins,  while  hordein  and  glycinin  seldom  caused 
fatal  reactions.  Nevertheless  the  minimum  sensitizing 
and  intoxicating  doses  of  edestin  and  squash-seed  globulin 
are  essentially  the  same.  The  influence  of  the  food  of  the 
guinea-pig  upon  the  anaphylactic  reaction  is  of  particular 
importance  in  experiments  with  vegetable  proteins,  since 
the  natural  food  of  the  guinea-pig  is  vegetable.  Experiments 
showed  that  continuous  feeding  with  a  vegetable  protein 
rendered  guinea-pigs  immune  to  this  protein,  so  that  they 
could  not  be  sensitized  to  it.  Although  brief  feeding  with 
animal  proteins  (cows'  milk,  foreign  sera,  egg  albumen) 

1  Jour.  Infect.  Dis.,  1911,  viii,  66. 


234  PROTEIN  POISONS 

renders  the  animal  sensitive  to  the  corresponding  animal 
protein,  probably  sufficiently  protracted  feeding  with 
animal  proteins  will  likewise  confer  immunity.  The  sensi- 
tization  through  feeding  is  specific  for  the  protein  food, 
showing  that  during  the  processes  preceding  and  including 
absorption  of  the  food  protein  no  change  takes  place  which 
robs  it  entirely  of  its  biological  specificity.  The  close 
similarity,  if  not  identity,  of  the  legumins  of  the  pea  and 
vetch  was  shown  by  the  interreaction  of  these  proteins, 
and  the  close  relation  to  vignin  from  the  pea  was  also 
indicated.  The  near  relation  or  probable  identity  of  the 
gliadins  from  wheat  and  rye  was  also  shown." 

This  is  in  accord  with  our  findings  of  some  years  ago, 
when  we  demonstrated  that  vegetable,  bacterial,  and  animal 
proteins  contain  the  same  poisonous  group. 

The  Sensitizing  Group  in  the  Protein  Molecule. — As  has 
been  stated  (Chapter  V)  we  have  split  proteins  into 
poisonous  and  non-poisonous  portions.  This  has  been  done 
with  proteins  of  most  diverse  origin,  bacterial,  vegetable, 
and  animal,  and  we  have  found  no  true  protein  which  has 
failed  to  undergo  this  cleavage.  Certain  pseudoproteins, 
like  gelatin,  do  not  respond  to  this  test,  but  all  true  proteins, 
so  far  as  tested,  have  been  split  into  poisonous  and  non- 
poisonous  portions.  This  is  the  foundation  stone  of  our 
theory  of  protein  sensitization.  All  true  proteins  are 
sensitizers,  and  so  far  it  has  not  been  shown  that  sensiti- 
zation can  be  established  by  any  non-protein  substance. 
All  sensitizers  develop  symptoms  of  poisoning  on  reinjec- 
tion.  These  symptoms  induced  by  reinjection  are  identical 
in  manifestation  and  sequence  with  those  induced  in  the 
fresh  animal  by  the  injection  of  the  poison  split  off  from 
the  protein  molecule  by  chemical  agents,  or  by  the  ferments 
in  the  serum  or  organ  extracts  of  sensitized  animals.  There- 
fore, we  have  concluded  that  anaphylactic  shock  is  due  to 
the  cleavage  of  the  molecule  of  the  protein  sensitizer  on 
reinjection,  and  the  liberation  of  the  protein  poison,  and 
this  cleavage  is  due  to  a  specific  proteolytic  enzyme  developed 
in  the  cells  of  the  animal  body  as  a  result  of  the  first  injec- 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      235 

tion.  We  have  repeatedly  shown  that  the  poisonous  group 
obtained  from  the  protein  molecule  by  cleavage  with 
chemicals  or  with  ferments  does  not  sensitize  animals. 
This  is  contrary  to  the  generally  accepted  view,  and  our 
claim  on  this  point  has  met  with  either  silence  or  denial, 
but  we  have  tested  this  matter  so  often  and  with  poisons 
obtained  from  so  many  and  such  a  variety  of  proteins  that 
we  have  no  hesitancy  in  affirming  that  the  poisonous  group 
in  the  protein  molecule  does  not  sensitize  animals.  But  it 
is  said  that  toxins  are  necessary  to  elaborate  antitoxins, 
and  that  the  latter  can  be  produced  in  no  other  way.  This 
is  true,  but  the  protein  poisons  are  not  toxins,  and  they 
lead  to  the  elaboration  of  no  antibodies.  The  toxins  are 
specific;  the  protein  poisons  are  not.  The  blood  serum  of 
an  animal  treated  properly  with  a  toxin  neutralizes  the 
toxin  both  in  vitro  and  in  vivo,  while  the  blood  serum  of  a 
sensitized  animal  renders  the  protein  with  which  the  animal 
has  been  treated,  when  brought  in  contact  with  it  under 
proper  conditions,  either  in  vitro  or  in  vivo,  poisonous.  It 
seems  to  us  that  it  has  been  positively  demonstrated  that 
the  sensitizing  and  toxic  groups  in  the  protein  molecule 
are  not  the  same.  It  might  be  argued  that  in  ordinary 
protein  mixtures,  such  as  blood  serum  and  egg-white,  one 
protein  may  contain  the  sensitizing  group  and  another  the 
toxic  group.  This  may  be  true,  but  when  pure  proteins, 
such  as  edestin,  are  used  the  two  groups  must  exist  in  the 
same  molecule.  The  specificity  of  proteins  is  demonstrated 
in  sensitization.  The  toxic  group  shows  no  specificity. 
This  property  characterizes  the  sensitizing  group,  and  it 
is  in  these  groups  that  the  fundamental  and  characteristic 
property  of  each  protein  resides.  The  exact  structure  and 
chemical  nature  of  neither  the  sensitizing  nor  the  poisonous 
groups  have  been  determined.  The  latter  seems  to  be 
physiologically  the  same  in  all  proteins,  the  former  is  specific 
in  every  protein.  By  our  method,  detailed  in  Chapter  V, 
the  poisonous  group  is  easily  obtained;  not  in  a  chemically 
pure  condition,  but  so  that  its  presence  can  be  demon- 
strated. The  poisonous  group,  being  the  same  in  all  proteins, 


236  PROTEIN  POISONS 

is  obtained  from  all  by  the  same  or  by  like  methods.  The 
sensitizing  group,  being  the  same  in  no  two  proteins,  cannot 
be  isolated  from  all  by  the  same  method.  We  have  been 
able  to  obtain  specific  sensitizing  groups  from  colon,  typhoid, 
and  tubercle  protein  quite  uniformly.  From  the  pneumo- 
coccus  and  related  organisms  we  have  never  succeeded  in 
obtaining  a  sensitizing  group.  From  egg-white  we  have 
rarely  succeeded,  generally  failed.  It  seems  evident  to 
us  that  the  sensitizing  groups  in  many  proteins  are  highly 
labile  bodies,  probably  of  such  delicate  structure  that  they 
easily  fall  to  pieces. 

If  sensitizers  are  ever  to  have  a  legitimate  place  in  the 
treatment  of  disease,  it  will  be  of  the  highest  importance 
to  obtain  them  free  from  the  poisonous  group.  Every 
time  an  unbroken  protein  is  introduced  into  the  body  it 
carries  with  it,  and  as  a  part  of  it,  a  poison.  From  the  very 
careless,  rash,  and  unwarranted  way  in  which  "vaccines" 
of  most  diverse  origin  and  composition  are  now  used  in 
the  treatment  of  disease,  this  matter  certainly  cannot  be 
understood  or  its  danger  appreciated  by  those  who  subject 
their  patients  to  such  risks.  It  should  be  clearly  understood 
that  all  proteins  contain  a  poisonous  group — a  substance 
which  in  a  dose  of  0.5  mg.  injected  intravenously  kills  a 
guinea-pig.  This  poison  is  present  in  all  the  so-called 
"vaccines"  now  so  largely  used,  and  it  is  not  strange  that 
death  occasionally  follows  the  use  of  "phylacogen"  or 
similar  preparations.  Not  only  do  these  proteins  contain 
a  poison,  but  when  introduced  parenterally  the  poison 
is  set  free,  not  in  the  stomach,  from  which  it  may  be  removed, 
but  in  the  blood  and  tissues.  It  is  possible  that  vaccine 
therapy  may  become  of  great  service  in  the  treatment  of 
disease.  Even  now  there  are  occasional  brilliant  results 
which  are  reported  while  the  failures  and  disasters  are  not 
so  widely  advertised.  But  before  sensitization  can  be  of 
great  service  in  a  therapeutical  way  we  must  secure  sensi- 
tizers free  from  poisonous  constituents.  Until  recently 
the  existence  of,  or  the  possibility  of  preparing  non-toxic 
sensitizers  has  been  made  evident  only  by  our  work. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     237 

Recently,  confirmation  of  our  studies  along  this  line  have 
come:  (1)  From  White  and  Avery,1  who  have  prepared 
by  our  method  a  sensitizing  group  from  tubercle  cell  sub- 
stance. (2)  From  Zunz,2  who,  as  the  result  of  a  most 
exhaustive  research,  has  shown  that  one  of  the  primary 
albumoses  (the  synalbumose  of  Pick)  sensitizes,  but  does 
not  induce  anaphylactic  shock  on  reinjection.  Zunz  states: 
Both  active  and  passive  anaphylaxis  can  be  induced  by 
the  three  so-called  primary  proteoses  (hetero-,  proto-,  and 
synalbumose),  but  not  by  thioalbumose,  nor  the  other 
so-called  secondary  proteoses,  nor  by  Siegfried's  pepsin- 
fibrin-peptone-j8,  nor  by  any  of  the  abiuret  products  of 
peptic,  tryptic,  or  ereptic  digestion. 

Animals  sensitized  with  hetero-,  proto-,  or  synalbumose 
develop  anaphylactic  shock  on  reinjection  with  the  original 
serum,  acid  albumin,  hetero-  or  proto-albumose,  but  not 
after  reinjection  with  synalbumose,  thio-albumose,  the  other 
secondary  proteoses,  pepsin-fibrin-peptone-|8,  or  any  of 
the  abiuret  products  of  peptic,  tryptic,  or  ereptic  digestion. 
The  hetero-  and  proto-albumoses  both  sensitize  and  induce 
anaphylactic  shock,  while  synalbumose  sensitizes  only. 
It  follows,  therefore,  that  sensitization  and  the  production 
of  anaphylactic  shock  are  due  to  different  groups  in  the 
protein  molecule. 

Wells  and  Osborne,3  working  with  the  purest  vegetable 
proteins  known,  hordein  from  barley,  glutinin  from  wheat, 
and  gliadin  from  both  wheat  and  rye,  find  that:  "Guinea- 
pigs  sensitized  with  gliadin  from  wheat  or  rye  give  strong 
anaphylactic  reactions  with  hordein  from  barley,  but  these 
are  not  so  strong  as  the  reactions  obtained  with  the  homolo- 
gous protein.  Similar  results  are  obtained  if  the  sensitizing 
protein  is  hordein,  and  the  second  injection  is  gliadin.  We 
have  here  a  common  anaphylaxis  reaction  developed  by 
two  chemically  distinct  but  similar  proteins  of  different 
biological  origin,  thus  indicating  that  the  specificity  of 

1  Jour.  Med.  Research,  1912,  xxvi,  317. 

2  Zeitsch.  f.  Immunitatsforschung,  1913,  xvi,  580. 

3  Jour.  Infect.  Dis.,  1913,  xii,  341. 


238  PROTEIN  POISONS 

the  reaction  is  determined  by  the  chemical  constitution  of 
the  protein  rather  than  by  its  biological  origin.  This  is 
in  harmony  with  the  fact  that  chemically  closely  related 
proteins  have,  as  yet,  been  found  only  in  tissues  biologically 
nearly  related. 

"From  the  results  of  these  experiments  it  seems  probable 
that  the  entire  protein  molecule  is  not  involved  in  the 
specific  character  of  the  anaphylaxis  reaction,  but  this  is 
developed  by  certain  groups  contained  therein,  and  that 
one  and  the  same  protein  molecule  may  contain  two  or 
more  such  groups." 

Evidently  the  view  that  the  protein  molecule  contains  a 
sensitizing  group,  one  or  more,  is  finding  strong  experi- 
mental support.  In  our  opinion  this  view  was  demonstrated 
by  Vaughan  and  Wheeler1  as  early  as  1907,  but  recent 
work,  such  as  that  by  Zunz,  Gay,  Wells  and  Osborne,  and 
others,  strengthens  the  evidence  then  offered.  According 
to  our  theory  every  protein  molecule  contains  a  chemical 
nucleus,  key-stone  or  archon.  This  is  the  protein  poison, 
and  is  physiologically  much  the  same  in  all  proteins.  One 
protein  differs  from  another  in  its  secondary  or  tertiary 
groups.  In  these  resides  the  biological  specificity  of  proteins. 
Biologically  related  proteins  contain  chemically  related 
groups,  and  in  these  are  found  the  sensitizing  agents.  The 
chemical  structure  of  the  protein  molecule  determines  its 
biological  differentiation  and  development.  It  is  not, 
therefore,  surprising  to  find  that  a  pure  protein  from  wheat 
sensitizes  to  another  closely  related  protein  from  such  a 
biologically  closely  related  grain  as  rye.  This,  however, 
does  not  indicate  that  the  proteins  from  the  two  grains  are 
wholly  identical  in  chemical  structure.  It  only  shows  that 
the  two  protein  molecules  contain  among  their  secondary 
groups  identical  or  closely  related  atomic  combinations. 
The  same  can  be  said  of  the  fact  that  certain  non-pathogenic 
acid-fast  bacteria  may,  at  least  partially,  sensitize  animals 
to  the  tubercle  bacillus.  Biological  relationship  is  deter- 

>Jour.  Infect.  Dis.,  iv,  476. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      239 

mined  by  the  chemical  structure  of  the  protein  molecule. 
We  hold  this  to  be  true  of  all  specific  biological  tests  for 
proteins,  whether  they  be  agglutination,  precipitin,  lytic, 
complement  deviation,  or  anaphylactic  tests.  The  chemical 
structure  of  the  protein  molecule  determines  all  of  these. 
The  form  and  function  of  every  cell  is  determined  by  the 
chemical  structure  of  its  constituent  proteins.  That  the 
sensitizing  agent  in  the  protein  molecule  resides  in  its 
secondary  groups  is  shown  by:  (a)  The  fact  that  sensitiza- 
tion  is  within  limits  specific;  (6)  The  fact  that  the  residues 
left  after  stripping  off  these  secondary  groups  by  proteo- 
lytic  digestion  or  by  the  action  of  dilute  bases  and  acids, 
do  not  sensitize.  Peptones,  polypeptids,  amino-acids,  and 
the  protein  poison  do  not  sensitize  to  either  themselves 
or  to  the  unbroken  proteins  from  which  they  have  been 
derived. 

The  Animal. — Guinea-pigs  give  the  most  striking  results. 
They  are  easily  sensitized  and  anaphylactic  shock  develops 
promptly  and  violently.  It  is  worthy  of  note  that  the 
work  of  Besredka  in  France  and  of  Rosenau  and  Anderson 
showed  great  difference  in  the  reaction  in  guinea-pigs 
in  the  two  countries.  In  this  country  practically  100  per 
cent,  of  the  animals  sensitized  to  horse  serum  die  on  the 
second  injection,  made  intraperitoneally ;  while  in  France 
the  highest  percentage  of  fatality  following  the  same  pro- 
cedure is  25.  It  was  at  first  supposed  that  this  difference 
is  due  to  the  race  of  horse  supplying  the  serum,  but  Rosenau 
and  Anderson,  using  Besredka's  serum,  obtained  the  same 
results  as  with  the  American  serum.  The  same  investi- 
gators say  that  the  difference  is  not  due  to  races  of  guinea- 
pigs.  In  our  work  with  egg-white  we  noted  a  much  higher 
percentage  of  mortality  with  short-haired,  smooth-coated 
animals  than  with  the  long,  curly-haired  ones. 

Doerr  and  Russ,1  using  ox  serum,  with  the  second  dose 
constant  at  0.2  c.c.,  found  the  following  comparative 
results  by  varying  the  sensitizing  dose,  both  doses  being 

1  Zeitsch.  f.  Immunitatsforschung,  ii,  109;  ibid.,  iii,  181. 


240  PROTEIN  POISONS 

administered  intravenously:  (1)  With  a  sensitizing  dose 
of  from  0.01  to  0.001  c.c.,  the  second  dose  was  followed 
uniformly  by  sudden  death.  (2)  With  the  sensitizing  dose 
reduced  to  from  0.0001  to  0.00001,  the  period  of  incubation 
was  prolonged,  but  after  this  the  results  of  the  second 
dose  were  the  same  as  in  the  former  instance.  (3)  When 
the  doses  were  further  reduced  to  0.000001  c.c.,  the  animals 
were  not  sensitized.  Pfeiffer  and  Mita  found  that  0.1 
c.c.  of  a  10  per  cent,  suspension  of  red  corpuscles  uniformly 
sensitized  guinea-pigs.  Vaughan  and  Wheeler  found  that 
1  c.c.  of  a  solution  of  egg-white  in  an  equal  volume  of  salt 
solution  sensitized  all  guinea-pigs,  and  the  result  was  death 
within  thirty  minutes  or  less  when  the  second  dose  con- 
sisted of  from  2  to  5  c.c.  of  the  same  solution.  Wells  found 
the  minimum  sensitizing  dose  of  the  purest  crystalline 
egg  albumen  which  he  obtained  to  be  0.05  mg.  Kraus 
and  Doerr  employed  from  \  to  1  loop  of  agar  cultures  of 
bacteria  for  the  sensitization  of  guinea-pigs,  but  Holobut 
and  Delanoe  found  repeated  injections  more  efficient. 

In  guinea-pigs  subcutaneous,  intraperitoneal,  and  intra- 
venous injections  of  soluble  proteins  are  practically  alike 
in  sensitization.  Rosenau  and  Anderson  sensitized  guinea- 
pigs  by  feeding  them  horse  serum.  The  question  of  sensi- 
tization by  way  of  the  alimentary  canal  will  be  discussed 
more  fully  later. 

Rabbits  are  not  so  easily  and  uniformly  sensitized  as 
guinea-pigs.  Friedemann1  recommends  the  following  method 
for  the  complete  sensitization  of  rabbits:  The  intravenous 
injection  of  1  c.c.  of  a  heterologous  serum  per  kilo;  repetition 
of  the  same  after  one  month,  and  the  giving  of  same  dose 
in  the  same  way  eight  days  later.  When  this  is  done  the 
animals  are  found  to  be  highly  anaphylactized.  We  have 
found  rabbits  highly  anaphylactized  when  treated  daily 
with  very  small  intravenous  injections  for  a  week,  and 
then  after  from  three  to  six  months  later  given  a  like 
intravenous  injection. 

1  Zeitsch.  f.  Immunitatsforschung,  ii. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      241 

Doerr  and  Russ  were  unable  to  sensitize  mice,  but  Braun 
succeeded  after  repeated  injections  and  Ritz  after  a  single 
treatment,  when  the  reinjection  was  made  intravenously. 

Goats  and  sheep  have  been  sensitized  by  Friedemann 
and  Isaac,1  and  horses  and  birds  by  Doerr. 

Dogs  are  not  easily  anaphylactized.  Friedemann  and 
Isaac,  also  Remlinger,2  failed,  and  the  former  suggested 
that  it  was  due  to  the  fact  that  this  animal  is  largely  car- 
nivorous, but  Biedl  and  Kraus,3  using  the  finer  method 
of  measuring  anaphylactic  shock  by  fall  in  blood  pressure, 
has  made  this  animal  of  great  value  in  studying  the  phe- 
nomena of  anaphylaxis.  The  dog  has  also  been  employed 
by  Arthus,4  by  Manwaring,  Pearce,  Edmunds,  and  others. 

Vaughan  and  Wheeler  failed  to  sensitize  cats  by  a  single 
intra-abdominal  injection  of  egg-white,  but  repeated  injec- 
tions were  not  tried.  More  recently  it  has  been  shown 
that  the  cat  can  be  easily  sensitized.  Up  to  the  present 
time  no  animal  thoroughly  tested  has  failed  to  respond 
to  protein  sensitization. 

We  are  without  sufficient  data  to  determine  with  any 
certainty  the  relative  susceptibility  of  man  to  protein 
sensitization.  As  with  other  animals,  man's  susceptibility 
evidently  varies  within  wide  limits  with  the  protein  supplying 
the  anaphylactogen.  Pirquet  and  Shick  have  shown  that 
a  high  degree  of  sensitization  may  result  from  a  single 
relatively  small  dose  of  a  heterologous  serum.  The  high 
degree  of  sensitization  shown  by  many  in  hay  fever  and 
in  susceptibility  to  certain  foods  raises  questions  which 
will  be  discussed  later. 

Period  of  Incubation. — By  the  period  of  incubation  we 
indicate  the  interval  of  time  between  the  introduction  of 
the  anaphylactogen  and  that  time  when  the  body  is  recog- 
nizably disturbed  by  a  reinjection  of  the  same  or  a  closely 
related  protein.  It  is  quite  properly  designated  as  the 

1  Zeitsch.  f.  Exp.  Path.  u.  Ther.,  i,  513. 

2  Compt.  rend,  de  la  Soc.  biol.,  Ixii,  23. 

3  Wien.  klin.  Woch.,  1909,  363. 

4  Compt.  rend,  de  1'Acad.  Sci.,  cxlviii,  1002. 
16 


242  PROTEIN  POISONS 

pre-anaphy lactic  state.  It  covers  the  time  necessary  for 
the  development  of  anaphylaxis.  The  first  injection  of 
the  foreign  protein  is  without  manifest  effect  upon  the 
animal,  but  in  reality  it  has  a  most  profound  effect.  It 
induces  changes  which  may  continue  throughout  life,  and 
may  be  transmitted  from  mother  to  offspring.  The  limits 
of  the  pre-anaphylactic  state  have  been  studied  only  in  the 
guinea-pig  sensitized  to  horse  serum.  In  these  studies 
it  appears  that  the  shortest  time  required  for  the  develop- 
ment of  the  anaphylactic  state  is  from  six  to  nine  days, 
and  the  usual  time  from  ten  to  twelve  days.  Otto,  Rosenau 
and  Anderson,  and  Gay  and  Southard  uniformly  found 
that  large  doses  of  the  anaphylactogen  (5  c.c.  or  more  of 
horse  serum)  prolonged  the  pre-anaphylactic  state  or 
delayed  the  full  development  of  sensitization.  Friedberger 
and  Burkhard1  have  apparently  contradicted  this  finding, 
but  since  the  maximum  dose  employed  by  the  latter  was 
only  1  c.c.,  we  fail  to  see  that  there  is  any  conflict.  Evi- 
dently there  is  a  maximum  amount  of  anaphylactogen 
which  the  body  cells  can  dispose  of  within  six  or  eight  days, 
and  that  this  for  horse  serum  in  the  guinea-pig  is  something 
more  than  1  c.c.,  and  something  less  than  5  c.c.  As  has 
been  stated,  Doerr  and  Russ  found  that  when  the  sensi- 
tizing dose  was  less  than  0.001  c.c.  of  ox  serum  the  pre- 
anaphylactic  stage  is  also  prolonged.  It  seems  rational 
to  conclude  from  all  the  evidence  at  hand  that  with  sensi- 
tizing doses  of  0.001  to  1  c.c.  of  serum  the  average  duration 
of  the  pre-anaphylactic  state  is  from  ten  to  twelve  days, 
with  a  minimum  period  of  six  days.  With  sensitizing  doses 
above  or  below  these  limits  the  period  of  incubation  may  be 
prolonged. 

The  Anaphylactic  State. — Rosenau  and  Anderson,  also 
Gay  and  Southard,  found  that  guinea-pigs  sensitized  to 
horse  serum  remain  in  this  condition  for  at  least  two  years. 
It  may  possibly  continue  throughout  life.  Vaughan  and 
Wheeler  found  that  guinea-pigs  lose  their  anaphylactic 

1  Zeitsch.  f.  Immunitatsforschung,  iv,  690. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     243 

state  to  egg-white  after  about  one  year,  and  after  this 
time  they  can  be  resensitized.  The  same  investigators 
found  that  guinea-pigs  sensitized  to  colon  or  typhoid 
proteins  begin  to  lose  their  sensitization  after  thirty 
days.  Pirquet  and  Shick  report  the  continuance  of  the 
sensitized  state  in  man  after  treatment  with  diphtheria 
antitoxin  up  to  more  than  seven  years,  and  Currie1  up  to 
five  years. 

The  Reinjection. — This  term  has  come  to  have  in  this 
connection  a  restricted  and  definite  meaning.  Repeated 
injections  may  be  employed  in  inducing  anaphylaxis,  but 
by  "  reinjection"  we  mean  the  one  made  after  the  anaphyl- 
actic  state  has  been  established.  While  subcutaneous, 
intra-abdominal,  and  intravenous  methods  of  adminis- 
tration are  alike  suitable  and  effective  in  inducing  anaphyl- 
axis, the  intravenous  reinjection  is  much  the  most  effective. 
Besredka  has  been  partial  to  the  intracerebral  introduction 
of  the  "reinjection,"  and  he  claims  that  it  has  advantages 
over  the  intravenous,  although  the  latter  is  the  most  effec- 
tive. With  a  set  of  highly  sensitized  guinea-pigs  and  with 
the  same  serum  he  obtained  the  following  comparative 
results  with  the  different  methods  of  administration:  Fatal 
dose,  intravenously,  TV  to  ^V  c-c-5  intracranially,  y1^  to  -§-; 
intraperitoneally,  only  about  one-half  the  animals  responded 
to  5  c.c.,  and  subcutaneously  this  amount  had  scarcely  any 
effect.  Besredka  prefers  intracranial  injections  because 
they  are  not  so  delicate  as  the  intravenous,  and  more  easily 
measured.  In  conjunction  with  Steinhardt2  he  has  developed 
a  method  of  standardizing  sera.  He  has  found  that  sera 
differ  widely  in  toxicity,  as  tested  on  sensitized  guinea-pigs 
by  intracerebral  injections,  the  fatal  dose  varying  from  \ 
c.c.  in  a  sample  thirteen  years  old,  to  y^-g-  c.c.  in  some  fresh 
samples.  This  variation  is  in  part  due  to  age  and  in  part 
to  the  horses  from  which  they  are  taken.  His  studies  on 
the  effects  of  age  on  the  toxicity  of  sera  as  tested  upon 


1  Jour,  of  Hygiene,  viii.  35. 

2  Ann.  d.  1'Institut  Pasteur,  1907,  157. 


244  PROTEIN  POISONS 

anaphylactized  animals  are  interesting  and  important. 
He  finds  that  sera  from  the  different  horses  of  the  Pasteur 
Institute,  all  of  the  same  race,  and  on  the  same  food,  show 
some,  but  not  marked,  variations.  On  the  day  that  it  is 
drawn,  horse  serum  is  highly  toxic  tested  by  this  method. 
During  the  first  ten  days  the  toxicity  rapidly  decreases, 
after  that  time  more  slowly.  In  making  these  tests  he 
uses  guinea-pigs  already  used  in  the  standardization  of 
diphtheria  antitoxin,  and  thus  saves  expense.  The  standards 
for  therapeutic  sera,  established  by  the  Frankfurt  Institute 
are  as  follows:  (1)  It  must  be  clear  and  contain  no  marked 
deposit.  (2)  It  must  not  contain  bacilli.  (3)  The  highest 
phenol  content  must  be  0.5  per  cent.  (4)  It  must  contain 
no  free  toxin,  especially  tetanus  toxin.  Besredka  thinks 
a  fifth  requirement  should  be  made,  namely,  that  the  D.  L. 
should  be  less  than  ^o  c-c->  as  tested  intracerebrally  on 
sensitized  guinea-pigs.  He  makes  the  following  statement 
concerning  the  average  serum  of  the  Pasteur  Institute: 
The  first  day  it  is  hypertoxic  (D.  L.  is  -^  c.c.);  on  the 
eleventh  day  it  has  fallen  to  one-half  (D.  L.  is  -£$  c.c.); 
by  the  forty-fifth  day  the  last-mentioned  dose  induces 
severe  symptoms,  but  does  not  kill;  after  two  months  it 
has  fallen  to  ^  c.c.,  and  after  this  the  decrease  is  very  slow. 
He  thinks  it  wise  not  to  use  a  serum  less  than  two  months 
old.  In  France  all  therapeutic  sera  are  heated  to  50°  before 
distribution,  and  Besredka  states  that  cases  of  serum 
disease  are  less  frequent,  and  when  they  do  occur,  less 
serious  than  in  countries  in  which  unheated  sera  are  em- 
ployed. The  temperature  cannot  be  raised  above  60° 
without  weakening  the  antitoxin.  Rosenau  and  Anderson 
have  tried  many  chemicals  and  ferments  with  the  hope  of 
destroying  the  anaphylactic  toxicity  of  therapeutic  sera 
without  injuring  the  antibody,  but  with  wholly  negative 
results.  Other  methods  of  averting  the  dangers  of  serum 
disease  will  be  discussed  elsewhere. 

Besredka  finds  that  milk  may  be  heated  to  100°  for  twenty 
minutes,  or  to  120°  for  fifteen  minutes  without  appreciable 
loss  in  its  anaphylactic  toxicity.  It  begins  to  lose,  however, 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     245 

when  the  temperature  reaches  130°;  at  135°  to  140°  it 
becomes  gelatinous  and  is  no  longer  toxic  when  tested  on 
sensitized  guinea-pigs. 

Symptoms. — The  symptoms  induced  by  the  reinjection 
of  a  homologous  or  closely  related  protein  into  an  anaphyl- 
actized  animal  vary  within  certain  limits  in  different  species 
of  animal,  but  in  the  same  species  are  constant,  whatever 
the  protein  used.  This  in  itself  is  a  strong  argument  in 
favor  of  the  claim  made  by  us  that  the  anaphylactic  poison 
is  the  same,  in  its  physiological  action  at  least,  whatever 
the  protein  be.  In  other  words,  it  is  strongly  in  favor  of  the 
view  that  all  proteins,  at  least  all  which  possess  the  capa- 
bility of  sensitizing  animals,  contain  the  same  poisonous 
group  and  the  symptoms  are  due  to  the  liberation  or 
activation  of  this  poison. 

When  a  sensitized  guinea-pig  receives  a  reinjection  of 
the  same  protein  to  which  it  has  been  sensitized  after  a 
proper  interval  of  time,  certain  characteristic  and  prac- 
tically invariable  symptoms  develop;  generally  within 
five  or  ten  minutes,  sometimes  as  late  as  thirty  or  forty 
minutes.  These  symptoms  develop  in  three  stages,  which 
are  best  studied  when  they  do  not  proceed  too  rapidly. 
For  this  reason  the  reinjection  should  be  made  intra- 
peritoneally.  When  given  intravenously  the  symptoms 
develop  so  rapidly  that  a  study  of  the  different  stages  may 
be  difficult  or  quite  impossible.  The  first  stage  is  that  of 
peripheral  irritation.  The  animal  is  excited  and  evidently 
itches  intensely,  as  is  shown  by  its  attempts  to  scratch 
every  part  of  its  body  that  it  can  reach  with  its  feet.  The 
second  stage  is  one  of  partial  paralysis.  The  animal  lies 
upon  its  side  or  belly,  with  rapid,  shallow,  and  difficult 
breathing.  It  is  disinclined  to  move,  and  when  urged  to 
do  so  shows  more  or  less  incoordination  of  movement,  and 
muscular  weakness,  with  partial  paralysis,  especially  obser- 
vable in  the  posterior  extremities,  which  it  drags.  Rarely 
the  animal  dies  in  this  stage.  The  third,  or  convulsive 
stage,  begins  with  throwing  the  head  back  at  short  intervals. 
The  convulsions  become  general,  more  frequent  and  violent, 


246  PROTEIN  POISONS 

and  the  animal  having  reached  this  stage,  usually  dies  in  a 
convulsion  or  immediately  following  one.  Expulsion  of 
urine  and  feces  is  frequent  in  the  convulsive  stage.  Recovery 
after  reaching  the  convulsive  stage  is  exceedingly  rare. 
When  this  stage  is  not  reached,  recovery  usually  occurs, 
and  is  so  prompt  and  complete  that  after  a  few  hours,  or 
at  most  by  the  next  day,  the  animal  cannot  be  distin- 
guished from  its  perfectly  healthy  fellows. 

This  is  an  exact  reproduction  of  the  picture  of  poisoning 
an  untreated  guinea-pig  with  the  protein  poison  of  Vaughan 
a"nd  Wheeler,  and  another  indication  that  this  and  the 
anaphylactic  poison  are  one  and  the  same. 

In  dogs  the  first  two  stages,  as  seen  in  the  guinea-pig, 
occur  with  some  variations.  The  first  stage  is  one  of  excite- 
ment. The  animal  moves  about  uneasily  and  cries.  He 
retches  and  sometimes  vomits.  Expulsion  of  urine  and 
feces  frequently  occurs.  In  the  second  stage,  one  of 
great  muscular  weakness,  he  lies  flat  on  his  side  or  belly, 
with  his  head  on  the  table.  When  placed  on  his  feet  he 
stumbles,  falls,  and  lies  with  extended  legs,  as  if  paralyzed. 
There  may  be  marked  expiratory  spasms  with  retching, 
and  repeated  expulsion  of  feces.  There  is  finally  suppres- 
sion of  urine.  The  animal  remains  in  this  state  of  depression 
for  many  hours,  and  then  dies  or  slowly  and  completely 
recovers,  so  that  the  next  day  it  seems  as  well  as  ever. 
Again,  this  is  a  duplication  of  the  poisoning  produced  in  an 
untreated  dog  with  the  protein  poison. 

Acute  anaphylactic  shock  is  seen  in  men  being  treated 
with  sera  or  other  albuminous  fluids.  We  saw  it  repeatedly 
some  years  ago  when  we  treated  tuberculosis  with  yeast 
nuclein,  and  the  tuberculin  reaction  is  one  of  sensitization. 
It  is  not  our  purpose  to  go  into  detail  concerning  the  ana- 
phylactic phases  seen  in  man  under  various  conditions. 
That  part  of  our  subject  will  be  dealt  with  later.  At 
present  we  are  to  speak  only  of  acute  anaphylactic  shock  in 
man.  When  the  homologous  protein  is  injected  into  a  man 
sensitized  by  disease  or  by  previous  treatments,  symptoms 
develop  promptly,  often  within  a  few  minutes,  usually 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     247 

within  a  few  hours.  The  stage  of  peripheral  irritation  is 
characterized  by  the  sudden  appearance  of  a  rash.  The 
rashes  that  occur  most  promptly  are  urticarial  or  erythema- 
tous.  We  have  seen  such  a  rash  rapidly  spread  over  the 
surface  like  a  blush  in  every  direction  from  the  point  of 
injection,  and  soon  involve  the  entire  surface.  The  lips  and 
tongue  seem  swollen,  and  often  the  backs  of  the  hands  are 
swollen.  The  individual  becomes  apprehensive,  says  that 
he  cannot  breathe,  and  falls  into  a  state  of  more  or  less 
marked  collapse.  In  extreme  instances  there  is  retching, 
and  occasionally  vomiting.  The  second  stage,  that  of 
great  muscular  weakness,  continues  for  a  variable  time  and 
usually  rapidly  passes  away.  In  rare  instances  speedy 
death  results. 

The  Mechanism  of  Anaphylaxis. — Gay  and  Southard1 
were  the  first  to  study  the  pathological  changes  induced 
by  anaphy lactic  shock  in  guinea-pigs.  They  reported 
minute  hemorrhages  in  the  pleura  and  in  the  mucous  mem- 
brane of  the  stomach,  and  showed  that  the  lungs  are  inflated 
after  death.  Auer  and  Lewis2  made  plain  that  death  in 
guinea-pigs  from  anaphylaxis  is  not  due  to  effects  on  the 
central  nervous  system,  but  is  due  to  tetanic  contraction  of 
the  smooth  muscles  of  the  bronchioles.  They  also  demon- 
strated that  these  spasms  could  be  averted  and  life  saved 
by  preventive  injections  of  atropine.  These  findings  have 
been  fully  confirmed  by  subsequent  researches,  especially 
those  of  Biedl  and  Kraus.  It  is  to  the  last-named  investi- 
gators that  we  owe  the  most  complete  and  satisfactory 
demonstration  of  the  mechanism  of  anaphylactic  shock. 
Biedl  and  Kraus3  have  summed  up  their  own  and  the 
researches  of  others  on  this  point  up  to  the  time  of  their 
writing  (1910).  We  will  first  follow  the  summary  and  then 
review  the  work  done  since  that  time. 

In  dogs,  fall  in  blood  pressure  is  a  characteristic  and 

1  Jour.  Med.  Research,  1908. 

2  Jour.  Arner.  Med.  Assoc.,  1909. 

3  Kraus  and  Levaditi,  Handbuch  d.  Technik  u.  Methodik  d.  Immuni- 
tatsforschung,  Erganzungsband,  i. 


248  PROTEIN  POISONS 

constant  result  of  the  reinjection.  When  sensitization  is 
not  complete,  fall  in  .pressure  may  be  the  only  symptom. 
In  all  cases  there  is  complete  parallelism  between  the 
clinical  symptoms  and  the  fall  in  blood  pressure.  As  the 
latter  proceeds  the  former  increase  in  intensity,  and  in 
recovery  rise  in  pressure  accompanies  the  return  to  the 
normal.  With  a  normal  pressure  of  from  120  to  150  mm. 
of  mercury,  soon  after  the  reinjection,  and  as  the  pulse 
grows  smaller  and  faster  and  the  general  depression  deepens, 
the  blood-pressure  in  the  femoral  artery  falls  to  80  or  60, 
sometimes  to  40  or  even  less.  The  character  of  the  curve 
changes,  the  effects  of  respiratory  movements  become  less 
marked,  and  cease  altogether  as  the  pressure  approaches 
the  lowyest  point.  Now,  only  the  movements  of  retching 
and  expiratory  spasms  cause  transitory  rises  in  the  curve. 
When  the  lowest  point  is  reached  the  depression  is  greatest 
and  recovery  is  indicated  by  and  accompanies  rise  in 
pressure.  The  corneal  and  cutaneous  reflexes  remain 
intact  throughout,  and  exclude  both  a  central  narcosis  and 
peripheral  muscular  paralysis.  The  absence  of  marked 
respiratory  disturbances  is  an  additional  indication  in  the 
same  direction,  and,  furthermore,  shows  that  the  respiratory 
function  of  the  red  corpuscles  is  not  at  fault. 

The  genesis  of  the  fall  in  blood  pressure  becomes  an 
interesting  question.  The  type  of  the  fall  and  the  accom- 
panying condition  of  the  pulse  show  that  it  is  not  due  to 
weakness  of  the  heart's  action.  More  than  fifty  years  ago 
it  was  shown  by  Marey  that  a  fall  in  blood-pressure  accom- 
panied by  increased  frequency  of  the  pulse  is  not  due  to  a 
decrease  in  the  strength  of  the  heart,  but  in  all  probability 
to.  decreased  peripheral  resistance.  Both  in  the  course  of 
the  fall  and  after  it  has  reached  the  lowest  point  there  is 
no  irregularity  in  the  action  of  the  heart.  On  the  contrary, 
while  in  non-narcotized  dogs  immediately  after  the  reinjec- 
tion the  heart  beat  becomes  slower  and  sometimes  irregular, 
as  the  pressure  falls  the  heart  becomes  and  remains  regular. 
That  the  heart  is  not  injured  is  furthermore  shown  by  the 
fact  that  with  spasmodic  expirations  in  which  the  abdominal 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     249 

viscera  are  compressed,  the  pressure  invariably  rises.  It 
follows  that  the  low  blood  pressure  in  anaphylactic  shock 
is  due  to  decreased  peripheral  resistance  from  marked 
peripheral  vasodilatation. 

The  next  question  is  to  determine  whether  the  dilata- 
tion is  due  to  paralysis  of  the  vasomotor  centre  or  that  of 
the  periphery.  At  first  it  seemed  that  the  trouble  might 
be  central,  because  stimulation  of  the  vasomotor  centre 
failed  to  increase  the  blood  pressure.  But  this  is  negatived 
by  the  fact  that  stimulation  of  the  peripheral  vasomotor 
apparatus  otherwise  than  through  the  centre  also  failed 
to  increase  the  pressure.  Not  only  did  irritation  of  the 
terminal  end  of  the  splanchnic  fail  to  raise  the  pressure, 
but  the  intravenous  injection  of  from  0.1  to  0.2  mg.  of 
adrenalin,  which  in  normal  animals  is  followed  by  marked 
increase  in  pressure,  in  anaphylactic  shock  is  either  wholly 
without  effect  or  has  but  slight  influence. 

It  is  generally  held  that  the  capability  of  increasing  blood 
pressure  possessed  by  adrenalin  is  due  to  its  action  on  the 
nervous  apparatus  in  the  vessel  walls,  and  possibly  in  part 
on  the  vessel  muscle.  It  follows  that  the  vasodilatation 
of  anaphylactic  shock  is  due  to  paralysis  of  the  peripheral 
vasomotor  apparatus.  Stimulation  of  the  vasomotor  centre 
naturally  fails  to  raise  the  pressure  because  the  end  appa- 
ratus does  not  work.  It  should  be  stated  that  this  failure 
of  adrenalin  to  raise  the  pressure  occurs  only  in  the  stage  of 
deep  depression  when  the  pressure  is  low.  If  the  pressure 
begins  to  rise,  as  recovery  begins,  then  the  administration 
of  adrenalin  carries  it  up  rapidly.  That  the  fall  in  blood 
pressure  in  anaphylactic  shock  is  due  to  a  transitory  paralysis 
of  the  peripheral  vasomotor  apparatus  seems  to  be  quite 
conclusively  demonstrated. 

As  has  been  shown  by  Boehm,  barium  chloride  causes  a 
marked  and  fairly  persistent  increase  in  blood  pressure, 
which  is  due  to  its  stimulating  effect  upon  the  smooth 
muscles  of  the  vessel  walls.  In  anaphylactic  shock  even 
when  the  pressure  has  fallen  to  the  lowest  point,  the  adminis- 
tration of  barium  chloride  causes  a  marked  rise.  Moreover, 


250  PROTEIN  POISONS 

when  barium  chloride  is  given  before  the  reinjection,  the 
latter  does  not  cause  a  fall  in  blood  pressure.  Still  more 
striking  is  the  fact  that  when  barium  chloride  is  given  in 
anaphylactic  shock,  as  the  pressure  rises  the  symptoms 
disappear;  also,  when  this  substance  is  given  in  doses  which 
cause  in  normal  animals  a  marked  and  peristent  increase 
in  pressure,  before  the  reinjection,  the  latter  induces  no 
anaphylactic  symptoms.  That  the  animals  upon  which 
these  observations  were  made  were  in  a  sensitized  state 
was  proved  by  inducing  passive  anaphylaxis  in  normal 
animals  with  their  sera.  It  will  be  seen  from  the  above 
that  in  experimental  anaphylaxis  in  dogs,  barium  chloride 
is  efficient  both  as  a  preventive  and  a  curative  agent. 

The  antagonistic  action  of  barium  chloride  demonstrates 
the  peripheral  genesis  of  anaphylactic  vasodilatation,  but 
it  does  not  wholly  settle  the  question  as  to  whether  the 
dilatation  is  due  to  the  effect  of  the  poison  on  the  nerves 
or  on  the  smooth  muscle.  The  failure  of  adrenalin  and 
the  success  of  barium  chloride  in  raising  the  pressure  in 
anaphylactic  shock  render  it  highly  probable  that  the 
anaphylactic  poison  lowers  the  blood  pressure  by  paralysis 
of  the  smooth  muscle  of  the  vessel  walls.  It  seems  quite 
certain  that  barium  chloride  and  the  anaphylactic  poison 
act  upon  the  same  peripheral  apparatus,  that  the  action  of 
the  former  is  stimulating,  and  that  of  the  latter  is  paralyzing, 
and  the  former  is  the  stronger  and  able  to  prevent  or  replace 
the  latter. 

Having  established  the  fact  that  fall  in  blood  pressure 
is  a  marked  and  constant  result  of  the  anaphylactic  poison, 
the  symptoms  become  easily  explainable.  The  resulting 
anemia  of  the  brain  explains  the  disturbances  of  respiration, 
the  retching,  the  expulsion  of  urine  and  feces,  the  great 
depression  and  muscular  weakness,  and  the  speedy  recovery, 
when  death  does  not  result. 

Biedl  and  Kraus  give  as  additional  phenomena  of 
anaphylactic  shock  the  following:  (1)  On  reinjection 
the  coagulability  of  the  blood  falls  markedly  or  wholly 
disappears.  Before  the  reinjection  is  made  the  blood  of  a 


•    PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     251 

sensitized  dog  coagulates  like  that  of  a  normal  animal,  while 
that  in  anaphylactic  shock  remains  fluid  for  hours  and 
even  days.  (2)  During  anaphylactic  shock  the  polynuclear 
leukocytes  wholly  disappear  from  the  blood,  while  the 
lymphocytes  and  platelets  remain.  (3)  A  second  reinjec- 
tion  made  in  the  depression  phase  or  some  hours  later,  or 
on  the  next  day  after  recovery,  is  wholly  without  effect. 
This  is  true  whether  the  amount  of  serum  employed  in  the 
second  reinjection  is  small  or  large.  Furthermore,  the 
animal  is  anti-anaphylactic  after  the  shock  has  been  either 
prevented  or  relieved,  by  the  administration  of  barium 
chloride.  However,  Biedl  and  Kraus  did  find  one  dog 
which  rapidly  recovered  under  barium  chloride  responsive 
to  a  second  reinjection  made  the  next  day. 

Biedl  and  Kraus  compare  anaphylactic  shock  with  the 
poisoning  of  normal  dogs  with  Witte's  peptone,  and  find 
that  even  in  the  minutest  details  they  are  not  only  similar, 
but  identical.  The  intravenous  injection  of  Witte's  peptone 
in  normal  dogs  in  doses  from  0.3  to  0.03  grams  per  kilo 
causes  fall  in  blood  pressure,  loss  of  coagulability  of  the 
blood,  the  disappearance  of  polynuclear  leukocytes,  and 
peptone  immunity.  Moreover,  peptone  poisoning  can  be 
prevented  or  relieved  by  injections  of  barium  chloride. 
They  conclude  that  anaphylactic  intoxication  is  caused  by 
a  poison  which  is  physiologically  identical  with  the  active 
constituent  of  Witte's  peptone.  This  is  of  the  highest 
importance  to  us  because  we  hold  that  the  protein  poison 
of  Vaughan  and  Wheeler  is  the  active  principle  of  Witte's 
peptone,  and  in  fact  of  all  proteins  which  contain  ana- 
phylactogens.  We  have  prepared  this  poison  from  Witte's 
peptone  as  well  as  from  other  proteins,  bacterial,  vegetable, 
and  animal.  We  will  return  to  this  point. 

It  should  be  understood  that  the  above  extracts  from  the 
researches  of  Biedl  and  Kraus  refer  only  to  serum  anaphyl- 
actic intoxication,  as  observed  in  dogs.  As  has  been  stated, 
the  cause  of  death  from  anaphylactic  shock  in  guinea-pigs 
was  discovered  by  Gay  and  Southard  and  more  fully  studied 
by  Auer  and  Lewis,  and  has  been  confirmed  by  subsequent 


252  PROTEIN  POISONS 

investigators.  There  is  spasmodic  contraction  of  the  muscles 
of  the  bronchioles.  This  is  independent  of  central  injury, 
or,  in  other  words,  is  due  to  peripheral  action,  and  is  pre- 
vented or  relieved  by  the  intravenous  administration  of 
atropine  in  doses  of  from  1  to  10  nig.,  provided  the  drug  is 
given  before  the  heart  stops.  In  anaphylactic  shock  in 
guinea-pigs  there  is  a  primary  rise  in  blood  pressure, 
which  after  an  intravenous  reinjection  lasts  from  thirty 
seconds  to  two  minutes.  This  is  followed  by  a  sudden 
fall  which  may  go  as  low  as  20  or  even  10  mm.  of  mercury. 
But  the  fall  in  blood  pressure  is  not  the  cause  of  death. 
It  is  on  account  of  the  difference  in  action  of  the  anaphyl- 
actic poison  in  guinea-pigs  and  dogs  that  the  symptoms  in 
the  two  species  vary.  The  sudden  onset,  the  stormy  progress, 
and  the  fatal  ending  of  the  symptoms  in  guinea-pigs  are 
seldom  or  never  seen  in  dogs.  In  the  former,  spasmodic 
contraction  of  the  bronchioles  prevents  the  expiration  of 
the  air,  and  when  the  lungs  are  laid  bare  they  are  seen  to 
fill  the  thoracic  cavity;  they  do  not  collapse,  and  are  pale 
and  bloodless.  Biedl  and  Kraus  have  shown  that  these 
conditions,  characteristic  of  anaphylactic  intoxication  in 
guinea-pigs,  result  also  when  guinea-pigs  are  poisoned  with 
peptone.  Besides,  fatal  poisoning  with  peptone  may  be 
prevented  by  the  intravenous  injection  of  atropine.  Thus, 
it  is  shown  that  in  these  animals  also  the  anaphylactic 
poison  is  identical,  physiologically  at  least,  with  the  active 
constitutent  of  Witte's  peptone.  In  dogs  this  poison  par- 
alyzes the  vessel  muscles  of  the  splanchnic  region,  while 
in  guinea-pigs  it  stimulates  the  constrictor  muscles  of  the 
bronchioles. 

Biedl  and  Kraus,  having  come  to  the  conclusion  that 
anaphylactic  intoxication  and  peptone  poisoning  are  iden- 
tical, discuss  the  poisonous  property  of  peptone.  Pick  and 
Spiro  state  that  there  are  peptones  which  do  not  lower  the 
blood  pressure  or  lessen  the  coagulability  of  the  blood,  and 
that  there  are  digestive  products  containing  no  albumose 
or  peptone,  or  only  traces  of  either,  which  do  induce  these 
poisonous  effects.  They  conclude  that  they  are  peptones 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      253 

devoid  of  peptone  action,  and  there  may  be  peptone  action 
without  peptone.  They  think  that  in  the  peptic  digestion 
of  proteins  there  is  formed  in  small  amount  a  highly  poison- 
ous body  for  which  they  propose  the  name  peptozym. 
Popielski,1  who  has  made  a  chemical  and  physiological 
study  of  Witte's  peptone,  states  that  the  albumose  con- 
tained in  it  is  without  effect,  and  that  peptone  prepared 
from  it  by  the  method  of  Pick  has  the  same,  but  less  marked, 
action  as  the  original  Witte's  peptone.  From  this  he  con- 
cludes that  the  active  agent  is  not  peptone.  He  also  con- 
cludes that,  in  peptic  digestion,  a  highly  poisonous  sub- 
stance is  formed  along  with  the  peptone,  and  on  account 
of  its  action  he  proposes  the  name  "vasodilatin."  This 
he  obtained  in  an  impure  state  by  fractional  precipitation 
of  aqueous  solutions  of  Witte's  peptone  with  hot,  absolute 
alcohol.  This  substance  is  highly  active  and  contains 
relatively  small  amounts  of  albumose  and  peptone,  and  no 
cholin.  This  agrees  well  with  our  own  work.  As  has  been 
stated,  we  have  prepared  our  protein  poison  from  Witte's 
peptone,  but  Nicolle  and  Abt2  could  not  obtain  it  by  our 
method  from  Defresne's  peptone,  and  we  subsequently 
confirmed  this.  Gastric  digestion  is  a  progressive  process, 
and  it  progresses  through  its  successive  stages  at  widely 
differing  rates.  When  it  is  arrested  as  in  the  manufacture 
of  peptones,  the  product  may  contain  the  poisonous  group, 
either  in  combination  or  free,  or  the  digestion  may  have 
continued  to  the  destruction  of  the  protein  poison.  This 
seems  a  simple  and  rational  explanation  of  the  above- 
mentioned  findings,  and  reconciles  their  apparent  contra- 
dictions. If  this  be  the  correct  explanation,  one  batch  of 
peptone  may  contain  the  poison,  while  another  from  the 
same  manufacturer  may  contain  no  trace  of  it.  The  protein 
poison  is  a  group  in  the  protein  molecule;  at  each  successive 
step  in  the  digestive  process  it  exists  in  a  smaller  and  more 
labile  molecule,  and  finally  it  itself  is  broken  up  and  rendered 
inert. 

1  Arch.  f.  Exp.  Path.,  Ivi;  Pfliiger's  Archiv,  cxxxvi. 

2  Ann.  d.  1'Institut  Pasteur,  1907. 


254  PROTEIN  POISONS 

We  have  tried  to  extract  the  protein  poison  from  Witte's 
peptone  by  long-continued  shaking  with  absolute  alcohol, 
but  with  only  negative  results.  It  is  probable  that  the 
poison  as  it  exists  in  peptone  is  in  the  form  of  a  larger 
molecule  than  is  split  off  by  our  method,  and  consequently 
is  not  soluble  in  absolute  alcohol. 

Passive  Anaphylaxis. — The  serum  of  a  sensitized  animal 
introduced  into  a  fresh  animal  renders  the  latter  suscep- 
tible. The  second  animal  may  be  of  the  same  or  another 
species.  In  the  former  case  the  condition  induced  by  the 
transference  of  the  serum  is  known  as  homologous,  and  in 
the  second  as  heterologous  passive  anaphylaxis.  On  account 
of  the  ease  and  completeness  with  which  it  is  sensitized  the 
guinea-pig  is  most  frequently  the  recipient,  whatever  be 
the  species  of  the  donor.  The  transfer  of  the  condition  of 
sensitization  from  the  mother  to  her  offspring  is  an  illus- 
tration of  homologous  passive  anaphylaxis.  This  has  been 
studied  especially  by  Rosenau  and  Anderson,  Gay  and 
Southard,  and  Otto.  The  latter  has  found  the  young 
sensitive  at  forty-four  days  after  birth.  Gay  and  Southard1 
were  the  first  to  demonstrate  experimental  passive  anaphyl- 
axis. These  investigators  found  their  recipients  first  sensi- 
tive on  the  fourteenth  day.  This  indicated  a  somewhat 
long  period  of  incubation  for  the  development  of  the  ana- 
phylactic  state  in  the  recipient,  and  this  was  not  easily 
explainable.  Otto  and  Friedemann2  injected  the  anaphyl- 
actic  serum  subcutaneously  and  the  antigen  intraperi- 
toneally  twenty-four  hours  later.  With  shorter  intervals 
they  failed  to  obtain  any  response.  Braun3  was  the  first 
to  inject  the  anaphylactic  serum  intravenously,  but  even 
with  this  method  a  short  period  of  incubation  seemed  to  be 
necessary.  By  injecting  both  sera  intravenously,  Doerr 
and  Russ4  cut  down  the  supposed  period  of  incubation  to 
four  hours,  but  their  most  striking  and  constant  results 

1  Jour.  Med.  Research,  1907,  xvi,  143. 

2  Munch,  med.  Woch.,  1907. 

3  Zeitsch.  f.  Immunitatsforschung,  1910,  iv. 
« Ibid.,  iii,  181. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     255 

at  that  time  were  obtained  by  injecting  the  anaphylactic 
serum  intraperitoneally  and  the  antigen  intravenously, 
twenty-four  hours  later.  Later  it  was  shown  by  Doerr 
and  Russ,  also  by  Biedl  and  Kraus,  that  acute  symptoms, 
with  death,  may  result  from  the  simultaneous  intravenous 
injection  of  anaphylactic  serum  and  antigen.  There  is 
therefore  no  period  of  incubation  in  passive  anaphylaxis, 
and  this  condition  loses  much  of  the  theoretical  importance 
which  was  attached  to  it  so  long  as  it  seemed  to  require  a 
period  of  incubation  for  its  development.  As  we  shall  see 
later,  the  serum  and  organ  extracts  of  sensitized  animals 
mixed  in  vitro  in  proper  proportions  with  the  antigen  produce 
a  poison  which  kills  fresh  animals  in  anaphylactic  shock. 

It  has  been  found  by  Gay  and  Southard  that  0.1  c.c.  of 
serum  suffices  to  render  the  recipient  passively  anaphylactic. 

In  reference  to  passive  homologous  anaphylaxis  in  rabbits, 
Friedemann1  makes  a  statement  of  which  the  following  is 
a  summary:  Among  the  old  authorities,  Weichard  with 
a  mixture  of  placental  cells  and  the  antiserum,  Batelli 
with  laked  blood  and  the  corresponding  hemolysin,  and 
Nicolle  with  horse  serum  and  anti-horse  serum,  succeeded. 
More  recent  authorities2  have  failed  to  secure  passive 
homologous  anaphylaxis  in  rabbits.  The  divergencies  are 
probably  explained  by  Friedemann,  who  recommends  the 
following:  (1)  Antigen  and  antiserum  should  be  injected 
intravenously  and  simultaneously.  When  the  antigen  is 
introduced  twenty-four  hours  after  the  serum  there  is  no 
marked  reaction.  (2)  There  is  an  optimum  proportion 
between  anti-serum  and  antigen.  By  employing  2.5  c.c. 
of  anti-serum,  Friedemann  obtained  no  results  when  the 
antigen  varied  from  2.5  to  0.25  c.c.,  but  did  obtain  positive 
effects  when  the  amount  of  antigen  was  reduced  to  from 
0.025  to  0.0025  c.c. 

This  corresponds  exactly  with  our  researches  on  anaphyl- 
axis in  vitro  (see  p.  274). 


1  Jahresbericht  u.  d.  Ergeb.  d.  Immunitatsforschung,  1910,  vi,  67. 

2  Braun,  Kraus,  and  others. 


256  PROTEIN  POISONS 

Doerr  and  Russ1  attempted  to  measure  the  antibody 
in  sera  by  the  following  methods:  (1)  A  series  of  guinea- 
pigs  received  intraperitoneally  1  c.c.  of  antiserum  and 
twenty-four  hours  later  decreasing  amounts  of  the  antigen 
intravenously.  By  this  method  they  determined  the 
smallest  amount  of  antigen  necessary  to  induce  sudden 
death.  (2)  A  series  of  guinea-pigs  received  intraperitoneally 
decreasing  doses  of  the  antiserum,  and  twenty-four  hours 
later  a  constant  dose  (0.01  to  1  c.c.)  of  the  antigen  intra- 
venously. In  this  way  they  determined  the  smallest  amount 
of  serum  necessary  to  induce  sudden  death.  In  our  opinion 
these  experiments,  while  of  great  value,  do  not  give  results 
which  justify  standardization  of  the  so-called  antibodies. 
Anaphylactic  shock  is  not  determined,  at  least  wholly,  by 
the  amount  of  antigen  given,  nor  yet  by  the  amount  of 
antibody  in  the  animal,  but  by  the  proportion  between 
the  two. 

We  have  stated  that  Gay  and  Southard  were  the  first 
to  discover  passive  anaphylaxis,  and  that  their  work  was 
done  with  guinea-pigs  as  both  donors  and  recipients.  Our 
French  confreres  generally  credit  Maurice  Nicolle2  with 
this  discovery,  and  in  his  work  rabbits  served  both  as 
donors  and  recipients.  It  seems  that  the  work  of  Nicolle 
was  done  before  that  of  Gay  and  Southard,  but  not  pub- 
lished until  after  the  work  of  the  Americans  appeared  in 
print.  This  is  the  statement  made  by  Levaditi.3  However, 
the  work  done  in  one  country  was  quite  independent  of  that 
done  in  the  other,  and  Nicolle  showed  that  rabbits  which 
were  being  treated  daily  by  intraperitoneal  or  intravenous 
injections  of  1  c.c.  of  horse  serum  furnished  a  serum  which, 
when  injected  into  the  peritoneal  cavity  of  a  fresh  rabbit, 
sensitized  the  latter,  as  was  demonstrated  by  the  subcu- 
taneous injection  twenty-four  hours  later  of  1  c.c.  of  horse 
serum,  this  injection  giving  rise  to  an  inflammatory  edema — 
the  Arthus  phenomenon  (see  p.  262).  When  the  reinjection 

1  Zeitsch.  f.  Immunitatsforschung,  iii,  181. 

2  Ann.  de  1'Institut  Pasteur,  1907,  xxi,  128. 

3  Jahresb.  u.  d.  Ergeb.  d.  Immunitatsforschung,  1907,  iii,  40. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      257 

was  made  into  the  brain,  sudden  death  resulted.  It  will 
be  seen  from  the  above  that  passive  anaphylaxis  may  be 
demonstrated  in  the  recipient  not  only  by  general  reac- 
tion, as  anaphylactic  shock,  but  by  local  reaction  also. 
Even  before  the  work  of  either  Gay  and  Southard,  or  that 
of  Nicolle,  v.  Pirquet  and  Shick1  had,  by  a  reversed  method, 
demonstrated  passive  anaphylaxis.  To  each  of  three  rabbits 
they  administered  10  c.c.  of  horse  serum;  twenty-four  hours 
later,  two  received  each  2  c.c.  of  rabbit-antihorse  serum, 
and  1  c.c.  of  rabbit  serum,  all  subcutaneously  in  the  ear. 
In  the  first  two,  edema  resulted,  while  in  the  third  there 
was  no  effect.  By  this  reversed  method  anaphylactic 
shock  may  be  induced.  Pick  and  Yamanouchi2  injected 
2  c.c.  of  ox-serum  subcutaneously  in  young  rabbits  of  about 
700  grams  weight,  and  14  days  later  5  c.c.  of  rabbit  anti-ox 
serum  intravenously,  causing  anaphylactic  shock  and  death. 

Weil-Halle  and  Lemain3  have  observed  both  local  and 
general  symptoms  of  anaphylaxis  in  both  guinea-pigs  and 
rabbits  when  simultaneously  on  one  side  rabbit-antihorse 
serum  and  on  the  other  normal  horse  serum,  were  injected. 
A  few  hours  after  such  injections,  the  tissue  around  the 
point  of  injection  of  the  antiserum  becomes  edematous, 
infiltrates,  and  becomes  necrotic.  Either  speedy  death 
follows,  or  the  animal  becomes  cachectic  and  dies  after 
two  or  three  weeks. 

Kraus,  Doerr,  and  Sohma4  found  that  the  blood-serum 
of  rabbits  sensitized  to  the  proteins  of  the  crystalline  lens, 
renders  the  recipients  anaphylactic  to  the  same  proteins, 
and  that  this  sensitization  is  strictly  specific,  and  Doerr  and 
Kraus5  have  made  a  similar  showing  in  bacterial  anaphyl- 
axis. Furthermore,  Richet6  has  made  like  demonstrations 
with  the  serum  of  animals  sensitized  to  mytilocongestion 
and  crepitin. 

1  Die  Serumkrankheit,  1906. 

2  Zeitsch.  f.  Immunitatsforschung,  i,  676. 

3  Compt.  rend,  de  la  Soc.  biol.,  1907. 
«  Wien.  klin.  Woch.,  1908,  No.  30. 

5  Ibid.,  No.  28. 

•  Ann.  de  1'Institut  Pasteur,  1907-08. 
17 


258  PROTEIN  POISONS 

The  following  additional  facts  concerning  passive  ana- 
phylaxis  are  of  importance  and  will  be  referred  to  again 
when  we  come  to  discuss  the  theories  of  anaphylaxis: 

1.  Otto1  has   shown   that   blood-serum    taken   from    an 
animal  in  the  pre-anaphylactic  stage  (eight  days  after  the 
first  injection  of  horse  serum)  sensitizes  recipients. 

2.  Gay  and  Southard  and  others  have  shown  that  the 
blood-serum   taken   from    animals    in    the    so-called   anti- 
anaphylactic  state  sensitizes  the  recipients. 

3.  Otto  has  shown  that  a  third  species  of  animal  may 
be  used   to  demonstrate  anaphylaxis;  thus,   a  guinea-pig 
which  has  been  treated  with  rabbit-antihorse  serum  mani- 
fests  anaphylactic   symptoms  when   subsequently  treated 
with  horse  serum.     We2  have  shown  that  under  proper 
conditions    the    anaphylactic    poison    may    be    generated 
in  vitro. 

Besredka3  in  discussing  passive  anaphylaxis  states  that 
results  are  inconstant.  In  our  opinion  this  is  due  to  the 
difficulty  in  securing  the  proper  proportion  between  the 
antigen  and  the  anaphylactic  serum  introduced  into  the 
recipient. 

Anti-anaphylaxis. — We  wish  to  join  Friedemann4  in  a 
protest  against  the  use  of  this  term.  However,  as  Friede- 
mann states,  it  is  so  deeply  embedded  in  the  literature  of 
anaphylaxis  that  it  cannot  be  omitted.  German  and 
American  authors  are  not  the  only  ones  who  object  to  the 
term  anti-anaphylaxis,  and  the  explanation  of  it  given  by 
Besredka  and  Steinhardt.  Levaditi5  has  produced  potent 
arguments  against  the  views  of  his  confreres.  The  work 
done  by  Besredka  and  Steinhardt  is  of  the  highest  value, 
but  we  differ  from  them  in  the  explanation  of  their  results. 
As  Levaditi  states,  we  owe  the  discovery  of  this  condition 
to  Otto  and  to  Rosenau  and  Anderson,  but  the  name  and 

1  Munch,  med.  Woch.,  1907,  No.  39. 

2  Zeitschr.  f.  Immunitatsforschung,  xi,  673. 

3  Kraus  and  Levaditi,  Handbuch  d.  Technik  u.  Methodik  d.  Immuni- 
tatsforschung, Erganzungsband  i,  240. 

4  Jahresbericht  u.  d.  Ergeb.  d.  Immunitatsforschung,  1910,  vi,  54. 
*  Ibid.,  iii,  37. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     259 

the  most  thorough  study  of  it  have  come  from  the  researches 
of  Besredka  and  Steinhardt. 

Theobald  Smith1  stated  that  the  guinea-pigs  which  had 
received  the  largest  doses  of  diphtheria  toxin-antitoxin 
mixture  more  frequently  survived  the  second  dose  than 
those  that  received  smaller  doses.  Rosenau  and  Anderson2 
found  that  animals  to  which  they  gave  reinjections  before 
the  end  of  the  period  of  incubation  (before  twelve  days) 
were  not  responsive  when  another  reinjection  was  made 
at  the  end  of  twelve  days,  and  that  a  longer  period  than 
another  twelve  days  had  to  pass  before  they  became  respon- 
sive. All  who  have  tested  this  point  in  serum  anaphylaxis 
have  found  that  the  larger  the  sensitizing  doses  employed, 
the  longer  must  be  the  period  before  typical  anaphylactic 
shock  appears  on  reinjection.  Moreover,  all  who  have 
experimented  with  anaphylactic  shock,  whatever  the  antigen 
employed,  and  whatever  the  avenue  of  administration, 
have  found  that  animals  which  survive  the  first  reinjection 
are  for  a  time  thereafter,  which  is  variable,  refractory  to  a 
second  reinjection.  Besredka  and  Steinhardt  observed 
that  it  was  easy  to  develop  the  refractory  state  in  guinea-pigs 
by  either  of  the  following  methods:  (1)  The  intracerebral 
injection  of  0.25  c.c.  of  horse  serum  before  the  expiration 
of  the  period  of  incubation  (twelve  days).  (2)  The  intra- 
cerebral injection  of  less  than  the  fatal  dose  (^5-  to  -f^  c.c.) 
after  the  period  of  incubation.  (3)  The  intraperitoneal 
injection  of  5  c.c.,  after  the  period  of  incubation.  The  last 
method  seems  to  apply  only  to  French  guinea-pigs,  which, 
as  we  have  already  stated,  are  not  so  easily  sensitized  as 
those  of  other  countries.  (4)  Rectal  injections.  The  rectum 
is  cleansed  with  a  glycerin-enema  and  then  10  c.c.  of  a 
dilution  of  serum  with  an  equal  volume  of  normal  salt 
solution  is  injected.  This  never  affects  sensitized  guinea- 
pigs,  and  after  twelve  hours  they  generally  prove  refractory 
to  poisonous  doses  given  intracerebrally.  Besredka3  secures 

1  Jour.  med.  Research,  1904,  xxii,  3. 

2  Hygienic  Laboratory  Bulletin,  1906.  No.  29. 

3  Kraus  and   Levaditi,   Handbuch   d.   Ergeb.   d.   Immunitatsforschung, 
Erganzungsband,  i. 


260  PROTEIN  POISONS 

the  refractory  state  in  sensitized  guinea-pigs  by  these 
methods.  He  accomplishes  a  similar  result  by  giving  the 
reinjection  while  the  animal  is  deeply  narcotized  with  either 
ether  or  alcohol.  He  finds  that  in  this  state  many  animals 
survive  the  reinjection  made  at  any  time  and  by  any 
method,  and  that  after  recovery  they  are  completely,  but 
only  temporarily,  refractory.  Of  these  methods  he  prefers 
the  rectal  injection,  or  better  still,  the  subcutaneous  injec- 
tion of  a  less  than  fatal  dose.  For  the  reinjection  he  prefers 
the  cerebral  method.  Since  these  investigations,  made  by 
Besredka,  are  of  the  highest  importance  both  theoretically 
and  practically,  we  must  study  them  more  in  detail.  They 
are  of  theoretical  importance  because  he  holds  that  mixed 
proteins  contain  not  only  sensitizing  and  toxic  substances, 
but  also  a  vaccinating  body  and  the  last  mentioned  of 
these  he  claims  vaccinates  sensitized  animals,  thus  ren- 
dering them  immune  to  reinjections.  This  part  of  his 
study  he  has  carried  out  most  thoroughly  with  milk,  and 
along  this  line  we  will  follow  him.  Milk  heated  for  fifteen 
minutes  at«120°  still  sensitizes  and  kills  on  reinjection,  but 
when  heated  for  fifteen  minutes  at  130°  it  neither  sensitizes 
nor  kills  sensitized  animals  on  reinjection,  but  does  render 
sensitized  animals  refractory  to  reinjections  of  milk  heated 
to  only  120°.  From  these  results  he  concludes:  (1)  That 
the  sensitizing  and  toxic  components  of  milk  behave  alike 
under  the  influence  of  the  temperatures  mentioned.  (2)  That 
the  vaccinating  component  can  be  separated  from  the  other 
two.  In  other  words,  the  milk  heated  to  130°  vaccinates  or 
immunizes  against  anaphylaxis,  while  it  can  neither  sensi- 
tize fresh  animals  nor  kill  sensitized  ones.  In  order  to 
determine  the  nature  of  the  vaccinating  substance  he  coag- 
ulates the  milk  with  Bulgarian  lactoferment  and  separates 
the  coagulum  from  the  whey  by  the  centrifuge  or  by  filtra- 
tion through  paper.  The  whey  immunizes  sensitized 
animals,  but  is  without  toxic  action.  The  whey  is  neutral- 
ized with  soda  and  the  flocculent  precipitate  which  forms 
is  separated  from  the  supernatant  fluid  by  decantation  of 
the  latter,  and  then  made  into  a  gelatinous  mixture  with 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     261 

salt  solution.  This  mixture  renders  sensitized  animals 
refractory,  and  Besredka  finally  concludes  that  the  vacci- 
nating constituent  of  milk  is  lactoprotein,  which  is  not 
destroyed  by  heating  to  130°  nor  removed  from  solution 
by  coagulation  with  the  Bulgarian  ferment.  Against  this 
conclusion  we  must  call  attention  to  the  following  facts 
demonstrated  by  Besredka  himself:  (1)  Milk  or  serum 
introduced  into  the  stomach,  rectum,  or  peritoneal  cavity 
of  sensitized  animals,  renders  them  refractory  to  a  cerebral 
reinjection.  (2)  Small,  non-fatal,  doses  of  milk  or  serum 
given  subcutaneously  or  in  any  other  way  has  a  like  effect. 
Why,  therefore,  is  it  not  reasonable  to  say  that  the  vacci- 
nating property  of  the  whey  or  of  the  precipitate  obtained 
from  it  is  not  due  to  the  small  amount  of  casein  which  it 
undeniably  contains?  Especially  is  this  query  pertinent, 
since  as  Besredka  states,  certain  food  authorities  hold 
that  casein  is  the  only  protein  in  milk.  We  must  conclude 
that  while  Besredka's  work  along  this  line  is  most  interest- 
ing and  valuable,  he  has  failed  to  prove  the  existence  of  a 
vaccinating  component  in  milk. 

Besredka's  work  on  the  refractory  state,  which  he  calls 
anti-anaphylaxis,  is  of  practical  value  in  pointing  out  a 
possible  way  by  which  sensitized  individuals  may  be  saved 
from  anaphylactic  shock  in  the  therapeutic  administration 
of  sera.  We  will  return  to  this  point  later  (see  Chapter  XV). 

We  wish  now  to  inquire  whether  there  is  any  con- 
dition that  may  properly  be  designated  as  anti-anaphyl- 
axis. In  discussing  passive  anaphylaxis  we  have  seen  that 
serum  taken  from  animals  while  in  the  refractory  state, 
whether  it  be  before  the  complete  development  of  sensiti- 
zation  or  after  recovery  from  a  non-fatal  reinjection,  and 
transferred  to  normal  animals  renders  the  recipient  sus- 
ceptible. It  hardly  seems  proper  to  say  that  an  animal 
has  been  desensitized  when  its  blood-serum  has  this  effect. 
If  the  blood-serum  of  refractory  animals  rendered  sensitized 
animals  refractory,  the  term  anti-anaphylaxis  might  be 
proper,  but  this  is  in  no  case  true.  The  refractory  animal 
is  still  sensitized,  but  the  degree  of  sensitization  has  been 


262  PROTEIN  POISONS 

so  lowered  that  it  no  longer  manifests  itself  in  anaphylactic 
shock  when  a  reinjection  is  made.  Moreover,  in  all  instances 
in  which  it  has  been  tested,  the  refractory  state  is  only 
temporary,  and  sooner  or  later  the  sensitized  condition  is 
sufficiently  restored  to  be  recognizable  by  anaphylactic 
shock.  We  will  take  up  this  point  again  when  we  discuss 
the  theories  of  anaphylaxis. 

