i

CELL PHYSIOLOGY AND
PHARMACOLOGY

^ z r^j^y

J/

CELL PHYSIOLOGY AND

PHARMACOLOGY

by
J.F.DANIELLI,

Ph.D. (LONDON AND C AM BRI DGE), D. Sc.

Professor of Zoology, King's College, London,

and Honorary Lecturer in Pharmacology, University College, London.

Formerly Reader in Cell Physiology, Royal Cancer Hospital, London,

and Fellow of St. John's College, Cambridge

ELSEVIER PUBLISHING COMPANY, INC,

NEW YORK AMSTERDAM LONDON BRUSSELS

1950

PRINTED IN THE NETHERLANDS
BV MEIJEr's BOEK- en HANDELSDRUKKERIJ - WORMERVEER

\ ■'

W,. i . - . ■ < V V

PREFACE '' ^^ %:

This book is based upon lectures given at University
College, London, at the invitation of Professor F.R.
WiNTON. It was not intended that these lectures should
constitute an exhaustive survey of the place of cell phy-
siology in pharmacology. They were intended to indi-
cate some of the more important factors, on the cellular
level, which must be considered by students of drug
action.

My interest in the mode of action of drugs has been
stimulated at various times by A.J.Clark, H.R.Ing,
J. C. Drummond, J. H. Gaddum, and F. R.Winton. Dur-
ing the war of 1939-45 this interest was turned towards
practical problems, and I was able to study at first hand
the problems which are encountered when a search is
being made for a new drug. The problems roughly se-
parate into three broad groups — corresponding to the
disciplines of organic chemistry, physical chemistry,
and biology. It was impressive to see that the chemical
problems could often be handled with ease and moder-
ate precision. But in practically all investigations it be-
comes necessary for the chemist to introduce theoris-
ation on the biological side at some stage. This is some-
times done as a guide to action, and sometimes as a
guide to inaction. But whatever the intention, the chem-
ist's biological theories are apt to be more surprising
than successful.

VI PREFACE

When, as sometimes occurs, the biologist is called in
to deal with the biological sides of the problem, his
success is frequently limited to pointing out errors in
the chemist's biological theories. The reason for this
limitation is the paucity of information on the biological
side. Whereas biology is probably more complicated
than chemistry, vastly more man-hours have been de-
voted to chemical research than to biological research.
When this situation is remedied, it will be possible for
the biological side of drug development to proceed in a
rational manner. This will eliminate the hit or miss tech-
nique, which is now the main basis of the search for new
drugs. For chemists this will mean much diminution of
effort — probably 90% or more of the time used for syn-
thesis of useless compounds can be eliminated. And for
biologists there will be a corresponding diminution in
the time spent in testing inactive compounds.

Consequently my hope is that these lectures will play
some little part in directing attention to the biological
aspects of drug action, and in showing that studies in
the biological field are practicable.

I am deeply indebted to Prof. A. Haddow for advice
and for reading the galley proofs.

J.F.Danielli

CONTENTS

Preface v

I. The Cell as a Physico-Chemical Unit

Introduction i

Cytochemical aspects 2

Cytoplasmic gels 4

Chromosome and gene structure 7

Physico-chemical aspects 11

Defence mechanisms 20

Self-reproducing bodies 22

Integration 23

References 24

II. Possible Actions of Drugs on Surfaces

Ionic interactions 25

Dipole interactions and complex formation 32

Specificity in surface reactions 34

Mechanism of lysis 36

The action of oestrogens on monolayers 37

The effect of micelle formation 39

Long-range forces 43

References 45

III, Membrane Permeability and Drug Action

Introduction 46

Membrane permeability and drug structure 48

Problems of the access of drugs to organs 63

Examples of the permeability factor in drug action .... 66

References 73

IV. Enzymes and Drug Action

Functions of enzymes 74

Possible functions of drugs in relation to enzymes .... 75
Problems in the analysis of the action of drugs on enzymes . 78
The action of drugs on respiration and glycolysis in muscle . 82
The action of various enzyme poisons on different physiologi-
cal processes 85

Classification of drugs according to their physiological effect 87

64433

VIU CONTENTS

The classification of drugs in terms of enzyme systems upon

which they act 8g

The mode of action of vesicants 91

Biological aspects of enzyme studies 94

References 96

V. The Actions of Narcotics

Introduction 97

Theories of actions upon surfaces 99

Theories based on oil-water partition effects ^05

Theories based on actions on enzymes 109

References 114

VI, Responses of Cells on the Biological Level

Introduction 119

The nature of biological responses 119

Artificial parthenogenesis I23

Mitotic poisons 123

Reproduction of bacteria and viruses 136

Nuclear and cytoplasmic drug action 139

Possible modes of drug action upon genes 141

The relationship between hormones and evocators .... 146

References I49

Author Index 151

Subject Index 153

Nucleolus

Chromocentre

(b)

(c)

Nucleic acid

Glycogen

Phosphatase

Bile canaliculi

(f)

Granules

Glycogen

Gel

Fig. I. Cytochemistry of hepatic cells, a. Hydrogen ions. b. Diffusible
SH. c. Pentose nucleic acid (red) and deoxypentose nucleic acid (green),
d. Glycogen, e. Alkaline phosphatase; in addition to the main sites
shown, there are low concentrations of alkaline phosphatase in the cyto-
plasm and in the chromocentres. f. Fat (red) and fatty aldehyde
(green), g. Centrifuged cell; fat at upper pole. h. Dividing cell; pentose
nucleic acid (green) and deoxypentose nucleic acid (red).

CHAPTER I

The Cell as a Physico- Chemical Unit

Introduction

In this chapter it is proposed to outHne the main aspects
of the cytological background which must be borne in
mind when the action of drugs is considered from the
point of view of cell physiology. In his books, The Mode
of Action of Drugs on Cells and General Pharmacology,
AJ .Clark drew particular attention to the importance
of the study of cell physiology in connection with drug
action. He wrote *'most of the functions of the body are
regulated by drug action, and hence the manner in
which drugs exert their action on cells has become one
of the most fundamental problems in physiology."
"(This study) is of course dependent on our knowledge
of the physical chemistry of cells." In the first instance,
we must make a study of cell morphology. But it is ab-
solutely essential that we should not confine ourselves
to the static morphology which is bound to be the result
of studies exclusively based upon the methods of classi-
cal histology. We need to consider every cell as a dyna-
mic organisation, as a system organised for activity, not
simply as a system which has a particular microscopic form.
Cell Physiology i

2 THE CELL AS A PHYSICO-CHEMICAL UNIT

Cytochemical Aspects

In Fig. I (facing page i) are given diagrams of the cyto-
chemical distribution of a number of different substan-
ces in the hepatic cells of the liver. The diagram shows
the distribution of hydrogen ions, SH groups, nucleic
acids, glycogen, alkaline phosphatase, acid phosphatase^
lipoids, and long chain aldehydes. These are the main
substances the distribution of which has so far been
revealed either by physico-chemical studies on cells,
or by cytochemical studies on cells. The distribution of
these various substances, and the concentration in par-
ticular places of these substances, is much affected by
the physiological condition of the animal, and by diet
and other factors extrinsic to the cell. Nor are all the
hepatic cells identical. Even in any particular lobule of
the liver, one sees marked differences between the cells
which are present at the periphery of the lobule and
those cells which are clustered round the central vein.
This difference is quite striking with pentose nucleic
acid, glycogen, alkaline phosphatase, and long chain al-
dehydes. On the other hand, there is comparatively
little difference in the desoxyribose nucleic acid con-
tent of hepatic cells in different positions in the lobule
and the total concentration of desoxyribose nucleic acid
is known to be comparatively little affected by extrinsic
factors.

Even when so few substances as those mentioned

CYTOCHEMICAL ASPECTS 3

above are considered, it becomes at once clear that the
physico-chemical structure of the cell is in a dynamic
condition. The fact that the broad outlines of structure,
as shown by classical histological methods, are not read-
ily changed, is indicative of certain stable elements
amongst those determining cell behaviour. But this
should not hide from us the great variation which may
occur on the chemical level. This variation occurs at
just that level on which we must expect most drugs to
act, so that in a sense the variability of the cell at this
chemical level is of more importance to us than the
relatively static nature of certain structural patterns.

It is possible to procedure redistributions of most
substances in the cell without irreversibly damaging the
cell. E.g. when a cell goes into division, the organisation
of some of the cell products is totally changed in the
production of a spindle. The sharp distinction between
nuclear sap and cytoplasmic material breaks down. Al-
ternatively, by centrifuging the cell, we can obtain a
new distribution of substances quite unlike that which
arises in a cell under normal conditions or when it is
dividing. If the use of centrifugal force is not excessive,
the cell can reconstitute the original organisation of sub-
stances, and continue with its normal functions unim-
paired. In the case of Ascaris eggs, the unfertilised eggs
may be centrifuged at 100,000 times gravity for four
days, and yet still develop normally when fertilised.
Thus particular adjuxtapositions of matter which are

-3/x

/^d'^)^

W

4 THE CELL AS A PHYSICO-CHEMICAL UNIT

shown up by c5rtochemical investigations are not all of
then; physically necessary for the maintenance of the
life of the cell. Insofar as there are absolutely unchange-
able centres of organisation in the cell, these must be
based upon the formed bodies of the cell such as chro-
mosomes, mitochondria, the cell and the nuclear mem-
branes, and the cell granules. The evidence at present
available shows that as long as these bodies are left
intact, the cell can recover from quite violent treatment
and disturbances of its normal chemical condition. But
this does not mean that the distribution of chemical
substances which we see in the cell under normal con-
ditions is of no consequence. Although cell constituents
may be rearranged by centrifugation or by micro-dis-
section procedures without killing the cell, there is a
certain amount of clear evidence showing that the acti-
vity of the cell is no longer carried on in the same way
during the period in which this redistribution exists.
Thus Bracket has shown that the cyanide-sensitive
respiration of amphibian liver is eliminated by strati-
fication in the centrifuge, but returns when the cell con-
stituents have been restored to their normal positions.

Cytoplasmic Gels

Except during mitosis, protoplasmic gels do not play
a very obviously prominent part in the activity of he-
patic cells. But in a variety of other cells experimental

CYTOPLASMIC GELS 5

procedures have shown that cytoplasmic gels are res-
ponsible for many characteristic activities. The normal
condition of these gels is essential for mitosis and cell
division, protoplasmic streaming, amoeboid movement,
phagocytosis and the maintenance of cell form. For exam-
ple the marine protozoon Ephelota coronata throws out
fine protoplasmic tentacles, the length of which is many
times greater than their diameter. This structure is
maintained by the cortical gel layer which lies just inside
the plasma membrane in the tentacles. When this gel
is liquified by application of high pressure, the operation
of surface tension forces immediately breaks the ten-
tacles up into a series of fluid droplets. The importance
of organised structures is also shown by some experi-
ments which were conducted by Moore on the slime
moulds. These moulds were able to grow through mem-
branes having a pore diameter of the order of i micron.
But if they are filtered by pressure through pores as
large as loo microns in diameter, the respiration of the
mould is decreased by 50%, and if the pore size is not
larger than 20 microns, the organism is killed. In consi-
dering the significance of these gels in the daily life of
the cell, it is important to remember that cell division,
protoplasmic streaming, amoeboid movement, phago-
cytosis, and the maintenance of cell form can go on
quite normally in the absence of the nucleus. In some
cases the absence of a nucleus does not exert a pronoun-
ced effect for several days after its removal from the cell.

THE CELL AS A PHYSICO-CHEMICAL UNIT

There is comparatively little information about the
details of structure, and so forth, of these gels in most
cells. In striated muscle cells, the gels become so pro-
minent as to be the characteristic feature of the cell, and
it is known that the component molecules are arranged
in organised fashion to produce definite intracellular
bands, discs, fibres and membranes. The most striking
of these features of the living cell is the distinction be-

Olycogen

Myosin
nisofropic band

~7

Isotropic band
A.T.P

Nucleus

Alkaline phosphatase
Acid phosphatase
Deoxypentose nucleic acid
Pentose nucleic acid

Fig. 2. Diagram of distribution of substances in a striated muscle fibre

tween anisotropic bands, which presumably contain the
greater part of the myosin of the cells, and the isotropic
bands. The former bands rotate the plane of polarisation
of plane polarised light, whereas the isotropic bands do
not. Cytochemical studies on muscle cells are at present
rather scanty. It is known that adenosine triphosphate
is concentrated in the isotropic bands, and that these
bands also contain a relatively labile alkaline phospha-
tase. Whereas acetone resistant acid phosphatase and
desoxyribose nucleic acid are both mainly present in the

CHROMOSOME AND GENE STRUCTURE 7

nucleus. This is indicated diagramatically on fig. 2. We
may be quite sure that this localisation of chemical sub-
stances has special chemical and physical significance.
One thing which is very clear is that often this locali-
sation keeps substrates apart from the enzymes which
would otherwise destroy them or utilise them prior to
the onset of the physiological activity for which this
process is useful. For example the co-existence of phos-
phate esters and of phosphatases which are active at
cellular p^ is only made possible by this physical sepa-
ration of the substances concerned. As soon as the bio-
chemist, by grinding or other maceration procedures,
breaks down this organisation on the chemical level,
phosphate esters are rapidly destroyed, and with them
the characteristic cycles which are involved in glucose
metabolism.

Chromosome and Gene Structure

The structures about which most details are available
are the chromosomes. There are two very different main
theories of the structure of the chromosome. One is
that it consists of at least one polypeptide chain running
the full length of the chromosome, with discrete bodies,
known as chromomeres, distributed along it. The chro-
momere, according to this theory, consists of material
additional to that contained in the long polypeptide
chains. Each chromomere has at least one gene and may

8 THE CELL AS A PHYSICO-CHEMICAL UNIT

contain several genes. According to the other theory, the
normal chromosome contains one or more polypeptide
chains extending the whole length of the chromosome.
The chromomeres, however, are believed to be not
additional material, but regions in which the main poly-
peptide chain is much folded upon itself. As in the first
theory, each chromomere is believed to consist of one
or more genes.

As Stedman & Stedman and Mirsky have shown,
chromosomes consist of at least two types of protein,
the basic proteins known as histones and acidic proteins
which Stedman calls chromosomin, and also two types
of nucleic acid, one of which contains a pentose and the
other a desoxypentose. The information about the phy-
sical structure of chromosomes has been very largely
derived from studies of exceptionally large chromo-
somes, such as those found in certain plant cells, and
the giant salivary chromosomes of the Diptera. Which-
ever of the theories of chromosome structure may be
correct, it is certain that the total amount of matter in
the chromomeres is greater than that in the parts of the
chromosomes lying between the chromomeres. Cyto-
chemical studies have shown that by far the greatest
part of the purine and pyrimidine, the desoxy sugar,
tyrosine, histidine, tryptophane, and alkaline phospha-
tase are present in the chromomeres. The interbands are
relatively lacking in these substances. Since genes are
known to be located in the chromomeres, it seems pro-

Plate I. Distribution of alkaline phosphatase on an X-chromosome of
Drosophila melanogaster (Danielli and Catcheside).

PHYSICO-CHEMICAL ASPECTS II

bable that the substances which have just been men-
tioned are essential constituents of genes: in fact it is likely
that a gene can be considered as a special array of en-
zyme molecules, organised so as to produce one chemi-
cal product, or a small family of chemical products. The
substances so produced are those involved in the medi-
ation of the effect of the gene. In addition to these sub-
stances, the gene has the capacity to reproduce itself
completely. The extent to which this is a different func-
tion from that of producing the substance or substances
concerned in the mediation of the genetic effect is un-
known.

The sharp localisation of chemical substances in the
chromomeres, and the variation in concentration of
these different substances from chromomere to chromo-
mere, must involve also, through the operation of Don-
nan equilibria etc., highly local variations in pH along
the chromosome and probably also in the SH content
of different parts of the chromosome. These two phy-
sico-chemical factors must be very important to consi-
der in relation to the enzymic activity of a gene.

Physico-Chemical Aspects

The foregoing discussion has been based mainly on
consideration of the cytochemical distribution of diffe-
rent substances. In forming a picture of the physico-
chemical system characteristic of the cell, we must con-

12 THE CELL AS A PHYSICO-CHEMICAL UNIT

sider a number of other properties of the molecular
types constituting the cell. The most important points
to be considered here are: i. the units of structure, 2.
the control of enzyme systems, 3. the dielectric proper-
ties, 4. membrane properties, 5. the dynamic condition
of cell constituents.

Units of Structure. Fig. 3 shows diagrams of a number
of types of units of structure which will be formed by

Adlineated chains

Tactoid

Leaflet

Micelle
Fig. 3. Units of cellular structure

molecules known to be present in the cell. The simplest
of these consists simply of adlineated polypeptide chains
such as are found in collagen fibres, keratin, muscle

PHYSICO-CHEMICAL ASPECTS I3

fibres and chromosomes etc. In the cases of the chromo-
some fibre, muscle fibre and collagen fibre it is known
that there is a good deal of differentiation along the
length of the chains. It is likely that this is also the case
with most other natural fibres. Except in the case of the
chromosome, it is not at all clear what significance
should be attached to the local differentiation along the
length of the fibre. One theory is that it assists in the
precise adlineation of polypeptide chains, thereby giving
rise to fibres of maximum strength.

A physically quite different type of structure, also
composed of adlineated protein molecules, is the type
found in tobacco mosaic virus tactoids. These consist of
needle-shaped molecules, which are oriented parallel to
one another, but which are not in contact at any point.
They are maintained in these positions, with the distance
between the long axes of the molecules relatively well
defined, by long range forces whose nature is not yet
clearly understood. Bernal has supposed that the
spindle and asters of a dividing cell are composed of
such tactoids. It is also possible that the forces acting
between daughter chromosomes during mitosis are of
this type, that cell adhesions and the form of cells are
in part determined by such forces, and even that the
differentiation of cells is influenced by forces of this type.
In addition to the gels or tactoids mentioned above,
there are the cortical gels which do not seem to consist
of very highly oriented molecules. They may have a

14 THE CELL AS A PHYSICO-CHEMICAL UNIT

closer resemblance to gelatin gels in which the indivi-
dual molecules appear to be dispersed more or less at

random.

When the lipoid molecules of the cell are also con-
sidered, we have units of structure of a fresh type aris-
ing, based on the micelle and on bimolecular leaflets of
fatty molecules. These types of structure are also shown
in Fig. 2. The stability of these structures is based on
the fact that the hydrocarbon parts of the molecules
concerned are in effect squeezed out of solution because
water attracts water more strongly than it attracts hy-
drocarbon. The polar groups of the molecules become
anchored in the interfaces between the hydrocarbon
parts of the molecules and the water. Owing to the very
high surface activity of protein molecules, the surfaces
of lipoid micelles and bimolecular leaflets must, under
biological conditions, always have adsorbed upon them
a monolayer of denatured protein. Under most circum-
stances there will also be adsorbed upon this primary
layer of denatured protein a secondary layer of globular
protein molecules. Thus a very complex structure may
arise simply as a result of the operation of adsorption
forces.

So far, the structure of the cytoplasmic granules found
in the cell has not been made sufficiently clear for de-
tailed discussion. Nor is it clear what part the diiferent
nucleic acids play in determining the structure of the
chromosomes and other bodies in which they are found.

PHYSICO-CHEMICAL ASPECTS I5

It is very important to notice that as a result of the pre-
sence of granules, chromosomes, membranes, nucleoli
and other formed bodies in the cell, there must necessari-
ly be great importance attributed to surface properties.
The main molecular constituents of cells, the proteins
and nucleic acids, are themselves such large molecules
that any reaction taking place with them, or in which
they are involved, is necessarily a surface action and not
a bulk reaction. And it must very often be the case that
two reacting molecules must be regarded as reacting in
the zone constituted by their overlapping surfaces. In
considering the basic physico-chemical nature of the
cell, we must therefore be particularly alive to the im-
portance of surface properties.

The Control of Enzyme Systems. Amongst the most im-
portant of the physico-chemical systems which form an
integral part of living cells are the systems controlling
the activity of intracellular enzymes. Among the factors
involved in controlling enzymes are hydrogen ion con-
centration, the concentration of SH groups, the concen-
trations of inhibitors and activators, and those processes
which control access of substrates to enzymes.

The hydrogen ion concentration is maintained con-
stant in cells partly by the buffering substances normal-
ly present, and partly by the active excretion of excess
of acidic or basic substances. The SH content of a cell
is important because many enzyme systems have maxi-

l6 THE CELL AS A PHYSICO-CHEMICAL UNIT

mal activity when either in the reduced state or in the
oxidised state. Usually the reduction can be carried out
by SH compounds, particularly by glutathione. Simi-
larly, the reduced forms of enzymes can usually be oxi-
dised with the oxidised form of glutathione. Thus the
presence within cells of glutathione constitutes a poising
system tending to maintain a given degree of reduction
of the enzyme systems of the cell, in just the same way
as the pH buffers tend to maintain a given degree of ioni-
sation of the enzyme systems.

Equally important with pn and SH content are the
factors of inhibition, activation and substrate access. But
extremely little is known of the variables which control
the operation of these factors.

However, the cell does not present a completely uni-
form environment with respect to pn and SH content.
Thus, when we consider a gene present in the nucleus,
one of the most striking properties is the high concen-
tration of ionising groups in the gene. As a result there
is a Donnan equilibrium between the nuclear sap and
the gene. Thus the pH in the gene will not be the same
as that in the nuclear sap. Even when we are dealing
with small bodies, like protein and nucleic acid mole-
cules, we encounter local variations in p^. This is be-
cause there are high concentrations of ionising groups in
the surface of proteins and nucleic acids, so that equili-
bria analogous to the Donnan equilibrium exist between
the fluid medium of the cytoplasm and the surfaces of

PHYSICO-CHEMICAL ASPECTS 17

the protein and the other colloidal particles dissolved
or suspended in it. The difference in pjj between such
surface and bulk phases may am,ount to i pH unit or
more. Thus the interior of a cell, although its average pn
value may be very constant, can in fact present an ex-
tremely variable pn in different sub-microscopic regions.
This permits of great variations in the degree of enzyme
activity, according to the particular sub-microscopic
environment with which individual enzyme molecules
are associated.

