BIOLOGY
I40RARY

NUTRITIONAL FACTORS IN THE GROWTH
OF CERTAIN YEASTS AND BACTERIA

By
Louis Freedman,....ph. B., M. S.

NEW YORK CITY
1922

DISSERTATION

submitted in partial fulfillment of the requirements for the degree of Doctor
of Philosophy in the Faculty of Pare Science, Columbia University, in the
City of New York-

NUTRITIONAL FACTORS IN THE GROWTH
OF CERTAIN YEASTS AND BACTERIA

By
Louis Freedman, Ph. B., M. S.

o
NEW YORK CITY

1922

DISSERTATION

submitted in partial fulfillment of the requirements for the degree of Doctor
of Philosophy in the Faculty of Pure Science, Columbia University, in the
City of New York.

; t-

CONTENTS.

PART I — VITAMINES

Page

Introduction 5

Experimental Procedure 5

Summary of Conclusions 15

Bibliography 16

PART II — PROTEIN HYDROLYSATES

Introduction 17

Experimental Procedure 18

Summary of Conclusions 27

Bibliography 28

502082

NUTRITION FACTORS IN YEAST AND BACTERIA

AUTHOR'S NOTE.

The two papers that constitute this dissertation were re-
printed from the Journal of Metabolic Research (1922, volume I,
No. 4, pages 457-480), where they were published under the
same titles in collaboration with Dr. Casimir Funk.

NUTRITIONAL FACTORS IN THE GROWTH OF CERTAIN
YEASTS AND BACTERIA.f

I. VlTAMINES.
By LOUIS FREEDMAN

Biochemical Laboratory of Columbia University at the

College of Physicians and Surgeons, New York City,
and the Research Laboratory of H. A. Metz, Ne<w York City.

INTRODUCTION.

Since the discovery of Wildier's yeast stimulating "bios" in
1901, rapid progress has been made in determining the nutri-
tional requirements of various yeasts and bacteria.

The early work of many investigators has shown that both
yeasts and bacteria require for their growth certain unknown
substances, the properties of which correspond closely to those
of our present day water-soluble B vitamine. Of late, however,
considerable discussion has arisen regarding the identity of the
substance that promotes the growth of yeast cells with that oi'
vitamine B, and also its relation to the substance that stimulates
the growth of bacteria.

In the present work, the authors have made an attempt to
remove some of the uncertainties regarding these factors, and
to show that a close nutritive analogy exists between yeasts
and bacteria.

EXPERIMENTAL PROCEDURE.

A strain of hemolytic streptococcus, which was kindly fur-
nished by Dr. Mueller,** was used. The medium for the growth

t Read before the Society for Experimental Biology and Medicine, New York City,
February IS, 1922.

** The authors are indebted to Dr. T. Howard Mueller of the Department of Bacteriology,
College of Physicians and Surgeons, for the use of a strain of streptococcus, and for a
description of the preparation of his media, and of the technique of carrying out the bac-
teriological tests.

6 NUTRITION FACTORS IN YEAST AND BACTERIA

of this organism was prepared as follows: A fresh beef-
heart, devoid of fat, and finely chopped, was boiled for five min-
utes with an equal weight of distilled water; the infusion was
strained through cloth and filtered, clear, through filter paper.
This beef-heart infusion was boiled for 15 minutes with two
per cent of its weight of norit charcoal; the decolorized liquid
(to be referred to hereafter as D. I.) was filtered clear, and
mixed with an equal volume of a glucose-salt solutionf (to be
termed G. S.). This mixture was boiled for several minutes,
filtered, adjusted to a PH of 7.4 and sterilized in lots of 250 cc.
at ten pounds pressure for ten minutes. This medium (referred
to hereafter as D.I.-G.S.), inoculated with streptococcus, failed
to give growth when used alone, whereas the original undecolor-
ized heart infusion was favorable for the growth of this
organism.

EFFECTS OF VARIOUS MEDIA ON GROWTH OF STREPTOCOCCI.

The tests on streptococci were made as follows: Duplicate
tubes were prepared containing nine cc. of the D.I.-G.S. solu-
tion plus one cc. of the solution to be tested, one tube being
inoculated and the other acting as a sterile control. One set of
duplicate tubes of D.I.-G.S. alone, and another set containing this
solution plus one per cent peptone, were used as additional sterile
and inoculated controls. The inoculations were made from a
24-hour bouillon culture previously inoculated from a pure blood
culture.