The  Arthus  Phenomenon. — This  phenomenon,  already 
referred  to,  deserves  more  detailed  study.  Arthus1  observed 
that  a  single  injection  of  horse  serum  into  rabbits,  whether 
the  amount  was  small  or  large,  whether  the  injection  was 
made  subcutaneously,  intraperitoneally,  or  intravenously, 
whether  the  serum  was  unheated  or  heated  to  57°,  had  no 
recognizable  effect  at  the  time  or  later,  but  that  repeated 
injections  were  followed  by  certain  constant  results.  When 
daily  subcutaneous  injections  of  5  c.c.  were  made,  the 
following  effects  were  observed:  After  the  first  three  injec- 
tions the  serum  was  readily  absorbed  in  a  few  hours;  after 
the  fourth  there  appeared  a  soft  infiltration  about  the 
point  and  this  disappeared  after  two  or  three  days;  after 
the  fifth  the  infiltration,  which  appeared,  was  hard,  edema- 
tous,  and  required  five  or  six  days  for  its  absorption;  after 
the  sixth  there  appeared  a  hard,  compact,  aseptic  mass 
which  remained  unchanged  for  weeks;  after  the  seventh, 
there  was  the  same  condition  much  accentuated.  The 
skin  over  the  swelling  became  red,  then  whitish  and  dry; 
the  tissue  became  gangrenous  and  finally  dropped  out, 
leaving  a  deep  wound  which  slowly  contracted  into  a  scar. 
This  local  reaction  became  more  marked,  and  more  extended, 
with  further  repetitions,  of  the  injections.  This  reaction  is 
strictly  specific,  inasmuch  as  animals  first  treated  with 
horse  serum  do  not  react  to  subsequent  treatment  with 
other  sera  or  with  milk.  Subsequent  studies  show  that 
this  reaction  can  be  obtained  in  the  same  way  in  guinea- 
pigs,  rats,  and  pigeons.  It  is  not  necessary  that  the  sensi- 
tizing and  the  reinjections  be  made  in  the  same  way.  If 

1  Compt.  rend,  de  la  Soc.  biol.,  1903,  817. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     263 

the  former  be  subcutaneous,  the  latter  may  be  intraperi- 
toneal  or  intravenous,  or  vice  versa.  However,  when  the 
animal  has  been  sensitized  by  several  (six  to  eight)  subcu- 
taneous or  intraperitoneal  injections,  and  the  reinjection 
consisting  of  2  c.c.  is  made  into  the  vein,  the  animal,  in 
many  instances,  dies  from  anaphylactic  shock.  Imme- 
diately after  the  reinjection  it  shakes  its  head,  and  evidently 
becomes  anxious.  Breathing  is  frequent,  polypnea  reaching 
sometimes  as  high  a,s  200  to  250  respirations  per  minute. 
There  is  expulsion  of  stool.  Immobility,  apnea,  and  exoph- 
thalmia  appear  and  the  animal  may  die  within  three  minutes 
after  receiving  the  reinjection.  Obduction  shows  the  heart 
in  systole  and  the  blood  fluid. 

The  rabbits  which  recover  after  manifesting  disturbances 
of  respiration  pass  into  a  long-continued  cachectic  state. 
Rabbits  which  have  received  daily  intravenous  injections 
and  then,  after  a  period  of  time  receive  the  reinjections 
by  the  same  route,  generally,  but  not  invariably,  die  in 
anaphylactic  shock.  Sensitization  by  repeated  intravenous 
injections  and  subcutaneous  reinjections,  invariably  result 
in  the  production  of  the  Arthus  phenomenon. 

Nicolle1  found  that  rabbits  in  which  Arthus'  phenomenon 
has  been  developed  become  highly  receptive  to,  and  readily 
succumb  to,  infection.  The  same  investigator  produced 
the  Arthus  reaction  in  guinea-pigs,  though  he  found  these 
animals  less  suitable  than  rabbits  for  the  demonstration 
of  this  form  of  anaphylaxis.  Lewis2  also  demonstrated 
in  guinea-pigs  the  phenomenon  of  Arthus.  Remlinger3  has 
done  similar  work.  He  injected  massive  doses  (10  c.c.) 
of  horse  serum,  at  intervals  of  one  week,  from  six  to  eight 
times,  and  one  month  after  the  last,  gave  a  reinjection. 
On  doing  this  his  animals  either  developed  general  symp- 
toms or  passed  into  a  cachectic  condition  and  died.  Vaughan 
and  Wheeler4  killed  guinea-pigs  by  daily  intraperitoneal 
injections  of  egg-white. 

1  Ann.  d.  1'Institut  Pasteur,  1907,  xxi,  128. 

2  Jour.    Exper.  Med.,  1908. 

3  Compt.  rend,  de  la  Soc.  biol.,  1907,  Ixii,  23. 

4  Jour.  Infect.  Dis.,  June,  1907. 


264  PROTEIN  POISONS 

Arthus  and  Brim1  studied  the  microscopic  changes  in 
the  tissues  in  this  reaction.  A  fluid  containing  a  few  poly- 
nuclear  leukocytes  first  infiltrates  the  subcutaneous  tissue. 
Later,  the  infiltration  approaches  the  surface,  and  forms  a 
line  of  cleavage  between  the  stratum  corneum  and  the 
stratum  lucidum.  The  subcutaneous  connective  tissue  is 
converted  into  a  homogeneous  mass,  and  there  are  extrava- 
sations of  blood.  Finally,  the  process  leads  to  necrosis  with 
a  sharp  line  of  demarcation.  "It  is  an  aseptic  necrosis 
which  first  involves  the  connective  tissue  and  vessels,  and 
finally  the  epidermis." 

From  a  study  of  the  Arthus  phenomenon  we  draw  two 
conclusions:  (1)  A  prolonged  period  of  incubation  is  not 
necessary  in  order  to  induce  the  anaphylactic  state.  Such 
a  period  is  necessary  in  order  to  secure  the  explosive  mani- 
festation of  anaphylaxis,  but  the  development  of  the 
specific  "antibody"  begins  soon  after  the  first  injection  of 
the  anaphylactic  protein.  (2)  There  is  no  such  thing  as  a 
condition  of  antianaphylaxis.  If  there  were,  certainly 
animals  which  are  receiving  daily  injections  should  manifest 
it,  but  the  only  effect  is  to  suppress  the  explosive  character 
of  anaphylaxis.  We  will  return  to  these  questions  when  we 
take  up  the  theories. 

Anaphylaxis  and  Toxic  Sera. — It  is  well  known  that  a 
single  injection,  even  in  very  small  amount,  of  certain 
sera  into  animals  of  another  species  proves  fatal.  One  of 
the  most  highly  poisonous  sera  is  that  of  the  eel,  a  very 
minute  quantity  of  which  injected  into  a  guinea-pig  causes 
death.  Doerr  and  Raubitschek2  have  studied  the  toxic 
and  anaphylactic  effects  of  eel  serum  on  guinea-pigs.  They 
find  that  heating  this  serum  to  58°  destroys  its  toxic  action. 
A  single  dose  of  this  heated  serum  has  no  apparent  effect 
upon  guinea-pigs,  but  does  sensitize  them  so  that  a  second 
dose  of  the  same  is  followed  by  anaphylactic  shock.  This 
demonstrates  that  the  toxin  and  the  anaphylactogen  of  eel 


1  Compt.  rend,  de  la  Soc.  biol.,  1903,  1478. 

2  Berl.  klin.  Woch.,  1908,  No.  33. 


PROTEIN  SENSITIZATION  OR  AN APR YL AXIS     265 

serum  are  distinct  substances,  the  former  being  thermo- 
labile  and  the  latter  thermostabile.  The  same  investigators 
demonstrated  the  same  thing  in  another  way.  The  toxin 
of  eel  serum  is  destroyed  by  acidifying  the  serum  with 
hydrochloric  acid,  and  is  not  restored  on  neutralization 
(differing  in  this  last  respect  from  certain  other  toxins, 
such  as  those  of  cobra  poison  and  of  diphtheria).  Eel 
serum  when  acidified  with  from  0.4  to  1  per  cent,  of  hydro- 
chloric acid  and  then  neutralized  has  no  poisonous  action 
in  a  single  dose  on  guinea-pigs,  but  does  sensitize  them  to  a 
second  dose  of  the  serum  treated  in  the  same  way.  Pre- 
cipitation of  eel  serum  by  saturation  with  ammonium 
sulphate  carries  down  both  the  toxin  and  the  anaphylac- 
togen.  Ox  serum  behaves  in  a  similar  way  with  eel  serum 
on  guinea-pigs,  and  it  also  is  robbed  of  its  toxic  property 
when  heated  to  60°.  The  blood  serum  of  guinea-pigs 
treated  with  unheated  eel  serum  contains  both  antitoxin  and 
the  substance  produced  by  the  anaphylactogen,  and  with 
this  serum  fresh  animals  may  be  protected  against  unheated 
eel  serum  and  anaphylactized  to  heated  serum. 

It  follows  from  these  researches  that  the  substance 
elaborated  in  the  organism  by  an  anaphylactogen  is  not 
an  antitoxin.  This  does  not  mean  that  the  animal  which 
dies  from  the  first  dose  of  a  toxic  serum  and  the  one  that 
dies  from  the  second  dose  of  a  heated  serum  do  not  die 
from  the  effects  of  the  same  poison.  The  toxic  serum  owes 
its  toxicity  to  a  ferment  which  splits  up  the  proteins  of 
the  animal's  body,  setting  a  poison  free.  The  unheated 
serum  leads  to  the  elaboration  in  the  animal's  body  of  a 
ferment  which  splits  up  the  protein  of  the  heated  serum 
on  the  second  injection,  setting  a  poison  free.  With  the 
toxic  serum  the  ferment  is  introduced  into  the  guinea-pig, 
and  splits  up  the  proteins  of  the  body.  When  the  heated 
serum  is  injected  the  cells  of  the  guinea-pig  elaborate  a 
ferment  which  splits  up  the  proteins  of  the  heated  serum 
on  its  second  injection.  In  both  instances  the  poison  is 
generated  by  the  parenteral  digestion  of  proteins,  and  in 
all  probability  is  the  same. 


266  PROTEIN  POISONS 

The  Toxogens. — There  has  been  marked  diversity  of 
opinion  concerning  the  nature  of  the  substance  developed 
in  the  body  under  the  influence  of  the  anaphylactogen. 
Most  German  authorities,  following  the  nomenclature 
introduced  by  Ehrlich  in  his  masterly  studies  of  toxins  and 
antitoxins,  have  designated  the  sensitizing  proteins  as 
antigen  and  the  substance  elaborated  in  the  organism  as 
antibody.  It  is  evident  that  if  antigen  be  appropriate  for 
the  sensitizing  substance,  the  substance  produced  under 
its  influence  must  be  the  "antibody."  We  have  already 
expressed  our  opinion  concerning  the  unfitness  of  the 
word  "antigen."  The  inappropriateness  of  the  term 
"antibody"  is  equally  evident.  The  anaphylactogen, 
instead  of  rendering  the  organism  resistant  to  subsequent 
injections,  renders  it  more  sensitive.  It  is  true,  as  we  shall 
see  later,  that  this  increased  sensitiveness  may  be  a  most 
delicate  and  efficient  means  of  subsequent  protection. 
It  sharpens  the  agents  of  defence,  but  it  does  not  blunt 
the  implements  of  attack.  It  prepares  the  body  cells  for 
subsequent  contests,  but  it  does  not  disarm  the  invader. 
It  places  in  the  hands  of  the  defender  more  efficient  means 
of  warfare,  but  it  does  not  impair  the  equipment  of  the 
attacking  force.  It  is  not  a  shield  for  protection,  but  a 
sharpened  sword  for  battle. 

The  blood  serum  of  an  animal  which  has  been  treated 
with  a  toxin,  mixed  in  vitro  with  the  toxin  in  proper  propor- 
tion, may  be  injected  into  a  fresh  animal  without  effect. 
The  blood  serum  of  an  animal  treated  with  an  anaphylac- 
togen, mixed  in  vitro  in  proper  proportion  with  the  anaphyl- 
actogen and  injected  into  a  fresh  animal  kills  it.  Surely 
there  is  no  justification  for  the  use  of  the  terms  "antigen" 
and  "antibody"  in  explaining  the  phenomena  of  sensiti- 
zation.  Moreover,  their  employment  confuses  and  misleads, 
and  in  our  opinion  they  should  be  discarded.  However,  like 
many  other  terms  improperly  used  in  scientific  research, 
they  have  become  so  deeply  engrafted  into  the  literature 
that  they  cannot  be  eliminated,  but  their  inappropriateness 
should  be  clearly  understood. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      267 

Not  all  the  German  authorities  have  used  the  term 
"antibody"  in  indicating  the  substance  elaborated  in  the 
organism  in  the  development  of  the  anaphylactic  state. 
Otto  calls  it  the  "reaction-body,"  and  to  this  there  can  be 
no  objection.  V.  Pirquet,  as  we  have  seen,  used  "allergy," 
meaning  altered  reaction  instead  of  anaphylaxis,  and 
"allergin"  for  the  substance  which  reacts  with  the  foreign 
protein  on  reinjection.  Besredka  calls  the  sensitizing  agent 
sensibilisinogen,  and  the  substance  developed  under  its 
influence,  sensibilisin.  Nicolle  uses  the  term,  albumino- 
lysin  and  Richet  the  word  toxogen.  All  of  these  are  free 
from  the  objections  which  we  have  urged  to  the  term 
"antibody."  We  have  adopted  Richet's  term,  but  this 
does  not  imply  condemnation  of  the  others.  Indeed,  the 
word  "toxogen"  needs  some  explanation  in  order  to  prevent 
error  following  its  use.  As  we  shall  see  later,  the  anaphyl- 
actic poison  is  not  a  toxin.  The  word  "toxogen"  is  used 
by  us  as  meaning  a  generator  of  poisons,  and  these  poisons 
are  not  toxins,  inasmuch  as  they  do  not  lead  to  the  elabora- 
tion of  antitoxins  when  introduced  into  the  animal  body. 
The  toxogen  is  a  ferment. 

Pfeiffer1  long  before  the  word  anaphylaxis  had  been 
coined  really  discovered  the  fundamental  fact  which  later 
research  has  confirmed.  This  is  known  as  Pfeiffer's  phe- 
nomenon. He  found  that  when  cholera  vibrios  are  injected 
into  the  abdominal  cavity  of  a  guinea-pig,  which  has  pre- 
viously been  immunized  by  repeated  injections  of  non-fatal 
doses  of  the  living  culture,  they  are  dissolved  like  sugar 
or  salt  in  water.  This  destruction  of  the  vibrios  can  be 
demonstrated  by  microscopic  study,  but  notwithstanding 
the  destruction  of  the  bacteria,  the  animal  is  poisoned,  and 
dies.  In  fact  the  more  powerful  the  lytic  serum  and  the 
more  rapid  and  complete  the  destruction  of  the  bacteria, 
the  more  certain  and  prompt  is  death.  Later,  it  was  shown 
by  Bordet  that  with  fresh  lytic  serum  the  vibrios  may  be 
dissolved  in  vitro.  Furthermore,  it  was  shown  by  Pfeiffer 

1  Zeitsch.  f.  Hygiene,  1903. 


268  PROTEIN  POISONS 

that  living  cultures  of  the  cholera,  typhoid,  colon,  and 
many  other  bacilli  secrete  no  toxin,  but  that  the  cellular 
proteins  of  these  organisms  are  themselves  poisonous. 
By  these  experiments  Pfeiffer  laid  the  foundation  of  our 
knowledge  of  lytic  immunity,  which,  as  we  shall  see,  is 
the  chief  protective  function  in  the  anaphylactic  state. 
The  anaphylactogen  which  he  used  was  the  cellular  protein 
of  the  cholera  bacillus.  This  caused  the  elaboration  of 
the  toxogen,  which  is  contained  in  his  lytic  serum,  and  this 
digesting  the  anaphylactogen  on  the  second  injection,  split 
it  up  with  the  liberation  of  the  poison.  From  these  researches 
Pfeiffer  developed  his  theory  of  endotoxins,  which  we  will 
discuss  later. 

The  next  important  work  done  along  this  line  was  that 
of  Weichardt.1  He  extracted  the  proteins  from  placental 
cells,  and  found  that  the  blood  serum  of  rabbits  which  had 
received  repeated  injections  of  such  extracts  when  mixed 
with  the  anaphylactogen  either  in  vitro  or  in  vivo,  produced 
a  poison  which  killed  rabbits  with  the  typical  symptoms  of 
anaphylactic  shock. 

The  toxogen  exists  in  the  blood  serum  and  in  the  tissues 
of  sensitized  animals,  and  with  the  former  it  may  be  trans- 
ferred to  normal  animals,  thus  establishing  passive  anaphyl- 
axis.  As  we  have  seen,  passive  anaphylaxis  may  be  induced 
in  either  homologous  or  heterologous  animals.  In  the  study 
of  anaphylactic  sera  one  observation  has,  in  our  opinion, 
led  several  authorities  astray.  It  has  been  found  that 
passive  anaphylaxis,  in  some  instances  at  least,  may  be 
induced  with  anaphylactic  serum,  either  unheated  or  heated 
(56°).  From  this  it  has  been  inferred  that  the  toxogen  is 
thermostabile.  In  fact,  the  toxogen  consists  of  amboceptor 
and  complement,  and  the  latter  is  destroyed  by  a  tempera- 
ture of  56°;  but  when  heated,  anaphylactic  serum  is  injected 
into  a  fresh  animal  the  recipient  does  or  may  furnish  the 
complement.  Whether  it  does  or  does  not,  determines  the 
degree  of  success  in  inducing  passive  anaphylaxis,  which, 
as  we  have  seen,  is  not  constantly  accomplished. 

1  Berl.  therap.  Woch.,  1903,  No.  1. 


Reference  has  already  been  made  to  the  endotoxin  theory 
of  Pfeiffer.  This  assumed  the  existence  of  a  preformed 
poisonous  body  in  the  cell,  and  cytolysis  resulted  in  setting 
it  free.  It  was  believed  to  be  an  intracellular  toxin,  an 
independent  and  separate  molecule,  and  not  a  group  in  a 
more  complex  molecule.  This  theory  was  applicable  only 
to  cellular  proteins.  An  endotoxin,  as  understood  by 
Pfeiffer,  could  not  exist  in  a  soluble  protein,  and  since  soluble 
proteins  are  most  efficient  as  anaphylactogens,  the  theo- 
retical endotoxin  cannot  be  the  anaphylactic  poison.  Indeed, 
there  is  now  no  reason  for  believing  in  the  existence  of  the 
endotoxin.  The  brilliant  work  of  Friedemann1  has  shown 
that  red  blood  corpuscles  may  be  dissolved  without  setting 
free  any  active  poison,  and,  on  the  other  hand,  the  poisonous 
group  from  the  hemoglobin  molecule  may  be  extracted 
without  dissolving  the  corpuscles.  Hemoglobin  is  not  an 
active  poison.2  Animals  are  not  affected  by  a  large  amount 
of  it  given  in  a  single  dose,  but  it  is  an  anaphylactogen 
which  means,  according  to  our  understanding  at  least, 
that  its  molecule  contains  a  poisonous  group  which  is 
liberated  on  reinjection  through  the  cleavage  action  of  the 
toxogen.  Friedemann  showed  'that  the  poisonous  group 
can  be  extracted  from  the  proteins  of  the  red  corpuscles 
without  dissolving  them.  He  used  3  c'.c.  of  a  heavy  sus- 
pension of  washed  ox-corpuscles.  To  this  he  added  an 
equal  volume  of  a  highly  active  anaphylactic  serum,  and 
after  a  time  separated  the  corpuscles  in  the  centrifuge. 
The  corpuscles  were  again  washed  and  incubated  for 
a  short  time  with  fresh  rabbit  serum,  then  placed  in  an 
ice-box,  then  centrifuged,  and  the  colorless  fluid  injected 
into  fresh  animals  induced  anaphylactic  shock.  By  this 
method  Friedemann  was  the  first  to  prepare  the  anaphyl- 
actic poison  in  vitro.  This  work  has  been  confirmed,  and 
it  has  been  shown  fully  that  the  formation  of  the  anaphyl- 
actic poison  is  quite  independent  of  hemolysis.  Thomsen3 

1  Zeitsch.  f.  Immunitatsforschung,  ii. 

2  This  is  Friedemann' s  statement,  not  ours.    We  have  found  hemoglobin 
quite  poisonous,  even  to  the  species  supplying  it. 

3  Zeitsch.  f.  Immunitatsforschung,  i,  741. 


270  PROTEIN  POISONS 

has  demonstrated  that  in  guinea-pigs  sensitized  with 
erythrocytes,  there  is  no  recognizable  hemolysis  on  reinjec- 
tion,  although  anaphylactic  shock  occurs.  It  is  generally 
believed,  as  first  taught  by  Bordet,  that  in  hemolysis  the 
stroma  only  is  involved.  When  unbroken  corpuscles  are 
used  the  anaphylactic  poison  may  come  from  either  the 
hemoglobin  or  the  stroma,  or  from  both.  We  have  anaphyl- 
actized  animals  with  hemoglobin  and  with  stroma.  The 
former  is  more  easily  done  on  account,  we  presume,  of  its 
more  ready  solubility.  We  have  found  the  stroma  difficult  to 
dissolve  without  using  so  much  alkali  that  the  preparation 
is  not  suitable  for  animal  injection,  and  suspensions,  as  a 
rule,  do  not  so  readily  sensitize  as  solutions.  Friedberger 
and  Vallardi1  have  found  that  only  by  having  stroma, 
amboceptor,  and  complement  in  proper  portions  can  they 
prepare  the  anaphylactic  poison,  an  excess  of  any  one  giving 
negative  results.  Moreover,  while  the  poison  is  quickly 
generated  under  proper  conditions  from  the  unbroken 
corpuscles  a  much  longer  time  is  required  when  the  stroma 
only  is  used.  It  has  been  shown  byNeufeld  and  Dold2 
that  the  anaphylactic  poison  can  be  extracted  from  bacteria 
without  cytolysis.  Furthermore,  they  have  found  that  the 
anaphylactic  poison  is  more  easily  extracted  from  those 
bacteria  which  are  least  susceptible  to  lytic  influences. 
For  instance,  the  pneumococcus  which  is  highly  resistent 
to  lytic  influences  easily  yields  its  anaphylactic  poison, 
even  at  0°,  while  no  poison  is  obtained  at  37°,  and  the 
cholera  bacillus,  which  is  highly  labile,  yields  the  poison 
with  more  difficulty.  Friedberger  and  Schiitze3  found  that 
the  tubercle  bacillus,  which  is  highly  resistent  to  lysis, 
readily  supplies  the  anaphylactic  poison.  We  have  shown 
that  the  tubercle  bacillus  from  which  the  protein  poison 
has  been  extracted,  leaves  a  residue  which  is  not  only  not 
poisonous,  but  sensitizes  fresh  animals,  and  this  has  been 
confirmed  by  the  later  researches  of  White  and  Avery.4 

1  Zeitsch.  f.  Immunitatsforschung,  vii,  94. 

2  Ueber  Bakterienempfindlichkeit  u.  ihre  Bedeutung  f.  Infektion. 
*.Berl.  klin.  Woch.,  1911,  No.  9. 

4  Jour.  Med.  Research,  1912,  xxvi,  317. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     271 

Friedberger's  excellent  work  on  the  extraction  of  the 
anaphylactic  poison  from  bacteria  shows  the  necessity  of 
attending  to  the  quantitative  proportions  between  the 
bacteria,  amboceptor,  and  complement,  also  that  the 
poison  may  be  destroyed  by  prolonged  digestion.  The 
amount  of  cellular  substance  necessary  to  supply  a  fatal 
dose  of  the  poison  is  smaller  than  the  lethal  dose  of  the 
living,  unbroken  cells.  This  confirms  our  work,  for  we 
demonstrated  some  years  ago  that  the  protein  poison  is 
only  a  part  of  the  larger  molecule  which  is  a  part  of  the 
cell.  The  greater  part  of  the  bacterial  cell  is,  after  the 
removal  of  the  poisonous  portion,  wholly  without  toxic 
action.  Neufeld  and  Dold  have  shown  that  there  is  no 
relation  between  the  amount  of  poison  in  a  given  bacillus 
and  its  pathogenic  action,  and  Friedberger  and  Goldschmidt1 
have  obtained  the  anaphylactic  poison  from  the  prodigiosus 
and  other  non-pathogenic  bacteria.  All  of  this  is  confirma- 
tory of  work  we  did  many  years  ago.  In  1902  we  published 
the  following  findings: 

THE  EFFECTS  OF  INTBAPEKITONEAL  INJECTIONS  OF  THE  AIR-DRIED  CELLS 
OF  THE  BACILLUS  PRODIGIOSUS  IN  GUINEA-PIGS 

No.  Weight  in  gm.       '  Dose  in  mg.  Result. 

1 260  50  + 

2 305  50  + 

3 287  20  + 

4  ......  272  10  + 

5 260  5  + 

6 270  3  + 

7 252  2 

8 252  1 

THE  EFFECTS  OF  INTRAPERITONEAL  INJECTIONS  OF  THE  AIR-DRIED 
CELLS  OF  THE  BACILLUS  VIOLACEUS  IN  GUINEA-PIGS 

No.  Weight  in  gm.  Dose  in  mg.  Result. 

1 220  30  + 

2 230  20  + 

3 255  15  + 

4 265  10  + 

5 210  5 

1  Zeitsch.  f.  Immunitiitsforschung,  vi,  299. 


272  PROTEIN  POISONS 

THE  EFFECTS  OF  INTRAPERITONEAL  INJECTIONS  OF  THE  AIR-DRIED  CELLS 
OF  SARCINA  AURANTIACA  IN  GUINEA-PIGS 

No.                               Weight  in  gm.  Dose  in  mg.  Result. 

1 240                             25  + 

2 300                             15  -;- 

3 305  10 

THE  EFFECTS  OF  INTRAPERITONEAL  INJECTIONS   OF  THE  FINELY-GROUND 
CELLS  OF  THE  COLON  BACILLUS  IN  GUINEA-PIGS 

No.                               Weight  in  gm.          Dose  in  mg.  Result. 

1 172  4.09  + 

2 170  4.05  + 

3 165  3.66  + 

4 195  2.60  + 

5 135  1.80  + 

6 145  1.45  + 

7 165  0.825  + 

8 200  0.10 

9 175  0.085 

10 162  0.081 

As  is  well  known,  rabbits  repeatedly  treated  with  some 
foreign  protein,  such  as  horse  serum,  furnish  a  serum 
which  precipitates  the  foreign  protein  in  vitro.  The  rabbit 
at  the  same  time,  and  by  the  same  treatment,  is  sensitized. 
Quite  naturally  one  suspects  that  toxogens  and  precipitins 
are  identical.  Friedemann  was  the  first  to  test  this  question 
experimentally.  He  mixed  the  blood  serum  of  sensitized 
rabbits  in  vitro,  in  varying  proportions  with  the  homologous 
anaphylactogen.  The  precipitate  which  formed  was  col- 
lected and  washed  in  the  centrifuge;  then  it  was  digested 
with  fresh  rabbit  serum  in  order  to  supply  the  complement. 
Such  preparations  after  varying  periods  of  digestion,  were 
injected  intravenously  into  rabbits,  but  with  negative  results. 
Friedberger,1  using  guinea-pigs  instead  of  rabbits,  succeeded 
fully  in  producing  the  anaphylactic  poison  by  this  method. 
For  two  reasons  the  guinea-pig  is  better  suited  for  this 
work  than  the  rabbit.  The  blood  of  the  former  is  richer 
in  complement,  and  this  animal  is  the  more  susceptible 
to  the  action  of  the  anaphylactic  poison.  However,  in 

1  Zeitsch.  f.  Immunitatsforschung,  iv,  636. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     273 

carrying  out  this  work,  even  with  the  guinea-pig,  the 
results  are  not  constant,  and  variations  in  the  quantitative 
relations  of  the  solutions  concerned  in  the  reaction  may 
lead  to  failure.  An  excess  of  the  anaphylactogen  prevents, 
apparently  at  least,  the  action  of  the  ferment,  and  no 
poison  is  formed.  Here  lies  an  important  question  which 
we  will  briefly  discuss.  Some  years  ago  when  we  were 
testing  the  lethal  doses  of  certain  bacterial  cellular  pro- 
teins,1 we  frequently  observed  that  a  small  dose  killed, 
while  two  or  three  or  more  times  this  amount  did  not;  or 
the  smaller  dose  killed  within  a  shorter  time  than  the 
larger.  The  proteins  were  administered  intra-abdominally, 
and  in  suspension.  Finally,  we  demonstrated  that  the  more 
finely  powdered  cell  substance  was  ground  the  more  poisonous 
it  became.  By  this  we  mean  that  smaller  doses  killed. 
Later  we  found,  much  to  our  surprise,  that  high  tempera- 
tures increased  the  toxicity  of  the  suspensions  of  cellular 
proteins.  We  came  to  the  following  conclusions:  (1)  The 
toxicity  of  the  bacterial  suspensions  is  determined  by  the 
rapidity  and  completeness  with  which  the  cells  are  split 
up  by  the  ferments  of  the  body.  (2)  Other  things  being 
equal,  the  rapidity  and  completeness  with  which  the  cells 
are  digested  depend  upon  the  proportion  of  surface  exposed 
to  the  action  of  the  ferment.  (3)  Grinding  the  powder  more 
finely  increases  the  surface  exposure  of  a  given  weight,  and 
therefore  leads  to  the  liberation  of  a  larger  amount  of  the 
poison  in  a  given  unit  of  time.  (4)  When  the  bacterial 
suspensions  are  heated  the  proteins  contained  in  the  cells 
are  partly  dissolved,  or  at  least  the  molecular  surface  is 
extended,  digestion  is  more  rapid  and  complete,  and  the 
substance  becomes  more  efficient  as  a  poison,  not  because 
more  poison  is  generated,  but  because  that  contained  in 
the  cell  is  made  more  available.  Later  in  our  work  on 
protein  fever2  we  came  upon  the  same  thing  in  a  new  guise. 
We  found  that  intra-abdominal  and  intravenous  injections, 
single  or  repeated,  of  egg-white  in  large  doses  in  rabbits 

1  Trans.  Assoc.  Amer.  Phys.,  1902. 

2  Zeitsch.  f.  Immunitatsforschung,  ix,  458, 
18 


274  PROTEIN  POISONS 

had  but  little  or  no  effect  on  the  temperature  of  the  animal, 
while  small  doses  frequently  repeated  caused  rapid  eleva- 
tion of  temperature,  and  death  within  ten  to  twelve  hours. 
Here,  again,  small  doses  kill  while  larger  ones  are  without 
visible  effect.  The  explanation  is  in  our  opinion  the  same 
as  that  given  for  the  bacterial  suspensions.  When  1  c.c. 
of  the  egg-white  dilution  is  injected  into  the  ear  vein  of 
the  rabbit  and  diluted  with  all  the  blood  in  the  animal 
body,  the  molecular  surface  of  the  foreign  protein  is  im- 
mensely greater  than  when  10  c.c.  of  the  egg-dilution  is 
injected.  The  egg-white  has  no  poisonous  action  until  it 
is  split  up  by  ferments,  and  the  rapidity  and  completeness 
with  which  this  is  done  is  determined  in  part  at  least  by 
the  extent  of  the  molecular  surface  of  the  substrate.  The 
same  thing  is  seen  in  the  action  of  the  precipitins.  An 
excess  of  the  antigen  prevents  precipitation.  We  believe 
the  matter  of  molecular  surface  exposure  to  be  of  great 
importance  in  the  various  phenomena  of  anaphylaxis.  The 
greater  it  is  in  the  anaphylactogen  the  more  potent  is  its 
action  both  in  sensitizing  and  on  reinjection. 

Friedberger  suggests  that  in  the  preparation  of  the 
anaphylactic  poison  in  vitro,  excess  of  the  anaphylactic 
serum  or  prolonged  time  exposure  may  carry  digestion 
beyond  the  formation  of  the  poison;  itself  being  split  up. 
This  is  in  accord  with  our  findings  as  reported  below. 

The  formation  of  the  anaphylactic  poison  from  soluble 
proteins  in  vitro  was  first  done  in  our  laboratory.1  The 
importance  of  this  matter  leads  us  to  reproduce  the  experi- 
mental part  of -our  report: 

Our  method  of  procedure  is  as  follows:  Experimentation 
has  so  far  been  confined  to  guinea-pigs.  The  chest  of  the 
etherized  animal  is  opened  and  the  blood  is  drawn  from 
the  heart  into  sterilized  tubes  and  thus  the  serum  is  obtained. 
The  animal  dies  from  bleeding.  The  organs  are  rubbed 
up  in  a  conical  glass  with  sand,  stirred  with  30  c.c.  of  physio- 
logical salt  solution,  and  allowed  to  stand  for  subsidence 

1  Zeitsch.  f.  Immunitatsforschung,  xi,  673. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     275 

one  hour.  The  supernatant  fluid  is  then  removed  and  is 
known  as  the  organ  extract.  Egg-white  and  horse  serum 
are  diluted  with  physiological  salt  solution  until  0.1  c.c. 
contains  the  amount  of  protein  desired  in  the  individual 
experiment,  but  in  case  more  than  10  mg.  of  protein  is 
used,  a  multiple  of  0.1  c.c.  constitutes  the  solvent.  These 
solutions  are  freshly  prepared  for  each  experiment,  and 
everything  is  done  aseptically.  The  volume  of  serum  or 
organ  extract  used  is  5  c.c.,  and  the  amount  injected  into 
the  heart  of  the  animal  is,  unless  otherwise  noted,  4  c.c. 
Further  details  will  appear  in  the  record  of  the  experiments. 

1.  One    milligram    of    egg-white    incubated    for    thirty 
minutes  in  5  c.c.  of  the  serum  or  organ  extracts  of  a  normal, 
unsensitized    guinea-pig    is    without    marked    effect    when 
injected  into  the  heart  of  another  unsensitized  guinea-pig. 

TABLE  XXIV 
No.                           Fluid.  Effect. 

1  ....      Serum  None 

2  ....      Liver  None 

3  ....     Kidney  Slight  scratching 

4  ....      Spleen  None 

2.  One    milligram    of    egg-white    incubated    for    thirty 
minutes  in  5  c.c.  of  the  serum  or  organ  extracts  of  a  guinea- 
pig  killed   three  days  after  sensitization  to   egg-white  is 
without  marked  effect  when  injected  into  the  heart  of  a 
fresh  guinea-pig. 

TABLE  XXV 

No.  Fluid.  Effect. 

1  ....  Serum  Slight  scratching 

2  ....  Liver  None 

3  ....  Kidney  None 

4  ....  Spleen  None 

5  ....  Brain  None 

3.  With  the  conditions  the  same  as  in  Table  XXV,  except 
that  the  animal  supplying  the   serum  and  organ  extracts 
had  been  sensitized  to  egg-white  fourteen  days  before  being 
killed,  the  effects  were  marked  as  showed  in  Table  XXVI. 


276  PROTEIN  POISONS 

TABLE  XXVI 

No.  Fluid.  Effect. 

1  .      .      .      .  Serum  (3  c.c.)  Dead  in  four  minutes 

2  ....  Liver  Dead  in  four  minutes 

3  ....  Kidney  Dead  in  four  minutes 

4  ....  Spleen  Convulsions,  recovered 

It  should  be  remarked  that  in  all  instances  in  which 
symptoms  followed  they  were  characteristically  those  of 
the  protein  poison. 

4.  With  the  serum  or  extract  of  the  same  animal  em- 
ployed in  Table  XXVI,  but  with  the  incubation  prolonged 
to  ninety  minutes,  the  effects  were  less  marked,  as  recorded 
in  Table  XXVII. 

TABLE  XXVII 

No.  Fluid.  Effect. 

1  ....  Serum  (3  c.c.)  First  and  second  stages 

2  ....  Liver  Convulsions,  recovery 

3  ....  Spleen  Slight 

4  ....  Kidney  First  and  second  stages 

We  infer  from  this  that  the  digestion  continued  until 
the  poison  was  in  part  destroyed. 

5.  With  the  serum  and  extracts  from  the  same  animal 
employed  in  Tables  XXVI  and  XXVII,  but  the  fluids  after 
the  addition  of   the  egg-white  kept  in  the  cold  room  for 
twenty-four  hours,  the  effects  were  not  marked,  as  shown 
in  Table  XXVIII. 

TABLE  XXVIII 

No.  Fluid.  Effect. 

1  ....  Serum  None 

2  ....  Liver  None 

3  ....  Kidney  Slight  scratching 

4  .      .      .      .  Spleen  None 

6.  With  the  serum  and  extracts  obtained  from  an  animal 
killed  seventeen  days  after  being  sensitized  to  egg-white, 
the  effects  varied  with  the  time  of  incubation,  as  recorded 
in  Table  XXIX. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     277 

TABLE  XXIX 

Time  of 

No.                       Fluid.  incubation.                     Effect. 

1  ...  Serum  15  min.  Dead  in  30  minutes 

2  ...  Serum  30  min.  Dead  in    6  minutes 

3  ...  Liver  15  min.  First  and  second  stgaes 

4  ...  Liver  30  min.  Dead  in  6  minutes 

5  ...  Kidney  15  min.  First  and  second  stages 

6  ...  Kidney  30  min.  Dead  in  5  minutes 

7  ...  Spleen  15  min.  First  stage 

8  ...  Spleen  30  min.  First  and  second  stages 

It  seems  from  this  that  fifteen  minutes  is  too  short  a 
time  for  the  full  development  of  the  poison. 

7.  The   ferment   passes   through   hardened   filter   paper. 
The  serum  and  organ  extracts  from  an  animal  killed  twenty 
days  after  sensitization  to  egg-white  were  filtered  through 
hardened  paper.     To  each  5  c.c.  of  the  filtrates  1  mg.  of 
egg-white  protein  was  added,  incubated  for  thirty  minutes, 
and    then    injections    were    made   intracardiacly    in   fresh 
guinea-pigs,  with  the  results  shown  in  Table  XXX. 

TABLE  XXX 

No.  Fluid.  Effect. 

1  ....  Serum  (3  c.c.)  Death  delayed  12  hours 

2  ....  Liver  First  and  second  stages 

3  ....  Kidney  Dead  in  8  minutes 

4  ....  Spleen  Dead  in  6  minutes 

8.  The  ferment  passes  through  a  Berkefeld  filter.     The 
serum  and  organ  extracts  of  an  animal  killed  twenty-two 
days  after  sensitization  to  egg-white  were  filtered  through 
a  Berkefeld  V.    To  each  5  c.c.  of  these  filtrates  1  mg.  of 
egg-white  protein  was  added,  incubated  for  thirty  minutes, 
and  then  intracardiac  injections  were  made  in  fresh  guinea- 
pigs  with  results  shown  in  Table  XXXI. 

TABLE  XXXI 
No.  Fluid.  Effect. 

1  ....     Serum  Dead  in  6  minutes 

2  ....     Liver  Dead  in  9  minutes 

3  ....     Kidney  Dead  in  4  minutes 

4  ....      Spleen  Dead  in  6  minutes 


278  PROTEIN  POISONS 

9.  The  poison  formed  by  the  action  of  the  ferment  on 
the  protein  passes  through   hardened   filter  paper.     One 
milligram  of  egg-white  protein  was  added  to  each  5  c.c. 
of  serum  and  organ  extracts  obtained  from  a  guinea-pig 
killed  twenty  days  after  sensitization  to  egg-white;  then 
these  portions  were  incubated  for  thirty  minutes,  filtered 
through   hardened   paper,    and   injected   intracardiacly   in 
fresh  guinea-pigs,  with  the  results  shown  in  Table  XXXII. 

TABLE  XXXII 
No.  Fluid.  Effect. 

1  ....     Serum  Convulsions,  recovery 

2  ....      Liver  Dead  in  12  minutes 

3  ....      Kidney  Dead  in    2  minutes 

4  ....     Spleen  First  and  second  stages 

10.  The   poison   passes   through   a   Berkefeld   V.     The 
serum  and  organ  extracts  of  a  guinea-pig,  killed  twenty- 
three  days  after  sensitization  to  egg-white,  were  treated 
with  1  mg.  of  egg-white  protein  to  each  5  c.c.,  incubated 
for  thirty  minutes,  filtered  through  a  Berkefeld  V,   and 
the  filtrates  injected   intracardiacly   in   fresh  guinea-pigs, 
with  the  results  shown  in  Table  XXXIII. 

TABLE  XXXIII 
No.  Fluid.  Effect. 

1  ....     Serum  (3  c.c.)  First  and  second  stages 

2  ....     Liver  Dead  in    6  minutes 

3  ....     Spleen  Dead  in  10  minutes 

4  ....     Kidney  Dead  in    9  minutes 

11.  Serum  and  organ  extracts  from  sensitized  animals 
are  inactivated  when  heated  to  56°  for  thirty  minutes. 

In  our  first  experiment  on  this  point  the  fluids  were 
placed  in  small  Erlenmeyer  flasks,  and  these  were  set  in 
water  at  56°  and  allowed  to  stand  for  thirty  minutes;  then 
5  c.c.  portions,  to  each  of  which  1  mg.  of  egg-white  protein 
was  added,  were  kept  in  the  incubator  for  thirty  minutes. 
The  result  was  that  the  inactivation  was  incomplete,  as 
shown  by  Table  XXXIV. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     279 

TABLE  XXXIV 

No.  Fluid.  Effect. 

1  ....  Serum  First  and  second  stages 

2  ....  Liver  Dead  in  4  minutes 

3  ....  Kidney  First  stage 

4  ....  Spleen  None 

As  a  control  to  the  experiments  of  Table  XXXIV,  animals 
were  treated  with  the  serum  and  organ  extracts  from  the 
same  animal  to  which  egg-white  had  been  added  and  incu- 
bated for  thirty  minutes  without  previous  subjection  to 
heat.  The  results  are  shown  in  Table  XXXIV  A. 


TABLE  XXXIV  A 

No.  Fluid.  Effect. 

1  ....  Serum  (3  c.c.)  Dead  in  6  minutes 

2  ....  Liver  Dead  in  6  minutes 

3  ....  Kidney  Dead  in  5  minutes 

4  ....  Spleen  First  and  second  stages 

In  a  repetition  of  this  test,  the  serum  and  extracts  were 
placed  in  thin  sealed  tubes  and  kept  submerged  for  thirty 
minutes  in  water  at  the  temperature  of  56°.  After  this 
the  fluids  were  treated  with  1  nag.  of  egg-white  protein, 
incubated  for  thirty  minutes,  and  injected  into  the  hearts 
of  normal  guinea-pigs,  with  the  results  shown  in  Table 
XXXIV  B. 

TABLE  XXXIV  B 

No.  Fluid.  ,     Effect 

1  ....  Serum  None 

2  ....  Kidney  None 

3  ....  Liver  None 

4  ....  Spleen  None 

As  a  control  to  the  experiments  of  Table  XXXIV  B,  the 
unheated  extracts  from  the  same  animal  were  employed, 
with  results  as  shown  in  Table  XXXIV  c.  There  was  not 
enough  serum  to  use  in  the  control. 


280  PROTEIN  POISONS 

TABLE  XXXIV  c 

No.  Fluid.  Effect. 

1  ....  Kidney  Dead  in  8  minutes 

2  ....  Liver  Dead  in  5  minutes 

3  ....  Spleen  Dead  in  4  minutes 

12.  Serum   and   organ   extracts   inactivated   by   heating 
to  56°  may  be  reactivated  by  the  addition  of  corresponding 
fluids  obtained  from  an  unsensitized  animal.     The  serum 
and  organ  extracts  obtained  from  an  animal  killed  twenty- 
seven  days  after  being  sensitized  to  egg-white  were  inacti- 
vated  by  being  heated  to  56°,   then  treated  with   equal 
volumes  of  serum  and  corresponding,  organ  extracts,  egg- 
white  added,  1  mg.  to  each  5  c.c.  of  fluid,  and  incubated 
for  thirty  minutes.     These  fluids  when  injected  into  the 
hearts  of  unsensitized  animals  produced  the  results  shown 
in  Table  XXXV. 

TABLE  XXXV 

No.  Fluid.  Effect. 

1  ....  Serum  (3  c.c,)  First  and  second  stages 

2  ....  Liver  Dead  in  34  minutes 

3  ....  Kidney  Dead  in  56  minutes 

4  ....  Spleen  Dead  in  45  minutes 

It  will  be  noticed  that  when  reactivated  fluids  were  used, 
death  was  not  so  speedy. 

As  a  control  to  the  experiments  of  Table  XXXV,  the 
inactivated  fluids  from  the  same  animal  were  used  without 
the  addition  of  complement,  with  results  shown  in  Table 
XXXV  A. 

TABLE  XXXV  A 

No.  Fluid.  Effect. 

1  ....  Serum  (3  c.c.)  None 

2  ....  Liver  None 

3  ....  Kidney  None 

4  ....  Spleen  None 

13.  The  serum  or  organ  extracts  of  animals  sensitized 
to  egg-white  do  not  produce  a  poison  when  incubated  with 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     281 

horse  serum  for  thirty  minutes.  The  serum  and  organ 
extracts  of  an  animal  killed  twenty-six  days  after  sensiti- 
zation  to  egg-white  were  incubated  with  horse  serum 
protein,  1  mg.  to  each  5  c.c.  of  fluid,  for  thirty  minutes, 
and  then  injected  into  the  hearts  of  unsensitized  animals, 
with  the  results  shown  in  Table  XXXVI. 

TABLE  XXXVI 

No.  Fluid.  Effect. 

1  ....     Serum  (2  c.c.)  None 

2  ....      Liver  None 

3  .      .       .      Kidney  None 

4  ....     Spleen  None 

14.  The  serum  and  organ  extracts  of  an  animal  sensitized 
to  horse  serum  do  produce  a  poison  when  incubated  with 
horse  serum.    The  serum  and  organ  extracts  of  an  animal 
killed  eleven  days  after  sensitization  with  horse  serum  were 
incubated  for  thirty  minutes  with  1  mg.  of  horse  serum 
protein  to  each  5  c.c.  of  fluid  and  then  injected  into  the 
hearts  of  fresh  animals,  with  the  results  shown  in  Table 
XXXVII. 

TABLE  XXXVII 

No.  Fluid.  Effect. 

1 Serum  Dead  in  24  minutes 

2  ....  Spleen  Dead  in  18  minutes 

3  ....  Liver  Dead  in  27  minutes 

4  ....  Kidney  Dead  in  60  minutes 

5  ....  Brain  Dead  in    4  minutes 

15.  The  serum  and  organ  extracts  of  an  animal  sensitized 
to  typhoid  bacilli  by  the  intensive  method   (described  in 
Zeitschrift  /.   Immunitatsforschnng,   vol.   ix,  p.   458)  when 
incubated   with  living  typhoid  bacilli   for   thirty  minutes 
do  produce  a  poison.     This  animal  had  been  thus  treated 
twenty-seven  days  before.     One-tenth  of  a  loop  from  an 
agar  slant  four  days  old  was  added  to  each  5  c.c.  of  fluid. 
The  results  are  shown  in  Table  XXXVIII. 


282  PROTEIN  POISONS 

TABLE  XXXVIII 

No.  Fluid.  Effect. 

1  .      Serum  Dead  in    4  minutes 


Spleen  Dead  in  16  minutes 

Brain  Dead  in    4  minutes 

Liver  Dead  in    4  minutes 

Kidney  Dead  in    6  minutes 


16.  A  like  experiment  with  the  serum  and  organ  extracts 
of  an  animal  treated  twenty-eight  days  before  with  the 
bacillus    of    cholera    gave    the    results    shown    in    Table 
XXXIX. 

TABLE  XXXIX 

No.  Fluid.  Effect. 

1  ....  Serum  Dead  in  14  minutes 

2  ....  Liver  Dead  in    6  minutes 

3  ....  Brain  Dead  in    8  minutes 

4  ....  Kidney  Convulsions,  recovery 

5  ....  Spleen  First  and  second  stages 

17.  Sensitization  with  egg-white  in  the  guinea-pig  does 
not  continue  indefinitely.     Vaughan  and   Wheeler  found 
that  these  animals,  when  the  second  injection  was  made 
four  hundred  days  or  longer  after  the  first,  proved  not  to 
be  in  a  condition  of  sensitization,  and  that  injections  made 
after  this  time  resensitized.    Apparently  in  cases  of  sensi- 
tization with  egg-white  the  serum  first  loses  the  specific 
ferment.    A  guinea-pig  killed  thirty-four  days  after  sensi- 
tization to  egg-white  furnished  a  serum  and  organ  extracts 
which  when  incubated  with  egg-white  and  injected  into  the 
hearts  of  fresh  animals  gave  the  results  shown  in  Table  XL. 

TABLE  XL 

No.  Fluid.  Effect. 

1  ....  Serum  (3  c.c.)  None 

2  ....  Liver  Dead  in    7  minutes 

3  ....  Kidney  Dead  in  18  minutes 

4  ....  Brain  Convulsions,  recovery 

5  ....  Spleen  Convulsions,  recovery 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     283 

18.  When  the  amount  of  egg-white  protein  added  to 
5  c.c.  of  the  fluid  before  incubation  was  varied,  the  results 
as  shown  in  the  promptness  and  intensity  of  the  effect  of 
the  poison  were  found  to  vary.  The  serum  and  organ 
extracts  of  two  guinea-pigs,  one  killed  twenty-eight  days 
and  the  other  thirty  days  after  sensitization  to  egg-white, 
were  divided  into  portions  of  5  c.c.,  to  which  varying  amounts 
of  egg-white  were  added,  incubated  for  thirty  minutes,  and 
injected  into  the  hearts  of  fresh  animals,  with  the  results 
recorded  in  Table  XLI. 


TABLE  XLI 


No. 

1 

2 

3 

4 

5 

6 

7 

8 

9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 


Amount  of 

Fluid. 

egg-white. 

Serum 

0.5  mg. 

Spleen 

0.5  mg. 

Liver 

0.5  mg. 

Kidney 

0.5  mg. 

Serum 

5.0  mg. 

Liver 

5  .  0  mg. 

Kidney 

5  .  0  mg. 

Spleen 

5.0  mg. 

Brain 

5.0  mg. 

Serum 

10.0  mg. 

Brain 

10.0  mg. 

Spleen 

10.0  mg. 

Kidney 

10.0  mg. 

Liver 

10.0  mg. 

Serum 

20.0  mg. 

Brain 

20.0  mg. 

Spleen 

20.0  mg. 

Kidney 

20.0  mg. 

Liver 

20.0  mg. 

Effect. 
Slight 

Dead  in  6  minutes 
Dead  in  4  minutes 
Dead  in  8  minutes 
Convulsions,  recovery 
Dead  in  2  hours 
Dead  in  5  minutes 
Dead  in  70  minutes 
Dead  in  2  hours 
Convulsions,  recovery 
First  and  second  stages 
First  and  second  stages 
Dead  in  80  minutes 
Dead  in  60  minutes 
None 

Slight  convulsions 
Dead  in  1  hour 
Dead  in  95  minutes 
Dead  in  65  minutes 


19.  The  protein  poison  prepared  from  egg-white  by 
cleavage  with  a  2  per  cent,  solution  of  sodium  hydroxide, 
and  still  by  no  means  pure,  kills  guinea-pigs  when  injected 
into  the  heart  in  doses  of  0.5  mg.  The  minimum  fatal  dose 
of  the  pure  poison  must  be  much  less  than  this.  The  symp- 
toms are  the  same  as  those  induced  by  injections  of  the 
serum  and  organ  extracts  from  guinea-pigs  sensitized  to  egg- 
white  after  incubation  with  egg-white  for  thirty  minutes. 


284  PROTEIN  POISONS 

It  will  be  seen  that  the  serum  and  organ  extracts  of 
sensitized  guinea-pigs  contain  an  agent  which  when  mixed 
with  homologous  anaphylactogens  in  vitro  in  proper  pro- 
portions and  incubated  for  the  proper  time  produces 
a  poison  which  when  injected  intracardiacly  into  fresh 
animals  causes  typical  anaphylactic  shock  and  sudden 
death.  This  poison-producing  agent  is  a  ferment  and  is 
inactivated  by  a  temperature  of  56°  and  reactivated  on  the 
addition  of  serum  or  organ  extracts  from  normal  animals. 
Like  the  toxins  and  many  other  ferments  this  one  consists 
of  amboceptor  and  complement.  The  latter  is  destroyed 
by  a  temperature  of  56°,  but  being  a  constituent  of  normal 
serum  and  organ  extracts,  its  loss  is  made  good  on  the 
addition  of  these  substances.  The  ferments  formed  in 
anaphylaxis  are  strictly  specific,  their  specificity  being 
determined  by  the  anaphylactogen  and  residing  in  the 
amboceptor.  The  ferment  elaborated  in  anaphylaxis, 
like  the  toxins,  consists  of  amboceptor  and  complement. 
The  anaphylactogen  is  not  a  toxin  and  the  substance 
produced  in  the  body  under  its  influence  is  not  an  antitoxin. 
The  anaphylactogen  does  not  even  contain  a  toxin  group; 
it  contains  a  poison,  and  it  is  this  that  is  set  free  on  reinjec- 
tion.  As  we  have  stated,  there  is  no  more  justification  in 
calling  the  anaphylactic  ferment  an  antibody  than  there 
would  be  in  designating  the  proteolytic  ferments  of  the 
alimentary  canal  antibodies. 

The  Poison. — While  anaphylactogens  and  anaphylactic 
ferments  are  specific  the  poison  is  not  specific.  It  is  one 
and  the  same  thing  whatever  the  anaphylactogen,  and  in 
our  opinion  it  is  the  poisonous  group  in  the  protein  mole- 
cule. Our  studies  on  the  protein  poison,  done  before  the 
phenomena  of  anaphylaxis  were  known,  demonstrated  the 
presence  of  a  poisonous  group  in  widely  diversified  proteins, 
and  it  probably  exists  in  all  true  proteins.  We  found  it  in 
bacteria,  both  saprophytic  and  pathogenic,  and,  as  has 
been  stated,  we  then  were  convinced  that  the  pathogenicity 
of  bacteria  bears  no  relation  to  the  poison  content  of  the 
molecule  of  its  cellular  protein.  The  pathogenicity  of  a 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     285 

bacterium  is  determined  by  its  capability  of  growing  in 
and  ultimately  sensitizing  the  animal  body.  Furthermore, 
we  demonstrated,  to  our  own  satisfaction  at  least,  that  the 
symptoms  and  lesions  of  the  infections  are  not  directly 
due  to  the  multiplication  of  the  bacteria  in  the  body,  but 
to  their  destruction  by  the  sensitized  cells  of  the  animal 
body,  because  at  the  time  when  the  growth  and  multipli- 
cation of  the  bacteria  proceed  most  rapidly — in  the  period 
of  incubation — there  are  no  symptoms  and  no  lesions.  The 
onset  of  the  disease  marks  the  time  when  sensitization 
becomes  manifest. 

We  found  the  protein  poison  in  diverse  animal  and  vege- 
table proteins,  and  with  the  few  substances  in  which  we 
did  not  find  the  protein  poison,  such  as  gelatin  and  some 
peptones,  we  were  not  able  to  sensitize  guinea-pigs. 

The  reasons  we  have  for  holding  that  our  protein  poison 
is  identical  with  the  anaphylactic  poison  may  be  stated 
as  follows: 

1.  The  protein  poison  exists  in  all  true  proteins,  so  far 
as   they  have   been   tested,   consequently   it   exists   in   all 
anaphylactogens. 

2.  Whatever  the  protein  from  which  the  poison  is  obtained, 
its  physiological  action  is  the  same.     While  there  may  be 
and  probably  are  chemical  differences  in  the  protein  poison 
as  obtained  from  diverse  proteins,  physiologically  there  is 
no  difference.     Likewise  the  symptoms  in  anaphylaxis  are 
the  same  whatever  the  anaphylactogen. 

3.  The  symptoms  induced  in  fresh  animals  by  the  protein 
poison  are  identical  in  every  detail  with  those  observed  in 
sensitized  animals  after  reinjection.    They  come  on  in  the 
same  time,  proceed  in  the  same  order,  and  terminate  alike. 

4.  Friedberger  has  shown  that  guinea-pigs  killed  with 
the  protein  poison  show  the  Auer-Lewis  phenomenon  in 
the  lungs. 

5.  Edmunds  has  shown  that  dogs  killed  with  the  protein 
poison  manifest  the  same  symptoms  as  those  studied  in 
anaphylactic   shock.     The   lowered   blood   pressure  found 
in  anaphylactic  shock  and  in  peptone  poisoning  in  dogs  is 


286  PROTEIN  POISONS 

just  as  marked  in  those  under  the  influence  of  the  protein 
poison. 

6.  Our  poison  is  the  active  principle  in  peptone,   and 
when  it  has  been  extracted  from  peptone  the  residue  is  no 
longer  poisonous. 

7.  When  the  poison  has  been  removed  from  an  anaphyl- 
actogen   the   residue   may   or   may   not   sensitize,   but   in 
no  case  does  it   induce  the  symptoms   of  anaphylaxis  on 
reinjection. 

8.  The  activity  of  the  protein  poison  is  progressively 
increased  to  a  certain  point  in  proteolytic  digestion.    Peptone 
is  more  poisonous  than  the  protein  from  which  it  is  formed, 
and  the  same  is  true  of  some  of  the  products  of  tryptic 
digestion.     The  protein  poison  in  ordinary  proteins  is  not 
active  because  it  is  combined  with  other  groups,  and  as 
these   groups   are   detached    it   becomes   more   and   more 
poisonous.      The    protein    molecule    is    a    highly    complex 
organic  compound  made  up  of  many  groups,  some  of  which 
are  basic,  and  some  acid  in  character,  and  at  least  one  which, 
when  detached  from  the  others,  is  highly  poisonous,  and  it 
is  poisonous  because  of  the  avidity  with  which  it  disrupts 
the  proteins  of  the  body.    To  make  it  simpler  we  may  say 
that  the  protein  molecule  is  a  neutral  or  basic  salt,  and  as 
the  basic  elements  are  split  off  it  becomes  an  acid  salt,  and 
finally  a  free  acid,  and  with  each  step  its  poisonous  action 
increases  because  its  capability  of  depriving  other  salts  of 
their  basic  elements  increases.     Finally  the  acid,  itself  a 
complex  body,  becomes  disrupted  and  looses  its  poisonous 
properties. 

9.  Since  proteolysis  is  a  progression  in  which  complex 
molecules  are  broken  into  simpler  and  still  simpler  ones,  in 
all  proteolytic  digestion  there  is  an  increase  in  the  activity 
of  the  protein  poison  up  to  a  given  point,  when  it  ceases  to 
be  a  poison.    It  follows,  therefore,  that  whatever  the  specifi- 
city of  the  proteolytic  ferment,  at  some  stage  in  the  process 
the  poison  is  more  or  less  freed  from  the  groups  which  tend 
to  prevent  its  action.     The  protein  molecule  has  definite 
lines  of  cleavage,  and  is  disrupted  only  along  these  lines, 


PROTEIN  SENSITIZATION  OF  ANAPHYLAXIS     287 

and  in  all  cases  its  poisonous  group  is  at  some  stage  of  the 
process  activated  as  it  were.  If  it  were  not  for  the  fact  that 
the  poisonous  group  is  not  readily  diffusible  through  animal 
membranes,  and  especially  through  the  walls  of  the  alimen- 
tary canal,  all  proteins  would  be  poisonous  to  us  even  when 
taken  by  the  mouth,  because  the  protein  poison  is  set  free 
in  alimentary  digestion,  but  not  being  readily  diffusible,  it  is 
split  up  and  rendered  inert  as  digestion  proceeds.  When, 
however,  digestion  is  parenteral,  escape  from  the  effect 
of  the  protein  poison  is  impossible,  and  the  ultimate  effect 
upon  the  organism  is  determined  wholly  by  the  amount  ren- 
dered active  at  one  time.  When  it  is  set  free  with  explosive 
rapidity  and  in  relatively  large  amount  it  induces  anaphyl- 
actic  shock,  and  possibly  death.  When  set  free  slowly  and 
in  small  amount,  we  have  fever  or  fall  in  temperature, 
according  to  the  amount  of  the  poison  liberated.  When  set 
free  either  in  the  circulating  fluid  or  when  it  passes  into 
this  fluid  immediately  we  have  systemic  effects.  When 
set  free  locally  we  have  inflammation  in  the  adjacent 
tissue.  Narrowly  used  the  term  anaphylaxis  refers  to  the 
symptoms  of  anaphylactic  shock.  In  a  wider  sense  it 
covers  all  the  phenomena  of  parenteral  protein  digestion. 
Some  think  that  parenteral  digestion  is  always  abnormal, 
either  artificially  induced  or  due  to  pathological  conditions. 
We  doubt  the  truth  of  this  assumption.  By  inhalation, 
through  abrasions  and  possibly  through  the  alimentary 
canal,  man  must  be  frequently,  almost  constantly,  taking 
into  his  blood  and  tissues  very  minute  traces  of  undigested 
proteins,  but  ordinarily  the  amounts  thus  taken  in  are  so 
infinitesimally  small  that  the  body  cells  are  not  sensi- 
tized, and  no  harm  comes.  While,  as  we  have  seen,  some 
anaphylactogens  sensitize  in  very  small  doses,  these  are  not 
infinitesimal,  and  there  are  measurable  doses  which  do  not 
sensitize.  The  limits  vary  with  the  protein  and  the  animal. 
Friedemann  and  Isaac,1  also  Pfeiffer  and  Mita,2  think 


1  Zeitsch.  f.  exp.  Path.,  1905,  i,  513;  1906,  ii;  1908,  iv,  830, 

2  Zeitsch.  f.  Immunitatsforschung,  1909,  iv. 


288  PROTEIN  POISONS 

that  the  poison  does  not  come,  solely  at  least,  from  the 
protein  used  in  the  reinjection.  Friedemann  says  that  it  is 
generally  held  that  the  poison  comes  from  the  antigen, 
but  that  this  is  pure  hypothesis.  He  holds  that  the  poison 
may  come  from  any  one  of  the  factors  in  the  reaction — 
the  anaphylactogen,  the  amboceptor,  and  the  complement. 
He  holds  that  the  minimum  killing  dose  of  the  protein 
on  reinjection  is  so  small  that  it  cannot  be  supposed  to 
furnish  a  fatal  quantity  of  the  poison,  and  he  thinks  that  the 
ferment  once  set  in  action  by  the  reinjection  may  go  on  and 
digest  the  proteins  of  the  animal  body.  Friedemann  has 
done  most  valuable  work  on  metabolism  in  anaphylaxis,  and 
he  holds  that  the  increase  in  the  nitrogen  output  is  greater 
than  all  of  this  element  contained  in  the  reinjection,  and 
therefore  he  thinks  that  the  evidence  that  the  whole  of 
the  poison  at  least  does  not  come  from  the  foreign  protein 
is  incontrovertible.  He  is  undoubtedly  right  in  his  finding 
that  nitrogen  metabolism  in  anaphylaxis  is  far  beyond 
that  which  can  be  accounted  for  by  the  nitrogen  in  the 
foreign  protein,  and  in  this  he  has  been  confirmed  by  others. 
Our  own  work1  proves  the  same  thing,  but  in  our  opinion 
this  does  not  show  that  the  poison  itself  has  any  other 
source  than  the  protein  of  the  reinjection.  In  the  first 
place,  as  we  have  seen,  the  minimum  of  the  protein  neces- 
sary to  produce  anaphylactic  shock  is  much  greater  than 
that  necessary  to  sensitize.  Rosenau  and  Anderson  sensi- 
tized one  guinea-pig  with  0.000001  c.c.  of  horse  serum, 
and  Besredka  found  sensitizing  doses  under  0.001  c.c. 
uncertain,  and  he  found  the  smallest  killing  dose  to  be 
^5-  c.c.  even  when  given  intravenously.  It  will  be  seen 
from  these  figures  that  there  is  a  big  difference  between 
the  sensitizing  and  the  killing  dose.  One-fortieth  of  a  cubic 
centimeter  of  horse  serum  contains  about  2  mg.  of  protein. 
We  found  in  our  work  that  serum  albumin  yields  about 
one-third  its  weight  of  poison,  then  2  mg.  would  yield 
0.66  mg.,  and  the  protein  poison  obtained  by  us,  in  a  crude 

1  Jour.  Amer.  Med.  Assoc.,  1909. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     289 

way  and  far  from  pure,  kills  guinea-pigs  when  injected 
intracardiacly  in  doses  of  0.5  mg.  The  minimum  fatal  dose 
of  the  pure  poison  as  split  off  by  the  ferment  in  the  body 
must  be  much  less  than  this.  It  will  be  seen  from  this 
that  the  proteins  of  the  reinjection,  even  when  the  smallest 
fatal  dose  is  used,  probably  contain  enough  of  the  poison 
to  kill.  In  the  second  place  the  ferment  developed  in 
anaphylaxis  is  specific.  It  splits  up  its  own  anaphylactogen 
and  no  other  protein.  There  is  no  reason  for  supposing 
that  it  can  digest  the  body  proteins.  It  seems  to  us  that 
this  supposition  is  wholly  untenable.  It  is  contrary  to  all 
we  know  about  the  specificity  of  the  anaphylactic  ferment. 
How  then  may  we  account  for  the  greatly  increased  nitrogen 
metabolism?  When  the  foreign  protein  is  split  up  the 
split  products  chemically  react  with  the  protein  molecules 
of  the  animal's  body.  The  liberated  poison  tears  off  the 
basic  groups  from  the  body  molecules,  and  this  goes  on 
to  the  extent  and  during  the  time  that  the  cleavage  con- 
tinues. We  see  this  in  its  most  marked  form  in  the  Arthus 
phenomenon,  for  in  this  the  process  is  more  localized.  We 
have  shown  that  foreign  proteins  injected  intravenously 
in  rabbits  soon  disappear  from  the  circulating  blood,  and 
after  this  they  may  be  detected  in  the  skin  and  in  other 
tissues.  We  have  already  seen  (p.  262)  to  what  extent 
destruction  of  tissue  may  occur  in  the  Arthus  phenomenon. 
There  is  an  additional  explanation  of  the  augmented 
nitrogen  metabolism.  The  protein  matter  resulting  from 
the  disruption  of  the  molecules  of  the  body  by  the  split 
products  from  the  anaphylactogen  is  digested  by  the  normal 
non-specific  parenteral  ferments,  and  in  this  way  nitrogen 
elimination  is  increased.  As  has  been  stated,  Friedberger 
has  obtained  the  poison  by  digesting  precipitates  and 
bacterial  cells  with  homologous  anaphylactic  sera.  As 
thus  obtained  and  injected  intravenously  it  induces  ana- 
phylactic shock  and  death.  He  has  obtained  like  results 
by  digesting  bacterial  cells  with  the  normal  serum  of  guinea- 
pigs.  It  might  be  assumed  from  this  that  Friedberger's 
poison  is  not  the  true  anaphylactic  poison,  but  in  our 
19 


290  PROTEIN  POISONS 

opinion,  it  is  the  poisonous  group  in  the  protein  molecule, 
and  this  is  the  anaphylactic  poison,  it  matters  not  what 
the  agent  be  which  has  detached  it  from  the  other  groups. 
This  agent  may  be  wholly  chemical,  such  as  we  have  used 
in  the  retort,  and  it  may  be  any  proteolytic  ferment,  the 
ferment  of  the  gastric  juice,  that  of  a  specific  or  a  non- 
specific serum.  As  we  have  stated,  since  proteolysis  consists 
in  the  successive  and  progressive  disruption  of  the  protein 
molecule,  in  at  least  one  stage  of  this  process,  whatever 
causes  it,  the  protein  poison  must  be  released  from  com- 
bination with  those  groups  which  in  the  original  molecule 
neutralize  it.  The  sera  of  many  animals,  possibly  of  all, 
contain  proteolytic  ferments;  some  are  more  active  than 
others;  some  act  upon  certain  while  some  act  upon  other 
proteins.  The  products  of  proteolysis  resulting  from  different 
ferments  certainly  differ,  and  even  the  poisonous  group  as 
detached  from  the  non-poisonous  groups  by  different 
ferments  probably  differs  in  its  molecular  structure,  but 
the  poisonous  principle  is  the  same  in  all  cases.  Even  its 
physiological  action  may  be  slightly  modified,  though  there 
is  no  evidence  of  this,  by  variation  in  the  lines  of  cleavage 
along  which  the  protein  molecule  is  disrupted.  The  pieces 
into  which  the  large  protein  molecule  is  split  depend  upon 
the  shape,  weight,  and  force  of  the  hammer  that  strikes 
it,  and  the  point  where  the  blow  falls.  Some  of  the  pieces 
are  large  and  some  small,  and  when  the  blow  is  especially 
effective  the  pieces  may  be  so  small  that  the  poisonous 
group  is  broken  and  rendered  inert;  but  even  when  protein 
is  fused  with  caustic  alkali  the  cyanogen  group  is  still  in 
evidence. 

It  seems  to  us  that  Friedberger's  work  has  only  confirmed 
our  contention,  first  published  in  1907,  that  the  anaphyl- 
actic poison  is  the  poisonous  group  in  the  protein  molecule. 
Friedberger  calls  his  poison  "anaphylatoxin."  We  join 
Friedemann  in  protesting  against  this  name.  The  substance 
is  not  a  toxin,  as  we  now  understand  that  word.  Fried- 
berger has  demonstrated  this  fact  himself,  inasmuch  as  he 
has  shown  that  his  poison  does  not  induce  immunity,  nor 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     291 

does  it  cause  animals  treated  with  it  to  elaborate  an  anti- 
toxin. 

That  the  anaphylactic  poison  is  a  protein  derivative  is 
certain;  whether  it  is  still  a  biuret  body  has  not  been  deter- 
mined. The  chemistry  of  the  protein  poison  has  been 
discussed  (p.  101). 

/3-iminazolylethylamin. — This  amin  is  produced  by  split- 
ting off  carbon  dioxide  from  histidin,  and  this  may  be  done 
by  either  chemical  or  bacterial  agencies.  It  was  first 
prepared  synthetically  by  Windam  and  Vogt,1  and  then 
by  Ackermann2  by  the  action  of  putrefactive  bacteria  on 
histidin.  About  the  same  time  it  was  detected  in  ergot, 
and  its  physiological  action  investigated  by  Barger  and 
Dale.3  In  the  same  year  Kutscher4  isolated  from  ergot  a 
substance  which  chemically  could  not  be  distinguished 
from  this  amin,  but  which  was  believed  to  have  a  some- 
what different  physiological  action.  /3-iminazolylethylamin, 
hereafter  designated  by  the  abbreviation  /3-i,  was  suggested 
as  a  possible  agent  in  inducing  anaphylactic  shock  by  Dale 
and  Laidlaw,5  who  made  a  thorough  study  of  its  physio- 
logical action.  It  is  highly  poisonous,  0.5  mg.  being  sufficient 
to  kill  a  guinea-pig,  with  all  the  symptoms  of  anaphylactic 
shock,  when  administered  intravenously.  Dale  and  Laidlaw 
describe  its  action  on  guinea-pigs  as  follows:  "In  large 
guinea-pigs,  weighing  800  to  1000  grams,  injection  of  0.5 
mg.  into  the  external  saphenous  vein  caused  death  in  a  few 
minutes.  The  immediate  effect  was  a  marked  respiratory 
impediment,  resulting  in  violent  but  largely  ineffective 
inspiratory  efforts,  during  which  the  lower  ribs  were  drawn 
in.  After  a  time  the  respiratory  convulsions  ceased,  and 
the  animal  lay  comatose,  though  the  heart  continued  to 
beat  for  some  time  longer.  Post  mortem:  The  lungs  were 
found  permanently  distended.  If  the  fatal  amount  were 

1  Berichte,  1907,  xl,  3691. 

2  Zeitsch.  f.  physiol.  Chem.,  1910,  xlv,  504. 

3  Proc.  Chem.  Soc.,  1910,  xxvi,  128. 

4  Zentralbl.  f.  Physiol.,  1910,  xxiv,  163. 
6  Journal  of  Physiology,  1911,  Ixi,  318. 


292  PROTEIN  POISONS 

given  more  slowly,  as  in  two  doses  of  0.25  mg.,  after  the 
second  of  which  death  ensued  rapidly,  the  final  condition 
of  pulmonary  distention  was  extreme.  Death  was  clearly 
due  to  asphyxia,  evidently  resulting  from  progressive 
obstruction  to  the  respiration,  sufficient  in  its  early  stages 
to  prevent  the  exit  of  air  sucked  into  the  lungs  by  the 
violent  inspiratory  spasms,  and  later  becoming  complete. 
The  larger  the  initial  dose,  and,  therefore,  the  earlier  the 
obstruction  became  complete,  the  less  pronounced  the 
distention  of  the  lungs.  Such  an  effect  could  only  be  due 
to  constriction  of  the  bronchioles  by  spasm  of  their  muscular 
coats,  though  the  effect  would  be  aided  by  increased  bron- 
chial secretion.  Preliminary  injection  of  atropine,  though 
it  did  not  abolish  the  action,  had  decided  protective  value. 
After  5  mg.  of  atropine  a  dose  of  1  mg.  of  /3-i  intravenously 
had  the  normal  effect,  but  another  guinea-pig,  which  received 
a  preliminary  injection  of  5  mg.  of  atropine,  recovered  from 
subsequent  intravenous  injections  of  0.5  mg.,  0.25  mg.,  and 
again  0.5  mg.  of  /3-i  given  in  fairly  rapid  succession;  whereas 
one  dose  of  0.5  mg.  was,  in  our  experience,  invariably 
fatal  when  given  intravenously  to  a  guinea-pig  untreated 
with  atropine.  Whether  atropine  actually  weakens  the 
bronchial  spasm,  or  merely  modifies  the  effect  by  preventing 
secretion,  must  remain  uncertain.  We  were  unable  to 
remove  the  obstruction  when  once  developed  by  a  sub- 
sequent injection  of  atropine." 

In  dogs  and  cats  /3-i  causes  a  marked  fall  in  blood-pressure, 
and  in  this  respect  also  agrees  with  the  anaphylactic  poison. 
On  the  smooth  muscle,  notably  on  that  of  the  virgin  uterus, 
it  has  a  markedly  stimulating  effect.  It  does  not,  according 
to  the  findings  of  the  English  investigators,  affect  the 
coagulability  of  the  blood. 