Very similar effects exist in connection with the con-
trol of the reduction of enzyme systems by glutathione.
In addition to the "diffusible" SH groups of glutathione,
there are other "indiffusible" SH groups permanently
fixed to protein molecules. The distribution of glutathi-
one between the surface of an enzyme, and the surround-
ing bulk phase, is in part determined by the total charge
on the protein and the degree of ionisation of the glu-
tathione. Hence there is a fairly complex equilibrium
determining the distribution of SH compounds between
the surface and bulk phases within a cell. Furthermore,
it may be important that whereas surface pn is almost
independant of the surface SH content, as the result of
the operation of electrostatic factors the surface SH con-
tent is not independant of pH-

Dielectric Constant of Cellular Systems. A number of
workers have pointed out that the presence of large po-

l8 THE CELL AS A PHYSICO-CHEMICAL UNIT

lar molecules, such as proteins, within living cells will
markedly raise the dielectric constant of the interior of
a cell, by comparison with the dielectric constant of the
external media. This increase in dielectric constant (as
measured by relatively low frequency alternating cur-
rent) is thought likely, by some workers, to have im-
portant consequences for the cell. But very little as yet
is known about these factors.

Membrane Properties. Extremely little is known of the
physico-chemical properties of the nuclear membrane,
but much is known about the properties of the plasma
membrane of the cell. The most striking properties for
our present purposes are i. its selective permeability,
2. its polarised condition, 3. its asymmetry and 4. the
"active patches" present within it.

In a later chapter the permeability of the cell mem-
brane will be dealt with in more detail. It is sufficient
for the moment to know that whereas diffusion is rapid
through the cytoplasm and through the nuclear sap,
permeation of the cell membrane can be a very slow
process, and it is possible that this is also true of the
nuclear membrane.

The polarised condition of certain cells, such as those
of nerve and muscle, is at present difficult to understand
in relation to the energy supplying systems which main-
tain this state of polarisation. But the development of
further understanding of this is of vital importance for

PHYSICO-CHEMICAL ASPECTS 19

the analysis of drug action, because as a result of this
polarisation, the cell membrane becomes a highly labile
system capable of giving almost explosive responses to
certain drugs such as acetylcholine, or alternatively of
losing its labile character in the presence of other drugs.
Possibly connected with the lability of polarised mem-
branes is the presence in many cells of small patches
which are selectively permeable to certain substances.
It may be that the action of acetylcholine is not an action
generalised over the whole surface of the responding
cell, but is confined to small active patches. Similarly
the stimulating action of light, which is mediated by
visual purple, may involve a sensitive system localised
in active patches in the cell membrane.

The Dynamic Condition of Cell Constituents. Studies
made with isotopes, though still very far from complete,
have shown that practically all the atoms present in the
apparently stable structures of a cell are being fairly
rapidly exchanged with other atoms of the same type.
Thus, although many features of cell architecture when
examined by histological or cytological methods may
appear stable, in fact every part of the cell appears to be
in a dynamic condition. Every enzyme molecule, every
protein molecule, every nucleic acid molecule, and prob-
ably every part of a gene is in a state of constant change
on the chemical level. It is very easy to see that drugs
may find sites of action of the greatest importance in

Cell Physiology 2

20 THE CELL AS A PHYSICO-CHEMICAL UNIT

places where they can interfere with these processes of
degradation and rebuilding of the constituent molecules
of the cell. It is probable that permanent interference
with any one stage in these processes will eventually
cause the death of the cell.

Defence Mechanisms

Many of the defence mechanisms of the cell are physico-
chemical in nature. They may in some measure be di-
vided into mechanisms of short term importance and
of long term importance. Among the short term mecha-
nisms available to the cell for dealing with foreign bodies
are concentration in vacuoles or granules, and detoxi-
cation processes. A foreign substance which becomes
introduced into the cytoplasm of the cell may be prac-
tically removed from most of the cell and concentrated
either in vacuoles, as is often the case with neutral red,
or in granules as is often the case with trypan blue. As
a result of these processes, most of the rest of the cell
can function unaffected by the foreign substance. Al-
though these processes are only obvious in the case of
coloured substances, there is, of course, no particular
importance in a substance being coloured. A colourless
substance is concentrated in the same ways as are col-
oured substances. Then, possibly after a substance has
been concentrated in the interior of a cell in this way, it
may be subjected to detoxication processes. Often these

DEFENCE MECHANISMS 21

processes involve the addition to the molecule of a polar
residue, such as glucuronic acid or sulphuric acid. As a
result of this type of procedure, the chemical and physi-
co-chemical character of the foreign substance may be
altered so as to make it comparatively innocuous. Thus
menthol in the body is largely converted into menthol
glucuronide, which is devoid of toxic properties and
also has much less surface activity than has menthol it-
self. Also the increase in polarity greatly reduces the
probability that the detoxicated molecules will re-enter
the cell once expelled after detoxication.

An alternative method of detoxication is frequently
encountered in the form of destructive reactions, in
which the foreign molecule is broken down into smaller
and less toxic compounds. These processes are usually
enzymic. Thus toxic amines are often detoxicated by
amine oxidases.

Amongst the longer term processes which may develop
on the physico-chemical level, there are included in-
creased efficiency in detoxication of a given foreign sub-
stance, the development of ''resistance" to the foreign
substance, and the development of antibodies.

The development of resistance probably takes many
forms. It may consist of an increase in ability to detoxi-
cate or to destroy a foreign body. This type of resistance
bears many formal semblances to the development of
adaptive enzymes in bacteria and yeasts. There is also
some evidence that the development of resistance may

22 THE CELL AS A PHYSICO-CHEMICAL UNIT

involve changes in the pernieabiUty of the cell mem-
brane, reducing the ease with which the foreign sub-
stance is able to penetrate into the cytoplasm.

The activity of antibodies may take two rather con-
spicuously different forms. The antibody may be re-
leased from the cell and combine with the foreign body
or foreign organism, before the latter reaches the cell.
Or alternatively the antibody may be present on the
outside of the cell membrane, and react with the foreign
body, to prevent its penetrating into the cell.

Two rather obvious conclusions follow from some of
the arguments advanced above. The first is that, as the
result of secretory activity of the cell, to give the average
concentration of a drug in a cell is practically meaning-
less, since it is very improbable that any drug is uniform-
ly distributed through the cell as a whole. Secondly,
there is a competition between the build-up of a sub-
stance inside a cell, and the rate at which it is removed
by detoxication or other processes. Both these points
must receive proper consideration if the action of drugs
is to be fully understood.

Self-reproducing Bodies

One of the most characteristic activities of biological
systems is their ability to reproduce themselves. The
cellular bodies which we at present anticipate may be
able to do this, or know are able to do this, are nuclear

INTEGRATION 23

genes, plasma genes, viruses, and adaptive and other en-
zymes. A drug may act upon these bodies, preventing
their self- rep reduction, or causing them to reproduce in
a new manner. Either of these activities would be ex-
pected to produce pronounced changes in the cell. The
problems involved in instances such as these, where
drugs are interfering with the genetical control of the
cell, will also be dealt with in more detail later.

Integration

Even from the incomplete account which has been
given of the physico-chemical organisation of the cell it
is clear that each particular region of the cell consists of
a complex interlocking of very many simultaneously
active physico-chemical systems. Each particular region
of the cell has its properties defined by a vast group of
variables, some of which are linked and some of which
are independant. Our understanding of these is very far
from complete. In some instances the necessary physics
and chemistry is almost completely unknown. In very
few instances are we able at present to deal quantitative-
ly with these variables. When sufficient information is
available to permit completely quantitative treatment,
it is likely that the system will be so complex that it will
be impossible to utilise this knowledge without the aid
of electronic calculating machines.

24 THE CELL AS A PHYSICO-CHEMICAL UNIT

REFERENCES

Adam, N. K., 1941: The Physics and Chemistry of Surfaces (Oxford

University Press, London).
Bernal, J. D. and Fankuchen, A., 1937: Nature 139, 923.
Bourne, G., 1950: Cytology and Cell Physiology (Clarendon Press,

Oxford).
Bracket, J., 1944: Emhryologie Chimique (Masson, Paris).
Clark, A. J., 1929: The Mode of Action of Drugs on Cells.
Clark, A. J., 1937: General Pharmacology.
Danielli, J. F. and Davies, J. T., 195 1: Advances in Enzymology, 11

(Acaderti'C Press, New York).
Darlington, C. D., 1937: Recent Advances in Cytology (Churchill,

London).
Davson, H. and Danielli, J. F., 1943: Permeability of Natural Mem-
branes (Cambridge University Press, London).
Gray, J., 1931: Experimental Cytology (Cambridge University Press,

London).
MiRSKY, A. E. and Ris, H., 1948: .5'- Gen. Physiol, 31, i.
Stedman, E. and Stedman, E., 1947: Symposia Soc. for Exp. Biol. I.
Symposia: Cold Spring Harbour Symposia (Darwin Press, NewBedford);
Society for Experimental Biology (Cambridge University
Press, London).
1947: Nucleic Acid.
1948: Growth.

1949: Selective Toxicity and Antibiotics.
Wilson, E. B., 1928: The Cell in Development and Heredity (Mac
Millan, New York).

CHAPTER II

Possible Actions of Drugs on Surfaces

Ionic Interactions

Ions may react with the charged groups of surfaces. The
molecules composing a surface may have charges upon
them which can be represented either as an electrostatic
dipole, a fully ionised group, or as a combination of
dipoles and fully ionised groups. When an interfacial
layer of molecules is composed mainly of dipolar mol-
ecules, the electrostatic potential difference between the
two phases is commonly of the order of 500 mV. Ions
may affect the packing of these molecules if they come
close to or are adsorbed upon the interface. The effect
of the ions is, of course, in such a case mainly mediated
by the interactions between the ionic charge and the
dipoles of the molecules. Where fully ionised groups
are present in the interface, such as carboxyl, phosphate
and amino groups, the action of ions is mainly mediated
by the interaction between the ion and the ionised group
of the interface. The effect of ions upon an ionised sur-
face is commonly greater than the effect of ions upon a
surface composed of dipolar molecules. As a result of
the interaction with ions the structure of a surface and

26

ACTIONS OF DRUGS ON SURFACES

its physical properties may be profoundly changed. For
example, there may be large changes in viscosity. A
monolayer of palmitic acid is a liquid, or even a gaseous
film upon an alkaline solution of sodium chloride. But
when the underlying cation is calcium, the monolayer is
an extremely viscous liquid or a solid. The effects of
ions upon monolayers can be readily studied by foU-

-O

■O
■O
-O
-O
-O

10- 20 A. U.

Fig. 4. Structure and thickness of an oil-water interface. The layer of
oil molecules oriented at the interface is 10-20 A.U. thick, and (in solu-
tions of uni-univalent electrolyte) the thickness of the electrical double
layer 6 is 3.i/>'c A.U., where c is the electrolyte concentration. The
lower part of the diagram indicates the variation in electrostatic potential
V due to the charge on the surface, at various distances from the surface.
At the point x=«5, xp=y)ole where e is the base of natural logarithms. The
excess of ions at any point x due to the charge on the surface is propor-
tional to ev: the greater part of the excess lies within the double layer

of thickness 6.

IONIC INTERACTIONS 27

owing the changes in viscosity of the surface film.

As a result of the forces operating between a charged
surface and the ions present in an underlying medium,
the ionic composition of the interfacial region may be
very different from that of the surrounding bulk phase.
To give an example: if a monolayer of palmitic acid is
spread upon a solution containing 2,000 sodium ions to
I calcium ion, in the interfacial region the ratio is of the
order of 0.3 sodium ions to one calcium ion (Danielli
and Webb). This large difference in the ionic ratio is
partly caused by the fact that the charge on the surface
attracts multivalent ions much more strongly than it
attracts univalent ions. But this is only part of the cause.
In the case of the sodium palmitate monolayer just men-
tioned, if the electrostatic effect were entirely responsible
for the difference in ratio between the surface and bulk
phases, the ratio in the surface would be 100 sodium
ions to I calcium ion : the difference between this figure
and the actual figure of 0.3 sodium ions to i calcium ion
is due to a second factor. The second factor is the for-
mation of unionised complexes between certain ions and
groups in the surface.

In the case of egg albumin molecules in solution it has
been possible to carry this analysis to a quantitative
conclusion (Danielli). The closed circles of Fig. 5,
show experimental values obtained for sodium : cal-
cium ratios in ovalbumin solutions. Ultrafiltrates were
prepared from the solutions, and the difference between

28

ACTIONS OF DRUGS ON SURFACES

these ultrafiltrates and the composition of the initial
solutions is due to the adsorption of ions at the surface
of the ovalbumin molecules. The crosses of Fig. 5 show
the results which would have been obtained if the excess
of ions at the surfaces of protein molecules had been

Fig. 5. The adsorption of calcium ions upon ovalbumin molecules, p^
is plotted horizontally, and the ratio: calcium in protein solution/
calcium in ultra filtrate, vertically, o experimental values; X calculated
for electrostatic binding of calcium ; ® calculated for formation of cal-
cium-carboxyl complex with electrostatic binding

produced by electrostatic factors only: clearly only part
of the effect can be attributed to electrostatic forces. The
open circles of Fig. 5 are calculated on the assumption
that the concentration of calcium at the surface of a pro-
tein molecule is raised by electrostatic interaction be-
tween the surface and the ions, and that the calcium ions
enter into an equilibrium of the following type with the

IONIC INTERACTIONS

29

carboxyl groups present at the surface of the ovalbumin
molecules.

Ca++ + R.COo- ±> R.C02Ca+

(I)

It will be seen that the results obtained in this way are in
excellent agreement with the experimental results.

I therefore suggest that in the mechanisms which we
have just studied we have a rational approach to some
aspects of the interaction of ions with surfaces. We shall
now consider two examples of this.

The sodiumi calcium ratio. It is very well known, indeed
so well known that of recent years no explanation has
been sought, that physiologically balanced salines, i.e.
salines which will support the life of tissues in a rela-
tively normal way, contain something of the order of
100 sodium ions to i calcium ion. It is difficult to con-
ceive any physico-chemical mechanism which would

TABLE I
THE RATIO [Na"'"]:[Ca"'""''] in physiological fluids, and at

THE SURFACE

OF CERTAIN CELLS

Ratio in fluid
medium

Ratio at surface

Red blood cell

Polymorphonuclear leucocyte
Arbacia egg
Astenas egg

50

25
2S

2.6
3.8
0.8
1.2

3© ACTIONS OF DRUGS ON SURFACES

permit loo sodium ions to enter into a directly balanced
relation with one calcium ion. But if the site of action
of the ions is not in a bulk phase, but at surfaces such
as the cell surface and the surface of protein molecules,
then the position is radically changed. For example, if
one calculates the ratio of sodium : calcium ions at the
surfaces of cells in their normal environments, one finds
that whereas the composition of the bulk phase surround-
ing the cells has a ratio of the order of loo sodiums
to I calcium, the ratio at the surface is of the order of
I : I (see Table i).

This then is a plausible theory for the explanation of
the balanced action of sodium and calcium in physiolo-
gical systems.

The Oligodynamic Effect of Heavy Metals. Let us sup-
pose that the toxicity of a heavy metal is caused by the
formation of unionised complexes between a surface such
as that of a protein molecule and the metal ion. Then in
the surface we have the following reaction:

Pr- + M+ ^ MPr (2)

[Pr-j[M+]

From the conditions we have just stated it follows that
equitoxic concentrations of different metals must give
the same value for [MPr]. In equation (3) the terms

IONIC INTERZCTIONS

31

[Pr~] and [MPr] will be constant for equitoxic concen-
trations of the different rnetals. Consequently if K^,
were known we could calculate the values of the equi-
toxic concentrations of the different metals, which are
represented by [M^] . Now the equilibrium constant is

rN^

*3

volts

+2

■H

° o^ i. °

\ dBi
• ■ o

^^«

oo

0.1 001 0001 0.0001 000001 0.000001

Molarity

Fig. 6. The toxicity of heavy metals to various systems. The logarithms
of the equitoxic concentrations of various metals are plotted against
their standard electrode potentials. ■ 50 % inhibition of papain; a 50 %
inhibition of catalase; A killing of Paramecium: % killing of Fundulus
egg; o killing oi Polycellis nigra (planarian).

related to the standard electrode potential of the metal
concerned (the reasons for this are given elsewhere,
Danielli, 1946). If we substitute for K^ in equation
(3) an appropriate term for the standard electrode po-
tential Em, and rearrange our equation we end up with

Em = 0.058 log [M] + constant, (4)

i.e. if we plot log[M] against E^ we should obtain a

32 ACTIONS OF DRUGS ON SURFACES

Straight line. Fig. 6 shows the action of heavy metals
on a number of biological systems. It is clear that the
relationship we have just obtained is approximately
true.

Thus we have at present a plausible theory of the oli-
godynamic action of heavy metals, and of the differences
between their actions. It is an integral part of this theory
that the action involves the formation of an unionised
complex with ionogenic groups^ at a surface. The types
of ionogenic groups which we may particularly consi-
der as likely to be involved are phosphate, carboxyl and
SH. Much remains to be done before this theory can be
regarded as established.

Dipole Interactions and Complex Formation

When molecules are present side by side in a monolayer
the organisation of the monolayer is profoundly affected
by the forces operating between adjacent molecules. If

^ It is of interest to consider mechanisms whereby a charge upon a
surface may arise. The mechanisms are four in number: i. The partition
of ions between the two phases; 2. The orientation of dipolar molecules
at the interface ; 3. The adsorption of ions; 4. The ionisation of ionogenic
groups.

Mechanisms 2, 3, and 4 are obvious, but a word of explanation is
required for mechanism i. The different ions have different partition
coefficients between say oil and water. But in order to preserve electrical
neutrality the actual concentration of positive and negative charges
in both bulk phases must be equal. This equality is brought about
by the building up of an electrostatic potential at the interface which
effectively modifies the partition coeflftcients of the different ions.

DIPOLE INTERACTIONS AND COMPLEX FORMATION 33

neighbouring molecules are ionised, the forces involved
may be rather large, whereas the weakest forces encoun-
tered are Van der Waals' forces. Usually the forces which
have to be considered can be ranked as follows: ion-ion
> ion— dipole > dipole— dipole > Van der Waals' inter-
actions. Since different molecular species have different
distributions of attractive and repulsive forces in their
molecules, at times the interaction between two adjacent
molecules may be sufficiently selective to entitle us to
regard the interaction as resulting in the formation of a
two-dimensional molecular complex (Schulman and
Rideal). As a result of such complex formation, changes
in at least five variables can be observed. These are: the
packing of molecules, the surface pressure, the surface
viscosity, the surface potential and the velocity of che-
mical reactions in the monolayer. The changes in these
variables quite clearly are likely to be of importance as
providing possible mechanisms of drug action.

Chemical reactions at interfaces have recently been
examined, with the emergence of some interesting points.
For example, the hydrolysis of trilaurin by hydroxyl
ions has been shown to be sensitive to surface pressure
changes. At a surface pressure of 5.4 dynes per cm the
activation energy of the reaction is 10,000 calories,
whereas at 16.2 dynes per cm the activation energy has
risen to 15,000 calories. The actual orientation of a mol-
ecule at the interface may also exercise a profound
effect upon the velocity of reaction. For example, with

34 ACTIONS OF DRUGS ON SURFACES

thesimple esters, the rate of hydrolysis may vary, accord-
ing to the configuration of the ester at the interface, so
as to have a velocity as high as 0.18 min"^, or as low as
0.005 min~i under a given set of conditions. Similar
results have been attained for the rates of oxidation
in interfaces of substances containing double bonds, by
permanganate.

We see, from consideration of these results, that the
interaction of an enzyme with its substrate may be affect-
ed by the orientation and other physical conditions of
the substrate and that these in turn may be affected by
drugs.

Specificity in Surface Reactions

Many drugs, such as adrenaline, veratrin, acetylcholine
and cocaine act in such very low concentration that it
is not at once apparent how a sufficient concentration of
these molecules can arise at their site of action to produce
a significant physico-chemical effect. However, if the
action of these drugs is at surfaces it is possible to see
both how an adequate concentration may arise, and why
there is a high degree of specificity in their action. Each
drug contains ionising groups, polar groups and practi-
cally non-polar hydrocarbon groups. Hence a drug has
an intrinsic capability of being adsorbed at a surface,
and when adsorbed, of altering the properties of the sur-
face. E.g. let us consider the concentration of adrenaline,

SPECIFICITY IN SURFACE REACTIONS 35

which may arise at a surface. This is given by Boltz-
mann's theorem,

C.

E
RT

e is)

where C^ = concentration at the adsorbing surface; Cj
= bulk concentration; E = energy of adsorption; R =
= gas constant; T = absolute temperature.

The energy E may be regarded as made up of three
components, one associated with the ionic groups, one
with the polar groups and one with the non-polar groups.
Minimal values for adrenaline are

Adrenaline ionic polar non-polar total E Cs/Cb

HO OH ^°° ^°°° ^^°° ^^°° 2x10^

H0()— CH-CH2-NH-CH3

The values taken for the energies are minimal. Yet when
these energies are summed, the concentration of adrena-
line which may arise at a surface is seen to be of the
order of 10^ times that which is found in the bulk phase.
Under favourable conditions the energies may well be
larger and could easily give rise to a relative concentra-
tion of the order of lo^^.

But before it is permissible to sum the energies in the
way in which we have just done, it is necessary to pro-
vide an interface at which all three mechanisms for ad-
sorption may become effective to an optimal degree.
This means that the surface at which the drug is ad-
Cell Physiology 3 . -::» r- 1 i-i

\2

^^4i^^j:y

36 ACTIONS OF DRUGS ON SURFACES

sorbed must present an organisation of ionising groups,
polar groups and non-polar groups as specific as that
which is to be found in the drug itself. If this criterion
is fulfilled the possible energy of adsorption is large.
But a group in the wrong position, or having the wrong
orientation, may readily prevent the dove-tailing of the
drug and the surface and thus prevent many of the sites
of potential adsorption becoming effective. Thus the
presence of a methyl group in the wrong place may read-
ily reduce the ease of adsorption of a drug by several
thousand calories, and cause it to be relatively inert.
This theory may be regarded as a plausible one. It is
very difficult to establish such theory.

Mechanism of Lysis

One of the aspects of drugs which has often attracted
attention is the ability of certain substances to cause cy-
tolysis. Cytolysis usually involves a direct action upon
the plasma membrane of the cell. When lytic substances
are considered from the point of view of surface ch em-
istry it is seen that they are likely to produce their action
partly by the formation of complexes with the mole-
cules constituting the cell membrane, and partly by dis-
solving in the lipoid layer of the plasma membrane. As is
indicated in Table ii, some of these substances probably
react primarily with proteins to form complexes, some
react primarily with lipoids and some react with both.