The original undecolorized heart infusion gave a profuse
growth when inoculated with streptococci, while the decolorized
medium, either alone or together with the glucose-salt solution,
failed to give growth. The decolorized medium, however, gave
growth when one percent peptone was added to it, while one
percent peptone solution alone or together with glucose-salt solu-
tion, failed to give growth. This confirms some of the results
obtained by Mueller.1

Autolyzed brewers' yeast, diluted 20 times, when added to
the D.I.-G.S. medium, supported the growth of streptococci ; but
in some cases the growth was retarded, due probably to the
strong acid reaction of the autolyzed yeast. When the reaction

f The glucose-salt (G. S.) solution consisted of an aqueous solution of: NaO— 1.0%;
C»C1«— 0.02%; MgSO4— 0.04%; K»HPO4— 0.20%; Glucose (Difco)— 0.20%.

LOUIS FREEDMAN 7

was adjusted to an alkaline range (Pn of 7.4-7.8), a more uni-
form growth of the microorganism was obtained. An added
factor may be the possible presence in autolyzed yeast of sub-
stances which are toxic to the specific organism. The above
results are given in detail in table I.

TABLE I.
Growth of streptococci on various media.

No. MEDIA Growth

1. 10 cc. Undecolorized beef -heart infusion ...................

2. 10 cc. Decolorized beef -heart infusion

3. 10 cc. Glucose-salt solution (G. S.)

4. 10 cc. D. I.-G. S. solution

6. 10 cc. D. I.-G. S. solution, containing 1% peptone

6. 10 cc. G. S. solution containing 1% peptone ................. —

7. 10 cc. 1% peptone solution alone ........................... —

8. 9 cc. D. I.-G. S. + 1 cc. 5% autolyzed yeast ............... +

9. 9cc. D. I.-G. S. + 1 cc. 6% autolyzed yeast (Pn 7.4-7.8)... +

-}- + + Denotes profuse growth.
4- + " moderate "

-f slight

— no

EFFECTS OF THE SAME MEDIA ON THE GROWTH OF YEAST CELLS.

For the test on the growth of yeast cells, we followed the
Funk-Dubin2 method, which is the most convenient and which,
briefly, is carried out as follows : A yeast suspension is prepared
by shaking a loopful of a 48-hour pure yeast culture in 100 cc.
Nageli solution on a shaking machine for three hours. Duplicate
sets of tubes are prepared containing

(1) 4 cc. yeast suspension + 5 cc. Nageli solution + 1 cc. water

(2) 4 cc. yeast suspension + 5 cc. Nageli solution + 1 cc. of the

solution to be tested.

The tubes are incubated for 20 hours at 30° C., and the
fermentation is then stopped by heating the tubes in a water
bath to 75 °C. The contents of the tubes are then transferred to
special centrifuge tubes, the bottom part of which is a capillary
2.5 cm. long and is graduated in millimeters. These tubes are
centrifuged for 15 minutes at about 2500 R.P.M., and the growth
of the yeast cells is read directly on the tube. The reading of
tube 1, which constitutes the blank, is subtracted from that of
tube 2 to give the net growth of yeast cells in millimeters. In all
tests, conditions were maintained which precluded the possibility
of bacterial contamination.

8 NUTRITION FACTORS IN YEAST AND BACTERIA

Undecolorized beef-heart infusion gives a growth of yeast
cells which compares favorably with that produced by a 5%
solution of autolyzed yeast itself. The addition of glucose to
autolyzed yeast has very little if any additional activating prop-
erty, other than that of slightly enriching the medium. One cc.
of decolorized heart infusion when added to the Nageli solution
with or without the glucose-salt solution, has little or no effect on
yeast cells; while the D.I.-G.S. medium containing one percent
of peptone, when added to the Nageli solution, stimulates the
growth of yeast cells just exactly as much as when one percent
peptone solution alone is added.

TABLE II.

Effects of various media on the growth of yeast cells.

Net yeast

Substance added to 9 cc. of standard Nageli solution Growth in

No. mm.

1. 1 cc. 5% autolyzed yeast (5 CQ. in 95 cc. water) 10.0

2. Ice. 5% autolyzed yeast -f- 1 cc. 10% glucose solution (Difco) 11.0

3. 1 cc. D. I. solution alone 0.