In  a  later  paper  Barger  and  Dale1  make  a  further  com- 
parison between  the  physiological  action  of  /3-i  and  peptone 
poisoning,  especially  with  the  action  of  the  "  vasodilatin" 
of  Popielski,  and  they  state  their  conclusions  as  follows: 

1  Journal  of  Physiology,  1911,  Ixi,  499. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     293 

"The  hypothetical  vasodilatine  must,  therefore,  be 
regarded  as  consisting  of  at  least  two  substances: 

"1.  /3-iminazolylethalamin,  causing  fall  of  blood-pressure, 
and  the  other  characteristic  effects  on  plain  muscle  and 
gland  cells,  but  not  affecting  coagulation  of  the  blood. 

"2.  Another  substance,  or  other  substances,  which  renders 
the  blood  incoagulable,  and  which  may  or  may  not  play 
some  part  in  the  other  effects." 

Friedberger  and  Moreschi1  conclude  from  its  behavior 
toward  alkalies  that  it  is  not  the  true  anaphylactic  poison 
which  they  believe  to  be  the  anaphylatoxin  of  Friedberger. 

Biedl  and  Kraus2  hold,  contrary  to  Barger  and  Dale, 
that  /3-i  does  delay  the  coagulation  of  blood  in  dogs.  They 
say:  "In  dogs  3  mg.  of  this  substance  causes  immediate 
fall  in  blood-pressure,  retards  the  coagulation  of  the  blood, 
and  induces  the  phenomena  of  anaphylaxis." 

With  our  poison,  /3-i  seems  to  agree  closely.  Both  induce 
bronchial  spasm  and  distention  of  the  lungs  in  guinea-pigs, 
and  cause  prompt  and  marked  fall  in  blood  pressure  in 
dogs.  Neither  destroys  the  coagulability  of  the  blood. 
In  the  purest  form  in  which  we  have  obtained  it  our  poison 
kills  guinea-pigs  intravenously  in  doses  of  0.5  mg.,  and  this 
is  the  fatal  dose  of  /3-i.  When  the  active  agents  in  our 
crude  poison  are  isolated  we  shall  not  be  surprised  if  /3-i 
or  some  closely  allied  body  is  among  them. 

/3-i  has  been  prepared  by  Barger  and  Dale  from  the 
mucosa  of  the  small  intestine  of  the  ox  by  boiling  with  0.1 
per  cent,  of  hydrochloric  acid,  and  further  treatment  with 
silver  nitrate,  and  excess  of  baryta,  according  to  the  method 
of  Kutscher.  In  regard  to  this  work  they  make  the  following 
statement:  "We  have  no  evidence  with  regard  to  the  origin 
of  the  /3-i  in  the  extract  of  intestinal  mucosa.  All  possible 
precautions  were  taken  to  avoid  putrefaction  before  the 
material  was  worked  up.  Moreover,  a  piece  of  intestine 
removed  immediately  after  death,  or  even  during  life,  from 


1  Berl.  klin.  Woch.,  1912,  No.  16. 

2  Zeitsch.  f.  Immunitatsforschung,  1912,  xv,  447. 


294  PROTEIN  POISONS 

an  anesthetized  animal,  washed,  scraped,  and  worked  up 
immediately  gives  an  extract  with  the  characteristic  physio- 
logical action  of  /3-i.  Bayliss  and  Starling  showed  that  the 
depressor  substance  could  be  extracted  from  fresh  mucous 
membrane  of  the  dog's  intestine  by  alcohol.  It  must 
probably,  then,  be  regarded  as  a  normal  product  of  intes- 
tinal mucosa,  though  whether  it  is  present  in  living  cells, 
or  only  formed  when  these  are  killed  and  disintegrated, 
remains  uncertain." 

/3-i  has  recently  become  a  commercial  product  under  the 
name  "ergamin;"  it  is  also  known  as  "histamin." 

Mellanby  and  Twort1  have  confirmed  Ackerman's  findings2 
that  histidin  is  converted  into  ergamin  by  bacterial  agencies, 
and  have  demonstrated  that  it  is  formed  in  this  way  in 
the  alimentary  canal.  They  have  isolated  a  bacillus  which 
causes  this  conversion :  "  It  is  a  small  bacillus  with  rounded 
ends,  non-motile,  and  Gram-negative.  It  will  grow  aero- 
bically  or  anaerobically  on  the  ordinary  laboratory  media. 
The  optimum  temperature  is  about  37°.  The  growth  on 
gelatin,  agar,  and  broth  is  similar  to  that  of  bacillus  coli. 
Milk  is  clotted  and  no  liquefaction  of  gelatin  takes  place. 
Acid  and  gas  are  produced  in  media  containing  glucose, 
lactose,  or  dulcite."  "In  the  alimentary  canal  of  a  guinea- 
pig,  at  least,  and  probably  in  that  of  most  mammals,  the 
bacillus  capable  of  producing  /3-i  from  histidin  is  present 
from  the  duodenum  downward.  It  is  legitimate,  therefore, 
to  assume  that  the  presence  of  the  histidin  base,  described 
by  Barger  and  Dale,  is  due  to  bacterial  decomposition 
going  on  in  the  intestine."  It  grows  and  produces  ergamin 
in  alkaline  Ringer's  solution  containing  0.1  per  cent,  of 
histidin.  When  the  concentration  of  the  histidin  is  greater, 
the  growth  is  not  so  prompt  nor  the  conversion  so  com- 
plete. "It  is  evident  that  the  toxic  symptoms  produced 
by  the  substance  together  with  its  presence  in  the  alimentary 
tract  must  bring  it  under  consideration  as  a  possible  cause 


1  Jour.  Physiol.,  1912,  xlv,  53. 

2  Zeitsch.  f.  physiol.  Chemie,  1910,  Ixv,  504. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      295 

of  pathological  conditions.  It  is  probable  that  under  normal 
conditions  the  liver  can  deal  adequately  with  /3-i,  as  it 
can  with  the  amins  of  tyrosin  and  tryptophan,  and  render 
it  innocuous;  but  if  this  defensive  mechanism  of  the  liver 
breaks  down  for  any  reason,  then  many  toxic  symptoms 
will  no  doubt  follow.  For  instance,  one  of  us  has  elsewhere 
suggested  that  the  condition  of  cyclic  vomiting  in  children 
may  be  due  to  the  excessive  accumulation  of  such  sub- 
stances as  j8-i  in  the  intestine,  causing,  from  time  to  time, 
an  exacerbation  of  symptoms.  In  any  case  a  fact  which 
would  appear  to  point  to  means  of  lessening  the  formation 
of  this  substance  in  the  alimentary  canal  is  worth  consid- 
eration. This  base  is  not  produced  in  an  acid  medium,  and 
this  fact  is  additional  support  to  the  medical  treatment, 
as  advocated  by  Metchnikoff,  involving  the  injection  of 
lactic  acid  producing  bacilli.  It  is  necessary,  however, 
to  point  out  that  the  colon  bacillus  responsible  for  the 
production  of  the  toxic  product  is  not  killed  by  the  acidity 
of  a  medium,  but  its  energies  are  only  directed  along  other 
lines,  so  that  as  soon  as  an  alkaline  reaction  returns  the 
production  of  the  histidin  base  continues." 

The  Kyrins.— These  bodies  have  been  studied  and  described 
by  Siegfried,1  who  regards  them  as  intermediate  products 
between  the  proteins  and  the  amino  acids.  His  method 
of  preparation  consists  in  digesting  the  protein  (fibrin, 
casein,  etc.)  for  three  weeks  at  38°  to  39°,  with  from  12  to 
16  per  cent,  hydrochloric  acid.  This  mixture  is  filtered 
and  the  filtrate  precipitated  with  phosphomolybdic  acid. 
The  precipitate  is  extracted  with  dilute  sulphuric  acid 
and  precipitated  with  alcohol.  Solution  and  precipitation 
with  these  reagents  are  repeated  about  fifteen  times,  but 
after  the  ninth  a  substance  of  constant  composition  is 
secured,  and  this  is  a  kyrin.  With  the  cleavage  process 
carried  one  step  farther  these  bodies  are  converted  into 
amino  acids.  It  will  be  seen  from  the  method  of  preparation 
that  they  are  closely  related  to  the  diamino  acids  (arginin 

1  Zeitsch.  f.  physiol.  Chem.,  1906,  xlviii,  54. 


296  PROTEIN  POISONS 

and  lysin).  The  precipitate  with  phosphomolybdic  acid 
may  be  crystallized,  and  the  picrate  differs  from  the  corre- 
sponding salts  of  arginin  and  lysin  by  its  solubility  in  alcohol. 
The  kyrins  give  the  biuret  reaction,  the  color  differing 
from  that  given  by  peptone  in  being  more  distinctly  a 
Bordeaux  red.  Siegfried  states  that  the  kyrin  formed 
from  fibrin  splits  on  further  cleavage  into  lysin,  arginin, 
and  glutamic  acid. 

The  kyrins  are  said  to  be  highly  poisonous,  and  Kam- 
mann1  has  suggested  that  they,  or  similar  bodies,  may  be 
the  active  agents  in  the  production  of  anaphylactic  shock. 
However,  we  have  been  unable  to  find  any  record  of  thor- 
ough studies  of  the  poisonous  action  of  these  cleavage 
products.  Schittenhelm  and  Weichardt2  have  studied 
two  kyrins,  one  prepared  from  hemoglobin  and  the  other 
from  gelatin.  The  latter  is  not  poisonous,  and  this  agrees 
with  our  work  in  which  we  failed  to  obtain  the  poisonous 
group  from  gelatin.  The  globinokyrin  is  moderately 
active,  but  not  so  poisonous  as  the  protamins. 

Anaphylatoxin. — Friedberger3  treated  rabbits  with  lambs' 
serum  until  he  obtained  abundant  precipitates  with  the  sera 
of  these  animals.  These  precipitates  were  deposited  in 
a  centrifuge,  washed  with  salt  solution,  and  then  digested 
with  normal  guinea-pig  serum  for  some  hours  in  the  incu- 
bator. When  this  was  done  and  the  serum  decanted  and 
injected  into  normal  guinea-pigs,  the  animals  promptly 
died  with  all  the  symptoms  of  anaphylactic  shock.  A 
poison  had  already  been  obtained  in  a  similar  manner  by 
Weichardt  from  placental  tissue,  and  by  Friedemann  from 
blood  corpuscles  (see  p.  269).  Friedberger  named  the  poison- 
ous substance  which  he  obtained  by  the  digestion  of  specific 
precipitates  with  normal  serum,  anaphylatoxin.  Desig- 
nating this  substance  as  a  toxin  might  be  criticized,  but 
at  that  time  Friedberger  believed  in  his  theory  of  sessile 
receptors,  and  it  is  plain  that  he  regarded  the  poison  which 


1  Zeitsch.  f.  Immunitatsforschung,  1911,  xi,  659. 

2  Ibid.,  1912,  xiv,  609.  3  Ibid.,  1910,  iv,  636. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     297 

he  had  prepared  as  a  toxin.  One  of  the  conclusions  stated 
in  the  paper  in  which  he  reported  this  work  is  as  follows: 
"Die  Bildung  eines  Antitoxin  gegen  das  Anaphylatoxin 
ist  mir  bischer  noch  nicht  einwandsfrei  gelungen,  jedoch 
ist  sie  wahrscheinlich."  He  also  concluded  that  the  poi- 
sonous action  of  this  substance  is  destroyed  by  a  temperature 
of  65°.  So  far  he  has  not  announced  the  successful  prepara- 
tion of  an  antitoxin,  and  further  research  has  convinced 
him  that  anaphylatoxin  is  thermostable.  In  these  respects, 
therefore,  anaphylatoxin  does  not  differ  from  the  poison 
obtained  by  the  cleavage  of  proteins  with  chemical  agents 
or  ferments.  Later,  Friedberger1  became  convinced  that 
anaphylatoxin  is  not  a  specific  body;  it  is  a  common  product 
of  the  cleavage  of  diverse  proteins.  This,  also,  distinguishes 
it  from  toxins,  one  invariable  characteristic  of  which  is  their 
specificity. 

One  of  the  most  important  contributions  made  to  the 
literature  of  anaphylaxis  is  the  paper  by  Friedberger  and 
Vallardi.2  In  this  contribution  Friedemann  is  properly 
credited  with  having  been  the  first  to  produce  the  anaphyl- 
actic  poison  by  ferment  action  in  vitro.  The  following 
statement  is  made:  "Concerning  the  nature  of  anaphyl- 
atoxin we  know  nothing,  but  we  are  justified  in  assuming 
its  close  relationship  to  the  split  product  obtained  by 
Vaughan  and  Wheeler  through  the  action  of  alkaline 
alcohol  on  proteins,  also  the  similarity  of  its  action  with 
that  developed  by  poisoning  with  peptone,  as  shown  by 
Biedl  and  Kraus,  and  Pfeiffer  and  Mita,  is  evident." 

With  specific  precipitates,  the  stroma  of  blood  corpuscles, 
and  with  whole  corpuscles,  under  the  action  of  ambo- 
ceptor  and  complement,  anaphylatoxin  was  developed, 
and  its  effect  on  fresh  animals  was  demonstrated.  The 
specific  precipitates  were  obtained  by  treating  rabbits  with 
lambs'  blood  and  then  mixing  the  sera  from  these  animals. 
Such  precipitates  were  collected  in  a  centrifuge,  washed 


1  Zeitsch.  f.  Immunitatsforschung,  1910.  vi.  179. 

2  Ibid.,  1910  vii,  94. 


298  PROTEIN  POISONS 

twice  with  salt  solution,  and  then  incubated  with  the 
serum  of  a  normal  guinea-pig.  The  ferment  in  the  serum 
split  up  the  precipitate  with  the  liberation  of  the  poison, 
and  when  the  serum  containing  the  poison  was  injected 
into  a  fresh  guinea-pig  the  animal  promptly  died,  with  all 
the  symptoms  of  anaphylactic  shock.  Washed  stroma 
treated  in  a  similar  way  yielded  the  same  poison.  With 
whole  red  corpuscles  the  results  are  complicated  by  the 
poisonous  action  of  the  liberated  hemoglobin,  which  is  an 
active  poison  without  further  cleavage.  Friedberger  desig- 
nates the  precipitates,  stroma,  and  corpuscles  subjected  to 
the  action  of  the  normal  serum  as  antigens.  We  have  a 
marked  antipathy  to  the  use  of  this  term  in  discussing 
the  phenomena  of  anaphylaxis,  and  would  designate  the 
substances  submitted  to  the  action  of  the  normal  serum  as 
substrates,  and  regard  the  serum  as  containing  the  ferment. 
It  might  be  said  that  the  terms  we  use  are  of  but  little 
importance,  and  the  meaning  is  the  one  important  thing. 
This  is  true,  but  we  use  words  to  express  ideas,  and  we 
hold  that  the  term  "antigen"  in  this  connection  confuses 
and  tends  to  lead  to  gross  misconception.  As  has  been 
shown  by  Friedberger  and  others,  anaphylatoxin  may  be 
obtained  by  incubating  various  proteins  with  normal 
serum,  and  why  should  we  call  the  substances  thus  split 
up  by  the  ferment  in  the  serum  an  antigen?  Would  it  not 
be  just  as  proper  to  denominate  starch  which  is  converted 
into  sugar  by  amylase  an  "antigen?"  But  this  is  a  digres- 
sion, and  we  will  return  to  the  work  of  Friedberger  and 
Vallardi.  They  found  that  when  specific  precipitates  and 
stroma  were  used  the  amount  of  the  poison  obtained  was 
not  in  proportion  to  the  amount  of  substrate  used.  With 
a  larger  amount  of  substrate,  that  of  the  serum  (ferment) 
being  constant,  they  obtained  no  poison.  With  the  amount 
of  substrate  very  small,  they  obtained  either  no  poison  or 
at  least  not  enough  to  demonstrate  its  presence  by  its 
effect  on  the  animal.  They  obtained  positive  results,  as 
shown  by  anaphylactic  death  only  when  the  amount  of 
substrate  but  slightly  exceeded  that  necessary  to  kill  a 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     299 

sensitized  animal  on  reinjection.  When  the  amount  of 
substrate  employed  was  very  small  they  obtained  either 
too  little  of  the  poison  to  affect  the  animal,  or,  what  is  more 
probable,  the  digestion  was  so  active  that  the  poison  itself 
was  destroyed.  When  whole  blood  corpuscles  were  used 
the  amount  of  poison  obtained  did  increase  with  an  increase 
in  the  substrate,  but  the  poison  thus  formed  in  greater 
abundance  was  hemoglobin.  At  least  this  is  our  explana- 
tion of  their  results.  On  the  theory  of  Friedberger,  his  own 
results  are  difficult,  or,  as  we  think,  impossible  of  explana- 
tion, while  on  our  theory  they  explain  themselves,  and, 
in  fact,  are  exactly  what  might  have  been  expected. 

Friedberger  and  Vallardi  find  that  both  the  subjective 
and  objective  symptoms  of  poisoning  with  anaphylatoxin 
are  identical  with  those  of  both  active  and  passive  anaphyl- 
axis,  and  in  this  we  quite  agree  with  them.  Biedl  and 
Kraus  have  held,  and  still  hold,  that  anaphylatoxin  cannot 
be  the  true  anaphylactic  poison  because,  as  they  claim, 
the  condition  of  the  lungs  after  death  in  guinea-pigs  from 
this  poison  is  not  the  same.  We  agree  with  Friedberger, 
who  holds  that  the  distention  of  the  lungs  after  anaphyl- 
actic death,  first  mentioned  by  Gay  and  Southard,  and 
more  fully  emphasized  by  Auer  and  Lewis,  is  found  after 
death  from  naturally  poisonous  sera,  from  poisonous  anti- 
sera,  from  peptone  as  shown  by  Biedl  and  Kraus,  from  the 
poison  of  Vaughan  and  Wheeler,  from  the  /3-i  compound  of 
Barger  and  Dale,  and  possibly  after  poisoning  from  other 
substances  as  well.  Lung  distention  due  to  constriction 
of  the  bronchioles  is,  in  guinea-pigs  at  least,  a  constant 
result  of  the  protein  poison,  but  should  not  be  considered 
as  pathognomonic  of  this  poison.  Friedberger  and  Jeru- 
salem1 attempted  to  isolate  and  study  the  physical  and 
chemical  properties  of  anaphylatoxin.  It  should  be  under- 
stood that  the  poison  as  they  had  it  is  in  guinea-pig  serum. 
They  state  their  conclusions  as  follows:  (1)  The  solution 
can  be  evaporated  to  dryness  (in  vacuo)  without  loss  of 

1  Zeitsch.  f.  Immunitatsforschung,  1910,  vii,  748. 


300  PROTEIN  POISONS 

toxicity.  (2)  By  evaporation  and  resolution  in  smaller 
volume  it  can  be  concentrated.  (3)  It  cannot  be  extracted 
from  the  serum  by  ether  or  chloroform.  (4)  It  can  be 
precipitated  without  loss  of  toxicity  by  alcohol.1  (5)  Ana- 
phylatoxin  is  not  a  globulin.  (6)  It  can  be  obtained  by  the 
action  of  complement  upon  heated  (as  well  as  unheated) 
precipitates. 

Biedl  and  Kraus  claim  that  anaphylatoxin  cannot  be 
the  true  anaphylactic  poison  because  it  does  not  induce 
anaphylactic  shock  when  injected  into  the  brain,  and 
Besredka  has  shown  that  with  serum  the  reinjection  of  a 
very  small  dose  into  the  brain  causes  the  shock.  To  this 
Friedberger  very  properly  replies  that  the  only  form  in 
which  he  has  obtained  the  poison  is  in  solution  in  guinea- 
pig  serum,  and  he  cannot  introduce  a  large  enough  quan- 
tity of  this  into  the  brain  without  doing  mechanical  injury, 
and,  moreover,  the  minimum  reinjection  dose  in  the  brain 
is  not  smaller  than  that  required  intravenously.  It  should 
always  be  borne  in  mind  that  Friedberger's  anaphylatoxin 
is  a  solution  of  a  poison  whose  physical  and  chemical 
properties  are  not  known  in  blood-serum. 

Friedberger2  states  that  after  death  from  anaphylatoxin 
the  blood  does  not  coagulate.  This  seems  to  complete  the 
identity  of  the  action  of  this  poison  with  that  formed  in 
anaphylaxis.  The  symptoms  and  all  the  postmortem 
findings  seem  to  be  identical. 

Friedberger3  and  his  students  showed  that  various  bacteria, 
such  as  the  vibrio  of  Metchnikoff,  the  prodigiosus,  the 
typhoid  and  tubercle  bacillus,  when  incubated  with  normal 
guinea-pig  serum,  furnish  a  soluble  poison  which,  when 
injected  into  normal  animals  intravenously,  causes  anaphyl- 
actic shock  and  death.  The  poison  obtained  from  these 
diverse  bacterial  proteins  as  well  as  that  obtained  from 

1  It  will  be  understood  that  the  proteins  of  the  serum  were  precipitated 
by  the  alcohol  and  the  poison  carried  down  with  the  precipitate.     It  does 
not  mean  that  the  poison  would  necessarily  be  precipitated  from  aqueous 
solution  by  alcohol. 

2  Zeitsch.  f.  Immunitatsforschung,  1910,  viii,  239. 

3  Ibid.,  1911,  ix,  369. 


PROTEIN  SENSITIZATION  OR  .ANAPHYLAXIS     301 

specific  precipitates,  blood  corpuscles,  stroma,  and  other 
proteins,  is  in  all  instances  the  same  in  its  physiological 
action.  It  matters  not  whether  the  bacteria  submitted 
to  the  action  of  the  serum  be  heated  or  unheated,  the 
result  is  the  same.  In  other  words,  Friedberger  and  his 
students  accomplished  with  the  proteolytic  ferments  in 
blood-serum  by  the  cleavage  of  proteins  just  what  we  did 
nearly  ten  years  earlier  by  chemical  agents.  We  demon- 
strated that  all  proteins,  living  or  dead,  formed  or  without 
form,  contain  a  poisonous  group,  and  that  the  physiological 
action  of  this  group  is  the  same  whatever  the  protein  from 
which  it  is  obtained.  We  split  up  the  pathogenic  and  non- 
pathogenic  bacteria,  vegetable  and  animal  proteins  of  the 
most  diverse  kind,  and  obtained  from  each  and  every  one 
the  same  poison,  and  now  the  same  has  been  accomplished 
by  ferments.  This  we  regard  as  a  confirmation  of  our 
statement  made  many  years  ago  that  the  protein  molecule 
contains  at  least  one  poisonous  group.  Besides,  the  poison 
obtained  by  us,  when  we  split  up  proteins  with  chemical 
agents,  is  the  same  or  very  closely  related  to  that  now 
obtained  by  the  cleavage  of  the  same  proteins  by  the  more 
delicate  agency  of  ferment  action.  As  we  stated  at  the 
time,  our  method  was  crude,  and  the  poison  was  obtained 
only  at  great  loss,  but  the  principle  is  the  same.  From  our 
work  we  developed  the  theory  of  the  relation  of  the  split 
protein  products  to  immunity  and  disease,  which  was 
formulated  in  1907,  and  which,  in  our  opinion,  is  confirmed 
in  every  particular  by  the  work  of  Friedberger  and  others. 
We  fail  to  see  why  he  and  our  German  confreres  in  general 
still  use  the  Ehrlich  nomenclature  in  discussing  the  protein 
split  products.  Bacterial  cellular  substance  is  submitted 
to  a  ferment  in  vitro,  and  broken  up,  and  why  should 
this  substance  be  called  "an  antigen"  and  the  ferment 
an  "antibody?"  The  theory  of  sessile  receptors,  the  only 
theory,  so  far  as  we  know  that  Friedberger  ever  originated, 
was  long  ago  demonstrated  to  be  false  by  his  own  work. 
He  has  adopted  another  theory,  one  which  he  certainly 
did  not  originate,  but  for  the  establishment  of  which  he 


302  PROTEIN  POISONS 

has  done  much,  and  still  he  employs  the  language  of  his 
own  theory  long  since  discarded  by  himself. 

Friedberger  and  Nathan1  showed  that  by  the  action  of 
normal  guinea-pigs'  serum  on  normal  horse  serum,  or  vice 
versa,  in  proper  proportions,  a  poison  is  set  free.  The  serum 
that  is  to  serve  as  substrate  is  inactivated  by  being  heated 
to  56°,  and  this  is  then  acted  upon  by  the  ferment  in  the 
unheated  serum.  It  will  be  understood  that  the  amount 
of  the  substrate  must  be  small.  In  fact,  it  was  found  that 
the  poison  is  produced  when  the  substrate  contained  not 
more  than  1  mg.  of  protein.  When  guinea-pig  serum  was 
used  as  the  ferment,  it  was  found  to  act  best  in  quantities 
of  about  6  c.c.  For  instance,  when  inactivated  horse  serum 
in  quantities  of  from  0.01  to  0.0005  c.c.  was  incubated  with 
6  c.c.  of  normal  guinea-pig  serum  for  eighteen  hours,  and 
then  4  c.c.  of  this  injected  intravenously  into  guinea-pigs 
of  about  200  grams,  the  animal  promptly  died  from  anaphyl- 
actic  shock.  On  the  other  hand,  when  guinea-pig  serum 
was  used  as  the  substrate  and  horse  serum  as  the  ferment, 
somewhat  larger  quantities  of  each  were  needed.  For 
instance,  with  0.1  c.c.  of  inactivated  guinea-pig  serum 
incubated  with  8  c.c.  of  horse  serum  for  twenty-four  hours 
a  fatal  amount  of  the  poison  was  obtained,  while  with  the 
substrate  reduced  to  0.01  c.c.  no  symptoms  were  induced. 
Friedberger,  in  reporting  this  work,  expressed  astonishment 
that  from  a  very  small  amount  of  protein,  enormous  quan- 
tities of  which,  in  its  unbroken  state,  could  be  injected 
into  animals  without  recognizable  effect,  there  could  be 
obtained  a  potent  poison,  and  still  years  before  we  had 
split  up  these  proteins  by  chemical  means  and  obtained 
the  same  poison.  What  we  had  done  with  chemical  agents, 
Friedberger  did  with  ferments.  We  claimed  ten  years  ago 
that  we  had  split  the  protein  molecule  along  definite  lines 
of  cleavage,  and  that  our  product  was  not  a  mere  degre- 
dation  body.  This  demonstration  that  ferments  split  the 
protein  molecule  along  the  same  lines  is  a  justification  of 
our  claim. 

1  Zeitsch.  f.  Immunitatsforschung,  1911,  ix,  567. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     303 

This  work  of  Friedberger,  in  our  opinion,  confirms  another 
claim  which  we  put  forth  some  years  ago.  There  are  two 
kinds  of  parenteral  proteolytic  enzymes  in  the  body,  or 
capable  of  being  developed  in  the  body.  One  of  these  is 
non-specific  and  the  other  specific.  The  former  is  found 
in  the  normal  blood  serum,  and  the  latter  is  developed  by 
protein  sensitization.  The  normal  serum  of  this  guinea- 
pig  under  proper  conditions  splits  up  most  diverse  proteins 
with  the  liberation  of  the  poisonous  group.  The  blood- 
serum  and  organ  extracts  of  the  sensitized  animal  contain 
specific  ferments  which  cleave  the  special  protein  to  which 
the  animal  has  been  sensitized.  The  non-specific  protein 
takes  care  of  the  small  amounts  of  foreign  protein  which  are 
constantly  finding  their  way  into  the  blood  without  having 
undergone  digestion.  Ordinarily,  these  enter  the  blood 
in  such  small  amounts  that  they  are  rapidly  and  fully 
digested  beyond  the  poisonous  stage  by  these  non-specific 
proteolytic  ferments.  Among  its  other  functions  the  blood 
is  a  digestive  fluid,  and  it  exercises  this  function  not  only 
on  the  unbroken  proteins  which  find  their  way  into  it  from 
the  outer  world,  but  also  upon  certain  substances  which  are 
constantly  coming  into  it  as  a  result  of  tissue  metabolism. 

Another  important  research  from  Friedberger's  laboratory 
is  reported  by  him  and  Girgolaff.1  Rabbits  and  guinea-pigs 
were  treated  with  homologous  proteins,  bacteria,  and  sera, 
and  after  the  animal  had  developed  the  specific  ferment 
(antibody)  it  was  bled  to  death  by  opening  the  aorta  and 
transfused  with  salt  solution  until  all  the  blood  was  washed 
out.  Then  a  portion  of  some  organ  from  this  animal  was 
implanted  in  the  abdominal  cavity  of  another,  and  after 
recovery  from  the  operation  this  animal  was  found  to  be 
sensitized.  A  few  illustrations  will  best  explain  this  work.  A 
guinea-pig  of  200  grams  received  1  c.c.  of  lambs'  serum  intra- 
venously. Fourteen  days  later  this  animal  was  exsanguinated 
and  washed  out  with  salt  solution.  Then  two  pieces  of  its 
spleen — about  half  of  this  organ— were  implanted  in  the 

1  Zeitsch.  f.  Immunitatsforschung,  1911,  ix,  575 


304  PROTEIN  POISONS 

abdominal  cavity  of  a  fresh  guinea-pig  of  about  the  same 
weight.  Fourteen  days  later  this  animal  received  intra- 
venously 1.5  c.c.  of  lamb  serum,  and  promptly  died  of 
anaphylactic  shock.  Autopsy  showed  the  lungs  distended, 
the  heart  still  beating,  and  the  blood  of  the  right  heart  had 
not  coagulated  at  the  expiration  of  ten  minutes. 

The  organ  may  be  implanted  into  another  species,  as  is 
shown  by  the  following:  A  rabbit  was  treated  with  lamb 
serum  until  the  precipitation  titer  of  the  serum  was  1  to 
1000;  then  the  rabbit  was  exsanguinated  and  perfused  with 
salt  solution,  and  two  pieces  of  its  spleen  implanted  in 
the  abdominal  cavity  of  a  guinea-pig.  Fourteen  days 
later  this  guinea-pig  received  intravenously  1.5  c.c.  of  lamb 
serum,  and  died  of  anaphylactic  shock.  Guinea-pigs  thus 
treated  were  found  to  be  sensitized  not  only  to  lambs,  but 
also  to  rabbit  serum,  thus  proving  that  the  implanted 
organs  continued  to  secrete  both  their  normal  and  their 
specifically  developed  proteolytic  ferments  (antibodies). 
Like  results  were  obtained  when  pieces  of  kidney  were 
transplanted.  Moreover,  the  animals  thus  sensitized  by 
the  receipt  of  transplanted  organs  retained  their  sensitized 
condition  for  a  time  at  least  after  the  removal  of  the  im- 
planted tissue.  The  following  is  an  illustration.  A  guinea- 
pig  received  intravenously  1  c.c.  of  lamb  serum.  Five 
days  later  it  had  two  subcutaneous  injections  of  1  c.c.  of 
lamb  serum.  Eight  days  later  it  was  killed  and  transfused, 
and  portions  of  its  spleen  and  kidney  implanted  in  a  fresh 
guinea-pig.  Six  days  later  the  implanted  tissues  were 
wholly  removed,  and  after  the  guinea-pig  had  fully  recovered 
from  this  operation  it  was  found  to  be  still  sensitized  to 
lamb  serum.  Evidently  the  implanted  tissue  had  not  only 
continued  to  develop  its  specific  ferment,  but  had  discharged 
it  in  part  at  least  into  the  blood.  These  experiments  are 
of  the  highest  value  for  two  reasons:  (1)  They  caused 
Friedberger  to  wholly  abandon  his  theory  of  sessile  receptors, 
and  (2)  they  show  that  the  specific  ferment  developed  in 
protein  sensitization  is  a  cellular  product,  and  that  the 
cells  of  the  spleen  and  kidney,  possibly  of  other  organs  as 
well,  elaborate  it. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      305 

Later  this  work  was  continued  by  Girgolaff,1  who  discusses 
the  possible  explanation  of  these  findings.  (1)  It  might 
be  possible  that  the  recipient  is  passively  anaphylactized 
by  the'  transfer  of  some  serum  from  the  donor.  This  suppo- 
sition is  held  untenable  for  two  reasons.  First,  the  washing 
out  is  so  thoroughly  done,  and  second,  the  volume  of  the 
piece  of  organ  transferred  is  too  small.  Moreover,  if  it  were 
passive  anaphylaxis,  the  recipient  should  be  most  respon- 
sive to  reinjection  within  a  day  or  two,  while  in  fact  it  is 
not  responsive  until  after  seven  or  eight  days.  (2)  It  might 
be  suggested  that  some  of  the  protein  used  in  sensitizing 
the  first  animal  is  carried  over  to  the  second  and  actively 
sensitizes  it.  This  is  highly  improbable  on  account  of  the 
small  amount  of  protein  used  in  sensitizing  the  first  animal; 
the  length  of  time  (in  some  cases  fourteen  days)  elapsing 
before  the  transfer  of  the  tissue,  and  the  thorough  washing 
out  given  the  first  animal.  Besides,  this  was  shown  to  be 
impossible  because  in  one  instance  a  rabbit  was  killed  and 
its  organs  transferred  to  another  rabbit  three  hours  after 
the  former  had  received  a  bacterial  suspension,  just  at  the 
time  when  the  bacilli  should  have  been  most  abundant  in 
the  tissue,  and  these  animals  were  not  sensitized.  (3)  The 
only  conclusion  which  seems  to  have  any  justification  is 
that  the  cells  of  the  tissue  removed,  having  acquired  a  new 
function  while  in  their  normal  location,  continue  to  exercise 
this  function  in  their  new  location.  We  regard  this  as  a  most 
complete  verification  of  our  theory  of  anaphylaxis,  in  which 
we  hold  that  a  new  function  is  developed  in  certain  cells  of 
the  animal  body  by  the  sensitizer.  Moreover,  it  does  seem 
that  this  work  should  lead  to  the  discarding  of  all  theories 
involving  an  "antigenrest,"  about  which  much  has  been 
said. 

Vaughan,  Vaughan,  Jr.,  and  Wright2  demonstrated  that 
the  serum  and  organ  extracts  of  normal  guinea-pigs  do 
not  form  a  poison  when  incubated  with  egg-white,  but 


1  Zeitsch.  f.  Immunitatsforschung,  1912,  xii,  401. 

2  Ibid.,  1911,  xi,  673. 
20 


306  PROTEIN  POISONS 

that  corresponding  preparations  from  guinea-pigs  sensitized 
to  egg-white  do  (p.  274). 

Friedberger  and  Mita1  show  that  summer  frogs  can  be 
anaphylactized.  From  0.1  to  0.5  c.c.  of  lamb  serum  was 
given  by  the  abdominal  vein,  or  into  the  dorsal  lymph  sac 
as  a  sensitizing  dose.  From  one  to  four  weeks  later  a  rein- 
jection  causes  characteristic  symptoms.  The  animal  soon 
becomes  stupid  and  lies  with  extended  limbs.  When  placed 
on  its  back  it  does  not  regain  its  normal  position.  Sudden 
death  does  not  follow,  and  the  animal  usually  survives 
from  twelve  to  twenty-four  hours.  When  the  heart  is 
watched  through  a  fenestrated  chest,  the  pulse  is  seen  to 
grow  slow  and  irregular,  and  finally  the  heart  stops  in 
diastole.  Anaphylatoxin  was  found  to  have  a  similar 
action  on  the  isolated  heart. 

Friedberger  and  Scymanowski2  show  that  the  presence 
of  leukocytes  lessens  the  formation  of  anaphylatoxin,  and 
apparently  destroys  it  when  abundantly  formed.  They 
question  whether  this  is  due  to  an  activity  of  the  leukocyte 
or  to  its  absorption  of  the  poison.  We  suggest  that  the 
leukocytes  destroy  the  poison  by  digesting  it  and  converting 
it  into  a  harmless  body. 

More  than  ten  years  ago  (see  p.  46)  we  showed  that  the 
poison  contained  in  the  cellular  substance  of  the  diphtheria 
bacillus  is  a  wholly  different  thing  from  the  toxin  elaborated 
by  the  same  organism.  This  convinced  us  that  the  protein 
poisons — substances  obtained  by  the  cleavage  of  the  protein 
molecule — are  not  toxins.  When  the  diphtheria  bacillus  grows 
it  elaborates  and  excretes  a  soluble  ferment  known  as  diph- 
theria toxin.  When  injected  into  animals  in  sufficient  doses 
this  toxin  kills  after  from  two  to  five  days.  When  repeatedly 
injected  in  smaller  doses  the  body  elaborates  an  antibody — 
an  antitoxin.  When  the  cellular  substance  of  the  diphtheria 
bacillus  is  split  up  by  our  method  a  poison  is  obtained. 
This  is  not  a  toxin,  but  a  poison.  When  injected  into 


1  Zeitsch.  f.  Immunitatsforschung,  1911,  x,  362. 
*  Ibid.,  1911,  xi,  485. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     307 

animals  in  sufficient  doses  it  kills  within  a  few  minutes — 
not  after  days.  When  repeatedly  injected  in  non-lethal 
doses  the  animal  body  does  not  elaborate  an  antibody — 
an  antitoxin.  Diphtheria  toxin  is  specific;  it  is  an  exclusive 
product  of  the  diphtheria  bacillus,  and  animals  treated 
with  it  in  the  proper  way  produce  a  specific  antibody. 
The  cellular  poison  of  the  diphtheria  bacillus  is  not  specific; 
the  same  poison  is  contained  in  other  proteins,  and  it  gives 
rise  to  no  specific  antibody.  We  convinced  ourselves 
many  years  ago  that  the  protein  poisons  and  the  toxins 
are  not  related  bodies,  and  we  demonstrated  that  diphtheria 
toxin  gives  no  protection  against  poisoning  with  the  active 
substance  contained  in  the  protein  molecules  making  up  the 
cell  substance  of  the  diphtheria  bacillus.  For  this  reason 
we  have  never  discussed  the  phenomena  of  sensitization 
due  to  protein  poisons  in  terms  applicable  only  to  toxin. 
Years  after  we  did  this  work  Friedberger  and  Reiter1  con- 
firmed it  by  showing  that  the  protein  poison  obtainable 
from  the  cellular  substance  of  the  dysentery  bacillus  is  a 
wholly  different  substance  from  the  toxin  of  the  same 
bacillus,  but  they  make  no  mention  of  our  work  in  this 
connection,  and  they  still  call  the  protein  poison  a  toxin 
and  speak  of  the  antibody. 

It  has  been  fully  demonstrated  that  the  protein  poison 
can  be  detached  from  its  combination  in  the  molecules  of 
specific  precipitates,  blood  corpuscles,  stroma,  many  bac- 
teria, etc.,  by  incubation  with  normal  guinea-pig  serum. 
Other  proteins  require  for  their  disruption  and  for  the 
liberation  of  the  poisonous  group,  sera  in  which  specific 
ferments  have  been  developed  by  sensitization.  The 
proteolytic  ferment  in  normal  guinea-pig  serum  is  not 
specific.  It  is  capable  of  digesting  many,  but  not  all, 
proteins.  When  a  guinea-pig  has  been  sensitized  to  a 
given  protein,  its  serum  contains  not  only  the  general, 
non-specific  proteolytic  ferment  normal  to  it,  but  in  addition, 
the  specific  ferment.  Whether  the  latter  is  a  wholly  new 

1  Zeitsch.  f.  Immunitatsforschung,  1911,  xi,  493. 


308  PROTEIN  POISONS 

product  or  is  due  to  a  modification  in. the  former  we  cannot 
say.  For  the  present  we  will  confine  our  attention  to  the 
general,  non-specific  proteolytic  ferment  in  normal  guinea- 
pig  serum.  Like  all  ferments,  it  is  supposed  to  consist  of 
two  parts.  (1)  A  thermostabile  part,  known  as  the  ambo- 
ceptor,  and  (2)  a  thermolabile  part,  known  as  complement 
or  alexin.  The  latter  is  destroyed  by  a  temperature  of  56°, 
and  serum  heated  to  this  point  is  inactivated.  It  has  been 
found  by  most  observers  that  inactivated  serum  from  the 
guinea-pig,  or,  in  fact,  any  inactivated  ferment  solution, 
does  not  function.  Seitz1  found  in  a  few  instances  that 
by  incubating  certain  bacteria  with  inactivated  serum  he 
obtained  a  free  poison,  but  this  is  contradicted  by  the 
experience  of  so  many  investigators  that  we  must  conclude 
that  his  technique  was  defective.2  There  are,  however, 
some  points  of  real  interest  in  connection  with  this  reaction. 
Since  bacteria  and  certain  other  proteins,  when  incubated 
with  normal  serum,  yield  a  soluble  and  active  poison,  why 
does  this  reaction  not  occur  when  these  proteins  in  the 
unbroken  condition  are  injected  directly  into  the  blood? 
The  only  answer  to  this  question  seems  to  be  that  the  ferment 
is  in  a  more  readily  available  form  in  the  serum  than  it  is 
in  blood.  The  ferment  in  the  serum  probably  comes  largely 
from  the  breaking  down  of  leukocytes.  When  an  unbroken 
protein  is  injected  into  an  animal  usually  the  first  effect 
upon  the  blood  is  a  leukopenia.  Certainly  there  is  for  a 
time  a  diminution  of  the  leukocytes  in  the  peripheral  blood. 
After  a  time  there  is  a  hyperleukocytosis,  and  this  is  generally 
believed  to  be  for  the  purpose  of  breaking  up  the  foreign 
protein.  There  is,  however,  another  possible  explanation. 
It  may  be  that  in  the  circulatory  blood  the  disruption  is 
carried  beyond  the  point  of  setting  free  the  poison.  It  may 
itself  be  disrupted  and  converted  into  relatively  harmless 
substances.  There  is  also  the  possibility  that  the  inclusion 
of  the  foreign  protein  by  the  phagocyte  may  delay  the 
disruption  of  the  former. 

1  Zeitsch.  f.  Immunitatsforschung,  1911,  xi,  588. 

2  See  article  by  Lura,  ibid.,  1912,  xii,  467. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     309 

It  seems  that  all  bacteria,  both  pathogenic  and  non- 
pathogenic,  at  least  so  far  as  tested,  yield  a  poison  when 
incubated  with  normal  serum  of  the  guinea-pig.  Some  are 
acted  upon  or  disrupted  more  promptly  and  quickly  than 
others,  but  with  prolonged  incubation  all  yield  enough 
of  the  poison  to  affect  animals  to  a  recognizable  extent. 
This  confirms  our  work  in  which  it  was  shown  that  all 
proteins,  so  far  as  tested,  yield  a  poisonous  fraction  when 
properly  disrupted  by  chemical  agencies.  It  was  this 
work  that  led  us  to  conclude  that  every  protein  molecule 
contains  a  poisonous  group.  Bold  and  Aoki1  have  obtained 
the  poison  by  incubation  with  serum  from  streptococci, 
meningococci,  gonococci,  b.  mallei,  pestis,  pneumonise, 
paratyphus,  chicken  cholera,  swine  erysipelas,  yeast  cells, 
actinomyces,  the  spirochetes  of  chicken  spirillosis,  and 
Russian  relapsing  fever.  They  did  fail  to  obtain  it  from 
the  spores  of  certain  molds,  but  this  does  not  prove  that 
these  spores  do  not  contain  a  poison.  It  simply  shows  that 
the  proteins  of  these  spores  are  resistant  to  the  cleavage 
action  of  the  ferment  contained  in  the  normal  serum  of 
the  guinea-pig. 

Boehneke  and  Bierbaum2  find  that  repeated,  alternate 
freezing  and  thawing  have  no  effect  upon  either  the  substrate 
or  the  ferment  in  the  production  of  the  poison,  and  no 
effect  on  the  poison  itself. 

Bessau3  sensitized  animals  simultaneously  with  ox  and 
horse  sera.  The  injections  were  made  on  each  side  of  the 
thorax  subcutaneously.  After  full  sensitization  had  been 
developed,  as  was  demonstrated  by  reinjection  of  controls 
which  had  been  sensitized  to  only  one  serum,  those  doubly 
sensitized  were  given  sublethal  reinjections  of  one  of  the 
sera,  and  after  recovery  from  the  effect  induced  they  were 
found  to  be  less  susceptible  to  reinjections  of  the  second 
serum.  He  also  determined  the  minimum  fatal  dose  of 
anaphylatoxin  prepared  from  typhoid  bacilli  on  fresh 

1  Zeitsch.  f.  Immunitatsforschung,  1912,  xii,  200. 

2  Ibid.,  1912,  xiv,  130. 

3  Centralbl.  f.  Bakteriol.,  1911,  Ix,  637. 


310  PROTEIN  POISONS 

guinea-pigs,  and  found  that  animals  sensitized  to  a  serum, 
and  given  a  non-fatal  reinjection  of  the  homologous  serum, 
after  recovery  survived  the  minimum  fatal  dose  of  anaphyl- 
atoxin.  From  these  experiments  Bessau  reached  the 
following  conclusions:  (1)  The  condition  or  state  of  anti- 
anaphylaxis  is  not  specific.  (2)  It  is  not  due  to  absorption 
of  the  ferment  (antibody).  (3)  It  is  due  to  increased  toler- 
ance of,  or  lessened  susceptibility  to,  the  poison.  Friedberger 
and  his  students1  have  taken  up  these  points,  and  by  exact 
quantitative  experiments  have  demonstrated  that  the 
state  of  anti-anaphylaxis,  like  that  of  anaphylaxis,  is 
strictly  specific,  but  that  it  is  true  that  increased  tolerance 
does  play  a  part  in  the  experiments  as  made  by  Bessau. 
When  an  animal  is  simultaneously  sensitized  to  two  sera, 
and  after  the  condition  of  sensitization  has  been  fully 
developed,  a  non-fatal  reinjection  of  one  of  these  sera 
renders  the  animal  after  recovery  absolutely  insusceptible 
to  any  dose  of  the  serum  which  has  been  employed  in  the 
reinjection,  but  leaves  it  still  susceptible  to  the  second  serum 
in  doses  only  slightly  larger  than  those  required  to  kill 
control  animals  sensitized  to  that  serum  only.  We  can 
make  this  plainer  by  the  following  statement.  When  an 
animal  is  sensitized  to  two  sera,  two  specific  proteolytic 
ferments  are  developed.  When  an  animal  in  full  sensitiza- 
tion to  both  sera  is  treated  with  a  non-lethal  reinjection  of 
one  of  them,  the  specific  ferment  for  this  serum  is  exhausted, 
and  a  certain  amount  of  the  poison  is  set  free,  not  enough 
to  kill  the  animal,  but  enough  to  give  the  animal  increased 
tolerance  to  the  poison.  Consequently  the  fatal  dose  of  the 
other  serum  necessary  to  kill  on  reinjection,  say,  twenty-four 
hours  later,  is  larger  than  the  minimum  fatal  dose  when  the 
animal  has  been  sensitized  to  only  one  serum.  We  demon- 
strated (see  page  139)  many  years  ago  not  only  that  tolerance 
to  the  protein  poison  can  be  increased,  but  that  resistance 
to  living  cultures  of  pathogenic  bacteria  may  be  increased 
by  repeated  doses  of  the  poison.  Furthermore,  we  showed 

1  Zeitsch.  f.  Immunitatsforschung,  1912,  xiv,  371. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      311 

that  in  neither  of  these  instances  is  the  action  specific, 
nor  does  the  poison  have  the  action  of  a  toxin,  nor  is  the 
increased  tolerance  of  it  due  to  the  production  of  antitoxin. 
Repeated  treatment  of  animals  with  the  poison,  beginning 
with  a  sublethal  dose  and  gradually  increasing  the  dose, 
may  enable  the  animal  to  bear  three  or  four  times  the 
minimum  lethal  dose,  as  tested  on  fresh  animals,  but  the 
effect  induced  is  never  quantitatively  comparable  to  that 
obtained  by  similar  treatments  with  increasing  doses  of  toxin. 
Besides,  we  were  never  able  to  find  any  evidence  of  the 
presence  of  an  antitoxin  in  the  blood  serum  of  the  treated 
animal.  For  these  reasons  we  decided  years  ago  that  the 
protein  poison  is  not  a  toxin.  Moreover,  we  found  that  the 
increased  resistance  to  typhoid  infection  came  just  as 
promptly  and  was  as  marked  when  the  animal  was  treated 
with  poison  obtained  from  egg-white  as  that  obtained  by 
repeated  treatments  with  the  poison  split  off  from  the 
cellular  substance  of  the  typhoid  bacillus.  This  demon- 
strated that  the  tolerance  obtained  to  the  protein  poison 
is  not  specific.  This  is  another  clear  proof  that  the  poisonous 
group  contained  in  the  protein  molecule  is  not  a  toxin. 
Years  after  our  work  had  been  done  and  reported,  Fried- 
berger1  found  that  after  a  guinea-pig  had  recovered  from 
severe  poisoning  with  his  anaphylatoxin,  it  would  bear  a 
certainly  fatal  dose  of  the  same,  and  at  that  time  he  thought 
that  he  had  secured  a  toxin-antitoxin  immunity.  Later 
still  H.  Pfeiffer2  found  that  the  urine  of  an  anaphylactized 
guinea-pig  is  highly  poisonous;  also,  that  treating  a  sensi- 
tized guinea-pig  with  such  urine  made  it  more  resistant 
on  reinjection;  also,  that  after  recovery  from  anaphylactic 
shock,  guinea-pigs  are  more  resistant  to  the  poison  in  the 
urine  of  anaphylactized  animals.  He  also  thought  that 
he  had  established  a  toxin-antitoxin  immunity,  but  if  we 
read  their  later  works  with  correct  interpretation,  neither 
Friedberger  nor  H.  Pfeiffer  now  believe  that  the  protein 


1  Zeitsch.  f.  Immunitatsforschung,  1910,  iv,  636. 

2  Ibid.,  1911,  x,  550. 


312  PROTEIN  POISONS 

poison  is  a  toxin,  in  the  sense  of  diphtheria  or  tetanus 
toxin,  though  both  continue  to  call  it  a  toxin.  The  work  of 
Bessau  and  Friedberger  confirms  ours  and  establishes 
beyond  any  doubt  that  the  increased  tolerance  brought 
about  by  repeated  administrations  of  the  protein  poison 
by  direct  injection  or  by  recovery  from  anaphylactic  shock 
is  not  of  the  nature  of  a  toxin-antitoxin  immunity. 

It  seems  to  us  that  there  is  one  point  about  anti-anaphyl- 
axis  which  both  Bessau  and  Friedberger  fail  to  see.  When 
a  sensitized  animal  is  reinjected  with  the  homologous 
protein  and  recovers,  it  immediately  loses,  for  a  time  at 
least,  its  responsiveness  to  the  same  protein.  As  Fried- 
berger has  shown,  injections  of  two  hundred  times  the 
amount  necessary  to  kill  the  sensitized  animal  is  without 
effect.  Indeed,  the  animal  seems  to  be  returned  suddenly 
to  the  condition  of  a  fresh  animal,  one  which  has  never 
received  a  protein  injection.  The  usual  explanation  is 
that  all  the  specific  ferment  in  the  sensitized  animal  has 
been  exhausted  by  the  non-fatal  reinjection.  Bessau 
thinks  it  due,  as  we  have  seen,  to  a  decreased  susceptibility 
to  the  poison,  or,  as  we  say,  to  an  increased  tolerance  of 
the  poison.  Both  of  these  are  undoubtedly  factors,  and 
they  may  be  the  most  important  factors  in  the  sudden 
development  of  the  anti-anaphylactic  state,  but  they  are 
not  the  only  factors,  and  we  are  inclined  to  the  opinion 
that  they  are  not  the  most  important.  That  the  specific 
ferment  (the  antibody  of  other  writers)  is  not  wholly 
exhausted  is  shown  by  the  fact  that  the  blood  serum  of 
an  animal  in  the  anti-anaphylactic  state,  when  transferred 
to  a  fresh  animal,  sensitizes  the  recipient.  This  could  not 
be  if  the  ferment  had  been  wholly  used  up.  There  must 
still  be  active  ferment  in  the  portion  of  blood  serum  trans- 
ferred. The  factor  which  we  suspect  of  being  of  greatest 
importance  in  the  production  of  the  anti-anaphylactic 
state  is  the  changed  relation  between  the  amount  of  ferment 
and  the  substrate.  This  is  one  of  the  most  important  and 
interesting  problems  connected  with  protein  sensitization. 
In  our  earliest  work  with  the  cellular  bacterial  poisons, 


PROTEIN  SENSIT1ZATION  OR  ANAPHYLAXIS     313 

when  suspensions  were  injected  into  the  abdominal  cavity 
we  found  that  large  doses  often  failed  to  kill,  or  killed 
slowly,  while  smaller  doses  killed  more  certainly  and  more 
promptly;  then  we  found  that  grinding  our  cellular  sub- 
stances more  finely  increased  their  toxicity;  later  we  found 
that  the  introduction  of  large  quantities  of  egg-white  into 
fresh  animals  was  without  visible  effect,  while  the  repeated 
injection  of  very  small  doses  produced  prompt  effects  and 
speedily  killed.  Later  still  we  ascertained  that  when  a 
small  amount  of  the  blood  serum  of  a  guinea-pig  sensitized 
to  egg-white  was  incubated  with  from  1  to  5  mg.  of  egg-white 
in  vitro  we  obtained  an  active  poison,  but  when  the  amount 
of  egg-white  present  was  greatly  increased  there  was  no 
evidence  of  the  production  of  a  poison.  Friedberger  has 
repeatedly  met  with  the  same  thing  in  preparing  his  ana- 
phylatoxin.  With  a  small  amount  of  ferment  and  an 
excessive  amount  of  substrate  the  reaction  is  impeded. 
Then  the  presence  and  accumulation  of  the  products  of 
fermentation  retard  the  fermentative  process.  The  con- 
centration of  the  ferment,  the  substrate,  and  the  products 
of  fermentation  all  influence  the  rapidity  wTith  which  the 
fermentative  process  proceeds,  and  all  of  these  are  altered 
when  a  few  cubic  centimeters  of  the  blood  serum  of  an 
animal  in  the  anti-anaphylactic  state  is  transferred  to  a 
fresh  animal  and  the  latter  receives  an  injection  of  the 
proper  protein.  Besides,  it  is  possible  that  in  the  prepara- 
tion of  the  serum  the  amount  of  available  ferment  is  increased 
by  disruption  of  the  leukocytes. 

Friedberger  at  first  stated  that  his  anaphylatoxin  is 
thermolabile.  If  this  be  true  it  cannot  be  identical  with  or 
very  closely  related  to  our  protein  poison,  which  is  ther- 
mostabile.  Later,  Friedberger  found  that  in  acid  solution 
anaphylatoxin  is  thermostabile.  It  is  well  to  see  how  these 
differences  can  be  reconciled.  It  must  be  understood  that 
anaphylatoxin  has  never  been  isolated,  not  even  partially, 
from  the  serum  in  which  it  is  formed,  and  of  course  the 
serum  is  alkaline.  Years  ago  we  showed  that  our  poison 
in  alkaline  solution  decreases  in  toxicity,  and  that  this 


314  PROTEIN  POISONS 

decrease  takes  place  rapidly  at  high  temperature.  We 
showed  this  by  making  aqueous  solutions  of  the  poison 
alkaline  with  sodium  bicarbonate  and  keeping  them  in  the 
incubator  for  varying  periods.  Friedberger  finds  that 
when  he  heats  the  alkaline  serum  containing  anaphylatoxin 
to  65°,  its  toxicity  decreases,  but  when  the  serum  is  made 
acid  it  may  be  heated  to  100°  without  appreciable  loss  in 
toxicity.  It  will  be  seen,  therefore,  that  the  two  substances 
behave  in  the  same  manner  when  heated  in  alkaline  solution. 
We  never  supposed  that  the  heat  destroyed  our  poison, 
but  that  on  combination  with  alkali,  which  combination  is 
hastened  by  heat,  it  becomes  less  poisonous,  and  Friedberger 
has  failed  to  show  that  this  is  not  true  of  anaphylatoxin. 

There  is  another  striking  point  of  similarity  between  our 
poison  and  Friedberger's  anaphylatoxin,  and  in  this  par- 
ticular both  substances  show  a  close  relationship  to  peptone. 
It  has  long  been  known  that  when  an  animal  is  quite  fully 
under  the  influence  of  peptone  the  further  administration 
of  peptone  has  but  little  effect.  This  is  true  of  both  our 
poison  and  anaphylatoxin.  We  have  designated  this  as 
tolerance,  but  it  must  be  admitted  that  it  is  an  unusual 
form  of  tolerance  and  it  needs  further  investigation. 

Besredka,  Strobel,  and  Jupilli1  refuse  to  accept  anaphyl- 
atoxin as  the  true  anaphylactic  poison  because  its  adminis- 
tration to  sensitized  animals  does  not  induce  the  so-called 
anti-anaphylactic  state.  Of  course  it  does  not  and  should 
not  be  expected  to  do  so.  The  anti-anaphylactic  state  is 
due  to  the  partial  exhaustion  of  the  specific  proteolytic 
ferment,  and  the  retarding  effects  of  the  products  of  digestion 
on  the  remaining  ferment.  Administration  of  the  poison 
itself,  already  formed,  uses  up  none  of  the  ferment,  and  the 
other  products  of  the  cleavage  action,  besides  itself,  are 
not  present.  Another  reason  the  French  investigators 
give  for  concluding  that  anaphylatoxin  is  not  the  true 
anaphylactic  poison  is  that  the  former  is  not  specific  in 
origin  and  may  be  obtained  equally  from  diverse  proteins; 

1  Zeitsch.  f.  Immunitatsforschung,  1913,  xvi,  250. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     315 

certainly  it  can  be  obtained  from  all  true  proteins,  as  we 
demonstrated  many  years  ago.  The  specificity  of  an  infec- 
tious disease  does  not  lie  in  the  poison  which  is  formed,  but 
in  the  ferment  by  which  it  is  formed.  The  same  poison  is 
contained  in  all  bacteria,  pathogenic  and  non-pathogenic, 
indeed,  in  all  proteins,  but  there  are  specific  ferments  which 
break  up  one  protein  more  readily  and  more  completely 
than  other  ferments.  The  specificity  lies  in  neither  the 
substrate,  except  that  it  must  be  a  protein,  nor  in  the 
cleavage  product,  but  in  the  agent  that  effects  the  cleavage. 

Physiological  Action  of  the  Protein  Poison. — Edmunds1 
has  made  the  most  thorough  study  of  the  protein  poison, 
as  prepared  by  Vaughan  and  Wheeler,  reported  up  to  the 
present  time.  •  His  experiments  were  made  on  dogs  and 
with  the  "crude  soluble  poison"  made  from  casein.  This 
preparation  contains  something  less  than  10  per  cent,  of 
the  poison  in  the  purest  form  in  which,  so  far,  we  have  been 
able  to  obtain  it,  and  this  is  not  chemically  pure.  On 
account  of  its  importance  we  make  the  following,  somewhat 
lengthy,  abstract  from  the  paper  by  Edmunds. 

Intravenous  injections  in  intact  dogs  are  reported  as 
follows:  "The  most  prominent  symptoms  were  a  marked 
depression,  disturbance  of  the  alimentary  canal,  and  some 
respiratory  disturbances,  the  latter  consisting  of  slight 
acceleration  with  a  slightly  labored  expiration.  In  some 
animals  the  respiratory  symptoms,  with  the  exception  of 
the  slight  acceleration,  were  scarcely  noticeable.  A  study 
of  these  symptoms  shows  that  they  resemble  closely  those 
exhibited  by  dogs  which  are  suffering  from  anaphylactic 
shock,  although  they  are  milder  than  those  described  by 
Pearce  and  Eisenbrey  and  others." 

The  effect  on  the  circulatory  system  was  studied  upon 
dogs  anesthetized  with  morphine  and  paraldehyde.  Blood 
pressure  was  measured  from  the  carotid,  and  the  respiration 
recorded  by  a  tambour  resting  against  the  chest  wall  and 
connected  with  a  second  one  by  which  the  movements  were 

1  Zeitsch.  f.  Immunitatsforschung,  1913,  xvii,  105. 


316  PROTEIN  POISONS 

traced  upon  blackened  paper.  Injection  into  the  external 
jugular  ,vein  of  the  soluble  portion  of  100  mg.  of  the  crude 
poison  was  followed  immediately  by  a  slow  decline  in  blood 
pressure,  amounting  to  from  6  to  8  mm.  Hg.;  after  about 
twenty  seconds  the  fall  became  rapid,  passing  from  a  normal 
of  about  72  mm.  to  about  20  mm.  Synchronously  with 
the  fall  in  pressure  the  heart-beat  was  at  first  slightly 
accelerated,  passing  from  138  or  140  to  144,  but  when  the 
pressure  reached  the  low  point  the  heart-rate  dropped  to 
92.  The  respiration  was  but  little  changed  in  rate,  becoming 
slightly  slower  with  the  fall  in  pressure,  but  the  strength 
was  considerably  decreased,  neither  inspiration  nor  expira- 
tion being  as  complete  as  normal. 

The  blood  pressure  was  slow  to  recover.  In  some  instances 
there  was  an  increase  of  only  a  few  millimeters  after  thirty 
minutes.  When  under  the  full  influence  of  the  poison, 
stimulation  of  either  the  sciatic  or  the  great  splanchnic 
nerve  with  the  induced  current  elicited  no  response,  showing 
peripheral  paralysis  of  the  vasomotors.  In  pithed  animals 
(with  the  brain  and  cord  destroyed)  the  effect  upon  blood- 
pressure  was  the  same  as  on  whole  animals,  only  that  the 
initial  pressure  being  small,  the  fall  did  not  measure  so 
many  millimeters.  "That  this  action  was  peripheral  to 
the  ganglia  along  the  course  of  the  constrictor  fibers  was 
proved  by  the  use  of  large  doses  of  nicotine,  sufficient  being 
given  to  paralyze  them.  When  this  stage  was  reached  an 
injection  of  the  poison  still  produced  the  characteristic 
fall. 

"The  localization  of  the  point  of  action  of  the  poison 
upon  nerve  ending,  receptive  substance,  or  muscle  wall 
was  studied  with  the  aid  of  nicotine,  epinephrin,  and  digitalis. 
The  action  of  nicotine  was  greatly  weakened  by  the  previous 
injection  of  the  poison.  Where  before  the  poisoning,  nico- 
tine had  given  a  marked  increase  in  pressure  in  the  charac- 
teristic manner,  following  300  mg.  of  the  poison  5  mg.  of 
nicotine  raised  the  pressure  from  16  mm.  to  only  62  mm. 
The  heart-rate  was  increased  by  the  nicotine  in  the  usual 
manner,  from  120  per  minute  to  216.  The  pressure  curve 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS      317 

from  the  nicotine  was  not  only  lower  than  that  usually 
seen  with  such  doses,  but  was  much  altered  in  shape,  there 
being  a  very  slow  rise  in  place  of  the  precipitous  increase 
commonly  obtained.  The  injection  of  nicotine  was  followed 
after  a  short  time  by  a  dose  of  epinephrin  which  raised  the 
pressure  from  22  to  220  mm.  Further  injections  of  large 
doses  of  the  poison  were  followed  by  repeated  injections 
of  epinephrin  which  raised  the  pressure  from  18  to  225  mm. 
These  experiments  seemed  to  point  conclusively  to  the 
nerve  endings  as  being  the  structure  primarily  acted  upon 
by  the  poison,  as  evidently  the  receptive  substance  which 
is  stimulated  by  the  epinephrin  had  not  been  paralyzed  by 
the  poison.  In  addition  to  epinephrin,  both  digitalis  and 
barium  chloride  raised  the  lowered  blood  pressure  very 
satisfactorily,  demonstrating  that  the  muscle  cell  in  these 
cases  was  not  affected  by  the  doses  of  poison  given.  In 
some  animals,  however,  the  use  of  large  doses  ,of  the  poison 
was  followed  by  a  lessened  response  to  epinephrin  and 
digitalis,  thus  showing  that  while  the  nerve  ends  are  first 
affected,  the  effect  of  large  doses  is  not  necessarily  confined 
to  these  structures,  but  may  spread  to  the  receptive 
substance  and  contractile  substance  proper." 

Small  doses  (25  mg.)  of  the  poison  have  but  little  effect 
upon  the  heart,  beyond  a  temporary  increase  in  systole  and 
diastole  in  both  auricle  and  ventricle.  This  increase  is 
followed  in  about  a  minute  by  some  weakening  in  both 
chambers,  most  marked  in  the  auricle,  both  systole  and 
diastole  being  decreased.  With  larger  doses  (100  to  150 
mg.)  the  same  changes  are  produced,  the  only  difference 
being  one  of  degree.  The  increase  in  systole  in  both  auricle 
and  ventricle  is  quite  marked  in  some  cases,  while  the 
change  in  the  extent  of  the  dilatation  is  not  so  great,  but 
is  present  in  most  cases.  These  changes  lead  to  an  increased 
amplitude  of  beat  which  lasts  usually  from  one  to  two 
minutes,  until  the  blood-pressure  has  reached  its  lowest 
limit.  The  fall  in  pressure  coming  on  while  the  strength  of 
the  beat  is  increased  finds  no  explanation  in  the  behavior 
of  this  organ. 


318  PROTEIN  POISONS 

"The  changes  which  the  heart  undergoes,  following  the 
increase  in  systole  described,  consist  in  a  weakening  in  both 
chambers,  while  the  extent  of  dilatation  may  be  still  further 
increased,  or  it  may  show  little  change.  Two  factors  come 
in  to  complicate  the  changes  produced  by  the  poison.  First, 
the  great  fall  in  blood  pressure  produced  by  it  decreases 
the  resistance  against  which  the  heart  has  to  contract;  and 
second,  the  changes  in  respiration,  which  at  times  produce 
a  mild  degree  of  asphyxia.  It  happened  in  some  cases, 
before  the  poison  was  given,  the  artificial,  respiration  had 
seemed  entirely  adequate,  after  its  administration  the 
lungs  did  not  inflate  so  well  and  the  blood  showed  signs  of 
deficient  aeration.  The  heart  changes  were  therefore  studied 
further  in  animals  in  which  all  other  organs,  save  the  lungs, 
were  excluded  from  the  circulation.  This  experiment  was 
carried  out  on  a  large  bulldog  anesthetized  in  the  usual 
way.  Under  artificial  respiration  the  sternum  was  cut 
lengthwise,  the  two  halves  pulled  apart,  and  the  heart 
exposed.  The  large  vessels  at  the  base  of  the  heart  were 
dissected  free  and  a  clamp  placed  in  position  on  the  aorta 
just  below  the  origin  of  the  left  subclavian  artery,  this  and 
the  right  subclavian  being  tied.  A  loose  clamp  was  also 
placed  on  the  inferior  vena  cava  just  above  the  diaphragm. 
Cannulas  were  placed  in  both  common  carotids  and  external 
jugular  veins.  The  cannulas  in  the  left  carotid  and  the 
right  jugular  were  paraffined  and  connected  by  a  short 
paraffined  rubber  tube,  thus  providing  a  channel  for  the 
blood  from  the  left  to  the  right  heart.  The  cannula  in  the 
right  carotid  was  connected  with  the  mercury  manometer  to 
record  the  blood  pressure,  and  that  in  the  left  jugular  was 
used  to  inject  the  poison.  The  clamps  on  the  aorta  and 
the  vena  cava  were  now  closed,  thus  shutting  off  the  circu- 
lation in  all  parts  of  the  body,  except  the  heart  and  lungs. 
The  pericardium  was  opened  and  sewed  to  the  sides  of 
the  chest  to  form  a  sort  of  cradle  for  the  heart.  The  myo- 
cardiograph  was  attached  to  the  right  auricle  and  ventricle 
in  the  usual  manner,  and  arranged  to  record  their  move- 
ments on  the  kymograph."  After  recovery  from  the  opera- 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     319 

tion  the  pressure  stood  at  105  mm.  Injection  of  the  poison 
caused  a  fall  of  only  2  or  3  mm.  Hg.  In  this  way  it  was 
shown  that  the  heart  and  lungs  were  not  responsible  for 
the  fall  in  pressure  observed  in  the  intact  animals. 

Isolated  organs  were  perfused  with  solutions  of  the 
poison  and  a  dilatation  of  the  vessels,  probably  due  to 
paralysis  of  the  vasomotor  mechanism,  was  observed. 
This  paralysis  did  not  disappear  after  subsequent  washing 
with  Ringer's  solution,  but  did  so  promptly  on  the  use  of 
epinephrin.  The  perfusion  experiments,  therefore,  indicate 
a  local  paralyzing  effect  upon  the  vessel  walls. 

Edmunds  demonstrated  by  careful  experimentation1 
that  the  liver  dilates  with  the  fall  in  blood-pressure.  This 
seems  to  settle  the  question  of  the  distribution  of  the  blood 
as  the  pressure  falls.  "The  fall  is  due  primarily  to  a 
peripheral  paralysis  of  the  vasomotors  running  in  the 
splanchnic  nerves."  The  spleen,  kidneys,  and  intestine 
do  not  show  increase  in  volume,  as  the  blood  is  drained 
from  these  organs  into  the  capacious  blood  channels  of  the 
liver.  Other  vascular  areas  besides  those  innervated  by 
the  splanchnics  are  affected.  This  was  shown  by  the  fact 
that  when  the  poison  was  injected  into  white  dogs  the  skin 
over  the  thorax  and  abdomen  and  down  on  the  legs  became 
bright  pink.  When  the  liver  was  excluded  from  the  circu- 
lation the  fall  in  blood  pressure  occurred,  but  less  promptly. 

Respiratory  changes  in  the  dog,  due  to  the  poison,  are 
not  marked.  The  usual  effects  are  slight  acceleration  and 
weakening.  With  the  chest  walls  open  and  under  artificial 
respiration,  there  would  be,  at  times,  signs  of  asphyxiation 
which  were  easily  relieved  by  a  slightly  stronger  pressure 
on  the  bellows.  The  most  marked  change  in  the  blood 
picture  observed  was  a  diminution  in  the  eosinophiles,  both 
relatively  and  absolutely. 

It  has  been  observed  by  all  who  have  studied  the  action 
of  peptone  and  the  protein  poison,  that  after  the  blood- 


1  The  details  can  be  found  in   Zeitsch.   f.   Immunitatsforschung,  1913, 
xvii,  105. 


320  PROTEIN  POISONS 

pressure  has  fallen  to  the  lowest  limit  the  further  adminis- 
tration of  the  peptone  or  poison  is  without  effect. 

Edmunds  closes  his  studies  with  the  following  conclu- 
sions: "The  toxic  portion  of  the  split  protein  molecule  as 
described  by  Vaughan  and  Wheeler  produces  in  dogs  when 
injected  intravenously  the  same  symptoms  as  are  seen  in 
these  animals  when  suffering  from  acute  anaphylactic 
shock.  An  analysis  of  the  changes  shows  the  same  rapid 
fall  in  blood  pressure  due  mainly  to  paralysis  of  the  vaso- 
motor  endings  of  the  splanchnic  nerves.  The  blood  does 
not  accumulate  at  the  time  of  the  fall  in  pressure  in  the 
intestines  or  kidneys,  but  is  drained  from  them  into  the 
liver,  and  probably  into  the  large  abdominal  veins.  There 
is  no  evidence  of  a  constriction  of  the  pulmonary  vessels, 
nor  of  lack  of  blood  to  the  left  side  of  the  heart.  In  these 
points  the  action  of  the  protein  poison  agrees  with  the 
changes  described  in  anaphylactic  shock,  but  whereas  with 
the  latter  the  ability  of  the  blood  to  coagulate  may  be  lost, 
this  is  not  affected  by  the  split  product." 

General  Physiological  Action  of  Proteins. — Schittenhelm 
and  Weichardt1  conclude  a  study  of  this  subject  as  follows: 
The  compound  proteins,  as  such,  are  relatively  inactive. 
In  the  doses  employed  they  give  rise  to  no  symptoms  and 
do  not  affect  blood  pressure.  Their  components  (the 
globulins,  histons,  and  protamins)  are  highly  poisonous 
compared  with  the  native  simple  proteins.  They  cause  a 
marked  fall  in  blood-pressure,  delay  blood  coagulation, 
influence  respiration  and  temperature,  and  in  small  doses 
may  cause  death.  This  is  true  even  when  they  are  of 
homologous  origin.  From  the  composition  of  the  protamins 
and  histons  it  has  been  inferred  that  their  poisonous  action 
is  connected  in  some  way  with  their  large  diamino  acid 
content,  but  the  globins  do  not  contain  a  large  amount  of 
these  acids.  On  the  other  hand,  as  has  been  stated,  the 
kyrins  contain  a  large  amount  of  diamino  acids  and  are 
not  so  poisonous  as  the  protamins  and  histons.  It  should 

1  Zeitsch.  f.  Immunitatsforschung,  1912,  xiv,  609. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     321 

be  remarked  that  while  the  globins  do  not  contain  a  large 
amount  of  the  diamino  acids,  they  are  rich  in  the  closely 
related  body,  histidin,  and  the  relation  of  this  to  the  highly 
poisonous  /3-i  body  of  Barger  and  Dale  has  been  mentioned. 
Certainly  there  is  reason  for  suspecting  that  the  poisonous 
group  or  groups  in  the  protein  molecule  has  some  close 
chemical  relationship  to  the  diamino  acids. 