ACTION OF OESTROGENS ON MONOLAYERS 37

There are a few substances, such as chloroform, which
probably act mainly by dissolving in lipoids. There ap-
pear to be no instances of lysis which do not appear to be
readily explicable in terms of the type of behaviour
which is found at interfaces.

TABLE II

POSSIBLE MODES OF ACTION OF LYTIC SUBSTANCES

React with
proteins

React with
lipoids

Dissolve in
lipoids and de-
nature proteins

Antibodies

+

Trypsin

+

Polyhydric phenols

+

Heavy metals

+

+

Soaps

+

+

Digitonin

+

Lysolecithin

+

Lecithinase

+

Bacterial toxins

+

Chloroform

+

Phenol

+

The Action of Oestrogens on Monolayers

It is well known that many dihydroxy stilbenes have
oestrogenic activity. Schulman and Rideal have stud-
ied the action of these substances on protein mono-
layers, and have claimed that there is a maximum of
oestrogenic activity coinciding with maximum activity
on the protein monolayer (Fig. 7). This is an instance of

38

ACTIONS OF DRUGS ON SURFACES

a theory where there is an interesting correlation be-
tween a complex physiological effect and a simple phys-
ico-chemical phenomenon. As Dale has pointed out,
this coincidence cannot be a sufficient explanation of the

u

K

^100

^^^^=4=

=^

^

^».,,_^^ ~^,.^__^

7~~ '

....,^__*^

200

-— ^£__^

300

,

"■ —^ ^

30

C

' 10

20

Minutes

Fig. 7. The relationship between reduction of surface potential of pro-
tein monolayers and oestrogenic activity in the stilbene series, a. 4,4'
dihydroxy stilbene (i : 150); b. 4,4' dihydroxy dimethyl stilbene
(i : 40,000) ;c. 4,4' dihydroxy ethyl stilbene ( I : 5,000) ;d. 4,4' dihydroxy
ethyl methyl stilbene (i : 1,000,000); e. 4,4' diethyl stilbene (i :
3,000,000); f. 4,4' dihydroxy he xadienediphenyl( I : 2,500,000) ;g. 4,4'
dihydroxy dipropyl stilbene (i : 100,000). The reduction in surface
potential increases steadily from a — f : substances e and f, which have
the maximum oestrogenic activity, also combine most readily with the
protein film. Substance g replades the protein film.

physiological action of oestrogens in producing oestrus,
for this change is restricted to certain types of cells of
the body, whereas all cells are lavishly provided with
protein monolayers. It is not, however, beyond the
bounds of possibility that the action of oestrogens on
monolayers is the basis, or part of the basis, of their

EFFECT OF MICELLE FORMATION 39

general capacity to promote cell division. Very much
more remains to be done, however, before these results
of ScHULMAN and Rideal can be adequately claimed to
be linked with physiology.

The Effect of Micelle Formation

In aqueous solution, when the properties of a solute are
dominated by its polar groups, as is the case with methyl
alcohol, the solute molecules are dispersed mainly as
single molecules. But when there is a great excess of non-
polar hydrocarbon groups in the molecule, practically
speaking no single molecules exist in an aqueous solu-
tion of the substance concerned. Instead, the molecules
are organised so that the non-polar parts are tucked
away into micelles, which may be spherical or take the
form of bimolecular sheets. The polar groups of the
molecules are mainly to be found in the interface with
the water, so that the non-polar part of the molecules
does not in fact necessarily come into direct contact
with water molecules. Substances forming solutions of
this type are stearic acid, cholesterol and tripalmitin.

Where there is a balance between polar and non-polar

+
groups, as in CH3(CH2)9 N Mcg, there is an equilibri-
um between single molecules and micelles, which is
quite dynamic. If the surface tension of such a solution
is studied as a function of concentration it is found that
as the concentration is increased, at first the surface

40

ACTIONS OF DRUGS ON SURFACES

tension falls very rapidly, until the concentration is
reached at which micelle formation commences (Fig. 8).
At this point, the curv^e rapidly flattens out, so that the
surface tension becomes almost independent of concen-
tration. Once the curve has become flat, the concentra-
tion of single molecules does not increase with increase

^■Concentration

Fig. 8. The relationship between concentration, surface tension and

+
micelle formation for a substance such as CH3(CH2)9NMe3

in concentration. What does increase is the concentration
of micelles.

As a result of the possibility of micelle formation, the
following consequences may occur.

I . If a substance can give rise to a micellar solution,
and if the physiological action of the substance is a
function of the concentration of single molecules in so-
lution, then over the range of existence of micelles the
action of a substance may be independent of its concen-
tration. This is because over the range of existence of

EFFECT OF MICELLE FORMATION 4I

micelles, the concentration of single molecules may be
independent of the total concentration, increases in con-
centration resulting merely in formation of more mi-
celles. The transition of micelles into single molecules of
solute may have a very high temperature coefficient.
Consequently the action of such a drug may also have
a very high temperature coefficient.

2. If the activity of a homologous series of drugs is
plotted against the number of carbons in the molecule,
it is frequently found that somewhere in the region of
nine carbon atoms a maximum of activity is reached. As
the number of carbons in a series is increased, so also the
ease of adsorption of the molecule increases, and conse-
quently its activity per molecule rises. But as the num-
ber of carbons increases, so does the ease of formation
of micelles, and it commonly happens with drugs con-
taining an aliphatic carbon chain that the ease of forma-
tion of micelles increases more rapidly than the surface
activity increases. Thus a point is reached at which mi-
celle formation occurs before a concentration of the drug
as single molecules is reached at which its physiological
activity can become manifest. Hence the maximum in
the curve of activity plotted against number of carbons

(Fig. 9)-

3. If a second substance is present, which can form
micelles, a drug may be inactivated by combination with
the micelles. This is illustrated by Fig. lo. In this dia-
gram the rate of penetration of hexyl resorcinol into

42

ACTIONS OF DRUGS ON SURFACES

Ascaris is Iplotted against the concentration of detergent
present in the same medium. At first the penetration
of the drug is increased by the presence of the detergent,
but as the concentration of the detergent is increased a
maximum is reached and the rate of penetration falls

n

Fig. 9. The relationship between physiological activity, micelle for-
mation and n, in a series CH3(CH2)nX, where the physiological acti-
vity is a function of the concentration of single molecules

off practically to zero as the detergent concentration is
still further increased (Trim and Alexander). In the
same figure, the surface tension of the solutions is plotted
also, and it will be seen that the point at which the rate
of penetration begins to fall coincides roughly with the
onset of micelle formation. It therefore seems very prob-
able that the decline in rate of penetration of the drug
is due to its forming a complex with the detergent when
the latter is present as micelles, and that this complex is
unable to penetrate the cuticle of Ascaris.

LONG-RANGE FORCES

43

It is plain that the phenomenon of micelle formation
enables us to furnish a plausible explanation of some of
the peculiar properties of certain drugs. It is necessary

Micelle formation

Surface tension

Permeability to
hexyl resorcinol

»- Concentration of soap

Fig. lo. The prevention of permeation of hexyl resorcinol into Ascaris
by micelle formation in a soap solution (After Trim and Alexander)

to remark here that the complete establishment of these
theories requires more than the demonstration of a coin-
cidence between micelle formation and the onset of a
change in the action of a drug.

Long-Range Forces

Recently the biologist has become interested in rel-
atively long-range forces, i.e. forces operating over dis-
tances of the order of 25 A.U. up to several ^. Bernal
and Fankuchen have shown that in some types of col-

44 ACTIONS OF DRUGS ON SURFACES

loidal solution molecules may be oriented parallel to
one another by long-range forces. More recently
ROTHEN has obtained results which can be interpreted
to mean that relatively specific forces, such as those be-
tween antibodies and antigens, may extend over a dis-
tance of at least loo A.U. The biologist is tempted by
such phenomena as the adlineation of chromosomes in
meiosis, and the reaction of cells to one another, to pos-
tulate similar forces extending up to several microns. If
the field of action of these forces is as extended as some
suppose, then they must be of fundamental importance
in such phenomena as differentiation, chromosome
mechanics, fibre adlineation and enzyme action. From
the scanty information which is available it is already
clear that the operation and the specificity of these for-
ces is greatly affected by the net charge on the molecules
concerned, and also by its detailed distribution. It is
therefore plain that we have here a fertile field for the
study of the action of drugs. But our knowledge is at
present so restricted that it is not possible to do more
than indicate the immense possibilities which exist here
for future research.

REFERENCES 45

REFERENCES

Abramson, H. a., 1934: Electrokinetic Phenomena (Chemical Catalog

Company, New York).
Adam, N. K., 1941: The Physics and Chemistry of Surfaces (Oxford

University Press, London).
Bernal, J. D. and Fankuchen, A., 1937: Nature, 139, 923.
Dale, H., 1943: Trans. Farad. Soc, 39, 320.
Danielli, J. F. and Webb, D. A., 1940: Nature, 146, 197.
Danielli, J. F., 1941: Biochem. J., 35, 470,
Danielli, J. F,, 1944: J. Exp. Biol., 20, 167.
Danielli, J. F. and Davies, J. T., 195 1: Advances in Enzymology, 11

(Academic Press, New York).
RiDEAL, E. K., 1945: J'. Chem. Soc, 423.
RoTHEN, A., 1947: y. Biol. Chem., 168, 89.

ScHULMAN, J. H. and Rideal, E. K., 1937: Proc. Roy. Soc, B 122, 29.
Symposia: 1949: Surface Chemistry (Butterworth, London).

1949: Selective Toxicity and Antibodies (Symposia Soc. for
Exp. Biol., III).
Trim, A. R. and Alexander, A. E., 1946: Proc. Roy. Soc, B 133, 220.

CHAPTER III

Membrane Permeability and Drug Action

Introduction

There have been many academic studies of the perme-
ability of natural membranes, but very few direct
studies on permeability to drugs. Consequently, most of
what can be said on this topic is based on permeability
to molecules which are not usually regarded as drugs. In
approaching this field we must distinguish between dif-
fusion and secretion. Diffusion is the movement of mol-
ecules produced by thermal agitation; it now has a
quantitative theory. Secretion is a process involving the
expenditure of energy by a living organism to move
molecules from one place to another: there" are no quan-
titative theories of secretion.

The importance of permeability studies may be seen
from consideration of the sites at which a drug may act.
Even when some process such as absorption from the in-
testine is not involved, a drug always has to penetrate a
cell membrane unless its action is on the external sur-
face of a cell. In general, a drug which has penetrated
a cell membrane may combine with a receptor group and
also may be detoxicated more or less rapidly. For effi-

INTRODUCTION

47

cient drug action the rate at which a drug penetrates into
a cell must be large compared with the rate of detox-
ication. Commonly drugs must also pass other membranes
in addition to those of cells, such as the complex mem-
branes composing the intestinal epithelium, the cuticle
of a parasite etc. It is thus clear that there is much point

Concentration of Ag*

Fig. II. The action of Ag on the invertion of sugar, (a) Inhibition of
purified invertase; (b) inhibition of invertion by yeast cells; (c) amount

of Ag taken up by yeast cells

in knowing how changes in the physical structure of a
drug may modify its ability to permeate various types of
membrane.

Even when we are studying the effect of drugs upon
simple cell suspensions, such as those of bacteria and
yeasts, the action may prove to be much more compli-
ated than when we are dealing with those homogenates
which are dear to the biochemists. Fig. ii shows the

48 MEMBRANE PERMEABILITY AND DRUG ACTION

action of silver on the invertion of sugar. Curve a shows
that the action of silver upon free invertase is immediate,
and always reduces the action of invertase. On the other
hand, as is shown by curve h, when silver is added to a
suspension of yeast cells, low concentrations of silver
actually increase the rate of invertion of sugar, and it
requires a relatively substantial concentration of sugar
to destroy the enzyme action. It is possible that the
first action of silver on the cells is to increase their per-
meability to sugar, and thus enable more substrate to
obtain access to the enzyme than would otherwise be
the case. When the concentration of silver is increased,
the enzyme itself is affected. These results show that the
permeability factor, even in a simple system, may be
involved in a relatively complicated manner. In the fol-
lowing pages we shall to a large degree ignore compli-
cations and deal with permeability problems exclusively.
But it is important to remember that in so doing we are
indulging in an artificial abstraction.

Membrane PeTmeability and Drug Structure

When considering the permeation of drugs into cells,
three major questions arise. These are:

1. Can the structure of the drug be modified without
destroying its therapeutic activity?

2. Do different cells differ in permeability to the same
substance ?

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 49

3. Can the structure of a drug be modified to give pre-
dictable changes in permeability ?
The answer to the first of these questions varies with
the type of a drug which is under consideration. It is
well known that the structure of the sulphonamides may

TABLE III

THE PERMEABILITY OF THE CELLS OF Chora TO DIFFERENT

SUBSTANCES

Living cell

Dead cell

Equal water
cylinder

Methyl alcohol

1-3

0.8

0.27

Urea

320

0.9

0-34

Acetamide

24

1.2

0.38

Glycerol

1,700

1.9

0.49

Trimethyl citrate

5-5

2.2

0.67

Sucrose

50,000

41

0.92

The figures given are the times taken for the average concentration inside

a cell to reach 50% of that outside. For comparison similar values are

given for dead cells, and calculated values for a water cylinder of the

same size as a Chora cell. (After Collander)

be altered within very wide limits without destruction of
activity. On the other hand among, say, the anti-malarial
drugs, only comparatively small changes in structure
are possible without loss of activity. Thus each group
of substances must be considered as a separate case. So
far as can be seen at the moment, unless the structure
of a drug can be modified without change of activity,
comparatively little can be done to make the best of the

50 MEMBRANE PEPMEABILITY AND DRUG ACTLON

differences in permeability which may exist between the
cells of the host and of the parasitic organism.
On the second question, as to whether cells differ in
their permeabilities, there are two groups of data to be
considered: the permeability of a given cell to different

TABLE IV

THE PERMEABILITY OF VARIOUS CELLS TO
DIFFERENT SUBSTANCES

13
u

u

X

O

bo
be

<3
-^

Chara
ceratophylla

Bact.
paracoli

Gregarina
sp.

Melosira
sp.

Beggiatoa
mirabilis

Trimethyl

citrate

6.7

3.0

Glycol

0.2

0.7

1.2

0.7

0.4

1-4

Urea

8

—

0.1

0.08

0.3

0.04

1.6

Malon-

amide

0.004

0,03

0.02

Glycerol

0.002

0.005

0.02

0.06

0.02

0.03

1.06

Erythritol

O.OOI

0.005

O.OI

0.8

Sucrose

0.0008

~^'~

0.006

0.1

substances and the permeability of different cells to the
same substance. Table iii indicates very wide differences
in permeability of the cells of Chara to different pene-
trating substances. It will be seen that the rate of per-
meation of the substances studied varies by a factor of
10*. But the range of permeabilities to different sub-
stances is in fact even greater than this, for it is incon-
venient to study the substances which penetrate ex-

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 51

tremely slowly. However, as in most instances substances
which penetrate very slowly will not have a drug action,
these more slowly penetrating substances are probably
of little practical importance.

Table iv shows the permeability of a number of dif-

TABLE V

THE PERMEABILITY OF DIFFERENT CELLS TO UREA AND

TO GLYCEROL

Urea

Glycerol

Ratio
urea/glycerol

Ox red cells

7.8

0.002

3,900

Chora ceratophylla

0.1

0.02

5

Plagiothecium denticulatum

0.004

0.003

1-33

Curcuma rebricaulis

0.002

0.002

I

Melosira sp.

0.04

0.03

1-33

Bacerium paracoli

0.08

0.06

1-33

Beggiatoa mirahilis

1.6

I.I

1-45

Gregarina sp.

0.7

0.02

35

ferent cells to various substances. In Table v some
figures are given for the permeability of certain cells to
urea and to glycerol. It will be seen from these results that
the permeability of different cells to the same substance
may vary by a factor of 100 or even 1000-fold. Moreover
the relative order of permeation of substances into dif-
ferent cells is not always the same. Thus some cells are
more permeable to urea than to glycerol, whereas others
are more permeable to glycerol than to urea. From this
it is apparent that considerable therapeutic advantages

Cell Physiology 4

52 MEMBRANE PERMEABILITY AND DRUG ACTION

can in principle be derived if the drug is constructed to
have permeabiHty properties which resemble those of
substances which readily penetrate into a parasitic cell,
but which penetrate much less readily into the cells of
the host.

Until recently, however, very few attempts have been
made, and these unsuccessfully, deliberately to modify
the structure of drugs so as to exploit the permeability
characteristics of different types of cell. The main
reason for this is that there was no general quantitative
theory correlating permeability of cells with the struc-
ture of penetrating substances. Without such a quanti-
tative theory it is really impossible to make much
headway. But recently theoretical studies have been
made which enable us to calculate the relative permeabil-
ities of cells to different molecular structures, to a first
approximation. It is convenient to divide the cases which
are met in practice into five groups.

The first group is that where permeation occurs by
bulk flow of fluid medium. As an instance of this we may
take the formation of the glomerular filtrate in the kidney
of mammals. Here all the crystalloical constituents of the
blood are filtered off through a membrane, the pore
size of which is very much larger than the diameter of
a crystalloid molecule. In such cases all the molecules
of a crystalloid character are carried along by the bulk
flow of the fluid medium. Consequently the only force
available to discriminate between different molecular

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 53

types is the electrostatic force which may arise by the
fact that the colloidal constituents of the blood are unable
to pass the glornerular membrane, whereas the crystalloid
ions which neutralise the charge on the colloids are able
to pass the membrane. From this it follows that where a
bulk flow occurs, as in an ultra-filtration, all molecules
of the same charge type display the same permeability,
provided they are small compared with the diameter of
the pores through which the bulk flow occurs. All un-
charged molecules have the same permeability. All uni-
valent positive ions have the same permeability, but
differ from the uncharged molecules. All univalent neg-
ative ions have the same permeability, but differ from
the uncharged molecules and from the univalent positive
ions, etc.

The second group is that in which permeation occurs
by thermal diffusion through membranes the pores of
which are large compared with the diameters of the
diffusing molecules. In this case the molecules diffuse
at different rates through the water filling the pores of
the membrane, whereas in the previous case all the
molecules were carried along by the bulk flow of the
fluid in which they were dissolved. When diffusing
through a membrane in this manner we find that the

relationship

PMVi = constant (6)

is obeyed (P = permeability, M = molecular weight
of diffusing molecule). Tables vi, Vii and viii show the

54 MEMBRANE PERMEABILITY AND DRUG ACTION

permeability of collodion membranes, chitin and of the
sulphur bacterium Beggiatoa ndrabilis. Two instances of
a collodion membrane are illustrated, one with a large
pore size and the other with a small pore size. It will
be seen that where pore size is large the permeability

TABLE VI

THE PERMEABILITY OF TWO COLLODION
MEMBRANES TO DIFFERENT MOLECULES

Membrane

Membrane

(a)

(b)

Methyl alcohol

6.9

5-2

Ethyl alcohol

7.8

2.0

Propyl alcohol

7-7

0.8

Butyl alcohol

7.3

0.7

Ethylene glycol

6.3

0.2

Glycerol

7.8

0.2

Glucose

7-3

<o.05

The values are of PM^^. Membrane (a) is
relatively permeable, and membrane (b) rel-
atively impermeable

falls off inversely as the square root of the molecular
weight, and equation (6) is obeyed to a first approxi-
mation. On the other hand, where the pore size is small,
the permeability falls off much more rapidly than is in-
dicated by equation (6). For membranes having a large
pore size it is clear that to a first approximation the
permeability may be calculated, provided that the per-
meability of the membrane to two or three other sub-

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 55

Stances is already known and that the molecular weight
of the drag is known. Table ix shows another interesting
case, where equation (6) is obeyed to a first approxi-
mation: this is the diffusion of anions through the walls
of the rumen of the sheep.

TABLE VII

THE PERMEABILITY OF CHITIN TO
DIFFERENT MOLECULES

P

pmv^

Formic acid

19

127

Acetic acid

13-5

105

Propionic acid

14

122

Butyric acid

13

123

The two groups which we have just considered are
those in which the molecule penetrating the membrane
does not in fact leave the water in which it is dissolved.
In the first case the permeating molecule was carried
along by bulk flow of the water and in the second case
thermal diffusion caused the permeating molecules to
move through the water filling the pores in the membrane.
The next two groups which we have to consider differ
from the first two groups in that the permeating mole-
cules actually pass from the aqueous phase in which they
are dissolved into the non-aqueous phase constituting the
membrane. In such cases the main resistance to perme-
ation may lie either at the membrane-water interfaces or

56 MEMBRANE PERMEABILITY AND DRUG ACTION

in the interior of the membrane. In the first case a diffus-
ing molecule finds its passage through the membrane
dominated by the difficulty of passing either from the
water into the membrane, or from the membrane into

TABLE VIII

THE PERMEABILITY OF THE BACTERIUM

Beggiatoa mirabilis to various

MOLECULES

P

pmv^

Glycol

1.4

II.O

Methylurea

1.2

10

Urea

1.6

12

Glycerol

I.I

10

Erythritol

0.84

9

Sucrose

0.14

2-5

TABLE IX

VALUES OF PM^^ FOR DIFFUSION OF ANIONS

THROUGH THE WALL OF THE RUMEN OF A

SHEEP

Acetate

Propionate

Butyrate

5.8
5-3

7-4

water. In the second case, passing the interfaces is not
the limiting factor: the difficulty is in passing through the
interior of the membrane. The group of cases in which
passage through the cell membrane interface presents
the main difficulty obeys the relationship

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 57

2500X
PMV^e RT /J5= constant (7)

where B = oil : water partition coefficient, and x =
number of unscreened CHg groups per molecule.
Table x shows the degree to which this relationship is

TABLE X

R MOLECULE

CELLS OF Chora ceratophylla

U 2 $00 X

VALUES OF PM^^e "^ FOR MOLECULES PENETRATING INTO THE

Erythritol

8.5

Urotropin

S-3

Methylolurea

10.9

Methylurea

8.3

Urea

5.8

Dicyandiamide

9-4

Glycerol

6.5

Lactamide

6.4

obeyed by molecules penetrating the cells of Chora cera-
tophylla. The adherence to the theoretical relationship
is sufficiently good for one to be able to calculate the
permeability to other molecules to a first approximation.
The fourth group of substances are those the main
resistance to the penetration of which lies in the in-
terior of the membrane. This group obeys the relation-
ship

PMVi/B = constant. ( 8)

Tables xi, xii and xiii show the degree to which this
relationship is in fact obeyed for the cells of Melosira,
Arhacia eggs and the rumen of the sheep.