4. 1 cc. D.-I.-G. S. solution 1.5

6. 1 cc. 1% peptone solution (Difco) 2.5

6. 1 cc. D. I.-G. S. solution containing 1% peptone 4.0

7. 1 cc. G. S. solution 0.

8. 1 cc. G. S. solution containing 1% peptone 2.0

9. 1 cc. Undecolorized meat infusion , 7.5

10. 1 cc. Undecolorized beef -heart infusion 12.0

It is apparent from the results shown in tables I and II,
that there is present in beef and beef-heart infusions, and in
peptone and autolyzed yeast, a substance that shows compar-
able growth-stimulation on hemolytic streptococci and yeast cells.
The question now naturally arises as to the nature of the sub-
stance that shows such a marked similarity of effect on these
two organisms. From the standpoint of the favorable action on
yeast cells, we are probably dealing here with one or more sub-
stances of the class of B vitamine. As for the growth-influencing
action on streptococci, this either is due to a substance of an
identical or similar nature; or, as suggested in the work of
Davis,3 Rivers and Poole,4 and others, we are dealing with two
unknown substances, one or both of which may belong to the class
of vitamines. These authors found that certain organisms, par-
ticularly those of the hemophilic type, require two substances^for

LOUIS FREEDMAN 9

growth, both of these substances being present in blood. Rivers
and Poole4 found that one of the substances was autoclave stable,
while the other was labile, and that the latter could also be found
in yeast.

It was quite logical to surmise, therefore, that by heating
beef-heart infusions with larger amounts of norit, we might
extract a second substance or even more substances, the re-
moval of which would render the D.I.-G.S. solution inactive, even
on addition of peptone or autolyzed yeast. Accordingly two
portions of an undecolorized beef-heart infusion were boiled
with five and ten percent by weight of norit respectively ; but no
markedly different results were obtained on the yeast or bacterial
growth, as can be seen by comparing table III below with tables
I and II.

This second unknown substance present in beef-heart,
thought to be in some way associated with blood, is therefore
not related to the class of B vitamine, as it is not extracted by
any of the adsorbing agents which have an affinity for this
class of vitamines.

TABLE III.

Growth of yeast and streptococci on beef-heart infusions,
decolorized with 5 and 10 norit.

Net yeast Bacterial

No. SUBSTANCE Growth* Growth

in mm.

1. 10 cc. Undecolorized beef -heart infusion .......... 17.5 + + +

2. 6cc. " " "

4- 5 cc. G. S. solution ......................... 18.0 4- + 4-

8. 10 cc. Beef -heart infusion, decolorized with 5%

norit ....................................... 1.0 —

4. 5cc. Beef -heart infusion, decolorized with 5%

norit, 4- 5 cc. G. S. solution ................... 1.5 —

5. 10 cc. Beef -heart infusion, decolorized with 10%

norit ....................................... 0. —

6. 5cc. Beef -heart infusion, decolorized with 10%

norit, 4- 5 cc. G. S. solution ................... 0.5 —

7. 10 cc. Beef -heart infusion, decolorized with 5%

norit, and containing 1% peptone .............. 3.0 4-

8. 5cc. Beef -heart infusion, decolorized with 5%

norit, and containing: 1% peptone, -f- 5 cc. G. S. 2.5f 4- 4-

9. 10 cc. Beef -heart infusion, decolorized with 10%

norit, and containing 1% peptone ............. 2.5 +

10. 5cc. Beef -heart infusion, decolorized with 10%

norit, and containing 1% peptone, + 5 cc. G. S. 2.0f 4- 4-

* For the yeast test, 1 cc. of each of the above mixtures was added to the standard
Nageli solution.

•f The slight decrease in activity here is due to the smaller amount of the beef-heart
infusion.

10 NUTRITION FACTORS IN YEAST AND BACTERIA

SHAKING OF AUTOLYZED YEAST WITH FULLERS EARTH AND NORIT.

Funk and Dubin5 have shown that at least two different sub-
stances can be separated from autolyzed yeast by means of
shaking with fullers earth. By this method it is now possible to
separate the vitamine active for yeast growth, which has been
provisionally called "vitamine D," from the anti-beriberi or B
vitamine.