Sensitization  is  Cellular. — Dale1  has  shown  that  the  plain 
muscle  of  a  sensitized  guinea-pig  contracts  when  touched 
with  a  dilute  solution  of  the  homologous  protein.  This 
demonstrates  that  sensitization  is  cellular.  Dale  states 
his  conclusions  as  follows:  "(1)  Plain  muscle  from  an 
anaphylactic  guinea-pig,  freed  from  all  traces  of  blood  and 
serum,  has  a  very  high  degree  of  sensitiveness  to  the  specific 
sensitizing  protein.  The  plain  muscle  of  the  virgin  uterus 
is  essentially  suited  to  the  demonstration  of  the  condition, 
and  exhibits  a  definite  rise  of  tonus  in  response  to  extreme 
dilutions  of  the  antigen.  (2)  The  effect  is  practically  imme- 
diate, i.  e.,  the  delay  is  not  obviously  more  than  can  be 
attributed  to  the  method  of  application  of  the  antigen. 
(3)  The  response  is  not  a  mere  exaggeration  of  the  reaction 
which  normal,  plain  muscle  gives  to  fresh  sera  in  general. 
Preparations  of  purified  protein  can  be  obtained  (e.  g., 
serum  globulin  precipitated  by  Gibson's  method,  or  egg 
albumen  crystallized  by  Hopkins'  method)  which  have  no 
effect  on  the  normal  plain  muscle,  but  are  as  toxic  for  the 
anaphylactic  plain  muscle  as  the  native  proteins.  (4) 
One  dose  of  the  specific  antigen,  in  sufficient  concentration 
to  produce  "maximal"  response  of  the  anaphylactic  plain 
muscle,  completely  desensitizes  the  latter  to  further  doses 
of  any  dimensions,  provided  that  the  experiment  is  not 
complicated  by  the  use  of  an  antigen  preparation  of  normal 
toxicity.  Either  normal  or  anaphylactic  plain  muscle  gives 
repeated  responses  to  successive  large  doses  of  a  normally 
toxic  serum  or  other  native  protein,  but  this  phenomenon 
is  not  anaphylactic  response.  (5)  When  sensitizing  doses 

1  Jour,  of  Pharm.,  1913,  iv,  167. 
21 


322  PROTEIN  POISONS 

of  several  antigens  are  given,  a  multisensitization  of  the 
plain  muscle  can  be  demonstrated.  Desensitization  of  the 
muscle  to  one  antigen  is  not  without  effect  on  its  sensitiveness 
to  the  others.  (6)  The  washed  plain  muscle  from  guinea-pigs 
immunized  to  an  antigen  by  a  series  of  injections,  is  sensi- 
tive to  the  antigen,  like  that  from  anaphylactic  pigs.  But 
the  sensitiveness  is  in  this  case  less  rigidly  specific,  e.  g., 
plain  muscle  from  a  guinea-pig  immunized  to  horse  serum 
showed  a  subsidiary  sensitiveness  to  sheep  serum.  (7) 
The  sensitiveness  of  the  washed,  plain  muscle  is  seen  in 
passive  as  in  active  anaphylaxis,  whether  the  serum  pro- 
ducing passive  sensitization  is  obtained  from  sensitive  or 
immune  guinea-pigs.  (8)  The  actively  or  passively  sensi- 
tized plain  muscle  after  being  desensitized  in  vitro  can  be 
resensitized  in  vitro  by  mere  contact  for  some  hours  with  a 
not  too  great  amount  of  sensitive  serum.  It  has  not  been 
found  possible  to  sensitize  normal  plain  muscle  in  exactly 
the  same  way;  but  perfusion  of  a  normal  uterus  for  five 
hours,  with  diluted  serum  from  sensitive  guinea-pigs,  pro- 
duced a  decided  passive  sensitization.  (9)  The  response 
to  the  specific  antigen  of  the  bronchioles  of  the  anaphylactic 
guinea-pig  is  not  impaired  by  excluding  the  abdominal 
viscera  and  the  brain  from  the  circulation,  and  is  produced 
with  apparently  undiminished  vigor  in  the  isolated  lungs 
perfused  with  Ringer's  solution." 

Dale  suggests  that  the  response  of  plain  muscle  to  its 
specific  sensitizer  might  be  used  for  medico-legal  purposes. 
The  suspected  material  might  be  used  to  sensitize  a  guinea- 
pig.  After  allowing  time  for  full  sensitization  the  uterus 
could  be  excised  and  a  suspended  horn  tested  with  various 
sera  until  the  one  giving  the  typical  response  was  detected. 
With  the  other  horn  the  limits  of  the  response  might 
be  determined.  A  second  method  might  be  as  follows: 
"Guinea-pigs  could  be  sensitized  with  a  small  injection  of 
known  serum  from  the  suspected  species,  e.  g.,  human 
serum.  After  the  usual  incubation  period  one  pig  would 
be  killed,  the  uterus  excised,  and  the  degree  of  sensitiveness 
of  the  first  horn  to  human  serum  tested.  If  the  sensitiveness 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     323 

were  not  of  a  high  order,  another  pig  could  be  tried  at  once, 
or  after  a  few  days  longer  incubation  period.  When  a 
uterus  was  found  which  gave  a  large  and  clear  response 
to,  say,  1  in  100,000  human  serum,  the  other  horn,  kept 
meanwhile  in  warm  oxygenated  Ringer,  could  be  suspended 
in  a  small  volume  such  as  10  c.c.  of  Ringer's  solution.  A 
dose  of  the  suspended  material  could  then  be  added,  and 
if  no  reaction  were  produced  it  would  be  clear  that  the 
dose  contained  less  than  0.0001  c.c.  of  human  serum,  which 
should  be  sufficient  evidence,  in  any  ordinary  case,  that 
the  blood  under  examination  was  not  human.  If,  on  the 
other  hand,  a  decided  response  were  produced,  it  would 
only  be  necessary  to  test  further  the  action  of  the  specimen 
on  a  normal  uterus,  so  as  to  exclude  primary  non-specific 
toxicity." 

It  has  been  demonstrated  by  Manwaring,1  and  confirmed 
by  Voegtlin  and-  Bertheim2  that  dogs  sensitized  to  horse 
serum  do  not  respond  on  reinjection  when  the  liver  is 
excluded  from  the  circulation. 

Theories. — Hamburger  and  Moro  at  one  time  suggested 
that  the  first  injection  leads  to  the  formation  of  precipitins, 
and  that  on  reinjection  precipitates  are  formed,  and  induce 
the  anaphylactic  symptom-complex  by  the  formation  of 
capillary  emboli.  The  formation  of  specific  precipitins  is 
a  reaction  which  occurs  in  vitro,  but  not  in  vivo.  Besides, 
the  symptoms  of  anaphylaxis  are  not  those  due  to  emboli, 
and  finally,  no  emboli  are  formed. 

Gay  and  Southard  thought  that  as  a  result  of  the  first 
injection  there  remains  in  the  circulation  a  protein  rest 
which  they  named  "anaphylactin,"  and  that  this  continues 
to  stimulate  the  cells,  creating  an  abnormal  affinity  for 
the  homologous  protein  which  on  reinjection  leads  to 
anaphylactic  shock.  The  transfer  of  this  "anaphylactin" 
to  a  fresh  animal  was  supposed  to  explain  passive  anaphy- 
laxis, a  phenomenon  first  studied  by  these  investigators. 


1  Zeitsch.  f.  Immunitatsforschung,  1910,  viii,  1. 

2  Jour.  Pharm.,  1911,  ii,  507. 


324  PROTEIN  POISONS 

Richet  held  that  sensitizers  contain  a  substance  which 
he  called  "congestin,"  and  that  this  develops  in  the  animal 
another  substance  known  as  "toxogenin."  The  reaction 
between  the  latter  and  the  homologous  protein  on  reinjec- 
tion  sets  free  a  poison  "apotoxin,"  which  on  account  of  its 
effect  on  the  nervous  system,  develops  the  symptoms  of 
anaphylaxis.  Considering  Richet's  toxogenin  a  ferment, 
we  can  accept  this  theory  as  essentially  correct. 

Besredka  taught  that  the  sensitizer  contains  two  sub- 
stances— "  sensibilisinogen"  and  "antisensibilisin."  On  the 
first  injection  the  former  develops  in  the  animal  body  a 
substance,  "sensibilisin,"  and  on  reinjection  the  sensibilisin 
and  the  antisensibilisin  combine  to  form  a  poison  which 
acts  on  the  nervous  system.  Besredka  has  not  been  able 
to  produce  satisfactory  proof  of  the  existence  of  antisensi- 
bilisin. His  work  along  this  line  has  already  been  referred 
to  (p.  260). 

For  reasons  which  will  become  evident  as  we  proceed,  it 
is  desirable  to  go  somewhat  into  detail  in  considering  the 
theory  of  Friedberger.  This  was  first  published  in  1909,1 
and  in  this  publication  Friedberger  clearly  and  unequivocally 
set  forth  his  theory.  It  may  be  known  as  the  theory  of 
sessile  or  fixed  receptors.  The  followng  is  an  abstract  of 
the  statement:  On  the  first  injection  the  protein  finds  but 
few  groups  with  which  it  can  combine,  and  for  this  reason 
it  is  not  poisonous,  even  in  large  doses,  just  as  happens 
when  tetanus  toxin  is  injected  into  a  naturally  immune 
animal.  During  the  period  of  incubation  the  animal  cells 
develop  specific  receptors  for  the  homologous  protein. 
With  frequent  injections  at  short  intervals,  as  when  the 
object  is  to  obtain  a  highly  active  precipitating  serum, 
the  newly  formed  receptors  are  in  large  part  cast  off  into 
the  blood.  When  a  single  small  dose  is  given,  as  in  sensi- 
tization,  less  receptors  are  cast  off  into  the  blood,  and  more 
remain  attached  to  the  cells.  In  this  way  an  organism 
relatively  insusceptible  to  a  given  foreign  protein  is  made 

1  Zeitsch.  f.  Immunitatsforschung,  ii,  208. 


PROTEIN  SENSITIZATION  OR  AN APR YL AXIS     325 

highly  susceptible,  and  on  the  second  injection  the  protein 
is  firmly  anchored  to  the  cell,  just  as  the  cells  of  an  animal 
susceptible  to  tetanus  anchor  the  tetanus  toxin.  The  only 
difference  is  that  in  the  latter  instance  the  receptors  are 
preformed,  while  in  the  case  of  sensitization  they  are 
developed  as  a  result  of  the  first  injection.  It  is,  as  Ehrlich's 
theory  explains,  the  same  substance,  the  receptor,  so  long  as 
it  remains  attached  to  the  cell,  that  is  the  cause  of  the 
poisoning,  and  which  becomes  the  cause  of  cure  when 
detached  from  the  cell,  and  cast  off  into  the  blood.  The 
only  difference,  as  has  been  stated,  is  that  the  substance 
attached  to  the  cell  (the  receptor)  is  not,  at  least  in  sufficient 
quantity,  preformed,  and  must  be  developed  by  the  first 
injection.  Protein  (toxin)  immunity  and  anaphylaxis, 
therefore,  are  alike  save  in  the  proportion  and  location  of 
the  antibodies.  When  the  precipitin  is  already  in  the  body 
fluids  the  injection  of  the  homologous  protein  is  without 
effect;  when  the  precipitin  is  still  attached  to  the  cell  in 
sufficient  quantity  the  reinjection  of  the  homologous  protein 
is  followed  by  the  phenomena  of  anaphylaxis.  The  anti- 
bodies exist  in  two  places:  (1)  As  free  antibodies  in  the 
serum  (known  as  precipitins  in  test-tube  experiments). 
(2)  As  sessile  antibodies  attached  to  the  cells.  In  cases  of 
local  sensitization,  as  in  Arthus  phenomenon,  the  local 
cells  only  are  affected  because  they  are  the  only  ones  which 
bear  the  sessile  receptors.  The  animal  escapes  anaphyl- 
actic  shock  because  the  cells  of  the  body  as  a  whole,  and 
especially  those  of  the  nervous  system,  do  not  carry  the 
specific  sessile  receptors.  Friedberger,  in  his  theory,  explains 
antianaphylaxis  as  follows:  An  animal  is  rendered  anti- 
anaphylactic  when  it  receives  a  large  reinjection  before 
the  period  of  incubation  is  complete.  In  this  case  the 
reinjection  uses  up  the  sessile  receptors  already  developed, 
but  these  are  not  enough  to  lead  to  anaphylactic  shock, 
and  at  the  end  of  the  period  of  incubation  the  new  crop  of 
sessile  receptors  is  not  sufficiently  developed  to  give  rise 
to  the  symptoms  of  anaphylaxis.  Again,  an  anaphylactized 
animal  may  be  rendered  antianaphylactic  by  a  small  dose 


326  PROTEIN  POISONS 

of  the  antigen.  This  is  explained  by  Friedberger  by  sup- 
posing that  the  small  dose  uses  up  a  part  of  the  sessile 
receptors,  and  that  there  are  not  enough  left  to  induce 
anaphylactic  shock  when  another  injection  is  made.  In 
short,  he  concludes:  "In  every  case  antianaphylaxis  is 
nothing  more  than  anaphylaxis  refraeta  dosi."  Passive 
anaphylaxis  is  explained  by  Friedberger  by  supposing  that 
the  free  receptors  in  the  blood  of  an  anaphylactized  animal 
become,  on  injection  into  a  fresh  animal,  anchored  to  the 
cells,  thus  forming  fixed  or  sessile  receptors.  This  is  Fried- 
berger's  theory.  It  is  clean  cut  and  clearly  stated  by  its 
distinguished  author,  but  at  present  it  has  no  support, 
and  is  clearly  out  of  harmony  with  known  facts,  some  of 
the  most  important  of  which  have  been  discovered  by  the 
researches  of  its  own  author.  It  was  an  attempt  to  make 
the  facts  of  anaphylaxis  fit  Ehrlich's  theory  of  the  action 
of  toxins  and  the  production  of  toxin  immunity,  while 
the  trend  of  later  research  is  to  show  that  the  two  sets  of 
phenomena  have  but  little  in  common.  Friedberger's 
theory  would  make  the  action  of  sensitizers,  such  as  serum 
albumin,  egg-white,  edestin,  bacterial  proteins,  etc.,  identical 
with  that  of  diphtheria  and  tetanus  toxin,  abrin,  ricin, 
the  venoms,  etc.  There  is  nothing  in  the  theory  about  the 
development  of  the  proteolytic  ferments,  and  the  liberation 
of  a  protein  poison  by  the  parenteral  digestion  of  the  sensi- 
tizer  on  reinjection.  If  we  read  his  works  with  correct 
interpretation,  Friedberger  has  abandoned  his  own  theory 
largely,  if  not  wholly.  Indeed,  in  Contribution  VI,1  Fried- 
berger plainly  discards  his  own  theory. 

According  to  Friedberger's  theory  all  sensitizers  act  like 
the  toxins;  although  at  first  only  mildly  toxic,  they  become 
more  so  by  developing  the  receptors,  and  thus  rendering 
the  animal  more  susceptible.  It  is  in  a  way  proper  for 
Friedberger  to  speak  of  anaphylaxis  as  a  "protein-anti- 
protein"  reaction.  Friedberger  calls  the  sensitizer  an 
antigen  and  the  substance  developed  in  the  animal  an 

1  Zeitsch.  f.  Immunitatsforschung,  1910,  vi,  179. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     327 

antibody.     According  to  our  theory  these  terms  are  not 
only  inappropriate,  but  are  confusing  and  misleading. 

The  theory  of  Vaughan  and  Wheeler  was  first  published 
in  1907, *  two  years  before  that  of  Friedberger,  and  while 
it  has  been  developed  and,  in  our  opinion,  confirmed  by 
later  investigations,  there  has  been  no  material  alteration 
in  it.  To  one  who  has  read  this  chapter  thus  far  the  essen- 
tials of  this  theory  must  be  already  fairly  understood,  but 
a  concise  statement  of  its  fundamental  points  must  be 
made  here  even  if  some  repetition  be  necessary.  The 
proteins  taken  into  the  alimentary  canal  are  broken  up 
by  the  digestive  ferments  into  non-protein  split  products, 
mostly  amino-acids.  During  or  after  absorption  these 
pieces  are  resynthesized,  in  part  at  least,  to  form  the  body 
proteins  peculiar  to  the  species.  The  precipitin  test  shows 
that  with  the  exception  of  the  proteins  of  the  crystalline 
lens,  those  of  all  the  fluids  and  tissues  of  the  body  are 
peculiar  to  the  species.  Those  of  one  species  differ  from 
those  of  all  other  species.  Just  where  this  synthesis  occurs 
in  the  animal  body  we  are  not  sure,  but  that  the  species 
proteins  are  formed  from  the  split  products  of  the  proteins 
of  the  food  has  been  positively  demonstrated.  Every 
protein  molecule  contains  a  poisonous  group.  In  the  whole 
molecule  this  group  is  saturated  with  other  non-poisonous 
groups.  As  the  whole  molecule  undergoes  cleavage  as  a 
result  of  enzyme  action,  the  poisonous  group  is  more  or 
less  completely  liberated,  and  in  this  process  it  becomes 
activated.  In  alimentary  digestion  the  poisonous  group 
becomes  most  active  at  or  about  the  stage  of  the  formation 
of  peptone.  As  the  digestive  process  proceeds,  the  poison- 
ous group  is  itself  disrupted,  and  ceases  to  be  a  poison. 
The  protein  poison  is  not  readily  diffusible,  and  for  this 
reason  it  is  retained  in  the  alimentary  canal  until  it  is  broken 
up  and  rendered  inert.  When  an  unbroken  or  undigested 
protein  finds  its  way  into  the  blood  or  tissues  it  must  be 
digested.  There  are  two  kinds  of  proteolytic  digestion: 

1  Jour.  Infect.  Dis.,  June,  1907. 


328  PROTEIN  POISONS 

(1)  Enteral,  (2)  parenteral.  Proteins  that  escape  enteral 
digestion  and  find  their  way  into  the  body  must  be  digested 
by  the  body  cells.  When  some  foreign  protein,  like  the 
blood  serum  of  another  animal,  milk,  egg-white,  bacterial 
proteins,  etc.,  are  injected  into  the  blood  or  tissues  they 
must  be  digested.  There  is  only  one  way  in  which  this 
can  be  accomplished,  and  that  is  by  the  cells  of  the  body, 
and  there  is  only  one  way  in  which  these  cells  can  do  this, 
and  that  is  by  elaborating  a  specific  proteolytic  ferment 
which  will  digest  and  destroy  the  foreign  protein. 

In  experimental  anaphylaxis  the  first  injection  introduces 
into  the  body  a  foreign  protein.  This  must  be  digested 
and  the  body  cells  slowly  elaborate  a  specific  proteolytic 
ferment  which  slowly  digests  it.  In  doing  this  certain 
body  cells  acquire  a  new  function.  The  protein  of  the  first 
injection  is  slowly  digested  usually  without  the  develop- 
ment of  recognizable  effects,  and  consequently  we  conclude 
that  the  animal  has  not  been  affected  or  had  its  functions 
altered  in  any  way.  But  this  is  a  mistake.  The  animal 
has  been  profoundly  affected.  It  has  developed  a  new- 
function  which  it  may  retain  quite  indefinitely,  and  which 
may  be  transmitted  from  mother  to  offspring.  The  foreign 
protein  is  digested  and  its  poisonous  group  set  free,  but 
this  has  been  done  so  slowly  and  gradually  that  the  effects 
have  not  come  within  the  range  of  our  powers  of  recog- 
nition. After  the  protein  of  the  first  injection  has  been 
disposed  of,  the  new  ferment  in  the  form  of  a  zymogen 
continues  to  be  formed  in  the  cells,  and  on  the  second 
injection  after  the  proper  interval,  this  zymogen  is  acti- 
vated and  splits  up  the  protein  so  promptly  and  so  abun- 
dantly that  the  liberated  poison  induces  the  symptoms  of 
anaphylactic  shock. 

The  following  statements  formulated  in  1907,  in  our 
opinion,  still  hold  good: 

1.  Sensitization  consists  in  developing  in  the  animal  a 
specific  proteolytic  ferment  which  acts  upon  the  protein 
that  brings  it  into  existence,  and  on  no  other. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     329 

2.  This   specific   proteolytic   ferment   stored   up   in   the 
cells  of  the  animal  as  a  result  of  the  first  treatment  with 
the  protein  remains  as  a  zymogen  until  activated  by  the 
reinjection  of  the  same  protein. 

3.  Our    conception    of    the    development    of    a    specific 
zymogen  supposes  a  rearrangement  of  the  atomic  groups 
of  the  protein  molecules  of  certain  cells,  or  an  alteration 
of  their  molecular  structure.     In  other  words,  we  regard 
the  production  of  the  specific  zymogen  not  as  the  formation 
of  a  new  body,  but  as  resulting  from  an  alteration  in  the 
atomic  arrangement  within  the  protein  molecule,  and  a 
consequent  change  in  its  chemism. 

4.  Some   proteins    in    developing   the    specific   zymogen 
produce  profound  and  lasting  changes  in  molecular  struc- 
ture, while  the  alterations  induced  by  others  are  slighter 
and  of  temporary  duration,  the  molecular  structure  soon 
returning  to  its  original  condition. 

5.  Bacteria  and  protozoa  are  living,  labile  proteins,  while 
egg-white,  casein,  serum  albumin,  etc.,  are  stabile  proteins. 
The  proteins  of  one  group  are  in  an  active,  while  those  of 
the  other  are  in  a  resting  state,  but  both  are  essentially 
proteins  made  up  of  an  acid  or  poisonous  chemical  nucleus, 
and    basic,    non-poisonous    groups.      Bacterial    immunity 
and    protein    sensitization,    apparently    antipodal,    are    in 
reality  the  same,  and  each  consists  in  developing  in  the 
animal  body  the  capability  of  splitting  up  specific  proteins. 
If  the  living  protein  be  split  up  before  it  has  time  to  multiply 
sufficiently  to  furnish  a  fatal  quantity  of  the  poison,  the 
animal  lives  and  we  say  it  has  been  immunized.     If  the 
stabile   protein    be   introduced   into    the   animal   body   it 
leads  to  the  development  of  a  specific  proteolytic  ferment, 
and  if  enough  of  it  to  supply  a  fatal  dose  be  reinjected  after 
this  function  has  been  developed,  the  animal  dies. 

6.  We   are   compelled   to   change   our  ideas   concerning 
the   causation   of   the   lesions   of   the   infectious   diseases. 
Formerly,  we  believed  the  structural  changes  to  be  due 
wholly    to    the    living,    growing,    feeding    microorganisms. 
For  instance,  we  were  sure  that  the  intestinal  ulcerations 


330  PROTEIN  POISONS 

of  typhoid  fever  are  caused  by  the  living  bacilli.  Now  we 
know  that  these  lesions  follow  the  intravenous  injection 
of  dead  proteins.  As  has  been  stated,  each  foreign  protein 
has  its  predilection  tissue  in  which  it  is  largely  deposited, 
whose  cells  it  especially  sensitizes,  and  where  it  is  disrupted. 
This  explains  the  characteristic  lesions  and  symptoms  of 
the  different  infectious  diseases.  Bacterial  inflammation 
is  essentially  a  chemical  process,  or  is  due  to  the  disruption 
of  cell  molecules  through  the  chemical  affinity  between 
certain  groups  in  the  bacterial  cell  and  certain  groups  in 
the  cell  of  the  animal.  So  long  as  the  bacterial  cells  are 
alive  the  chemism  that  holds  the  living  molecule  together 
tends  to  resist  this  process  of  disintegration.  The  patho- 
genic bacterium  assimilates  the  nutritious  constituents 
of  the  fluids  of  the  animal  body,  builds  them  into  its  own 
tissue,  converts  them  into  substances  foreign  to  the  host, 
and  finally,  when  the  bacterial  cell  goes  to  pieces  either 
from  spontaneous  dissolution,  or  through  the  aggressive 
action  of  some  animal  cell,  these  reconstructed  chemical 
groups  are  set  free  and  poison  the  animal,  inducing  lesions 
in  various  tissues,  and,  in  many  instances,  so  interrupting 
the  vital  functions  as  to  cause  death.  It  is  in  harmony  with 
these  statements  that  Friedberger  has  been  able  to  induce 
aseptic  pneumonia  by  spraying  horse  serum  into  the  lungs 
of  guinea-pigs  sensitized  with  the  same,  and  Schittenhelm 
and  Weichardt  have  established  "enteritis  anaphylactica" 
by  the  reinjection  of  egg-white  into  sensitized  dogs.  It 
is  more  than  probable  that  cholera  infantum  and  the  kindred 
summer  diarrheas  result  from  the  absorption  of  undigested 
milk  and  consequent  sensitization.  The  designation  "  protein 
diseases"  might  be  used  to  cover  the  majority  of  bacterial 
and  protozoal  diseases,  and  many  of  these  hitherto  regarded 
as  autogenous. 

7.  It  seems  to  be  a  physiological  law  that  the  specific 
ferments  elaborated  by  living  cells  are  determined  by  the 
proteins  brought  into  contact  with  them,  but  as  yet  we 
know  but  little  concerning  these  bodies  which  we  call 
ferments.  That  they  are  labile  chemical  bodies  resulting 


PROTEIN  SENSITIZATION  OF  ANAPHYLAXIS     331 

from  intramolecular  rearrangement  in  the  protein  molecules 
of  the  cell  seems  a  plausible  theory,  but  at  present  it  is 
only  a  theory.  We  know  but  little  of  the  action  of  these 
so-called  ferments  upon  their  homologous  proteins.  Our 
knowledge  of  the  chemistry  of  protein  sensitizers  is  exceed- 
ingly limited,  and  as  we  have  pointed  out,  it  is  highly  desirable 
that  work  in  this  direction  should  be  prosecuted  with 
vigor,  because  we  need  sensitizers  free  from  the  poisonous 
group.  Furthermore,  there  is  the  question  why  small 
doses  of  protein  induce  fever  while  large  doses  have  no  such 
effect.  At  present  we  have  no  satisfactory  answer  to  this 
question.  If  it  could  be  conclusively  demonstrated  that 
the  toxins  are  ferments,  the  subject  of  the  etiology  of  disease 
would  be  greatly  simplified.  We  have  elsewhere  (see  Chapter 
XV)  given  our  reasons  for  holding  that  the  toxins  are 
ferments,  and  at  this  point  we  wish  to  formulate  what  we 
believe  to  be  two  biological  laws: 

(a)  When  the  body  cells  find  themselves  in  contact  with, 
or  permeated  by,  foreign  proteins  they  tend  to  elaborate 
specific  ferments  which  digest  and  destroy  the  foreign 
proteins. 

(6)  When  body  cells  are  attacked  by  destructive  ferments 
they  tend  to  elaborate  antiferments  the  function  of  which 
is  to  neutralize  the  ferments  and  thus  protect  the  cells. 

Zunz1  finds  the  proteoclastic  (protein-splitting)  properties 
of  blood-serum,  as  tested  on  the  sensitizing  protein,  increased 
in  the  anaphylactic  state.  This  increase  becomes  measurable 
in  the  pre-anaphylactic  stage,  usually  about  the  fifth  day 
after  the  injection,  and  continues  to  be  measurable  to  from 
the  twentieth  to  the  sixtieth  day.  It  is  not  recognizable 
in  blood-serum  taken  during  or  soon  after  anaphylactic 
shock.  Zunz  concludes  that  the  increased  proteoclastic 
property  of  the  blood  serum  is  not  sufficient  to  fully  account 
for  the  phenomena  of  anaphylaxis.  In  this  conclusion  we 
fully  agree  with  the  distinguished  Belgian  investigator. 
In  our  opinion  the  failure  of  the  blood  serum  taken  during 

1  Zeitsch.  f.  Immunitiitsforschung,  1913,  xvii,  241. 


332  PROTEIN  POISONS 

or  soon  after  anaphylactic  shock  to  show  measurable  proteo- 
clastic  effect  is  due  to  the  accumulation  of  the  cleavage 
products  in  the  blood.  That  all  the  phenomena  of  anaphyl- 
axis  are  not  due  to  cleavage  ferments  in  the  blood-serum 
must  be  evident  to  all  who  have  followed  us  thus  far.  The 
work  of  Pfeiffer  and  Mita,  that  of  Zunz,  our  own,  and  that 
of  others  agree  in  showing  that  the  property  of  splitting  up 
the  sensitizing  protein,  in  measurable  quantity,  at  least, 
disappears  from  the  blood-serum  of  the  sensitized  animal 
long  before  the  anaphylactic  state  disappears.  When  a 
guinea-pig  is  sensitized  to  horse  serum,  the  blood-serum  of 
this  animal  looses  the  power  to  split  up  horse  serum  in  vitro 
in  appreciable  amount  after  from  twenty  to  sixty  days,  but 
the  animal  remains  sensitive  to  horse  serum  for  at  least  two 
years,  as  shown  by  Rosenau  and  Anderson,  and  probably 
so  long  as  the  animal  lives.  We  must  therefore  agree  with 
Zunz  that  the  increased  proteoclastic  property  of  the  blood- 
ferum  of  the  sensitized  animal  is  not  sufficient  to  account 
sor  all  the  phenomena  of  sensitization.  Our  theory  of  sensi- 
tization  takes  this  into  account.  We  hold  that  sensitization 
develops  in  certain  body  cells  a  new  function— that  of 
elaborating  a  new  specific,  proteoclastic  ferment.  The 
duration  of  this  new  function  varies  with  the  sensitizing 
protein  and  with  the  cells  in  which  it  is  developed.  In  a 
given  cell  this  function  must  be  limited  by  the  life  of  the 
cell.  We  do  not  know  just  what  cells  develop  this  new  func- 
tion, but  we  do  know  that  the  animal  may  remain  in  a 
sensitized  state  long  after  the  blood-serum  fails  to  show  any 
cleavage  action  on  the  sensitizing  protein  in  vitro.  It  may 
be  that  the  specific  ferment  present  in  the  blood-serum  of 
recently  sensitized  animals  comes  from  the  white  corpusc'es, 
or  it  may  come  in  part,  or  altogether;  from  fixed  cells.  It 
seems  justifiable  to  conclude  that  the  ferment  which  manifests 
its  action  in  animals  long  after  it  is  absent  from  the  blood 
must  come  from  fixed  cells,  and  that  these  are  stimulated  to 
elaborate  this  ferment  only  when  the  specific  protein  is 
brought  into  contact  with  them,  probably  only  when  they 
are  permeated  by  the  specific  protein.  All  the  facts  which 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     333 

have  been  ascertained  in  regard  to  this  matter  indicate  that 
sensitization  is  secured  only  by  alteration  in  the  cell  and  that 
in  some  cells  the  newly  developed  function  is  more  persistent 
than  in  others. 

The  facts  of  cross-sensitization  seem  in  harmony  with  our 
view  that  the  protein  molecule  contains  one  or  more  special 
sensitizing  groups.  The  white  of  hen's  eggs  sensitizes  to  itself, 
less  perfectly  to  the  white  of  duck's  eggs,  and  very  imper- 
fectly or  not  at  all  to  the  white  of  robin's  eggs.  The  serum 
of  man's  blood  sensitizes  to  itself  and  less  fully  to  that  from 
the  ape.  Horse  serum  sensitizes  to  itself,  and  less  fully 
to  that  of  the  donkey.  Certain  non-pathogenic  acid-fast 
bacteria  sensitize  in  some  degree  to  the  tubercle  bacillus. 
In  short,  the  phenomenon  of  sensitization,  like  that  of  pre- 
cipitation, may  be  employed  to  show  biological  relationship. 
All  this  seems  in  harmony  with  the  view  that  the  specificity  of 
sensitization  depends  upon  the  similarity  or  dissimilarity  in  the 
chemical  structure  of  protein  molecules  from  different  sources. 

Doerr1  makes  the  following  statement  concerning  Fried- 
berger's  anaphylatoxin :  So  far  as  the  matrix  of  the  poison 
is  concerned  it  is  highly  improbable  that  it  comes  from  the 
bacteria  or  other  antigens.  That  the  most  diverse  proteins 
should  yield  the  same  poison  seems  improbable.  The  theory 
that  the  poison  comes  from  the  amboceptor,  as  held  by 
Wassermann  and  Keysser,  is  still  less  probable.  That  the 
anaphylatoxin  comes  from  the  blood-serum,  the  one  constant 
factor  in  all  the  experiments  in  its  production,  is  most  prob- 
able. During  its  formation  or  in  the  process  of  blood  coagu- 
lation, a  poison  is  formed,  but  the  serum  obtained,  after 
coagulation  is  complete,  is  not  poisonous  on  account  of  the 
presence  of  antibodies.  When  the  serum  is  digested  with 
bacteria,  precipitates,  etc.,  the  latter  absorb  the  antibodies 
and  thus  the  serum  again  becomes  poisonous.  That  ana- 
phylatoxin can  be  obtained  when  there  are  no  formed  elements 
present,  as  for  instance  when  inactivated  horse  serum  is 


1  Handbuch  d.  path.  Mikroorganismen  by  Kolle  and  Wassermann,  second 
edition,  ii,  947. 


334  PROTEIN  POISONS 

digested  with  fresh  guinea-pig  serum,  does  not  contradict  this 
theory  because  the  colloidal  bodies  may  serve  as  absorption 
agents. 

This  explanation  of  the  easy  production  of  anaphylatoxin, 
given  by  Doerr,  is  worthy  of  consideration.  His  failure  to 
understand  how  the  same  poison  can  be  obtained  from  the 
most  diverse  proteins  has  no  weight  with,  us  since  wTe  have 
prepared  a  poison  which  has,  grossly  at  least,  the  same 
physiological  action,  from  the  most  diverse  proteins,  bacterial, 
vegetable,  and  animal.  The  protein  molecule,  wherever 
found,  must  have  some  common  nucleus,  and  this  we  believe 
to  be  the  poison.  But  that  a  poison  may  be  liberated  in  the 
process  of  blood  coagulation  does  not  seem  to  us  to  be  beyond 
the  range  of  possibility.  Blood  coagulation  is  a  fermentative 
process  and  that  there  is  no  cleavage  in  the  protein  molecule 
in  this  process  has  not  been  shown. 

As  Doerr  points  out,  so  long  ago  as  1877  Kohler  showed 
that  fresh  defibrinated  blood,  whether  homologous  or  heter- 
ologous,  is  an  active  poison.  This  has  been  confirmed  by 
others,  and  recently  it  has  been  reinvestigated  by  Moldovan, 
who  has  shown  that  blood  freshly  defibrinated  by  shaking 
with  glass  beads  causes  acute  death  when  injected  intra- 
venously into  guinea-pigs  and  rabbits.  In  the  former  animals 
the  typical  anaphylactic  lungpicture  after  death  is  seen.  When 
the  dose  is  slightly  sublethal  there  is  marked  fall  in  -tempera- 
ture with  subsequent  fever.  When  the  doses  are  smaller 
there  is  marked  fever.  On  standing  for  fifteen  to  forty-five 
minutes  defibrinated  blood  looses  in  toxicity.  Serum  obtained 
by  rapid  centrifugation  of  defibrinated  blood  is  poisonous. 
The  same  is  true  of  the  deposited  and  once-washed  cor- 
puscles. When  coagulation  is  delayed  by  the  presence  of 
sodium  citrate,  neither  the  supernatant  fluid  nor  the  cor- 
puscles are  poisonous,  but  both  become  so  when  coagulation 
has  been  induced  by  shaking  with  porcelain  beads.  Later, 
Doerr  has  shown  that  blood  received  in  paraffined  vessels 
becomes  poisonous;  but  when  coagulation  is  complete  the 
toxicity  disappears.  When  coagulation  is  made  to  proceed 
slowly  by  the  addition  of  hirudin  solution  or  a  0.7  per  cent. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     335 

solution  of  colloidal  silicic  acid,  it  retains  its  toxicity  for 
several  hours.  The  source  of  the  poison  in  coagulation 
blood  has  been  discussed  and  variously  explained.  Kohler 
thought  the  fibrin  ferment  is  the  poison,  but  this  was  shown 
not  to  be  true  by  Boggs.  Studzinski  suggested  that  the 
poison  might  come  from  the  mechanical  disruption  of  the 
red  corpuscles,  but  Moldovan  and  Doerr  have  shown  that 
the  plasma  or  serum  may  be  absolutely  free  from  both 
hemoglobin  or  cells  and  still  be  poisonous.  Freund  thought 
that  the  poison  might  come  from  the  disrupted  blood  plate- 
lets, but  in  rapid  centrifuges  even  the  platelets  may  be  re- 
moved and  even  then  the  plasma  or  serum  may  be  poisonous. 
However,  the  role  that  the  platelets  play  in  coagulation  still 
suggests  that  they  may  furnish  the  matrix  for  the  poison. 

The  toxicity  of  extracts  of  normal  tissue,  especially  of 
the  lungs  and  lymph  nodes,  is  most  interesting  in  this 
connection.  If  the  normal  lungs  of  a  rabbit  or  guinea-pig 
be  digested  for  two  hours  in  physiological  salt  solution,  the 
solution  kills  promptly  on  intravenous  injection.  Homolo- 
gous organ  extracts  are  more  active  than  heterologous. 
Rabbits  are  the  most  susceptible  animals  used  so  far.  The 
poison  seems  to  be  destroyed  when  heated  to  70°;  it  does  not 
pass  through  a  Berkefeld  filter,  and  is  absorbed  by  kaolin 
as  is  anaphylatoxin.  The  addition  of  serum  either  homolo- 
gous or  heterologous  seems  to  destroy  or  neutralize  the 
poison  after  contact  of  one  hour  or  more.  Doerr  states  that 
the  fresh  vaccine  growth  just  scraped  from  a  calf  with  a 
curette  furnishes  an  extract  which  kills  rabbits  instantly 
on  intravenous  injection.  These  extracts  seem  to  owe  their 
effects  to  a  coagulating  ferment. 

Section  of  a  dying  animal  shows  the  left  heart,  and  the 
pulmonary  arteries  and  veins  filled  with  coagula.  Doerr 
admits  that  poisoning  with  organ  extracts  from  normal 
animals  is  quite  unlike  death  from  anaphylactic  shock. 
In  the  former  the  blood  is  coagulated;  in  the  latter  it  is 
fluid.  But  Doerr  holds  that  the  formation  of  thrombi  during 
life  is  not  the  sole  cause  of  death  after  the  administration 
of  extracts  from  normal  tissues  or  after  the  intravenous 


336  PROTEIN  POISONS 

injection  of  fresh  defibrinated  blood.  He  states  that  Bianchi 
failed  to  find  intravascular  thrombi  after  sublethal  doses  of 
the  extracts,  and  that  Moldovan  met  with  a  like  observation 
after  poisoning  with  defibrinated  blood.  To  us  the  difference 
between  poisoning  with  normal  tissue  extracts  and  the 
effects  of  anaphylactic  shock  seem  quite  clear.  The  organ 
extracts  do  not  contain  a  chemical  poison,  but  a  ferment. 
This  ferment  coagulates  the  blood  and  leads  to  the  forma- 
tion of  thrombi.  This  is  a  process  of  protein  digestion,  and 
whether  a  protein  poison  is  set  free  in  it  remains  for  future 
research  to  determine.  In  anaphylactic  poisoning  the 
ferment  is  in  the  body  cell  and  splits  up  the  protein  introduced 
with  the  liberation  of  a  protein  poison. 

In  this  connection  the  work  of  Blaizot1  is  of  interest. 
When  dog's  serum  is  treated  with  an  extract  from  the  intes- 
tinal mucosa  of  the  dog,  or  rabbit's  serum  with  an  extract 
of  the  intestinal  mucosa  of  the  rabbit,  after  a  few  minutes 
of  contact,  if  either  preparation  be  injected  intravenously 
into  a  guinea-pig,  acute  death  results.  Extensive  thrombi 
are  found  in  the  heart  and  large  vessels.  The  serum  of  the 
guinea-pig  is  not  rendered  poisonous  by  homologous  extracts, 
but  is  made  poisonous  by  heterologous  extracts. 

We  must,  however,  admit  with  Doerr,  that  the  matrix 
of  Friedberger's  anaphylatoxin  remains  undetermined,  with 
much  probability  in  favor  of  the  possibility  of  its  being  in 
the  so-called  complement,  or  the  serum  of  the  guinea-pig, 
the  one  constant  factor  in  its  production.  This  does  not 
mean  that  it  is  not  a  protein  poison.  It  must  be  borne  in 
mind  that  anaphylatoxin  is  recognized  only  by  its  effect, 
and  it  has  never  been  even  partially  isolated  from  the  serum. 
Our  protein  poison  comes  certainly  from  the  protein  molecule. 
It  cannot  be  a  ferment  as  we  understand  ferments  at  present. 
It  is  thermostabile,  and  it  elaborates  no  antibody  and  yet 
it  may  be  identical  with  anaphylatoxin,  for  whether  the 
latter  comes  from  bacterial  cells  or  from  the  serum  it  is  of 
protein  origin. 

1  Compt.  rend.  Soc.  biol.,  1910-11-12. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     337 

Friedberger,  Mita,  and  Kumagi1  have  prepared  anaphyl- 
atoxin  by  the  action  of  the  normal  serum  of  the  guinea-pig 
on  the  crude  toxins  of  tetanus  and  diphtheria,  and  the 
venom  of  the  cobra.  They  assume  that  the  poison  comes 
from  the  cleavage  of  the  toxins,  an  assumption  which  seems 
to  us  wholly  without  warrant.  Some  years  ago  we  precipi- 
tated the  crude  toxins  of  tetanus  and  diphtheria  with  alcohol 
and  split  up  the  precipitate  with  chemical  agents  by  our 
method,  and  obtained  the  protein  poison,  but  we  never 
felt  justified  in  even  supposing  that  the  poison  came  from 
the  toxins.  Crude  toxins  and  venoms  are  complex  protein 
substances,  and  because,  the  protein  poison  can  be  obtained 
by  the  cleavage  of  these  is  far  from  proof  that  the  poison 
comes  from  the  toxin  constituent  of  such  a  mixture.  Indeed, 
one  does  not  know  that  the  active  principle  in  these  mixtures 
is  a  protein.  Until  the  toxins  have  been  obtained  in  some- 
thing like  a  pure  state  it  is  useless  to  speculate  concerning 
their  split  products. 

Bordet2  has  shown  that  anaphylatoxin  can  be  obtained 
by  the  action  of  the  normal  serum  of  the  guinea-pig  on  agar, 
and  this  has  been  confirmed  by  Nathan.3  One-half  gram  of 
agar  is  added  to  100  c.c.  of  0.85  per  cent,  salt  solution  and 
sterilized  by  boiling.  From  0.5  to  1  c.c.  of  this  agar  solution 
is  incubated  with  5  c.c.  of  normal  serum  from  the  guinea- 
pig  at  37°  for  from  one  to  twenty-four  hours  and  then  centri- 
fuged.  Many  tubes  may  be  employed  and  the  supernatant 
fluid  from  these  mixed.  From  3  to  5  c.c.  of  this  fluid  injected 
intravenously  into  a  guinea-pig  of  about  250  grams  kills  with 
typical  anaphylactic  symptoms  in  from  three  to  five  minutes. 
Even  0.1  c.c.  of  the  agar  solution  furnishes  enough  poison 
to  kill.  When  the  amount  of  agar  solution  employed  is 
greater  than  1  c.c.  or  less  than  0.1  the  amount  of  poison 
formed  is,  as  a  rule,  not  sufficient  to  kill,  but  may  induce 
anaphylactic  symptoms  of  varying  intensity.  When  the 

1  Zeitsch.  f.  Immunitatsforschung,  1913,  xvii,  506. 

2  Compt.  rend.  Soc.  biol.,  1913,  Ixxiv,  No.  5. 

3  Zeitsch.  f.  Immunitatsforschung,  1913,  xvii,  478. 
22 


338  PROTEIN  POISONS 

serum  is  inactivated  by  being  heated  for  one-half  hour  at 
55°  no  poison  is  formed.  Experiments  of  this  kind,  as  well 
as  those  with  kaolin  already  referred  to,  have  led  to  various 
suggestions.  They  have  caused  many  to  suspect  that  the 
poison  comes  from  the  serum.  It  has  been  suggested:  (a) 
That  the  agar  or  kaolin  or  bacteria  absorbs  the  complement 
from  the  serum  and  that  this  renders  the  serum  poisonous. 
(6)  That  the  poison  is  preformed  in  the  serum  but  that  its 
action  is  neutralized  by  some  other  constituent  of  the  serum 
which  is  absorbed  by  the  agar  or  kaolin,  (c)  That  the  absorp- 
tion of  some  constituent  of  the  serum  by  the  agar,  kaolin,  or 
bacteria  leads  to  a  disturbance  of  the  equilibrium  of  the 
protein  constituents  of  the  serum  which  as  a  consequence 
break  up  with  the  liberation  of  the  poison.  These  suggestions 
assume  that  the  poison  comes  from  the  serum,  and  this 
may  be  true.  Further  experimentation  will  determine  this, 
but  it  must  be  borne  in  mind  that  agar  contains  small 
amounts  of  protein,  and  this  has  a  large  surface  exposure 
and  is  in  a  physical  state  most  favorable  to  the  action  of  a 
proteolytic  ferment  in  the  serum. 

Zinsser1  has  shown  that  the  action  of  complement  "upon 
typhoid  bacilli  strongly  sensitized  or  not  at  all  sensitized 
may  be  carried  on,  at  body  temperature,  for  considerably 
longer  than  twelve  hours,  without  leading  to  a  destruction 
of  the  poisons,  and  that  this  is  true  when  the  quantities  of 
the  bacteria  used  vary  within  the  wide  range  of  from  one 
to  twelve  agar  slants.  It  has  been  found,  in  fact,  that  in 
the  case  of  this  microorganism  prolonged  exposure  at  the 
higher  temperature  of  considerable  quantities  of  bacteria 
constitute  an  unfailing  method  of  regularly  obtaining  power- 
ful poisons.  The  results  obtained  by  the  use  of  smaller 
quantities  and  the  less  vigorous  action  at  low  temperatures 
are  far  less  regular  or  satisfactory."  He  thinks  that  his 
results  throw  more  weight  on  the  assumption  that  anaphyl- 
atoxins  are  responsible  to  a  large  extent  for  the  toxemic 
manifestations  of  typhoid  fever.  He  also  states:  "If  we 

1  Jour.  Exp.  Med.,  1913,  xvii,  117 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     339 

leave  out  of  consideration  bacteria  which,  like  the  diphtheria 
bacillus,  produce  true  secretory  poisons,  it  would  be  the 
ability  to  gain  a  foothold  in  the  body,  the  degree  of  invasive 
power,  the  predilection  in  the  choice  of  a  path  of  entrance, 
and  the  specific  local  accumulation,  upon  which  the  speed  and 
quantity  of  toxin  production  and  absorption  would  depend, 
and  which  consequently  would  give  character  to  variations 
in  the  clinical  pictures  of  different  diseases.  Besides, 
simplifying  considerably  our  comprehension  of  bacterial 
toxemia,  the  point  of  view  suggested  by  this  work  again 
brings  out  the  great  importance  of  the  work  of  Vaughan 
and  Vaughan  and  Wheeler  on  the  non-specific  poisonous 
fraction  obtained  by  hydrolysis  of  bacterial  and  other  pro- 
teins, and  makes  it  desirable  that  the  particular  conditions 
of  anaphylatoxin  and  endotoxin  production  in  the  case  of 
individual  pathogenic  bacteria  should  be  carefully  studied." 

We  regard  the  work  of  Jobling  and  Bull1  as  confirmatory 
of  our  studies  in  every  particular.  These  investigators 
have  studied  the  action  of  the  cellular  substance  of  the 
typhoid  bacillus  and  its  split  products,  produced  by  the 
action  of  a  proteolytic  ferment  obtained  from  leukocytes, 
and  state  their  findings  as  follows:  "Freshly  washed, 
unheated  typhoid  bacilli  intravenously  injected  into  dogs 
cause  the  development  of  definite  symptoms  as  early  as 
twenty  minutes  after  the  injection.  Boiling  for  ten  minutes 
does  not  destroy  the  toxic  effects  of  a  freshly  washed  bac- 
terial emulsion.  Complete  solution  of  the  bacteria  (in  dilute 
alkali)  of  a  fresh  emulsion  does  not  prevent  the  removal  of 
the  toxic  substance  with  the  coagulable  proteins.  The  action 
of  leukoprotease  splits  the  toxic  substance  to  a  non-coagulable 
state,  the  digested  mixtures  being  toxic  after  removing  the 
coagulable  portion.  The  mere  presence  of  the  leukocytic 
ferment  is  not  responsible  for  the  toxicity  of  the  filtrate 
from  the  digested  mixture,  and  continued  digestion  destroys 
the  toxicity  of  a  previous  toxic  mixture.  From  these  observa- 
tions it  is  concluded  that  the  toxic  properties  of  freshly 

1  Jour.  Exp.  Med.,  1913,  xvii,  453. 


340  PROTEIN  POISONS 

washed  typhoid  bacteria  are  not  entirely  due  to  preformed 
secretory  toxic  bodies  that  are  stored  in  the  bacterial  bodies, 
but  that  these  properties  are  due  largely  to  products  formed 
by  hydration  of  the  bacterial  proteins  through  the  agency 
of  ferments  present  in  the  circulation  of  the  animal  previous 
to  the  injection,  or  which  become  mobile  subsequent  to 
the  entrance  of  the  foreign  bodies  into  the  blood-stream. 
Since  leukocytic  ferments  can  attack  the  bacterial  proteins 
in  vitro,  it  is  possible  that  the  leukocytes  are  a  source  of  the 
ferments  which  are  active  in  experimental  and  natural  cases 
of  intoxication  with  the  whole  bacteria." 

Nolf1  has  stated  a  theory  of  anaphylaxis  which  has  come 
to  be  known  as  "the  physical  theory."  It  supposes  that 
the  active  constituent  of  proteins  is  a  thromboplastic  sub- 
stance which  disturbs  the  colloidal  equilibrium  of  the  blood 
and  leads  to  the  deposition  on  the  surfaces  of  the  leukocytes 
and  the  endothelial  cells  of  capillaries,  of  a  delicate  film  of 
fibrin.  Thus  stimulated,  these  cells  pour  out  an  unusual 
amount  of  antithrombin.  On  account  of  the  consumption 
of  a  part  of  the  fibrinogen  and  the  increased  formation  of 
antithrombin  the  blood  fails  to  coagulate  after  anaphylactic 
shock  or  peptone  poisoning.  On  account  of  the  coagulation 
deposits  on  the  endothelial  cells  the  viscosity  is  increased 
and  the  leukocytes  adhere  to  the  vessel  walls,  thus  accounting 
for  the  leukopenia  observed  after  protein  injections.  The 
endothelial  cells  are  injured  and  the  walls  of  the  capillaries 
become  more  permeable,  thus  accounting  for  the  local  edema 
often  seen  in  anaphylaxis.  The  fine  capillaries  of  a  given  area 
may  be  occluded  by  thrombi,  and  this  explains  the  necrosis 
characteristic  of  the  Arthus  phenomenon.  The  irritation 
of  the  endothelial  cells  extends  to  the  smooth  muscle,  and 
this  leads  to  vasoparalysis,  and  the  characteristic  fall  in 
blood-pressure.  The  affinity  of  the  endothelial  cells  for 
the  protein  is  stimulated  by  the  first  injection,  and  acts  in 
a  fulminating  way  on  reinjection,  and  thus  the  suddenness 
of  anaphylactic  shock  is  explained. 

1  Archiv.  intern,  de  physiol.,  1910. 


PROTEIN  SENSITIZATION  OR  ANAPHYLAXIS     341 

This  is  a  plausible  and  attractive  statement,  and  we  are 
inclined  to  believe  that  there  is  truth  in  it,  but  we  fail  to 
see  any  reason  for  designating  it  as  a  "physical  theory." 
It  starts  out  with  the  assumption  that  proteins  contain  a 
poison  and  the  theory  is  an  explanation  of  the  modus  operandi 
of  the  chemical  poison.  The  endothelial  cells  are  sensitized 
and  pour  out  a  ferment,  antithrombin,  in  increased  quantity. 
That  the  endothelial  cells  are  involved  in  sensitization  we 
held  as  long  ago  as  1907.  That  the  permeability  of  the  walls 
of  the  capillaries  is  increased  under  the  action  of  protein 
poison  has  been  frequently  demonstrated  by  the  general 
diapedesis.  We  found  this  true  even  when  rabbits  died  as 
the  result  of  a  single  large  dose  of  egg-white. 

There  is,  however,  one  very  important  point  in  which 
the  theory  of  Nolf  differs  from  ours.  In  his  theory  the  cause 
of  death  on  reinjection  is  not  due  to  the  cleavage  of  the 
protein  introduced,  but  is  due  to  the  action  of  the  antithrom- 
bin on  the  blood.  He  holds  that  the  fact  that  intravenous 
reinjections  are  so  much  more  effective  both  in  dose  and  in 
time  than  intraperitoneal  and  subcutaneous  administrations 
is  in  favor  of  his  theory,  and  we  are  inclined  to  agree  with 
him  on  this  point.  However,  anaphylactic  shock  cannot  be 
due  wholly  to  rendering  the  blood  non-coagulable,  because 
this  may  be  done  by  injections  of  hirudin  without  shock. 
Doerr  objects  to  our  theory  on  the  ground  that  anaphylactic 
shock  follows  reinjection  so  quickly  that  there  is  not  time 
for  a  ferment  to  split  up  the  injected  protein  and  liberate 
a  poison;  but  Nolf's  theory  also  depends  upon  ferment 
action.  The  sensitized  endothelial  cells  must  be  awakened 
by  the  reinjection,  must  pour  out  their  abnormally  accumu- 
lated ferment,  and  this  must  act,  not  on  the  injected  protein, 
as  we  suppose,  but  on  the  blood.  According  to  either  theory, 
certain  cells  are  sensitized  and  store  up  zymogen,  which  is 
activated  on  reinjection.  This  ferment  acts  either  upon  the 
injected  protein  or  on  the  proteins  of  the  blood.  It  seems 
to  us  that  the  time  objection  made  by  Doer  to  our  theory 
is  quite  as  applicable  to  that  of  Nolf.  The  latter  is,  after 
all,  only  a  modification  of  the  former. 


CHAPTER    XII 

THE  PARENTERAL  INTRODUCTION 
OF  PROTEINS1 

FOR  a  long  time  it  was  thought  that  the  proteins  of  our 
food  undergo  but  slight  modification  before  absorption 
through  the  walls  of  the  alimentary  canal.  The  studies 
of  Beaumont  laid  the  foundation  of  the  scientific  investi- 
gation of  proteolytic  digestion,  and  soon  it  was  shown  that 
the  digestive  juices  convert  proteins  into  peptones. 

After  experiments  had  demonstrated  that  peptone  is 
formed  in  alimentary  digestion  and  had  shown  the  com- 
paratively ready  diffusibility  of  the  digestive  products, 
several  questions  arose.  Among  these  may  be  mentioned 
the  following:  (1)  Is  all  the  protein  converted  into  peptone 
in  the  alimentary  canal,  or  is  part  of  it  absorbed  in  unaltered 
form?  (2)  What  is  the  fate  of  peptone  after  absorption? 

Briicke,2  whose  studies  on  pepsin  and  its  action  made  him 
one  of  authority  in  this  matter,  held  that  only  a  part  of 
the  protein  is  converted  into  peptone  in  alimentary  diges- 
tion, and  that  much  of  the  soluble  protein  of  the  food  is 
absorbed  unchanged.  Furthermore,  he  taught  that  the 
fate  of  the  two  after  absorption  is  different.  The  peptone, 
he  taught,  is  rapidly  oxidized  and  serves  as  a  source  of 
energy,  but  is  not  utilizable  in  the  building  of  tissue,  the 
latter  function  devolving  solely  on  the  protein  absorbed  in 
unaltered  form.  Briicke's  arguments  in  support  of  this 
theory  may  be  briefly  stated  as  follows: 

1  This  is  largely  taken  from  an  article  by  Vaughan,   Gumming,   and 
McGlumphy  (Zeitsch.  f.  Immunitatsforschung,  1911,  ix,  16). 

2  Sitzungsber.  d.  k.  Akad.  d.  Wissensch.  zu  Wien,  1859,  Band  xxxvii, 
ibid.,  Band  lix. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      343 

(a)  At  best  the  gastric  juice  forms  peptone  slowly,  and 
the  time  during  which  the  food  is  detained  in  the  stomach 
does  not  permit  of  its  complete  peptonization.  It  will  be 
understood  that  the  action  of  the  pancreatic  juice  was 
not  known,  nor  had  erepsin  been  discovered  when  Briicke 
wrote.  (6)  In  an  animal  killed  while  in  digestion,  Briicke 
found  forty-eight  hours  after  death  coagulable  protein  not 
only  in  the  chyle  vessels  in  the  intestinal  walls,  but  in  the 
intestinal  villi,  and  he  concluded  that  this  could  come  only 
from  the  absorption  of  unaltered  protein,  (c)  Briicke  argued 
that  the  absorption  of  unbroken  protein  is  quite  as  possible 
as  that  of  fat,  since  the  molecule  of  the  former  could  not 
be  larger  than  that  of  the  latter.  This  argument  assumed 
that  the  absorption  of  both  proteins  and  fats  is  simply  a 
process  of  filtration. 

Diakonow1  supported  the  theory  of  Briicke  because 
peptone  cannot  be  found  in  large  amount  in  the  blood. 
Voit  and  Bauer2  and  Eichhorst3  concluded  that  unaltered 
protein  is  absorbed  because  they  found  that  the  introduction 
of  protein  in  the  large  intestine  is  followed  by  increased 
elimination  of  urea.  This  certainly  is  proof  that  the  protein 
is  absorbed,  but  not  proof  that  it  is  absorbed  unaltered. 
Eichhorst  showed  that  a  glycerin  extract  of  the  mucous 
membrane  of  the  large  intestine  had  no  digestive  action, 
but  he  did  not  show  that  the  large  intestine  did  not  contain 
any  pancreatic  juice  and  this  might  have  digested  the 
proteins.  Fick4  took  an  aqueous  solution  of  peptone  and 
precipitated  it  with  alcohol,  then  dissolved  the  precipitate 
in  water  and  injected  it  into  nephrotomized  dogs.  He  found 
that  the  blood  after  this  treatment  yielded  a  larger  amount 
of  nitrogenous  material  soluble  in  alcohol  and  precipitable 
with  mercuric  nitrate,  and  he  concluded  that  peptone 
introduced  into  the  blood  is  speedily  converted  into  urea 
without  being  employed  in  tissue  building,  while  unaltered 


1  Hoppe-Seyler's  med.  chem.  Untersuchungen,  1867. 

2  Zeitsch.  f.  Biol.,  1869,  v 

3  Pfliiger's  Arch.,  1871,  iv,570;  4  Ibid.,  1872,  v,  40. 


344  PROTEIN  POISONS 

protein  is  used  in  tissue  metabolism.  Fick's  conclusions 
support  Briicke's  hypothesis.  Maly1  pointed  out  a  possible 
error  in  Fick's  work,  showing  that  while  peptone  may  be 
precipitated  from  aqueous  solution  by  alcohol,  it  is  not 
wholly  insoluble  in  this  menstruum,  and  the  increase  of 
alcohol  soluble  nitrogen  in  the  blood  might  be  due  to  peptone 
and  not  to  the  conversion  of  this  into  urea. 

Evidently  if  Briicke's  theory  were  true,  the  animal  body 
could  not  maintain  its  health  and  vigor  if  fed  exclusively 
on  peptone,  which  according  to  the  theory  is  not  utilizable 
by  the  animal  in  the  repair  of  tissue.  Plosz2  fed  animals 
exclusively,  so  far  as  their  nitrogenous  food  is  concerned, 
on  peptones  and  found  that  they  did  not  loose  weight  or 
suffer  in  any  detectable  way.  Maly3  confirmed  this  finding 
which  has  been  repeated  many  times  and  under  divers 
conditions,  so  that  now  Briicke's  contention  that  the 
absorption  of  unaltered  protein  is  essential  to  health  has 
no  support. 

The  second  question,  What  is  the  fate  of  the  absorbed 
peptone,  became  for  a  time  one  of  much  importance.  Plosz 
and  Gyergyai4  injected  from  200  to  300  c.c.  of  a  10  per 
cent,  solution  of  peptone  into  the  stomachs  of  fasting  dogs 
and  after  periods  of  from  one  to  four  hours  searched  for 
the  peptone  in  the  blood  and  tissues  of  various  organs.  The 
method  of  recognizing  peptone  consisted  in  the  applica- 
tion of  the  biuret  and  Millon  tests  to  the  filtrate  after  the 
removal  of  all  the  coagulable  protein  by  acid  and  heat. 
They  found  the  largest  amount  of  peptone  in  the  mesenteric 
veins  and  in  extracts  of  the  mesentery,  much  less  in  the 
liver,  and  only  traces  in  hepatic  and  carotid  blood.  Next, 
they  injected  dogs  and  cats  intravenously  with  a  10  per 
cent,  solution  of  peptone,  employing  from  100  to  200  c.c., 
and  introducing  it  at  the  rate  of  from  2  to  3  c.c.  per  minute. 
A  dog  received  200  c.c.  during  one  and  one-half  hours,  and 
after  three  hours  the  carotid  blood  showed  only  a  small 


1  Pfluger's  Arch.,  ix,  605.  2  Ibid.,  325. 

3  Loc.  cit.  4  Pfluger's  Arch.,  1875,  x,  536. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS     345 

amount  of  peptone  which  had  wholly  disappeared  after 
four  hours.  When  larger  amounts  were  used  a  small  quantity 
appeared  in  the  urine,  but  the  proportion  eliminated  in  this 
way  was  only  a  small  part  of  that  injected.  They  also 
transfused  certain  organs  and  tissues  with  blood  to  which 
peptone  has  been  added,  and  found  that  the  peptone  soon 
disappeared  from  the  blood.  These  investigators  concluded 
that  peptone  is  soon  so  changed  in  the  organism  that  it 
can  no  longer  be  detected  by  the  method  which  they 
employed.  Whether  it  is  changed  directly  into  albumin 
or  is  so  altered  by  cell  activity  by  combining  with  other 
substances,  they  could  not  decide.  They  were  of  the 
opinion  that  the  capability  of  effecting  this  change  is  not 
confined  to  any  one  organ  or  tissue,  but  that  it  may  occur 
in  the  liver,  muscle,  or  other  tissue.  They  were  quite  con- 
vinced, however,  that  the  conversion  is  not  essentially  one 
of  oxidation,  since  the  amount  of  oxygen  in  the  blood  did 
not  affect  it. 

Schmidt-Muhlheim1  injected  from  5  to  10  grams  of 
peptone  into  the  jugular  vein  of  dogs  and  found  that  the 
peptone  disappeared  from  the  blood  within  sixteen  minutes 
after  completing  the  injection.  He  also  concluded  that  the 
injected  peptone  undergoes  a  rapid  conversion  into  albumin 
and  globulin. 

According  to  Hofmeister,2  peptone,  when  injected  into 
the  blood,  does  quickly  disappear  from  that  fluid,  but  is 
not  converted  into  albumin  or  globulin.  It  quickly  diffuses 
through  all  the  tissues  undergoing  a  dilution  which  is 
determined  by  the  total  fluid  in  the  body,  and  which  is 
so  great  that  its  detection  by  chemical  tests  is  impossible. 
Diffusion  into  the  brain  results  in  certain  characteristic 
symptoms,  the  most  marked  of  which  are  muscular  weak- 
ness and  somnolence.  These  symptoms  may  be  observed 
in  a  10  kg.  dog  after  the  subcutaneous  injection  of  from 
0.2  to  0.4  gram  of  peptone,  but  the  fatal  dose  is  large;  as 


1  Du  Bois  Raymond's  Arch.  f.  Phys.,  1880. 

2  Zeitsch.  f.  phys.  Chemie,  1881,  v,  127. 


346  PROTEIN  POISONS 

high  as  1  gram  per  kilo.  Injections  of  peptone  lead  to 
lowered  blood  pressure  and  much  of  the  peptone,  according 
to  Hofmeister's  finding,  may  be  deposited  in  the  tissue 
where  it  may  be  detected  at  a  time  when  there  is  none  in 
the  blood.  He  found  one-seventh  of  the  peptone  injected 
into  the  blood  in  the  kidneys,  which  organs  were  equiva- 
lent to  only  yir  °f  the  body  weight,  and  concludes  that  the 
peptone  injected  into  the  circulation  has  a  special  predilec- 
tion for  renal  tissue.  When  the  amount  of  peptone  injected 
is  large,  Hofmeister  recovered  as  much  as  84  per  cent,  of  it 
from  the  urine. 

Neumeister1  reviewed  the  literature  of  this  subject  up 
to  that  time,  and  made  some  additional  contributions. 
He  stated  that  some  proteins  are  absorbed  unchanged, 
that  others  need  only  to  be  dissolved,  and  that  still  others 
must  be  digested.  He  stated  that  the  compound  proteins, 
as  casein  and  hemoglobin,  when  injected  into  the  blood, 
act  like  foreign  bodies,  and  are  eliminated  in  the  urine,  while 
the  simple  and  denatured  proteins,  when  injected  into  the 
blood,  do  not  cause  albuminuria.  Stockvis  did  not  observe 
albuminuria  after  injecting  dog,  rabbit,  or  frog  serum  into 
dogs  or  rabbits,  but  did  when  he  used  egg  albumen.  Leh- 
mann  invariably  induced  albuminuria  by  injecting  egg 
albumen  intravenously  in  dogs,  but  failed  to  do  so  when 
he  employed  sodium  albuminate,  or  syntonin,  prepared 
from  frogs'  muscle,  myosin,  or  fibrin.  Ponfick  found  that 
dogs  bear  astounding  amounts  of  lamb  serum,  free  from 
corpuscles,  when  the  injections  are  made  slowly,  and  under 
gentle  pressure.  "The  amount  of  urine  was  not  appreciably 
increased,  although  the  color  became  darker,  owing  to  the 
greater  concentration,  while  not  a  trace  of  albumin  could 
be  detected."  There  is  no  statement  concerning  the  effect 
upon  the  elimination  of  nitrogen;  Forster  injected  large 
amounts  of  horse  serum  into  dogs,  while  the  urine  remained 
free  from  albumin.  Neumeister  injected  into  the  jugular 
veins  of  dogs  without  inducing  albuminuria,  the  following 

1  Zeitsch.  f.  Biol.,  1891,  xxvii,  309 


proteins  in  large  amounts:  Syntonin  and  albuminate  from 
egg  albumen,  syntonin  from  ox  flesh,  crystalline  phyto- 
vitellin  from  pumpkin  seed,  and  pure  serum  albumin  from 
the  ox. 

According  to  Neumeister,  Salviolo  was  the  first  to  show 
that  peptone  is  transformed  by  the  living  intestinal  wall, 
but  this  investigator  only  demonstrated  that  peptone 
disappears  when  placed  in  an  intestinal  loop  and  cannot 
be  found  in  either  the  blood  from  the  part,  or  within  the 
loop.  The  nature  of  the  transformation  was  not  determined. 
The  fact  that  peptone  is  synthesized  into  albumin  seems 
to  have  been  first  suggested  by  two  women,  students  of 
Kronecker,  Nadine  Popoff  and  Julia  Brinck.  It  was  thought 
by  these  investigators  that  this  synthesis  is  accomplished 
partly  by  the  epithelial  cells  of  the  intestinal  wall,  and 
partly  by  a  microorganism,  to  which  Julia  Brinck  gave 
the  name  micrococcus  restituens.1  Hofmeister2  suggested 
that  the  leukocytes  in  the  intestinal  wall  might  combine 
with  peptone  much  as  hemoglobin  does  with  oxygen  in  the 
lungs,  and  Heidenhain3  thought  that  the  leukocytes  might 
play  a  part  in  the  absorption  of  peptones,  but  that  it  could 
not  be  as  suggested  by  Hofmeister,4  otherwise  the  leuko- 
cytes in  the  circulating  blood  would  combine  with  peptone 
injected  intravenously. 

As  early  as  1874  Tschiriew,5  working  under  Ludwig's 
direction,  found  that  dog  serum  transfused  into  another 
dog  increased  the  elimination  of  urea  much  more  slowly 
than  when  given  to  the  dog  by  mouth,  but  Forster6  found 
that  horse  serum  affects  the  urea  output  in  dogs  equally, 
both  in  amount  and  time,  whether  given  intravenously  or 
by  mouth. 

Zunz   and   von  Meering7   injected  solutions  of   peptone 


1  Zeitsch.  f.  Biol.,  1889,  vii,  427,  453. 

2  Zeitsch.  f.  phys.  Chemie,  1881,  v,  151. 

3  Pfltiger's  Arch.,  1888,  liii.  4  Loc.  cit. 
6  Arb.  a.  d.  phys.  lust,  zu  Leipzig. 

6  Zeitsch.  f.  Biol.,  1895,  ii,  496. 

7  Pfluger's  Archiv,  1883,  xxxii,  173. 


348  PROTEIN  POISONS 

and  egg-white  intravenously  in  rabbits.  The  injections 
were  single  and  too  large  to  be  followed  by  marked  rise, 
and  too  small  to  result  in  marked  depression  of  temperature. 
However,  that  they  noticed  the  ill  effects  of  these  injec- 
tions is  shown  by  the  following  quotation :  "  The  poisonous 
action  of  peptone  was  unknown  to  us  at  the  time  when  the 
experiments  were  made,  but  some  of  our  rabbits  died  soon 
after  rather  large  injections  of  peptone.  Moreover,  un- 
changed egg-white  is  not  an  indifferent  substance,  and  it 
has  long  been  known  that  its  direct  introduction  into  the 
blood  may  cause  albuminuria  and  deep-seated  changes 
in  the  kidneys." 

Giirber  and  Hallauer1  employed  casein  for  the  reason 
that  it  may  be  distinguished  from  other  proteins  by  the 
action  of  rennin.  Solutions  of  this  protein  were  injected 
intravenously  into  rabbits  and  the  casein  was  detected  in 
the  bile.  Evidently  the  foreign  protein  was  on  its  way  to 
the  intestines  where  it  might  be  properly  digested.  These 
authors  quite  properly  point  out  that  because  a  protein 
injected  into  the  blood  does  not  appear,  or  appears  only 
in  part,  in  the  urine,  is  no  proof  that  it  has  been  assimilated 
in  unchanged  form  by  the  tissues,  because  it  may  have 
been  carried  to  the  intestine  and  there  properly  digested. 
They  also  brought  out  another  point,  confirmed  later 
by  our  own  work,  that  when  the  foreign  protein  diffuses 
from  the  blood  it  carries  with  it  some  of  the  blood  proteins. 
This  is  true  whether  it  is  poured  out  into  the  intestine  or 
eliminated  through  the  kidneys. 

The  results  obtained  by  Giirber  and  Hallauer  have  been 
confirmed  by  Burckardt,2  who  injected  hemielastin  intra- 
venously and  found  it  in  the  wall  of  the  small  intestine. 
He  concluded  that  it  had  been  brought  to  this  locality 
preparatory  to  its  being  properly  digested  and  fitted  for 
assimilation. 

Friedemann  and  Isaac,3  after  extensive  experimentation, 

1  Zeitsch.  f.  Biol ,  1904,  xlv,  372. 

2  Zeitsch.  f.  physiol.  Chem.,  1907,  li,  506. 

3  Zeitsch.  f.  exp.  Path.  u.  Ther.,  1907,  iv,  830. 


concluded  that  both  homologous  and  heterologous  proteins 
when  parenterally  administered  are  digested  and  probably 
assimilated.  They  say:  In  fasting  animals  the  parenteral 
administration  of  protein  leads  to  increased  protein  metab- 
olism. The  increased  nitrogen  elimination  is  the  same 
whether  the  injected  protein  be  the  serum  of  the  same  or 
of  a  different  species,  or  egg  albumen.  In  dogs  in  nitrogen 
equilibrium  protein  parenterally  administered  behaves 
the  same  as  that  given  by  mouth.  Carbohydrates  hinder 
increased  nitrogen  elimination  while  on  a  carbohydrate 
free  diet  there  is  increased  nitrogen  elimination.  In  her- 
bivorous animals  (goats  and  sheep)  there  seems  to  be  a 
tendency  to  retain  some  of  the  nitrogen  given  parenterally, 
but  the  results  obtained  were  not  constant.  We  cannot 
speak  of  a  toxic  protein  metabolism  unless  symptoms  imme- 
diately follow  the  administration.  If  heterologous  proteins 
be  poisonous,  when  administered  parenterally,  we  have 
not  been  able  to  demonstrate  it  in  fasting  animals.  If 
this  does  not  exclude  a  toxic  metabolism  it  renders  its 
assumption  wholly  hypothetical.  In  our  experiments 
increased  nitrogen  elimination  corresponds  to  increased 
administration,  and  a  toxic  protein  metabolism  is  char- 
acterized by  the  fact  that  nitrogen  ingestion  and  elimination 
bear  no  relation  to  each  other.  The  parenteral  adminis- 
tration of  protein  has  an  advantage  over  the  enteral,  because 
in  the  former  we  know  just  how  much  protein  enters  the 
blood.  When  blood  serum  is  introduced  there  can  be  no 
increased  concentration  of  proteins  in  the  blood,  but  in 
fasting  animals  the  introduction  of  serum,  either  homolo- 
gous or  heterologous,  does  lead  to  increased  nitrogen 
elimination.  The  transfusion  of  blood  even  in  large  amount 
from  one  dog  to  another  is  not  followed  by  any  marked 
nitrogen  elimination  by  the  recipient.  It  seems,  therefore, 
that  even  homologous  serum  behaves  like  a  foreign  protein, 
possibly  on  account  of  the  changes  that  have  taken  place 
in  it  during  coagulation.  In  sensitized  animals  the  parenteral 
introduction  of  the  homologous  protein  leads  to  explosive- 
like  increase  in  nitrogen  elimination. 


350  PROTEIN  POISONS 

In  the  light  of  later  research,  some  criticism  of  the  above 
given  conclusions  reached  by  Friedemann  and  Isaac  may 
be  made.  Foreign  proteins  when  introduced  parenterally 
are  poisonous.  In  some,  the  poisonous  action  is  due  to 
ferments,  but  all  are  poisonous,  even  when  the  ferment 
action  has  been  destroyed  by  heat.  It  is  only  a  question 
of  dosage.  Even  egg-white  or  horse  serum  injected  intra- 
venously in  sufficient  doses  will  kill  dogs  and  rabbits. 
Saturat  on  of  the  body  cells  with  any  foreign  protein  inter- 
rupts their  function.  The  protein  foods  of  the  body  cells 
are  carefully  prepared  physiologically.  The  foreign  pro- 
teins eaten  by  the  animal  are  broken  into  non-protein 
bodies,  and  these  pieces  are  put  together  again  after  a 
model  which  is  peculiar  to  that  species  of  animal  to  which 
the  feeder  belongs.  In  this  way  the  specific  albumins  and 
globulins  of  the  blood  of  each  species  are  constructed, 
and  these  supply  the  normal  protein  foods  for  the  body 
cells.  Foreign  proteins  parenterally  introduced  are  digested, 
if  the  amount  be  not  too  large.  This  digestion  is  carried 
out  in  part  in  the  intestines,  and  other  body  cells  acquire 
the  function  of  digesting  a  limited  amount  of  the  protein 
introduced.  Certainly  the  digestive  products  formed  in 
the  intestine  are  fit  for  assimilation,  and  it  may  be  that 
those  formed  in  other  parts  are  also  utilizable,  but  the 
capability  of  the  body  of  taking  care  of  proteins  parenterally 
introduced  is  limited,  and  in  large  doses  all  foreign  proteins 
thus  administered  are  poisonous.  Whether  this  is  also 
true  of  homologous  sera  we  do  not  have  sufficient  data  to 
determine. 