Several comments must be made upon these results.
First, the constants in the various equations just given

58 MEMBRANE PERMEABILITY AND DRUG ACTION

are not the same for every cell type. A different constant
is obtained for every type of cell, and the constant varies
with species as well as with cell type. Thus for any par-
ticular cell, before the permeability to a drug can be
calculated, it is necessary to know the permeability to

TABLE XI

VALUES OF PM^^^/B FOR VARIOUS SUBSTANCES PERMEATING INTO

Melosira cells

Propionamide 7.2

Acetamide 2.0

Glycol 6.1

Glycerol 4.5

Methylurea 2.3

Urea 2.2

several other molecules. A second point which must be
emphasised is that as drastic changes are made in the
structure of the molecule, it is likely that one will pass
from molecules falling into one of the groups given
above to molecules falling into another of the groups.
For example, group four consists mainly of molecules
which permeate cells rather rapidly whereas group three
consists of molecules which penetrate rather slowly.

TABLE XII

VALUES OF PM^'^-jB FOR VARIOUS SUBSTANCES PERMEATING INTO

Arbada ova

Butyramide 5.4

Propionamide 5 . 9

Acetamide 9.2

Glycol II. 7

1.8

1 :2 Dihydroxy-

propane
1:3 Dihydroxy-

propane 2.8

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 59

With those cells which have a lipoid membrane deter-
mining their permeability, it commonly happens that as
the structure of the penetrating molecule is changed so
as to make it penetrate more slowly, a transference is
made from group four to group three. Another case

TABLE XIII

VALUES OF PM'^V^ FOR FATTY ACIDS P ASSING^THROUGH CELLS

OF THE SHEEP RUMEN

P

PM'^^IB

Acetic acid
Propionic acid
Butyric acid

3.5

10.8
27.6

2.7
3.3
3.4

which may be of practical importance is that of trans-
ference from group three to group two. This is illus-
trated by the results shown for the diffusion of sub-
stances through the sheep rumen. The sheep rumen
membrane consists of cells between which lies an inter-
cellular cement. Molecules such as the free fatty acids,
which can penetrate the cells rapidly, permeate the mem-
brane mainly by passing through the cells, though, of
course, some free fatty acid also passes through the
pores. The fatty acids passing through the cells encounter
the main resistance to permeation in the interior of the
cell membranes they pass through, and thus fall into
group three. But the fatty acid anions permeate the cell
membranes much less readily. Consequently passage

6o MEMBRANE PERMEABILITY AND DRUG ACTION

through the pores is much more important in the case
of the anions, and the anions to a first approximation
fall into group two.

In the fifth group we collect together all those instan-
ces where substances, or groups of substances, do not
conform to groups one to four or with the transition
stages between the different groups. These are the in-
stances in which secretory activity plays a part in deter-
mining the passage of the molecules across a membrane.
In such cases, directly or indirectly, the cells concerned
are expending energy to promote the passage of mole-
cules across a membrane. Usually the permeability of
cells in such cases is a non-linear function of the con-
centration of the molecule concerned. As the concen-
tration of the molecules which are secreted increases in
the cell environment, the secretory mechanism tends
to become saturated, and the rate of secretion increases
much more slowly than does the concentration in the
environment. The existence of secretion is also often
revealed by a degree of anoxia of the tissues concerned,
or by the moderate use of enzyme poisons which inter-
fere with the metabolic processes which provide the
energy for secretion, or even in some cases possibly with
the mechanism of secretion itself. At present our knowl-
edge of secretion is much too slight for us to be able to
predict with any confidence what the quantitative effect
will be of a change in molecular structure upon the rate
of secretion.

MEMBRANE PERMEABILITY AND DRUG STRUCTURE 6l

From the facts which have just been presented we
may draw the general conclusion that, subject to certain
reservations, the permeability of a cell to a drug can be
calculated to a first approximation. E.g. when the struc-
ture of a drug is changed by adding or subtracting
chemical groupings such as CHg, COOH, or NHg etc.,
we can calculate the order of magnitude of the change
in permeability which will ensue. In theory we can cal-
culate the permeability of a membrane provided we
know the details about its structure. Thus, for a porous
membrane, we need to know the size of the pores, the
thickness of the membrane and the numbers of the
pores. With a lipoid membrane, we need to know the
thickness of the membrane, the lipoid : water partition
coefficients of the permeating molecule and the effective
viscosity of the lipoid composing the membrane. Table
XIV shows the permeability of several cells compared
with the permeabilities calculated for a lipoid layer and
for a water layer of the same thickness as that of the cell
membrane. It will be seen that whereas the permeabil-
ity of the lipoid membrane is of the same order of perme-
ability as that of the cells, a water layer is about lo^
times more permeable. Since there is no known method
by which the viscosity of the interior of a cell membrane
can be determined directly, in practice one has to deter-
mine it indirectly by determining the permeability to
two or three substances. This, however, is in any case
necessary before it is possible to decide to which group

62 MEMBRANE PERMEABILITY AND DRUG ACTION

or groups the drugs in which we are interested belong.
The reservations mentioned above are of two kinds.
If a substance happens to fall into one of the types which
are secreted, we usually cannot calculate the effect of a
change in structure. Then also there are small patches

TABLE XIV

THE PERMEABILITY OF VARIOUS CELL MEMBRANES TO CERTAIN
SUBSTANCES, COMPARED WITH THE PERMEABILITY CALCULA-
TED FOR AN EQUIVALENT THICKNESS OF W^ATER, AND OF OIL
HAVING A VISCOSITY lOO TIMES THAT OF WATER

Arbacia
egg

Chara
cerato-
phylla

Plagio-

thecium

denticu-

latum

oil

5m/f
water

Propionamide

2-3

3.6

0.2

3.0

1.4 X 10"

Acetamide

I.O

1-5

0.7

0.8

1.8 X IO»

Glycol

0.73

1.2

0.3

0.7

1.7 X 10^

Urea

O.II

0.004

0.02

1.8 X IO»

Glycerol

0.005

0.02

0.0003

0.005

1.4 X 10*

Malonamide

0.004

0.0008

0.002

1.4 X IO»

Erythritol

O.OOI

0.00007

0.00007

1.2 X 10®

on the surfaces of at least some cells which appear to be
specially adapted to permit rapid permeation by a given
molecular species without permitting a similar advantage
to other molecular species. These patches may, of course,
be part of the cellular secretory mechanisms, but until
we know more about them it is wiser to assume that they
may be distinct from active secretory processes. The
fact that these reservations must be made should not be

ACCESS OF DRUGS TO ORGANS 63

taken as indicating that there need be any hesitation in
applying the conclusions derived from studies of cell
permeability to practical problems. What is necessary in
such circumstances is to keep an alert eye for the com-
plications which may arise from failure to obey the laws
of thermal diffusion across membranes.

Problems of the Access of Drugs to Organs

The permeability problems involved in the study of drug
action are far from limited to those encountered in the
study of the permeability of cell membranes. Frequent-
ly, the problem of access to an organ, or of absorption
from the digestive tract, constitutes a more important
difficulty than permeation into cells of either the host
or the parasite. The main practical problems which tend
to arise are: i. inadequate adsorption from the digestive
tract, 2. peculiar permeability properties which prevent
a drug reaching a particular organ, e.g. the brain, 3. se-
curing an effective concentration in one organ may in-
volve a toxic concentration elsewhere.

From these difficulties there are at present three gene-
ral procedures to which resort may be made. The first
of these is to modify the rate at which the drug may
permeate an organ by simple diffusion. Thus if a drug
is needed to penetrate the central nervous system, an
increase in its lipoid solubility should be sought, where-
as if permeation of the central nervous system is an

64 MEMBRANE PERMEABILITY AND DRUG ACTION

undesirable feature, increasing the polar character of the
drug, for example by its administration as a glucoside,
is suggested. The second possibility is to modify the
structure of the drug so as to increase the probability
that it will, or will not, fit into the secretory pattern of
certain organs. The combination of a drug with cholic
acid may secure a heavy secretion by the hepatic cells.
Increasing the polar-nonpolar asymmetry of the struc-
ture of a drug is very likely to increase its secretion by
the kidney. It may also secure its penetration into the
central nervous system, and there are indications that
permeation into the mammary gland may be favoured in
this way. A thorough study of the so-called blood-brain
barrier from this point of view is likely to lead to valuable
results, and the economic problem of mastitis in cattle
might well yield to a similar study of the secretory ac-
tivity of the mammary gland. A third possible mechanism
is to administer a drug so that it shall be inactive except
at selected sites of action. This mechanism has been
used, for example, for drugs involving the grouping

/ \as=As/ \ Compounds of this type are able to pene-
trate into the central nervous system, whereas the sim-
ple arsenoxides R^~^AsO cannot do so readily. To get

a therapeutic concentration of arsenoxide in the central
nervous system is likely to involve a toxic concentration
elsewhere. But if the drug is given as R/ \as=As/ ^R^
it penetrates into the central nervous system relatively

ACCESS OF DRUGS TO ORGANS 65

readily, and is there transformed into the therapeutically
effective arsenoxide. There are probably many instances
where such procedures could be adopted intentionally,
instead of by chance, as was the case with the arsenic
compounds. Thus a drug administered as a phosphate
ester would be likely to be inactive until the phosphate
had been hydrolysed away from the rest of the drug.
Thus such a compound might well display a high activ-
ity only at sites having a relatively high concentration
of phosphatase, such as the kidney and bones. If it is
true, as has been suggested recently, that tumours have
a higher concentration of glucuronidase than normal
tissues, it is possible that a drug relatively selective in its
action could be obtained by administration of a toxic
substance as a glucuronide. In the same way, since the
cells of tumours of the prostate are commonly rich in
acid phosphatase, a phosphate ester might increase the
specificity with which a drug can act upon this tumour.
Instances of this sort could be multiplied indefinitely.
Optimal results are likely to be obtained in any indi-
vidual instance by combining several of the devices
mentioned above. For example, if one wishes to obtain
a drug which will penetrate relatively well into the central
nervous system, one would tend to study the effect of
increasing its lipoid solubility, increasing its basic char-
acter and increasing its polar-nonpolar asymmetry.

66 MEMBRANE PERMEABILITY AND DRUG ACTION

Examples of the Permeability Factor in
Drug Action

To conclude this chapter I shall give three examples in
which the importance of the permeability factor in the
study of drug action is readily made apparent from the
practical point of view.

Drugs acting on Ascaris. A number of substances are
known to be of practical value as anthelminthics. These
include thymol and hexyl resorcinol. But resorcinol

TABLE XV

EXPERIMENTAL AND CALCULATED VALUES OF
THE PERMEABILITY OV Ascaris CUTICLE

P observed

P calculated

Resorcinol

0.7

0.05

Butyl resorcinol

30

(3.0)

Hexyl resorcinol

15

26

Heptyl resorcinol

37

80

Thymol

10

II

Chloroform

20

7

Nicotine

0.1

0.1

itself is of no practical value. Trim has recently studied
the permeability of Ascaris to various drugs. The results
are shown in Table xv. The rate of permeation into
Ascaris obeys the relationship

PM^^IB=o.2 approximately,

i.e. the membrane controlling the permeation of sub-

PERMEABILITY FACTOR IN DRUG ACTION 67

Stances into Ascaris belongs to group three of the types
of permeation given on p. 52. The permeability of As-
caris cuticle to a drug can be readily calculated from the
equation P = o.zBjMy^ The figures obtained for the
permeability are sufficiently close to the experimental
ones to indicate the relative likelihood of various sub-
stances permeating in concentrations sufficient to have
a toxic action. For example, the results show that the
active substance hexyl resorcinol penetrates into Ascaris
much more rapidly than does the inactive substance re-
sorcinol.

The Toxicity of Arsenoxides to Trypanosomes. The toxi-
city of arsenoxides to trypanosomes has been the subject
of an intensive study by a number of investigators, par-
ticularly King and Hawking. A great number of arsen-
oxides have been synthesised and their toxicities deter-
mined by finding the lethal dilution (L.D.), i.e. the
number of litres in which i gram mol. of each substance
must be dissolved to obtain a given degree of killing
of trypanosomes in a given time. The substances con-
cerned are of a type which we should expect to permeate
trypanosomes readily, and to conform to group four
of the permeation groups given on p. 52. The dilutions
in which these drugs are effective are very high, and it
seems likely from the studies of Hawking in particular
that practically every molecule of arsenoxide which
permeates into a trypanosome becomes fixed by a re-
Cell Physiology 5

68 MEMBRANE PERMEABILITY AND DRUG ACTIOlSr

ceptor grouping such as SH. Thus a trjrpanosome
is killed when a given proportion of its receptors is
saturated, i.e. when a given amount of arsenoxide^

TABLE XVI

VALUES OF THE LETHAL DILUTION (l.D.) FOR

the killing of trypanosomes by certain

arsenoxides, and calculated values of

(l.d.)M'/V5

L.D.

(L.D.)M^V-B

/ NasO

1900

2500

ho/ \AsO

530

6900

H^n/ ^AsO

33

580

COaH.CHa.NH^^AsO

3.8

5400

CO2H/ NasO

0-39

2000

Variation

5000 fold

1 2 fold

The relatively small variability of (L.D.)M^'^/B in-
dicates that most of the variation in L.D. is caused
by differences in the rates of permeation of the
arsenoxides into the trypanosome.

irrespective of its structure, has penetrated into the
trypanosome. If this is true, the permeability of the
trypanosome must be the main factor determining the
value of the lethal dilution. Assuming that this is so,
and that the permeation group is group four, it is easy

PERMEABILITY FACTOR IN DRUG ACTION 69

to show that the following relationship should hold:

'' — = constant approximately.

B

Table xvi shows a small selection of the experimental
data, including the most toxic, the least toxic and several
intermediate compounds of the arsenoxide type.Whereas
the value of the lethal dilution varies by a factor of 5,000
fold, the value of (L.D)M^^^IB variesby only twelve fold,
i.e. of the variation in the L.D., practically the whole
is accounted for by the variation in permeability. Thus the
organic chemists, in synthesising a wide range of arsen-
oxides, were unwittingly studying the permeability of
trypanosomes and little else.

The Chemotherapy of Lezoisite Poisoning. In a series of
biochemical studies Peters, Stocken and Thompson
found evidence that lewisite exercises its toxicity by
combining with the SH groups of enzymes. By studying
the ease with which lewisite may be detached from the
compound it forms with kerateine, they concluded that
lewisite must often combine not merely with one, but
with two thiol groups of the protein molecule. Conse-
quently they argued that to obtain a substance which
would compete efficiently with tissue enzymes for lewi-
site, and thus constitute an efficient therapeutic reagent ,
it would be necessary to have a dithiol grouping. Voegtlin
and QuASTEL had both shown earlier that monothiols can
prevent or reverse some of the toxic action of arsenoxides.

70 MEMBRANE PERMEABILITY AND DRUG ACTION

The danger from lewisite usually arises from skin con-
tamination. Consequently to prevent vesication by lewi-
site, it is necessary to have a dithiol which will readily
penetrate the skin. To attain reasonably fast permeation
of skin, it is necessary to have a small molecule, having
a moderate oil-water partition coefficient. Glycerol a, /5-
dithiol, later known as b.a.l.^ conforms to this specifica-
tion and was found to be very effective as an antidote to
the vesicant action of lewisite.

B.A.L. also has value as an agent for systemic arsenical
poisoning. But this is limited by the toxicity of b.a.l.
itself. The maximum dose of b.a.l. which may be ad-
ministered to man is 4 mg/kg/4 hrs. This is consider-
ably below the amount of dithiol which would be needed
to secure efficient therapeutic action in a serious case
of systemic lewisite poisoning. Consequently an attempt
was made to obtain a substance which would be an effi-
cient antidote for systemic poisoning. The points in the
molecular specification for such a substance were:

1 . It should inactivate arsenic.

2. It should have a low toxicity.

3. It should remove arsenic from cells into which arsenic
has penetrated.

4. It should penetrate the whole of the lymph and
vascular spaces.

5. The complex formed with arsenic should be readily
excreted.

^ B.A.L. for British Anti-Lewisite.

PERMEABILITY FACTOR IN DRUG ACTION 71

In chemical terms these specifications can be met in
the following ways. A dithiol grouping in the molecule
will provide for point i . Increasing the polarity of the
compound will comply with point 2 by reducing the
rate of penetration into cells: to increase the polarity
it is necessary to increase the number of OH, COOH or
SO3H groups in the molecule. Point 3 can be met by
having a relatively high concentration of the dithiol in the
blood stream: intracellular arsenic is not completely
incapable of diffusing out of cells, and if there is a high
concentration of external dithiol, every arsenic molecule
which diffuses out of a cell will be trapped by the dithiol.
To meet requirement 4 it is necessary that the molecule
should not be very large: for example, thiostarch or thio-
glycogen are excluded by this condition. The last re-
quirement, ready excretion, is achieved by the same
means as requirement 2.

Thus the specification of the molecule appeared as fol-
lows:

C.SH 2 COOH i

I + 3 OH > as alternatives

C.SH or I SO3H )

With a Specification as precise as this it was not necessary
to investigate many compounds. In fact, only two di-
thiols were studied, dithioadipic acid and the glucoside
of B.A.L. The glucoside of b.a.l. was the better, having an
L.D.50 of the order of 7.5 g/kg. When administered to
rabbits which had received an L.D.95 of lewisite, the

72 MEMBRANE PERMEABILITY AND DRUG ACTION

glucoside would ensure ioo% survival with treatment
begun not less than 4 hours after contamination and
50% survival with treatment begun not less than 6^
hours after contamination. Untreated animals die in
about 12 hours.

In the first experiments with this compound, point 3
of the specification was assumed to be sufficiently met
by the diffusibility of the arsenic itself, which would
cause the arsenic to be trapped by the dithiol in the
blood stream. Later an attempt was made to increase the
diffusibility of the arsenic by administering small amounts
of B.A.L. This can penetrate readily into cells and there
combines with the arsenic which was hitherto combined
with intracellular proteins. It was thought that as a result
of this process arsenic would diffuse much more readily
into the blood stream and would there be trapped by
the large concentration of b.a.l. glucoside. It was in fact
found that amounts of b.a.l. which were too small to
affect the percentage of survival when given alone, and
which were too small to exercise a toxic effect, would
markedly increase the efficiency of the glucoside.

It will be apparent from what has been said in this
chapter that, even with our present limited knowledge of
the permeability of mammalian tissues, it should often
be possible to reach valuable conclusions about the
action of a drug and the design of new drugs. In particu-
lar it is often possible to decide whether the permeability
factor is limiting the effectiveness of a drug, and to decide

REFERENCES 73

what changes in permeability would be profitable. The
systematic employment of this information should pre-
vent much waste of time in the synthesis and testing of
drugs. The great weakness in our present understanding
in this field is our limited knowledge of secretory pro-
cesses.

REFERENCES

Clark, A. J., 1937: General Pharmacology.
CoLLANDER, R. , 1937: Trans. Faraday Soc, 33, 985.
CoLLANDER, R. and Barlund, a., 1933: Actu. Bot. Fenn., 11, i.
Danielli, J. F., McAnally, M. and Phillipson, J., 1946: J. Exp.

Biol., 20, 417.
Danielli, J. F. and others, 1947: Biochem. J., 41, 325.
Davson, H., 1940: X Cell. Comp. Physiol., 15, 317.
Davson, H. and Danielli, J. F., 1943: The Permeability of Natural

Membranes (Cambridge Press, London).
LiLLiE, R. S., igzy. Protoplasmic Action and Nervous Action (Chicago).
Peters, R. A., ig^y: Nature, 159, 149.
Peters, R. A., Stocken, J. R. and Thompson, R. H. S., 1945: Nature,

156, 616.
QuASTEL, J. H., ig47: Nature, 159, 824.

Symposia: Faraday Society, 1937: Permeability, Trans. Faraday Soc;
1940: Mode of Action of Drugs, Trans. Faraday Soc.
Society for Experimental Biology, 1949: Selective Toxicity
and Antibiotics.
VoEGTLiN, F. R., 1925: Physiological Reviews, 5, 63.
WiNTERSTEiN, H., 1926: Die Narkosc (Berlin).

CHAPTER IV

Enzymes and Drug Action

Functions of Enzymes

In recent years many biologists have emphasised the
likelihood that a large part of the action of drugs upon
cells is to be explained mainly by the action of the drugs
on cellular enzymes. In England, for example, this was
particularly emphasised by A.J.Clark, R.A.Peters,
and D.Keilin. It may not be immediately apparent
vv^hy this should be so, but reasons become clear enough
if we consider the functions of enzyme systems in cells.
These functions include i. the synthesis of substances
which act as an immediate source of potential energy
for the physiological activity of the cell, e.g. the syn-
thesis of adenosine triphosphate; 2. the conversion of
potential energy to mechanical work, as is seen in mus-
cular contraction; 3. the protection of the cell against
invasion by foreign bodies — thus foreign proteins are
destroyed by proteases, J-amino acids by ^-amino oxi-
dase, hydrogen peroxide by catalase; 4. secretory ac-
tivity is dependent, directly or indirectly, upon enzyme
activity; 5. evidence has arisen, both from cytochemical
studies and from the study of mutations of yeasts.

FUNCTIONS OF DRUGS 75

moulds and bacteria, that enzymes may be concerned
in the mediation of genetic effects. In fact, there are singu-
larly few activities of living cells in which one of the key
positions is not occupied by one or more enzyme systems.

Possible Functions of Drugs in Relation

to Enzymes

Accepting the fact that enzymes are of vital importance
in the activity of cells, we must now consider the various
ways in which the activity of enzymes may be modified
by a drug. The possible modes of activity are quite
numerous. They include the following.

1. Action as carriers. Substances such as methylene
blue and pyocyanin may act as carriers between atmos-
pheric oxygen and dehydrogenases: in so doing the nor-
mal carrier systems, such as cytochrome, are short-
circuited. Such an action may not at first sight appear to
have serious consequences. But in practice the consequen-
ces may be quite dramatic: for example, sea urchin eggs
have their respiration raised by about 200% by addition
of pyocyanin, and this rise is accompanied by an almost
complete cessation of the processes of cell division.

2. Action as activators. Some substances are able to
modify the structure of an enzyme and thus modify its
activity. For example, reducing agents, such as b.a.l. and
HCN, are able to activate SH enzymes.

3. Action as chemical inhibitors. A number of sub-

76 ENZYMES AND DRUG ACTION

Stances such as the arsenoxides, the nitrogen mustards,
iodoacetate, fluoride and iodine, are able to inhibit the
action of enzymes by forming a chemical compound
with chemical groups which are essential for the mainten-
ance of enzyme activity.