Several preparations of autolyzed yeast were subjected to
shaking with various amounts of both fullers earth and norit,
as follows : The clear filtrate of autolyzed yeast, after removal
of the protein by heat coagulation, was first shaken for three
hours with 50 grams of fullers earth per liter, and filtered.
The filtrate showed very little loss of growth-stimulating effect
on yeast cells as compared with the original material, while
there was no observable loss of stimulation on bacterial growth.
This filtrate was again subjected to shaking with fullers earth,
this time with 100 grams per liter. The filtrate from this shak-
ing gave a yeast growth-increase about one half as great as that
given by the original 5% autolyzed yeast solution. This filtrate
gave as a rule no growth of streptococci, although in one or two
cases a slight growth was obtained. This second filtrate was
again shaken for three hours with 100 grams of fullers earth
per liter, the filtrate from this final shaking having almost no
stimulating effect on yeast growth and none on bacterial growth.

We obtained practically similar results by subjecting auto-
lyzed yeast solutions to shaking with norit. The first filtrate
after shaking with 50 grams of norit per liter was still strongly
active on both organisms, while the second filtrate, obtained
after shaking the first with 100 grams per liter, showed a slight
effect on yeast growth, but was entirely negative on the growth
of streptococci, as was to be expected. The third filtrate, ob-
tained after again shaking the previous filtrate with 100 grams
of norit per liter, had lost all activity on the growth of both
organisms.

A few experiments were also tried with Lloyd's reagent, the
results comparing favorably with those obtained with the other
two adsorbing reagents, although Lloyd's reagent is less ener-
getic an adsorber than fullers earth or norit.

LOUIS FREEDMAN 11

EXTRACTION OF THE ADSORBED MATERIAL FROM FULLERS EARTH

AND NORIT BY MEANS OF BARYTA AND ACETIC ACID

RESPECTIVELY.

Following the method of Seidell,6 the fullers earth contain-
ing the adsorbed vitamine-like material was shaken for two
to three minutes with one and a half volumes of a 10% solution
of baryta at 60 °C. The solid was then filtered off, washed with
water, and the combined filtrate and washings quickly neutral-
ized with 20% sulphuric acid. The filtrate, after removal of the
barium sulphate, was concentrated in vacuum and made up to a
volume equal to the original volume of the medium before treat-
ment with the adsorbing agent. The solution was then adjusted
to a PH of 7.4 and sterilized. Unless otherwise noted, one cc. of
the solution was used in all tests.

The baryta extract of the fullers earth from the first shak-
ing with autolyzed yeast (50 grams per liter) showed a com-
paratively strong growth stimulating action on the growth of
both yeast cells and streptococci, as can be seen in table IV
below. This extract contains nearly all of the vitamine B from
autolyzed yeast and a small part of the newly isolated vitamine
D. Its stimulating effect on yeast growth compares with that
obtained by Funk and Dubin5 as shown in the second column
of table IV. It also showed a comparatively strong stimulation
of the growth of bacteria.

The baryta extract of the fullers earth from the second
shaking with autolyzed yeast (100 grams per liter) had an
effect on yeast growth almost equal to that of the first baryta
extract; while the action on the growth of streptococci was
only slightly diminished as shown by 2 (a) and 3 (a). This
extract had no curative effect on polyneuritic pigeons, showing
the apparent absence of vitamine B. The fullers earth, during
the third shaking, failed to extract any more of the active sub-
stance, as its baryta extract showed no effect on yeast and
bacterial growth.

Extraction of the norit ivith acetic acid: Following the
method of Eddy, Stevenson, Johnson and Heft,7 the material
adsorbed by the charcoal was extracted from it by heating the
charcoal on a water bath for three hours with ten parts (by
weight) of glacial acetic acid; the solution was filtered, and
evaporated in vacuum to dryness; the residue was taken up
in water, and the solution again evaporated. This procedure

12

NUTRITION FACTORS IN YEAST AND BACTERIA

was repeated several times to remove most of the acid. The
last traces of acetic acid were neutralized by addition of nor-
mal sodium hydroxide solution. The neutralized solution was
made up to a volume equal to the original volume of the autolyzed
yeast, adjusted to the proper PH (7.4) and sterilized. One cc.
of this extract was used in each test.

These acetic acid extracts from the charcoal gave very in-
teresting results. As shown in table IV, Nos. 5 (a) and 6 (a),
they are practically in accord with the results obtained from
the baryta extracts from the fullers earth after adsorption of
the vitamine-like substance. The first two extracts stimulate
both yeast and bacterial growth, while the extract from the
third shaking [7 (a) ] shows a very slight stimulating effect on
yeast growth and no effect on the growth of streptococci.

TABLE IV.