Bankowski  and  Szymanowski1  find  that  normal  human 
blood,  when  injected  intravenously  into  guinea-pigs,  kills 
with  the  symptoms  of  anaphylactic  shock  in  doses  of  0.5 
per  cent,  of  the  body  weight  of  the  animal.  In  typhoid 
fever  the  minimum  fatal  dose  falls  to  0.25  and  in  scarlet 
fever  and  measles  to  0.13  per  cent,  of  the  body  weight  of 
the  animal.  Fetal  human  blood  is  well-nigh  atoxic;  when 

1  Zeitsch.  f.  Immunitatsforschung,  1913,  xvi,  330 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      351 

the  dose  is  as  much  as  2.5  per  cent,  of  the  body  weight  it 
kills  slowly.  The  mother's  blood  kills  in  doses  of  0.5  per 
cent.,  and  fetal  blood  in  doses  of  2.5  per  cent.  This  is  not 
due  to  biological  differences  between  the  blood  of  the 
mother  and  that  of  the  fetus,  because  one  sensitizes  to 
the  other  as  well  and  in  as  small  doses  as  to  itself,  but  is 
due  to  the  relative  freedom  of  the  fetal  blood  from  ferments. 
When  human  blood  is  injected  into  the  blood  of  a  guinea- 
pig  the  former  carries  the  ferment  and  the  latter  supplies 
the  substrate.  In  the  infectious  diseases  the  proteolytic 
ferment  in  the  blood  is  increased  and  consequently  the 
minimum  fatal  dose  is  decreased. 

Dehne  and  Hamburger1  state  that  white  mice  do  not 
produce  a  precipitin  when  treated  with  horse  serum,  and 
Celler  and  Hamburger2  find  that  white  rats  fail  to  respond 
to  ox  serum  by  elaborating  a  precipitin. 

Uhlenhuth3  and  Michaelis  and  Oppenheimer4  found 
that  when  rabbits  are  repeatedly  fed  through  a  tube  with 
egg-white  or  serum,  they  develop  precipitins.  Celler  and 
Hamburger  say  that  this  may  be  due:  (1)  To  injury  of 
the  esophagus  or  stomach  by  the  tube,  and  the  introduction 
of  the  protein  in  this  way.  (2)  To  the  failure  of  the  secre- 
tion on  account  of  the  unnatural  method  of  feeding.  (3)  To 
the  direct  introduction  of  the  protein  into  the  intestine 
where,  according  to  Oppenheimer  and  Rosenberg,5  serum 
proteins  resist  tryptic  digestion. 

Celler  and  Hamburger  found  that  in  forced  or  tube 
feeding  the  protein  may  be  absorbed  unchanged,  on  account 
of  the  lack  of  digestive  juice. 

Chiray6  has  studied  the  effects  of  the  administration  of 
heterologous  proteins.  The  intravenous  injection  of  a  very 
small  amount  of  egg-white  in  rabbits  causes,  after  from  ten  to 
thirty  minutes,  a  transitory  albuminuria  with  increase  in  the 

1  Wien.  klin.  Woch.,  1904,  No.  29.  2  Ibid.,  1905,  xviii,  271. 

3  Deutsch.  med.  Woch.,  1900,  p.  734. 

4  Arch.  f.  Phys.,  phys.  Abt.,  Suppl.-Band,  p.  336. 
6  Hofmeister's  Beitrage,  1903,  v,  412. 

6  These  de  Paris,  1906;  Jahresber.  d.  Tierchemie,  1907,  xxxvi,  805. 


352  PROTEIN  POISONS 

volume  of  urine,  but  without  glycosuria,  hemoglobinuria,  or 
hematuria.  The  intravenous,  subcutaneous,  or  intraperito- 
neal  injection  in  increasing  doses,  even  with  long  intervals, 
causes  in  rabbits  a  gradual  decrease  in  weight.  When  the 
egg-white  is  injected  into  the  portal  vein  the  albuminuria 
appears  much  later  than  when  the  injection  is  made  into 
the  general  circulation.  Subcutaneous  injections  induce 
an  albuminuria  which  appears  later,  is  less  marked,  and 
persists  longer  than  when  the  injection  is  made  intravenously. 
The  intramuscular  injection  of  2  c.c.  of  egg-white  in  man 
was  without  effect,  but  when  tried  upon  one  who  had  renal 
deficiency,  albumin  appeared  in  the  urine  from  fourteen 
to  twenty-four  hours  later.  In  cases  of  marked  albuminuria 
injections  of  egg-white  did  not  materially  affect  the  excre- 
tion of  albumin.  In  rabbits,  dogs,  and  men  the  introduction 
of  large  amounts  of  egg-white  into  the  stomach  was  fol- 
lowed by  albuminuria,  though  this  did  not  invariably 
occur  in  the  men.  The  injection  of  egg-white  into  the 
rectum  of  rabbits  was  followed  by  an  albuminuria  which 
appeared  later  and  was  more  persistent  than  when  it  was 
given  by  the  stomach.  Egg-white  administered  by  rectum 
to  man,  especially  to  convalescents  from  infectious  diseases, 
was  followed  by  an  albuminuria,  but  this  did  not  occur 
when  the  albumin  was  mixed  with  an  active  trypsin  before 
injection.  In  some,  not  in  all,  the  administration  of  peptone 
by  rectum  was  followed  by  albuminuria  as  well  as  peptonuria. 
Injections  of  egg-white  in  rabbits  decrease  the  proteins  of 
the  blood,  as  shown  by  the  refractometer.  It  appears  that 
not  only  a  part  of  the  foreign  protein,  but  also  a  part  of 
the  blood  proteins  passes  into  the  muscles.  All  of  the 
injected  protein,  probably  not  the  greater  part,  'does  not 
pass  through  the  kidney,  at  least  not  within  the  time  of 
an  observation,  but  much  is  withdrawn  from  the  blood 
and  held  in  the  tissues.  Repeated  injections  of  egg-white 
lead  to  marked  structural  changes  in  the  kidneys.  Alimen- 
tary albuminuria  is  not  due  to  poisons  resulting  from  the 
splitting  of  the  protein,  but  the  absorption  and  elimination 
of  the  unbroken  protein  which  is  a  foreign  and  poisonous 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      353 

body.  When  a  milk  diet  is  presented,  casein  may  in 
some  instances  be  found  in  the  urine.  The  prohibition  of 
egg  diet  in  albuminuria  is  justified. 

The  statement  that  the  injection  of  a  foreign  protein 
leads  to  the  exudation  of  the  normal  proteins  of  the  blood, 
as  made  by  Chiray,  is  interesting,  and  if  confirmed  it  may 
be  found  to  be  of  marked  importance.  It  has  been  prac- 
tically confirmed  by  Wolf,1  who  found  that  the  proteins 
of  the  plasma  were  diminished  by  the  injection  of  Witte's 
peptone  in  11  out  of  14  tests. 

Oppenheimer2  estimated  that  as  much  as  49  per  cent, 
of  egg-white  injected  intravenously  or  intra-abdominally 
in  rabbits  is  eliminated  in  the  urine.  However,  he  does  not 
claim  any  great  exactness  for  this  work,  because  he  is  aware 
of  the  fact  that  all  the  protein  in  the  urine  does  not  come 
from  that  injected,  and  that  a  part  of  it  is  serum  albumin. 
It  is  probable  that  the  latter  makes  up  the  larger  part. 

Castaigne  and  Chiray3  hold  that  heterologous  proteins 
injected  subcutaneously  are  absorbed  and  eliminated  in 
the  urine  unchanged.  They  act  as  poisons,  causing  destruc- 
tion of  the  proteins  of  the  blood  and  increased  elimination 
of  nitrogen,  urea,  and  sulphur.  The  decrease  in  the  normal 
proteins  of  the  blood  may  be  as  high  as  from  1  to  3  per 
cent.  This  is  not  due  to  hydremia  as  shown  by  determina- 
tion of  total  solids.  Repeated  injections  of  heterologous 
proteins,  either  subcutaneously  or  intravenously,  lead  to 
cachexia. 

Nobecourt4  introduced  egg-white  into  the  alimentary 
canal  of  rabbits.  He  used  46  animals,  31  adults,  weighing 
from  1650  to  2270  grams  and  15  young,  weighing  from 
320  to  4030  grams.  These  received  by  the  stomach  or 
rectum  from  5  to  13  c.c.  of  egg-white,  at  each  injection 
at  intervals  of  one,  three,  seven,  ten,  and  fifteen  days. 
The  mortality  was  as  follows: 

1  Arch.  int.  Phys.,  iii,  343. 

2  Hofmeister's  Beitrage,  iv,  263. 

3  Compt.  rend.  Soc.  biol.,  1906,  Ix,  218. 
*  Ibid.,  1909,  xlvi,  850. 

23 


354  PROTEIN  POISONS 

Per  cent,  of  mortality: 

In  adults:  In  young: 

Intervals.             Stomach.      Rectum.  Stomach.  Rectum. 

Every  day  ...         0                   0  100  100 

Every  three  days  .      100                   0  100  100 

Every  seven  days  .        50                  50  33  33 

Every  ten  days       .50                    0  33  33 

Every  fifteen  days           0                 50  33  33 

From  the  above  test,  24  animals,  20  adults  and  4  young 
survived.  After  a  rest  of  from  twenty  to  forty-four  days 
each  of  these  received  rectal  injections,  every  seven  days 
of  from  3.3  to  9.6  c.c.  of  egg-white. 

Per  cent,  of  mortality. 

Interval  in  first  series.      Method  of  administration  in  first  series. 
Stomach.  Rectum. 

Every  day 66  0 

Every  three  days  ...     66  adults  0 

Every  seven  days  .      .      .    f50  adults  75  adults 

\50  young  100  young 

Every  ten  days      ...       0  50 

Every  fifteen  days       .      .       0  0 

From  the  second  ordeal,  13  animals,  12  adults  and  1 
young,  survived.  After  a  rest  of  from  seventeen  to  forty 
days  these  received  every  seven  days  rectal  injections  of 
from  5.2  to  8.9  c.c.  of  egg-white,  with  the  following  results: 

Per  cent,  of  mortality. 

Intervals  in  first  series.       Method  of  administration  in  first  series. 
Stomach  Rectum. 

Every  day 0  50 

Every  three  days  ...          0  100 

Every  seven  days  /  50  adults  100 

\100  young 

Every  ten  days      ...      100  100 

Every  fifteen  days       .      .      100  0 

From  this  test,  4  animals,  all  adults,  survived.  After 
a  rest  of  from  twenty-four  to  forty-five  days  these  received 
every  seven  days  from  4.6  to  7.5  c.c.  of  egg-white.  Three 
died. 

It  appears  from  the  experiments  of  Uhlenhuth  and  Nobe- 
court  that  egg-white  is  absorbed,  at  least  in  some  instances, 
unchanged  from  the  stomach  and  intestines  of  rabbits. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      355 

When  tubes  are  used  for  the  introduction  of  the  egg-white 
it  is  possible  that  a  small  amount  of  the  material  may  be 
introduced  through  some  slight  wound  or  abrasion  in  the 
mucous  membrane.  Sensitization  might  be  induced  in 
this  way,  but  it  is  hardly  conceivable  that  subsequently 
enough  would  be  introduced  in  this  way  to  kill  the  animal. 
It  seems,  therefore,  that  we  must  conclude  that  in  forced 
feeding,  at  least,  unbroken  egg-white  may  be  absorbed  from 
the  alimentary  canal  of  the  rabbit.  It  must  be  understood, 
however,  that  apart  from  any  injury  to  the  mucous  mem- 
brane, the  conditions  of  forced  feeding  are  not  exactly  the 
same  as  those  in  natural  feeding.  Celler  and  Hamburger 
have  called  attention  to  this  point.  By  continued  tube- 
feeding  of  rabbits  with  the  serum  and  blood  of  the  ox,  in 
only  one  instance  did  they  obtain  a  precipitin  for  the  serum, 
and  an  hemolysin  for  the  corpuscles,  while  they  found 
that  rabbits  after  fasting  took  the  serum  and  blood  willingly 
when  mixed  with  milk,  and  in  none  of  these  was  there  any 
evidence  of  absorption  without  digestion.  They  admit  the 
possibility  of  wounding  the  mucous  membrane  with  the 
tube,  or  of  carrying  the  material  through  the  tube  into 
the  intestine,  but  they  are  inclined  to  the  opinion  that  in 
the  unnatural  tube  feeding  the  digestive  secretions  are  not 
poured  out  so  freely  or  are  less  effective  than  in  natural 
feeding.  This  is  in  accord  with  the  findings  of  Pawlow, 
who  holds  that  desire  for  food  is  an  important  factor  in 
securing  thorough  digestion. 

With  this  brief  and  imperfect  review  of  the  literature  of 
the  subject,  we  turn  to  our  own  experimental  work.  Our 
method  is  to  inject  egg-white  into  the  animals  and  test  for 
its  presence  in  the  blood  and  extracts  of  tissue  by  sensitizing 
guinea-pigs,  having  first  demonstrated  that  the  blood  of 
the  rabbit  and  extracts  from  its  tissue  do  not  sensitize 
guinea-pigs  to  egg-white.  The  details  of  the  method  will 
be  developed  in  the  report  of  the  experiments.  Our  findings 
are  as  follows: 

1.  Egg-white  injected  into  the  stomach  of  a  rabbit  may 
be  in  part  absorbed  unchanged. 


356  PROTEIN  POISONS 

December  7,  1909,  at  9.30  A.M.,  50  c.c.  of  egg-white  was 
introduced  through  a  tube  into  the  stomach  of  a  rabbit  that 
had  been  kept  without  food  for  two  days.  Neither  at  the 
time  nor  subsequently  did  this  have  any  recognizable 
effect  upon  the  rabbit.  3  c.c.  of  blood  was  drawn  from  the 
heart  of  this  rabbit  at  10.30  and  11.30  A.M.,  and  at  12.30 
2.30,  and  4.30  P.M.,  and  each  of  these  portions  of  blood  was 
injected  intraperitoneally  in  a  fresh  guinea-pig.  January 
3,  1910,  each  of  these  pigs  received  intra-abdominally 
5  c.c.  of  a  dilution  of  egg-white  with  an  equal  volume  of 
physiological  salt  solution. 

Only  one  of  these  pigs  developed  symptoms  of  sensitization 
and  this  one  received  blood  drawn  from  the  rabbit's  heart 
three  hours  after  the  introduction  of  the  egg-white  into  the 
stomach.  Neither  the  blood  drawn  earlier  nor  that  drawn 
later  sensitized  guinea-pigs. 

January  8,  1910,  at  8  A.M.,  50  c.c.  of  a  dilution  of  egg- 
white  with  physiological  salt  solution  (1  to  1)  was  intro- 
duced through  a  tube  into  the  stomach  of  a  rabbit  which 
had  not  been  kept  without  food. 

Hourly  2.5  c.c.  of  blood  was  drawn  from  the  heart  of  this 
animal,  and  injected  intra-abdominally  into  guinea-pigs. 

January  22,  1910,  these  pigs  were  treated  each  with  5 
c.c.  of  the  egg-white  dilution  intra-abdominally.  The  first, 
second,  and  third  hour  pigs  showed  no  sensitization;  the 
fourth  and  fifth  hour  ones  were  sensitized,  while  the  sixth, 
seventh,  and  eighth  were  not. 

That  absorption  from  the  stomach  of  the  fed  animal 
should  have  been  more  tardy  than  from  the  fasting  one  is 
easily  understood. 

2.  Egg-white  injected  into  the  rectum  of  a  rabbit  may 
be,  in  part  at  least,  absorbed  unchanged. 

January  8,  1910,  at  8  A.M.,  50  c.c.  of  egg-white  diluted 
with  physiological  salt  solution  (1  to  1)  was  introduced 
through  a  tube  into  the  rectum  of  a  rabbit.  Hourly,  2.5 
c.c.  of  blood  was  drawn  from  the  heart  and  injected  intra- 
abdominally  into  guinea-pigs. 

January  22,  1910,  these  pigs  were  tested  and  all  from  the 


first  to  the  seventh  hour  were  found  to  be  sensitized  to 
egg-white. 

It  appears  from  this  that  egg-white  may  be  absorbed 
from  the  rectum  of  a  rabbit  without  being  so  far  altered 
as  to  destroy  its  specific  sensitizing  properties  and  that 
absorption  into  the  blood  begins  within  the  first  hour  and 
continues  for  at  least  seven  hours. 

3.  Egg-white  injected   into  the  peritoneal   cavity   of  a 
rabbit  may  be  absorbed  unchanged. 

December  7,  1907,  at  9.30  A.M.,  a  rabbit  received  intra- 
peritoneally  50  c.c.  of  a  dilution  of  egg-white  with  an  equal 
volume  of  physiological  salt  solution.  Hourly,  2.5  c.c. 
of  blood  was  drawn  from  the  heart  of  this  animal  and 
injected  intra-abdominally  into  guinea-pigs. 

January  3,  1910,  these  pigs  were  treated  with  the  egg- 
white  dilution  given  intraperitoneally. 

All,  from  the  first  to  the  fourth  hour,  died,  the  first  two 
in  fifteen  and  the  latter 'in  twenty  minutes.  The  fifth  hour 
one  was  not  sensitized.  It  should  be  stated  that  in  all  these 
experiments  guinea-pigs  found  not  to  be  sensitized  to  egg- 
white  were  subsequently  tested  and  found  to  be  sensitized 
to  the  blood  serum  of  the  rabbit. 

4.  Egg-white  injected  intravenously  in  rabbits  quickly 
disappears  from  the  circulating  blood. 

January  3,  1910,  a  rabbit  received  intravenously  50  c.c. 
of  a  dilution  of  egg-white  with  physiological  salt  solution 
(1  to  1).  Every  half  hour  blood  was  drawn  from  the  heart 
of  this  animal  and  injected  intra-abdominally  into  guinea- 
pigs. 

January  12,  1910,  these  pigs  were  tested  with  the  egg- 
white  dilution. 

The  first  two  were  found  to  be  sensitized  to  egg-white 
while  the  others  were  not.  The  pig  that  received  blood 
drawn  at  the  end  of  the  first  hour  died  in  a  typical  way 
within  thirty  minutes,  while  the  blood  drawn  at  the  expira- 
tion of  one  and  one-half  hour  failed  to  sensitize. 

December  6,  1909,  a  rabbit  received  intravenously  50 
c.c.  of  the  egg-white  dilution  (1  to  1).  Hourly  blood  was 


358  PROTEIN  POISONS 

drawn  from  the  heart  and  injected  into  guinea-pigs.  The 
first  two  were  found  to  be  sensitized  while  the  others  were 
not. 

5.  Egg-white  injected  intravenously  in  rabbits  may  be 
detected  in  the  peritoneal  cavity  after  it  has  disappeared 
from  the  circulating  blood. 

January  3,  1910,  a  rabbit  received  intravenously  50  c.c. 
of  the  egg-white  dilution.  Two  and  one-half  hours  later 
and  after  the  egg-white  had  disappeared  from  the  heart's 
blood,  as  was  shown  by  a  subsequent  test,  some  physio- 
logical salt  solution  was  injected  into  the  peritoneal  cavity, 
withdrawn  and  injected  into  a  guinea-pig,  which  later  was 
found  to  be  sensitized  to  egg-white. 

6.  Egg-white   injected    intravenously   into   rabbits   may 
be  detected  in  the  bile. 

November  15,  1909,  a  rabbit  received  intravenously  50 
c.c.  of  the  egg-white  dilution,  one  and  one-half  hours  later 
the  abdominal  cavity  was  opened,  the  animal  being  under 
ether,  and  small  amounts  of  bile  and  washings  from  the 
small  intestine  were  injected  into  guinea-pigs,  all  of  which 
later  were  found  to  be  sensitized  to  egg-white.  Of  four 
pigs  thus  treated  all  but  one  died,  and  this  one  developed 
marked  symptoms  when  treated  with  the  egg-white  dilution. 

7.  Egg-white  when  injected  intravenously  into  a  rabbit 
may  be  detected  by  the  sensitization  test  in  certain  organs 
after  it  has  disappeared  from  the  circulating  blood. 

December  6,  1909,  at  8.45  A.M.,  a  rabbit  received  intra- 
venously 50  c.c.  of  a  dilution  of  egg-white  with  an  equal 
volume  of  physiological  salt  solution.  5  c.c.  of  blood  was 
drawn  from  the  heart  of  this  animal  at  9.45,  10.45,  and 
11.45  A.M.,  and  at  1.45  P.M.  Each  of  these  portions  was 
injected  into  the  abdomen  of  a  guinea-pig  and  one  hour 
after  the  last  blood  was  taken  the  rabbit  was  killed  with 
ether,  and  extracts  of  the  brain,  liver,  kidney,  and  spleen, 
with  physiological  salt  solution  were  made  and  injected 
into  other  fresh  pigs.  December  17,  1909,  each  of  the  pigs 
that  had  been  treated  with  the  blood  of  the  rabbit  had 
5  c.c.  of  egg-white  dilution  (1  to  1)  intra-abdominally. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      359 

The  only  pig  that  gave  any  evidence  of  sensitization  was 
the  one  that  had  received  the  first  blood,  drawn  one  hour 
after  the  injection  of  the  egg-white.  The  symptoms  in 
this  animal  were  slight  and  transitory.  The  other  pigs 
showed  no  indications  of  having  been  sensitized.  It  seems 
from  this  that  after  one  hour  there  was  no  egg-white  in  the 
circulating  blood  of  the  rabbit.  All  of  these  pigs  had  been 
sensitized  to  the  proteins  of  the  rabbit's  blood  as  was  shown 
by  treating  them  with  rabbit  serum. 

January  4,  1910,  the  guinea-pigs  that  had  received  the 
extracts  of  the  organs  were  treated  with  the  egg-white 
dilution.  All  were  affected  within  a  few  minutes. 

The  one  that  had  received  the  kidney  extract  was  most 
seriously  disturbed  and  passed  to  the  convulsive  stage,  but 
ultimately  recovered.  The  one  that  had  the  spleen  extract 
came  next  in  the  severity  of  the  symptoms  developed.  The 
first  and  second  stages  were  well-marked  in  this  animal, 
somewhat  less  so  in  the  pig  that  had  received  the  extract 
from  the  brain.  Much  to  our  surprise,  the  pig  that  had 
received  the  extract  from  the  liver  was  least  affected. 
However,  failure  to  sensitize  with  the  extract  from  an 
organ  does  not  necessarily  mean  that  the  tissue  of  the  organ 
contained  none  of  the  foreign  protein.  It  may  combine 
with  certain  tissues  so  firmly  that  it  is  not  removed  by  a 
simple  solvent,  like  physiological  salt  solution.  The  fact 
that  from  certain  organs  the  extracts  did  sensitize  the  pigs 
shows  that  these  tissues  had  absorbed  the  egg-white,  but 
failure  to  sensitize  or  to  sensitize  so  fully  with  other  extracts 
does  not  conclusively  show  that  such  tissues  have  not 
absorbed  the  protein. 

8.  Egg-white  carried  into  the  tissue  after  intravenous 
injection  may  be  washed  back  into  the  blood  current  by 
transfusion  with  salt  solution. 

A  rabbit  received  intravenously  50  c.c.  of  the  egg-white 
dilution.  After  one  hour  egg-white  had  disappeared 
from  the  circulating  blood.  Two  and  one-half  hours  after 
the  injection  of  the  egg-white,  the  animal  was  transfused 
with  physiological  salt  solution.  During  the  transfusion, 


360  PROTEIN  POISONS 

2  c.c.  portions  of  the  fluid  were  drawn  from  the  heart  and 
injected  into  guinea-pigs.  The  last  of  these  portions  was 
drawn  after  one  liter  of  the  salt  solution  had  passed  through. 
All  of  these  portions  sensitized  the  guinea-pigs.  After  the 
transfusion,  the  brain,  liver,  spleen,  kidney,  and  the  deltoid 
muscle  were  removed,  and  rubbed  up  with  physiological 
salt  solution.  2  c.c.  of  each  of  these  extracts,  after  filtra- 
tion, was  injected  into  guinea-pigs.  The  one  having  the 
brain  extract  was  not  affected  by  the  subsequent  injection 
of  egg-white.  The  one  having  the  liver  extract  was  in 
convulsions  within  five  minutes  after  receiving  the  egg- 
white.  The  ones  having  the  extracts  from  the  spleen  and 
muscle  developed  first  and  second  stages  of  sensitization, 
but  recovered,  while  the  one  that  received  the  kidney 
extract  was  not  affected. 

We  conclude  from  this  that  the  brain  and  kidney  were 
washed  free  of  the  egg-white  by  the  transfusion,  while  the 
muscle,  liver,  and  spleen  held  the  egg-white  more  tena- 
ciously. 

9.  The  injection  of  egg-white  intravenously  in  rabbits 
decreases  after  a  few  hours  the  total  protein  in  the  blood. 

In  reviewing  the  literature  we  have  referred  to  the  finding 
of  Chiray  that  the  intravenous  injection  of  foreign  proteins 
decreases  the  total  proteins  of  the  blood.  His  experiments 
were  made  with  a  refractometer.  We  deemed  this  of 
sufficient  importance  to  justify  further  study.  On  one 
day  blood  was  drawn  from  a  rabbit  and  the  serum  obtained. 
On  the  next  day  this  animal  received  50  c.c.  of  the  filtered 
egg-white  dilution  intravenously.  Each  cubic  centimeter 
of  this  dilution  of  egg-white  contained  26  mg.  of  protein, 
as  calculated  from  a  nitrogen  determination.  In  other 
words,  with  the  dilution  there  was  introduced  into  the  blood 
of  the  rabbit  1.3  grams  of  foreign  protein.  On  the  day  after 
the  injection  more  blood  was  drawn  and  the  serum  from 
this  secured.  The  total  nitrogen  in  these  sera  was  deter- 
mined and  the  protein  content  calculated  with  the  following 
results: 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS   361 

Per  cent,  of  protein  in  the  blood  serum  before  the 

injection  of  egg-white 10 . 50 

Per  cent,  of  protein  in  the  blood  serum  after  the 

injection  of  egg-white 8.18 

Loss 2.32 

This  experiment  was  repeated  on  a  second  animal  with 
the  following  results: 

Per  cent,  of  protein  in  the  blood  serum  before  the 

injection  of  egg-white 9 . 33 

Per  cent,  of  protein  in  the  blood  serum  after  the 

injection  of  egg-white 7.36 


Loss 1.97 

< 

In  a  third  animal  the  following  results  were  obtained: 

Per  cent,  of  protein  in  the  blood  serum  before  the 

injection  of  egg-white 7.90 

Per  cent,  of  protein  in  the  blood  serum  after  the 

injection  of  egg-white 6.30 


Loss 1.40 

It  appears  from  these  figures  that  the  injection  of  egg- 
white  intravenously  in  rabbits  is  followed  by  the  disap- 
pearance of  an  appreciable  amount  of  the  normal  proteins 
from  the  circulating  blood.  This  confirms  the  finding  of 
Chiray. 

10.  The  injection  of  a  large  amount  of  egg-white  intra- 
venously in  rabbits  proves  fatal. 

No.  1.  35  c.c.  of  undiluted  egg-white  filtered  through 
cotton  was  slowly  injected  into  the  ear  vein.  The  respira- 
tion was  immediately  embarrassed,  and  with  a  slight  con- 
vulsive movement  the  animal  died  before  it  could  be  removed 
from  the  table.  On  opening  the  thorax  the  heart  was  found 
to  be  still  beating  and  irregularly  distended.  The  right 
side  was  dilated  and  filled  with  dark  fluid  blood.  Markedly 
anemic  areas  were  plainly  seen  in  the  lungs. 

No.  2.  32  c.c.  of  the  same  was  injected  more  slowly  and 
through  a  finer  needle.  The  result  was  practically  the  same. 


362  PROTEIN  POISONS 

No.  3.  40  c.c.  was  injected.  The  respiration  became 
difficult  and  the  animal  quite  limp.  The  right  side  was 
found  to  be  paralyzed,  but  the  animal  lived  for  two  hours, 
when  it  died  with  failure  of  respiration,  and  without  a 
movement.  The  heart  was  dilated  and  contained  dark, 
fluid  blood.  Anemic  areas  were  seen  in  the  lungs  and  the 
muscles  also  were  anemic. 

Van  Alstyne  and  Grant1  injected  dilute  egg-white  intra- 
venously into  a  dog  and  sensitized  guinea-pigs  with  blood 
drawn  from  one-quarter  to  seventy-two  hours.  Pearce2 
injected  foreign  proteins  intravenously  into  rabbits,  and 
sensitized  guinea-pigs  with  organ  extracts.  His  conclusions 
are  stated  as  follows:  "Extracts  of  the  kidneys  of  normal 
rabbits  prepared  one,  two,  three,  and  four  days  after  the 
intravenous  injection  of  egg  albumen  and  horse  serum  have 
the  power  to  sensitize  guinea-pigs  to  a  second  injection  of 
these  proteins.  The  sensitization  by  first-  and  second-day 
extracts  was  constant  and  intense,  that  by  the  third-day 
extracts  was  less  marked  and  sometimes  was  not  evident, 
and  that  by  the  fourth-day  extracts  was  only  occasional, 
and  when  present  was  always  weak.  Comparative  studies 
of  the  power  of  the  blood,  liver,  and  kidney  to  sensitize 
indicate  that  this  sensitization  depends  upon  the  content 
of  the  foreign  protein  in  the  circulatory  blood  and  not  upon 
its  accumulation  or  fixation  in  the  tissues  of  an  organ.  This 
opinion  is  supported  by  other  experiments  in  which  the 
sensitizing  power  of  the  blood  and  of  the  extracts  of  unwashed 
kidneys  was  compared  with  the  sensitizing  power  of  washed 
kidneys.  The  weak  sensitizing  power  of  washed  kidney 
extract  is  taken  as  evidence  that  foreign  proteins  of  the 
kinds  used  are  not  held  in  the  tissues  of  the  kidney,  and  if 
these  results  may  be  applied  to  nephrotoxic  proteins,  it 
follows  that  nephritis  is  not  due  to  selection  and  persisting 
fixation  of  a  protein  by  the  renal  cells,  but  is  due  to  the 
action  of  such  proteins  merely  during  the  process  of  elimina- 


1  Jour.  Med.  Research,  1911,  xxv,  399. 

2  Jour.  Exp.  Med.,  1912,  xvi,  349. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS   363 

tion.  In  experimental  acute  nephritis  of  the  type  due  to 
uranium  nitrate,  the  power  of  sensitization  to  egg  albumen 
is  prolonged  for  twenty-four  hours,  and  in  the  chromate 
type  for  forty-eight  hours,  thus  indicating  that  in  nephritis 
of  the  acute  type  at  least,  the  elimination  of  a  foreign  protein 
is  delayed." 

In  our  opinion  the  possibility  of  harm  coming  to  the 
kidney  or  any  other  organ  from  the  deposition  of  a  foreign 
protein  in  it  is  not  due  to  any  directly  poisonous  effect  of 
the  foreign  protein  but  to  the  liberation  of  the  poisonous 
group  when  the  body  cells  become  sensitized  and  split  up 
the  foreign  protein. 

Abderhalden1  has  shown  by  both  dialysis  and  by  the 
polariscope  that  foreign  proteins  injected  into  animals  are 
digested  by  ferments.  However,  he  does  not  find  evidence 
that  specific  proteolytic  ferments  are  formed.  Indeed,  it 
still  remains  a  question  whether  the  sensitizer  leads  to  the 
development  of  an  entirely  new  ferment  or  causes  the 
common  non-specific  proteolytic  ferment  of  the  blood  to 
develop  specific  properties.  We  regard  this  question  as 
only  of  academic  value.  In  either  case  the  proteolytic 
ferment  becomes  specific,  whether  formed  by  an  altered 
rearrangement  in  the  molecules  of  the  cells  or  by  alteration 
in  the  molecular  structure  of  a  non-specific  proteolytic 
ferment.  Abderhalden  believes  that  the  ferment  is  always 
present  in  the  blood,  and  that  it  is  a  secretion  of  the  leuko- 
cytes. We  agree  with  him  insofar  as  the  non-specific 
proteolytic  ferment  of  the  blood  is  concerned.  The  blood 
is  a  digestive  fluid,  but  we  believe  that  specific  ferments 
are  developed  in  various  fixed  cells  under  the  influence  of 
foreign  proteins  or  sensitizers.  Abderhalden  holds  that 
the  ferment  is  always  present  in  the  blood,  and  that  the 
ferment  and  the  sensitizer  may  both  be  present  as  they 
are  on  first  injection,  but  that  for  the  production  of  ana- 
phylactic  shock  a  third  and  unknown  factor  is  necessary. 
He  seems  to  be  influenced  in  this  belief  largely  by  the 

1  Schutzfermente,  1912. 


364  PROTEIN  POISONS 

phenomena  of  so-called  antianaphylaxis,  but  he  admits 
that  rapid  digestion  in  this  state  may  be  prevented  by  the 
accumulated  products  of  digestion.  He  says:  "From  recent 
studies  we  know  that  the  ferment  forms  a  compound  with 
the  substrate  before  the  equilibrium  of  the  latter  is  destroyed. 
After  the  cleavage  the  ferment  is  again  free,  so  far  as  it  is 
not  bound  by  the  cleavage  products."  It  seems  to  us  that 
this  is  all  that  is  necessary  to  explain  the  known  facts  in 
so-called  antianaphylaxis. 

In  his  work  on  the  digestive  action  of  blood  serum  Abder- 
halden  has  largely  employed  polypeptids  and  purified 
peptones.  Of  course,  he  does  not  expect  these  denatured 
proteins  to  act  as  sensitizers  and  lead  to  the  development 
of  specific  ferments,  but  they  are  especially  suited  for  diges- 
tive experiments,  because  the  split  products  as  soon  as 
formed  are  easily  recognized  by  their  effect  on  the  rotation 
of  light.  In  this  way  he  has  shown  that  peptones  and  poly- 
peptids are  quickly  split  into  their  constituent  amino-acids 
by  the  proteolytic  ferment  normally  present  in  serum.  In 
other  instances  he  has  employed  native  proteins  as  sensi- 
tizers. In  one  case  he  divided  a  lot  of  guinea-pigs  sensi- 
tized to  egg-white  into  three  .groups.  The  members  of 
the  first  group  while  in  the  sensitized  state  were  bled,  and 
the  serum  thus  obtained  was  digested  with  egg-white,  and 
it  was  demonstrated  both  by  dialysis  and  the  optical  method 
that  the  egg-white  was  digested  by  the  serum.  Now,  had 
this  been  done  with  the  serum  of  normal  guinea-pigs  there 
would  have  been  no  recognizable  digestion.  It  must  follow, 
therefore,  so  far  as  we  can  see,  that  the  blood-serum  of  the 
sensitized  guinea-pig  contains  a  ferment  which  is  not 
present  in  the  blood-serum  of  the  fresh  guinea-pig.  Further- 
more, had  it  been  tried,  it  would  have  been  found  that  the 
blood-serum  of  the  guinea-pig  sensitized  to  egg-white  would 
either  have  no  digestive  action,  or  but  slight  effect,  on  other 
proteins.  It  follows,  therefore,  that  the  sensitized  animal 
differs  from  the  unsensitized  in  the  fact  that  its  body  cells 
elaborate  a  specific  ferment  which  digests  the  protein  by 
which  it  was  called  into  existence,  and  no  other.  It  will, 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      365 

of  course,  be  understood  that  the  non-specific  proteolytic 
ferments  are  capable  of  digesting  a  more  or  less  extended 
group  of  proteins,  but  with  this  ferment  the  digestive  process 
proceeds  slowly,  and  it  is  not  supposable  that  all  proteins 
would  be  digested  by  it. 

From  the  second  set  of  sensitized  guinea-pigs  Abder- 
halden  took  the  serum  and  dialysed  it  by  itself,  the  purpose 
being  to  see  if,  while  in  the  sensitized  state,  the  blood-serum 
contains  any  diffusible  biuret  body.  The  serum  from 
six  sensitized  guinea-pigs  was  tested  in  this  way,  and  only 
in  one  instance  did  the  dialysate  respond  to  the  biuret 
test.  On  the  eighteenth  day  after  sensitization  the  remaining 
guinea-pigs  (6)  were  reinjected  and  blood  was  taken  five, 
fifteen,  thirty,  forty-five,  sixty,  and  ninety  minutes  after 
the  reinjection.  The  serum  obtained  from  the  samples 
of  blood  was  dialyzed  and  the  dialysate  subjected  to  the 
biuret  test.  The  samples  taken  five  and  fifteen  minutes 
after  reinjection  failed  to  yield  dialysates  which  responded 
to  the  biuret  test,  while  the  remaining  four  did.  This 
experiment  demonstrates  that  at  the  very  moment  when 
anaphylactic  shock  is  being  developed,  peptone-like  bodies 
are  being  formed  in  the  blood.  It  is  probable  that  the 
specific  digestion  of  the  protein  begins  at  the  very  moment 
that  the  reinjection  is  made,  but  that  further  time  is  neces- 
sary for  the  digestive  product  to  accumulate  in  a  few  cubic 
centimeters  of  blood  serum  in  recognizable  amount.  The 
biuret  test  is  not  a  highly  delicate  means  of  recognizing 
proteins. 

H.  Pfeiffer  and  Jarisch1  have  repeated  diffusion  experi- 
ments. The  method  of  procedure  may  be  briefly  described 
as  follows:  Guinea-pigs  are  sensitized  with  horse  serum, 
and  after  varying  intervals  they  are  bled  and  serum  obtained. 
The  serum  of  the  sensitized  animals  is  mixed  with  varying 
amounts  of  horse  serum,  and  the  mixture  incubated  in 
small  dialyzers.  If  digestion  takes  place,  peptone-like 
products  are  formed,  diffuse  through  the  membrane,  and 

1  Zeitsch.  f.  Immunitatsforschung,  1912,  xvi,  38. 


366  PROTEIN  POISONS 

are  detected  in  the  dialysate  by  the  biuret  test.  First,  it 
was  shown  that  the  serum  of  normal  guinea-pigs  with  or 
without  mixture  with  horse  serum  does  not  supply  biuret 
bodies.  After  guinea-pigs  had  been  sensitized  to  horse 
serum  for  six  days,  then  for  the  first  time  serum  obtained 
from  them  and  mixed  with  horse  serum  and  the  mixture 
dialysed,  did  the  biuret  test,  when  applied  to  the  dialysate, 
prove  positive.  This  does  not  mean,  as  we  take  it,  that 
the  formation  of  the  specific  ferment  begins  on  the  sixth 
day  after  the  injection  of  the  sensitizer.  It  means  that 
with  the  amount  of  the  sensitizer  employed,  the  specific 
ferment  had  accumulated  sufficiently  and  was  efficient 
enough  when  brought  into  contact  with  horse  serum  in  vitro 
to  digest  it  enough  to  show  its  action  by  the  biuret  test  when 
applied  to  the  dialysate.  The  serum  of  sensitized  animals 
continues  to  digest  the  homologous  protein  (that  to  which 
the  sensitization  is  due)  up  to  about  the  thirtieth  day. 
This  does  not  mean  that  the  formation  of  the  ferment 
ceases  after  this  time.  We  know  that  this  is  not  the  case, 
because  Rosenau  and  Anderson  have  shown  that  guinea- 
pigs  sensitized  to  horse  serum  remain  in  this  condition  for 
two  years  at  least,  and  probably  throughout  life.  This  is 
an  important  point  and  one  which  H.  Pfeiffer  nowhere 
discusses,  so  far  as  we  can  find,  although  it  has  been  brought 
out  by  his  work  more  prominently  than  by  anyone  else. 
In  order  that  it  may  be  understood,  we  will  try  to  state  it 
plainly.  Guinea-pigs  sensitized  to  horse  serum  furnish, 
from  about  the  sixth  to  about  the  thirtieth  day,  serum 
which  in  vitro  digests  horse  serum,  as  is  shown  by  the  forma- 
tion of  diffusible  biuret  bodies.  After  the  thirtieth  day  or 
thereabouts,  the  serum  of  the  sensitized  animal  no  longer 
has  this  digestive  action  on  horse  serum  in  vitro;  at  least 
such  action  is  not  demonstrable.  And  yet  the  guinea-pig 
remains  sensitive  to  horse  serum.  This,  in  our  opinion,  is 
due  to  the  fact  that  certain  fixed  cells  in  the  animal  body 
remain  sensitive  and  responsive  to  reinjection  long  after 
the  leukocytes  lose  their  sensitization.  Pfeiffer  and  Jarisch 
found  that  in  the  so-called  antianaphylatic  state  the  blood- 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS   367 

serum  of  the  guinea-pig  does  not  digest  the  horse  serum, 
at  least  not  to  the  extent  of  supplying  the  dialysate  enough 
of  the  digestive  product  to  be  detectable  by  the  biuret 
test,  and  that  no  response  to  this  reaction  can  be  obtained 
until  the  third  or  fourth  day  after  the  reinjection.  This  is 
not  due  to  the  absence  of  the  specific  ferment  in  the  blood- 
serum,  but  is  due  to  the  accumulation  of  the  digestive 
products,  leading  to  an  increase  in  the  antitryptic  titer 
of  the  serum.  It  has  been  known  for  some  time  that  the 
addition  of  normal  blood-serum  to  a  mixture  of  casein  and 
trypsin  prevents  or  arrests  the  digestive  action  of  the  latter 
on  the  former.  This  phenomenon  has  been  investigated 
by  Rosenthal,1  who  concluded  that  the  antitryptic  action 
of  blood-serum  is  not  due  to  the  presence  of  antiferment. 
His  reasons  for  this  conclusion  may  be  stated  as  follows: 
(1)  It  takes  at  least  twenty-four  hours  to  produce  an  anti- 
ferment,  and  this  effect  of  blood-serum  on  tryptic  digestion 
is  immediate.  (2)  The  antitryptic  action  of  blood-serum 
is  not  increased  by  ligature  of  the  pancreatic  duct,  and  it 
should  be  if  it  were  due  to  increased  formation  of  anti- 
trypsin.  (3)  The  antitryptic  constituent  of  blood-serum 
is  thermostabile  and  non-specific,  and  it  would  be  thermo- 
labile  and  specific  were  it  an  antiferment.  (4)  The  anti- 
tryptic action  of  blood-serum  is  increased  in  full  digestion, 
and  in  those  diseases  in  which  there  is  excessive  protein 
metabolism,  and  is  decreased  in  hunger.  Rosenthal  con- 
cluded that  the  antitryptic  action  of  blood  serum  is  due  to 
the  presence  of  digestive  products.  When  these  are  abun- 
dant the  digestive  effect  of  blood  serum  is  decreased  or  its 
antitryptic  property  is  increased.  When  the  blood  is 
relatively  poor  in  the  products  of  protein  metabolism  the 
digestive  property  of  this  fluid  is  increased  or  its  antitryptic 
property  is  decreased.  This  is  a  striking  illustration  of  at 
least  one  of  the  ways  in  which  the  parenteral  digestion  of 
proteins  is  regulated,  and  it  seems  to  us  quite  sufficient 
to  explain  the  phenomena  of  so-called  antianaphylaxis. 

1  Folia  Serologica,  1910,  vi,  285. 


368  PROTEIN  POISONS 

Rusznjak1  has  shown  that  when  a  sensitized  animal  recovers 
from  a  reinjection  or  is  in  the  so-called  antianaphy lactic 
state,  its  blood  is  laden  with  digestive  products.  He  has 
demonstrated  this  by  showing  that  as  early  as  thirty  minutes 
after  the  reinjection  the  antitryptic  titer  of  the  blood-serum 
is  greatly  increased.  In  this  way  the  animal  body  strives 
to  protect  itself  from  the  effects  of  unusual  parenteral 
digestion.  Parenteral  digestion  is  a  normal  process.  It  is 
continuous  and  the  protein  poison  in  small  amount  is  being 
formed  constantly  in  the  body,  and  in  part  converted  into 
a  harmless  substance  by  further  digestion,  and  in  part 
eliminated  as  such  in  the  urine.  After  anaphylactic  shock 
it  is  found  in  the  urine  in  unusual  quantity.  The  regulation 
of  the  formation  and  disposition  of  this  poison  is  dependent 
upon  the  fine  adjustment  between  cell  metabolism  and  the 
digestive  action  of  the  blood.  Parenteral  digestion,  as  a 
physiological  process,  is  carried  on  by  the  non-specific 
proteolytic  ferments  of  the  blood,  and  tissues.  When  a 
substance  easily  acted  upon  by  this  non-specific  proteo- 
lytic ferment  is  suddenly  thrown  into  the  blood,  life  may  be 
endangered.  This  is  apparently  the  case  when  a  hemo- 
lytic  ferment  is  injected  into  the  body  in  the  form  of  a 
foreign  active  serum,  or  a  venom.  The  hemolysis,  thus 
caused,  results  in  the  liberation  of  a  large  amount  of  protein 
substance  which  is  readily  split  up  by  the  normal,  non- 
specific proteolytic  ferment,  and  the  poison  thus  formed 
may  destroy  life.  However,  if  the  dose  be  not  overwhelm- 
ingly great  the  digestive  products  retard  the  action  of 
the  ferment  and  tend  to  conserve  life.  H.  Pfeiffer  and 
Jarish  attempt  to  distinguish  between  primary  and  second- 
ary protein  toxicoses.  When  the  protein  poison,  preformed, 
as  in  peptone,  urinary  residue  or  /3-i  (ergamin)  is  injected 
into  an  animal,  the  antitryptic  titer  of  the  serum  is  decreased. 
On  the  other  hand,  when  the  protein  is  broken  up  in  the 
animal  body,  as  in  anaphylactic  shock  or  hemolytic  poison- 
ing, the  products  of  the  cleavage  increase  the  antitryptic 

1  Deutsch.  med.  Woch.,  1912,  No.  4. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS      369 

titer  of  the  serum,  or,  in  other  words,  lessen  the  digestive 
action  of  the  blood,  and  they  propose  to  distinguish  between 
primary  and  secondary  protein  toxicoses  by  determining 
the  antitryptic  titer  of  the  blood  serum.1  They  state  that 
the  curve  of  the  antitryptic  serum  titer  in  retention  uremia 
after  double  nephrectomy  in  rabbits  is  similar  to  that  of 
anaphylactic  shock  and  hemolytic  poisoning.  Before  the 
development  of  symptoms  of  poisoning  the  antitryptic 
titer  rises  above  the  normal  (owing  to  the  accumulation 
of  digestion  products). 

It  must  be  evident  that  the  presence  of  a  specific  proteo- 
lytic  ferment  is  not  necessary  in  all  cases  to  split  up  proteins 
with  the  liberation  of  the  protein  poison.  As  has  been 
stated,  Friedberger  and  many  others  have  found  that  the 
protein  poison  is  liberated  from  bacterial  proteins  by  diges- 
tion with  the  normal  serum  of  guinea-pigs.  In  this  there 
can  be  no  question  of  a  specific  ferment.  The  animals 
supplying  this  serum  have  not  been  sensitized  with  bacterial 
or  any  other  proteins.  They  are  normal,  untreated  animals; 
besides,  there  is  nothing  specific  in  this  reaction,  since  the 
same  poison  is  obtained  from  diverse  bacterial  proteins. 
As  we  have  held  for  years,  every  protein  molecule  contains 
a  poisonous  group,  and  whenever  and  by  whatever  agent 
the  protein  molecule  is  disrupted,  the  poisonous  group 
may  be  set  free.  The  disrupting  agent  may  be  a  chemical 
substance,  a  specific  or  a  non-specific  ferment.  Failure 
to  grasp  this  point  has,  in  our  opinion,  led  more  than  one 
investigator  into  error.  At  one  time  Friedberger  stated 
that  our  poison  cannot  be  the  true  anaphylactic  poison 
because  its  formation  is  not  specific.  If  this  be  true  of  our 
poison  it  is  also  true  of  his  so-called  anaphylatoxin.  When 
a  protein  is  digested  or  split  up  there  is  one  stage  in  the 
process  when  the  poisonous  group  is  liberated.  This  may 
not  always  be  evident  because  when  the  cleavage  is  carried 
one  step  farther  the  poison  itself  is  destroyed.  This  is  true 


1  The  details  of  this  procedure    are  given   in    their  paper,   Zeitsch.   f. 
Immunitatsforschung,  1912,  xvi,  38. 
24 


370         .  PROTEIN  POISONS 

of  our  poison  and  of  Friedberger's  anaphylatoxin.  As  we 
have  stated  more  than  once,  parenteral  digestion  is  a 
normal,  physiological  process,  and  in  this  process  the 
protein  poison  is  liberated.  There  are  many  proteins,  not 
all,  which  are  digested  by  the  normal,  non-specific  proteo- 
lytic  ferment  of  the  blood  and  tissues.  In  this,  in  our 
opinion,  lies  the  explanation  of  the  results  obtained  by 
Szymanowski,1  who  has  found  that  the  intravenous  injection 
of  varied  protein-precipitating  agents,  such  as  copper 
nitrate,  copper  sulphate,  mercuric  chloride,  lead  acetate, 
phosphomolybdic  acid,  tannin,  and  picric  acid  in  small 
doses,  may  cause  all  the  symptoms  of  acute  anaphylactic 
shock  and  death.  In  our  opinion,  the  most  probable  explan- 
ation of  this  is  that  the  precipitates  formed  in  the  blood  by 
these  substances  act  like  foreign  proteins  and  are  digested 
by  the  non-specific  proteolytic  ferment  of  the  blood  with 
the  liberation  of  the  protein  poison. 

There  has  been  some  difference  of  opinion  as  to  the 
source  of  the  protein  poison  in  anaphylaxis.  At  one  time 
HU  Bfeiffieri  thought  that  it  must  come  from  the  proteins  of 
t&e-^body.  kHe  was  led  to  this  conclusion  by  the  smallness 
exfihthB)od®be  of  the  anaphylactogen  necessary  to  induce 
shock  on  reinjection,  but  in  his  latest  paper 
the  poison  comes  from  the  anaphylactogen 
(antigafi)).fi  ^or  like  reason  Friedemann2  was  inclined  to 
fchuliepiniointhat  the  poison  is  furnished  by  the  serum  of 
the  srarfdtiHed  animal,  but  he  seems  now  to  think  that  it 
the  anaphylactogen  (antigen).  Wassermann 
thought  that  the  source  of  the  poison  is  in 
the  fiermertt9(amboceptor).  They  shook  horse  serum  with 
kadlfn  aondxthen  separated  the  kaolin  from  the  horse  serum 
arfJanceigsJfeuge  and  digested  the  kaolin  with  guinea-pig 
iygmaniEiiirfid  .Obtained  a  poison.  They  explained  this  by 
feapposmg  9$E#t  the  kaolin  absorbed  the  amboceptor  from 
fcheJ  koEseCEeriim,  and  when  this  was  acted  upon  by  the 

.1  i.2eitffi&.  ,fo( Jhpmunitatsforschung,   1912,  xvi,   1.  *  Ibid.,  ii,  591. 

3  Folia  Serologica,  1911,  vii. 


THE  PARENTERAL  INTRODUCTION  OF  PROTEINS  371 

complement  of  the  guinea-pig  serum  the  poison  was 
formed.  But  Friedberger  showed  that  kaolin  does  not 
absorb  amboceptor,  but  does  absorb  the  protein  of  horse 
serum,  and  this,  when  acted  upon  by  the  ferment  in  the 
guinea-pig  serum,  furnishes  the  poison.  Bauer1  thinks 
that  the  digesting  serum  becomes  poisonous  from  the  loss 
of  its  complement;  thus,  when  bacteria  are  digested  with 
the  normal  serum  of  the  guinea-pig  the  bacteria  absorb 
the  complement  from  the  serum  and  on  account  of  this 
loss  the  serum  becomes  poisonous.  Bauer  thinks  that  de- 
complemented  serum  acts  as  an  anaphylactic  poison.  This 
claim  is  deserving  of  further  study.  The  ease  with  which 
Friedberger  and  his  students  prepare  anaphylatoxin  from 
all  kinds  of  bacteria  by  digesting  them  with  normal  guinea- 
pig  serum  has  caused  some  to  suspect  some  flaw  in  the 
experiment.  Besredka  and  Strobel2  claimed  that  the 
poisonous  effect  obtained  is  due  to  traces  of  peptone  trans- 
ferred from  the  medium  on  which  the  bacteria  have  been 
grown,  and  that  when  bacteria  grown  on  peptone-free 
media  were  employed  the  results  were  negative.  They  also 
found  that  when  peptone-agar  was  digested  with  normal 
guinea-pig  serum  the  latter  became  poisonous.  In  answer 
to  this  communication,  Lura3  stated  that  bacteria  grown 
on  peptone-free  agar  and  those  grown  on  potato  furnish 
the  poison,  when  digested  with  serum,  just  as  abundantly 
as  those  grown  on  peptone-agar.  The  work  of  Lura  has 
been  confirmed  by  others,  and  if  there  be  a  flaw  in  the 
preparation  of  anaphylatoxin  by  digesting  bacteria  with 
normal  guinea-pig  serum  it  has  not  been  detected  up  to 
the  present  time. 

Pearce  and  Eisenbrey4  showed  by  the  following  experi- 
ment that  the  specific  ferment  formed  in  sensitization  is, 
under  certain  conditions  at  least,  a  product  of  the  fixed 
cells:  "Our  procedure  has  been  to  exsanguinate,  under 

1  Berl.  klin.  Woch.,  1912,  344 

8  Compt.  rend,  de  la  Soc.  biol.,  1911,  Ixxi. 

3  Zeitsch.  f.  Immunitatsforschung,  1912,  xii,  701. 

4  Journal  of  Infectious  Diseases,  1910,  vii,  565. 


372  PROTEIN  POISONS 

ether  anesthesia,  a  small  normal  dog  (A),  and  to  transfuse 
this  animal  by  Crile's  method,  with  the  blood  of  a  larger 
sensitized  dog  (B),  until  the  blood  pressure  reached  approxi- 
mately its  original  level.  After  sufficient  blood  has  been 
obtained  from  B  to  raise  the  pressure  of  A,  the  sensitized 
dog  is  then  bled  to  exsanguination  and  transfused  from  a 
third  normal  dog  (C)  until  its  pressure  reaches  its  previous 
normal  level.  At  the  proper  moment,  the  normal  dog 
containing  the  blood  of  the  sensitized  dog  and  the  latter 
containing  the  blood  of  the  normal  dog,  each  receives 
intravenously  the  toxic  dose  of  horse  serum.  In  the  former, 
a  fall  in  pressure  does  not  occur,  and  in  the  latter  it  does, 
thus  proving  that  the  phenomenon  of  anaphylaxis  is  due 
to  a  reaction  in  the  fixed  cells,  and  not  either  primarily  or 
secondarily  in  the  blood."  That  the  blood  does  under 
certain  conditions  at  least  contain  the  specific  ferment  is 
shown  by  the  production  of  passive  anaphylaxis. 


CHAPTER  XIII 
PROTEIN  FEVER1 

IT  is  interesting  and  instructive  to  read  the  older  litera- 
ture on  fever  in  the  light  of  the  knowledge  which  has  been 
gained  in  the  study  of  sensitization.  It  has  long  been 
known  that  the  parenteral  introduction  of  proteins  in  small 
amounts,  and  especially  repeated  introduction,  leads  to 
fever.  The  older  literature  on  this  subject  as  well  as  an 
account  of  his  own  work  was  given  in  1883  by  Roques.2 
In  1888,  Gamaleia3  showed  quite  clearly  that  fever  accom- 
panies and  results  from  the  parenteral  digestion  of  bacterial 
proteins,  and  a  year  later  Charrin  and  Ruffer4  confirmed 
this  work  and  extended  it  to  non-bacterial  proteins.  In 
1890  Buchner5  produced  the  characteristic  phenomena  of 
inflammation — calor,  rubor,  tumor,  and  dolor — by  the 
subcutaneous  injection  of  diverse  bacterial  proteins.  In 
1895,  Krehl  and  Matthes6  induced  fever  by  the  parenteral 
introduction  of  albumoses  and  peptones,  but  they  did  not 
obtain  constant  results,  which  we  now  know  are  secured 
only  by  regulation  of  the  size  and  frequency  of  the  dosage. 
In  1909,  Vaughan,  Wheeler,  and  Gidley7  demonstrated  that 
any  desired  form  of  fever  (acute  fatal,  continued,  inter- 
mittent, or  remittent)  can  be  induced  in  animals  by  regu- 
lating the  size  and  frequency  of  the  doses  of  foreign  protein 
administered  parenterally,  and  in  1911,  Vaughan,  Gumming, 
and  Wright  extended  the  details  of  this  work. 

1  This  chapter  is  taken  in  part  from  an  article  by  Vaughan,  Gumming, 
and  Wright,  Zeitsch.  f.  Immunitatsforschung,  1911,  ix,  458. 

2  Substances  Thermogenes,  Paris,  1883. 

3  Ann.  de  1'Institut  Pasteur,  xii,  229. 

4  Compt.  rend.  Soc.  de  biol.,  1889,  63. 
o  Berl.  klin.  Woch.,  1890,  216. 

6  Arch.  f.  exp.  Path.  u.  Pharm.,  1895,  xxxv,  232. 

7  Jour.  Amer.  Med.  Assoc.,  August  21,  1909. 


374  PROTEIN  POISONS 

When  a  man  drinks  water  containing  typhoid  bacilli  and 
proves  susceptible,  he  does  not  immediately  manifest  symp- 
toms of  this  disease.  There  is  a  period  of  incubation  which 
in  typhoid  fever  is  about  ten  days.  During  this  time  the 
bacilli  are  multiplying  in  the  man's  body  in  great  numbers, 
and  are  converting  his  proteins  into  bacterial  proteins.  The 
period  of  incubation  stops  and  that  of  the  active  disease 
begins  when  the  cells  of  the  man's  body  become  sensitized, 
elaborate  a  specific  proteolytic  ferment,  and  with  this  begin 
to  split  up  the  foreign  protein. 

The  Production  of  Continued  Fever  in  Rabbits  by  Repeated 
Subcutaneous  Injections  of  Dilutions  of  Egg-white. — If  the 
above  be  true  it  should  be  possible  to  cause  a  continued 
fever  in  animals  by  repeated  injections  of  a  foreign  protein 
in  small  doses.  This  was  first  tried  on  rabbits  with  egg- 
white,  and  the  results  are  shown  in  Fig.  12. 

This  animal  was  kept  under  observation  and  its  tempera- 
ture taken  for  six  days  before  the  injections  were  begun. 
The  temperature  of  this  fore  period  varied  from  101.8°  to 
102.5°  F.  The  injections  consisted  of  egg-white  with  an 
equal  volume  of  0.5  per  cent,  phenol  solution,  and  were 
freshly  prepared  each  day.  The  injections  were  made 
under  the  skin  over  the  back  and  repeated  at  intervals  of 
two  hours  from  7  A.M.  to  9  P.M.  The  urine  was  collected, 
measured,  its  specific  gravity  taken  with  a  picnometer, 
and  its  nitrogen  content  determined  by  the  Kjeldahl  method. 
The  animal  was  weighed  once  a  day.  The  first  injection 
of  2  c.c.  of  the  egg-white  dilution  was  made  at  1  P.M.,  May 
27,  1909.  This  dose  was  continued  at  the  intervals  stated 
until  3  P.M.,  June  1,  when  it  was  doubled,  and  again  doubled 
at  7  A.M.,  June  4.  The  animal  received  40  doses  of  2  c.c. 
each,  20  doses  of  4  c.c.  each,  and  82  doses  of  8  c.c.  each  of 
the  egg-white  dilution;  in  all  816  c.c.  Albumin  appeared 
in  the  urine  when  the  dose  was  increased  to  4  c.c.  The  last 
dose  was  given  at  9  A.M.,  June  15,  after  which  the  albumin 
in  the  urine  gradually  diminished  and  wholly  disappeared 
June  26.  The  day  before  the  first  dose  was  given  the  animal 
weighed  2525  grams,  and  on  the  day  of  the  last  dose  it 


375 


weighed  2180  grams.  After  discontinuing  the  treatment, 
June  15,  the  weight  continued  to  decrease  until  June  21, 
when  it  reached  its  lowest,  1850  grams.  After  this  the 
weight  gradually  increased  until  June  26,  after  which  it 
was  not  taken. 

The  following  figures  give  the  weight  of  the  animal,  the 
amount  of  urine,  the  specific  gravity,  percentage  of  ash  and 
percentage  of  N  in  the  urine : 


Date, 

Weight, 

Urine, 

Specific 

Ash, 

N. 

1909. 

gm. 

c.c. 

gravity. 

per  cent. 

per  cent. 

'  May  26 

.  2525 

May  27  . 

.   2450 

260 

1.0106 

1.61 

0.16 

May  28  . 

.   2440 

125 

1.0124 

2.12 

0.15 

May  29  . 

.   2380 

54 

0.16 

May  30  . 

.  2330 

275 

1.0110 

1.73 

0.19 

May  31  . 

.  2270 

266 

1.0130 

2.41 

0.23 

June  1  . 

.  2355 

104 

1.0294 

3.75 

0.53 

June  2  . 

.  2335 

370 

1.0114 

1.89 

0.16 

June  3  . 

.  2320 

290 

1  .  0093 

1.26 

0.30 

June  4  . 

.  2290 

155 

1.0101 

1.55 

0.28 

June  5  . 

.  2290 

225 

1.0129 

1.20 

0.42 

June  6  . 

.  2250 

95 

1.0148 

1.71 

0.56 

June  7  . 

.  2320 

100 

1.0165 

2.00 

0.66 

June  8  . 

.  2350 

52 

1.0267 

2.02 

2.53 

June  9  . 

.  2355 

115 

1.0134 

1.56 

1.00 

June  10  . 

.  2345 

155 

1.0106 

1.66 

0.56 

June  11  . 

.  2230 

165 

1.0100 

1.64 

0.60 

June  12  . 

.  2330 

95 

1.0170 

.1.56 

0.92 

June  13  . 

.  2280 

190 

1.0140 

1.81 

0.54 

June  14  . 

.  2250 

115 

1.0148 

1.15 

0.98 

June  15  . 

.  2180 

140 

1.0179 

2.60 

0.65 

June  16  . 

.  2130 

155 

1.0194 

2.99 

0.66 

June  17  . 

.  2080 

170 

1.0139 

2.76 

0.50 

June  18  . 

.  2040 

205 

1.0184 

2.10 

0.43 

June  19  . 

.   1960 

150 

1.0198 

2.80 

0.64 

June  20  . 

.   1880 

22 

.  . 

2.90 

June  21  . 

.   1850 

10 

June  22  . 

.  1885 

55 

June  23  . 

.   1900 

June  24  . 

.   1980 

June  25  . 

.  2070 

June  26  . 

.  2120 

It  must  be  admitted  that  this  chart  bears  a  striking 
resemblance  to  one  of  typhoid  fever.  Indeed,  we  have 
here  a  condition  practically  identical  with  an  infectious 


376 


PROTEIN  POISONS 


fever,  and  yet  without  infection.  In  the  infectious  diseases 
the  invading,  multiplying  cell  supplies  the  foreign  protein; 
in  this  experiment  the  supply  of  foreign  protein  has  been 
kept  up  by  the  frequency  of  the  injections.  The  amount  of 
urine  is  variable,  but  averages  less  than  normal  as  it  does 
in  the  continued  fever  of  man.  The  elimination  of  nitrogen 
is  increased.  During  the  twenty-four  hours  immediately 
following  the  first  dose  the  temperature  kept  below  the 
normal  average,  and  each  time  the  dose  was  doubled  a 
marked  rise  followed  in  twenty-four  hours.  There  is  the 


5-2O-0 

1    £ 

-21 

fr 

•It 

s 

i:; 

5 

x 

24 

x 

0- 

.!.-, 

ft. 

21! 

5 

27 

4 

5 

a 

2M 

."> 

2'J 

j 

:.:> 

5- 

x 

:n 

( 

-1 

x 

<;- 

2 

i 

C,- 

S 

G- 

4 

<;-:• 

<      Q- 

/^ 

. 

I 

\ 

/ 

, 

r 

S 

^  J 

\ 

__T 

\ 

^r 

1 

" 

\ 

r 

v- 

\_ 

^ 

^ 

E   F 

Eh 

>ID 

r 

t 

^ 

-\ 

^ 

^ 

-_/ 

V 

-J 

+FIR3T  INJECTION  OF  SERIES 


i-.-ld     r.-l  i     G-12    6-13    6-14    6-15    6-1G    6-17   C-1S 


C-20    C-21    6-22 


= 

a{ 

7 

> 

j 

^ 

i 

> 

^ 

^ 

< 

^ 

^ 

? 

T 

•s 

I 

• 

* 

J 

^ 

^ 

S. 

• 

x 

a 

x. 

X 

x 

*. 

a. 

x. 

a    a 

r^ 

V 

V 

\ 

r^ 

•^ 

'VV, 

A 

-T 

\ 

__  f 

y 

1  \ 

^ 

f" 

^ 

A 

^ 

'  

~J- 

'    " 

^A 

^ 

N 

yv 

^/ 

^ 

- 

k^- 

101 

* 

+  LA3T  INJECTION  OF  SERIES 


FIG.  12. — The  production  of  continued  fever  in  a  rabbit  by  repeated 
subcutaneous  injections  of  egg-white. 

gradual  rise  in  the  morning  temperature,  so  frequently  seen 
in  typhoid  fever  and  the  fact  that  for  the  most  part  the 
highest  temperature  for  the  day  falls  in  the  afternoon  is 
also  interesting.  When  the  injections  were  discontinued 
the  temperature  gradually  fell  and  remained  somewhat 
below  the  normal,  as  is  often  observed  in  convalesence 
from  typhoid  fever. 

We  wished  to  ascertain  the  source  of  the  albumin  in  the 
urine;  did  it  consist  wholly  of  egg  albumen  or  serum  albumin, 
or  did  it  contain  both?  In  order  to  determine  this  we 
injected  2  c.c.  of  the  filtered  urine  on  June  18,  intra-abdomi- 


PROTEIN  FEVER  377 

nally,  into  each  of  5  guinea-pigs  for  the  purpose  of  sensitizing 
them  to  whatever  proteins  the  urine  might  contain.  Twelve 
days  later  2  of  these  animals  received  intra-abdominally, 
each  5  c.c.  of  egg-white  dilution  (with  an  equal  volume  of 
water);  2  others  had,  each  5  c.c.  of  fresh  serum  from  a 
rabbit,  and  the  fifth  had  a  mixture  of  2.5  c.c.  of  each  of 
these  fluids.  All  were  found  to  be  sensitized,  thus  showing 
the  presence  of  both  egg-white  and  serum  protein  in  the 
urine  of  the  febrile  rabbit. 

It  is  worthy  of  note  that  while  these  animals  developed 
the  three  stages  characteristic  of  protein  sensitization,  the 
second  and  third  stages  were  unusually  prolonged  and  less 
acute  than  those  generally  observed  in  sensitized  guinea- 
pigs.  Of  the  two  treated  with  egg-white,  one  died  at  the 
end  of  two  hours  and  the  other  fifteen  minutes  later.  Of 
the  two  treated  with  rabbit  serum,  one  died  at  the  end  of 
one  hour,  while  the  other  lived  for  three  hours.  The  one 
that  had  the  mixture  of  proteins  developed  the  symptoms 
more  promptly  than  any  of  the  others,  but  did  not  die. 

A  continued  fever  was  maintained  in  another  rabbit  by 
injections  of  the  same  strength  of  egg-white  solution  from 
April  30  to  May  18,  1909.  In  this  instance  the  size  of  the 
dose  was  not  altered.  The  animal  received  four  doses  daily 
from  April  30  to  May  11,  after  which  five  were  given  until 
May  15,  and  then  for  three  days  we  returned  to  four  doses 
daily.  The  fever  continued,  after  the  injections  were 
discontinued,  until  the  evening  of  May  20,  when  it  fell  by 
crisis  below  the  normal,  slowly  returning  to  the  normal. 
The  urine  was  collected  and  nitrogen  determined  as  in  the 
other  instance,  but  the  charts  are  so  similar  that  we  do 
not  consider  it  necessary  to  present  the  second  one. 

The  Production  of  Continued  Fever  in  Rabbits  by  Repeated 
Subcutaneous  Injections  of  the  Poisonous  Group  of  the  Typhoid 
Protein. — Fig.  13  shows  the  effects  of  repeated  subcutaneous 
injections  of  sublethal  doses  of  the  poisonous  group  split  off 
from  the  cellular  substance  of  the  typhoid  bacillus  with  a  2 
per  cent,  solution  of  sodium  hydroxide  in  absolute  alcohol. 

The  material  used  was  the  crude  soluble  poison  containing 


378 


PROTEIN  POISONS 


about  10  per  ceixt.  of  the  poison  in  the  purest  form  in  which 
we  have  been  able  to  obtain  it.  This  crude  soluble  poison 
was  administered  every  two  hours  from  7  A.M.  to  9  P.M 
from  May  3  to  May  18,  1909.  Each  dose  consisted  of 
200  mg.  of  the  crude  poison,  300  mg.  being  a  fatal  quantity 
for  rabbits  of  the  size  used. 

When  a  fatal  dose  of  the  protein  poison  is  administered 
the  temperature  rapidly  falls,  but  with  smaller  repeated 
doses  a  continued  fever  results.  The  more  nearly  the  dose 
approaches  the  fatal  amount  the  more  speedily  will  the 

5-3.09     .'.-I         5-5        .-,-;!       5-7        5-8        .,.'.)       5-19    5-11       5-12    5-13       5-14 


X 

z 

Z 

z 

s 

2 

T 

i 

z 

z 

£ 

2 

z 

2 

z 

z 

z 

? 

z 

z 

2 

I 

2       Z 

A 

A 

,A 

a 

u 

i 

~l  , 

. 

A 

z 

p 

\J 

V 

V 

/V 

N  / 

-s/ 

•Vs 

\; 

k/^ 

\ 

j 

31— 

t 

v 

V, 

\J 

1(1". 

I 

* 

V 

+  FIRST  INJECTION  OF  SERIES. 


5-15      5-10      :.-17      5-18      5-19      5-20    5-21      5-22      5-23     5-24      5-25      5-20 


X       Z 

m 

^ 

5 

a 

^ 

a 

• 

2 

Z 

• 

2 

£ 

2 

2 

£ 

Z 

2 

3: 

2 

2 

X     X 

\l* 

s/ 

/\ 

A 

^ 

7  L 

SA 

N 

\r 

^-v 

VL 

_—  •*-' 

^ 

^ 

,> 

<\ 

-^ 

+  LAST  INJECTION  OF  8FRIE8. 

FIG.  13. — The  production  of  continued  fever  in  a  rabbit  by  repeated 
subcutaneous  injections  of  the  poisonous  group  from  the  typhoid  bacillus. 

animal  succumb.  Death  may  be  sudden  and  under  a  high 
temperature,  or  it  may  be  slow  and  preceded  by  a  fall  of 
several  degrees  below  the  normal.  From  sublethal  doses 
of  the  poison,  animals  recover  quickly  and  apparently, 
completely.  This  seems  to  indicate  that  the  poisonous 
effects  are  quickly  neutralized  in  the  animal  body,  but  we 
are  of  the  opinion  that  this  neutralization  is  secured  by 
more  or  less  chemical  disintegration  in  the  protein  molecule 
of  certain  cells  in  the  body.  The  protein  poison  is  acid  to 
litmus,  and  we  have  secured  continued  fever  in  rabbits  by 
repeated  injections  of  either  the  acid  solution  or  the  same 


PROTEIN  FEVER  379 

after  neutralization  with  sodium  bicarbonate.  Fig.  13 
needs  no  further  explanation. 

The  Effects  of  Intra-abdominal  Injections  of  Egg-white. — 
Large  single  or  repeated  doses  of  egg-white  injected  intra- 
abdominally  in  non-sensitized  rabbits  have  but  little  effect 
on  the  temperature.  Generally  the  temperature  runs  slightly 
subnormal  after  such  injections. 

August  22,  1909,  we  injected  the  whites  of  three  eggs 
into  the  abdominal  cavity  of  a  rabbit.  The  highest  tem- 
perature of  the  fore  period  was  100.9°.  After  the  injection 
the  temperature  was  taken  every  two  hours  from  8  A.M. 
to  6  P.M.  up  to  September  6.  The  animal  was  weighed 
each  day  and  its  urine  measured  and  tested  for  albumin. 
There  was  no  fever;  indeed,  the  morning  temperature 
fell  some  days  to  97°  and  one  day  to  96.6°.  The  animal  lost 
in  weight,  slightly  more  than  one-fifth  of  its  original  weight. 
The  volume  of  urine  averaged  normal,  and  at  no  time  did 
it  contain  albumin. 

On  the  other  hand,  0.05  c.c.  of  egg-white,  filtered  through 
cotton,  injected  intra-abdominally  every  half  hour  from 
8  A.M.  until  4  P.M.  produced  the  following  results: 

Time.  Dose  in  c.c.               Temperature. 

8.00  A.M.                   0.05  102.8° 

8.30  0.05  101.0° 

9.00  0.05  102.6° 

.9.30  0.05  102.9° 

10.00  0.05  102.6° 

10.30  0.05  103.2° 

11.00  0.05  103.4° 

11.30  0.05  103.5° 

12.00  0.05  103.4° 

12.30  P.M.                   0.05  104.4° 

1.00  0.05  104.8° 

1.30  0.05  105.1° 

2.00  0.05  105.3° 

2.30  0.05  105.3° 

3.00  0.05  105.8° 

3.30  0.05  105.8° 

4.00  0.05  105.8° 

4.30  0  106.6° 

5.00  0  106.6° 

5.30  0  106.1° 

8.00  0  104.1° 

8.00  A.M.                      0  103.0° 


380  PROTEIN  POISONS 

The  Production  of  Fever  in  Rabbits  by  Repeated  Intra- 
venous Injections  of  Egg-white. — The  injection  of  a  large 
amount  of  egg-white  intravenously  in  a  single  or  in  repeated 
doses  in  non-sensitized  rabbits  does  not  cause  any  marked 
elevation  of  temperature. 