4. Action as physical (competitive) inhibitors. In this
category we may mention malonate, which acts as an
inhibitor for succinic dehydrogenase; glyceraldehyde,
which acts as an inhibitor for triose phosphate dehydro-
genase, and the sulphonamides, which are believed to
compete with ^-aminobenzoic acid for enzyme systems
concerned in the metabolism of the latter.

5. Action as prosthetic groups. Certain substances are
able to act as the prosthetic groups of enzymes, thus
activating previously existing apoenzyme molecules.
Examples are vitamin Bj, which is concerned as a
prosthetic group in the pyruvic oxidase system, pyri-
doxal, which is a prosthetic group for some decarboxy-
lases, and vitamin Bg, which is a constituent of flavo-
protein systems.

6. Action as coenzymes. As an example of this may be
mentioned nicotinic acid, which is incorporated into the
molecule of coenzyme II.

7. Action as cosuhstrates. No examples of this are
known, but abnormal cosubstrate activity could clearly
be a serious source of trouble to a cell. It is but com-
paratively recently that Bergman introduced the con-
ception of cosuhstrates, so that it is not surprising that

FUNCTIONS OF DRUGS 77

no examples of such activity have yet come to light^.
8. Action as substrate removers. In the living cell the
course of metabolism of a particular substance is in part
determined by the presence of a suitable chain of en-
zyme systems. The functioning of such a chain of en-
zyme systems is dependent upon the product of the
action of one enzyme passing on to another enzyme for
which it is a specific substrate. Serious interference may
occur by modification of the structure of an intermediate,
in the course of its passage by diffusion from one enzyme
to another. For example, the chain of enzymes con-
cerned in the anaerobic metabolism of glucose may have
their action disrupted in this way by HCN, or by
H3ASO4. The action of HCN is to form a cyanhydrin
with phosphoglyceraldehyde, thus removing the sub-
strate for phosphoglyceraldehyde dehydrogenase.
H3ASO4 excercises its toxicity in part by combining
with phosphoglyceraldehyde under the action of triose
phosphate dehydrogenase, so that as a result of the
activity of this enzyme a phosphoarsenoglyceric acid
results instead of diphosphoglyceric acid. The phospho-

^ Cosubstrate activity is best understood by considering an example.
When trypsin is added to a solution of glycyl-leucine, the dipeptide is
not split by the enzyme. But if, now, to the solution is added a little
acetyl-phenyl-alanyl-glycine, synthesis occurs of a little of the substance
acetyl-phenyl-alanyl-glycyl-glycyl-leucine. Following this synthesis,
trypsin splits off first leucine and then glycine, leaving acetyl-phenyl-
alanyl-glycine as a residue which is not attacked. The compound acetyl-
phenyl-alanyl-glycine is said to have cosubstrate activity in the splitting
of glycyl-leucine by trypsin.

78

ENZYMES AND DRUG ACTION

arsenoglyceric acid decomposes spontaneously, so that no
diphosphoglyceric acid is available as substrate for the
next enzyme in the series.

Problems in the Analysis of the Action of
Drugs on Enzymes

When a drug is acting upon a simple solution of an en-
zyme, the analysis of the effect of the drug may be
relatively simple. For example, when the percentage
inhibition of an enzyme is plotted against the logarithm
of the concentration of the inhibiting drug, a linear curve
is quite commonly obtained. When drugs are acting

200

-100

-4 -3 -2

log concentration of phenol

Fig. 12. The respiration of yeast (a) and the fermentation of sugar by

yeast (b) as affected by phenol

upon cells, the action-concentrations curves are some-
times linear, sometimes non-linear. And even when they
are linear it does not follow necessarily that the action

ANALYSIS OF DRUG ACTION

79

of the drug on a given enzyme system is the same as it
would be on the same enzyme in aqueous solution.
The analysis is further complicated by the fact that the

100

200

100

Concentration x 10

Fig. 1 3. The eflfect of dichlorophenol upon the respiration (a) and cleavage

(b) of Arbacia ova. Note that there is little effect on cleavage until the

increase in respiration is almost complete.

effect of an enzyme poison may affect different cellular
processes in different ways. As is shown by Fig. 12
phenol increases the rate of fermentation of sugar by
yeasts, but decreases the rate of respiration. Fig 13 shows

0.8 1.2

"/o phenol in solution

Fig. 14. The relationship between the uptake of phenol by yeast (a), and
the lethal action of phenol on yeast (b)

8o

ENZYMES AND DRUG ACTION

that whilst dichlorophenol increases the rate of respiration
of Arhacia eggs, it decreases their rate of cleavage. In
such instances, when there are two or more cellular pro-
cesses which are interfered with in different ways by the
same drug, which is to be taken as the index of activity
on cellular enzymes ? There is no simple answer to this
question. Furthermore, if we endeavour to correlate the

-2

log [hCN]

Fig. 15. The relationship between concentration of HCN and inhibition

of various physiological processes, (a) Assimilation of CO2 by Chlorella;

(b) oxygen comsumption of frog's ventricle; (c) mechanical response of

frog's ventricle; (d) lethal action on Tribolium confusum

amount of drug taken up with the change in a particular
activity, it is commonly found that there is no close
correlation. Fig. 14 shows, for example, that when phe-
nol is acting on yeast there may be a considerable uptake
of phenol before a significant change in physiological
activity is observed. The first moiety of the phenol taken
up appears to be inactive.

It is thus clear that the action of drugs on intracellular
enzymes must be difficult to analyse.

ANALYSIS OF DRUG ACTION

8i

The analysis is further compHcated by the occurrence
of very marked species differences. For example, when
we take the inhibitory action of HCN on cellular processes
or say the lethal action of HgS, it is tempting to work
on the hypothesis that in all cases the drug is acting on
the same enzyme system, but Figs. 15 and 16 show that
the concentrations required to produce a given degree of

100

log [HiS]

Fig. 16. The relationship between concentration of H2S and its lethal
action on the spores of eight different species of fungi, (a) Venturiain-
equalis and Uromyces caryophyllinus; (b) Puccinia antirhini; (c) Sclerotina
americana: (d) Macrosporidium sarcinaeforme; (e) Pestolatia stellata; (f)
Glomerella cingulata; (g) Botrytis

action vary widely from species to species. Various ex-
planations of this are possible. It may be that the drug
does not produce its effect by acting on the same en-
zyme in all species. Or it may be that a given enzyme in
different species varies in its susceptibility to a given
drug. A third possibility is that the effective concen-
tration of the drug which arises in the vicinity of the
enzyme is different in different cells, although the ex-
ternal concentration of drug is the same.

82 ENZYMES AND DRUG ACTION

The Action of Drugs on Respiration and
Glycolysis in Muscle

From the evidence which can be derived from studies on
living cells it is clear, as we have just seen, that there
are great difficulties in analysing the action of a drug
into terms of activities on specific enzyme systems. An
alternative approach to the problem can be made by
studying the enzyme systems involved in a particular
physiological process on the test-tube scale, i.e. using tis-
sue extracts. This procedure has been carried out in great
detail in the case of the respiration and glycolysis of
muscle. The results so obtained seem likely to be repre-
sentative of the type of conclusion which will be reached
when a completely satisfactory analysis of drug actions
on enzymes is available. At present we must have some
reservations about the theories put forward by the bio-
chemists, because the enzyme systems have in most
cases been shown to perform the functions ascribed to
them in vitro only. In the case of the cytochrome system,
Keilin and others have provided direct evidence that
the intracellular systems are behaving in the way which
is postulated from test tube experiments. But so far as
glycolysis is concerned, whilst the picture built up by
biochemical studies is very plausible, we still await con-
clusive evidence from studies on living tissues that the
chain of events is identical with that postulated.

With these, and certain other reservations which will

ACTION ON RESPIRATION AND GLYCOLYSIS 83

be mentioned later, we may proceed to examine the action
of drugs on the respiration and glycolysis of muscle.

Fig. 17 gives a rough outline of the main steps involv-
ing the use of oxygen and glucose by muscle. Also in-
cluded are the probable points of action of carbon mon-

co^

HCN)
A/3 J

0? ^

Cytochrome oxidase\

Glucose

Mustard 90s
Lewisite

^

Cytochrome

Dehydro^nase.

Pyocyanm

Glucose phosphate

Glucose diphosphate

Olyceraldehyde

Urethan

lodoacefate

Lewisite

¥

Phosphoglyceraldehyde -^^HCN

Diphosphoglyceric acid

I -E!l

Pyruvate

Pyruvic oxidase^*

Mustard gas
Lewisite
Vitamin B

CO2 etc.
Fig. 17. A condensed scheme of the enzyme-catalysed steps in the res-
piration and glycolysis of muscle, with the points of action of some drugs

oxide, hydrocyanic acid, azide, urethan, iodoacetate,
lewisite, mustard gas, glyceraldehyde, arsenic acid,
fluoride and vitamin B^. As is indicated in the figure,
we have strong reasons for believing that the enzyme
systems concerned in respiration and glycolysis are
organised to give rise to a chain of reactions in which
each reaction produces a reaction product which is the
substrate for the subsequent reaction. As is indicated

Cell Physiology 6

84 ENZYMES AND DRUG ACTION

by the diagram, there is a strong tendency for each drug
to act rather selectively on a particular enzyme system.
But this is no more than a tendency, and with the enzyme
systems with which we are concerned here some of the
drugs act at several points. For example, HCN can act
both on the cytochrome system and also by combining
with phosphoglyceraldehyde : lewisite can act on hexokin-
ase, on triose phosphate dehydrogenase and on the
pyruvic oxidase system: mustard gas can also act at more
than one point. From these observations it is clear
that a drug may have the potentiality of acting at several
stages, even in one chain of biochemical events. Conse-
quently the analysis of the action of a drug on the intra-
cellular enzymes is extremely complicated. In the sys-
tem which we have been considering there are about
twenty enzyme systems only. In the living cell there
must be thousands of enzyme systems. To decide which
of these enzyme systems is of importance in the medi-
ation of the effect of any particular drug is obviously a
task of the first magnitude. In a few cases, where the
action of the drug is very rapid, as is the case with HCN,
the position is simplified because it is clear that the
physiological effects of the drug must be produced on
those enzyme systems, such as those involved in respi-
ration, damage to which can produce an immediate re-
sult. But where there is a serious delay between admin-
istration of a drug and the emergence of its action, no
such simplification is possible and the whole enzyme

ACTION OF ENZYME POISONS 85

organisation of the cell must be suspect, including even
those enzymes concerned in the mediation of genetic
effects.

The case which we have been considering is the one
in which we have the greatest knowledge of the organi-
sation of enzymic processes. We cannot proceed very
much further with the elucidation of the action of drugs
in this manner until much more elaborate studies have
been made on the enzymes concerned.

The Action of Various Enzyme Poisons on
Different Physiological Processes

An alternative method of exploring the action of enzyme
poisons is by considering their effect on different phys-
iological processes. Table xvii shows the action of four
different drugs on eight different physiological processes.
Of the four drugs, colchicine is the only one whose
effect may be restricted to one physiological process
only. The other drugs interfere with several physiolog-
ical activities. As we have already noted, HCN, iodo-
acetate and mustard gas all have a strong action upon
the metabolism of glucose, and theories have been put
forward from time to time that their physiological effect
is produced by interference with the metabolism of glu-
cose. But when we consider the action of these substances
on different physiological processes it is seen that their
actions are not always the same. One of the most striking

86

ENZYMES AND DRUG ACTION

TABLE XVII

THE ACTION OF FOUR ENZYME POISONS ON EIGHT DIFFERENT

CELLULAR ACTIVITIES

Colchi-
cine

M/iooo
HCN

M/iooo
iodoacetate

Mustard gas

Cellular respiration

Often 90%
inhibition ;
sometimes
none

Usually less
than HCN

No action

•

Muscle contraction

—

None

Stopped

No initial

initially;

almost

action ;

lactic acid

immediately

rigor later

rigor later

Amoeboid

None

?

No initial

movement

initially

action;
stopped later

Ciliary movement

?

None
initially

None
initially

?

Cell division

Spindle

None

Stopped

Spindle and

formation

initially;

chromosome

inhibited

stopped
later

abnormalities

Glycolysis

Stopped

Stopped
eventually

Renal Secretion

Stopped

Stopped

No initial
effect

Skin arterioles of

Dilated

Constricted

No initial

frog

effect

PHYSIOLOGICAL EFFECT OF DRUGS 87

examples of this is in the action of cyanide, iodoacetate
and mustard gas on the arterioles of the frog. Mustard
gas produces practically no effect at all, HCN produces
an almost immediate vasodilation and iodoacetate an
almost immediate vasoconstriction. It is no doubt pos-
sible by the exercise of sufficient ingenuity, by taking
account of the fact that these drugs act at different points
upon the metabolism of glucose, to maintain that it is
in fact interference with glucose metabolism which is the
key point in the attack of all these substances upon cells.
But this attitude smacks considerably of special pleading,
and a great deal more analysis on the physiological as
well as upon the biochemical level must be obtained be-
fore it can be accepted.

Classification of Drugs according to their
Physiological Effect

The possibility exists that some useful principle might
emerge from grouping drugs together which produce
the same physiological change in cellular behaviour. But
it is soon seen that a simple classification along these
lines is not very enlightening. As examples we may take
the inhibition of muscular contraction, the inhibition of
mitosis, and the induction of lachrymation.

Of course, muscular contraction can in practice be in-
hibited by substances which act upon the neuromuscular
junction. But as we are here concerned rather with cellu-

88 ENZYMES AND DRUG ACTION

lar units, we shall not consider junctional inhibitors, but
only the substances acting directly upon the muscle cell
itself. As outstanding examples there are iodoacetate and
potassium. Both of these substances rapidly give rise to
a condition in which muscle will no longer contract. But
they do so by quite different mechanisms, iodoacetate
by preventing the synthesis of adenosine triphosphate,
and potassium by reducing the excitability.

The formation of a properly functional spindle in mi-
tosis can be prevented by a variety of substances. These
include colchicine, urethan, arsenite, dithioglycerol,
the |S-chloroethylamines, and the sulphonamides. It
seems extremely improbable that substances with such
diverse chemical structures can be acting upon the same
cellular mechanism. The action of the /?-chloroethylam-
ines, for instance, depends upon the loss of a chloride
ion from the molecule with the consequent formation of
a carbonium ion. There appears to be no analogous pro-
cess possible with urethan. Then consider the two sub-
stances arsenite and dithioglycerol: the action of the
former can actually be neutralised by a moderate amount
of the latter.

Lachrymation is produced as a result of the action of
substances upon the nerve endings in the conjunctiva.
Substances producing this effect include soap, osmic
acid and ethyl iodoacetate. It seems very improbable
that all these substances can be acting upon the same
enzyme system.

ACTION OF DRUGS UPON ENZYME SYSTEMS 89

It therefore seems clear that not all the substances
having a common physiological effect exercise that effect
by an action upon a single enzyme system.

The Classification of Drugs in Terms of Enzyme
Systems upon Which They Act

Whilst very little valuable information comes from classi-
fying drugs according to the physiological effect they
produce, since it soon becomes apparent that the sub-
stances producing a common physiological effect must
do so by acting on a diversity of receptor systems, it
still remains possible that substances acting specifically
upon a certain enzyme system may produce a standard
physiological effect. As examples we may take substances
which act particularly upon SH enzymes, substances
which are inhibitors of choline esterase, and substances
which are inhibitors of hexokinase.

It seems likely, from the work of DixON and his
colleagues, that substances acting specifically on SH
groups, such as those of triose phosphate dehydrogenase,
are always lachrymators. In particular substances con-
taining the groupings

R.CO.CH2X, where X = halogen, and R.CO.CH=CH2

are often rather specific agents for SH groups. They
include many of the substances which have been found
useful in chemical warfare as lachrymators. On the other

90

ENZYMES AND DRUG ACTION

hand, not all lachrymators are capable of combining with
SH groups: for example, soap does not.

Substances such as eserine act on choline esterase, and
prevent the hydrolysis of acetyl choline, probably by
competitive inhibition. As a result of this process marked

TABLE XVIII

THE ACTION OF VARIOUS SUBSTANCES, AS POISONS TO HEXO-
KINASE, AND AS VESICANTS IN MAN

Vesicancy

% Inhibition

S (CHaCH^COa

+ +

80

OS (CHaCH^COa

0

S (CHC1.CH3)2

0

S (CHjCOa

0

S (CHj.CHa.CHaCOa

0

02S(CH=CH2)2

+

60

OS(CH=CH2)2

0

S (CH2.CH20H)2

0

S (CH2.CH3)(CH2.CH2C1)

+

45

N (CHa.CHaCDa

+

70

AsCl2(CH=CH2Cl)

+ +

100

As ClgCCHa.CHaCl)

0

As ClaCCHa.CHa)

+

45

CHsBr

+ +

90

CHa^C.O.CO

+

40

CH-CH

The inhibitory action was studied in the presence of M/i 50 glucose.

physiological effects ensue which include constriction of
the pupil of the eye. Most of these substances act in a
reversible manner, but studies by Saunders, Adrian

ACTION OF VESICANTS 91

and Dixon have shown that the alkyl fluorophosphonates
act upon choUne esterase in a relatively irreversible
manner and are also myotics. It thus seems likely that
all substances which can produce a sufficient degree of
inhibition of choline esterase have myotic activity.

Table xviii is taken from the work of Dixon and
Needham. It shows the action of vesicants on hexokinase
in the presence of M/150 glucose. It will be seen that the
substances having vesicant activity all poison hexokinase
under these conditions, whereas those substances having
no vesicant activity are not good poisons for hexokinase.
The conclusion appears that all substances inhibiting
hexokinase are likely to be vesicant.

From evidence of the type just given it seems likely
that substances having a selective action upon a particu-
lar enzyme system are likely to display the same physio-
logical activity.

The Mode of Action of Vesicants

•

As was noted in the previous section, the action of vesi-
cants upon enzyme systems has been the subject of
intensive study, both in England and in America. We
shall examine these studies in a little more detail to show
the great difficulty which is encountered in reaching a
final conclusion as to the actual mode of action of any
particular vesicant. We shall particularly consider lewi-
site and mustard gas, since these two substances have

92 ENZYMES AND DRUG ACTION

been the subject of the most consistent attack, both by
enzymologists and by others. In the first instance we
must distinguish between the local vesicant effect and the
systemic effect of these substances. A survey of the liter-
ature shows that there is no common agreement between
different workers as to the mode of action of these two
substances. Peters is of the opinion that the primary
effect of mustard gas is upon the surfaces of skin cells,
and possible upon the enzymes in those surfaces. The
result of this action is the liberation of proteases which
among other effects produce leucotaxin. On the other
hand Peters considers that the primary effect of lewisite
is upon pyruvic oxidase.

Dixon and Needham, on the other hand, tend to the
view that both substances exercise their primary effect,
in common with other vesicants, upon hexokinase and
perhaps other phosphokinases.

American workers in this field have reached somewhat
different conclusions, in which they have been influenced
by their observation that some proteases are as sensitive
to these substances as are the phosphokinases. Cori and
Cannan appear to favour the view that the primary
effect of mustard gas is upon the cell surface, and results
in liberation of enzymes and changes in permeability.
A recent American review concludes that "the specific
chemical lesion is not yet defined by studies on inacti-
vation of enzymes."

The effect of vesicants on the skin is so dramatic that

ACTION OF VESICANTS 93

for long the systemic effects of mustard gas and of
lewisite tended to escape study. Attention was particu-
larly directed towards the systemic effects by Cameron,
who pointed out that in the long run the systemic effect
might be much more important than the skin lesion.
Subsequently many studies have been made of the basis
of the systemic effects. Peters inclines towards the view
that the main effect is produced by the inhibition of
pyruvic oxidase, whereas DixoN and Needham incline
to the view that various phosphokinases may be con-
cerned, amongst which may be part of the pyruvic oxidase
system. Various cytological approaches have also been
made to this problem. Robson, Auerbach and Roller
have indicated that the action of mustard gas may be
primarily on the nucleus of a cell. They have shown that
mustard gas is mutagenic and prevents mitosis. American
workers have emphasised the action of mustard gas on
cell membranes — red blood cells, Nitella, leucocytes, and
lung cells have been studied and all appear to have their
membrane properties significantly changed by amounts
of mustard gas which would not exercise any profound
effect upon intracellular enzymes. It has also been pointed
out that there are many points of resemblance between
mustard gas poisoning and radiation sickness.

From this survey of studies on vesicants it must be
plain that whilst studies on enzyme systems are often
very suggestive, yet it is often extremely difficult to obtain
incontrovertible proof that a particular inhibition of one

94 ENZYMES AND DRUG ACTION

or more enzyme systems is in fact the result of a parti-
cular enzyme poison.

Biological Aspects of Enzyme Studies

When we look at the results of enzyme studies from the
point of view of a biologist, we find a number of difficul-
ties in applying these studies to cellular systems. These
are:

1 . Among the characteristic activities of many cells is the
capability of concentrating foreign substances in va-
cuoles or granules. The biochemist is accustomed to
adding a certain amount of enzyme poison to his en-
zyme solution or suspension, and then considering
that the poison is uniformly distributed in the mate-
rial under observation. But there is no such uniform
distribution likely to occur in cellular systems, and
the calculation of average concentrations for cellular
systems is often fruitless, and even misleading. As a
result of the concentration of drugs in particular lo-
calities in cells, a drug may never come into contact
with the cellular systems which in vitro are the most
sensitive to it.

2. Up to now the enzymes which have been mostly stud-
ied by biochemists have been cytoplasmic systems.
It is much more difficult to make adequate studies on
the enzymes of nuclei. Yet substances acting in very
low concentrations are at least as likely to exercise

BIOLOGICAL ASPECTS OF ENZYME STUDIES 95

their effect upon nuclear as upon cytoplasmic systems.

3. In many cases it is possible to show a decline in the
concentration of active enzyme in a particular tissue,
subsequent to the administration of a drug. But usu-
ally this effect can only be demonstrated after the
lapse of an interval from the time of administration of
the drug. In such circumstances it is difficult to be
sure whether the inactivation is produced by direct
action of the drug upon the enzyme, or is an indirect
effect ensuing from the action of the drug on some
other enzyme system.

4. When the action of a given substance is studied on the
enzyme systems of different organs, it is often found
that the main effect appears to be selective for different
enzyme systems in different organs.