After  keeping  a  rabbit  under  observation  for  three  days 
and  finding  that  its  temperature  at  no  time  reached  102°, 
we  injected  into  its  ear  vein  every  two  hours  from  8  A.M.  to 
6  P.M.  4  c.c.  of  a  dilution  of  egg-white  with  an  equal  volume 
of  physiological  salt  solution,  which  dilution  had  been 
passed  through  a  Berkefeld  filter,  and  each  cubic  centimeter 
of  which  contained  26  mg.  of  protein  as  ascertained  by  a 
Kjeldahl  determination.  This  dosage  was  continued  for 
six  days.  During  the  greater  part  of  this  time  the  tempera- 
ture, which  was  taken  before  each  injection,  remained 
normal,  sometimes  subnormal,  and  only  once  did  it  reach 
102°.  Then  the  dose  was  increased  to  10  c.c.  and  con- 
tinued for  four  days.  Twenty-four  hours  after  the  increase 
in  dose  there  was  an  irregular,  but  not  marked  elevation 
of  temperature,  the  highest  point  reached  being  104.4°. 
During  the  whole  of  the  time  the  animal  seemed  quite  well. 
Its  greatest  weight  was  observed  during  the  time  when  the 
largest  injections  were  being  given,  and  at  the  same  time 
the  daily  elimination  of  urine  greatly  increased,  from  114 
c.c.,  the  average  of  the  fore  period,  to  as  much  as  650  c.c. 
at  the  time  of  the  largest  injections.  The  urine  was  tested 
daily  for  albumin  with  negative  results.  This  experiment  was 
continued  from  August  6  to  19,  1909.  In  March,  1910, 
a  single  injection  of  5  c.c.  of  the  dilution  of  egg-white  was 
followed  by  a  gradual  rise  in  temperature  to  105.8°  within 
a  few  hours. 

In  order  to  induce  fever  in  rabbits  by  the  intravenous 
injection  of  dilutions  of  egg-white  the  doses  must  be  small 
and  the  most  striking  results  are  obtained  when  the  size 
of  the  dose  is  gradually  increased.  We  have  made  many 
experiments  along  this  line  and  some  of  them  will  be 
detailed. 


PROTEIN  FEVER 


381 


Group  /.—The  initial  dose  in  this  group  was  1  c.c.  of 
egg-white  diluted  with  an  equal  volume  of  physiological 
salt  solution  and  passed  through  a  Berkefeld  filter.  Each 
cubic  centimeter  of  this  dilution  contained  26  nag.  of  protein. 
The  doses  were  increased  by  1  c.c.  at  each  repetition,  which 
was  hourly. 

In  rabbit  No.  2  the  highest  temperature  of  the  fore  period 
was  101.4°.  The  following  table  shows  the  results: 


Time. 


8.00 

9.00 

10.00 

11.00 

12.00 


3.00 


5.00 


Dose  in  c.c. 
1 
2 
3 
4 
5 
6 
7 
8 
9 
0 
0 


Temperature. 
100.8° 
103.2° 
104.3° 
104.6° 
105.4° 
105.6° 
106.1° 
105.2° 
102.2° 
100.5° 
Death 


This  animal  died  with  a  sudden  convulsive  movement. 
The  urine  of  the  day  and  that  in  the  distended  bladder  after 
death  was  tested  for  albumin  with  negative  results.  Fig.  14 


106 

2  F 

M. 

4  P.M. 

6  P.M. 

1  CC. 

9  A.M. 
t  CC. 

10  A.M. 
3  CC. 

11  A.M. 
4  CC. 

I  CC. 

1  P.M. 
6CC. 

t  P.M. 
7  CC. 

3  :>.M. 
8  CC. 

4  P.M. 
9  CC. 

5  P 

M. 

^J^- 

\ 

105 
10-1 
1C3 
102 

101 
11)0 

/ 

.  

\ 

t 

/ 

f 

/ 

\ 

/ 

\ 

/ 

\ 

/ 

/ 

\ 

/ 

\ 

S 

~—~ 



X 

/ 

\ 

"^ 

\ 

h- 

XFIRST  INJECTION  CF  SERIES 


-{-DEATH 


FIG    14. — Acute  fatal  fever  produced  in  a  rabbit  by  repeated 
intravenous  injections  of  egg-white. 


382  PROTEIN  POISONS 

shows  the  temperature  curve  of  this  animal,  including  the 
long  fore  period. 

In  No.  5  the  highest  temperature  of  the  fore  period  was 
101.1°. 

Time.  Dose  in  c.c.               Temperature. 

8.00  A.M.                      0  100.1° 

9.00  1  100.4° 

10.00  2  101.3° 

11.00  3  102.2° 

12.00  4  102.7° 

1.00  PM.                      5  102.8° 

2.00  6  103.5° 

3.00  7  104.7° 

4.00  8  105.1° 

5.00  9  105.6° 

6.00  10  104.0° 

7.00  11  103.8° 

8.00  12  104.0° 

8.30  0  Death 

This  animal  showed  no  symptoms  until  the  final  con- 
vulsive movement. 
The  following  is  the  record  of  No.  34. 


Time. 

Dose  in  c.c. 

Temperature. 

9.00     A.M. 

1 

102.8° 

10.00 

2 

104.0° 

11.00 

3 

105.4° 

12.00 

4 

105.2° 

1.00     PM. 

5 

105.4° 

2.00 

6 

105.4° 

3.00 

7 

106.0° 

4.00 

8 

105.4° 

5.00 

9 

105.3° 

6.00 

10 

105.8° 

By 

10  P.M.  the  temperature  had  fallen  to  102.2°. 

The 

record  of  No. 

35  is  shown  by 

the  following: 

Time 

Dose  in  c  c. 

Temperature. 

9.00    A.M. 

1 

102.8° 

10.00 

2 

103.6° 

11.00 

3 

105.4° 

12.00 

4 

105.6° 

1.00     P.M 

5 

106.2° 

2.00 

6 

106.7° 

3.00 

7 

106.6° 

4.00 

8 

106.8° 

5.00 

9 

106.6° 

PROTEIN  FEVER 


383 


By  10  P.M.  the  temperature  had  fallen  to  103.4°  and  the 
next  morning  the  animal  was  apparently  normal.  Twenty- 
six  days  later  this  animal  was  treated  in  the  same  way, 
with  fatal  results.  Fig.  15  shows  the  curve  for  both 
treatments. 


1 

107 
100 
105 
104 
103 
102 

O-20-O9  10-27-09 

10-28-O9 

I  f 

M. 

1  CO. 

2 

C. 

3  C 

c. 

4  CC. 

5  CC. 

6  CC. 

7  1 

C. 

1 

C. 

91 

C. 

to 

M. 

,c. 

/ 



,  —  - 

~~- 

-—  _ 

--—  - 

.  1 

/ 

/ 

\ 

^. 

X 

\ 

' 

/ 

/ 

/ 

r 

ORE 

PER 

OC 

/ 

^ 

/ 

x 

\ 

/ 

N 

/ 

FIG.  15. — Acute  fever  produced  in  a  rabbit  by  intravenous  injections 
of  egg-white.  The  continuous  line  represents  the  temperature  in  the 
non-sensitized  animal.  The  broken  line  represents  the  temperature  in 
the  same  animal,  sensitized. 


Group  II. — The  dilution  used  contained  13  mg.  of  protein 
in  each  cubic  centimeter. 


384  PROTEIN  POISONS 

The  following  is  the  record  of  No.  15: 

Time.  Dose  in  c.c.                Temperature. 

9.00  A.M.                      1  102.6° 

10.00  2  104.0° 

11.00  3  104.9° 

12.00  4  105.8° 

1.00  P.M.                      5  105.1° 

2.00  6  105.6° 

3.00  7  105.6° 

4.00  8  106.2° 

5.00  9  105.6° 

6.00  10  104.2° 

The  temperature  gradually  fell,  and  the  next  morning 
at  10  it  was  101.8°. 
The  following  is  the  record  of  No.  16: 

Temperature. 
103.5° 
104.1° 
103.4° 
104.2° 
104.5° 
104.5° 
103.8° 
103.4° 
103.2° 
104.2° 

The  next  morning  the  temperature  was  101.8°. 
Group  III. — Each    cubic    centimeter    of    the    dilution 
contained  6.5  mg.  of  protein. 
The  following  is  the  record  of  No.  18: 

Time.  Dose  in  c.c.                Temperature. 

8.00  A.M.                      1  103.2° 

9.00  2  102.0° 

10.00  3  102.5° 

11.00  4  103.8° 

12.00  5  104.0° 

1.00  P.M.                     6  104.6° 

2.00  7  104.0° 

3.00  8  103.6° 

4.00  9  104.0° 

5.00  10  103.8° 

6.00  11  103.8° 


Time. 

Dose  in  c.c. 

9.00     A.M. 

1 

10.00 

2 

11.00 

3 

12.00 

4 

1.00     P.M. 

5 

2.00 

6 

3.00 

7 

4.00 

8 

5.00 

9 

6.00 

10 

PROTEIN  FEVER 


385 


The  temperature  was  normal  the  next  morning. 

After  an  interval  of  174  days  this  animal  was  treated  in 
the  same  way,  with  a  fatal  ending.  Fig.  16  shows  the  curve 
for  both  treatments. 


10-11-O9     10-12-09 


M 

M. 

-,. 

Cr- 

-* 

- 

-« 

-^ 

10-1 

X 

^ 

x 

\ 

^ 

---J 

X 

,oc 

/ 

^-- 

•—  «, 

^- 

<^ 

^ 

/ 

-^, 

.-~~ 

\ 

/ 

.in-.; 

\ 

^ 

^ 

+FIRST   INJECTION 


N— 


FIG.  16. — Acute  fever  produced  in  a  rabbit  by  the  intravenous  injection 
of  egg-white.  The  continuous  line  represents  the  temperature  in  the  non- 
sensitized  animal.  The  broken  line  represents  the  temperature  in  the 
same  animal,  sensitized. 

Group  IV. — Each  cubic  centimeter  of  the  dilution  con- 
tained 3.25  mg.  of  protein. 
The  following  is  the  record  of  No.  20: 


Time. 


9.00 

10.00 

11.00 

12.00 

1.00 

2.00 

3.00 

4.00 

5.00 


Dose  in  c.c. 
1 
2 
3 
4 
5 
6 
7 
8 
9 


Temperature. 
103.0° 
103.0° 
103.4° 
103.8° 
104.0° 
103.8° 
104.2° 
104.4° 
104.4° 


25 


386  PROTEIN  POISONS 

The  following  is  the  record  of  No.  21 : 

Time.  Dose  in  c.c.                Temperature. 

9.00  A.M.                     1  103.0° 

10.00  2  101.8° 

11.00  3  102.4° 

12.00  4  102.8° 

1.00  P.M.                     5  103.2° 

2.00  6  104.0° 

3.00  7  103.6° 

4.00  8  103.8° 

5.00  9  104.0° 

After  an  interval  of  140  days  this  animal  was  again 
treated  with  the  following  record: 

Time.  Dose  in  c.c.                Temperature. 

9.00  A.M.                     1  102.0° 

10.00  2  103.4° 

11.00  3  105.2° 

12.00  4  105.8° 

1.00  P.M.                     5  106.6° 

2.00  6  106.4° 

3.00  7  105.6° 

4.00  8  106.2° 

5.00  9  105.6° 

6.00  10  106.4° 

7.00  0  106.2° 

8.00  0  105.6° 

9.00  0  106.4° 

10.00  0  103.4° 

After  the  second  injection  Cheyne-Stokes  respiration 
appeared  and  continued  through  the  day,  but  the  injections 
were  continued  until  6  P.M.,  and  the  animal  recovered. 

Group  V, — In  this  group  but  one  injection  was  made 
each  day.  The  dilution  contained  26  mg.  of  protein  in 
each  cubic  centimeter.  The  beginning  dose  was  3  c.c. 
and  each  day  it  was  increased  by  2  c.c.  In  No.  6  the  dose 
was  given  each  day  at  10  A.M.,  and  was  repeated  for  seven 
consecutive  days.  The  beginning  dose  was  3  c.c.,  two  of 
the  larger  doses  being  repeated.  In  this  way  a  well-marked 
intermittent  fever  of  mild  type  was  established.  After 


PROTEIN  FEVER  387 

each  injection  the  temperature  arose  within  from  two  to 
four  hours,  and  returned  to  normal  during  the  evening, 
and  continued  so  until  the  next  injection.  With  increase 
in  the  size  of  the  dose  the  tendency  was  to  take  the  remittent 
type,  so  that  at  no  time  of  the  day  did  the  fall  quite  reach 
the  normal  limit. 

Group  VI. — We  have  found  some  rabbits  which  do  not 
respond  to  the  pyrogenic  effect  of  intravenous  injections 
of  egg-white  until  the  size  of  the  dose  is  reduced.  On 
receiving  a  new  consignment  of  rabbits,  we  were  surprised 
to  find  that  these  animals  did  not  develop  fever  on  receiving 
repeated  doses  of  a  dilution  of  egg-white  (1  to  1)  with  salt 
solution.  The  following  illustrates  our  experience: 

No.  101.  This  animal,  weighing  about  2600  grams,  was 
treated  with  increasing  doses  of  the  dilution  (1  to  1): 


Time.  Dose  in  c.c.  Temperature. 

9.00  A.M.  1  102.0° 

10.00  2  100.4° 

11.00  3  100.5° 

12.00  4  100.2° 

1.00  P.M.  5  100.0° 

2.00  6  96.4° 

3 . 00  0  Death 


The  abdominal  cavity  was  filled  with  bloody  fluid.  The 
urine  found  in  the  bladder  contained  a  small  amount  of 
albumin. 

No.  102.  Thinking  that  No.  101  was  an  individual 
exception,  No.  102  was  treated  with  the  same  dilution: 


Time.  Dose  in  c.c.                Temperature. 

10.00  A.M.                      1  101.2° 

11.00  2  101.4° 

12.00  3  101.5° 

1.00  P.M.                     4  101.5° 

2.00  5  101.8° 

3.00  6  101.3° 

4.00  7  101.5° 

4.45  0  Death 


388  PROTEIN  POISONS 

No.  103.  In  this  experiment  the  egg-white  dilution  was 
reduced  (1  to  2) : 

Time.  Dose  in  c.c.                Temperature. 

10.00  A.M.                      1  101.4° 

11.00  2  101.7° 

12.00  3  102.0° 

1.00  P.M.                     4  102.5° 

2.00  5  102.8° 

3.00  6  102.4° 

4.00  7  101.8° 

5.00  0  97.6° 

6.00  0  97.4° 

6 . 30  0  Death 

No.  104.  In  this  experiment  the  dilution  was  further 
reduced  (1  to  3). 

Time.  Dose  in  c.c.                Temperature. 

7.30  A.M.                      1  101.6° 

8.30  2  101.2° 

9.30  3  101.7° 

10.30  4  101.8° 

11.30  5  102.6° 

12.30  P.M.                     6  103.4° 

1.30  7  103.8° 

2.30  8  104.1° 

3.30  9  104.4° 

4.30  0  104.5° 

5.30  0  105.2° 

These  differences  in  response  to  the  injections  of  egg- 
white  seem  to  be  characteristic  of  certain  groups  or  batches 
of  animals  and  not  of  individuals  within  the  group.  Our 
laboratory  buys  its  rabbits  in  lots  of  from  twenty-five  to 
one  hundred.  If  one  out  of  a  given  lot  responds  to  a  certain 
dose  of  egg-white  we  have  found  that  others  of  the  same 
lot  respond  in  much  the  same  way.  Whether  this  is  due 
to  differences  in  food  or  in  breed  we  have  not  determined, 
but  are  inclined  to  believe  that  breed  has  much  to  do  with 
it.  Possibly,  age  is  an  individual  factor. 

A  Brief  Statement  of  the  Autopsy  Findings  in  Acute  Poison- 
ing of  Rabbits  with  Egg-white. — This  work  has  been  turned 
over  to  our  colleagues  of  the  department  of  pathology,  and 


PROTEIN  FEVER  389 

Morse  has  been  kind  enough  to  make  a  few  autopsies  for 
us.  We  make  a  short  abstract  of  his  report: 

Rabbit  A  received  intravenously  four  doses  of  10  c.c. 
each  of  a  dilution  of  egg-white  with  an  equal  volume  of 
physiological  salt  solution.  The  doses  were  administered 
at  intervals  of  one  hour.  During  the  administration  of  the 
fourth  dose  the  animal  died  in  convulsions.  A  gray  rabbit 
of  average  size  and  well-nourished;  external  orifices  are 
normal;  mucous  membranes  cyanotic;  body,  cold;  rigor 
mortis,  moderate.  The  peritoneal  cavity  contains  a  small 
amount  of  blood-tinged,  serous  fluid.  Superficial  inspection 
reveals  nothing  else  of  note.  There  is  no  displacement  of 
organs,  no  peritonitis,  no  area  of  hemorrhage  in  the  serosa. 
The  pericardial  sac  is  normal.  There  is  no  increase  of 
pericardial  fluid  and  the  pleural  cavities  are  dry.  In  the 
anterior  mediastinum  there- is  a  moderate  amount  of  pale 
fat  with  a  few  petechial  hemorrhages.  The  thymus  is 
large,  swollen,  and  edematous.  It  spreads  over  the  anterior 
mediastinum  covering  the  great  vessels  and  it  contains 
hundreds  of  miliary  hemorrhages.  There  are  no  large 
areas  of  blood  in  the  tissue.  The  heart  is  moderately 
dilated  and  filled  with  red  clot.  There  is  no  imbibition 
of  hemoglobin  in  the  intima  of  the  great  vessels.  The 
heart  valves  are  normal,  the  myocardium  is  darker  than 
normal,  and  drips  blood  too  freely.  The  lungs  are  reddish 
pink,  though  slightly  darker  than  normal  and  rather  moist 
on  section.  There  is  no  pneumonia  and  no  solid  areas  are 
seen  in  the  lung  tissue.  The  spleen  is  slightly  congested 
and  darker  than  normal.  The  kidneys  are  dark,  congested, 
and  drip  blood  on  section.  The  adrenals  are  apparently 
normal.  The  stomach  and  intestine  show  no  abnormality. 
The  bladder  is  empty  and  normal  in  appearance.  The  liver 
is  large,  dark,  and  bleeds  freely  on  section.  In  the  retro- 
peritoneum  there  is  a  moderately  large  suffusion  of  blood 
through  the  cellular  tissue  and  partially  involving  the  head 
of  the  pancreas.  The  brain  appears  normal,  but  section 
shows  the  tissue  somewhat  congested  and  moist. 

The  chief  microscopic  findings  may  be  stated  as  follows: 


390  PROTEIN  POISONS 

The  myocardium  shows  slight  increase  in  hemofuscin, 
and  there  seems  to  be  an  excessive  fragmentation  of  the 
muscle  bundles  (myocardite  segmentaire  of  Renault). 
However,  this  may  be  due  to  the  fixing  fluid.  The  most 
striking  thing  is  the  occurrence  of  numerous  miliary  hemor- 
rhages into  the  muscle  substance,  forcing  the  fibers  apart 
in  places.  There  is  a  diffuse  distribution  of  blood  throughout 
the  heart  muscle,  blood  cells  being  found  here  and  there 
outside  the  capillaries  in  the  muscles,  forcing  the  fibers 
apart  as  though  there  had  been  a  general  diapedesis.  The 
ventricular  cavity  shows  a  homogeneous  red  clot.  The 
lungs  show  marked  acute,  passive  congestion  and  localized 
areas  of  moderate  congestion.  There  are  a  few  small 
hemorrhages  near  the  veins  along  the  bronchi  and  also 
beneath  the  pleura.  The  liver  shows  extreme  passive 
congestion,  all  the  capillaries  being  gorged  with  blood. 
The  whole  liver  substance  appears  as  though  soaked  in 
blood,  which  lies  everywhere,  between  the  liver  cells  and 
in  the  bile  capillaries.  Some  of  the  smaller  ducts  contain 
blood.  The  liver  cells  have  a  cloudy  appearance  and  the 
nuclei  are  farther  apart  than  normal,  due  to  the  engorgement 
with  blood.  The  kidneys  show  some  cloudy  swelling  and 
marked  acute  passive  congestion.  A  condition  similar  to 
that  seen  in  the  heart  and  liver  is  found  throughout  the 
kidney.  The  renal  tissue  is  full  of  miliary  hemorrhages, 
and  appears  to  be  soaked  in  blood.  Everywhere  between 
the  tubules  and  scattered  throughout  the  parenchyma  are 
red  blood  cells.  Many  of  the  glomeruli  have  red  blood 
cells  lying  free  within  Bowman's  capsules,  and  the  capillaries 
of  the  tufts  are  extremely  dilated.  The  epithelium  of  the 
proximal  convoluted  portion  of  the  tubules  is  markedly 
desquamated.  Many  of  the  collecting  tubules  contain 
pale  blood  cells  which  have  lost  their  hemoglobin.  The 
pelvic  fat  is  in  part  displaced  by  large  suggillations  of  blood, 
and  there  are  a  few  areas  of  coagulated  blood  around  the 
kidney  capsule.  The  pancreas  is  passively  congested  and  a 
portion  of  this  organ  has  been  included  in  a  large  clot  in  the 
retroperitoneal  region  and  is  wholly  necrotic.  There  is  also 
marked  fat  necrosis  and  infiltration  of  the  tissue  with  blood. 


PROTEIN  FEVER  391 

In  sensitized  rabbits  killed  by  injections  repeated  after 
six  months,  Morse  has  found  the  microscopic  lesions  of  the 
same  character,  but  much  less  marked  than  those  described 
above  as  resulting  from  acute  poisoning.  In  fresh  rabbits 
hemolysis  and  hemorrhage  seem  sufficient  to  account  for 
death,  but  this  does  not  appear  to  be  the  case  in  sensitized 
animals  dying  suddenly  from  relatively  small  doses. 

The  Effects  of  Intravenous  Injections  of  Laked  Human  Red 
Corpuscles  on  the  Temperature  of  Rabbits. — The  blood  was 
drawn  into  a  solution  of  sodium  citrate,  and  the  corpuscles 
thrown  down  in  a  centrifuge.  The  corpuscles  were  repeatedly 
washed  with  physiological  salt  solution,  and  then  dissolved  in 
distilled  water  and  diluted  to  the  volume  of  the  original  blood. 

One  dose  of  5  c.c.  of  this  solution  was  injected  into  the 
ear  vein  of  rabbit  No.  7.  The  highest  temperature  of  the 
fore  period  was  101.8°.  The  effects  of  this  injection  are 
shown  by  the  following  figures: 

Time.  Dose  in  c.c.                Temperature. 

8.00  A.M.                     0  101.8° 

9.00  5  101.5° 

11.00  0  104.2° 

12.00  0  105.2° 

1.00  P.M.                     0  106.0° 

2.00  0  105.6° 

3.00  0  105.6° 

4.00  0  104.7° 

5.00  0  104.2° 

6.00  0  104.0° 

8.00  A.M.                     0  101.8° 

In  No.  10,  2  c.c.  of  the  same  solution  had  the  effect 
showrn  in  the  following: 

Time.  Dose  in  c.c.                Temperature. 

8.00  A.M.                     0  103.0° 

10.00  2.  103.1° 

11.00  0  104.4° 

12.00  0  105.6° 

1.00  P.M.                     0  107.0° 

2.00  0  106.4° 

3.00  0  106.2° 

4.00  0  105.6° 

6.00  0  104.6° 

8.00  A.M.                     0  102.8° 

12.00  0  102.6° 


392 


PROTEIN  POISONS 


It  will  be  observed  that  this  animal  had  a  temperature 
of  103.1°  before  the  injection  was  made.  Fig.  17  gives  this 
record. 

In  No.  11,  1  c.c.  of  the  same  solution  had  the  following 
effect : 


Time.                 Dose  in  c.c.                Temperature. 
8.00    A.M.                     0                               102.3° 
9.00                                1                                102.0° 
10.00                               0                               104.2° 
11.00                               0                               104.4° 
12.00                               0                               105.  4~ 
1.00    P.M.                     0                               104.0° 
2.00                               0                               103.0° 
3.00                               0                               102.2° 
4.00                               0                               102.6° 

9-294)9                    !i.:;ii.iiit                                                                                            10-1-09                  1(1-24)9          10-3-0 

4PM. 

(.A 

M 

l! 

M 

4t 

M. 

M 

M. 

M 

M 

10* 

11A 

.V 

" 

w. 

7 

^k 

tf 

M. 

3t 

M. 

M. 

V 

M. 

SA 

.M. 

'• 

"' 

4C 

H. 

CA 

M. 

:i 

M.     4P 

t 

' 

"  —  . 

'  . 

/ 

^ 

\ 

/ 

\ 

/ 

s 

/ 

\ 

/ 

s 

? 

—  _ 

\ 

X 

—  —  ' 

_± 

\ 

1  ^ 

X 

1  

^^ 

X 

^^~~ 

+DOSE  2  C.C. 

FIG.  17. — Acute  fever  produced  in  a  rabbit  by  an  intravenous  injection 
of  washed  human  blood  cells  hemolyzed. 


Fig.  18  gives  the  curve  in  this  case. 


100 
lU'. 
H4 

m 

100 
101 
HO 

9-33- 

*.    i 
S    ? 

99    9-24         9-25        9-26        9-27         9-28         9-29         9-80                  .      10-t                         10-3 

1 

\ 

/ 

\ 

^ 

\ 

1 

\ 

1 

\ 

/ 

s 

\ 

X 

\ 

s 

/ 

\ 

/ 

\ 

/ 

\ 

/ 

x 

\x 

/ 

/ 

\ 

/ 

s 

s 

^ 

/• 

/ 

/ 

\ 

/ 

/ 

\ 

/ 

\ 

/ 

\ 

/ 

\ 

+  DOSE  1  C.C. 

FIG.  18. — Acute  fever  produced  in  a  rabbit  by  an  intravenous 
injection  of  washed  human  blood  cells,  hemolyzed. 


PROTEIN  FEVER 


393 


In  No.  12,  0.5  c.c.  of  the  same  solution  produced  the 
following  effects: 


Time. 

8.00    A. 
10.00 
11.00 
12.00 

1.00    P. 

2.00 

3.00 

4.00 


Dose  in  c.c. 

0 

0.5 
0 
0 
0 
0 
0 
0 


Temperature. 
102.4° 
102.0° 
103.0° 
104.0° 
105.2° 
104.4° 
103.4° 
102.4° 


The  highest .  temperature  in  the  fore  period  covering 
seven  days  was  102.6°. 

In  No.  17,  10  c.c.  of  the  same  solution  caused  a  precipitate 
fall  in  temperature,  and  death  in  seven  hours. 

The  laked  blood  corpuscles  of  either  man  or  rabbit  after 
filtration  cause  an  elevation  of  temperature  when  injected 
into  rabbits  either  intra-abdominally  or  intravenously. 

In  rabbit  No.  55,  rabbits'  corpuscles  prepared  as  already 
stated  and  filtered  were  injected  intra-abdominally  as 
shown  by  the  following  figures: 


Time. 


8.00 

9.00 

10.00 

11.00 

12.00 


.00 
.00 


3.00 
4.00 
5.00 


Dose  in  c.c. 
0.1 
0.2 
0.3 
0.4 
0.5 
0.6 
0.7 
0.8 
0.9 
0 


Temperature. 
102.8° 
102.2° 
102.8° 
102.8° 
104.3° 
105.2° 
105.1° 
104.8° 
104.4° 
104.6° 


394 


PROTEIN  POISONS 


In  No.  56  the  unfiltered  laked  corpuscles  were  injected 
intra-abdominally : 


Time. 


8.00 

A.M. 

8.30 

9.00 

9.30 

10.00 

10.30 

11.00 

11.30 

12.00 

12.30 

P.M. 

1.00 

1.30 

2.00 

2.30 

3.00 

3.30 

4.00 

4.30 

5.00 

5.30 

6.00 

6.30 

7.00 

7.30 

8.00 

8.30 

9.00 

10.00 

10.00 

A.M. 

2.00 

P.M. 

4.00 

6.00 

8.00 

A.M. 

9.00 

Dose  in  c.c. 
0.05 
0.10 
0.15 
0.20 
0.25 
0.30 
0.35 
0.40 
0.45 
0.50 
0.55 
0.60 
0.65 
0.70 
0.75 
0.80 
0.85 
0.90 
0.95 


.00 

.05 

.10 

.15 

.20 

.25 

.30 

.35 

0 

0 

0 

0 

0 

0 

0 


Temperature. 
102.0° 
101.8° 
101.6° 
101.8° 
101.8° 
101.8° 
101.8° 
102,4° 
102.4° 
103.0° 
103.2° 
M)3.2° 
103.8° 
104.6° 
105.0° 
105.2° 
105.4° 
105.4° 
105.4° 
105.6° 
105.4° 
105.7° 
104.0° 
105.0° 
105.2° 
105.2° 
105.0° 
105.6° 
105.2° 
105.0° 
103.6° 
104.6° 
103.4° 
103.6° 


The  Destination  of  Egg-white  Introduced  into  the  Circulating 
Blood  of  the  Rabbit. — We  have  shown  (p.  357)  that  egg-white 
injected  into  the  ear  vein  of  a  rabbit  soon  disappears  from 
the  circulating  blood  and  diffuses  through  the  various 
tissues,  from  which  it  may  be  extracted  with  physiological 
salt  solution  and  its  presence  demonstrated  by  sensitizing 
guinea-pigs.  Since  reporting  on  this  we  have  found  that 
the  distribution  of  egg-white,  injected  into  the  blood, 
through  the  tissues  extends  not  only  to  the  organs  mentioned 


PROTEIN  FEVER  395 

in  the  article  referred  to,  but  also  to  the  skin  and  walls  of 
the  alimentary  canal. 

We  have  also  attempted  to  determine  how  long  after 
injection  egg-white  can  be  detected  in  the  tissues. 

Three  rabbits  received  intravenously  25  c.c.  of  egg-white 
dilution  (1  to  1).  One  was  killed  twenty-four  hours  later, 
and  sections  of  its  skin,  kidney,  brain,  liver,  spleen,  and 
intestinal  and  stomach  walls  rubbed  up  with  salt  solution. 
After  standing  in  the  cold  room  overnight  these  emulsions 
were  filtered  and  the  filtrates  injected  intra-abdominally 
into  guinea-pigs.  The  second  rabbit  was  killed  after  forty- 
eight  hours,  and  the  third  after  seventy-two  hours,  and  their 
tissues  treated  in  the  same  way.  All  the  guinea-pigs  that 
received  extracts  from  the  first  and  second  rabbits  were 
found  to  be  sensitized,  though  none  died.  Choking  symp- 
toms were  very  marked,  most  pronounced  in  those  that 
received  extracts  from  the  spleen  and  kidney.  The  symptoms 
were  quite  as  marked  in  those  that  received  the  extracts 
from  the  second  rabbit  as  in  those  treated  with  the  extracts 
from  the  first.  The  pigs  that  received  the  extracts  from  the 
third  rabbit  showed  absolutely  no  symptoms. 

From  these  experiments  we  conclude  that  egg-white 
diffused  through  the  tissues  after  injection  into  the  blood 
becomes,  sometime  between  two  and  three  days,  either 
so  far  changed  as  to  loose  its  identity  or  so  fixed  in  the 
tissue  that  it  cannot  be  washed  out  with  salt  solution.  This 
time  interval  probably  varies  with  the  kind  and  amount 
of  foreign  protein  introduced,  and  in  different  species  of 
animals. 

The  Digestive  Action  of  the  Blood  Serum  of  Rabbits  in  Which 
Fever  Has  Been  Induced  with  Egg-white. — The  following 
illustrates  some  of  our  experiments  on  the  digestive  action 
of  the  blood  serum:  The  temperature  of  two  rabbits  was 
raised  to  106°  by  hourly  intravenous  doses  of  a  dilution 
of  egg-white  (1  to  1).  One  hour  after  the  last  injection 
both  of  these  animals  were  bled  to  death  from  the  jugular 
vein  and  the  serum  obtained. 

Two  cubic  centimeters  of  this  fever  serum  without  any 


396  PROTEIN  POISONS 

addition,  after  standing  for  twenty-four  hours  in  the  incu- 
bator, was  diluted  to  10  c.c.  with  normal  salt  solution  and 
deprived  of  normal  proteins  by  acetic  acid  and  heat.  The 
filtrate  gave  a  slight  biuret  test  but  no  Millon.  At  11.15, 
2.5  c.c.  of  the  filtrate  was  injected  intracardiacally  into  a 
guinea-pig.  Temperature  before  the  injection  was  98.8°. 
At  11.30,  97.9°  and  at  11.40,  98.4°.  The  animal  was  not 
visibly  disturbed,  with  the  exception  of  slight  tremor. 

A  second  sample  of  2  c.c.  of  this  serum  which  had  been 
mixed  with  2  c.c.  of  milk  and  kept  in  the  incubator  for 
twenty-four  hours  was  treated  in  the  same  way.  The  biuret 
was  slight  and  the  Millon  negative.  The  guinea-pig  was 
not  disturbed  nor  the  temperature  lowered. 

A  third  portion  of  the  serum  mixed  with  an  equal  volume 
of  a  2  per  cent,  solution  of  Witte's  peptone  was  tested  in 
the  same  way.  The  filtrate  gave  a  beautiful  biuret,  but 
no  Millon.  The  temperature  of  the  pig  fell  2.6°  in  ten 
minutes,  but  otherwise  the  animal  was  not  affected. 

A  fourth  portion  of  the  serum  mixed  with  an  equal 
volume  of  a  dilution  of  egg-white  was  tested  in  the  same 
way.  The  filtrate  gave  a  splendid  biuret  and  also  a  good 
Millon.  The  pig  received  only  1.25  c.c.  of  the  filtrate,  half 
the  quantity  given  to  the  others,  but  it  immediately 
developed  the  symptoms  characteristic  of  the  protein 
poison  and  died  within  five  minutes.  Postmortem  examina- 
tion showed  no  injury  and  the  heart-apex  still  beating. 

We  took  a  mixture  of  2  c.c.  of  the  fever  serum  and  10  c.c. 
of  the  egg-white  dilution  (1  to  1),  which  had  stood  in  the 
incubator  for  five  days.  This  was  diluted  to  20  c.c.  and 
heated,  after  being  made  distinctly  acid  with  acetic  acid. 
After  the  removal  of  the  normal  blood  proteins  the  filtrate 
gave  both  the  biuret  and  the  Millon  tests  very  distinctly. 
Five  cubic  centimeters  of  the  filtrate  was  evaporated  on 
the  water-bath  and  the  yellowish  residue  extracted  with 
20  c.c.  of  absolute  alcohol.  The  portion  insoluble  in  alcohol 
was  extracted  with  5  c.c.  of  salt  solution.  The  part  soluble 
in  salt  solution  responded  feebly  to  both  the  biuret  and  the 
Millon  tests.  The  part  insoluble  in  both  alcohol  and  salt 


PROTEIN  FEVER  397 

solution,  when  suspended  in  water,  did  not  give  the  biuret, 
but  did  give  an  intense  Millon  reaction,  while  the  part 
soluble  in  alcohol  gave  neither.  A  duplication  of  this 
experiment  gave  identical  results. 

We  are  not  ready  to  conclude  from  our  work,  which  has 
been  more  extensive  than  detailed  here,  that  the  ferment 
of  the  fever  serum  is  strictly  specific.  An  exhaustive  meas- 
urement of  its  specificity  will  take  much  time  and  close 
work.  The  digestive  products  probably  vary  much,  in 
amount  at  least,  with  conditions,  and  a  close  study  of 
parenteral  digestion  offers  a  promising  field  for  research. 
We  are  inclined  to  think  that  the  rabbit  is  a  good  animal 
in  which  to  study  parenteral  digestion,  and  we  suspect 
that  this  form  of  digestion  is  not  altogether  abnormal  in 
this  animal.  We  have  already  shown  (p.  355)  that  egg-white 
introduced  into  the  stomach  or  rectum  of  a  rabbit  is,  in 
part  at  least,  absorbed  unchanged  into  the  blood.  Besides, 
we  have  observed  that  our  laboratory  rabbits  when  abun- 
dantly supplied  with  food  often  show  a  rectal  temperature 
of  103°  or  over,  and  when  the  food  supply  is  limited  the 
temperature  is  lower  and  more  constant. 

The  Production  of  Acute  Fever,  Followed  by  Immunity,  by 
Repeated  Intra-abdominal  Injections  of  Bacterial  Suspensions. 
In  this  group  of  experiments  guinea-pigs  have  been  used. 
The  usual  method  has  been  to  take  a  standard  loop  from 
an  agar  slant  four  days  old,  suspend  this  in  10  c.c.  of  normal 
salt  solution,  and  with  a  beginning  dose  of  0.1  c.c.  of  this 
suspension,  the  dose  is  increased  by  0.1  c.c.  each  time  and  is 
repeated  every  half  hour.  As  a  basis  for  these  experiments 
two  guinea-pigs  were  treated,  in  the  manner  described,  with 
the  salt  solution  alone.  The  result  in  one  of  these  is  shown 
in  Fig.  19.  The  total  range  in  this  case  covers  2.7°,  and  it 
is  possible  that  these  injections  stimulate  parenteral  diges- 
tion slightly.  The  general  agreement  of  the  curve  in  kind 
with  those  to  be  presented  later  is  quite  as  interesting, 
though  not  so  striking  as  its  difference  from  them  in  quantity. 
It  is.  worthy  of  note  that  during  the  continuance  of  these 
injections  the  temperature  did  not  fall  below  the  initial. 


398 


PROTEIN  POISONS 


Figs.  20,  21,  22,  and  23  are  illustrations  of  the  results 
obtained  by  this  mode  of  treating  guinea-pigs  with  sus- 


TIME 


s  . 


FIG.  19. — Showing  effect  of  repeated  injections  of  physiological  salt 
solution. 


"1:00 

2:00 

3:00 

4:00/ 

^5:00 

6:00 

7:00 

8:00 

9:00 

10:00 

/ 

^ 

\ 

/ 

\ 

/ 

\ 

/ 

\ 

/ 

\ 

/ 

—  \ 

\ 

. 

\ 

/ 

^ 

\ 

/ 

V 

\ 

\ 

/ 

\ 

/ 

\ 

\ 

/ 

\ 

/ 

FIQ.  20. — Showing  effect  of  repeated  injections  of  bacillus  subtilis 


FIG.  21. — Showing  effect  of  repeated  injections  of  bacillus  prodigiosus. 


PROTEIN  FEVER 


399 


pensions  of  living  bacteria.  All  the  animals  whose  tem- 
peratures are  shown  in  these  curves  recovered.  Additional 
information  concerning  these  experiments  is  given  in  tables 
XLII  to  XLV.  The  first  one  or  two  animals  of  each 
group  were  treated  with  a  beef-tea  culture  of  the  bacillus 


a  10:00 

11:00 

12:00 

Jl:00 

2:00 

3:00 

4:00 

5:oo 

6:00 

7:00 

/ 

\ 

/ 

\ 

/ 

\ 

"\ 

/ 

\ 

\ 

/ 

\ 

x 

t 

^ 

/ 

\ 

. 

^ 

\ 

/ 

\ 

1 

\ 

/ 

\ 

/ 

N 

1 

\ 

/ 

FIG.  22. — Showing  effect  of  repeated  injections  of  bacillus  cholerse. 


".11:00 

12:00 

».i  :oo 

2:00^ 

AJ:oo 

4:oo 

6:00 

6:00 

7:00 

8:00 

9:00 

10:00 

11:00 

/ 

\ 

/^ 

ss 

1 

s 

j 

\ 

7 

\ 

\ 

/ 

\ 

/ 

/\ 

/\ 

\ 

/ 

\, 

,/ 

\ 

7 

\ 

/ 

V^ 

/ 

\ 

/ 

/ 

\ 

/ 

\ 

s   / 

FIG.  23. — Showing  effect  of  repeated  injections  of  bacillus  typhosus. 

twenty-four  hours  old.  After  this,  we  changed  to  the 
suspension  in  salt  solution  and  the  exact  dilution  used  in 
each  animal  is  shown  in  the  tables.  The  average  bacterial 
content  of  each  loop  is  also  given  in  the  tables,  and  the 
first  dose  contained  one  one-hundredth  of  this  number, 
in  case  the  dilution  was  10  c.c. 


400 


PROTEIN  POISONS 


i 

-Q        T3  T3 

?'  u 

3    33 

Second  injection. 

• 

g     2  § 
.2     .2.S 
o      o  a 

.a      M  >> 
O      OH 

Q     QQ 
i-i     33 
S     SS 

<N        CO  CO 

«>  $  • 

*i 

<N        00  00 

Result  cf  series. 

Died  after  last  injection 
Recovered 

Died  12  hrs.  after  series 
Recovered 

"o.S 

.   O 

O   0) 

fc'3 

•»J<t-Tj<<Ol-Hl-ll-HO 

<Ni-H<N<Ni-Hi-i<N<N 

•a 
Rd 

OOCOOINOOlNt^ 

8"3 

«2~ 

OOCO'*'-<O5Tj<CT> 

1 

COCDOSOOOiOO 

2 

rt^H<N'<l<'-iO<NO 

•g- 

OOt^^TfiOINOO 

.S  a 

(JC,S-1 

OOO^HOOOO 

"3ft 

OOtOOO-<l<00-<lilN 

•3  a 

tS-s 

O3O5O5--IOOOO 
O5O3C35OOOOO 

cjd------ 

•ength  of 
pension. 

5  6- 

•S22S2222 

o 

•°.s 

-*j  — 

CO  g 

5  §  

T(t    ~" 
C^i—  1>—  li—  (i-Hi—  1   i-l   i-l 

6 
K 

iOOOO^H(N-*<C 
^H  ^  (N  CO  C<5  C«5  CO 

I-H   i-H   i-H   i-H 

1 

K 

•O  73 
II 

TO 

3 

S 
•43 

0)     33 

O    *i 

^    CO 

a"o 

0 

| 

HO 

T5 

a 
o 

qq 

(N  <N 

Days 
interval 

oo  oo 

)0  bacteru 
BTILIS 

1 
B 

a 
o 
•43 
o 

:l 

0  P 
uf  » 

1 

"8 

1 

o  ® 

»  5 

CO  j 

1 

•a  h  •& 

•  JK   w 

1  1 

O  T3    O  "     "     "     " 
u   5    o 

o  ,_; 

o  B 

O,  *"H. 
O  I-] 

2  M 

O  & 
•  "o 

O  V 

fc'fl 

1C  t^   CD   "^   CO   r*    i-l 

IN  rt  (N  <N  iM  (N  (N 

a  H 

TJ 

CD  O  CO  CD  t-  i-i  O5 

§  5? 
£  H 

§•2 

CO 

CO  CD  <N  IN  ^H  Tj<  Tj< 

CO 

i 

•«}<  t*  00  O  t^  00  CD 

H 

a 

CM    ^H   i-H   (N   O    rH    ^H 

2=3 

00  O  t~  •*  0  "5  00 

s- 

O  CN  O  O  O  O  O 

•3  ft 

a>  •*  oo  •*  CD  CM  •* 

11 

00  O5  O  O  00  CM  I-H 
O5  O5  O  O  O5  O  O 

I-H    ^H             1-H    I-l 

o3         c5  -     -    - 

"o  ej 

J3.2 

^            0  " 

a  § 

£§• 

asg 

j.     a:    :    :    : 

-=       o 

N   CM    i-H    i-H    i-H    i-l    i-l 

6 
% 

•*  CD  01  10  a>  >o  co 

i-H    i-H   CO   CO 

i—  I    i-H 

PROTEIN  FEVER 


401 


nj 


H 


if 


« 

«  a 


31 


e  <u 
g  a 


T3  T3-OT3T3 

<D  0>Q^OO 

>  >>>> 

3  3333 


Q  Q  Q      Q      Q  Q  Q  Q 
3     3333 


T3 

"o 
A 
a, 
>> 


C 


g  T3 
" 


W 


B 


-OOT3O 
m  O  m  O 
i,  O  •— 


•  SO)  .i,    O  •—    i) 

Q  tf          Q  pj  Q  tf 


26 


402 


PROTEIN  POISONS 


H 
1 

i 

o 


i 

w 


5 

x 

ia 
J 

cq 


1* 

§2 


2s 


p      pp      P      PP 


CO        CN  CN        CO        •*  »O 


pp 


CO  t»         CO  CO         »H         I-H  l> 


So 


- 


ser 


2       2: 


OS  OmO  SuSo 

•SI  :     B  >a  S     a;  •-.   o       -S   oj  -S   o 

Q      tfP      tfPtf      PtfPtf 


•iT    -g    '  jj 

CO        CN  <N    [S  £ 

CO            i-H    i-H     £  .  03 

T3        -O         O  "  -o 

flJ            ^3  ^        w  fl) 

(3    3"  «  P 


Oi  co  co  co  cf  ^  Tt*  ^  ^^  ^^  *~^  f^  i^  ^^ 


T3 

rt 

X  CO  00  IO  O 

CD  fH    TjH    T-H    1C   CO 

00        T}<  10  SN  ^H  O 

rH  iH          CNCOCOCNCO  ^ 

3 
o  o  ^  co  co 

O        O  O  1-1  i-H  CN 

00  t>  CN  CO  00  - 

OCOOO5'-iO3XOOOOOX^HO5O        O5        O  O  OS  -H  O 
Ok  9  Q  Oft  O  Ofr  Cv  Ov  C&  O  O)  O  O)  O        ~        OOO5OO 

•^     -^  1 

d 

•3 

*3 
p] 
OJ  C----,     ------  O          O.     .     .     , 

o  o  o  o  o  o  o 
.3  2 .2  -<  .9 

Q<;;;:::::;::'-|D<mO>:::: 
o  m  o  o  o 

o  £  o  -o  o 

—^  Q      ^H  ^-<  j- 

C^Cqr-li— I^H.— li— I^H^H^HrHr-H^H.— I'D— •  — •  — • 


i-Hi-iCNCNCN-*-*-*-*-*iO 


CO  >-H    CM   CO   Tfl   IO 


H 


PROTEIN  FEVER  403 

It  will  be  seen  that  in  the  large  majority  of  the  animals 
the  first  effect  is  a  slight  fall  in  temperature.  We  designate 
this  as  the  primary  or  short  fall.  It  should  be  understood 
that  we  estimate  the  falls  and  rises  from  the  initial  tempera- 
ture. The  short  fall  is  followed  by  the  primary  rise.  In  a 
few  instances  the  primary  fall  does  not  occur,  or  what  is 
more  probable,  has  not  been  detected.  In  one  animal 
(No.  136,  Table  XLII)  neither  primary  fall  nor  primary  rise 
was  detected.  The  rise  is  followed  by  the  secondary  or 
long  drop.  We  call  attention  to  some  of  the  great  drops; 
for  instance,  Nos.  131  and  134,  Table  XLII;  No.  140,  Table 
XLIV;  and  Nos.  46, 141,  and  142,  Table  XLV.  We  did  not 
suspect  that  this  treatment  would  give  immunity,  and  for 
this  reason  in  our  earlier  work  the  recovered  animals  were 
not  tested,  nor  have  we  as  yet  determined  the  limits  of  the 
immunity  secured  by  these  treatments.  As  is  shown  by  the 
tables,  the  immunity  is  not,  qualitatively  at  least,  specific. 
Animals  treated  with  subtilis  or  prodigiosus  bear,  at  least 
from  two  to  three  M.  L.  D.'s  of  cholera  or  typhoid  bacilli, 
and  the  immunity  secured  by  the  latter  is  interchangeable. 

When  treated  animals  are  inoculated  with  living  cultures 
they  become  sick  some  hours  before  the  controls,  and  we 
are  reminded  of  the  immunity  induced  some  years  ago 
in  this  laboratory  with  the  haptophor,  or  non-poisonous 
groups  of  the  colon  and  typhoid  bacilli,  and  reported  by 
V.  C.  Vaughan,  Jr.  (p.  144),  by  Vaughan  and  Wheeler 
(p.  157),  and  by  Vaughan.1 

The  Production  of  Fever  by  Repeated  Injections  of  Vegetable 
Proteins. — As  was  shown  in  this  laboratory  some  years 
ago,  the  vegetable  proteins  contain  the  same  poisonous 
group  found  in  bacterial  and  animal  proteins;  conse- 
quently there  seemed  no  reason  why  these  should  not 
induce  fever,  and  such  we  have  found  to  be  true.  We  will 
give  here  only  one  illustration.  We  extracted  10  grams  of 
oat-meal  with  100  c.c.  of  normal  salt  solution,  and  with  a 
beginning  dose  of  0.1  c.c.  of  the  slightly  opalescent  fluid 

1  Zeitsch.  f.  Immunitatsforschung,  1909,  i,  263. 


404  PROTEIN  POISONS 

thus  obtained  we  have  induced  fevers  similar  to  those 
already  described  in  this  chapter. 

We  have  made  many  experiments  on  the  production  of 
fever  with  non-protein  bodies,  giving  special  attention  to 
the  amino-acids,  the  xanthin  group,  inorganic  salts  of 
ammonia,  and  certain  carbohydrates,  but  a  report  upon 
these  findings  must  be  postponed. 

General  Conclusions. — Protein  fever,  and  this  includes 
the  great  majority  of  clinical  fevers,  results  from  the  paren- 
teral  digestion  of  proteins.  Bouillaud1  was  practically 
right  when  he  said:  "La  fievre  est  une  maladie,  dont  la 
nature  est  toujours  la  meme."  Proteins,  living  and  dead, 
occasionally  find  their  way  into  the  body.  They  may  come 
from  without  or  from  within.  Crushed  bone,  muscle,  or 
other  tissue,  on  being  deprived  of  its  vitality  or  detached 
from  its  normal  surroundings,  becomes  foreign  material 
and  must  be  broken  up  preparatory  to  its  elimination. 
Under  certain  conditions  proteins  taken  into  the  alimentary 
canal  escape  enteral  digestion  and  are  in  part  absorbed 
unbroken.  When  this  happens,  they  are  disposed  of  by 
parenteral  digestion.  In  a  finely  divided  form,  as  in  the 
pollen  of  plants,  proteins  are  absorbed  from  the  respiratory 
tract  and  give  rise  to  the  condition  designated  as  hay  or 
rose  fever.  But  in  the  great  majority  of  instances  proteins 
gain  entrance  to  the  body  in  unbroken  form,  as  living 
proteins,  bacteria,  or  protozoa.  The  parenteral  proteo- 
lytic  ferments  are  of  two  kinds,  non-specific  and  specific. 
The  former  are  normally  present  in  the  blood  and  tissues, 
especially  in  the  former,  of  all  animals.  They  differ  in 
kind  in  different  species  and  in  amount  and  efficiency  in 
individuals.  Their  purpose  is  to  break  up  foreign  proteins 
that  find  their  way  into  the  blood  and  tissues.  They  are, 
within  limits,  general  proteolytic  ferments,  as  are  those 
of  the  alimentary  canal;  though  the  variety  of  proteins 
upon  which  they  can  act  is  more  limited.  They  constitute 
the  most  important  factor  in  racial  and  individual  immunity. 

1  Trait6  clinique  et  experimentale  des  Fifevries,  1826. 


PROTEIN  FEVER  405 

Man  is  immune  to  most  bacteria,  not  because  they  do  not 
elaborate  poisons,  for  every  protein  molecule  contains  its 
poisonous  group,  but  because  they  are  destroyed  by  the 
general  proteolytic  enzymes  as  soon  as  they  enter  the 
tissue  and  consequently  are  not  permitted  to  multiply 
in  man's  body.  These  non-specific,  parenteral  proteolytic 
enzymes  are  probably  secretions  of  certain  specialized 
cells.  Under  natural  conditions  these  enzymes  are  capable 
of  digesting  those  proteins  upon  which  they  do  act  only  in 
small  amounts,  but  the  cells  which  elaborate  them  may 
be  stimulated  to  increased  activity  by  proper  treatment, 
and  the  method  detailed  in  this  paper  seems  to  accomplish 
this  purpose.  Whether  or  not  these  enzymes  become 
qualitatively  specific  under  such  treatment  as  we  have 
detailed  can  be  determined  only  by  further  study.  The 
immunity  secured  by  these  enzymes  is  limited  in  extent 
and  transitory  in  duration. 

The  specific,  parenteral  proteolytic  ferments  are  not 
normal  products  of  the  body  cells,  but  are  brought  into 
existence  under  the  stimulation  of  those  proteins,  intro- 
duced into  the  blood  and  tissues,  which  on  account  of  their 
nature  or  amount  escape  the  action  of  the  non-specific 
ferments.  It  is  to  the  development  of  these  ferments  that 
the  phenomena  of  sensitization  (wrongly  called  anaphyl- 
axis)  are  due.  A  protein  introduced  into  the  blood  and 
not  promptly  and  fully  digested  by  the  non-specific  enzymes 
is  discharged  from  the  blood  current  and  deposited  in  some 
tissue,  the  cells  of  which  after  a  time  develop  a  specific 
ferment  which  splits  up  this  protein  and  is  not  capable  of 
acting  upon  any  other.  For  certain  proteins  there  are 
certain  predilection  organs  and  tissues  in  which  they  are 
stored,  either  exclusively  or  most  abundantly:  the  pneu- 
mococcus  in  the  lungs;  the  typhoid  bacillus  in  the  mesen- 
teric  and  other  glands;  the  viruses  of  the  exanthematous 
diseases  in  the  skin,  etc.  For  the  development  of  the  specific 
proteolytic  ferments  time  is  required,  and  this  varies  with 
the  protein  and  probably  with  the  tissue  in  which  it  is 
deposited.  The  development  of  these  ferments  necessitates 


406  PROTEIN  POISONS 

changes  in  the  chemical  constitution  of  the  protein  mole- 
cules of  the  cell,  and  by  this  means  the  cell  acquires  a  new 
function,  which  subsequently  is  brought  into  operation 
only  by  contact  with  that  protein  to  which  its  existence  is 
due.  As  a  result  of  this  rearrangement  in  molecular  struc- 
ture, the  .cell  stores  up  a  specific  zymogen  which  is  activated 
by  contact  with  its  specific  protein.  This  explanation  of 
the  phenomena  of  sensitization  originated  in  this  laboratory,1 
and  was  not  simply  a  fortunate  guess,  as  has  been  assumed 
by  some.  The  same  is  true  of  the  statement  made  at  the 
same  time,  that  protein  sensitization  and  bactericidal 
immunity  are  identical,  and  not  antipodal,  as  they  may 
appear  to  the  superficial  observer.  A  close  study  of  the 
split  products  of  bacterial,  vegetable,  and  animal  proteins, 
and  especially  of  the  poisonous  group  found  in  all  proteins, 
had  already  been  made  in  this  laboratory.  A  study  of  the 
symptoms  induced  by  the  protein  poison  and  of  those 
following  a  second  administration  to  the  sensitized  animal 
was  certainly  good  and  sufficient  ground  for  concluding,  as 
then  stated,  that  the  man  that  dies  from  the  administra- 
tion of  morphine  and  the  one  that  dies  from  opium  both 
owe  their  death  to  the  same  poison.  The  most  valuable 
experiments  of  Friedberger  and  his  assistants,  and  of 
Pfeiffer  and  Mita,  have,  in  our  opinion,  fairly  established 
the  validity  of  this  explanation.2 

Whether  the  products  of  digestion  with  the  non-specific 
ferments  and  those  elaborated  by  the  specific  enzymes  are 
identical  or  not  remains  to  be  ascertained.  The  presence 
of  a  poisonous  group  in  the  protein  molecule  is  disclosed  in 
both  enteral  and  parenteral  digestion,  as  well  as  by  our 
process  of  splitting  up  the  protein  with  dilute  alkali  in 
absolute  alcohol.  In  the  first  case  it  appears  in  the  peptone 

1  Jour,  of  Infect.  Dis.,  June,  1907. 

*  In  1910  Friedberger  (Berl.  klin.  Woch.,  Nos.  32  and  42)  made  very 
plain  the  relation  between  sensitization  and  the  infectious  diseases,  and 
in  his  address  at  the  meeting  of  the  German  naturalists  at  Konigsberg  in 
September,  1910  (Munch,  med.  Woch.,  1910,  Nos.  50  and  51),  he  dwelt 
most  instructively  upon  the  wide  application  of  the  facts  learned  in  his 
studies  of  sensitization. 


PROTEIN  FEVER  407 

molecule,  which  is  large  and  complex.  By  the  chemical 
process  it  is  obtained  as  a  less  complex,  more  diffusible,  and 
consequently  more  active  body.  Between  the  two  there 
are  probably  several  intermediate  substances.  The  prompt- 
ness in  action  manifested  by  all  proteolytic  ferments  is 
determined,  in  part  at  least,  by  the  proportion  between  the 
surface  of  the  substrate  and  the  mass.  We  observed  some 
years  ago1  that  the  more  finely  divided  the  cellular  sub- 
stance of  bacteria  is,  the  smaller  the  dose  which  proves 
fatal.  This  is  due  to  the  greater  surface  exposure,  and  the 
same  apparently  holds  good  for  colloids  in  solution.  When 
soluble  proteins  are  expelled  from  the  blood  and  diffused 
throughout  the  animal  body,  the  conditions  for  their  rapid 
cleavage  are  most  favorable,  and  consequently  the  fulmina- 
ting phenomena  observed  after  the  second  injection  into 
a  sensitized  animal. 

When  a  protein  deposited  in  mass  is  rapidly  acted  upon 
by  the  parenteral  enzymes,  more  or  less  marked  inflamma- 
tion results.  This  may  be  demonstrated  by  injecting  sus- 
pended, dead,  bacterial  cellular  substance  into  the  peritoneal 
cavity  of  a  guinea-pig  when  a  diffuse  peritonitis  results, 
and  we  wish  to  suggest  that  the  exanthems  are  due  to  the 
rapid  digestion  of  proteins  deposited  in  the  skin.  We  admit 
that  this  is  largely  theoretical,  but  we  have  found,  as  already 
stated,  that  egg-white  is  in  part  deposited  in  the  skin  of 
rabbits  after  intravenous  injection.  This  may  be  an 
explanation  of  the  Arthus  phenomenon. 

The  fact  that  every  protein  molecule  contains  a  poisonous 
group  does  not  mean  that  the  products  of  protein  digestion 
must  contain  a  poison,  for  the  poison  itself  may  be  split  up 
and  rendered  inert,  as  happens  when  the  proteins  in  the 
alimentary  canal  are  broken  up  into  amino-acids.  It  may 
therefore  happen  that  in  certain  forms  or  stages  of  parenteral 
digestion  no  poison  is  formed. 

The  low  temperature  seen  in  some  of  our  charts 
undoubtedly  indicates  the  liberation  of  the  poisonous 

1  Trans.  Assoc.  Amer.  Phys.,  1902. 


408  PROTEIN  POISONS 

group,  and  consequently  the  subnormal  as  well  as  the  high 
temperature  is  a  result  of  parenteral  digestion,  and  it  is 
in  this  stage  that  the  greater  danger  to  the  life  of  the  animal 
lies,  as  is  plainly  shown  in  our  results.  However,  there  is 
danger  to  life  in  the  high  temperature  in  and  of  itself.  A 
rabbit  is  not  likely  to  survive  a  temperature  above  107°, 
and  this  was  reached  in  at  least  one  of  our  experiments,  and 
closely  approached  in  many  others. 

Fever  must  be  regarded  as  a  conservative  process,  although 
like  many  of  nature's  processes  it  often  leads  to  disaster. 
But  its  purpose  is  the  disposal  of  foreign  and  dangerous 
material,  and  therefore  must  be  regarded  as  beneficent. 

In  parenteral  digestion  the  following  sources  of  heat 
production  must  be  evident:  (1)  The  unaccustomed 
stimulation  and  consequent  increased  activity  of  the  cells 
which  supply  the  enzymes  must  be  the  source  of  no  incon- 
siderable increase  in  heat  production.  (2)  The  cleavage  of 
the  foreign  protein  means  the  liberation  of  heat.  (3)  The 
reaction  between  the  products  of  the  digestion  and  the 
tissues,  especially  when  an  active  and  irritant  poison  is 
liberated,  must  lead  to  increased  heat  production.  We 
regard  the  first  and  last  of  these  as  the  more  important 
sources  of  the  overproduction  of  heat  in  the  febrile  state. 

Special  Conclusions. — 1.  Large  doses  of  unbroken  protein 
administered  intra-abdominally,  subcutaneously,  or  intra- 
venously have  no  effect  upon  the  temperature;  at  least, 
do  not  cause  fever. 

2.  Small  doses,   especially  when  repeated,  cause  fever, 
the  forms  of  which  may  be  varied  at  will  by  changing  the 
size  and  the  interval  of  dosage. 

3.  The  effect  of  protein  injections  on  the  temperature  is 
more  prompt  and  marked  in  sensitized  than  in  fresh  animals. 

4.  The  intravenous  injection  of  laked  blood  corpuscles 
from  either  man  or  the  rabbit  causes  in  the  latter  even  in 
very  small  quantity,   either  in   single  or  repeated  doses, 
prompt  and  marked  elevation  of  temperature. 

5.  Laked   corpuscles   after   removal    of   the   stroma   by 
filtration  have  a  like  effect. 


PROTEIN  FEVER  409 

6.  Protein  fever  can  be  continued  for  weeks  by  repeated 
injections,  giving  a  curve  which  cannot  be  distinguished 
from  that  of  typhoid  fever. 

7.  Protein  fever  is  accompanied  by  increased  nitrogen 
elimination  and  gradual  wasting. 

8.  Protein  fever  covers  practically  all  cases  of  clinical 
fever. 

9.  Animals  killed  by  experimentally  induced  fever  may 
die  at  the  height  of  the  fever,  but,  as  a  rule,  the  temperature 
rapidly  falls  before  death. 

10.  Fever  induced   by   repeated   injections   of  bacterial 
proteins  and  ending  in  recovery  is  followed  by  immunity. 

11.  The  serum  of  animals  in  which  protein  fever  has 
been  induced  digests  the  homologous  protein  in  vitro. 

12.  Fever  results  from  the  parenteral  digestion  of  proteins. 

13.  There  are  two  kinds  of  parenteral  proteolytic  enzymes, 
one  specific  and  the  other  non-specific. 

14.  The  production  of  the  non-specific  ferment  is  easily 
and  quickly  stimulated. 

15.  The  development  of  the  specific  ferment  requires  a 
longer  time. 

16.  Sensitization  and  lytic  immunity  are  different  mani- 
festations of  the  same  process. 

17.  Foreign  proteins,  living  or  dead,  formed  or  in  solu- 
tion, when  introduced  into  the  blood  soon  diffuse  through 
the  tissues  and  sensitize  the  cells.    Different  proteins  have 
predilection  places  in  which  they  are  deposited  and  where  they 
are,  in  large  part  at  least,  digested,  thus  giving  rise  to  the 
characteristic  symptoms  and  lesions  of  the  different  diseases. 

18.  The   subnormal   temperature   which   may   occur   in 
the  course  of  a  fever  or  at  its  termination  is  due  to  the 
rapid  liberation  of  the  protein  poison,  which  in  small  doses 
causes  an  elevation,  and  in  larger  doses  a  depression  of 
temperature. 

19.  Fever  per  se  must  be  regarded  as  a  beneficient  phe- 
nomenon, inasmuch  as  it  results  from  a  process  inaugurated 
by  the  body  cells  for  the  purpose  of  ridding  the  body  of 
foreign  substances. 


410  PROTEIN  POISONS 

20.  The  evident  sources  of  excessive  heat  production  in 
fever  are  the  following:  (a)  That  arising  from  the  unusual 
activity  of  the  cells  supplying  the  enzyme;  (6)  that  arising 
from  the  cleavage  of  the  foreign  protein;  (c)  that  arising 
from  the  destructive  reaction  between  the  split  products, 
from  the  foreign  protein  and  the  proteins  of  the  body. 

In  1910  Friedberger1  studied  the  effects  of  graduated 
doses  of  foreign  proteins  on  the  temperature  of  both  normal 
and  sensitized  animals.  With  lambs'  serum  intravenously 
administered  to  normal  guinea-pigs  he  obtained  the  following 
results : 

5.0      c.c.  equal  fatal  dose. 

0.5      c.c.  equals  limit  for  fall  in  temperature. 

0.01    c.c.  equals  upper  constant. 

0.005  c.c.  equals  fever  plane. 

0.001  c.c.  equals  lower  constant. 

In  sensitized  guinea-pigs  the  above  figures  were  changed 
to  the  following: 

0.005        c.c.  equals  fatal  dose. 
0.0005      c.c.  equals  limit  for  fall. 
0.00001    c.c.  equals  upper  constant. 
0.000005  c.c.  equals  limit  for  fever. 
0.000001  c.c.  equals  lower  constant. 

In  1911  Schittenhelm,  Weichardt,  and  Hartmann2 
experimented  upon  the  effect  of  the  parenteral  administra- 
tion of  diverse  proteins  on  animal  temperature  and  came 
to  the  following  conclusion,  which  in  our  opinion  is  well 
stated :  "  In  severe  experimental  anaphylaxis  there  is  a 
fall  in  temperature;  in  the  lighter  manifestations  there  is 
fever."  We  regard  this  as  a  confirmation  of  our  conclusion 
reached  some  years  earlier.  "Small,  especially  repeated, 
doses  of  the  protein  poison  cause  fever,  while  large  doses 
depress  the  temperature." 

Some  years  ago  Friedmann  and  Isaak3  showed  that  after 
the  parenteral  introduction  of  foreign  proteins  the  increase 
in  nitrogen  elimination  is  greater  than  can  be  accounted 

1  Berl.  klin.  Woch.,  1910,  No.  42. 

2  Zeitsch.  f.  exp.  Path.  u.  Ther.,  1911.  3  Ibid.,  1905,  1906,  and  1908. 


PROTEIN  FEVER  411 

for  by  the  protein  injected.  This  has  been  confirmed  by 
the  work  of  Schittenhelm  and  Weichardt,1  and  as  has  been 
stated,  we  found  the  same  in  protein  fever.  Our  explanation 
for  the  marked  increase  in  nitrogen  elimination  has  been  given. 

In  intermittent  and  remittent  fevers  and  in  relapses  in 
all  infectious  diseases  the  phenomena  of  protein  sensitiza- 
tion  are  fully  demonstrated.  In  the  different  forms  of 
malaria,  chill  and  fever  correspond  to  the  discharge  of 
foreign  protein  into  the  blood,  just  as  promptly  as  anaphyl- 
actic  symptoms  follow  the  injection  of  the  homologous 
protein  in  a  sensitized  animal.  The  moment  the  blood 
cells  rupture  and  the  protozoal  protein  is  disseminated 
the  sensitized  cells  discharge  the  lytic  ferment  by  which 
the  foreign  protein  is  disrupted  and  destroyed,  but  in  this 
process  the  poison  is  liberated. 

Local  sensitization  is  frequently  established  in  the  mucous 
membrane  of  the  air  passages  and  of  the  alimentary  canal, 
also  in  the  skin  for  two  reasons.  In  the  first  place,  foreign 
proteins  are  frequently  brought  into  direct  contact  with 
these  tissues,  and  in  the  second  place,  foreign  proteins  intro- 
duced into  the  blood  are  frequently  deposited  in  the  skin 
and  in  the  walls  of  the  alimentary  canal.  These  local  sensi- 
tizations  characterize  many  of  the  infectious  diseases.  The 
work  of  Dunbar  and  Weichardt  on  hay  fever  is  a  good 
illustration.  These  investigators  injected  each  other  sub- 
cutaneously  with  minute  quantities  of  pollen  suspension. 
Immediately  Dunbar,  being  a  hay-fever  subject,  became 
dizzy,  and  within  a  few  minutes  began  to  sneeze,  then  a 
whooping-like  cough  began.  The  eyes  were  congested,  and  an 
abundant  secretion  flowed  from  the  nose.  The  face  became 
swollen  and  cyanotic,  and  soon  the  body  was  covered 
with  an  urticarial  rash.  After  twenty-four  hours  these 
symptoms  subsided.  Weichardt,  not  being  a  hay-fever 
subject,  was  not  affected.  That  this  and  kindred  affections 
are  not  benefited  by  antisera  was  abundantly  and  positively 
demonstrated  by  the  failure  of  the  so-called  hay-fever 

1  Zentralbl.  f.  d.  ges    Physiol.  u.  Pathol.  d.  Stoffwechsel,  1910. 


412  PROTEIN  POISONS 

serum,  which  was  found  in  no  instance  to  be  of  special 
value,  and  in  some  it  greatly  intensified  the  symptoms. 

Our  common  colds  are  instances  of  local  sensitization. 
Schittenhelm  and  Weichardt  tell  of  a  man  who  was  so 
deeply  sensitized  by  the  inhalation  of  Witte's  peptone  that 
he  could  tell  on  entering  the  laboratory  whether  the  peptone 
flask  was  open  or  closed,  and  some  moist  peptone  painted 
on  the  skin  caused  the  area  covered  to  become  red.  The 
high  degree  of  susceptibility  to  odors  from  the  horse  shown 
by  some  people  has  already  been  referred  to.  It  seems  in 
some  instances  that  this  susceptibility  is  transmitted  from 
mother  to  child. 

A  volume  might  be  filled  with  citations  of  cases  of  food 
and  medicine  idiosyncrasies.  That  these  are,  in  large  part 
at  least,  instances  of  protein  sensitization  has  been  demon- 
strated by  rendering  animals  susceptible  to  the  same  food 
or  medicine  by  injecting  them  with  the  serum  of  the  sus- 
ceptible individual.  In  other  words,  passive  anaphylaxis 
has  been  established  in  the  animal.  In  this  way  Briick  has 
sensitized  animals  to  iodoform  and  antipyrin  with  the  sera 
of  persons  especially  susceptible  to  these  agents. 

Thiele  and  Embleton1  have  made  an  extensive  study 
of  temperature  variations.  They  employed  well-fed  guinea- 
pigs.  First,  they  tried  non-protein  bodies,  and  with  these 
reached  the  following  conclusions: 

1.  Sodium   chloride    varies   in    its    effects    according  to 
its    degree   of   concentration   when   injected    into  feeding 
animals : 

As  normal  saline  it  produces  only  a  slight  rise. 

As  2  to  2.5  per  cent,  saline  it  produces  a  marked  rise. 

As  3  per  cent,  saline  it  produces  a  fall. 

As  5  per  cent,  saline  it  produces  a  rise. 

2.  Calcium  salts  intraperitoneally  produce  a  fall. 

3.  Ringer's  fluid  has  no  effect  on  temperature. 

4.  Alkalies,    very  dilute,  produce  a  rise,   and  stronger, 
produce  a  fall. 

1  Zeitsch.  f.  Immunitatsforschung,  1913,  xvi,  178. 


PROTEIN  FEVER  413 

5.  Acid,  same  effects. 

6.  Lecithin  injected  intraperitoneally  as  a  water  emulsion 
has  no  effect  on  the  temperature. 

7.  Charcoal  (animal)  very    fine    powder;    in  suspension 
intraperitoneally  in  large  doses  produces  a  marked  fall,  and 
in  small  doses  a  slight  rise. 

They  notice,  confirmatory  of  the  findings  of  Krehl  and 
Matthes,  that  fasting  animals  do  not  easily  get  febrile 
reactions  with  these  substances.  This  agrees  with  Hirsch 
and  Roily,  who  state  that  glycogen-free  animals  cannot 
become  febrile  with  sodium  chloride  injections.  Kulz 
showed  that  injections  of  sodium  chloride  cause  glycosuria. 
This  was  confirmed  by  Fischer,  who  believed  the  glycosuria 
to  be  due  to  irritation  of  the  central  nervous  system,  and 
showed  that  it  is  abolished  by  cutting  the  splanchnics,  and 
that  marked  glycosuria  results  when  direct  injection  into 
the  cerebral  vessels  is  made.  Calcium  salts  inhibit  both  the 
glycosuria  and  the  fever  induced  by  sodium  chloride.  Freund 
showed  that  adrenalin  causes  glycosuria,  and  in  large  doses 
a  fall,  and  in  small  doses,  a  rise  in  temperature.  Again, 
calcium  salts  inhibit  both  the  glycosuria  and  pyrexia  due 
to  adrenalin.  The  conclusion  is  reached  that  the  fever 
caused  by  these  non-protein  substances  is  due  to  sympathetic 
irritation  and  consequent  increased  glycogen  metabolism. 

Turning  now  to  the  work  done  by  Thiele  and  Embleton 
with  protein  bodies,  we  find  a  complete  confirmation  of  the 
results  obtained  by  previous  investigators.  They  say: 
"With  regard  to  endotoxin  proper,  by  which  we  mean  a 
toxic  substance,  which  is  specific  to  the  bacteria  or  protein, 
there  is  but  little  evidence.  The  virulent  and  non-virulent 
bacteria,  as  well  as  simple  egg  albumen,  appear  to  have 
the  same  temperature  effects  when  inoculated  into  healthy 
animals  in  the  way  mentioned  above,  and  also  performed 
by  Vaughan  and  Wheeler.  The  so-called  endotoxin  has 
been  liberated  from  protein,  from  bacteria,  etc.,  by  Vaughan 
and  Wheeler  and  others,  also  by  Schittenhelm  and  Weichardt. 
In  all  these  experiments  the  protein,  whether  simple  or 
bacterial,  has  been  treated  for  some  time  with  caustic 


414  PROTEIN  POISONS 

alkali  either  in  watery  solution  or  in  alcohol,  so  that  a 
certain  amount  of  protein  degradation  has  occurred  which 
is  apparently  just  sufficient  to  form  the  toxic  bodies  which 
cause  the  acute  toxic  symptoms  or  temperature  variations 
under  discussion,  according  to  the  size  of  the  dose.  The 
work  of  Friedberger,  Abderhalden,  and  others  with  ana- 
phylatoxin  production  in  vitro,  and  the  demonstration  of 
the  presence  of  proteolytic  degradation  bodies  during  the 
process  of  the  formation  of  toxic  substance,  is  in  favor  of 
the  view  that  the  toxin  (poison)  is  purely  a  degradation 
product  of  the  protein  simple  or  bacterial,  as  again  there 
is  no  specificity  in  the  toxins  (poisons)  produced  from  the 
various  antigens  (sensitizers)  used.  The  reason  why  bacteria 
have  a  much  more  potent  action  in  such  small  quantities 
would  appear  to  be  in  the  chemical  composition  of  their 
bodies,  and  in  the  presence  of  normal  specific  ferments  to 
them.  Thus,  according  to  Schittenhelm  and  Weichardt, 
and  our  own  observations  from  the  protamine  injections, 
etc.,  it  appears  that  some  proteins  of  the  normal  animal 
body  are  much  more  toxic  than  others,  and  the  toxic  ones 
are  those  which  have  a  high  percentage  of  the  diamino 
bases.  In  the  experiments  brought  forward  here  the  toxicity 
of  protamine  as  regards  producing  temperature  variations 
is  almost  as  great  as  that  of  the  tubercle  bacillus.  This 
is  important  in  view  of  Ruppel's  work  showing  that  the 
tubercle  bacillus  has  a  large  amount  of  diamino  bases. 
Further,  the  formation  of  toxic  substances  in  the  case  of 
bacteria  suspended  in  normal  saline  would  appear  to  be 
due  to  the  formation  of  cleavage  bodies  from  autolytic 
changes,  just  as  occurs  when  tissues  undergo  autolysis.  A 
final  argument  that  cleavage  products  of  protein  and  bac- 
teria are  the  causation  of  temperature  reactions,  etc.,  is 
the  observation  of  Matthes,  who  showed  that  in  a  digesting 
tuberculous  animal,  albumose  injection  gave  rise  to  hyperemia 
of  the  small  intestine  and  around  the  tuberculous  foci  just 
as  tuberculin  injection  does,  as  we  have  noted  in  our 
present  experiments.  Here  we  have  a  protein  cleavage 
product  giving  rise  to  the  same  effects  as  the  specific  antigen. 
Hence  it  would  appear  that  the  cleavage  products  of  the 


PROTEIN  FEVER  415 

antigen  are  the  cause  of  the  reaction,  cleavage  going  on 
continually  locally  at  the  tuberculous  foci,  and  the  addition 
of  a  little  more  cleavage  body  producing  a  cumulative 
effect,  and  in  the  intestine  where  cleavage  is  also  going 
on  in  the  cells  during  digestion,  the  same  occurs  whether 
the  reacting  dose  is  from  simple  or  bacterial  protein." 

These  investigators  agree  with  us  that  it  is  the  same 
substance  which  in  larger  doses  causes  a  fall  in  temperature, 
and  in  small  doses  a  rise.  They  say:  "(1)  In  sensitized 
animals,  owing  to  the  presence  of  a  specific  enzyme,  the 
homologous  antigen  undergoes  more  rapid  degradation 
than  in  non-sensitized  ones,  and  consequently  certain 
degradation  products  are  liberated  in  sufficient  quantities 
from  relatively  small  amounts  of  the  antigen  to  cause 
temperature  depression  in  these  animals,  and  from  still 
further  amounts  to  cause  fever.  (2)  The  pyrogenic  bodies 
are  not  a  further  stage  in  the  degradation  of  the  antigen,  but 
the  same  degradation  body  or  bodies  cause  depression  or  ele- 
vation according  to  the  amounts  present  at  any  given  time." 

The  relative  effects  of  egg-white  and  tubercle  protein  on 
fresh  and  sensitized  animals  are  shown  by  Thiele  and 
Embleton  as  follows: 

EGG-WHITE 

Normal  animal,  Sensitized, 

Limits  of  grams.  grams. 

Temperature  fall    ....     0.05  0.005 

Constant  temperature       .      .     0 . 02  0 . 0002  to  0 . 0001 

Temperature  rise   .      .      .      .     0.01  to  0.001  0.0001  to  0.000002 

TUBERCLE  EMULSION 

Normal  animal,  Sensitized, 

Limits  of  grams.  grams. 

Temperature  fall    .      .      .      .     0.005  to  0.002  0.0005 

Constant  temperature       .      .     0.002  to  0.001  0.0001 

Temperature  rise  ....     0.001  to  0.00001  0.00001  to  0.000001 

In  another  interesting  way  these  investigators  confirm 
some  of  our  earliest  work  (Chapter  III)  when  they  say: 
"The  more  finely  divided  the  bacterial  protoplasm  is,  the 
more  rapid  are  the  temperature  effects  and  the  more  toxic 
is  the  bacillary  substance." 


CHAPTER  XIV 
SPECIFIC  FERMENTS  OF  THE  CANCER  CELL1 

THE  chief  distinction  between  a  living  protein  and  any 
more  or  less  complex  chemical  group  or  dead  cell  is  that 
the  living  protein  contains  within  itself  different  ferments 
or  enzymes.  Many  of  these  ferments  are  common  to  several 
different  forms  of  living  protein  or  at  least  bear  a  close 
resemblance  to  each  other,  but  it  is  also  probable  that  each 
and  every  form  of  living  cell  contains  within  itself  a  specific 
enzyme  which  is  distinctive  for  that  given  form  of  protein, 
and  is  present  in  no  other. 

While  as  yet  ferments  have  not  been  isolated  in  a  chemic- 
ally pure  state,  and  we  have  not  an  exact  knowledge  of 
their  chemical  nature,  yet  they  are  commonly  regarded  as 
albuminous,  and  their  functions  are  specific  and  fairly 
well  known.  This  at  least  is  true  of  the  soluble  ferments 
such  as  the  amylases,  proteoses,  and  lipases.  Again,  there 
are  examples  of  ferments  of  similar  or  nearly  like  consti- 
tution which  are  formed  by  more  than  one  variety  of  cell 
(thus  ptyalin  formed  by  the  salivary  glands  and  diastase 
formed  by  the  pancreas  are  both  starch-splitting  ferments). 

The  above  examples  are  all  of  ferments  which  are  water 
soluble,  and  which  are  active  outside  of  the  cell  that  pro- 
duces them.  They  are  ferments  which  are  excreted  by  the 
cell  and  are  used  by  the  more  highly  developed  forms  of 
animal  life  for  the  purpose  of  converting  complex  chemical 
bodies  into  more  simple  forms,  so  that  they  may  be  of 
service  to  the  whole,  through  their  reaction  with  other 
ferments  contained  in  other  body  cells. 

1  This  chapter  is  contributed  by  J.  Walter  Vaughan  and  the  work  has 
been  done  by  the  aid  of  the  Chase  Cancer  Fund  in  the  research  laboratory 
of  Harper  Hospital,  Detroit,  Michigan 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     417 

It  is  well  known,  however,  that  all  unicellular  forms  of  life 
contain  ferments  which  it  is  safe  to  divide  into  two  classes. 
These  we  would  designate  as  soluble  and  insoluble,  or  extra- 
and  intracellular  ferments;  a  soluble  ferment  being  one 
that  is  excreted  by  the  cell  and  which  performs  its  ferment 
action  without  the  cell,  and  an  insoluble  ferment  being  an 
enzyme  that  acts  only  within  the  cell  structure.  Obviously, 
with  our  rather  limited  comprehension  of  the  nature  of 
ferments,  our  knowledge  is  confined  chiefly  to  the  soluble 
variety.  In  this  class  of  ferments  formed  by  unicellular 
organisms  might  be  mentioned  zymase  formed  by  the 
yeast  cell,  and  the  putrefactive  ferments  formed  by  many 
bacteria.  V.  C.  Vaughan  would  also  classify  such  sub- 
stances as  diphtheria  toxin  and  tetanus  toxin  in  this  group, 
since  it  is  his  belief  that  these  substances  are  not  poisons 
in  themselves,  but  liberate  poisons  through  their  ferment 
action. 

While  the  known  functions  of  most  enzymes  are  those  of 
decomposition,  either  through  such  processes  as  hydrolyza- 
tion  or  oxidation,  yet  we  also  know  that  ferment  action 
may  be  one  of  construction;  and  both  processes  may  be 
carried  on  by  one  and  the  same  ferment.  Thus  we  know  that 
ethyl  butyrate  may  be  converted  by  the  action  of  lipase 
into  alcohol  and  butyric  acid,  while  with  a  change  in  the 
acting  masses,  alcohol  and  butyric  acid  plus  lipase  will 
form  the  ester. 

Inasmuch  as  we  do  know  of  the  presence  of  constructive 
ferments  belonging  to  the  soluble  variety,  in  which  class 
we  would  naturally  suppose  the  greater  number  to  be 
destructive,  since  their  function  is  to  reduce  complex 
proteins  into  simpler  forms  for  the  easier  assimilation  by 
the  cell,  it  seems  reasonable  to  suppose  that  most  intra- 
cellular or  insoluble  ferments  are  of  the  constructive  variety, 
and  that  it  is  through  their  aid  that  simple  chemical  sub- 
stances are  converted  into  the  complex  distinctive  proteins 
of  the  cell  itself. 

While  much  of  the  above  is  theoretical,  yet  there  is 
sufficient  knowledge  concerning  some  of  the  facts  to  make 
27 


418  PROTEIN  POISONS 

the  theoretical  portion  logical,  and  we  have  previously 
advanced  the  following  theory  as  to  the  etiology  of  malig- 
nant disease  with  the  foregoing  in  view. 

Every  living  cell  has  within  itself  a  constructive  ferment 
whose  specific  action  is  to  construct  proteins  of  the  same 
specific  composition  as  the  cell  itself.  Through  the  aid  of 
this  ferment  a  sufficient  amount  of  cell  protein  is  formed, 
so  that  one  cell  may  form  two  daughter  cells  and  these  in 
turn  may  again  do  likewise.  As  an  example  of  this  may  be 
cited  the  growing  of  the  typhoid  bacillus  upon  an  agar  slant. 
In  the  original  few  organisms,  which  are  spread  upon  the 
media,  is  contained  a  ferment  which  is  capable  of  constructing 
the  specific  typhoid  protein  as  long  as  sufficient  required 
chemical  substances  are  present  in  the  agar  medium.  In 
the  case  of  a  unicellular  organism  like  this  the  property  of 
cell  division  and  multiplication  must  be  inherent  in  the 
cell,  and  the  process  of  cellular  reproduction  progresses  as 
long  as  suitable  media  are  furnished.  When,  however,  we 
come  to  consider  more  complex  organisms,  where  the 
functions  and  relations  of  one  cell  are  dependant  upon 
outside  cells,  we  have  a  somewhat  different  and  more  com- 
plex problem  to  consider.  Such  cells  cannot  continue  to 
multiply  without  limit,  else  the  cell  that  reproduced  the 
fastest  would  soon  predominate  and  outstrip  all  others. 
Here  the  growth  of  cells  must  be  governed  by  outside 
circumstances  to  some  extent,  although  it  would  appear 
reasonable  to  suppose  that  the  property  of  reproduction  is 
inherent  in  the  cell  as  well  as  in  the  case  of  the  simple 
unicellular  organisms.  This  property,  however,  must  lie 
dormant  at  times,  only  to  be  aroused  by  outside  stimuli. 
A  more  lucid  and  concise  way  of  expressing  this  idea  would 
be  to  state  that  the  reproductive  ferment  was  normally 
stored  up  within  the  cell  as  a  zymogen  or  inactive  ferment, 
which  becomes  active  when  called  upon  by  outside  stimuli. 
Even  in  the  unicellular  organisms  it  is  a  well-known  fact 
that  cell  division  can  be  hastened  or  retarded  by  outside 
conditions,  such  as  heat  and  cold  or  various  chemical  or 
electrical  stimuli.  With  regard  to  the  bearing  of  the  fore- 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     419 

going  upon  the  cancer  cell,  we  would  state  that  it  seems 
possible  to  conceive  that  a  cancer  cell  is  one  which  has  lost 
its  power  of  forcing  its  reproductive  ferment  back  into  an 
inactive  stage.  It  is  a  cell  whose  chemical  nature  has  become 
so  altered  that  the  reproductive  ferment  is  uppermost  and 
can  no  longer  be  influenced  by  outside  stimuli.  Just  like 
the  typhoid  bacillus  in  the  test-tube,  its  sole  purpose  is  now 
one  of  active  reproduction,  and  the  reproduction  will 
persist  until  the  cancer  cell  has  used  Up  all  protein  that  it 
is  capable  of  transferring  into  its  own  specific  protein,  or 
until  it  itself  is  destroyed  through  the  formation  by  the 
body  cells  of  an  antiferment  in  sufficient  amount  to  destroy 
all  cancer  cells.  How  often  the  body  itself  does  this  it  is 
difficult  to  surmise,  but  we  do  know  that  once  such  cells 
have  multiplied  so  as  to  form  a  palpable  tumor,  the  body 
is  very  seldom  able  to  cope  with  it. 

With  the  foregoing  in  mind  the  following  experiments 
will  throw  some  light  upon  the  subject,  especially  with 
regard  to  the  formation  of  an  antiferment  for  the  cancer  cell. 

As  has  been  brought  out  in  a  previous  article,1  certain 
definite  and  characteristic  blood  changes  are  brought  about 
by  the  injection  of  small  amounts  of  dead  cancer  protein 
into  a  living  host.  Two  forms  of  vaccine  have  been  used 
in  this  work,  cancer  residue  and  a  vaccine  made  of  the 
cancer  cell  in  its  entirety. 

Cancer  residue  is  prepared  by  dissecting  as  freely  as 
possible  the  cancer  material  from  all  surrounding  tissues. 
The  cancer  material  is  next  ground  in  a  meat  grinder,  then 
is  washed  with  water,  diluted  salt  solution,  alcohol,  and 
lastly  ether.  This  process  removes  salts,  fats,  wax,  several 
protein  bodies,  and  traces  of  carbohydrates.  The  remaining 
substance  is  then  heated  in  a  flask  with  a  reflux  condenser, 
with  from  fifteen  to  twenty  times  its  weight  of  a  2  per  cent, 
solution  of  sodium  hydroxide  in  absolute  alcohol,  and  by 
this  means  it  is  split  into  a  toxic  and  a  non-toxic  group. 
The  toxic  portion  is  soluble  in  the  alcohol,  the  non-toxic 

1  Jour.  Amer.  Med.  Assoc.,  November  16,  1912. 


420  PROTEIN  POISONS 

is  insoluble  in  alcohol,  but  soluble  in  water.  This  portion 
is  dissolved  in  normal  saline  solution  in  sufficient  quantities 
to  form  a  1  per  cent,  solution,  and  is  used  in  the  same 
manner  as  a  vaccine. 

Cancer-celled  vaccine  is  prepared  by  grinding  cancer  tissue 
finely  in  a  meat  grinder,  after  which  it  is  rubbed  up  as  a 
suspension  in  alcohol  in  a  sterile  mortar.  Next,  it  is  rubbed 
through  a  very  fine-meshed  sieve,  the  alcohol  filtered  off, 
and  the  collected  cell  substance  air-dried.  This  is  weighed 
and  then  placed  in  normal  saline  solution,  making  a  2  per 
cent,  cell  suspension.  To  this  0.5  per  cent,  phenol  is  added 
for  the  purpose  of  rendering  it  less  likely  to  become 
contaminated.  In  the  preparation  of  both  residues  and 
vaccines  it  has  been  found  that  satisfactory  blood  changes 
can  be  obtained  only  when  the  tumor  is  of  firm  consistency 
and  without  necrotic  or  infected  areas.  The  average  injec- 
tion of  a  1  per  cent,  residue  is  from  5  to  20  minims;  that  of 
a  2  per  cent,  cancer-cell  emulsion  is  from  5  to  10  minims. 

Sheep  and  rabbits  have  been  injected  intravenously, 
intra-abdominally,  and  subcutaneously  with  both  cancer 
residue  and  cancer  vaccine,  and  frequent  blood-counts 
made.  In  over  600  animals  the  blood  changes  have  been 
practically  uniform  except  in  about  10  animals  in  which 
the  vaccine  used  had  been  allowed  to  stand  too  long.  The 
accompaning  charts  show  that  while  the  percentage  of 
polymorphonuclear  and  small  mononuclear  leukocytes  are 
not  affected  with  any  degree  of  regularity,  the  proportion 
of  large  mononuclear  cells  is  invariably  increased  from  100 
to  400  per  cent,  within  from  twenty-four  to  forty-eight 
hours.  This  increase,  however,  is  of  short  duration  and 
recedes  with  rapidity  after  reaching  its  height. 

Fig.  24  illustrates  the  average  blood  change  obtained 
through  the  injection  of  an  active  residue.  The  preparation 
used  was  a  1  per  cent,  sarcoma  residue  which  had  been 
made  from  a  small  round-celled  sarcoma  of  the  mediastinum. 
Three  subcutaneous  injections  of  5  minims  each  were  given 
at  hourly  intervals,  the  first  blood  count  being  made  before 
the  injections  were  commenced.  The  second  count,  made 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     421 

seven  hours  later  showed  an  increase  in  polymorphonuclear 
cells  from  21  to  37  per  cent.,  and  a  corresponding  decrease 
in  small  mononuclear  leukocytes.  The  third  count  made, 
twenty-five  hours  after  the  first,  showed  the  characteristic 
increase  of  large  mononuclear  cells  which  were  here  regis- 


8-14-12          2-15-12    2-10-12 


MAST    ( 


FIG.  24. — Rabbit:  5  minims  of  1  per  cent,  sarcoma  residue  injected 
subcutaneously  at  8,  9,  and  10  A.M.  The  line  designated  as  P  indicates, 
in  this  and  the  following  charts,  count  of  polynuclears;  S  indicates  small, 
and  L  large  mononuclears. 

tered  at  27  per  cent.,  with  a  normal  at  the  first  injection  of 
10  per  cent.  Two  and  one-half  hours  later  the  percentage 
of  this  form  of  cell  had  returned  to  normal,  which  was 
again  followed  by  a  slight  increase.  It  is  probable  that  the 
highest  registration  of  large  mononuclear  cells  occurred 
between  the  second  and  third  counts. 


422 


PROTEIN  POISONS 


Fig.  25  represents  four  rabbits  which  were  given  0.5  c.c. 
each  of  cancer-cell  vaccine  by  different  methods.  The  first 
was  given  a  simple  subcutaneous  injection  of  rectal  adeno- 
carcinoma.  The  fourth  represents  the  same  amount  of 
the  same  tumor  given  intraperitoneally,  as  does  also  the 
third.  The  second  shows  the  injection  of  vaccine  made 
from  a  rapidly  growing  round-celled  sarcoma  of  the  neck. 


B 


11-1-1111-2-11  11-3-11 


ll-2-llll-3-llll-4.il 


1C 

S 
P 

L- 

C 

-10-11  10-17-1110-18.il 

/\ 

^ 

\ 

\ 

* 

\ 

A 

/ 

V 

/- 

/ 

-X 

x 

D 

10.10-1110-17-1110.18-11 


FIG.  25. — A,  rabbit  given  subcutaneous  injection  of  0.5  c.c.  rectal 
adenocarcinoma  (Morris) ;  B,  rabbit  injected  with  0.5  c.c.  small  round-celled 
sarcoma;  C,  rabbit  given  intraperitoneally  0.5  c.c.  rectal  adenocarcinoma 
(Morris) ;  D,  rabbit  given  intraperitoneally  0.5  c.c.  rectal  adenocarcinoma 
(Morris). 


From  these  charts  it  can  be  seen  that  the  intraperitoneal 
method  of  injection  gives  a  more  rapid  reaction  than  the 
subcutaneous. 

Fig.  26  illustrates  the  blood  changes  occurring  in  a  sheep 
following  the  injection  of  0.5  c.c.  of  sarcoma  residue. 

Figs.  27  and  28  show  more  frequent  differential  counts 
following  repeated  subcutaneous  injections. 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     423 

Fig.  29  illustrates  daily  counts  following  a  single  intra- 
peritoneal  injection  of  cells  from  a  breast  adenocarcinoma. 
This  is  rather  an  extreme  reaction,  inasmuch  as  the  pro- 
portion of  large  mononuclear  cells  is  increased  from  6  per 
cent,  to  35,  about  500  per  cent. 

Fig.  30  shows  a  very  slight  reaction  following  the.  injec- 
tion of  a  breast-cancer  residue  which  has  almost  lost  its 
activity. 


Day 
1          2          3 

POLY 

\ 

SMALL\ 

\ 

\> 

s 

£ 

^1 

/ 

/ 

!GE  / 

FIG.  26. — Sheep  injected  with  0.5  c.c.  sarcoma  residue. 

Figs.  31,  33,  and  34  illustrate  rabbits  giving  the  average 
reactions.  Fig.  32  represents  the  effect  of  giving  an 
intraperitoneal  injection  of  10  c.c.  of  breast-carcinoma 
cell-emulsion  into  a  rabbit  sensitized  to  sarcoma  vaccine, 
the  injection  being  given  when  the  percentage  of  large 
mononuclear  cells  was  at  its  highest  point.  Death  resulted 
in  from  eight  to  ten  hours. 

In  order  to  ascertain  just  what  bearing  this  change  in 
percentage  of  large  mononuclear  leukocytes  had  to  the 
formation  of  a  specific  ferment,  several  rabbits  were  sensi- 


424 


PROTEIN  POISONS 


tized  to  the  cancer-cell  and  then  at  varying  percentages 
10  c.c.  of  cancer-cell  emulsion  was  injected  intravenously 
into  the  animals.  In  unsensitized  rabbits  there  was  no 
noticeable  effect.  Rabbits  with  a  percentage  of  above 


FIG.  27 


<--l-9-12--X 1-10-12 **-l-ll-12--> 


FIG.  28 


-1-3-12--X 1-4-12-  - 

.19  4  8          10.15  a 

.M.         P.M.          A.M.        A.M.          P.M. 


8- 


\ 


7 


V 


FIGS.  27  and  28. — A,  rabbit  injected  subcutaneously  with  5  minims  of 

1  per  cent.  s.  r.  sarcoma  at   12  M.,   2   P.M.  and  3   P.M.  ;  B,  rabbit  injected 
subcutaneously  with  0.1  c.c.  of  2  per  cent,  mixed  residue  at  12.15  P.M., 

2  P.M.,  and  4  P.M. 

30  of  large  mononuclear  cells  usually  died  within  one  to 
three  hours,  and  rabbits  with  a  lower  percentage,  but  with 
a  marked  increase,  were  made  sick,  but  recovered. 

Sickness  is  immediate  after  injection;  the  animal  at  once 
falls  on  its  side  and  begins  violent  scratching.    The  respira- 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     425 

tion  is  labored  and,  when  death  ensues,  always  stops  before 
the  heart-beat.  Rabbits  which  die  from  a  dose,  after  being 
sensitized,  usually  have  a  stage  of  apparent  rest  from  one- 
half  to  three  hours  after  the  first  stage  of  excitability,  which 
lasts  for  from  three  to  five  minutes.  A  possible  explanation 


DAY 
1 2          3 


DAY 


GO 
55 
50 
45 
40 
35 
30 
25 
20 
15 
•m 

/ 

/* 

/ 

/ 

SMALL 

-POI 

Y 

\ 

\ 

\ 

\ 

/ 

/* 

<£ 

GE 

FiG.  29. — Rabbit  given  a  single 
intraperitoneal  injection  of  cells 
from  a  breast  adenocarcinoma. 


FIG.  30. — Rabbit  injected  with 
a  breast-cancer  residue  which  has 
almost  lost  its  activity. 


of  this  is  that  the  first  stage  of  excitement  is  due  to  the 
destruction  of  cancer  cells  and  consequent  liberation  of 
their  toxic  radical  by  the  specific  ferment  present  in  the 
blood  serum,  while  the  fatal  result  which  follows  later  is 
due  to  the  reaction  between  the  cancer  cell  and  the  large 
mononuclear  leukocytes. 


426 


PROTEIN  POISONS 


The  increased  percentage  of  large  mononuclear  leukocytes 
is  but  a  transitory  affair,  however,  which  usually  lasts  from 
four  to  ten  hours,  and  it  is  impossible  to  produce  fatal 
results  in  rabbits  by  intravenous  injections  of  cancer-cell 


FIG.  31 


1-SO-18       1-81-12  2-1-18 

P.M. 8  A.M. 


FIG.  32 


FIGS.  31  and  32. — A,  rabbit  167  W.,  given  5  minims  of  1  per  cent,  sar- 
coma at  8.30,  9.30,  and  10.30  A.M.;  B,  rabbit  54  T.,  given  5  minims  of  1 
per  cent,  sarcoma  at  1,  2,  and  3  P.M.;  at  12.15  P.M.,  10  c.c.  of  breast  car- 
cinoma-cell emulsion  (Mel.)  was  injected  intraperitoneally ;  rabbit  died 
in  eight  to  ten  hours. 

emulsion  after  this  stage  is  passed.  Consequently  we 
have  applied  the  term  "transitory  sensitization"  to  this 
phenomenon. 

If  we  bleed  a  rabbit  at  the  height  of  this  transitory  sensi- 
tization and  obtain  the  serum,  this,  when  mixed  with  cancer- 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL      427 

cell  emulsion  and  incubated  for  one  hour  will  produce 
marked  symptoms  of  poisoning  when  injected  intravenously 
into  a  normal  rabbit.  The  severity  of  symptoms  depends 
upon  both  the  amount  of  cancer-cell  emulsion  and  serum. 


FIG.  33 
A 

1-23-12        1-24-12    1-25-12 

2.49      4.30       8.30 


4   P.M.  11   A.M. 


FIG.  34 
B 

1-17-13         1-18-12          1-19-12 

1  P.M. 4  P.M. 8  P.M. 4  P.M. 8  A.M. 12  IK 


p 


s. 


7\ 


N 


/ 


A 


FIGS.  33  and  34. — A,  rabbit  54  Br.,  given  5  minims  of  1  per  cent,  sarcoma 
at  1,  2,  and  3  P.M.;  B,  rabbit  54  Bl.,  given  5  minims  of  1  per  cent,  sarcoma 
at  1,  2,  and  3  P.M. 


Given  5  c.c.  of  cancer-cell  emulsion  and  10  c.c.  of  the  serum, 
a  rabbit  of  from  500  to  1000  grams  will  be  sick  but  will 
invariably  recover. 

The  next  step  was  to  ascertain  whether  the  specific 
ferment  could  be  removed  from  the  large  mononuclear  leuko- 
cytes. For  this  purpose  rabbits  were  sensitized  to  the 


428  PROTEIN  POISONS 

cancer  cell  and  when  the  percentage  of  large  mononuclear 
leukocytes  was  at  its  highest  point,  the  blood  was  collected 
under  sterile  conditions,  in  0.333  per  cent,  acetic  acid. 
This  was  next  centrifugated  until  the  leukocytes  were 
thrown  down  and  the  supernatant  fluid  decanted.  The 
leukocytes  were  then  placed  in  a  sterile  mortar  and  mixed 
with  a  sufficient  quantity  of  sterile  quartz  sand  to  cover. 
This  was  then  ground  up  with  vigor  for  fifteen  minutes, 
with  the  frequent  addition  of  sterile  normal  salt  solution, 
until  five  times  the  volume  of  the  leukocytes  obtained  had 
been  added.  Next,  this  normal  saline  extract  was  separated 
by  passing  through  a  Berkefeld  filter  and  tests  made  for 
the  presence  of  the  specific  ferment  by  adding  varying 
amounts  of  the  leukocyte  extract  to  cancer-cell  emulsion, 
incubating  for  one  hour  and  then  injecting  intravenously 
into  rabbits.  Through  repeated  experiments  it  was  ascer- 
tained that  5  c.c.  of  2  per  cent,  cancer-cell  emulsion,  plus 
10  c.c.  of  leukocyte  extract,  which  was  obtained  when  the 
percentage  of  large  mononuclear  leukocytes  was  25  or 
above,  would,  when  injected  intravenously  into  a  rabbit 
of  from  500  to  1000  grams,  kill  within  from  one  to  five 
minutes.  This  is  well  illustrated  in  the  accompanying 
table. 

TABLE  XL VI. — SHOWING  RESULTS   OF   INJECTING   INTRAVENOUSLY   INTO 

RABBITS  CANCER-CELL  EMULSION  PLUS  LEUKOCYTE  EXTRACT 

INCUBATED  ONE  HOUR. 

Weight,          Rabbit  leukocyte 

Animal.          gm.                       extract.  +     Vaccine.  Result. 

343  G.           646.5  10  c.c   care.  100  5.0  c.c.  100  Died  1  min. 

167  G.           750.0  10  c.c.  sarc.  51  res.  5.0  c.c.  100  Died  1  min. 

157  W.        1500.0  10  c.c.  sarc.  100  5.0  c.c.  100  Very  sick  £ 

hr. ;  rec. 

53  Bl.        1050.0           7  c.c.  sarc.  51  res.  5.0  c.c.  100  Died  5  min. 

96  G.          1317.0           7  c.c.  sarc.  51  res.  2.5  c.c.  103  Died  4J  hr. 

251  W.          645.0  10  c.c.  sarc.  51  res.  Not  sick. 

251  Br.         842.0  15  c.c.  care.  100  Not  sick. 

167  W.          700.0  10  c.c.  normal  sal.  5.0  c.c.  100  Not  sick. 

The  entire  mass  of  leukocytes,  when  removed  from  the 
outside  of  the  Berkefeld  filter  and  mixed  with  5  c.c.  of 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     429 

cancer-cell  emulsion  and  incubated  one  hour,  failed  to  have 
any  effect  on  a  rabbit  when  injected  intravenously,  thus 
showing  that  the  specific  ferment  was  soluble. 

Again,  10  c.c.  of  leukocyte  extract,  prepared  in  the  same 
manner  from  a  normal  rabbit  which  had  not  been  sensitized, 
added  to  5  c.c.  of  cancer-cell  emulsion  and  incubated  one 
hour,  produced  no  effect  when  injected  intravenously  into 
a  normal  rabbit.  This  would  show  that  the  sensitized 
animal  possesses  some  specific  chemical  substance  which 
reacts  with  cancer  tissue  and  which  the  normal  rabbit  does 
not  possess.  That  this  substance  is  specific  for  malignant 
cells  is  further  shown  by  the  fact  that  10  c.c.  of  leukocyte 
extract  from  a  sensitized  animal,  plus  5  c.c.  of  a  2  per 
cent,  normal  skin  vaccine,  incubated  one  hour,  produces 
no  effect  when  injected  intravenously  into  a  rabbit  of 
450  grams. 

The  same  table  shows  also  that  sarcoma  residue  or 
vaccine  sensitizes  to  carcinoma  as  well  as  sarcoma,  and 
vice  versa,  so  that  the  probable  conclusion  is  that  the  chemical 
change  within  the  cell  is  the  same  for  both  sarcoma  and 
carcinoma. 

Organ  extracts  from  kidney,  liver,  brain,  spleen,  and 
heart,  made  by  grinding  these  organs  in  normal  saline  on 
successive  days,  after  sensitization  to  cancer  protein,  have 
no  effect  when  mixed  with  cancer-cell  emulsion,  incubated 
one  hour;  and  injected  intravenously  into  normal  rabbits. 

When  as  small  an  amount  as  1  c.c.  or  more  of  leukocyte 
extract  is  injected  directly  into  the  tumor  of  a  cancer 
patient  it  may  cause  sudden  and  severe  symptoms.  In  four 
cases,  when  this  procedure  was  adopted,  the  patient  has 
complained  within  from  one  to  five  minutes  of  difficulty 
of  respiration.  Next,  he  would  lose  consciousness,  which 
would  be  accompanied  by  rather  violent  muscular  twitchings 
and  lowered  pulse-rate.  This  stage  would  last  from  five 
to  ten  minutes,  and  would  be  followed  by  a  period  of  rest, 
from  fifteen  minutes  to  one  hour  in  duration,  which  in  turn 
would  be  followed  by  a  violent  chill,  and  temperature 
ranging  from  103°  to  106°  F.  This  would  last  from  one  to 


430  PROTEIN  POISONS 

six  hours  and  would  be  followed  by  from  twenty-four  to 
forty-eight  hours  of  extreme  exhaustion.  At  no  time  has 
such  a  reaction  been  obtained,  even  with  doses  of  10  c.c., 
when  given  subcutaneously  or  intravenously  away  from  the 
tumor;  although  a  chill  from  one  to  three  hours  after  injec- 
tion has  been  observed.  We  wish  now,  however,  to  call 
attention  only  to  the  animal  experiments,  the  above  being 
mentioned  simply  because  it  is  additional  proof  of  the 
presence  of  a  specific  ferment. 

It  is  of  interest  to  note  here  also  that  a  vaccine  prepared 
from  human  carcinoma  gives  a  much  higher  percentage  of 
large  mononuclear  leukocytes  when  injected  into  a  rabbit 
or  sheep  than  when  injected  into  a  human  being.  This 
has  been  observed  regardless  of  whether  the  human  being 
had  malignant  disease  or  not.  I  have  injected  cancer 
vaccine  into  myself,  and  the  highest  resulting  percentage 
of  large  mononuclear  leukocytes  was  15.  This  fact  is  of 
interest  when  we  consider  that  with  experimental  cancer 
the  animal  injected  must  always  be  of  the  same  family  as 
the  one  that  furnishes  the  tumor  in  order  to  obtain  a 
"take." 

It  should  not  be  understood  that  we  consider  the  ferment 
causing  reproduction  in  the  cancer  cell  to  be  of  exactly  the 
same  chemical  nature  as  the  active  ferment  of  a  reproducing 
normal  cell,  but  rather  that  we  are  dealing  with  a  chemically 
altered  constructive  ferment,  a  fact  that  we  will  demon- 
strate later. 

While  attempting  to  use  leukocyte  extract  from  animals 
sensitized  to  the  cancer  protein  in  a  therapeutic  way,  it 
was  found  that  extract  prepared  by  filtering  through  a  hard 
filter  paper  would  upon  repeated  usage  in  the  same  case 
cause  symptoms  of  sensitization;  while  the  use  of  extract 
prepared  by  Berkefeld  filtration  would  not  be  followed  by 
these  characteristic  phenomena.  At  the  same  time,  as 
previously  mentioned,  all  of  the  specific  ferment  passes 
with  ease  through  the  filter. 

In  order  to  ascertain  what  constituents  were  removed 
from  the  leukocyte  extract  by  passage  through  the^Berkefeld 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     431 

filter,  definite  amounts  were  analyzed  by  the  Scherer  and 
Hammersten  methods  for  total  albumins  and  globulins. 

Determination  of  Albumin  and  Globulin. — 25  c.c.  of  leuko- 
cyte extract  which  had  been  centrifugated  to  throw  down 
all  corpuscles  and  sand  used  in  its  preparation  was  passed 
through  a  filter  paper.  This  was  rendered  acid  with  acetic 
acid,  4  grams  of  NaCl  added,  and  then  heated  for  one-half 
hour  on  a  water-bath.  After  coagulation  had  occurred  this 
was  filtered  through  a  previously  dried  and  weighed  filter 
and  washed  with  hot  water  until  the  filtrate  ceased  to  give 
a  reaction  for  chlorides.  Next,  the  residue  was  washed 
with  absolute  alcohol  and  then  ether,  after  which  it  was 
dried  at  130°  to  a  constant  weight. 

Paper  and  albumin  plus  globulin        .      .      .      1 . 66658  grams 
Weight  of  paper 1.61150  grams 


Weight  of  albumin  and  globulin   .      .      .      .     0.05508  gram 

The  same  method  was  applied  to  25  c.c.  of  the  same  lot 
of  leukocyte  extract  after  passing  through  a  Berkefeld 
filter. 

Weight  of  filter  plus  albumin  and  globulin   .      1 . 55000  grams 
Weight  of  filter 1.53125  grams 

Weight  of  albumin  and  globulin   ....     0.01875  gram 

In  order  to  make  a  separate  determination  of  globulins 
and  albumins  the  following  method  of  Hammersten  was 
adopted:  25  c.c.  of  leukocyte  extract  before  filtration 
through  a  Berkefeld  was  added  to  30  grams  of  pulverized 
magnesium  sulphate.  This  was  warmed  to  30°  with  fre- 
quent stirring,  and  then  placed  in  the  cold  for  twenty-four 
hours.  This  was  then  filtered  through  a  weighed  filter, 
previously  dried  at  110°,  and  washed  with  magnesium 
sulphate  until  the  filtrate  ceased  to  give  a  reaction  for 
albumin  when  heated  with  acetic  acid.  Next,  the  filter 
was  dried  for  four  hours  at  110°  to  coagulate  the  globulin, 
after  which  the  magnesium  sulphate  was  washed  out  with 


432  PROTEIN  POISONS 

hot  water.     It  was  then  washed  with  alcohol  and  ether, 
and  dried  to  a  constant  weight  at  110°. 

Weight  of  globulin  and  filter   ....'.      1.5864  grams 
Weight  of  filter 1.5766  grams 


Weight  of  globulin 0 . 0098  gram 

25  c.c.  of  the  same  leukocyte  extract   after   Berkefeld 
filtration  was  treated  in  the  same  manner. 

Weight  of  globulin  and  filter 1.4913  grams 

Weight  of  filter 1.4766  grams 


Weight  of  globulin 0.0147  gram 

By  subtracting  these  globulin  weights  of  before  and 
after  filtration  from  the  above-given  albumen  plus  globulin 
weights  we  arrive  at  the  albumin  weights: 

Albumin  and  globulin  before  filtration     .      .      0.05508  gram 
Globulin  before  filtration 0.00980  gram 


Albumin  before  filtration 0.04528  gram 

Albumin  and  globulin  after  nitration       .      .     0.01875  gram 
Globulin  after  filtration 0.01470  gram 


Albumin  after  filtration 0 . 00405  gram 

From  the  above  it  can  be  seen  that  we  have  removed  a 
large  percentage  of  the  albumin  through  the  passing  of  the 
extract  through  the  Berkefeld  filter.  This  is  represented 
by  the  difference  between  0.04528  gram  and  0.00405  gram, 
which  is  0.0412  gram. 

The  difference  in  the  globulin  weight  gives  an  apparent 
increase  of  5  mg.  in  the  after-globulins,  an  amount  so  small 
that  it  can  be  neglected  when  compared  with  the  albumin 
difference. 

From  this  it  can  be  seen  that  the  sensitizing  portion  of 
the  extract  is  in  all  probability  contained  within  the  albumin, 
and  that  the  specific  enzyme  is  not  of  albuminous  nature. 


SPECIFIC  FERMENTS  OF  THE  CANCER  CELL     433 

While  the  globulins  are  filterable  this  does  not  prove  that 
the  ferment  is  a  globulin,  but  only  that  it  filters  through 
with  the  globulin.  The  enzyme  itself  may  be  of  much 
simpler  construction. 

In  conclusion  I  may  state  that  the  most  valuable  deduc- 
tions to  be  drawn  from  the  work  as  so  far  conducted  are 
as  follows: 

1.  Transitory  sensitization.     The  fact  that  an   animal 
may  be  sensitized  to  certain  proteins,  and  that  such  sensi- 
tization is  active  for  only  a  few  hours  is  of  extreme  interest 
and  importance. 

2.  The  active  transitory  ferment  formed  by  the  intro- 
duction of  cancer  protein  into  an  animal  is  not  an  albumin. 
It  is  either  a  globulin  or  of  simpler  chemical  structure. 

3.  Sensitization,  in  the  use  of  leukocyte  extract  at  least, 
is  probably  caused  by  the  albumin  content  of  the  solution. 

Other  Methods  Used. — In  order  that  a  more  definite 
knowledge  of  the  chemistry  of  this  specific  ferment  might 
be  obtained,  the  following  experiment  was  made:  120  c.c. 
of  sensitized  leukocyte  extract,  after  filtration  through  a 
Berkefeld  filter,  was  mixed  with  an  equal  volume  of  satu- 
rated ammonium  sulphate  solution.  This  was  allowed  to 
stand  overnight  and  the  precipitate  was  then  filtered  off. 
The  precipitated  globulins,  while  still  slightly  moist,  were 
removed  from  the  filter  paper  and  dissolved  in  a  saturated 
solution  of  sodium  chloride.  This  was  next  rendered  slightly 
acid  with  0.25  per  cent,  acetic  acid,  which  again  precipitated 
the  globulins.  When  the  precipitation  was  complete  and 
had  settled  to  the  bottom,  for  which  twelve  to  twenty-four 
hours  should  be  allowed,  the  globulin  was  collected  upon 
a  hard  filter  paper.  From  this  it  was  removed,  while  still 
slightly  moist,  to  a  sterile  watch-glass,  where  it  was  allowed 
to  dry.  The  globulin  was  next  weighed  and  dissolved  in 
normal  saline  solution  in  the  proportion  of  1  mg.  of  globulin 
to  1  c.c.  of  normal  saline. 

The  above  method  is  given  after  using  many  different 
modifications  of  the  same.  In  many  experiments  I  have 
precipitated  the  globulins  with  semisaturated  ammonium 
28 


434  PROTEIN  POISONS 

sulphate  solution,  then  dissolved  in  water,  equal  in  amount 
to  the  original  leukocyte  extract,  and  then  precipitated  again 
with  ammonium  sulphate  before  dissolving  in  saturated 
sodium  chloride  solution,  but  such  a  procedure  is  unneces- 
sary when  the  extract  has  been  previously  filtered  through  a 
Berkefeld,  inasmuch,  as  previously  shown,  four-fifths  of  the 
albumin  is  removed  at  this  time. 

Again,  it  was  my  custom  to  filter  after  dissolving  in 
saturated  sodium  chloride  solution,  but  this  has  been 
abandoned  because  of  the  fact  that  if  insufficient  saturated 
sodium  chloride  is  used,  much  globulin  may  be  retained 
by  the  filter,  and,  inasmuch  as  the  excess  of  ammonium 
sulphate  is  all  that  is  desired  to  be  rid  of,  both  time  and 
material  can  be  saved  by  not  filtering.  The  saturated 
sodium  chloride  solution  is  always  filtered  before  using. 

Again,  it  was  my  custom  to  dialyze  through  parchment 
paper  after  collecting  the  precipitate  thrown  down  by  0.25 
per  cent,  acetic  acid,  but  repeated  experiment  has  shown 
that  the  percentage  of  acetate  when  the  globulins  are 
dissolved  in  the  proportion  of  1  mg.  to  1  c.c.  of  normal 
saline  is  so  small  as  to  be  of  no  moment,  and  it  is  much 
preferable  to  have  a  solution  of  normal  saline  than  one  of 
sterile  water  for  injection  in  the  patient. 

It  should  be  added  that  precipitation  is  aided  by  allowing 
hot  water  to  pass  over  the  outside  of  the  flask  until  the 
temperature  of  the  contained  fluid  is  about  30°  C. 

The  following  experiment  has  been  repeated  many  times, 
so  that  the  possibility  of  error  is  slight. 

(a)  10  c.c.  of  soluble  globulins  from  sensitized  leukocyte 
extract  plus  5  c.c.  normal  saline  were  incubated  one  hour, 
and  10  c.c.  of  this  injected  intravenously  into  a  small  rabbit. 
There  was  no  change  in  the  animal. 

(6)  10  c.c.  of  soluble  globulins  from  sensitized  leukocyte 
extract  plus  5  c.c.  of  2  per  cent,  cancer-cell  emulsion  were 
incubated  one  hour  and  then  injected  intravenously  into 
a  moderate-sized  rabbit.  The  rabbit  died  in  one-half 
minute. 

From  the  above  it  can  be  seen  that  the  specific  ferment 


is  in  all  probability  a  part  of  the  globulin  content,  and  its 
activity  is  much  increased  by  obtaining  it  in  the  more 
purified  state. 

It  is  now  my  practice  to  test  the  strength  of  each  batch 
prepared  by  the  above  experiment  before  using  it  in  any 
given  case.  Sensitization  phenomena  have  been  entirely 
lacking  whenever  used  therapeutically,  which  was  not 
true  of  any  former  preparations.  The  filtered  leukocyte 
extract  when  used  in  small  amount  subcutaneously  rarely 
produced  sensitization  phenomena,  but  when  injected 
intravenously  in  increased  dosage  this  symptom  complex 
was  frequently  observed.  The  unfiltered  leukocyte  extract 
frequently  produced  sensitization  even  with  small  amounts 
given  subcutaneously,  so  the  conclusion  previously  arrived 
at;  that  the  albumin  content  was  responsible  for  these 
untoward  symptoms  is  apparently  confirmed.  It  must  not 
be  understood,  however,  that  the  globulins  do  not  sensitize. 
The  work  so  far  simply  shows  that  it  is  possible  to  remove 
the  albumin  which  contains  no  specific  ferment,  and  inas- 
much as  four-fifths  of  the  protein  has  been  removed,  much 
larger  doses  must  be  given  before  the  phenomena  of  sensiti- 
zation can  be  observed. 

The  above  substance,  because  of  its  specific  ferment 
action  and  its  apparent  chemical  nature,  I  would  designate 
as  anticancer  globulin. 


CHAPTER  XV 
THE  PHENOMENA  OF  INFECTION 

IT  may  be  of  interest  to  go  somewhat  into  detail  con- 
cerning our  ideas  of  the  phenomena  of  infection.  In  all 
infections  there  are  two  principal  factors — one  the  infecting 
virus  and  the  other  the  body  cell.  In  addition  to  these 
there  is  the  environment  in  which  the  struggle  for  supremacy 
between  the  virus  and  the  body  cell  takes  place.  This  con- 
sists of  the  unorganized  fluids  of  the  body,  and  is  of  great 
weight  in  determining  the  result  of  the  contest.  In  the  first 
place,  wrhat  do  we  know  of  the  infecting  virus?  As  we 
have  seen,  bacteria  are  particulate,  specific  proteins.  Since 
they  are  particulate,  we  speak  of  them  as  bacterial  cells. 
It  is  not,  however,  essential  that  an  infecting  virus  be 
particulate  in  the  sense  that  it  be  possessed  of  substance 
and  form  recognizable  to  our  limited  sense  of  sight  even 
when  aided  by  the  most  perfect  microscope.  There  are' 
many  filterable  viruses.  Some  pass  through  our  finest 
porcelain  filters  and  cannot  be  deposited  from  the  fluids  in 
which  they  exist  even  when  kept  for  hours  in  the  most 
efficient  centrifuge  manufactured.  Theoretically,  there 
is  no  reason  why  a  virus  may  not  exist  in  any  degree  of 
lability  of  structure.  The  bacteria  are  particulate  and 
solid,  which  means  that  their  structure  is  so  radically 
different  physically  from  the  medium  in  which  they  exist 
that  they  can  be  recognized  by  our  sight,  aided  by  proper 
magnifying  lenses,  but  viruses  may  be  semi-  or  wholly 
fluid.  In  such  instances  their  structure  is  not  sufficiently 
differentiated  from  the  medium  that  we  can  recognize 
them.  According  to  our  conception,  a  living  protein  does 


THE  PHENOMENA  OF  INFECTION  437 

not  necessarily  possess  a  form  recognizable  to  our  limited 
sense  even  when  aided  by  the  most  perfect  lenses. 

One  of  the  most  important  results  of  our  work,  in  our 
opinion,  is  the  demonstration  that  bacteria  are  chemically 
not  simple,  but  quite  complicated  in  structure.  Morpho- 
logically, they  show  but  little  or  no  differentiation  in  struc- 
ture, but  chemically  they  are  quite  as  complicated  and 
complex  as  many  of  the  cells  of  the  higher  animals.  They 
contain  carbohydrates,  nuclein  bodies,  and  polymers  of  the 
mono-  and  diamino-acids.  They  are  glyconucleoproteins. 
We  interpret  this  as  signifying  that  functionally  they  are 
highly  developed. 

While  an  infecting  virus  may  be  solid,  semisolid,  gela- 
tinous, or  liquid,  we  will,  in  the  further  consideration  of 
the  phenomena  of  infection,  take  the  particulate  type,  the 
bacterium,  as  an  example  of  an  infecting  agent. 

What  are  some  of  the  capabilities  of  a  bacterial  cell?  In 
the  first  place  it  possesses  that  attribute  which  distinguishes 
and  characterizes  all  living  matter — the  capability  of  growth 
and  reproduction.  In  order  to  grow  and  multiply  its 
molecular  structure  must  be  labile — in  a  sate  of  constant 
change.  Some  bacteria  under  certain  conditions  may  pass 
into  a  resting  state  characterized  by  the  formation  of  spores, 
but  these  are  awakened  into  life  when  the  environment 
becomes  fit,  and  the  spore  develops  into  the  active  form 
when  it  infects.  In  all  instances  the  active,  infecting  agent 
is  a  living  protein,  capable  of  growth  and  multiplication. 
In  order  to  do  this  it  must  carry  on  a  constant  exchange 
in  matter  with  the  medium  in  which  it  exists.  It  must 
assimilate  and  eliminate.  It  must  absorb  groups  from  the 
molecules  about  it,  and  cast  out  those  which  it  has  already 
used.  Stop  this  process  and  the  continuation  of  life  is 
impossible.  Every  living  cell,  be  it  bacterial,  vegetable, 
or  animal,  must  feed  or  cease  to  exist.  Besides,  a  cell  is 
limited  in  its  food  supply  by  that  which  lies  within  its 
reach.  There  must,  therefore,  be  a  certain  supporting 
relation  between  the  bacterial  cell  and  the  medium.  The 
groups  derived  from  the  medium  must  fit  into  the  molecular 


438  PROTEIN  POISONS 

structure  of  the  cell,  otherwise  they  would  be  of  no  service 
to  it.  This  necessitates  the  cleavage  of  the  molecules  of 
the  medium  along  definite  lines.  Many  kinds  of  cells  may 
live  in  the  same  or  like  media,  but  for  each  kind  of  cell  the 
cleavage  of  the  medium  must  be  specific.  From  this  it 
follows  that  the  agent  by  which  the  cleavage  products  are 
secured  must  be  supplied  by  the  cell  itself,  and  must  be 
peculiar  to  that  kind  of  cell.  These  cleavage  agents  which 
prepare  foods  for  the  cell  from  the  medium  are  known  as 
ferments,  and  each  kind  of  cell  has  its  own  characteristic 
and  specific  ferments.  As  to  the  real  nature  of  ferments, 
we  know  little  or  nothing,  but  that  every  kind  of  cell  has 
its  specific  ferment  or  ferments,  we  do  know.  The  same 
ferment  may  not  be  able  to  break  up  all  proteins.  In  this 
respect  there  are  great  variations  in  the  proteolytic  fer- 
ments. Some  digest  a  wide  variety  of  proteins  while  others 
are  capable  of  acting  only  on  one  specific  protein.  There 
must  be  a  relation  between  the  ferment  and  its  substrate. 
As  Fischer  once  said,  the  former  must  fit  the  latter  as  a 
key  fits  into  the  lock,  and  as  there  are  master  keys  that 
open  many  doors,  so  there  are  general  proteolytic  ferments, 
and  as  there  are  special  keys  that  fit  only  one  lock,  so  there 
are  specific  proteolytic  ferments.  It  will  be  observed  that 
we  have  used  the  word  "  specific"  in  two  senses  in  speaking 
of  proteolytic  ferments.  Each  kind  of  cell  has  its  specific 
ferment,  and  each  protein  may  have  its  specific  ferment. 
This  double  use  of  the  term  "specific"  should  be  borne  in 
mind,  since  there  seems  to  be  no  way  to  avoid  it. 

It  follows  from  what  has  been  said  that  a  bacterium 
placed  in  a  medium  in  which  its  ferment  is  ineffective  cannot 
grow  and  multiply.  A  bacterium  which  cannot  grow  and 
multiply  in  the  animal  body  cannot  cause  an  infection.  Its 
inability  to  grow  and  multiply  in  the  animal  body  may  be 
due  to  the  fact  that  its  ferment  or  ferments  cannot  digest 
or  properly  break  up  the  proteins  of  the  animal  body.  This 
is  one  of  the  reasons  why  the  great  majority  of  bacteria 
are  non-pathogenic  or  are  harmless.  These  organisms 
when  grown  on  suitable  media  produce  just  as  much  poison 


THE  PHENOMENA  OF  INFECTION  439 

as  the  pathogenic  bacteria,  but  not  being  able  to  feed  upon 
the  proteins  of  the  body  they  die.  This,  however,  is  not 
the  sole,  and  probably  not  the  most  important,  cause 
of  the  failure  of  so  many  varieties  of  bacteria  to  do  harm 
to  the  higher  animals.  What  has  been  said  about  the  pro- 
duction of  ferments  by  the  bacterial  cell  is  equally  true  of  the 
body  cell.  In  fact,  it  is  true  of  every  living  cell.  The  body 
cell  has  its  specific  ferments,  and  the  bacterial  cell  being 
protein  substance  is  liable  to  be  digested  by  the  'ferments 
elaborated  by  the  body  cells. 

In  the  inability  of  the  bacterial  cell  to  grow  in  the  animal 
body  either  because  it  cannot  feed  upon  the  proteins  of 
the  body,  or  because  it  is  itself  destroyed  by  the  ferments 
elaborated  by  the  body  cells  lies  the  fundamental  expla- 
nation of  all  forms  of  bacterial  immunity  either  natural 
or  acquired.  Toxin  immunity  needs  further  explanation. 
Certain  bacteria,  of  which  the  diphtheria  bacillus  may  be 
taken  as  a  type,  elaborate  soluble,  extracellular  substances 
known  as  toxins.  These  are  probably  ferments  or  closely 
allied  bodies.  They  resemble  ferments  in  the  following 
particulars:  (1)  They  are  destroyed  by  heat.  (2)  They 
act  in  very  dilute  solution.  (3)  When  repeatedly  injected 
into  animals  in  non-fatal  doses  they  cause  the  body  cells 
to  elaborate  antibodies  which  neutralize  the  toxin  both 
in  vivo  and  in  vitro.  (4)  In  the  development  of  their  effects 
a  period  of  incubation  is  required.  (5)  It  has  been  shown 
by  Abderhalden,  by  optical  methods,  that  they  have  a 
cleavage  effect  upon  proteins.  They  split  complex  proteins 
into  simpler  bodies.  In  other  words,  they  have  a  proteo- 
lytic  action.  (6)  They  are  specific  in  two  senses,  (a)  They 
are  specific  according  to  the  cell  which  produces  them. 
Diphtheria  toxin  is  elaborated  by  the  diphtheria  bacillus 
and  by  no  other  organism.  The  toxin  of  snake  venom  is  a 
specific  product  of  the  poisonous  gland  of  the  snake,  and 
this  is  further  specific  inasmuch  as  that  produced  by  the 
glands  of  one  species  is  different  from  that  elaborated  in 
another  species.  (6)  They  are  specific  in  the  antibody 
elaborated  in  the  animal  body  after  repeated  injections  of 


440  PROTEIN  POISONS 

non-fatal  doses.  Diphtheria  antitoxin  protects  only  against 
diphtheria  toxin,  and  not  against  that  of  the  tetanus  or 
dysentery  bacillus,  or  that  of  snake  venom. 

The  side-chain  theory  evolved  by  the  genius  of  Ehrlich 
best  explains  the  action  of  toxins  and  the  production  of 
antitoxins.  Without  subscribing  to  all  the  details  of  this 
theory,  we  believe  that  it  is  a  biological  law  that  when  a 
living  cell  is  attacked  by  a  destructive  ferment  or  toxin  it 
tends  to  elaborate  an  antiferment  or  antibody.  This  is 
one  of  the  ways  in  which  the  living  cell  may  protect  itself. 
The  formation  of  such  antibodies  in  multicellular  animals 
is  one  of  the  factors  in  the  fine  adjustment  essential  to 
harmony  of  action  between  different  tissues  and  organs. 
It  best  explains  the  fact  that  the  digestive  organs  do  not 
harm  themselves,  and  the  antitryptic  action  of  blood-serum 
is  one  of  the  most  interesting  and  important  phases  of 
parenteral  digestion. 

The  number  of  pathogenic  bacteria  which  produce 
toxins,  at  least  in  appreciable  quantity,  is  small,  and  the 
action  of  toxins  and  antitoxins  in  infections  due  to  those 
organisms  which  do  not  produce  such  bodies  is  of  minor 
importance.  Since  all  bacteria,  and  in  fact  all  living  cells 
produce  ferments,  and  since  every  ferment,  so  far  as  we 
know,  may  lead  cells  acted  upon  by  them  to  produce  anti- 
ferments,  there  may  be  some  toxin  and  antitoxin  action 
in  all  infections,  but  in  most  bacterial  infections  such 
action  is  overshadowed  by  processes  much  more  powerful 
in  their  effects. 

In  our  opinion  the  action  of  the  diphtheria  bacillus  may 
be  stated  as  follows:  The  organism  finds  lodgement  and 
the  conditions  for  growth  favorable  in  the  upper  air  pas- 
sages. Here  it  grows  in  mass  and  may  kill  by  mechanical 
obstruction.  It  produces  its  soluble,  diffusible  toxin, 
which  has  the  properties  of  a  ferment  and  splits  up  the 
proteins  of  the  body,  setting  free  the  protein  poison.  In 
case  of  recovery  or  in  the  production  of  antitoxin  in  animals, 
the  body  cells  elaborate  an  antiferment  or  antitoxin  which 
neutralizes  the  toxin  and  prevents  its  cleavage  action. 


THE  PHENOMENA  OF  INFECTION  441 

The  bacilli  in  the  throat  are  not  destroyed  by  natural 
recovery  or  by  cure  with  antitoxin,  but  the  action  of  the 
toxin  is  prevented  by  the  antibody.  It  is  not,  in  our  opinion, 
the  toxin  itself  which  kills,  but  a  cleavage  product  which 
results  from  the  action  of  the  toxin  on  the  proteins  of  the 
body. 

All  ferments  are  of  cellular  origin.  This  does  not  mean 
that  ultramicroscopic  forms  of  life  or  non-particulate  living 
organisms,  if  there  be  such,  do  not  produce  ferments.  It 
would  probably  be  better  to  say  that  all  ferments  are  the 
products  of  living  organisms  and  that  there  can  be  no  living 
organism  which  does  not  produce  its  specific  ferment. 
We  cannot  conceive  of  life  without  ferment  action,  because 
all  living  things  must  feed  and  food  assimilation  without 
ferment  action  is  inconceivable.  Food  must  be  fitted  for 
assimilation,  and  this  is  dependent  upon  ferment  action. 

Ferments  are  intra-  and  extracellular.  All  are  formed 
within  the  cell,  but  some  diffuse  into  the  medium  while 
others  do  not.  In  some  instances  at  least  cell  permeation 
by  the  pabulum  is  essential  to  the  feeding  of  the  cell.  In 
other  instances  it  is  highly  probable  that  the  ferment  is 
accumulated  on  the  cell  surface  and  there  acts  upon  the 
pabulum.  In  still  other  instances  the  ferment  diffuses 
into  the  medium  more  or  less  widely  from  the  cell  which 
elaborates  it.  Many  cells  produce  both  intra-  and  extra- 
cellular ferments,  and  these  are  not  necessarily  the  same. 
In  some  instances,  probably  in  most  cells,  the  intracellular 
ferment  cannot  be  extracted  from  the  cell  or  obtained  in 
soluble  form  without  destruction  of  the  cell.  This  does 
not  mean  that  it  must  exist  in  the  soluble  form  before  it 
can  manifest  its  cleavage  action.  The  pabulum  may  per- 
meate the  cell  and  in  this  location  be  split  up  by  the  intra- 
cellular ferment.  We  have  insisted  upon  this  as  an  explana- 
tion of  the  well-established  fact  that  soluble  proteins 
sensitize  much  more  readily  and  completely  than  insoluble 
ones. 

It  will  be  well  to  illustrate  what  we  have  said  about 
cellular  ferments  by  a  condensed  sketch  of  the  work  that 


442  PROTEIN  POISONS 

has  been  done  on  the  germicidal  properties  of  the  blood. 
As  early  as  1872  Lewis  and  D.  Cunningham1  demonstrated 
that  non-pathogenic  bacteria  injected  into  the  circulation 
soon  disappear.  In  the  blood  of  12  animals  thus  treated 
bacteria  could  be  found  after  six  hours  in  only  7.  Of  30 
animals,  bacteria  were  found  in  the  blood  of  only  14  after 
twenty-four  hours,  and  of  17  animals  bacteria  were  found 
in  the  blood  of  only  2  when  the  examination  was  made 
from  one  to  seven  days  after  the  injection.  In  1874, 
Traube  and  Geschiedlen2  found  that  arterial  blood,  taken 
under  aseptic  precautions  from  a  rabbit  into  the  jugular 
vein  of  which  1.5  c.c.  of  a  fluid  rich  in  putrefactive  bacteria 
had  been  injected  forty-eight  hours  previously,  failed  to 
undergo  decomposition  when  kept  for  months.  These 
investigators  attributed  the  germicidal  properties  of  blood 
to  its  ozonized  oxygen.  Like  results  were  obtained  by 
Fodor3  and  Wysokowicz.4  The  latter  accounted  for  the 
disappearance  of  the  bacteria,  not  through  the  germicidal 
action  of  the  blood,  but  by  supposing  that  they  found 
lodgement  in  the  capillaries.  The  first  experiments  made 
with  extravascular  blood  were  conducted  by  Grohmann5 
under  the  direction  of  A.  Schmidt  in  his  researches  on  the 
coagulation  of  the  blood.  It  was  found  that  the  virulence 
of  anthrax  bacilli,  as  demonstrated  by  their  effect  on  rabbits, 
was  diminished  by  being  kept  in  blood.  He  supposed  that 
the  bacilli  were  altered  in  some  way  by  the  process  of 
coagulation.  In  1887  Fodor6  made  a  second  contribution 
on  this  subject,  in  which  he  combated  the  retention  theory 
•  of  Wysokowicz.  One  minute  after  the  injection  of  1  c.c. 
of  an  anthrax  culture  into  the  jugular  vein,  in  eight  samples 
of  blood,  Fodor  found  only  one  colony  of  the  bacillus.  He 


1  Eighth  Annual  Report  of  the  Sanitary  Commission  of  the  Government 
of  India. 

2  Schlesische  Gesellschaft  f.  Vaterland.  Cultur. 

8  Archiv  f.  Hygiene,  iv,  1886.  4  Zeitsch.  f.  Hygiene,  i,  1886. 

8  Ueber  die  Einwirkung  d.  Zellenfrien  Blut-plasma  auf  einige  pflanzliche 
Mikroorganismen,  Dorpat,  1884. 
6  Deutsch.  med.  Woch. 


THE  PHENOMENA  OF  INFECTION  443 

also  took  blood  from  the  heart  with  a  sterilized  pipette, 
and  added  anthrax  bacilli  to  it.  This  was  kept  at  38°,  and 
plates  made  from  time  to  time  showed  rapid  diminution 
in  the  number  of  bacteria,  until  after  a  time,  when  the  blood 
having  lost  its  germicidal  properties,  the  number  rapidly 
increased.  In  1888  Nuttall,1  working  under  the  direction 
of  Fliigge,  used  defibrinated  blood  taken  from  various 
species  of  animals,  rabbits,  mice,  pigeons,  and  sheep,  found 
that  the  blood  destroyed  the  bacillus  anthracis,  b.  subtilis, 
b.  megatherium,  and  staphylococcus  aureus.  He  also  con- 
firmed the  finding  of  Fodor  that  after  a  while  the  blood 
loses  its  germicidal  properties  and  becomes  a  suitable 
culture  medium.  Continuing  this  work,  Nissen2  reached 
the  following  conclusions:  (1)  The  addition  of  small 
quantities  of  salt  solution  or  bouillon  to  the  blood  does  not 
destroy  its  germicidal  properties.  (2)  The  bacilli  of  cholera 
and  typhoid  fever  are  easily  destroyed  by  fresh  blood.  (3) 
For  a  given  volume  of  blood  there  is  a  maximum  number 
of  bacilli  that  can  be  destroyed.  (4)  Blood  whose  coagu- 
lability has  been  destroyed  by  peptone  injection  is  still 
germicidal.  (5)  Blood  in  which  coagulation  is  prevented 
by  the  addition  of  25  per  cent,  of  magnesium  sulphate 
has  its  germicidal  properties  decreased.  (6)  Filtered  blood 
plasma  from  the  horse  is  germicidal.  Behring3  attributed 
the  germicidal  action  of  the  blood  of  the  white  rat  on  the 
anthrax  bacillus  to  its  great  alkalinity.  In  1890,  Buchner 
and  his  students4  published  their  first  contribution  on  the 
germicidal  properties  of  blood  serum.  At  first  Buchner 
believed  that  the  germicidal  constituent  of  serum  is  the 
serum  albumin  and  the  conclusions  were  stated  as  follows: 
(1)  The  germicidal  action  of  blood  is  not  due  to  the  phago- 
cytes, because  it  remains  after  destruction  of  the  leuko- 
cytes by  alternating  freezing  and  thawing.  (2)  The  germi- 
cidal properties  of  the  cell-free  serum  must  be  due  to  its 


1  Zeitsch.  f.  Hygiene,  iv,  353. 

2  Ibid.,  1889,  vi,  487.  3  Ibid.,  1889,  vi,  1.7. 
4  Archiv  f.  Hygiene,  1890,  x,  84,  101,  121,  149. 


444  PROTEIN  POISONS 

soluble  constituents.  (3)  Neither  neutralization  of  the 
serum,  nor  the  addition  of  pepsin,  nor  the  removal  of 
carbon  dioxide  gas,  nor  treatment  with  oxygen  has  any 
effect  upon  the  germicidal  properties  of  the  blood.  (4) 
Dialysis  of  the  serum  against  water  destroys  its  activity, 
while  dialysis  against  0.75  per  cent,  salt  solution  does  not. 
In  the  diffusate  there  is  no  germicidal  substance.  The 
loss  by  dialysis  with  w^ater  must  be  due  to  the  withdrawal 
of  the  inorganic  salts  of  the  serum.  (5)  The  same  is  shown 
to  be  the  case  when  the  serum  is  diluted  with  water,  and 
when  it  is  diluted  with  the  salt  solution.  In  the  former 
instance  the  germicidal  action  is  destroyed,  while  in  the 
latter  it  is  not.  (6)  The  inorganic  salts  have  in  and  of  them- 
selves no  germicidal  action.  They  are  active  only  insofar 
as  they  affect  the  normal  properties  of  the  albuminates  of 
the  serum.  The  germicidal  properties  of  the  serum  reside 
in  the  albuminous  constituents.  The  difference  in  the 
effects  of  the  active  serum  and  that  which  has  been  heated 
to  55°  is  due  to  the  altered  condition  of  the  albuminate. 
The  difference  may  possibly  be  a  chemical  one  (due  to 
cleavage  within  the  molecule)  or  it  may  be  due  to  changes 
in  mycelial  structure.  The  albuminous  bodies  act  upon 
the  bacteria  only  when  the  former  are  in  an  active  state. 
Vaughan  and  McClintock1  called  attention  to  a  contra- 
diction between  Buchner's  work  and  his  conclusions,  in 
the  following  language:  "We  wish  at  this  point  to  call 
attention  to  an  inconsistency  between  the  results  obtained 
by  Buchner  and  the  conclusions  that  he  draws:  In  experi- 
ment 45  he  renders  the  serum  slightly  acid,  and  adds  0.1 
gram  of  pepsin  to  each  5  c.c.  of  serum  (showing  by  a  side 
experiment  that  this  pepsin  actively  digests  coagulated 
egg  albumen  in  neutral  solution)  and  finds  that  the  digestive 
action  of  the  pepsin  does  not  lessen  the  germicidal  properties 
of  the  serum.  In  fact,  he  states  this  in  his  conclusion,  but 
his  ultimate  opinion  and  the  one  held  by  him  in  his  latest 
contribution,  is  that  the  germicidal  constituent  of  the 

1  Med.  News,  December  23,  1893. 


THE  PHENOMENA  OF  INFECTION  445 

blood  is  the  serum  albumin.  How  much  serum  albumin 
remains  in  blood  serum  after  it  has  been  thoroughly  digested 
with  pepsin  ?  He  could  scarcely  have  chosen  a  more  positive 
method  of  demonstrating  that  the  germicidal  constituent 
is  not  serum  albumin.  Either  his  pepsin  was  not  active 
and  on  this  supposition  his  experiment  was  without  value, 
or  the  active  constituent  of  the  blood-serum  is  a  substance 
that  is  not  destroyed  or  materially  altered  by  peptic  diges- 
tion. We  know  that  the  peptones  not  only  have  no 
germicidal  properties,  but  that  they  belong  to  that  class 
of  proteins  that  is  most  favorable  to  the  growth  of  bacteria. 
We  recognize  this  fact  when  we  add  peptones  to  the  various 
artificial  media  on  which  we  cultivate  bacteria."  We  will 
return  to  this  point  after  proceeding  farther  with  the 
chronological  order  in  which  this  research  has  developed. 

Prudden1  found  that  ascitic  and  hydrocele  fluids  restrain 
the  development  of  certain  bacteria.  Rovighi2  reported 
that  the  germicidal  action  of  the  blood  is  increased  in 
febrile  conditions.  Pekelharing3  enclosed  anthrax  spores  in 
bits  of  parchment  and  introduced  these  under  the  skin 
of  rabbits.  Thus  treated  the  spores  soon  lost  their  viru- 
lence and  finally  their  capability  of  growth.  The  destruc- 
tion of  these  spores  could  not  have  been  due  to  phagocytes 
which  did  not  penetrate  the  parchment,  but  must  have 
been  caused  by  soluble  substances.  Behring  and  Nissen4 
found  that  the  serum  of  the  white  rat,  the  dog,  and  the 
rabbit  destroy  anthrax  bacilli,  while  serum  obtained  from 
the  mouse,  sheep,  guinea-pig,  chicken,  pigeon,  and  frog 
has  no  such  action.  It  will  be  observed  that  in  this  there 
is  no  constant  relation  between  the  germicidal  action  of 
the  blood  of  animals  of  different  species  and  their  suscep- 
tibility to  the  infection.  Thus  the  rabbit  is  highly  suscep- 
tible to  anthrax  notwithstanding  the  fact  that  its  blood 
destroys  large  numbers  of  this  organism.  On  the  other 
hand  the  chicken  is  immune  to  anthrax  from  the  moment 

1  Medical  Record,  1890. 

2  Atti  della  Accad.  Med.  di  Roma,  1890. 

3  Ziegler's  Beitrage,  viii.  4  Zeitsch.  f.  Hygiene,  1890,  viii,  412. 


446  PROTEIN  POISONS 

when  it  comes  from  the  shell,  and  yet  the  bacillus  grows 
luxuriantly  in  the  extravascular  blood  of  the  chick.  Hankin1 
was  one  of  the  first  to  show  that  other  cells,  besides  the 
leukocytes,  contain  germicidal  substances.  He  made  several 
contributions  to  the  study  of  so-called  "defensive  proteins," 
which  he  believed  to  be  globulins.  This  is  interesting  in 
view  of  the  fact  that  ferments  are  often  carried  down  with 
globulins  on  precipitation  with  neutral  salts.  Bitter2  was 
unable  to  confirm  Hankin's  work,  but  it  is  needless  to  go 
into  this  because  we  now  know  that  many  cells  elaborate 
germicidal  substances.  Christmas3  prepared  a  germicidal 
substance  from  the  spleen  and  other  organs  by  the  following 
method:  The  animal  was  killed  with  ether,  opened  under 
aseptic  precautions,  the  organ  removed,  cut  into  fine  pieces, 
covered  with  50  c.c.  of  glycerin,  and  allowed  to  stand  for 
twenty-four  hours  and  then  filtered.  The  filtrate  is  treated 
with  five  times  its  volume  of  alcohol  and  the  precipitate 
is  immediately  collected  and  washed  with  absolute  alcohol. 
Traces  of  alcohol  are  removed,  so  far  as  possible,  by  pressure, 
and  the  precipitate  is  dissolved  in  25  c.c  of  distilled  water, 
and  air  is  blown  through  the  solution  to  destroy  last  traces 
of  alcohol;  then  the  fluid  is  filtered  and  its  germicidal 
action  tested.  Bitter4  strove  hard  to  find  fault  with  this 
agent  and  its  method  of  preparation.  He  found  it  a  power- 
ful germicide,  but  he  could  not  reconcile  the  fact  that  the 
preparation  of  Christmas  still  proved  a  powerful  germicide 
after  it  had  been  heated  to  65°,  while  blood-serum  loses  its 
germicidal  effect  when  heated  to  55°.  Buchner5  had  the 
following  to  say  on  this  point :  "  A  method  given  by  Christ- 
mas for  the  preparation  of  germicidal  solutions  from  the 
organs  of  normal  rabbits  has  also  been  tested  by  Bitter. 
Germicidal  solutions  were  indeed  obtained,  which,  however, 
differed  materially  from  active  serum,  for  in  three  experi- 
ments, notwithstanding  heating  to  65°,  the  germicidal 

1  Centralbl.  f.  Bakt.,  1891,  ix,  336. 

2  Zeitsch.  f.  Hygiene,  1892,  xii,  328. 

3  Annales  de  1'Institut  Pasteur,  1891,  v,  487. 

4  Loc.  cit.  6  Archiv  f.  Hygiene,  1893,  xvii,  112. 


THE  PHENOMENA  OF  INFECTION  447 

action  remained."  We  have  gone  into  this  detail  concerning 
the  preparation  of  Christmas  for  the  following  reasons: 
(1)  It  remains  to  day  a  good  method  of  preparing  a  germi- 
cidal agent  from  the  spleen  or  other  tissue.  (2)  Its  method 
of  preparation  indicates  that  it  is  a  ferment.  (3)  It  is  an 
illustration  of  the  fact  that  the  degree  of  heat  borne  by  a 
ferment,  without  being  inactivated,  is  dependent  in  part  at 
least  on  the  character  of  the  solvent  in  which  the  ferment 
is  found. 

Emmerich  and  his  students1  made  the  following  experi- 
ments: A  serum  was  dialyzed  against  distilled  water  until 
its  globulin  was  precipitated.  The  globulin-free  serum 
was  precipitated  with  alcohol,  and  the  serum  albumin 
thus  thrown  down  was  dissolved  in  0.05  per  cent,  solution 
of  potassium  hydroxide.  This  solution  was  found  to  be 
markedly  germicidal,  and  the  conclusion  reached  was  that 
the  germicidal  constitutent  of  blood  serum  was  an  alkaline 
albuminate. 

Vaughan  and  his  students2  published  their  first  paper 
upon  the  germicidal  properties  of  nuclein.  In  their  first 
contribution  they  showed  that  nucleins  prepared  from 
testes,  thyroid  gland,  and  yeast  cells  are  markedly  germi- 
cidal to  both  pathogenic  and  non-pathogenic  bacteria. 
In  1894  Kossel3  quite  independently  announced  the  dis- 
covery of  the  germicidal  action  of  nuclein  and  nucleic 
acid.  Vaughan  not  only  demonstrated  the  germicidal 
action  of  nuclein  in  vitro,  but  also  showed  (1)  that  it 
protected  rabbits  against  subsequent  inoculation  with 
the  pneumococcus;  (2)  it  also  protected  a  considerable 
percentage  of  rabbits  against  inoculation  with  the  bacillus 
tuberculosis;  (3)  that  it  had  a  curative  effect  on  rabbits 
already  inoculated  with  tuberculosis;  and  (4)  that  it 
apparently  benefited  initial  tuberculosis  in  man. 