5. In many cases some doubt as to an interpretation in
terms of enzyme activity appears because a similar
physiological effect can be produced by a mechanism
which is primarily non-enzymic in its action. Thus
vesication, which is produced by substances which
are strong enzyme poisons, is also produced by fric-
tion, by heat and by cold.

6. If we attempt a straight-forward application of the
results of enzyme studies, we should be tempted to
say that drugs acting on the same enzyme system must
produce the same effect. But this is seldom true. For
example, vesicant substances all appear to poison
hexokinase under certain conditions, and we might

96 ENZYMES AND DRUG ACTION

expect the vesicles to be identical. But in fact on the
cytological and histological levels, the vesicles pro-
duced by different substances appear very different.
The general conclusion which we can reach at present
is that drugs often exercise one or more of their modes
of action through enzyme systems. Some drugs may
exercise their action on enzymes only, or even on one
enzyme only, but this degree of specificity is very rare.

REFERENCES

AuERBACH, C. and Robson, J. M., jg44: Nature, 154, 81.

Baldwin, E., 1946: Dynamic Biochemistry (Cambridge University
Press, London).

Clark, A. J., 1937: General Pharmacology (Handbuch der Exp. Pharm,
IV).

CuLLUMBiNE, H., ig4j: Nature, 159, 151.

Dixon, M., ig^^: Nature, 161, 226.

Dixon, M. and Needham, D. M., 1946: Nature, 158, 432.

Dixon, M., 1948: Multi-Enzyme Systems (Cambridge Univ. Press,
London).

Oilman, A. and Phillips, E. S., 1946: Science, 103, 409.

Keilin, D., 1929: Proc. Roy. Soc, B. 104, 206.

Roller, P. C, 1947: Symp. Soc. Exp. Biol., I, 270 (Cambridge Uni-
versity Press, London).

Peters, R. A., 1947: Nature, 159, 149.

Work, T. S. and Work, E., 1948: The Basis of Chemotherapy (Oliver
& Boyd, London).

CHAPTER V

The Actions of Narcotics

Introduction

In this chapter we shall deal with the action of narcotics
on cells. This group of substances has not been studied
as intensively by such a large number of chemists as have
vesicants, but on the other hand it has been studied by
people with much more diverse backgrounds. As a result
there has been a very fruitful interplay of physical and
chemical theories and a great variety of working hypo-
theses has been considered. Most of these hypotheses
fall into three broad groups. The first of these involves
an action on the cell surface or some other biologically
important surface. The second emphasises the partition
of narcotics between an aqueous phase and other phases
of a lipoid character. The third considers narcotics as
substances whose action is mediated primarily through
enzyme systems.

Before proceeding in further detail, it will be as well
to get some idea of what is referred to under the heading
of narcosis. It is not a word with a single precise meaning.
It includes such phenomena as the loss of consciousness,
inhibition of a reflex, inhibition of the contractility of,

98

ACTIONS OF NARCOTICS

say, heart muscle, inhibition of cell division, inhibition
of ciliary movement, inhibition of respiration etc. Usu-
ally a given narcotic substance will produce all these

c
.o

o

2 /A

o / / /

o / / /

/ / /c
S. -3 / / /

-2- / //X

-1 - ////

0- -^ /

10

12

H

Fig. 1 8. Some narcotic actions of various primary alcohols. The con-
centration of alcohol producing a standard degree of narcosis is plotted
against the number of carbon atoms in the alcohol, (a) Inhibition of
swimming of tadpoles; (b) inhibition of hog ventricle; (c) concentrations
reducing the surface tension of water by lo dynes/cm ;(d) concentrations
lethal to B. typhosus; (e) immobilisation of Paramoecium

effects, but at different concentrations. Fig. i8 shows
some characteristic results for the aliphatic primary alco-
hols. It will be seen that the different effects are produced
by quite different concentrations of alcohol, and that the
curves are not even all parallel to one another. From the
facts already considered it is improbable that all the nar-

ACTIONS UPON SURFACES 99

cotic actions of a given substance are produced by the
sanie mechanism. Hence it does not follow that what is
established for a narcotic in one connection is necessarily
involved in any other action involving the same narcotic.

Theories of Actions upon Surfaces

Permeability to narcotics. It is very common to find that
the effectiveness of a narcotic increases as the oil-water
partition coefficient is increased: the greater the relative
concentration in oil, the greater is the narcotic action.
The question naturally arises as to whether this is due
to differences in cell permeability, for permeability
usually also increases as the oil-water partition coefficient
increases. This, however, is very unlikely to be the case,
for most effective narcotics have a structure which allows
them to penetrate quite rapidly into cells. For example,
the aliphatic alcohols display an activity which usually
increases by between 2.4 and 4.5 fold for each additional
CH2 group in the molecule. But all the aliphatic alcohols
penetrate very rapidly into cells and the differences be-
tween the rates at which they permeate are far too small
to allow for the increments in activity for each CHg group
added to the molecule. It is probably quite seldom that
the differences in narcotic activity in a homologous series
can be attributed to the differences in rates at which the
different members of the series penetrate into cells.

Cell Physiology 7

lOO ACTIONS OF NARCOTICS

Is the action restricted to the cell surface? There is a good
deal of evidence showing that substances, such as mag-
nesium, cocaine and curare, act primarily on the cell
surface, in some way modifying the excitability of the
cell. Although the evidence as yet available is in no case
conclusive, the best working hypothesis in the study of
the action of these substances is probably that the action
is restricted to the cell surface.

From time to time it has been suggested that the action
of many other drugs is also restricted to the cell surface.
Experiments which have received much attention are
those of Brinley, Hiller and Marsland, who studied
the action of narcotics and other substances by injecting
them into the interior of Amoebae. They found that
HCN, HgS, picric acid and various of the conmion
narcotics had no action when they were injected into the
interior of Amoebae. This was true even if a narcotic was
injected at a concentration which was sufficient to cause
complete narcosis when an Amoeba was placed in the
narcotic solution. An extreme example was that found
with HCN. Amoebae placed in M/3000 HCN are killed
in 24 hours, whereas the injection of M/ioo HCN had no
effect. The authors therefore argued that the action of
these substances must be restricted to the external sur-
face of the cell.

But there is a serious source of experimental error in
work of the type just mentioned. All of the substances
mentioned, including HCN, HgS, picric acid and the

ACTIONS UPON SURFACES lOI

narcotics, are of a type which pass through cell mem-
branes very rapidly indeed. As a result, within a few
seconds of making an injection into an Amoeba, prac-
tically the whole of the injected substance has diffused
into the medium surrounding the Amoeba. Consequently
experiments of this type are useless unless the substances
injected can be relied upon to stay within the cell which
is injected.

There is one set of experiments which is not invali-
dated by the diffusibility of the narcotic concerned.
These were experiments made with paraffins as the nar-
cotic substances. Mars land found that when a drop
of olive oil is brought against the surface of Amoeba dubia
it forms a cap on the surface. When an appropriate amount
of a paraffin is dissolved in the olive oil, the Amoeba is
fairly rapidly narcotised and ceases to move. But when
droplets of this paraffin solution are injected into the
interior of the Amoeba, no narcosis results. Hence in this
case one may justly conclude that the narcotic effect of
paraffins is exercised on the cell membrane.

Do narcotics change permeability to metabolites? A num-
ber of workers, particularly Hober, Lillie and Winter-
stein, have concluded that narcotics exercise their action
by decreasing the permeability of cells to essential meta-
bolites. At the time at which the theory was put forward
there was comparatively little experimental evidence
available on the influence of narcotics on cell permeabil-

102 ACTIONS OF NARCOTICS

ity. But a certain number of studies are now available.
The rnore important of these are those of Jacobs and
Parpart (1937), Barlund (1938), and Davson (1940).
Barlund showed that ether tends to decrease cell per-
meability and that the decrease varies with the molecule
which is considered. The effect is never very large.
Jacobs and Parpart showed that butyl alcohol de-
creases permeability of the red cells of man, rat and rabbit
to glycerol, but increases the permeability of the red
cells of ox, sheep and dog to glycerol. Davson found
that various narcotics increased the permeability of cat red
cells to potassium and simultaneously decreased their
permeability to sodium. These and various other results
in the literature indicate that the effect on cell permea-
bility of a given narcotic is sometimes to decrease the
permeability, sometimes to increase it: the effect varies
from cell type to cell type, species to species, and mole-
cule to molecule. As the narcotic substances do not
display any common effect on cell permeability it seems
unlikely that the changes in permeability which they
may induce are often prominent as effective modes of
action in narcosis.

7^ Traube's Theory of adsorption correct? Traube noted
that there is a close parallel between the effect of narcotic
substances on surface tension and the physiological effect
of the substances. As the number of CHg groups is in-
creased, so surface tension is reduced and narcotic activ-

ACTIONS UPON SURFACES

103

ity increases. The effect shows up in sonie cases rather
clearly when the surface tension is studied of solutions
of substances, all of which have roughly the same narcotic
activity. There is a strong tendency for the solutions
of equi-narcotic activity to have the same surface tension.
Figure 19 shows an example of the parallel between

10

Concenfrafion

Fig. 19. TTie action of ethyl urethane in reducing the respiratory reflex in
the cat (•) and in reducing the tension at the air-water interface ( X )

reduction in the surface tension at the air-water inter-
face and reduction in a reflex of the cat. Traube there-
fore proposed that the narcotic substances exercise their
effects primarily on surfaces, and that the effect is due
to the reduction of surface tension. The strength of his
case rested upon the fact that when a given substance
is studied at different concentrations there is a parallel
between narcotic activity and reduction in surface ten-
sion. And when different members of a homologous

104 ACTIONS OF NARCOTICS

series are studied there is also a close parallel between the
reduction of surface tension and the narcotic effect.

However there are three main criticisms which must
be levied against this rather simple theory. These are:

1. The parallel between surface tension reduction and
narcotic activity may be misleading. The reason for
this is that in a homologous series the relative reduc-
tion in surface tension obtained with different mem-
bers of the series is merely a reflection of the differ-
ences in the hydrocarbon moiety ofthe molecule. There
are many other physico-chemical properties of mole-
cules which also are a reflection of the hydrocarbon
content of the molecule. For example, changes in the
oil-water partition coefficient run closely parallel to
changes in surface tension, for the simple reason that
both properties closely follow the hydrocarbon content
ofthe molecule. Thus from the studies of narcosis, one
is justified in concluding that there is a close parallel
between the narcotic effect and the amount of hydro-
carbon in the molecule. But one is not justified in
concluding that any particular individual physico-
chemical property which runs parallel to the hydro-
carbon content is necessarily the one through which
the physiological effect is mediated.

2. Even if the effect of narcotic substances is exercised at
an interface, studies on the air-water interface may
very well be irrelevant. The important interfaces of a
cell are oil-water, protein-water, nucleic acid-water.

OIL-WATER PARTITION EFFECTS I05

etc. Consequently the parallel between air- water sur-
face tension and narcotic effect noted by Traube is
quite likely to be fortuitous.
3. As Meyer has shown, some narcotics have no action
on the air-water surface tension, for example methane
and nitrous oxide are both without action upon the
air-water surface tension, although they can exercise
apotent narcotic effect. It is, therefore, clear that whilst
some narcotics may exercise their effects upon an inter-
face having properties similar to the air-water interface,
other narcotics must act in a quite different manner.
Consequently one can conclude that a simple theory
of absorption such as that of Traube is unlikely to be
of major importance, except in peculiar cases. This how-
ever, must not blind us to the fact that adsorption at
interfaces may indeed be a very important part of the
action of many narcotics. But if so the mechanism is not
so simple as Traube has suggested.

Theories based on oil-water partition effects

A number of workers have put forward the hypothesis
that many narcotic substances exercise their physiolo-
gical effect by virtue of changes which occur in the or-
ganisation of essential structures of the cell as a result
of the dissolving of the narcotics in oily phases which
are part of these structures. For example, suppose we
consider the substances

Io6 ACTIONS OF NARCOTICS

CHCls ^_ /NH-COv /Et

CO

NO. A^'^'^^Y

CH3.CH2OH V_yNH2 \NH-CoAEt

First inspection of these formulae suggests that it would
be impossible for these substances to act through the
same chemical mechanism, but if we make the hypoth-
esis that these substances act after dissolving in a lipoid
phase, then the first thing we must do is calculate what
are the relative concentrations of these substances in the
lipoid for solutions having the same narcotic activity. It
is surprising to find that although the structures of these
substances are quite different, and although their equi-
narcotic concentrations in water are quite different, the
concentration of these substances in lipoid is often prac-
tically the same.

Tables xix and xx show two examples of calculations
made by K.H.Meyer (1937), in which he develops
the Overton-Meyer hypothesis in a fairly precise
manner. One column in these tables shows concentration
of material which produces a given degree of narcosis.
The other column shows the concentrations of the diffe-
rent substances in an oily phase which would be in equi-
librium with the narcotic concentration in air or water,
as the case may be in the two tables. For the results shown
in Table xix where the narcotic is present in air, the con-
centrations of different narcotics producing a standard
effect sho w a variation of 740 fold, whereas the corre-
sponding equilibrium concentrations in oil show a vari-

OIL-WATER PARTITION EFFECTS

107

ation of only 1.8 fold. Similarly in Table xx the concen-
trations of the different narcotic substances in water
which produce a given degree of narcosis show a 7000
fold variation. But the corresponding equilibrium con-
centrations in oil show a variation of only 2.5 fold. It is

TABLE XIX

THECONCENTRATIONSIN AIR OF VARIOUS NARCOTICS
REQUIRED TO PRODUCE A GIVEN DEGREE OF NARCOSIS
IN MICE, AND THE CORRESPONDING EQUILIBRIUM
CONCENTRATIONS OF THE NARCOTICS IN OLIVE-OIL

Concentration

in air; vols.

per cent

Corresponding
concentration

in olive oil;
mol/litre x lo*

Methane

370

8

Nitrous oxide

100

6

Acetylene

56

5

Ethyl chloride

5

7

Ether

3-4

9

Methylal

2.8

8

Carbon disulphide

I.I

7

Carbon tetrachloride

0.6

7

Chloroform

05

9

Range of variation

740 fold

1.8 fold

thus clear that there is a very close correspondence be-
tween the production of a given degree of narcosis, and
the production of a standard concentration of molecules
in a lipoid phase. The detailed chemical structure of the
molecules does not appear to be of any importance. It
is the features of the molecules which determine their

io8

ACTIONS OF NARCOTICS

oil-water partition coefficient which also determine nar-
cosis.

The partition of a substance between oil and water can

TABLE XX

THE CONCENTRATIONS IN WATER OF VARIOUS SUB-
STANCES REQUIRED TO PRODUCE A GIVEN DEGREE OF
NARCOSIS IN TADPOLES, AND THE CORRESPONDING
EQUILIBRIUM CONCENTRATIONS OF THE SUBSTANCES

IN OIL

Concentration

in water;
mols per litre

Corresponding

concentration

in oil ; mols per

litre X lo^

Ethyl alcohol

0-33

3-3

Propyl alcohol

O.II

3.8

Butyl alcohol

0.03

2.0

Valeramide

0.07

2.1

Antipyrin

0.07

2.1

Pyramidon

0.03

3-9

Ether

0.024

S-O

Benzamide

0.013

3-3

Salicylamide

0.0033

2.1

Luminal

0.008

4.8

o-Nitraniline

0.0025

3.5

Carbon disulphide

0.0005

3-0

Chloroform

0.00008

2.6

Thymol

0.000047

4-5

Range of variation

7000 fold

2.5 fold

be regarded as determined by the sum of the differences
in free energy of its component groups in the two media.
From the data just given it is evident i . that there is a
parallel between equi-narcotic action and concentration

ACTIONS ON ENZYMES I09

in a lipoid phase which is not upset by variation in either
polar or non-polar groups, and 2. that no structural spe-
cificity in the molecules is involved. It, therefore, seems
very improbable that adsorption at an interface is the
mode of action in the two instances of narcosis which
have just been given. Equally, it seems highly probable
that narcosis is produced in these two instances by the
production of a critical concentration of foreign molecules
in a bulk phase consisting mainly of hydrocarbon.

At this point, we must ask the question, where is the
lipoid phase in the cell ? It may perhaps be the interior
of the cell membrane, the lipoid micelles, or possibly
a conglomeration of the hydrocarbon residues of protein.
Possibly all three sites may be involved in different in-
stances of narcosis. And no doubt there are some sites
which have not yet been thought of.

It is important to note that this theory is not incom-
patible with the involvement of enzymes in narcotic
action. The action upon any of the three sites just men-
tioned may well produce an ultimate action upon the
degree of activity of an enzyme system.

Theories Based on Actions on Enzymes

We have previously noted that urethane acts upon de-
hydrogenases and will inhibit respiration for this reason.
But the linkage between respiration and other cellular
effects is complex, as is well illustrated by the fact the

110

ACTIONS OF NARCOTICS

the same physiological effect can sometimes be achieved
either by increasing or by decreasing the respiration
rate! For example, dichlorophenol and urethan both
stop cell division at a certain minimal concentration,
which is of course different for the two substances. But

TABLE XXI

THE NARCOTIC ACTION ON GUINEA PIGS, AND THE INHIBITION
OF OXYGEN UPTAKE BY GUINEA PIG BRAIN, OF CERTAIN

BARBITURATES

Narcotic
action

/NH-COvyCHMca
CO C

\nH-CoAcHo-CH= CHa

^NH-COx/CHMcj

CO C

\NH-CO/^CHo-CBr= CHo

/NH-CO\/CHMe2
CO C

^NH-Co/^CHa-CH Br-CH,

+ +

+ +

negligible

% Inhibition

of O2 uptake

by brain

40

SO

dichlorophenol at the concentration at which it stops cell
division increases the respiration of Arhacia eggs by more
than 100%, whereas urethan at the concentration at
which it inhibits the division, reduces the respiration by
about 75% !

Amongst the more striking studies of the eflPects of

ACTIONS ON ENZYMES III

narcotics upon respiration is that of Quastel upon the
respiration of the brain tissue. Table xxi shows some
of his results. Inspection of this Table indicates that
with the substances he was investigating a difference of
one double bond in structure was sufficient to cause
profound modification of narcotic activity, and an
equally profound modification in the ability to inhibit
oxygen uptake by brain tissue. It is clear that the very
small difference in partition coefficient produced by a
difference in structure of one double bond could not
possibly account for the differences of activity of these
different molecules. One therefore wonders whether this
is an instance in which a specific structure of the mole-
cule producing narcosis is important, and whether this
structure is not specific for one of the enzyme systems
involved in respiration.

Some doubt however is thrown upon this by the
results given in Table xxii. This shows in one column the
concentration of narcotic which is necessary to produce
a given degree of narcosis in the rat, and in the other
column, the degree to which the uptake of oxygen by
brain tissue is inhibited by the narcotic concentration.
If all the narcotics are acting in the same manner, one
would except them to inhibit the respiration of brain
by the same amount when present in equi-narcotic con-
centrations. But, in fact, urethan produces only one-
fifth of the inhibition produced by an equi-narcotic con-
centration of avertin. The evidence, therefore, is not very

112

ACTIONS OF NARCOTICS

convincingly in favour of the idea that narcotics are
acting upon a single site when acting upon brain tissue.
It is possible that the results would have conformed
more precisely to theory had the brain tissue been in a
normal condition, instead of being in slices. But, never-

TABLE XXII

THE CONCENTRATIONS OF NARCOTICS PRO-
DUCING A GIVEN DEGREE OF NARCOSIS IN THE
RAT (i.e. EQUINARCOTIC CONCENTRATIONS),
AND THE PERCENTAGE INHIBITION OF THE
RESPIRATION OF RAT BRAIN SLICES PRODUCED
BY THESE CONCENTRATIONS OF NARCOTIC

Equi-narcotic

% Inhibition

concentration

of brain

in rat

respiration

Ethyl urethan

0.022

6

Chloral hydrate

0.0013

10

Luminal

0.00079

15

Chloretone

O.OOIO

20

Avertin

0.00106

31

theless, one cannot help being highly suspicious of a
simple theory of action on one site.

Further investigations by Quastel have shown that
the effects of narcotics on respiration are probably pro-
duced by their effect on carbohydrate metabolism, which
is of exceptional importance in the respiration of brain.
But he found that in the low concentrations producing
narcosis in mammals, there is no significant direct effect

ACTIONS ON ENZYMES II3

of the narcotics on the dehydrogenases themselves.
The effect of the narcosis is on an earlier step as in-
dicated by the following diagram:

O2 — Cytochrome oxidase

I

Cytochrome

I I Region sensitive to low

Flavoprotein ) -*- concentrations of narcotics

Coenzyme

Substrate — Dehydrogenase ■<- Region sensitive to high

concentration of narcotics

The views which Quastel has developed, outlined
above, refer only to the action of narcotics upon the
central nervous system of mammals, particularly of ro-
dents.

Johnson and his colleagues have made some very
stimulating studies from a quite different point of view
on the effect of narcotics on bacteria. They found that
narcotic action can be antagonised by high pressure. It
had been noted that when protein denatures there is an
increase in volume. If this is prevented by compression,
no denaturation occurs. Johnson, therefore, argues that
since the action of narcotics can be reversed by high
pressure, the narcotics act by denaturing proteins and
not by adsorption on specific active centres of enzymes.
This view is, of course, quite compatible with the oil-
water partition coefficient hypothesis of Overton and
Meyer, for it is quite clear that lipoid substances can

114 ACTIONS OF NARCOTICS

denature proteins. They probably do this mainly by
changing the organisation of the lipoid residues in the
polyptide chains which make up the proteins. This
theory of Johnson's has the advantage that it also allows
for species variation, since when a given enzyme is taken
from different organisms, it usually shows marked vari-
ation in ease of denaturation, according to the source
from which it has been taken.

REFERENCES

Barlund, H., 1938: Protoplasma, 30, 70.

Brinley, F. J., igzS: y. Gen. Physiol., 12, 201.

Clark, A. J., 1937: General Pharmacology (Handbuch der Exp. Pharm.

IV).
Davson, H., 1940: y. Cell. Camp. Physiol., 15, 317.
Davson, H. and Danielli, J. F., 1943: The Permeability of Natural

Membranes (Cambridge University Press, London).
HiLLER, S., 1927: Proc. Soc. Exp. Biol, and Med. ,25, 305.
HoBER, R., 1945: Physical Chemistry of Cells and Tissues (Churchill,

London).
Johnson, F. H., Brown, D. E. and Marsland, D., 1942: y. Cell.