It  now  turns  out  that  the  germicidal  action  attributed 
by  Vaughan  and  Kossel  to  nuclein  was  probably  not  due  to 

1  Centralbl.  f.  Bakt.,  1892,  xii,  364. 

2  Medical  News,  May  20,  1893. 

3  Archiv  f.  Anat.  u.  Physiol.,  Physiolog.  Abtheilung,  1894,  194. 


448  PROTEIN  POISONS 

this  agent,  but  to  ferments  which  came  out  with  the  nuclein 
from  the  cells  used  in  its  preparation.  This  should  have 
been  known  at  the  time  the  work  was  done,  because  both 
of  these  investigators  were  aware  of  the  fact  that  a  tempera- 
ture short  of  boiling  destroyed  the  germicidal  properties 
of  their  solutions,  but  we  did  not  know  so  much  about 
cell  ferments  then  as  we  do  now.  This,  of  course,  does  not 
mean  that  all  the  results  obtained  with  preparations  of 
nuclein,  such  as  an  increase  in  the  number  of  leukocytes, 
were  due  to  the  ferment  contained  in  the  preparations, 
but  it  is  more  than  probable  that  the  germicidal  action  was 
due  to  the  ferment.  The  whole  matter  demands  reinves- 
tigation. 

It  should  be  stated  that  Vaughan  and  McClintock1 
demonstrated  the  presence  of  nuclein  in  blood-serum.  This 
was  done  by  precipitating  a  large  amount  of  serum,  obtained 
under  aseptic  conditions,  wyith  alcohol  and  digesting  the 
precipitate  with  artificial  gastric  juice  so  long  as  digestion 
proceeded,  the  completion  of  digestion  being  indicated  by 
failure  to  respond  to  the  biuret  test.  The  small  amount 
of  protein  material  which  wholly  resisted  gastric  digestion, 
and  which  could  be  only  nuclein,  was  dissolved  in  0.12 
per  cent,  potassium  hydroxide,  and  its  germicidal  action 
demonstrated  on  the  bacillus  of  cholera,  anthrax  bacillus, 
typhoid  and  colon  bacilli,  and  on  various  cocci.  At  the 
same  time  it  was  shown  that  a  0.5  per  cent,  solution  of  the 
alkali  was  without  effect  upon  these  organisms. 

It  then  seemed  that  the  whole  question  of  the  germicidal 
action  of  the  blood  was  practically  settled.  The  leukocytes 
contain  large  quantities  of  nuclein.  The  blood  serum 
contains  small  quantities  of  the  same  substance.  That  of 
the  serum  comes  from  the  leukocytes  either  in  the  form  of 
a  secretion  or  as  a  result  of  the  breaking-down  of  the  cells, 
and  nuclein  is  a  powerful  germicide.  The  phagocytes 
destroy  bacteria  either  by  engulfing  and  then  digesting 
them,  or  through  the  action  of  the  nuclein  dissolved  in  the 

1  Loc.  cit. 


THE  PHENOMENA  OF  INFECTION  449 

blood.  We  say  all  this  seemed  clear  and  probably  it  is  all 
right,  except  now  it  seems  probable  that  although  nuclein 
is  abundant  in  the  leukocytes  and  present  in  small  amount 
in  plasma  and  serum,  it  is  not  the  germicidal  agent  in 
either.  The  germicidal  agent  in  the  cell  and  that  dissolved 
in  the  plasma  or  serum  are  both  most  likely  ferments,  the 
one  intra-  and  the  other  extracellular,  and  the  two  are  not 
identical.  Metschnikoff's  phagocytic  theory  and  Buchner's 
alexin  theory  are  both  in  a  way  right,  but  whether  the 
germicidal  substance  in  the  serum  is  a  secretion  of  the 
phagocyte  or  a  disintegration  product  of  the  cell  is  an 
interesting  question. 

The  germicidal  constituent  of  blood  serum,  studied  by 
Buchner  and  named  alexin  by  him,  is  inactivated  by 
heating  the  serum  to  55°,  while  the  germicidal  substance 
obtained  by  Kossel  and  Vaughan  from  cells,  and  believed 
by  them  at  the  time  to  be  nucleic  acid,  required  a  tempera- 
ture of  85°  to  render  it  inert.  Evidently  these  must  be  two 
quite  different  bodies,  or  if  the  same  substance,  their  behavior 
under  the  influence  of  temperature  must  be  markedly 
affected  by  the  conditions  under  which  they  were  tested. 
The  researches  of  Schattenfroth1  showed  further  differences 
between  the  intra-  and  extracellular  germicidal  constituents 
of  the  blood.  The  former  has  no  hemolytic  action  on  the 
red  corpuscles  of  other  species,  while  the  latter  may  have. 
The  intracellular  germicide  is  not  affected  by  the  salt 
content  of  the  medium,  retaining  its  activity  in  a  salt-free 
menstruum,  while  the  extracellular  substance  is  inactivated 
by  the  removal  of  salt  from  the  serum  by  dialysis.  Daubler2 
came  to  the  conclusion  that  the  germicidal  constituents 
of  the  serum  and  of  the  leukocytes  are  not  identical,  the 
latter  remaining  active  after  being  heated  to  60°.  He  also 
found  that  the  germicidal  substances  obtained  from  the 
leukocytes  of  different  species  differ  in  measurable  degree 
as  tested  upon  the  same  bacteria.  Many  other  investi- 


1  Archiv  f.  Hygiene,  1897,  xxxi,  1;  ibid.,  1899,  xxxv,  135. 

2  Centralbl.  f.  Bakt.,  1899,  xxv,  129. 
29 


450  PROTEIN  POISONS 

gators  produced  evidence  of  the  fact  that  the  intra-  and 
extracellular  germicidal  constituents  of  the  blood  are  not 
identical,  but  since  the  literature  of  this  subject  has  been 
collected  by  Kling,3  we  will  not  go  into  detail  but  will 
content  ourselves  with  the  reproduction  of  the  summary  as 
given  by  this  author.  It  should  be  stated  that  Petterssen 
designates  the  intracellular  germicidal  constituent  of  leuko- 
cytes and  other  cells  as  "endolysins."  Kling's  conclusions 
from  the  .work  of  others  and  himself  are  stated  substantially 
as  follows:  (1)  The  germicidal  substances  (endolysins) 
of  the  polymorphonuclear  leukocytes  may  be  obtained 
from  the  protoplasm  by  the  following  methods:  (a)  By 
digesting  the  cells  for  half  an  hour  at  50°  in  bouillon.  (6) 
by  extracting  the  cells  with  weak  acid  or  alkali,  or  (c)  by 
alternating  freezing  and  thawing  of  the  cells.  They  cannot 
be  obtained  by  digesting  with  bouillon  at  37°,  nor  with 
physiological  salt  solution,  nor  with  5  per  cent,  "inacti- 
vated" serum.  (2)  As  tested  on  bacillus  subtilis,  the 
endolysin  bears  a  temperature  of  65°  without  recognizable 
effect  on  its  germicidal  action,  and  it  is  not  until  the  tem- 
perature is  increased  to  75°  that  any  such  effect  is  noticed. 
The  endolysins  can,  in  daylight  at  room  temperature,  and 
in  the  dark  at  37°,  be  evaporated  to  dryness,  and  in  this 
state  they  may  be  heated  for  half  an  hour  at  100°  without 
being  destroyed.  The  serum  alexins  may  be  obtained  in 
the  dry  state  in  the  same  manner,  but  when  heated  to  this 
temperature  they  are  inactivated.  The  endolysin  as  tested 
on  the  subtilis  does  not  pass  through  a  Pukall  filter,  while 
the  serum  alexin  does.  The  endolysins  as  tested  on  the 
subtilis,  the  anthrax,  and  the  typhoid  bacillus  are  destroyed 
by  the  Rontgen  ray,  while  the  serum  alexins  are  not.  The 
endolysins  cannot  be  extracted  with  ether,  but  are  not 
injured  by  ether,  while  the  serum  alexins  are  destroyed  by 
ether.  (3)  The  activity  of  an  inactivated  extract  of  the 
leukocytes  of  the  rabbit,  as  tested  on  the  subtilis,  may 
be  restored  by  the  addition  of  a  small  quantity  of  the  same 

3  Zeitsch,  f,  Immunitatsforschung,  1910,  vii,  1. 


THE  PHENOMENA   OF  INFECTION  451 

extract  in  a  fresh  state.  Likewise,  an  inactivated  normal 
serum  of  the  rabbit  or  the  inactive  serum  of  the  guinea-pig 
may  be  complemented  by  the  addition  of  a  small  amount 
of  the  leukocytic  extract  from  the  rabbit  or  guinea-pig, 
respectively.  Furthermore,  an  inactivated  leukocytic 
extract  from  a  guinea-pig  can  be  activated  by  the  addition 
of  a  small  amount  of  the  normal  serum  of  a  rabbit.  (4) 
Extracts  from  the  polymorphonuclear  leukocytes  of  rabbits, 
guinea-pigs,  and  cats  destroy  in  vitro  the  timothy  bacillus, 
the  grass  bacillus  II,  Korn's  acid-fast  bacillus  I,  and  Rubner's 
butter  bacillus.  The  extract  from  rabbits'  leukocytes  has 
a  bactericidal  action  on  the  bacillus  tuberculosis  of  man. 
Extracts  of  rabbit,  guinea-pig,  and  cat  macrophages  do  not 
destroy  these  acid-fast  bacilli  in  vitro.  The  same  is  true  of 
the  extracts  from  the  thymus  gland  of  the  rabbit.  Living 
polymorphonuclear  leukocytes  injected  into  guinea-pigs 
decrease  the  virulence  of  the  human  tuberculosis  bacillus. 
The  leukocytes  of  the  guinea-pig  do  not  have  this  effect. 
These  experiments  do  not  permit  us  to  draw  positive  con- 
clusions concerning  the  action  of  living  macrophages  and 
lymphocytes  on  tubercle  bacilli,  but  it  appears  that  rabbit 
macrophages  may  have  a  protective  action  against  these 
organisms.  (5)  Extracts  from  rabbit,  guinea-pig,  and  cat 
macrophages  have  no  hemolytic  effect  upon  the  erythrocytes 
of  chickens,  goats,  rabbits,  or  guinea-pigs.1 

1  This  is  interesting  in  view  of  the  statement  made  by  Vaughan  (Med. 
News,  December  15  and  22,  1894),  from  which  the  following  quotation  is 
taken:  "On  March  19,  1894,  I  inoculated  rabbits  1,  2,  3,  4,  5,  6,  a  and 
6  with  a  virulent  culture  of  the  (tubercle)  bacillus.  Animals  from  1  to  6 
inclusive  had  had  previous  treatments  with  a  1  per  cent,  solution  of  nucleinic 
acid  as  follows: 

March 9          10         13         14        15       16       17       19 

Amount    of    solution    in 

cubic  centimeter      .      .     0.3       0.5       0.6       0.7         1111 

a  and  b  had  had  no  nuclein.  All  of  the  animals  were  half-g  own,  and  weighed 
respectively:  No.  1,  714  grams;  No.  2,  724  grams;  No.  3,  740  grams;  No. 
4,  729  grams;  No.  5,  647  grams;  No.  6,  614  grams;  a,  709  grams;  b,  705 
grams.  On  July  6,  1894,  I  killed  No.  6,  a  and  b.  No.  6  weighed  at  this 
time  1557  grams.  I  found  a  nodule  the  size  of  a  pea  at  the  point  of  inocu- 
lation. In  all  other  respects  this  animal  was  normal,  I  could  find  no 


452  PROTEIN  POISONS 

The  above-mentioned  facts,  ascertained  by  experimental 
study,  have  been  cited  to  show  that  the  existence  of  both 
intra-  and  extracellular  germicidal  substances  in  the  blood 
has  been  demonstrated.  These  substances  have  been 
called  alexins,  antibodies,  and  by  other  names.  It  seems 
to  us  that  at  present  they  should  be  classed  as  ferments. 
As  has  been  said,  we  do  not  know  much  about  ferments, 
but  it  is  evident  that  these  bodies  have  a  lytic  action.  They 
break  up  complex  molecules  into  simpler  bodies.  Their 
primary  function  seems  to  be  to  supply  the  cells  which 
elaborate  them  with  food.  In  doing  this  they  also  protect 
the  cells,  to  which  they  belong,  by  the  destruction  of  harmful 
bodies  both  particulate  and  formless,  both  animate  and 
inanimate.  The  digestive  ferments  of  our  alimentary 
canals  serve  the  same  double  purpose.  Any  unbroken 
foreign  protein  having  found  its  way  into  the  blood  is  a 
poison,  but  in  the  alimentary  canal  it  is  broken  up  and 
prepared  as  a  food  for  the  body  cells.  Every  living  cell 
has  such  a  ferment  or  such  ferments.  Their  presence  and 

bacilli  in  the  nodule,  which  was  rubbed  up  with  beef-tea  and  injected  into 
the  abdominal  cavity  of  guinea-pig  No.  186,  weighing  385  grams.  On 
October  10,  1894,  I  killed  this  pig,  and  found  a  nodule  the  size  of  a  pea  at 
the  point  of  inoculation.  Three  small  tubercles  were  found  in  the  peri- 
toneum; the  omentum  and  liver  were  filled  with  tuberculous  nodules.  One 
testicle  was  tuberculous.  This  is  an  interesting  case,  showing  that  the 
germ,  which  had  not  spread  in  the  rabbit,  had,  when  transferred  to  the 
more  susceptible  guinea-pig,  induced  a  widespread  tuberculosis 

"Rabbit  a  weighed  1030  and  b  1100  grams.  In  both,  nodules  as  large 
as  filberts  were  found  at  the  point  of  inoculation,  and  smaller  nodules  in 
the  omentum.  On  October  10,  I  killed  No.  1,  weight,  2134  grams.  This 
animal  was  found  to  be  wholly  free  from  tuberculosis.  On  October  4,  I 
killed  No.  2,  weight,  2150  grams,  which  was  found  perfectly  normal.  No.  3 
was  found  dead  October  2.  Postmortem  examination  showed  a  pear-shaped 
tumor  in  the  omentum.  This  was  three  inches  long  and  one  and  one-half 
inches  in  diameter  at  the  base.  It  consisted  of  three  cysts,  which  contained 
very  fetid  pus,  in  which  were  found  a  short  bacillus  and  a  large  micrococcus. 
There  was  no  evidence  of  tuberculosis.  No.  4  was  killed  October  10, 
weight  1990  grams.  I  found  a  small  nodule  at  the  point  of  inoculation. 
This  was  not  attached  to  the  abdominal  wall,  but  was  in  the  connective 
tissue,  between  the  skin  and  the  muscle.  I  could  find  no  germ.  In  all  other 
respects  this  rabbit  was  normal.  No.  5  was  killed  October  10,  weight  2000 
grams,  and  found  perfectly  normal 

"These  experiments  indicate  that  rabbits  may  be  rendered  immune  to 
tuberculosis  by  previous  treatment  with  yeast  nucleinic  acid  " 


THE  PHENOMENA  OF  INFECTION  453 

activity  distinguish  living  from  non-living  matter.  We 
have  taken  the  leukocyte  as  an  illustration,  but  the  bac- 
terium is  also  supplied  with  its  ferments,  some  of  which  are 
intra-  while  others  are  extracellular.  We  do  not  know  that 
all  cells  elaborate  both  kinds  of  ferments,  but  all  have  at 
least  one  kind. 

Before  proceeding  further  it  may  be  well  to  call  special 
attention  to  some  of  the  properties  of  these  ferments.  The 
extracellular  ferments  are  diffusible.  They  not  only  pass 
out  of  the  cells  in  which  they  are  prepared,  but  they  diffuse 
more  or  less  widely  through  the  medium  which  surrounds 
the  cell.  This  suggests  that  in  molecular  structure  they 
are  relatively  simple.  At  least  some  of  them  may  pass 
through  membranes  and  collodion  sacs,  as  is  shown  by  the 
fact  that  bacteria  and  other  proteins  enclosed  in  such 
receptacles  and  left  in  a  body  cavity  are  destroyed. 
The  extracellular  ferments  are,  in  part  at  least,  filterable, 
passing  with  more  or  less  readiness  through  porcelain. 
In  their  activities  they  are  easily  affected  by  modification 
in  the  medium  through  which  they  diffuse.  The  alexin 
of  the  blood  serum  is  highly  sensitive  to  the  salt  content 
of  the  serum,  and  by  variations  in  this  the  activity  of  the 
ferment  may  be  hastened,  lowered,  or  wholly  arrested. 
The  same  is  true  of  bacterial  ferments.  In  one  species  of 
animal  a  given  bacterium  multiplies  with  great  rapidity; 
in  another  it  grows  slowly,  while  in  a  third  it  cannot  grow 
at  all.  There  are  like  variations  in  individuals  of  the  same 
species.  The  extracellular  ferments,  at  least  some  of  them, 
are  susceptible  to  slight  changes  in  temperature.  It  is 
believed  that  every  ferment  has  its  optimum  temperature, 
but  the  range  in  which  continued  activity  is  possible  is 
narrow  with  some  and  relatively  wide  with  others. 

The  intracellular  ferments  are  non-diffusible,  or  at  least 
less  diffusible  than  the  extracellular.  They  remain  in  the 
cells  in  which  they  are  elaborated.  They  cannot  be  extracted 
from  the  cell  by  indifferent  solvents.  As  a  rule,  they  can 
be  obtained  from  the  cell  only  after  partial  or  complete 
destruction  of  the  cell.  Some,  probably  most,  are  best 


454 


PROTEIN  POISONS 


extracted  from  the  cell  with  dilute  alkali,  while  others  are 
best  obtained  by  dilute  acid.  In  either  case  the  reagent 
must  not  be  strong  enough  to  destroy  the  ferment  itself. 
They  are  non-filterable,  or  pass  through  filters  slowly  and 
imperfectly.  We  suspect  that  their  molecular  structure 
is  relatively  complex,  or  that  they  are  more  colloidal  than 
the  extracellular  ferments.  Under  natural  conditions  the 
intracellular  ferments  act  only  on  those  bodies  which  are 
taken  into  the  cell.  The  inclusion  of  bacteria  by  phago- 
cytes is  essential  to  the  digestion  of  the  former  by  the 
intracellular  ferment  of  the  latter.  This  is  a  phenomenon 
which  may  be  seen,  but  cell  permeation  by  foreign  bodies 
is  certainly  necessary  before  such  bodies  can  be  acted  upon 
by  the  intracellular  ferments,  and  occurs  with  soluble 
proteins  as  well  as  with  participate  ones.  The  intracellular 
ferment  bears  a  wider  variation  in  temperature,  and  is  not 
so  easily  and  delicately  influenced  by  variations  in  the 
composition  of  the  medium  in  which  the  cell  exists.  So 
far  as  we  know  the  intracellular  ferments  do  not  diffuse 
from  living  cells.  They  are,  however,  recognizable  in  the 
fluids  of  abscess  cavities  as  the  leukocytes  disintegrate. 
We  are  of  the  opinion  that  they  are  essential  constituents 
of  the  chemical  structure  of  cells.  The  reason  for  this 
belief  will  be  developed  later.  The  extracellular  ferments 
may  be  regarded  as  secretions  of  cells.  Much  has  been 
written  about  cellular  and  humoral  theories.  In  our  opinion 
every  living  thing  has  a  chemical  structure,  which  we  may 
designate  as  a  cell  if  we  wish,  understanding  that  a  cell  is 
not  necessarily  something  that  can  be  seen,  and  that  it 
may  possess  widely  different  degrees  of  lability,  but  we  are 
quite  certain  that  there  is  no  ferment  which  is  not  the 
product  of  life  processes.  We  have  been  somewhat  sur- 
prised to  find  it  stated  that  our  own  theory  of  protein 
sensitization  or  anaphylaxis  is  a  humoralistic  doctrine. 

All  ferments  are  products  of  life  processes,  and  all  life 
processes  are  more  or  less  responsive  to  outside  influences, 
to  change  in  environment.  In  our  opinion  the  most  valuable 
fact  that  we  have  learned  in  the  study  of  protein  sensi- 


THE  PHENOMENA  OF  INFECTION  455 

tization  is  that  life  processes  manifested  through  ferment 
action  are  modified  and  may  be  modified  at  will  by  changes 
in  environment.  The  blood-serum  and  organ  extracts  of 
normal  guinea-pigs  do  not  digest  egg-white,  but  these 
fluids  from  an  animal  sensitized  to  this  protein  do  have 
this  action.  The  virus  of  smallpox  is  pathogenic  to  the 
man  who  has  never  had  smallpox,  and  has  not  been  vacci- 
nated, but  to  the  man  who  has  had  the  disease  or  been 
properly  vaccinated  the  virus  of  smallpox  is  non-pathogenic. 
We  explain  this,  and  in  our  opinion,  the  experiments  of 
Pirquet  have  so  demonstrated,  that  this  is  due  to  the  fact 
that  the  ferments  of  the  man's  body  cells  have  been  so 
influenced  by  the  disease  or  by  vaccination  that  they  have 
acquired  a  new  function — that  of  digesting  and  thus 
destroying  the  virus  of  the  disease.  If  this  explanation  be 
true,  it  opens  up  a  wide  field  for  the  possible  extension  of 
the  beneficial  effects  of  preventive  treatment. 

There  is  another  point  of  difference  between  intracellular 
and  extracellular  ferments,  which  is  of  the  greatest  impor- 
tance in  a  study  of  the  phenomena  of  infection.  The  extra- 
cellular ferments  are  comparable  to  those  of  the  digestive 
juices  of  the  alimentary  tract  in  the  higher  animals.  They 
roughly  prepare  foods  for  the  cells.  Their  function  is 
solely  a  lytic  one.  They  break  up  complex  proteins  into 
simpler  bodies,  but  the  products  thus  formed  are  not, 
without  further  treatment,  ready  to  be  built  into  the  struc- 
ture of  the  cell.  Proteins  in  the  medium  are  rendered 
soluble  by  the  extracellular  ferments.  They  are  so  altered 
that  they  may  be  taken  into  the  cell,  but  they  are  not  so 
patterned  that  they  are  ready  to  be  built  into  the  structure. 
They  are  fitted  for  absorption,  but  are  not  ready  for  assimi- 
lation. The  extracellular  ferments  are  in  a  sense  destructive 
agents.  They  break  down  complex  molecules  into  simpler 
structures.  The  intracellular  ferments  are  constructive. 
They  are  cell  builders.  They  shape  the  material  brought 
them  and  fit  it  into  place.  They  build  up  specific  proteins. 
They  convert  the  raw  material  brought  them  into  specific 
proteins,  bacterial,  vegetable,  or  animal.  This  does  not 


456  PROTEIN  POISONS 

mean  that  the  intracellular  ferments  have  no  cleavage 
action.  They  chip  the  rough  stone  so  that  it  fits  in  at  the 
right  place.  It  is  by  virtue  of  their  activity  or  through  their 
agency  that  cells  grow  and  multiply.  In  case  of  an  infec- 
tious disease  the  intracellular  ferment  of  the  infecting 
organism  during  the  period  of  incubation  converts  man's 
proteins  into  bacterial  proteins,  and  continues  to  do  this 
with  more  or  less  success  during  the  course  of  the  disease. 
This  seems  to  be  accomplished  in  pome  diseases,  at  least, 
like  typhoid  fever,  without  any  marked  disruption  of  the 
cells  of  the  man's  body.  The  bacteria  multiply  rapidly 
during  the  period  of  incubation,  and  at  this  time  the  man  is 
unconscious  of  the  fact  that  his  body  is  serving  as  a  culture 
flask.  We  must  conclude  from  this  that  the  conversion  of 
human  proteins  into  typhoid  proteins  in  the  growth  of  the 
infecting  agent  is  not  accompanied  by  the  liberation  of  the 
poisonous  group  in  the  protein  molecule.  This  group, 
probably  attached  to  other  groups,  or  as  a  constituent  of 
a  more  complex  group,  is  used  in  the  construction  process. 
The  poisonous  group  is  common  to  all  proteins.  The  syn- 
thesis of  specific  proteins  from  other  specific  proteins  is 
accomplished  without  the  -liberation  of  the  poisonous 
portion.  It  is  one  of  the  building  stones,  and  changes  in 
specificity  do  not  occur  in  this,  but  in  the  secondary  or 
characteristic  groups.  This  is,  in  our  opinion,  the  explana- 
tion of  the  fact  why  incubation — a  period  of  rapid  repro- 
duction in  the  infecting  agent — proceeds  without  any 
recognizable  disturbance  in  the  health  of  the  host.  The 
typhoid  bacillus  therefore  does  not  feed  upon  the  cells  of 
the  man's  body,  but  upon  the  formless,  soluble  proteins. 
Cell  building  is  accompanied  by  the  absorption  of  the 
poisonous  group  in  the  proteins  serving  as  food.  However, 
when  the  body  cells  become  sensitized  and  elaborate  a 
ferment  which  breaks  down  the  bacterial  cells,  the  poisonous 
group  in  the  proteins  of  the  latter  is  set  free,  and  it  is  the 
effect  of  this  poison  that  develops  the  symptom  complex 
of  the  disease.  The  symptoms  of  one  infectious  disease 
differ  from  those  of  another  largely  according  to  the  organ 


THE  PHENOMENA  OF  INFECTION  457 

or  tissues  in  which  the  infecting  agent  is  located.  In  acute 
miliary  tuberculosis  and  in  typhoid  fever,  both  conditions 
arising  from  a  bacteremia  caused  by  different  organisms, 
the  symptoms  are  only  too  frequently  identical,  and  it  is 
only  by  bacteriological  methods,  a  suggestive  history,  or 
the  finding  of  a  preexisting  tuberculous  focus  in  some  part 
of  the  body  that  a  differential  diagnosis  may  be  reached. 
A  cholecystitis  is  the  same,  not  only  in  symptomatology, 
but  frequently  in  gross  pathology  as  well,  whether  the 
infecting  organism  be  the  pneumococcus,  the  streptococcus, 
the  colon,  or  the  typhoid  bacillus.  The  most  skilful  diag- 
nostician cannot  tell  from  the  symptoms  alone  the  specific 
bacterial  cause  of  a  meningitis. 

During  the  period  of  incubation  of  an  infectious  disease, 
the  infecting  organism  supplies  the  ferment,  the  body  pro- 
teins constitute  the  substrate,  the  process  is  essentially 
constructive,  no  poison  is  set  free,  and  there  are  no  recog- 
nizable clinical  symptoms.  During  the  active  progress  of  an 
infectious  disease,  the  body  cells  supply  the  ferment,  the 
infecting  organism  constitutes  the  substrate,  the  process  is 
essentially  destructive,  the  protein  poison  is  set  free,  the 
symptoms  of  disease  appear-  and  life  is  placed  in  jeopardy. 

Our  work  seems  to  show  that  the  body  cells,  when  over- 
whelmed with  a  foreign  protein  of  the  blandest  kind — such 
as  egg-white — may  fail  to  function  and  death  may  result. 
There  is  no  reason  for  suspecting  that  in  these  cases  there 
is  any  cleavage  of  the  foreign  protein  or  the  liberation  of 
any  poison.  The  body  cells  are  simply  clogged  with  the 
foreign  protein  and  fail  to  function.  We  are  not  sure  that 
this  phenomenon  has  any  parallel  in  the  infectious  diseases. 
There  is,  however,  something  closely  related  to  it  in  cholera 
infantum,  cholera  nostras,  and  Asiatic  cholera. 

We  have  already  referred  to  the  fact  that  ferments  may 
be  modified  in  their  activities.  These  modifications  may 
be  so  radical  that  it  is  generally  believed  that  cells  may  be 
trained,  as  it  were,  to  develop  new  ferments.  There  can 
be  no  doubt  that  change  in  environment  does  alter  activity 
as  manifested  through  the  ferments.  As  we  have  stated, 


458  PROTEIN  POISONS 

it  seems  to  be  a  biological  law  that  when  a  living  cell  is 
brought  in  contact  with  or  permeated  by  a  foreign  protein, 
it  tends  to  furnish  a  ferment  which  will  digest  and  destroy 
the  foreign  body.  The  ferments  of  the  cells  of  man's  body 
may  be  modified  or  new  ones  developed  by  (a)  disease, 
(6)  vaccination,  and  (c)  sensitization.  Many  of  the  infec- 
tious diseases  give  immunity  to  subsequent  exposure.  In 
some  of  the  chronic  infectious  diseases  the  altered  behavior 
of  the  body  cells  to  the  infecting  agent  is  evident  even 
while  the  disease  continues. 

That  the  tuberculous  animal  behaves  differently  from 
the  non-tuberculous  on  receiving  injections  of  the  tuber- 
culin protein,  whether  it  be  in  the  form  of  the  living  bacillus, 
in  dead  cells,  or  in  solution,  has  been  abundantly  demon- 
strated. Before  Koch  gave  us  tuberculin,  Arloing  and 
Courmont  had  come  to  the  conclusion  that  the  tubercle 
bacillus  produces  soluble  substances  which  reduce  the 
natural  resistance  of  the  body  and  render  it  more  susceptible 
to  reinfection.  This  corresponds  closely  with  the  first 
impression  made  by  observation  of  the  phenomena  of  ana- 
phylaxis;  the  impression  that  led  Richet  to  select  this 
term.  In  1891,  Koch  described  a  perfect  example  of  protein 
sensitization  as  we  understand  it  today.  He  stated  that 
when  a  healthy  guinea-pig  is  inoculated  with  the  living 
tubercle  bacillus  there  is  no  change  at  the  site  of  inoculation 
until  from  ten  to  fourteen  days  later,  when  a  hard  lump 
forms,  finally  opens  and  ulcerates,  and  continues  until  the 
animal  dies.  On  the  other  hand,  when  a  tuberculous  guinea- 
pig  is  Inoculated  with  the  living  bacillus,  on  the  second  or 
third  day  a  lump  forms,  soon  becomes  necrotic,  falls  out, 
ulcerates  for  a  time,  and  finally  heals  without  any  infection 
of  the  neighboring  lymph  glands.  In  1897  Trudeau  observed 
that  when  healthy  rabbits  receive  injections  of  virulent 
cultures  in  the  eye,  there  is  little  to  be  seen  for  about  four- 
teen days,  when  with  increasing  vascularity  tubercles  form 
in  the  iris,  after  which  inflammation  extends  and  the  eye 
is  practically  destroyed  within  from  six  to  eight  weeks. 
Like  treatment  of  tuberculous  rabbits  develops  an  iritis 


THE  PHENOMENA  OF  INFECTION  459 

within  from  two  to  five  days,  but  at  the  end  of  the  second 
or  third  week,  at  a  time  when  the  controls  begin  to  develop 
destructive  changes,  the  inflammation  begins  to  subside. 
Later  studies  have  confirmed  and  amplified  these,  and  it 
has  been  found  that  death  may  be  induced  within  twenty- 
four  hours  by  injecting  a  large  amount  of  a  living  culture 
into  a  tuberculous  animal. 

The  same  difference  between  healthy  and  tuberculous 
animals  has  been  observed  in  their  response  to  injections 
of  dead  cultures  of  the  tubercle  bacillus.  The  first  observa- 
tion along  this  line,  so  far  as  we  know,  was  made  by  Strauss 
and  Gamaleia,  who  found  that  when  large  numbers  of  dead 
tubercle  bacilli  are  injected  into  tuberculous  animals  death 
results,  while  similar  amounts  are  without  immediate  effect 
upon  healthy  animals. 

When  we  come  to  tuberculin,  every  phase  of  its  action 
or  its  failure  to  act  is  explainable  on  the  ground  that  the 
tuberculous  animal  is  a  sensitized  one.  Koch  found  that 
0.5  gram  of  his  preparation  killed  tuberculous  guinea-pigs, 
and  induced  no  symptoms  in  healthy  ones.  A  fraction  of 
1  mg.  may  cause  marked  symptoms  in  a  tuberculous  man, 
while  many  times  this  amount  is  borne  easily  by  a  healthy 
man.  The  inflammatory  reaction  about  local  tuberculous 
lesions  caused  by  injections  of  tuberculin  is  explained  by 
the  fact  of  the  high  degree  of  sensitization  in  their  localities, 
and  the  cleavage  of  the  bacilli.  The  ophthalmic,  cutaneous, 
subcutaneous,  and  intravenous  tests  with  tuberculin  are 
all  typical  sensitization  reactions.  Even  in  the  failure  to 
respond  to  tuberculin  seen  in  advanced  tuberculosis  we  have 
the  condition  known  as  anti-anaphylaxis,  which  simply 
means  that  the  anaphylactic  ferment  is  partially  exhausted 
by  the  large  amount  of  material  supplied  by  the  bacilli 
in  the  body. 

There  is  a  second  factor  in  the  failures  of  advanced  cases 
of  tuberculosis  to  respond  to  the  tuberculin  test,  which  has 
been  generally  overlooked,  but  to  which  we  have  already 
referred  in  our  discussion  of  so-called  anti-anaphylaxis. 
This  is  the  fact  that  in  such  cases  the  body  is  saturated  with 


460  PROTEIN  POISONS 

the  products  of  the  digestion  of  tuberculoproteins.  It  is  a 
well-established  fact  that  the  accumulation  of  fermentation 
products  retards  and  finally  arrests  the  fermentative  process. 
The  instance  in  point  is  a  perfect  illustration  of  this  law  of 
fermentation.  It  must  be  evident  from  this  how  unscientific 
it  is  to  treat  advanced  tuberculosis  with  tuberculin.  It 
has  been  argued  that  the  tuberculin  reaction  is  not  an 
example  of  sensitization,  because  as  the  treatment  proceeds, 
larger  and  larger  doses  of  tuberculin  are  necessary  in  order 
to  induce  the  reaction,  as  shown  by  the  development  of 
fever.  To  anyone  who  has  followed  the  evidence  that  we 
have  given  so  far,  the  explanation  must  be  plain.  It  lies 
in  two  facts,  of  each  of  which  we  think  that  we  have  given 
abundant  proof.  In  the  first  place  we  have  shown  that  a 
certain  degree  of  tolerance  for  the  protein  poison  is  easily 
and  quickly  established.  In  the  second  place,  the  accumu- 
lation of  fermentation  products  retards  fermentation. 
Tuberculosis,  in  most  instances  at  least,  begins  as  a  strictly 
local  infection.  This  is  true  even  when  the  first  recognition 
of  it  has  been  in  the  acute  miliary  form.  There  has  been 
a  previous  focal  infection.  Only  those  body  cells  in  the 
immediate  vicinity  of  the  infection  are  sensitized,  and  only 
these  supply  a  ferment  capable  of  digesting  the  tuberculo- 
protein.  It  may  well  be  that  in  this  stage  benefit  may  be 
secured  by  the  proper  use  of  tuberculin,  which  may  act  as 
a  sensitizer,  and  develop  more  of  the  ferment  to  split  up  and 
destroy  the  tubercle  bacilli.  It  should  be  always  borne  in 
mind  that  tuberculin  contains  a  poison  and  should  be  used 
with  caution. 

There  is  another  line  of  evidence  that  in  tuberculosis  there 
is  a  condition  of  specific  protein  sensitization.  This  is  to  be 
found  in  the  fact  that  this  disease  is  much  more  deadly  in 
lands  and  among  people  who  have  recently  come  under  its 
influence  than  it  is  where  it  has  prevailed  for  many  genera- 
tions. In  other  words,  the  widespread  and  long-continued 
existence  of  the  disease,  slowly,  and  at  the  cost  of  much 
sickness,  and  many  deaths,  brings  a  certain  degree  of 
immunity.  The  readiness  with  which  the  North  American 


THE  PHENOMENA  OF  INFECTION  461 

Indian  has  succumbed  to  this  disease  is  a  striking  illustration, 
and  Calmette  has  recently  collected  additional  evidence 
on  this  point.  He  states  that  tuberculosis  is  being  widely 
disseminated  among  peoples  who  have  until  recently  been 
free  from  it.  The  world-wide  wanderings  of  the  white  man 
are  carrying  the  disease  to  every  people,  from  the  Laplander 
and  Esquimaux  of  the  Arctics  to  the  negroes  and  Malays 
of  the  tropics.  Iceland,  the  Faroe  Islands,  and  the  steppes 
of  Russia  are  being  infected,  and  in  these  new  regions 
tuberculosis  exists  in  its  most  speedily  fatal  forms.  The 
same  author  points  out  that  recently  discovered  methods 
for  the  recognition  of  this  disease,  even  in  latent  states, 
shows  that  among  Europeans  not  more  than  7  or  8  per  cent, 
reach  more  than  twenty  years  of  age  without  receiving  the 
infection.  Those  who  survive  the  first  infection  become 
more  or  less  immune,  and  after  that  develop,  when  they  do 
acquire  the  disease,  the  more  chronic  forms. 

Romer1  concludes  that  the  less  widely  tuberculosis  is 
distributed  among  a  people  the  greater  is  the  case  mortality, 
and  the  wider  the  distribution  the  smaller  is  the  case 
mortality. 

Still  another  fact  of  importance  is  that  the  most  speedily 
fatal  forms  of  tuberculosis,  such  as  the  miliary  and  menin- 
geal,  are  more  frequent  among  children  than  among  adults. 

There  is  another  matter  of  much  importance  in  this 
connection  which  we  must  discuss.  We  have  found  the 
tubercle  bacillus  highly  resistant  to  lytic  agents,  and  it 
appears  that  its  long  experience  as  a  parasite  has  led  it  to 
protect  itself  with  deposits  of  wax  and  fat,  but  proteolytic 
enzymes  digest  the  most  firm  proteins.  Friedberger  has 
found  that  at  least  some  strains  of  this  bacillus  are  digested 
by  the  serum  of  healthy  guinea-pigs,  and  the  researches  of 
Markl,  Bail,  and  Kraus  and  his  students  have  shown  that 
tubercle  bacilli  placed  in  the  peritoneal  cavity  of  tuberculous 
animals  respond  to  Pfeiffer's  reaction.  Some  strains  are 
dissolved  in  the  peritoneum  of  healthy  guinea-pigs,  but 

1  Beitrage  z.  klinik  d.  Tuberk.,  1912,  xxii,  301. 


462  PROTEIN  POISONS 

dissolution  occurs  more  promptly  and  more  completely  in 
the  peritoneum  of  a  tuberculous  animal.  The  healthy  animal 
may  have  to  depend  upon  its  phagocytes  to  combat  the 
invading  bacillus,  but  the  tuberculous  animal  supplies  a 
specific  proteolytic  enzyme,  and  to  this  the  fresh  invader 
succumbs. 

Nature  is  slowly  immunizing  the  white  man  to  tuber- 
culosis, and  the  question  arises  whether  or  not  the  process 
employed  by  Nature  -can  be  aided  in  any  way.  There  is 
before  the  medical  profession  at  this  time  no  greater  question 
than  this:  Is  it  possible  to  aid  in  eradicating  tuberculosis 
by  vaccination?  As  Romer  says,  the  problem  of  securing 
immunity  to  tuberculosis  with  a  non-infective  virus  is  of 
great  practical  importance,  and  recent  work  brings  the 
possibility  of  doing  this  more  and  more  to  the  front.  What 
we  need  is  a  vaccine.  Various  methods  of  modifying  the 
tubercle  bacillus  so  that  it  could  be  used  as  a  vaccine  have 
been  tried.  The  bovo  vaccine  of  Von  Behring  was  tried, 
but  the  increased  resistance  given  by  it  was  found  to  be  of 
short  duration.  Attempts  to  reduce  its  virulence  by  age, 
heat,  chemicals,  and  by  submitting  it  to  ultraviolet  and 
other  rays  and  emanations  have  been  made.  What  we 
need  is  a  tubercle  protein  sensitizer.  It  should  be  soluble, 
and  it  should  be  free  from  the  poisonous  group  in  the  protein 
molecule.  In  our  opinion  the  nearest  approach  to  this 
desired  substance  is  the  non-poisonous  portion  of  the  tubercle 
protein.  So  far  we  have  not  been  able  to  secure  a  uniform 
product.  Some  preparations  seem  to  fill  every  requirement. 
They  sensitize  animals  to  the  unbroken  bacillus,  dead  or 
alive,  and  in  surface  tuberculous  lesions  they  cause  inflam- 
mation about  the  tuberculous  area,  and  we  have  seen 
the  tuberculous  tissue  slough  off  and  complete  recovery 
result;  but  other  preparations  made  from  the  same  cellular 
substance  by  the  same  method  seem  inert.  We  have  had 
similar  difficulties  with  the  sensitizing  groups  from  other 
proteins.  Some  preparations  from  egg-white  sensitize  to 
unbroken  egg-white,  while  others  seem  wholly  without 
effect,  and  still  all  are  prepared  from  the  same  material  and 


THE  PHENOMENA  OF  INFECTION  463 

in  the  same  way.  Evidently  the  sensitizing  group  in  the 
protein  molecule  is  a  highly  labile  body  and  susceptible 
to  influences  which  so  far  we  have  not  been  able  to  recog- 
nize. We  have  no  difficulty  in  obtaining  the  poisonous 
group  uniformly,  but  it  is  otherwise  with  the  sensitizing  body. 
Further  work  along  this  line  is  needed,  and  if  an  efficient 
and  uniformly  reliable  sensitizer  for  the  tuberculous  protein, 
free  from  the  poisonous  group,  can  be  secured,  all  children 
should  be  vaccinated  for  tuberculosis;  then  with  protection 
against  natural  infection  the  restriction  of  tuberculosis 
will  be  as  completely  under  man's  control  as  is  that  of 
smallpox.  It  should  be  clearly  understood  that  the  pro- 
tection afforded  by  vaccination  is  relative  and  not  absolute. 
The  studies  inaugurated  by  Wright  have  demonstrated 
that  vaccination  is  of  service  not  only  in  prevention,  but 
also  in  cure.  Bacteria  and  protozoa  are  particulate,  and  in 
many  diseases  they  are  confined  to  limited  localities.  As 
we  have  seen,  sensitization  may  also  be  local.  No  body 
cell  is  sensitized  against  a  foreign  protein  until  the  latter 
comes  in  contact  with  the  former  and  penetration  of  the 
body  cell  is  probably  essential  to  the  most  efficient  sensi- 
tization. The  microorganisms  of  acne  are  located  in  the 
cutaneous  tissue,  and  being  particulate  and  not  in  solution, 
the  area  sensitized  by  them  is  small,  if  there  be  any  sensiti- 
zation at  all.  By  vaccine  therapy  the  area  of  sensitization  is 
greatly  extended  and  the  amount  of  lytic  agent  formed  and 
made  available  is  greatly  increased.  This  being  in  solution 
and  diffusible,  digests  and  destroys  the  bacteria  located  in 
the  skin.  The  same  is  true  of  the  treatment  of  localized 
tuberculosis,  or  of  any  other  localized  infectious  disease. 
In  vaccine  therapy,  as  in  vaccination,  the  great  need  is  for 
soluble  sensitizers  free  from  poisonous  content.  When  these 
are  secured,  and  not  until  then,  we  may  develop  a  vaccine 
therapy  along  scientific  lines,  and  expect  to  secure  important 
results  with  it.1 


1  The  following  pages  are  taken  with  but  little  change  from  an  article  by 
Vaughan,  Jr.,  in  "International  Clinics." 


464  PROTEIN  POISONS 

The  importance  of  sensitization  as  a  factor  in  the  case  of 
tuberculosis  is  evident  in  the  widespread  use  of  tuberculin 
as  a  diagnostic  measure.  The  various  reactions  of  the 
body  to  tuberculin,  whether  they  occur  as  the  general 
reaction  following  subcutaneous  injections,  or  as  the  more 
local  reaction  following  the  vaccination  of  the  skin  with 
tuberculin,  the  application  of  a  tuberculin  containing 
ointment  to  the  skin,  or  the  instillation  of  tuberculin  into 
the  conjunctival  sac,  are  all  evidences  of  the  sensitization 
of  the  body  of  the  tuberculous  individual  to  tuberculin. 
Thus,  when  a  small  amount  of  tuberculin  is  injected  into 
the  fluids  and  tissues  of  a  normal  individual,  no  effects  are 
noticeable,  since  the  enzyme  which  causes  proteolysis  of 
tuberculin  is  not  present  in  the  body.  When,  however,  the 
same  amount  of  tuberculin  is  injected  into  the  tuberculous 
individual,  it  practically  corresponds  to  a  second  injection 
of  this  foreign  protein.  The  enzyme  present  in  the  body  of 
the  tuberculous  individual  attacks  the  tuberculin,  liberating 
the  poisonous  cleavage  products,  which  in  turn  give  rise 
to  the  well-known  symptom-complex  designated  as  the 
tuberculin  reaction.  In  addition  to  the  general  symptoms, 
such  as  fever,  which  accompany  the  presence  of  protein 
poisoning  within  the  body,  poisonous  proteins  have  a 
decidedly  irritant  local  effect  upon  the  tissues  with  which 
they  are  brought  directly  in  contact.  This  is  seen  in  the 
hyperemia  and  inflammation  of  the  peritoneum  in  cases  of 
infection  within  the  abdominal  cavity,  and  is  also  evidenced 
by  the  local  reaction  of  inflammatory  type  following  the 
application  of  tuberculin  to  the  mucous  membrane  or  the 
abraded  skin  of  the  tuberculous  individual. 

Sensitization  to  tuberculin  may  be  either  local  or  general 
in  type,  as  is  quite  evident  to  anyone  who  has  employed 
the  conjunctival  test  as  a  means  of  diagnosis  in  tuberculous 
disease.  This  test  consists  in  the  application  of  a  1  per  cent, 
solution  of  specially  prepared  tuberculin  to  the  conjunctival 
sac.  The  reaction  following  this  method  of  applying  the 
tuberculin  test  may  be  divided  into  two  distinct  types, 
the  first  of  which  we  may  call  the  reaction  of  general  sensi- 


THE  PHENOMENA  OF  INFECTION  465 

tization,  or  the  tuberculous  reaction,  in  contradistinction 
to  the  second,  or  the  reaction  of  local  sensitization. 

When  a  solution  of  tuberculin  is  applied  to  the  conjunc- 
tival  sac  of  a  tuberculous  individual,  no  changes  are  usually 
noticed  for  an  interval  varying  from  six  to  forty-eight 
hours.  At  the  end  of  this  time  there  is  a  slight  smarting 
or  gritty  sensation  complained  of,  the  patient  often  referring 
to  it  as  a  sensation  of  sand  in  the  eye.  The  examination  of 
the  conjunctiva  at  this  time  reveals  a  reddening  and  swelling 
of  the  mucous  membrane  of  the  lower  lid  and  the  caruncle. 
This  inflammatory  reaction  gradually  increases  in  intensity 
until  from  ten  to  fifteen  hours  have  elapsed,  at  which  time 
it  has  usually  reached  a  maximum,  and  after  which  a  gradual 
recession  occurs,  until  at  the  end  of  from  two  to  four  days, 
occasionally  after  a  longer  interval,  the  conjunctiva  has 
again  regained  its  normal  appearance.  At  the  height  of 
the  reaction,  and  on  awakening  in  the  morning,  it  is  not 
uncommon  to  observe  a  slight  fibrinous  or  fibrinopurulent 
exudate  accompanying  the  inflammatory  reaction. 

When  a  solution  of  tuberculin  is  applied  to  the  eye  of  a 
normal  individual,  no  reaction  is  obtained.  If,  however,  a 
second  instillation  is  made  in  the  same  eye  after  an  interval 
of  seven  days,  a  reaction  will  be  observed  in  a  large  propor- 
tion of  cases.  This  reaction  is  quite  distinct  from  that 
previously  described  as  occurring  in  the  eye  of  the  tuber- 
culous individual.  The  reaction  is  rapid  in  appearance, 
explosive  in  type,  and  subsides  with  great  rapidity.  Thus, 
it  is  not  rare  to  find,  as  a  result  of  a  second  instillation,  within 
from  three  to  four  hours  after  the  application,  a  highly 
inflamed  conjunctiva  associated  with  considerable  chemosis 
of  the  lids  and  a  profuse  purulent  discharge.  The  symp- 
toms, however,  in  spite  of  their  severity,  rapidly  subside. 
These  differences  in  type  of  reaction  find  a  satisfactory 
explanation  if  we  consider  the  fact  that  in  the  tuberculous 
reaction  we  are  dealing  with  what  may  be  termed  a  phenom- 
enon of  general  sensitization.  In  this  case  the  cleavage 
of  the  tuberculin  introduced  within  the  conjunctival  sac 
is  brought  about  through  the  action  of  the  proteolytic 
30 


466  PROTEIN  POISONS 

enzyme  which  has  been  developed  in  the  body  of  the  tuber- 
culous individual  as  a  result  of  his  disease.  During  this 
cleavage  certain  poisons  are  liberated  which  act  as  irritants 
to  the  conjunctival  mucous  membrane,  and  the  degree  of 
irritation  present  will  be  directly  proportionate  to  the  amount 
of  toxic  cleavage  product  present  at  a  given  time.  However, 
the  amount  of  cleavage  product  present  at  a  given  time 
will  be  determined  by  the  rate  of  proteolysis,  which  depends 
in  turn  upon  the  quantity  of  proteolytic  enzyme  directly 
available.  Since  this  enzyme  is  available  only  in  such 
proportions  as  may  be  present  in  the  circulating  fluids  of  the 
conjunctiva,  it  necessarily  follows  that  only  a  small  amount 
can  be  operative  at  a  given  time.  The  result  is  that  we 
have  a  foreign  protein  slowly  broken  up,  with  the  liberation 
of  a  small  quantity  of  irritant  poison  over  a  considerable 
interval  of  time.  For  this  reason  the  reaction  of  general 
sensitization  is  slow  in  its  development,  maintained  at  its 
maximum  for  a  considerable  period,  and  subsides  gradually. 
When  tuberculin  is  instilled  into  the  eye  of  a  normal 
individual,  no  apparent  result  is  obtained,  since  no  ferment 
is  present  in  the  body  capable  of  splitting  up  tuberculin. 
However,  as  a  result  of  the  instillation  itself,  certain  cells 
of  the  mucous  membrane  are  stimulated  to  produce  a 
specific  ferment  which  will  be  stored  up  as  a  zymogen  for 
future  use.  If  subsequently  a  solution  of  tuberculin  is 
brought  in  contact  with  these  sensitized  cells,  the  zymogen 
is  activated,  liberated  in  a  concentrated  form,  and  splits 
up  at  once  all  of  the  tuberculin  introduced.  The  result  is 
that  we  obtain  the  reaction  characteristic  of  local  sensitiza- 
tion, which  is  rapid  in  onset,  comparatively  severe  in  type, 
and  disappears  with  great  rapidity.  Owing  to  the  high  grade 
of  inflammatory  reaction  obtained  in  connection  with  the 
second  instillation,  it  is  well  to  use  a  more  dilute  solution 
of  tuberculin  and  to  ask  the  patient  to  present  himself  for 
examination  within  from  two  to  four  hours  following  the 
instillation.  At  this  time,  if  any  noticeable  redness  is 
present,  the  eye  should  be  thoroughly  washed  out  with  a 
saturated  solution  of  boracic  acid  in  order  to  remove  any 


THE  PHENOMENA  OF  INFECTION  467 

excess  of  tuberculin  which  may  be  present.  The  information 
obtained  through  the  employment  of  the  second  instillation 
in  an  individual  who  has  previously  failed  to  react  is  of 
value  in  that  it  indicates  that  the  body  cells  of  the  patient 
are  capable  of  producing  a  ferment  which  will  split  up 
tuberculin,  and  consequently  we  should  have  obtained  a 
primary  reaction  provided  the  individual  was  actively  tuber- 
culous. Failure  of  the  ophthalmo-reaction  occurs  under 
the  following  conditions:  (1)  In  early  cases  in  individuals 
who  are  incapable  of  producing  the  specific  ferment.  Such 
individuals  will  fail  to  react  to  the  second  instillation. 
(2)  Normal  individuals  who  are  not  afflicted  with  tuber- 
culous disease  will  fail  to  react  to  the  primary  instillation. 
They  may,  or  may  not,  react  to  the  second  instillation, 
depending  on  whether  or  not  their  body  cells  are  capable 
of  producing  the  specific  ferment.  (3)  Patients  suffering 
from  acute  tuberculous  disease  or  advanced  cases  fail  to 
react  to  either  the  first  or  the  second  instillation.  In  these 
cases  the  failure  of  the  reaction  is  due  to  the  exhaustion  of 
any  specific  ferment  which  may  have  been  present  through 
the  overwhelming  of  the  system  with  tuberculin,  or  to  the 
accumulation  of  split  products,  as  has  been  stated. 

While  the  importance  of  sensitization  in  connection  with 
the  infectious  diseases  is  not  as  yet  thoroughly  appreciated, 
later  investigations  have  been  conducted  largely  along 
these  lines.  Thus,  sensitization  is  undoubtedly  an  important 
factor  in  the  treatment  of  bacterial  diseases  through  the 
employment  of  vaccine  therapy.  This  is  true  whether  the 
vaccine  employed  consists  of  the  whole  bacterial  cell  or 
the  split  products,  such  as  those  obtained  after  our  method. 

The  injection  of  foreign  proteins  as  such  into  the  body 
always  represents  an  abnormal  condition.  The  symptoms 
of  sensitization  following  the  administration  of  horse  serum 
in  man  may  be  divided  into  two  classes,  according  to  the 
interval  of  time  elapsing  between  the  administration  of  the 
serum  and.  the  development  of  symptoms.  In  general,  it 
may  be  stated  that  symptoms  of  sensitization,  provided  they 
occur  at  all,  show  themselves  either  very  shortly  after  the 


468  PROTEIN  POISONS 

administration  of  the  serum,  or,  if  not  at  this  time,  on  the 
seventh  to  the  tenth  day  following  the  injection.  In  instances 
in  which  effects  are  not  noticeable  until  from  seven  to  ten 
days  following  the  injection,  the  symptoms  are  largely 
confined  to  those  of  peripheral  irritation,  as  evidenced  by 
urticarial  lesions  accompanied  by  intense  itching.  On  the 
other  hand,  in  cases  in  which  a  reaction  follows  within 
twenty-four  hours,  the  symptoms  of  poisoning  are  more 
pronounced,  and  where  occasionally  a  fatal  result  follows 
it  occurs  usually  within  an  hour  after  the  injection.  In 
these  cases,  which  are  fortunately  rare,  we  find  that  the 
symptoms  are  very  similar  to  those  obtained  through  the 
injection  of  the  poison  obtained  by  Vaughan  through 
protein  cleavage.  Thus,  Gillette1  reports  the  case  of 
an  asthmatic  fifty-two  years  old,  to  whom  he  gave  2000 
units  of  antitoxin  globulin,  administered  under  the  left 
scapula.  While  dressing,  following  the  injection,  the 
patient  complained  of  a  prickling  sensation  in  the  chest  and 
back  of  the  neck.  He  at  once  sat  down  in  the  chair  and 
complained  of  inability  to  breathe.  The  physician  felt  his 
pulse  and  found  it  full  and  regular.  Immediately  thereafter 
the  patient  was  seized  with  a  tonic  spasm,  during  which 
death  ensued,  the  whole  interval  elapsing  between  the 
injection  and  the  fatal  outcome  not  exceeding  five  minutes 
in  duration.  In  spite  of  the  rapidity  with  which  death 
occurred  in  this  case,  we  can  still  recognize  evidences  of 
the  three  stages  characteristic  of  fatal  protein  poisoning: 
the  stage  of  peripheral  irritation  indicated  by  itching 
sensations  in  the  skin,  the  stage  of  partial  paralysis  or 
weakening  of  the  lower  extremities,  and  the  convulsive 
stage,  during  which  death  occurred. 

In  cases  of  sudden  death  following  within  a  few  minutes 
after  the  injection  of  horse  serum,  it  is  not  infrequent  that 
one  of  the  stages  is  absent  or  ill-defined.  Thus,  in  the 
instance  cited  above,  the  loss  of  ability  to  move  the  lower 
limbs  was  not  specifically  mentioned,  although  in  other 

1  Jour.  Amer.  Med.  Assoc.,  January  4,  1908,  p.  40. 


THE  PHENOMENA  OF  INFECTION  469 

reported  cases  patients  have  before  death  remarked  on  their 
inability  to  walk. 

It  is  quite  evident  from  a  study  of  the  untoward  results 
following  the  administration  of  horse  serum,  that  the 
apparent  differences  existing  between  immediate  mani- 
festations and  those  occurring  after  an  incubation  period 
of  from  seven  to  ten  days  are  of  degrees  of  intensity  rather 
than  of  character  of  the  poisoning.  Thus,  in  the  instances, 
fortunately  rare,  in  which  death  occurred  within  thirty 
minutes  following  the  injection,  the  symptoms  are  due  to 
the  liberation  of  a  fatal  amount  of  poisonous  substance  at 
once,  and  in  instances  in  which  alarming  but  not  fatal 
symptoms  arise  shortly  after  injection,  recovery  from  the 
intoxication  is  usually  prompt  and  complete.  On  the  other 
hand,  where  symptoms  appear  only  after  an  interval  of 
from  seven  to  ten  days,  and  are  confined  to  those  of  peripheral 
irritation,  as  evidenced  by  the  development  of  urticaria, 
we  find  that  complete  recovery  is  slow  and  tedious. 

That  such  differences  should  exist  appears  but  natural 
when  we  consider  the  mechanism  involved  in  sensitization, 
and  the  fact  that  immediate  effects  are  due  to  the  injection 
of  the  serum  into  a  sensitized  individual,  whereas  the  remote 
effects  are  to  be  looked  upon  as  a  manifestation  of  the 
sensitization  of  the  patient  as  the  result  of  the  injection 
itself.  In  the  first  instance  the  individual  has  stored  up  in 
his  body  cells  a  ferment  which,  liberated  by  the  injection 
of  the  serum,  splits  up  the  foreign  protein  introduced  at 
once,  and  sets  free  all  of  the  poison  contained  therein  imme- 
diately. The  symptoms  resulting  therefrom  are  necessarily 
acute  in  character,  sudden  in  development,  and  transitory 
in  nature,  since  the  effects  of  the  poison  rapidly  disappear. 
In  the  individual  developing  symptoms  after  an  incubation 
period  of  from  seven  to  ten  days,  conditions  are  decidedly 
different.  In  this  case  no  special  ferment  capable  of  pro- 
ducing proteolysis  of  the  foreign  proteins  contained  in 
the  serum  is  present  within  the  body  at  the  time  of  injec- 
tion, and  as  a  result  the  foreign  proteins  continue  to 
exist  as  such  within  the  body  for  a  certain  length  of  time. 


470  PROTEIN  POISONS 

However,  under  their  influence  certain  body  cells  are  stimu- 
lated to  produce  a  ferment  which  will  split  up  the  foreign 
substances  into  simpler  non-specific  bodies.  The  fact  that 
animals  injected  with  serum  do  not  become  hypersensitive 
to  a  second  injection  until  after  the  lapse  of  from  seven  to 
ten  days  indicates  clearly  the  length  of  time  necessary  for 
the  new  ferment  to  be  formed  in  appreciable  amount. 
Symptoms  developing  after  an  incubation  period  are, 
therefore,  to  be  explained  by  the  fact  that  the  foreign 
proteins  still  existing  in  the  tissue  are  acted  upon  by  the 
enzyme  called  forth  by  their  presence.  Under  these  con- 
ditions, however,  no  large  amount  of  ferment  will  be  active 
at  any  given  time,  and  consequently  the  amount  of  poison 
liberated  through  protein  cleavage  at  any  one  period  will 
be  small  in  amount,  although  the  cleavage  itself  will  con- 
tinue over  a  comparatively  long  interval  and  the  resulting 
poisoning  will  be  more  chronic  in  type.  This  affords  a 
plausible  explanation  of  the  fact  that  late  manifestations 
of  serum  sickness  are  milder  in  character,  being  confined 
for  the  most  part  to  those  of  peripheral  irritation,  and  also 
of  longer  duration. 

Provided  the  theory  advanced  above  is  correct,  one  would 
expect  to  find  a  difference  in  the  time  interval  elapsing 
between  the  injection  of  the  foreign  serum  and  the  subse- 
quent appearance  of  symptoms  of  poisoning,  depending  on 
whether  the  individual  had  previously  been  treated  with 
serum  or  not.  While  it  is  true  that,  in  many  instances  in 
which  alarming  symptoms  develop  immediately,  a  history 
of  previous  treatment  is  unobtainable,  the  results  quoted 
by  Pirquet  are  interesting  as  bearing  on  this  point.  Thus 
it  has  been  found  that  of  214  individuals  who  developed 
symptoms  after  the  first  injection  of  serum  for  therapeutic 
purposes,  111,  or  51.8  per  cent.,  manifested  symptoms  of 
poisoning  on  from  the  seventh  to  the  tenth  day  inclusive; 
while  in  172  patients  who  received  a  second  injection,  89, 
or  51.7  per  cent.,  showed  signs  of  poisoning  within  the  first 
forty-eight  hours. 

As  has  been  previously  mentioned,  alarming  symptoms 


THE  PHENOMENA  OF  INFECTION  471 

following  the  use  of  diphtheria  antitoxin  and  other  thera- 
peutic sera  are  fortunately  so  rare  that  they  should  not  be 
considered  too  seriously  when  indications  for  the  use  of 
such  sera  arise.  However,  there  are  certain  precautions 
which  can  and  should  be  employed,  and  which  may  aid 
materially  in  avoiding  the  untoward  effects  following  the 
administration  of  these  remedies.  Much  has  been  accom- 
plished by  the  efforts  of  various  pharmaceutical  houses  to 
prepare  an  antitoxin  from  which  a  large  proportion  of  the 
foreign  albumin  contained  in  horse  serum  has  been  removed, 
and  such  products  should  be  used  exclusively  whenever 
possible.  It  wrould,  furthermore,  appear  to  be  well  to  make 
a  preliminary  test  with  regard  to  the  sensitiveness  of  any 
given  individual  to  the  serum  employed.  This  may  be 
done  by  the  injection  of  a  very  small  test  dose  (0.05  c.c.)  of  the 
serum  and  watching  for  rapid  evidences  of  toxic  action  in 
the  patient.  Alarming  signs,  if  they  occur,  develop  usually 
within  an  hour  after  treatment;  and  if  no  sign  of  poisoning 
occurs  within  this  time,  it  may  be  safely  assumed  that  the 
individual  does  not  contain  within  his  body  the  special 
ferment  required  to  split  up  the  material  injected,  and  a 
second  injection  may  be  made  with  impunity,  provided 
the  interval  of  time  elapsing  between  each  is  not  sufficiently 
long  to  admit  of  the  development  of  sensitization.  In  a 
disease  such  as  diphtheria  this  is,  of  course,  a  matter  which 
does  not  enter  into  consideration  in  the  treatment  of  any 
given  attack.  The  preliminary  injection  of  atropine  has 
been  advised  by  Auer,  who  found  that  18  out  of  25  sensi- 
tized guinea-pigs  which  had  been  given  atropine  sulphate 
recovered  from  the  second  injection,  while  of  24  untreated 
controls,  only  6  survived. 

When  symptoms  of  sensitization  appear  immediately  or 
soon  after  the  injection,  the  use  of  ether  by  inhalation  is 
to  be  recommended,  as  Besredka  found  that  in  experi- 
mental sensitization  animals  narcotized  with  ether  did  not 
succumb  to  the  second  injection.  When  it  is  ascertained 
that  a  given  individual  is  sensitive,  and  nevertheless  the 
use  of  therapeutic  sera  is  imperative,  it  may  be  given  by 


472  PROTEIN  POISONS 

fractioning  the  total  amount  of  serum,  and,  instead  of 
using  a  single  dose,  give  several  doses  at  frequent  intervals 
over  a  considerable  period  of  time.  In  this  way  it  may  be 
possible  to  exhaust  the  ferment  present  in  the  body  for 
the  time  being,  after  which  further  amounts  may  be  given 
with  impunity.  A  temporary  exhaustion  of  the  special 
ferment  would  explain  cases  of  diphtheria  such  as  are 
reported,  in  which,  although  alarming  symptoms  followed 
the  first  injection,  on  account  of  the  condition  of  the  patient, 
a  second  injection  seemed  advisable,  and  was  given  within 
a  few  hours  without  any  untoward  effects  whatever.  Experi- 
ences such  as  this  find  their  analogy  in  experimental  work 
in  the  fact  that  sensitized  animals  which  have  recovered 
from  the  poisoning  following  a  second  injection  are  not 
again  susceptible  to  that  particular  protein  until  after  the 
lapse  of  several  days. 

The  idiosyncrasies  which  certain  individuals  possess 
with  regard  to  certain  protein  articles  of  diet  would  appear 
to  be  explainable  on  the  ground  that,  through  some  abnor- 
mal condition  of  the  intestinal  mucosa,  certain  protein 
substances  are  allowed  to  enter  the  body  in  an  unchanged 
state.  The  symptoms  which  develop  are  certainly  strikingly 
suggestive  of  those  described  as  appearing  in  connection 
with  sensitization.  The  symptom  most  constant  in  appear- 
ance is  an  urticaria,  more  or  less  generalized  in  extent  and 
accompanied  by  intense  itching.  Thus,  Bruick  reported 
the  case  of  a  man  who  reacted  with  urticaria  every  time 
after  eating  pork.  Smith  reports  the  case  of  an  individual 
who  developed  a  severe  urticaria  within  a  short  time  after 
partaking  of  any  article  of  diet  which  contained  buckwheat. 
Numerous  cases  of  severe  urticaria  accompanied  by  dyspnea, 
and  occasionally  incoordination  of  the  lower  extremities, 
have  been  reported  as  occurring  in  susceptible  individuals 
after  partaking  of  any  food  containing  egg  albumen,  and 
one  of  us  has  recently  had  opportunity  to  observe  an  indi- 
vidual who  developed  a  generalized  urticaria  accompanied 
by  marked  edema  and  intense  itching  within  half  an  hour 
after  partaking  of  peas  as  an  article  of  diet. 


THE  PHENOMENA  OF  INFECTION  473 

The  most  striking  peculiarity  mentioned  in  connection 
with  the  above  idiosyncrasies  is  the  rapidity  with  which 
the  symptoms  of  poisoning  develop  after  the  introduction 
of  the  attending  cause  within  the  alimentary  tract.  As 
has  been  mentioned,  these  individual  peculiarities  are 
possibly  best  explained  by  the  supposition  that  the  individual 
has  become  sensitized  to  certain  specific  proteins,  the  sensi- 
tization  arising  from  the  fact  that  the  particular  protein 
has  gained  entrance  into  the  body  through  the  intestinal 
mucosa  in  an  unchanged  state.  When  this  occurs  a  foreign 
protein  is  present  in  the  tissues  and  fluids  of  the  body,  and 
to  counteract  the  abnormal  condition  thus  produced  certain 
body  cells  are  called  on  to  develop  a  proteolytic  ferment 
which  will  have  for  its  function  the  cleavage  of  the  particular 
protein  present  in  any  given  case.  This  ferment,  once 
formed,  is  stored  up  in  certain  cells  as  a  zymogen  for  future 
use.  The  same  protein  cleavage  then  occurs  as  normally 
takes  place  within  the  intestinal  canal,  with  the  important 
difference  that  the  toxic  substances  formed  are  liberated 
within  the  body  itself  and  consequently  are  capable  of 
exerting  their  harmful  action.  That  these  ferments  are 
not  present  in  the  body  in  inexhaustible  amount  is  shown 
experimentally  by  the  fact  that  animals  which  have  recovered 
from  the  effects  of  sensitization  following  a  second  injection 
of  egg  albumen  are  not  subsequently  sensitive  to  this 
protein  until  after  the  lapse  of  several  days.  This  is  un- 
doubtedly due  to  the  fact  that  the  ferment  has  been  largely 
used  up  in  bringing5  about  the  cleavage  of  this  particular 
protein  and  time  must  be  allowed  for  the  body  cells  to 
produce  an  additional  amount.  In  other  words,  it  is  possible 
in  a  susceptible  individual  to  destroy  their  susceptibility 
with  regard  to  any  particular  protein  through  an  exhaustion 
of  the  special  ferment  present  in  their  body.  This  is  well 
illustrated  clinically  by  the  following  example:  A  woman 
who  was  fond  of  strawberries,  but  developed  an  intensely 
disagreeable  urticaria  after  each  indulgence,  was  accustomed 
to  eat  this  fruit  two  or  three  times  during  the  season. 
Finally,  being  firmly  convinced  that  the  rash  was^simply  a 


474  PROTEIN  POISONS 

nervous  manifestation,  she  determined  to  eat  them  con- 
tinuously, in  order,  as  she  said,  "to  break  herself  of  the 
nervous  habit."  After  the  first  week  no  unpleasant  symp- 
toms whatever  were  noted  following  the  daily  use  of  this 
article  of  diet.  This  appeared  to  the  patient  to  be  an 
entirely  satisfactory  proof  of  the  effectiveness  of  Christian 
Science,  and  yet  the  phenomenon  is  explainable  on  a 
rational  basis.  The  daily  use  of  strawberries  had  led 
to  an  exhaustion  of  the  special  ferment  in  her  body,  and 
subsequent  indulgence  was  consequently  not  followed  by 
untoward  symptoms.  Whether  or  not  the  experience  was 
repeated  during  the  succeeding  summer  we  have  not  been 
able  to  ascertain. 

In  conclusion,  we  may  state  that  sensitization  primarily 
represents  an  important  phenomenon  of  lytic  immunity. 
Sensitization  occurs  whenever  a  foreign  protein  as  such 
gains  entrance  into  the  fluids  and  tissue  of  the  body,  and 
results  from  the  development  within  the  body  of  a  special 
ferment  which  will  attack  the  particular  protein  introduced. 
When  individuals  become  sensitized  through  the  introduc- 
tion of  dead  protein  substances,  such  as  egg  albumen  or 
horse  serum,  the  results  obtained  prove  unfavorable  to  the 
individual.  In  these  cases  our  attempt  should  be  to  bring 
about  a  desensitization  of  the  individual  through  the 
exhaustion  of  the  special  ferment.  On  the  other  hand, 
sensitization  occurring  as  a  result  of  the  entrance  of  bacterial 
cells  into  the  body  represents  a  beneficial  process,  and 
plays  an  important  part  in  the  development  of  active 
immunity  to  the  specific  infections.  Since,  under  ordinary 
circumstances,  pathogenic  bacteria  represent  the  only 
proteins  which  gain  entrance  into  the  body  in  an  unchanged 
state,  we  may  conclude  that  sensitization  arises  as  an 
attempt  of  nature  to  protect  the  individual  against  bacterial 
disease. 


INDEX 


AMINO  acids,  74,  78 
Anaphylactic  state,  242 
Anaphylatoxin,  296 
Anaphylaxis,  214 

in  vitro,  274 

mechanism  of,  247 

passive,  254 
Animals,  action  on,  119 
Anthrax  protein,  189 
Arthus  phenomenon,  262 
Anti-anaphylaxis,  258 


BACILLUS,  action  of,  119 
colon,  37 
diphtheria,  38 
participate  proteins,  17 
pathogenicity  of,  21 
reducing  action  of,  62 
Bacterial  cellular  substance,  37 
/3-iminazolyle  thylamin,  291 


CANCER  cell,  specific  ferments  of,  416 
Cellular  protein,  hydrolysis  of,  81 
substance,  37 

action  of,  121 

carbohydrates  in,  57,  66 

chemistry  of,  52 

diamino-acids  of,  74 

fats  of,  60 

immunization  with,  138 

mono-amino  acids  of,  78 

nucleins  in,  73 

of  pneumococcus,  205 

proteins  of,  52 


Cleavage  of  proteins,  95 
Cultures,  massive,  29 


DEFIBRINATED  blood,  the  poisonous 

action  of,  334 
Diamino-acids  in  cellular  substance, 

74 
Diseases,  infectious,  23 


E 


EGG-WHITE,  cleavage  of,  98 

the  disposition  of,  when   intro- 
duced parenterally,  355 
Ergamin,  291 


FEVER,  acute  fatal,  381 
continued,  374 

digestive  action  of  blood  in,  395 
followed  by  immunity,  397 
intermittent,  386 
protein,  373 
remittent,  387 


GASTRIC  juice,  action  of,  42 
Group,  poisonous,  19 
sensitizing,  234 


HAPTOPHOR,  properties  of,  112 
Histamin,  291 


476 


INDEX 


IMMUNITY,  natural,  24 

toxin,  27 

with  split  products,  157 
Immunization    with    poisonous   por- 
tion, 138 

with  residue,  144 
Infection,  the  phenomena  of,  436 


KYRINS,  the,  295 


M 

MONO-AMINO-ACIDS  in  cellular  sub- 
stance, 78 


N 


NUCLEUS,  chemical,  20 


PARENTERAL  digestion,  342 
Peptone,  the  fats  of,  342 
Pneumococcus,  cellular  substance  of, 

205 
Poison,  action  of,  125 

crude  soluble,  101 

immunization  with,  138 

physiological  action  of,  315 

protein,  17,  284 
Protein  fever,  373 

poisons,  nitrogen  in,  111 

sensitization,  25,  214 
Proteins,  cleavage  of,  95 

particulate,  17 

physiological  action  of,  320 


RED  corpuscles,  production  of  fever 

with,  391 

Reinjection,  the,  243 
Residue,  immunization  with,  144 


SENSITIZATION,  cellular,  321 

period  of  incubation,  241 

protein,  214 

symptoms  of,  245 
Sensitizers,  219 

volatile,  231 


TANKS  for  massive  cultures,  30 
Theories,  323 

Theory  of  Friedberger,  324 
of  Nolf,  340 

of  Vaughan  and  Wheeler,  327 
Toxic  sera,  264 
Toxicity    of    extracts    from    normal 

tissue,  325 
Toxogens,  266 
Trypsin,  action  of,  43 
Tubercle  bacillus,  cellular  substance 

of,  164 

cleavage  of,  165 
poison  in,  165 
residue  of,  165 
split  products  of,  164 
toxophor  of,  178 
Tuberculosis  and  sensitization,  181 


VACCINATION,  24 
Vaccines,  26 
Vegetable    proteins, 
fever  with,  403 


production    of 


*w 

•a 
3 


-P 
o 


PH 


University  of  Toronto 
Library 


DO  NOT 

REMOVE 

THE 

CARD 

FROM 

THIS 

POCKET 


Acme  Library  Card  Pocket 

Under  Pat.  "Ref.  Index  File" 
Made  by  LIBRARY  BUREAU