Comp. Physiol., 20, 269.
LiLLiE, R. S., 1923: Protoplasmic Action and Nervous Action (Chicago).
Marsland, D., 1934: J. Cell. Comp. Physiol., 4, 9.
Meyer, K. H., 1937: Trans. Faraday Soc, 32, 1062.
QuASTEL, J. H., 1943: Trans. Faraday Soc, 39, 348.
Traube, L, 1935: Biochem. Z., 277, 39; 282, 444.
Traube, I., 1937: Trans. Faraday Soc, 32, 1066.
WiNTERSTEiN, H., 1926: Die Narkose (Berlin).

^

'-y^

'.. Vv

ym

V .'■-

'''^^^m>'^^0^-

^mt' -.

a

Plate II. Vesicles in the skin of the frog, produced by different agents,
a is a normal skin. In h the prickle cells adhere firmly to oneanother,
and to the dermis; the cornified layer splits away from the prickle cells.

In c the whole epidermis is spHtting away from the dermis, and in d

the spHt is occuring in the middle of the prickle cell layer. The three

types of vesicle b, c and d are characteristic of the different agents

producing them (see p. 120 and 121).

CHAPTER VI

Responses of Cells on the Biological Level

Introduction

It can be concluded from the previous chapters that
if the attempt is to be made to develop new drugs on a
rational basis, rather than on the hit or miss principle,
two types of research unit are required. One of these
types is necessary to study the general physico-chemical
properties of the cells of mammals and of parasites, in
particular such properties as permeability, secretion
mechanisms, excitability phenomena and adsorption
effects. The other type is more biochemical in character,
and is particularly required for the study of enzyme
effects. But this type of research in only a part of that
which can be deduced to be necessary from the cytological
point of view. There is also a great need for more frankly
biological studies of cells. Altogether too little is known of
what may, by analogy, be called the natural history and
ecology of cells, and their responses on the biological level.

The Nature of Biological Responses

The biological responses of cells to drugs are occasionally
highly specific and can only be elicited by a very small
Cell Physiology 8

120 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

range of compounds. But more often the response may
be evoked by a wide variety of substances and indeed
by a wide variety of types of stimuli. For example, the
characteristic response of nerve or muscle may be elicited
by heat, cold, pressure, cutting, potassium, various or-
ganic bases and by electrical shocks. Similarly, the
complex series of cellular events resulting in vesication
may be elicited by heat, cold, friction, bacterial toxins,
arsenoxides and /3-chloroethylamines. It is very tempting
to conclude that all the agents producing a given effect
must be acting upon the same mechanism, and that by
contemplation of what system would have the ability to
respond to all of these agents it will be possible to deduce
the nature of the reactive mechanism in the cell. Howev-
er, more careful analysis of the biological facts frequently
indicates that the different agents acting upon a cell may
be reacting with quite different systems, although there
may be a great deal in common in the final results of
such action. For example, a muscle may be stimulated to
contract both by treatment with potassium chloride and
by treatment with acetyl choline: but the quantities of
these substances which are required to produce con-
traction are so remarkably different that it is impossible
to credit that they act upon the same system. Then again,
when the details of vesication are considered , one finds that
there are marked differences between the vesicles pro-
duced by different agents. Plate II (pages ii6 and 117)
shows three types of vesicle, produced by three different

NATURE OF BIOLOGICAL RESPONSES 121

types of agent acting upon the skin of the frog. In the one
case, the skin has split between the cornified layer and the
prickle cells; in the second case, the skin has split in the
middle of the prickle cell layer; and in the third case,
the prickle cell layer and the cornified layer have become
detached from the dermis. If the action of these three
agents had been assessed as either "vesication" or "not
vesication", all three would have been assessed as vesi-
cants and presumed to act through the same mechanism.
But the details of their biological action are in fact so
diflterent that it must be concluded that only part of
their action at most can be exerted upon a common
mechanism.

From these examples, as from many others which
could be adduced, it must be concluded that a cell or
tissue is commonly designed to fulfil a particular pur-
pose, and that it responds to stimuli of very diverse types
in a manner which is characteristic of the cells involved,
and not necessarily of the natureof the stimulus. Differ-
ent stimulating agents may act upon quite different
cellular systems, but the design of a cell is commonly
such that these mechanisms are funnelled to result in the
elicitation of a response characteristic of the particular
design of cell. With this point in mind we can proceed to
examine a number of types of biological response in
rather more detail. These will include artificial partheno-
genesis and mitotic abnormalities, and the responses of
genetic systems to drugs.

122 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

Artificial Parthenogenesis

In Fig. 20 we see arranged a list of agents which are com-
petent to cause artificial parthenogenesis. The first group
including substances such as fatty acids, saponin and
organic bases, have in conamon the possibility that they
may act upon the surface of the cell causing, as J.Loeb

Superficial cytolysis
(fatty acidi, saponin, etc)

Pricking

Ultraviolet light

». Permeability change

(change in resting potential?)

Dehydration by
hypertonic!

Dehydration after_
hypotonicity

Cold-

Gelation in
endoplasm

■Aster formation
Aster growth

Cell division
Fig. 20. Mode of action of agents causing artificial parthenogenesis

suggested, a superficial cytolysis to arise, or to become
incipient. With some species activation may also be
secured by totally different methods such as pricking,
the use of ultra-violet light, treatment with hypertonic or
hypotonic solutions, and treatment at low temperatures.
So far as is known, there is no common mechanism which
can be acted upon by these diverse agents and processes.
But all the reagents ultimately lead to cell division, and

MITOTIC POISONS I23

it seems probable that they can do so in the absence
of the cell nucleus. The characteristic required for cell
division appears to be a process involving an activity of
the cortical gel layer of the egg and also the formation of
two asters in the cytoplasm. The reagents also have in
common the fact that if used somewhat excessively they
cause not two asters but many asters to appear in the
cytoplasm. It seems probable, therefore, that the various
agents act at different stages in a chain of processes which
leads up to gelation of the endoplasm and a concomitant
activity of the cortical gel. Beyond this it is not possible
to proceed at present. If the mechanism of partheno-
genesis is to be understood it will probably be necessary
to work backwards from the biological response of aster
formation to the details of the mechanism of activation
involved in each particular type of reagent.

Mitotic Poisons

A considerable number of drugs have been classed as
mitotic poisons, and in the chemotherapy of cancer in-
creasing attention is being paid to the action of drugs
on mitosis. Mitosis is a complex process and can be
regarded as consisting normally of at least eight steps.
These are:

1. Division or duplication of chromosomes (involving
duplication of each gene).

2. Spiralisation of the chromosome accompanied by con-

124 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

densation upon it of nucleic acid, and a fading away
of nucleoli.

3. Division of the centrosome and breakdown of the
nuclear membrane.

4. Formation of the spindle and of the equatorial plate.

5. Division of the equatorial plate and movement of
chromosomes toward the centrosomes.

6. Division of the cytoplasm.

7. Reformation of nuclear membranes.

8. Despiralisation and loss of nucleic acid from the chro-
mosomes, and reconstitution of nucleoli.
Obviously mitosis is a very complex process, and if

mitotic poisons act directly it is obvious that derange-
ment will occur at many points, and through interference
with many quite distinct phenomena. It is, of course,
quite possible that some, if not all, of the known mitotic
poisons exercise their effect, not by acting upon the spe-
cific processes concerned exclusively in mitosis, but on
one or more of the processes which, amongst other
things, supply the necessary energy for the different
stages of mitosis.

Among the commoner effects of mitotic poisons are:

1 . The adhesion of chromosomes to one another (com-
monly called "stickiness") and the failure of daughter
chromosomes to separate completely from one another.
Such phenomena commonly lead to

2. The breaking of chromosomes, and the formation of
chromatin fragments which may not become at-

MITOTIC POISONS 125

tached to the spindle at subsequent divisions. These
phenomena lead to unequal distribution of chromatin
between daughter cells.

3. Failure of spindle formation. This may lead to poly-
ploidy, and on occasion, if the chromosomes have
failed to aggregate, to the formation of micronuclei,
each organised by one or a small group of chromo-
somes.

4. The formation of multi-polar spindles, perhaps due to
multiple division of the centrosome, or perhaps to the
formation of new centrosomes. The usual result of
this process is the uneven partition of chromatin be-
tween daughter cells, or occasionally the formation
of a multinucleate cell.

5. Chromosomes adhesions may prevent the complete
separation of the nuclei, and the result may be that
cell division does not go to completion.

6. All the processes through which the spindle goes may
be slowed down.

7. Over-spiralisation, and failure to despiralise after cell
division, may occur.

8. As a result of fragmentation and breaking, transloca-
tion may occur.

9. Cell division may fail to go to completion, resulting in
the formation of a bi- or multinucleate cell.

In view of the complexity of the phenomena involved
in mitosis and the great variety of abnormalities which
may occur, it is obviously insufficient to classify drugs

126 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

as mitotic poisons because they have the common effect
of deranging of mitosis. At present our knowledge of
the biochemical, biophysical and biological character-
istics of the various known mitotic poisons is quite in-
sufficient for an analysis of their actions, but some hints
of the type of sub-groupings that may be expected are
available. For example, colchicine differs frommostof the
other mitotic poisons in that, in low concentrations, its
action appears to be almost entirely restricted to sup-
pression of spindle-formation. It is therefore probable
that it attacks the cell in a different way from, say, some
of the nitrogen mustards, with which bridge formation
and fragmentation are very prominent phenomena.

OsTERGREN has shown that a very wide variety of or-
ganic compounds may cause stickiness (adhesion of
chromosomes) and suppression of spindle-formation.
The equitoxic concentrations of these substances are
roughly proportional to their oil-water partition coeffi-
cients, and it may be that in such cases the toxic effect
is produced by a mechanism similar to that discussed
in the chapter on narcosis. A small number of substances
are active as mitotic poisons in concentrations much
smaller than would be suggested by their oil-water par-
tition coefficients. This suggests that their action is a
specific action on the cell at some point more intimately
related to mitosis than is the generalised effect of organic
compounds which is proportional to the oil-water par-
tition coefficients. It does not, however, necessarily

t

1

1*'

I

Plate III. A comparison of the distributions of deoxypentose nucleic
acid and alkaline phosphatase in cells of the rat Walker sarcoma,
poisoned with a nitrogen mustard, a Failure of spindle formation -

Feulgen. b Failure of spindle formation - phosphatase, c Pycnotic de-
generation - alkaline phosphatase, d Pycnotic degeneration - Feulgen.

(seep. 134).

MITOTIC POISONS I31

follow that the specificity represents an exceptional po-
tency for reaction with, say, a chromosome or spindle
constituent. The apparent specificity may in fact be due
to the operation of a process not even distantly related
to mitosis. Thus, those compounds which readily fit into
the secretory patterns of cells, and so are found in dis-
proportionately high concentrations inside cells by com-
parison with other appaiently similar compounds, would
be expected to display an apparently abnormally high
toxicity if these high concentrations of drug are effective
in those regions of the cell which are intimately con-
nected with mitosis.

Among the substances having a disproportionately
high activity by comparison with their oil-water partition
coefficients are urethan and some members of the ni-
trogen mustard series. It has been suggested that ure-
than acts by interfering with nucleic acid metabolism,
and that the nitrogen mustards may also act in this way.
The evidence in support of this hypothesis is at present
very slender.

An interesting example of the difficulties which may
be encountered is found in the action of arsenical com-
pounds on mitosis. Substances such as sodium arsenite
and phenyl arsenoxide are fairly strong mitotic poisons,
and share this property with iodoacetamide. One of the
few properties which these three substances have in
common is that of combining vigorously with SH
groups, and it has therefore been suggested by Rapkine

132 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

that SH groups are of particular importance in mitosis.
It is not clear whether the SH groups concerned are to be
regarded as part of the chromosomes, part of the spindle
proteins or part of the general enzyme systems of the
cell. In support of the contention that SH groups are
particularly involved, it has been found that b.a.L. (dithio-
glycerol) is able to reverse the toxic effects produced by
sodium arsenite and phenyl arsenoxide, but not the toxic
effect of iodoacetamide. The position, however, is more
than a little complicated by the fact that b.a.l. is itself a
mitotic poison, and its toxic action can be reversed with
arsenoxide.

The present position is that whilst it is clear that sub-
stances having a fairly high degree of specificity for SH
groups are mitotic poisons, it is not yet clear that the true
point of attack of these substances is in fact upon SH
groups. And beyond this, it is still less clear whether the
supposed SH groups are, as some contend, part of the
spindle proteins, or whether they may not even be so
distantly related to the actual processes of mitosis as to
be the SH groups of enzyme systems concerned in the
mobilisation of energy, such as the SH groups of phos-
phokinases.

Substances which are generally regarded as having an
action upon respiratory mechanisms involved in the mo-
bilisation of energy, quite commonly have a very pro-
found effect upon mitosis. This is true of HCN, of
phenols such as hydroquinone, of urethan and of the

MITOTIC POISONS I33

quinones. Whilst certain members of these groups, such
as urethan, may on some cells act in such low concen-
tration as to suggest that an action upon non-respiratory
mechanisms is involved, others, such as phenyl urethan,
act at a level of concentration which appears to be similar
to that involved in the inhibition of respiration.

The ^-chloroethylamines. It is now well-known that many
compounds containing two /9-chloroethyl groups are po-
tent mitotic poisons. The first of these to be examined
was mustard gas itself. This was shown by Koller, in
studies on Tradescantia^ to produce chromosome breaks,
failure of division, the lagging of chromosomes and the
formation of bridges between the separating groups of
daughter chromosomes in anaphase and telephase.
Hughes and Fell have made a particular study of these
phenomena in tissue culture, and show that spindle ab-
normalities, such as tripolar spindles, are very common.
More recently attention has been focussed upon the
use of the so-called nitrogen mustard compounds of the
general type R.N:(CH2.CH2C1)2. These compounds in
some instances have a relatively selective effect upon the
growth of some types of tumours, and as Koller has
shown, this effect is probably largely produced by the
action of these compounds upon mitosis. Evidence has
recently been obtained by Revell that the hetero-chro-
matic regions of the resting nucleus are particularly sus-
ceptible to attack. On the chemical level, as with the

134 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

sulphur mustard compounds, activity seems to be asso-
ciated v^dth the abihty of the chloride ion to dissociate
from the jS-chloroethyl group, leaving a positively charged
carbonium ion which has the capacity to react with
many cell components. Although many speculations
have been put forward about the biochemical mechanism
through which these compounds act, the evidence so far
available is too sparse to enable any definite conclusion
to be drawn.

It is possible that a useful guide to the mode of action
of many mitotic poisons, including the nitrogen mus-
tards, will be obtained from cytochemical studies. At
present such studies are very few, and so far have been
more indicative of the degree of involvement of chemical
processes in the biological response itself than of the
biochemical mechanism initiating the biological response.
For example, some comparative studies have been
made of the Feulgen reaction for deoxypentose nucleic
acid and of the reaction of Takamatsu and Gomori for
alkaline phosphatase. Some of these results are shown
in Plate ill (pages 128 and 129). It will be seen that the
anomalies produced in the distribution of nucleic acid
are strikingly similar to those produced in the distri-
bution of phosphatase.

It must, of course, be remembered that the identifi-
cation of the biochemical mode of action of a compound
such as nitrogen mustard may prove to be extraordinari-
ly difficult by the biochemical techniques which are yet

MITOTIC POISONS I35

available. On theoretical grounds one would be inclined
to predict that most of the mitotic abnormalities which
are seen as a result of the action of a dose of nitrogen
mustard could originate as the result of the failure of one
or a few genes either to reproduce, or to function. The
reproduction, or functioning, of a single gene can prob-
ably be inhibited by combination with one molecule of
nitrogen mustard. Particularly significant in this con-
nection is some of the recent work of Herriot on viruses,
the reproduction of which he shows to be much more
sensitive to mustard than is any other biological process
so far examined. If it should prove to be true that the
most important site of action is upon the genes and that
the action is exerted by a small number of molecules of
nitrogen mustard, then quite exceptional methods will
be necessary to detect the exact site of action. Further-
more, the identification of the chemical action exerted
upon the gene will be very difficult to ascertain, partly
because of the quantities involved, and partly because
the reactions between the gene and the drug need not
necessarily be those which are regarded as in the main
course of chemical reactivity of the drugs concerned.
Where a few molecules only out of a relatively large
dosage can have a definitive action upon a biological proc-
ess, it is perfectly possible for the key reaction involved
to be amongst those which are normally classified by the
organic chemist as "side reactions", and be usually even
less understood than the characteristic reactions.

136 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

Reproduction of Bacteria and Viruses

A number of substances which have a very potent effect
upon the multiplication of some types of bacteria and
plants have also the property of producing at least some
of the phenomena which have been mentioned as charac-
teristic of the substances which have been classified as
mitotic poisons. Amongst the bacteriostatic substances
are, for example, the sulphonamides, which have been
shown to prevent cell division and cause polyploidy in
onion root tips. It may very well be that examination of
such cases would assist considerably in understanding
the mechanisms which may be involved in mitotic poi-
soning. But the analysis of these cases is not likely to be
particularly simple. Thus, according to the views de-
veloped by Woods and Fildes, the sulphonamides usu-
ally exercise their bacteriostatic effect by preventing the
utilisation of para-aminobenzoic acid: and Gale has
recently suggested that the primary action of penicillin
is to prevent the uptake of glutamic acid. The question
therefore arises as to whether the action of such sub-
stances as mitotic poisons involves the same inhibitions
as are supposed to be concerned in bacteriostasis, or
whether some other quite different actions are involved
when they act as mitotic poisons.

It is, of course, possible that the bacteriostatic action
is, in fact, identical with the action upon mitosis. But it
cannot at present be assumed that the phenomena of

REPRODUCTION OF BACTERIA AND VIRUSES I37

multiplication of genetically active units in bacteria in-
volve all the steps concerned in mitosis, particularly
when these events are viev^ed from the morphological
level. When we turn to the multiplication of bacterio-
phages and viruses it is possible that we are dealing with
a different or at all events much simplified process, with
which the possible routes of interference are more re-
stricted. It is quite likely that the reason why the common
bacteriostatic agents and mitotic poisons seem to be of
little use in the treatment of virus diseases is that some
of the specialised processes involved in mitosis simply
do not occur in the reproduction of viruses, and that it is
these rather specialised processes which are primarily
attacked by the known poisons.

If it is indeed the case that virus reproduction is a
simpler process than mitosis, it may be necessary to look
for rather different physico-chemical phenomena as pos-
sible modes of attack in the designing of substances which
will prevent the production of viruses. For example, the
study of viruses in vitro may well lead to the discovery of
types of substances which are selectively adsorbed upon
them: such substances may well prevent the reproduc-
tion of viruses. Then certain viruses have been shown to
attach themselves to particular points on the surfaces of
cells, and it has also been shown that particular sub-
stances may be necessary to secure this attachment.
From this, two possibilities arise for the design of com-
petitive substances: one type of substance which would

138 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

be of interest would be those compounds which can
adhere more vigorously than does the virus to the regions
of the cell surface which are specifically virus-adsorbing;
then, if a substance like tryptophane is required for the
adhesion of a virus to a cell, a satisfactory inhibition of
virus activity might be obtained by using a competitive
substance such as methyl tryptophan.

The relationship of mitotic poisons to mitotic stimu-
lators is also one which would probably repay more de-
tailed examination. Whilst it is possible that some stim-
ulating substances may be rather highly specific, other
substances which act as mitotic stimulators have a rel-
atively generalised effect. Xanthopterin has an effect
which is mainly restricted to the kidney and a smaller
effect upon the bone marrow. But oestrone has a rela-
tively generalised effect, producing an increase in mi-
totic rate in practically all cell lines capable of mitosis.
Despite this, oestrogens act as mitotic poisons for the
cells involved in cancer of the prostate and post meno-
pausal breast tumours. One wonders whether the mech-
anisms involved in mitotic poisoning and mitotic stim-
ulation are related or not. It is, of course, no new thing
to find that a substance may apparently have diametri-
cally opposite physiological effects under different cir-
cumstances: for example, adrenaline is vaso-constrictor
for some arterioles and vaso-dilator for other arterioles.

NUCLEAR AND CYTOPLASMIC DRUG ACTION I39

Nuclear and Cytoplasmic Drug Action

It is not unreasonable to suppose that some substances
will exercise their main action upon the cytoplasm of
cells and others will have their main action on the nuclei.
We can form an initial impression of what the differences
may be between these two lines of approach by consider-
ing studies upon cells from which the nuclei have been
removed. From such experiments we know that cells
without nuclei may retain their form, conduct electrical
impulses, exhibit the phenomena of amoeboid move-
ment, phagoc5^osis and intracellular digestion and even
divide. But such cells cannot differentiate and their life
appears to be restricted to about 20 days or less. From
these phenomena it is tempting to suggest that drugs
which have an immediate action exercise their effect pri-
marily upon cytoplasmic processes, whereas those with
a delayed action have a primary action upon the nucleus.
But whilst the first hypothesis may well be correct, more
doubt must attach to the second, insofar that there are
now reasons for supposing that in addition to the genet-
ical activity of the nucleus we must also consider genet-
ically active particles in the cytoplasm.

As an example of the type of phenomenon which has
to be considered may be mentioned the case of lewisite.
By using transparent chambers for the study of the skin
similar to those designed by Clark, it has been shown
that large or moderate doses of lewisite cause cell death

Cell Physiology 9

140 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

within a few hours. But when a very small dose of lewi-
site is applied to the skin, there is a transient effect which
passes off after about two hours, and may be followed
several days later by a much more profound effect which
may result in the death of cells after about a week (Fig.
21). It is very probable indeed that the destruction of

* Large rfose
'^ Small dose

/

Compjeie stoppage
of circulation
\Many clots in 'arger
vessels
ISome clots in larger
ivessels

} Massive diapedesis
.red cells
L J ] Slight diapedesis

I I \red cells

I * / • ^Stasis in capillaries

.^.^, / ^Dilatation of larger

\vessels

J Gross Capillary
dilatation

\ -^ -Capillary dilatation
Xj-x '£— I I Normal

0.1 0.5 10 5 10 50 100 200

Hours

Fig. 21. The response of the skin of the ear of rabbits to large and

small doses of lewisite

cells caused by relatively large doses of lewisite, and the
transient interference with cellular activity observed
soon after slight contamination, can be attributed to
cytoplasmic damage. But whether the secondary phe-
nomena developing with slight contamination are due to
cytoplasmic or nuclear damage is much more difficult
to decide. The secondary phenomena are probably closely
related to systemic poisoning by arsenical compounds,
so that the solving of this problem is not necessarily of
academic interest only.

DRUG ACTION UPON GENES I4I

Possible Modes of Drug Action upon Genes

At least three consequences may follow the action of a
drug upon a gene. i. The normal physiological action
of the gene may be reduced or abolished. 2. Mutation of
the gene may occur. 3. Reproduction of the gene may be
inhibited. If a cell changes its behaviour or nature under
the action of a drug it may or may not return to its initial
condition when the drug is removed. If the action of the
drug is reversible, its action is commonly said to be phys-
iological, whether the action is upon a gene or not.
If the action is irreversible it is thought to involve a mu-
tation, although it is usually only possible to prove this
when sex cells are involved.

It is of much interest that the same or similar substances
may often be simultaneously i . morphogenetic evocators,
2. carcinogenic and 3. hormones with a physiological
function. For example, members of the oestrogen series
have all three of these activities. At present one may well
be inclined, on theoretical grounds, to suggest that i. and
2. are in fact similar processes involving the mutation of
genes.

The situation is complicated by the fact that, although
most of the genetic phenomena with which we are accus-
tomed to deal are mediated by genes attached to chro-
mosomes and obeying the Mendelian laws, evidence is
steadily accumulating to show that some transmissbile
characteristics are carried, not necessarily by nuclear

Cell Physiology 9*

142 RESPONSES OF CELLS ON THE BIOLOGIC AL LEVEL.

genes, or not exclusively by nuclear genes, but also by
genetically active bodies in the cytoplasm, or plasma-
genes. Thus, even in the cases where it is suspected tht
mutation has occurred in sex cells, it is not always possi-
ble to test for this by the normal procedures which caa
be used for studying mutation of nuclear genes.

Haddow has suggested that mutations may rather
commonly take place under the action of growth inhib-
itors and that this mutation may take the form of per-
mitting the escape of a cell from the action of an inhib-
itor. In the normal animal growth in most organs in the
adult is restrained to just that degree which is necessary
to permit replacement of cells which have died. The proc-
ess by which this control is established is very far from
fully understood. But it is known that there are some
substances present in animal tissues which promote cell
growth and cell division, and others which inhibit this
process. It seems likely that the growth-promoting sub-
stances are sometimes in some sense used up by the
cells upon which they act, so that in the case of the sub-
stances which promote cell division there tends to be an
equilibrium established between the concentration of the
substance and the number of cells acting upon it. The
action of inhibitors upon this process is likely to be in the
direction of shifting the equilibrium position so that
fewer cells are in equilibrium with a given concentration
of growth promoter. There are a great many possible ways
in which the inhibitory substances might produce such

DRUG ACTION UPON GENES I43

an effect, amongst the n^ore entertaining of which is the
possibiHty that the inhibitors may in fact be substances
causing a more rapid turn-over of the growth-promoting
substance in the cells. It is a characteristic of this bal-
ance of control that it is adjusted so that normally mu-
tations permitting escape from this control are infrequent.
But when additional foreign growth inhibitors are present,
mutation appears to be more frequent, and may involve
not only escape from the foreign growth inhibitor but
also from the normal endogenous growth inhibitors.
Cells which suffer such mutations may give rise to
tumours. There is a good deal of evidence available now
suggesting that mutations, whether they be of nuclear
genes, or of plasma genes, may commonly take place
under the action of drugs, such as penicillin, the sulphon-
amides, arsenicals, etc. and thus give rise to strains of
cells (usually of micro-organisms) which are resistant
to the drug concerned.

If we are seriously to adopt the point of view which
has just been suggested, namely, that mutation may be
a fairly common event under appropriate conditions, we
must reconsider our attitude towards the stability of
genes. It has usually been supposed that genes are re-
markably stable bodies. This point of view has arisen be-
cause the occurrence of mutations is normally a very
infrequent process. However, when we consider the prob-
able chemical composition of genes, i.e. a combination
of deoxypentose nucleic acid and protein, there appears

144 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

to be no intrinsic probability that genes are by nature
very stable. On the contrary both the proteins and the
deoxypentose nucleic acids are very unstable towards
suitable reagents. Early attempts to produce mutations
by the action of chemicals were a failure, and these
failures tended to reinforce the view that genes were re-
markably stable. But consideration of the chemicals
which were studied shows that practically without ex-
ception they were substances which either would never
get into a cell without first killing it by destroying the
permeability of the plasma membrane, or else were sub-
stances which were certain to undergo almost instan-
taneous reaction with the cytoplasm after entering the
cell. I.e. the substances studied were almost all singularly
unlikely to make contact with the nuclear genes. More
recent experiments, starting with those of Robson,
AuERBACH and Roller on mustard gas, have shown that
chemical substances which have physical properties
which will both enable them to penetrate into the cell
nucleus and to react with the components of genes are
remarkably effective in producing mutations, thus con-
firming the evidence obtained by the study of radiations,
which also have the property of being able to penetrate
into the nucleus and secure a reaction with gene com-
ponents. We may thus conclude that the stability nor-
mally exhibited by genes is not an intrinsic refractor-
iness towards change, but is attributable to the genes
being present in a very stable environment. Towards

DRUG ACTION UPON GENES I45

suitable reagents we may expect genes to be very un-
stable, and it is possible that the morphogenetic evocators
and the more potent carcinogens are such reagents.

If the hypothesis just formulated should prove to be
correct, we should have a profitable line of approach to
a number of biological problems. As a particularly in-
teresting example we may take the work of Berenblum
on the induction of cancer by croton oil and dimethyl
benzanthracene. When croton oil is applied alone to the
skin of a mouse hyperplasia rapidly develops, but nor-
mally no tumours will develop during the natural life
of the animal. If dimethyl benzanthracene alone is
applied to the skin of a mouse hyperplasia results and
after some period tumours commonly develop. If both
croton oil and dimethyl benzanthracene are applied
together, tumours appear significantly more rapidly than
when dimethyl benzanthracene is applied alone. Even
more striking is the fact that if just one application of
dimethyl benzanthracene is made, tumours appear after
a rather protracted period. But if, after the application
of dimethyl benzanthracene, croton oil is applied, tu-
mours develop much more rapidly and there is a strong
tendency for the appearance of the tumour to occur at
a standard time after the application of croton oil, rather
than after the application of dimethyl benzanthracene.
At present the only possible interpretation of these re-
sults is that dimethyl benzanthracene rather readily
causes an irreversible change in the cells to which it

146 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

is applied, which remains latent until the damage is
revealed by an irritant such as croton oil. We could read-
ily understand these results if the irreversible damage
caused by the dimethyl benzanthracene is a mutation.

The Relationship between Hormones and Evocators

At present it is supposed that hormones are usually in
some way connected with the activities of enzymes, and
exercise their physiological effect through this connection.
The hormones may be prosthetic groups, coenzymes or
inhibitors etc. of the enzymes concerned. We also sus-
pect that genes consist of a specific array of enzymes.
This being so, we are inclined to suspect that hormones
and genes may have some fairly direct relationships.
A possible source of the relationship between the evocator
action and the physiological hormone action of a given
substance can be traced if we consider the circumstances
of gene reproduction.

One of the most striking characteristics of genes is
their ability to reproduce themselves. We do not know
a great deal about the conditions which are necessary for
the control of reproduction of genes. But two possible
extreme cases can be distinguished. Some genes are
perhaps in a cell quite intact at all times, and reproduce
themselves in the general environment provided by the
cell. Other genes may need additional specific substan-
ces present, which may be called primer substances.

RELATIONSHIP BETWEEN HORMONES AND EVOCATORS 147

before they can reproduce. Now if a gene combines with
a hormone it is no longer exactly the same body as it was
before that combination took place, and it is obvious
that it may be unable to reproduce itself during the pe-
riod in which it is changed by combination with the hor-
mone. Another possibility is that if it does reproduce, it
will not reproduce exactly as it was before combination
with the hormone, but as a new body: i.e. it may repro-
duce as a mutant of the original gene. With these points
is mind we can see that the nature of the action of a
hormone upon a gene may depend upon the fraction
of time with which gene is combined with hormone.
If the concentration of hormone is less than a roughly
defined concentration, the hormone will affect only the
physiological activity of the gene. But if the concen-
tration of the hormone exceeds this rough level, the
gene may be unable to reproduce, or may reproduce a
mutant gene. The possibility of a hormone producing
a mutation of a gene by combination with it is itself very
interesting. When we consider the case of a primer sub-
stance also being required for gene reproduction a num-
ber of related alternative mechanisms for the production
of mutant genes appear.

Similarly, if the action of the hormone is directly or
indirectly to prevent gene reproduction, then at least
two possibilities emerge. One is that the gene or primer
may be destroyed in the cell during the period in which
it is unable to reproduce. In this case it is treated virtu-

148 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

ally as a foreign body. Alternatively, cell division may
proceed until a cell is produced which lacks the gene or
prirner which cannot reproduce itself. Both possibilities
result in the appearance of a cell which is lacking the
gene upon which the hormone acts, i.e. deletion of the
gene has occurred.

Mutation or deletion of a gene need not become ob-
vious immediately. For example, let us consider the
case of a chromosome gene, which exercises its physio-
logical effect by producing a product which is itself com-
petent to reproduce, e.g. the product is a plasmagene.
Then, destruction of the chromosome gene may only
appear if, i . later circumstances also prevent self- repro-
duction of the plasmagene, or 2. later circumstances re-
sult in mutation of the plasmagene. In fact damage to
a gene caused by one circumstance may become appar-
ent only very much later, as the result of some other
circumstance quite unrelated to the first, and perhaps
many years may elapse before the second circumstance
occurs. It may be that it is phenomena such as these
which account for the appearance of a tumour after the
elapse of many years from the time of exposure to a car-
cinogenic agent. And phenomena of this type may be
involved in results such as have been described above
with croton oil and dimethyl benzanthracene. I.e. croton
oil may perhaps reveal the damage caused to a gene
by dimethyl benzanthracene, through an action upon
a plasmagene which is the normal product of the

REFERENCES I49

gene upon which diniethyl benzanthracene has its effect.
There are many other problems which might well be
investigated from points of view analogous to that which
we have been discussing. For example, some types of
leukemia respond well to treatment with urethan for a
period, but after that period become resistant to ure-
than. Is some phenomenon of mutation or deletion in-
volved here ? Then we might, for example, ask whether
a sulphonamide-induced agranulocytosis is an instance
of gene deletion ? But space does not permit of going into
further details in such matters.

REFERENCES

AuERBACH, C. and Robson, J. M., 1944: Nature, 154, 81.

Berenblum, I. and Shubik, P., 1947: Brit. J. Cancer, 1, 379.

Blakestee, a. F. and Avery, A. G., 1937: J. Hered., 28, 393.

Clark, A. J., 1933 : Mode of Action of Drugs on Cells (Arnold, London).

Clark, A. J., 1937: General Pharmacology (Handbuch der Exp.
Pharm., IV).

Danielli, J. P., 1940: In Cytology and Cell Phys. Edit. Bourne (Cla-
rendon, Press, Oxford).

Danielli, J. F. and Catcheside, D. G., 1945: iVafwre, 156, 294.

Darlington, C. D. and Roller, P. C, ig4.y: Heredity, 1, 187.

DusTiN, A. P., 1925: C r. Soc. Biol., 93, 465.

DuSTiN, P., 1947: Nature, 159, 794.

Eynny, H., Johnson, F. H. and Gensler, R. L., 1946: J. Phys. Chem.,

50, 453-
Gale, E. F., 1949: Symp. Soc. Exp. Biol. Ill, 233.
GoMORi, G., 1939: Proc. Soc. Exp. Biol, and Med., 42, 23.
Herriot, R. M., 1948: y. Gen. Physiol, 32, 221.
Hughes, A. and Fell, H. B., 1949: Quart. J. Micros. Sci., 90, 37.
Johnson, F. H., Brown, D. E. and Marsland, D., 1942: J. Cell.

Comp. Physiol., 20, 247, 269.
Karush, F. and Siegel, B. M., 1948: Science, 108, 107.

150 RESPONSES OF CELLS ON THE BIOLOGICAL LEVEL

Roller, P. C, 1947: Symp. Soc. Exp. Biol., I, 270 (Cambridge Univ.

Press London).
Roller, P. C, 1947: Brit. J. Cancer, 1, 38.
Needham, J., 1942: Biochemistry and Morphogenesis (Cambridge Univ.

Press, London).
OsTERGREN, G. and Lev AN, A., ig^^: Hereditas, 29, 496.
OsTERGREN, G., 1944: HcTcditas, 30, 429.
Revell, S. and Loveless, A., ig^g: Nature, 164, 938.
RoTHEN, A. 1946: J. Biol. Chem., 163, 345.
RoTHEN, A. 1947: jf. Biol. Chem., 167, 299.
Takamatsu, H., 1939: Trans. Soc. Path. Japan, 29, 492.
Woods, D. D. and Nimmo-Smith, R. H., 1949: Symp. Soc. Exp. Biol.

Ill, 177.

Author Index

Abramson, H. A., 45
Adam, N. K., 24, 45
Adrian, E. D,, 90
Alexander, A. E., 42, 45
Auerbach, C, 93, 96, 144, 149
Avery, A. G., 149

Baldwin, E., 96
Barlund, A., 73, 102, 114
Berenblum, I., 145, 149
Bergman, 76
Bernal, J. D., 13, 24,45
Blakestee, A. F., 149
Bourne, G., 24
Brachet, J., 4, 24
Brinley, F. J., 100, 114
Brown, D. E., 114, 149

Cameron, 93

Cannan, 92

Catcheside, D. G., 9, 149

Clark, A. J., 24, 73. 74, 96, 114,

139, 149
Collander, R., 49, 73

Cori, 92

Cullumbine, H., 96

Dale, H., 38, 45

Danielli, J. P., 9. 24, 27, 31, 45,

114, 149
Darlington, C. D., 24, 149
Davies, J. T., 24, 45
Davson, H., 24, 73, 102, 114
Dixon, M., 89, 91, 92, 93, 96
Dustin, A. P., 149
Dustin, P., 149

Eynny, H., 149

Fankuchen, A., 24, 43, 45
Fell, H. B., 149
Feulgen, 134
Fildes, 136

Gale, E. P., 136, 149
Gensler, R. L., 149
Gilman, A., 96
Gomori, G., 134, 149
Gray, J., 24

Haddow, A., 142
Hawking, 67
Herriot, R. M., 135, I49
Hiller, S., 100, 114
Hober, R., loi, 114
Hughes, A., 149

Jacobs, 102

Johnson, P. H., 113, 114, I49

Karush, P., 149
Keilin, D., 74
King, 67

Roller, P. C, 93, 96, i33, i44,
149, 150

Levan, A., 150

Lillie, R. S., 73, loi, 114

Loveless, A., 150

Marsland, D., 100, loi, 114, i49

McAnally, M., 73

Meyer, K.H., 105, 106, 113, 114

Mirsky, A. E., 8, 24

Moore, 5

152

AUTHOR INDEX

Needham, D. M,, 91, 92, 93, 96,

150
Nimmo-Smith, R. H., 150

Ostergren, G., 126, 150
Overton, 106, 113

Parpart, 102

Peters, R. A., 69, 73, 74, 96
Phillips, G. S.,96
Phillipson, J., 73

Quastel, J. H., 69, 73, iii, 112,
114

Rapkine, 131

Revell, S., 133, 150

Rideal, E. K., 33, 45

Ris, H., 24

Robson, J. M., 93, 96, 144, 149

Rothen, A., 45, 150

Saunders, 90
Schulman, J. H., 33, 45
Shubik, P., 149
Siegel, B. M., 149
Stedman, E., 8, 24
Stocken, J. R., 69, 73

Takamatsu, H., 134, 150
Thompson, R. H. S., 69, 73
Traube, I., 102, 105, 114
Trim, A. R., 42, 45, 66

Voegtiin, F. R., 69, 73

Webb, D. A., 27, 45
Wilson, E. B., 24
Winterstein, H., 73, loi, 114
Woods, D. D., 136, 150
Work, E., 96
Work, T. S.,96

Subject Index

Activators, 15, 75
Active patches, 19
Acetylcholine, 19
Adrenaline, 34, 138
Agranulocytosis, 149
Alcohols, 98

Amoeboid movement, 5, 86
Anthelminthics, 66
Antibodies, 21, 22
Arsenic acid, 77, 83
Arsenical poisoning, 70
Arsenicals, 143
Arsenite, 88

Arsenoxides, 64, 67, 76, 131
Arterioles, skin, 86
Asters, 13, 123
Azide, 83

Bacteriostasis, 136
B.A.L., 70, 132
— , glucoside, 72
Blood-brain barrier, 64
Brain, 63, no
— tissue, 112
Butyl alcohol, 102

Calcium, 26, 28, 29, 30
Cancer, 138, 141, 145
Carbon monoxyde, 83
Carriers, 75
Cell division, 5, 39, 86
Cell form, 5
Cellular structure, 12
Centrifugal force, 3
Centrifugation, 4
Centrosome, 124

/?-Chloroethylamines, 88
Cholic acid, 64
Choline esterase, 89, 90, 91
Chromomeres, 7, 8
Chromosomes, 7, 8, 15, 44, 123,

131, 133, 141, 148
Ciliary movement, 86
Cocaine, 100
Coenzymes, 76, 146

— 11, 76

Colchicine, 88, 126
Cortical gels, 13, 123
Croton oil, 145, 148
Curare, 100

Cytochemical studies, 2, 6, 8
Cytochemistry of hepatic cells, x
Cytochrome, 75

— system, 82
Cytolysis, 36

Defence mechanisms, 20
Dehydrogenases, 75, 76, 77, 89
Detoxication, 20
Dichlorophenol, no
Dielectric constant of cellular

systems, 17
Differentiation, 44
Diffusion, 46, 53, 54, 63
Dimethyl benzanthracene, 145,

148
Dithioglycerol, 88
Dithiol, 69

Enzyme(s), 15, 17, 19, 21, 23, 69,

76, 78, 109, 134, 146
— , action, 44, 48

154

SUBJECT INDEX

Enzyme, cellular, 74
— poisons, 60
Equatorial plate, 124
Eserine, 90
Ethyl iodoacetate, 88
Evocators, 141, 146

Fluoride, 76, 83
Fluorophosphonates, 91

Gel, 5

— , cortical, 13, 123

— , protoplasmic, 4

Genes, 8, 11, 16, 19, 23, 123,

134, 137, 141, 143. 144, 146,

148
Glucose, 83
Glyceraldehyde, 76, 83
Glycolysis, 82, 86
Granules, 14, 15, 20

Heavy metals, 32

— , oligodynamic effect, 30

Hexokinase, 89, 9°. 9i> 92, 95

Hexyl resorcinol, 43

High pressure, 113

Hormones, 141, 146, 148

Hydrocyanic acid, 75> 77> 81, 83,

84,85, 100, 132
Hydrogen sulphide, 81, 100

Inhibitors, 15, 75, 142, 146
— , competitive, 76
Iodine, 76
lodoacetamide, 131
Iodoacetate, 76, 83, 85, 88
Ions, 25, 27, 29, 30, 53, 102
Isotopes, 19

Lachrymation, 88
Lachrymators, 89
Leucotaxin, 92

Leukemia, 149

Lewisite, 69, 70, 83, 84, 91, 92,

139
Lipoid membrane, 61

— molecules, 14

— solubility, 63
Lipoids, 36
Lysis, 36

Magnesium, 100
Malonate, 76
Mastitis in cattle, 64
Membrane, cell, loi
— , lipoid, 61
— , nuclear, 18, 124
— , plasma, 36

— properties, 93
Membranes, 15
— , natural, 46
Metals, heavy, 32
Methylene blue, 75
Micelles, 14, 42

— formation, 39, 40
Mitosis, 4, 13,93, 123, 126, 133

136, 137
Monolayer, 26, 32
— , protein, 37, 38
Muscle, 82, 120

— cell, 88

— cells, striated, 6

— contraction, 86
Mustard, 131, 134

Mustard gas, 83, 84, 85, 9 1, 93, 144
Mutation, 141, 146, 148
Mutations, 74, 142, 143
Myosin, 6
Myotics, 91

Narcotics, 97
Nerve, 120
Nicotinic acid, 76
Nitrogen mustards, 76

SUBJECT INDEX

155

Nuclear membrane, 18, 124

Nuclei, 94, 139

Nucleic acids, 2, 6, 14, 15, 16,

19, 124, 143
Nucleoli, 15, 124
Nucleus, 5, 16, 93, 133

Oestrogens, 37, 38
Oestrone, 138
Oligodynamic action, 32

— effect of heavy metals, 30
Osmic acid, 88

Oxygen, 83

Paraffins, loi
Parthenogenesis, 122
Partition coefficients, 126

— effects, 105
Penicillin, 136, 143
Permeability, 18, 22, 48, 50, 52,

54, 58, 60, 63, 66, 68, 92, 99,
loi, 144

PH, II

— buffers, 16
Phagocytosis, 5
Phenol, 79
Phenols, 132
Phosphokinases, 92
Picric acid, 100
Plasmagenes, 142, 148
Plasma membrane, 36
Polarisation, 18
Potassium, 88
Prosthetic groups, 76, 146
Proteases, 92

Proteins, 8, 14, 15, 16, 19, 28,36,

74, 113, 143
— , monolayer, 37, 38

— , spindle, 132
Protoplasmic streaming, 5
Prussic acid, see Hydrocyanic
acid

Purple, visual, 19
Pyocyanin, 75
Pyridoxal, 76
Pyruvic oxidase, 92, 93

Quinones, 133

Radiation sickness, 93
Radiations, 144
Reactions at interfaces, 33
Removers, substrate, 77
Renal secretion, 86
Resistance, 21, 143
Respiration, 82, no
— , cellular, 86
Rumen, 59

Secretion, 46, 60, 94

Secretory activity, 22

Self-reproduction, 23

Silver, 48

Slime moulds, 5

Soap, 88, 90

Spindle(s), 13, 124, 131, 133

— proteins, 132
Structure, cellular, 12
Substrates, 15
Substrate removers, 77
Sulphonamides, 76, 88, 136, 143,,

149
Surface action, 15

— tension, 102
Surfaces, 99, 103, 104

— of skin cells, 92

Tactoids, 13
Trypanosomes, 67, 68
Tryptophan, 138

Urethan, 83, 88, no, 131, 132^
149

156 SUBJECT INDEX

Vacuoles, 20 Visual purple, 19

Vesicant substances, 95 Vitamin Bi, 76, 83

Vesicants, 90, 91 Vitamin Bj, 76

Vesication, 70, 120
Viruses, 23, 134, i37 Xanthopterin, 138

y