Effect of Shaking of Autolyzed Yeast with Fullers Earth

and Norit.

No.

FULLERS EARTH

1. Autolyzed yeast (5% solution)

2. Autolyzed yeast shaken with 50

grams per liter

2 (a) Baryta extract of fullers earth

(from 2)

3. Autolyzed yeast (filtrate from 2)

shaken with 100 grams per liter

3 (a) Baryta extract of fullers earth

(from 3)

4. Autolyzed yeast (filtrate from 3)

shaken with 100 grams per liter

4 (a) Baryta extract of fullers earth

(from 4)

Net Yeast Growth*
in mm. obtained by
Funk and Our-

Dubin selves

14.5 12.5

NOHIT

5.

Autolyzed yeast shaken with 50
grams per liter

5 (a) Acetic acid extract of norit

(from 5)

6. Autolyzed yeast (filtrate from 5)

shaken with 100 grams per liter

6 (a) Acetic acid extract of norit

(from 6)

7. Autolyzed yeast (filtrate from 6)

shaken with 100 grams per liter

7 (a) Acetic acid extract of norit

(from 7)

12.0
4.0
6.0
3.5
0.5
0.

13.5
3.0
3.0
3.0
0.5
0.

9.5

5.5
5.0
4.0
0.5
0.

10.5
4.0
3.0
3.0
0.
0.

Bac-
terial
growth

44-
4-4-
4-4

4-4-

4-
4-4-

4-

* Slight variations between the results obtained by Funk and Dubin and by ourselves
are due to the use of different preparations of autolyzed yeast.

LOUIS FREEDMAN 13

Our results as detailed in table IV, show that we can obtain
from autolyzed yeast by fractional adsorption with fullers earth
and norit, and subsequent extraction of the adsorbents, a con-
centrated solution of vitamine D, almost free from vitamine B.
This vitamine D extract is active in stimulating the growth of
both yeast cells and streptococci, but is not curative for beriberi.
We were, however, unable to obtain vitamine B entirely free
from D as this fraction which was found to be curative for beri-
beri was also active in stimulating the growth of yeast cells and
bacteria. We have also shown that the extraction of these active
substances from the adsorbents with baryta and acetic acid is
almost quantitative.

SHAKING OF BEEF-HEART INFUSIONS WITH FULLERS EARTH AND

NORIT.

A beef-heart infusion, prepared as described in the early
part of this paper, when shaken for three hours with 50 grams
of fullers earth per liter, lost most of its stimulating effect on
yeast growth, while it still retained enough of the active sub-
stance to promote the growth of streptococci. The nitrate, after
a second shaking with fullers earth, (100 grams per liter)
showed no activity on the growth of either yeast or bacteria.

The baryta extract of the fullers earth from the first shak-
ing, peculiarly showed only a very slight growth-stimulating
activity on yeast; even the addition of 2 cc. of the extract to
the medium increased the growth of yeast cells only slightly.
The baryta extract of the fullers earth from the second shaking
was correspondingly less active on yeast growth, showing a net
growth of only 1 mm. ; while it was only slightly active on bac-
terial growth. The effects on the two organisms, however, were
comparable, as shown in table V.

Norit removed from beef-heart infusions practically all of
the substance which stimulates the growth of yeast when the
infusion was shaken with 2% of its weight (approximately 20
grams per liter) for three hours. The infusion was, however,
still effective for the growth of streptococci. A second shaking,
this time with 50 grams per liter, removed all activity on yeast
and bacterial growth.

The acetic acid extract of the norit from the first shaking
showed great growth stimulation on both organisms, this ex-
tract giving almost as much stimulation as the original me-
dium. The net growth stimulating activity of the extract on

14 NUTRITION FACTORS IN YEAST AND BACTERIA

yeast varied from 5.5 to 9 mm. with an average of about
7.5 mm. This activity compared favorably with that of 1%
peptone solutions. Extraction of the norit from the second
shaking failed to give any action on yeast or bacteria. Table V
gives in detail the results of these extraction experiments with
beef-heart infusions.

An interpretation of these results shows that the growth-
stimulating substances are present in smaller amounts in beef-
heart infusions than in autolyzed yeast; and that they are more
easily adsorbed from the former than from the latter. This
adsorption is more complete, at least from beef-heart infusions,
with norit than with fullers earth. Baryta fails to extract any
active substance from the norit after adsorption, while glacial
acetic acid is ineffective on fullers earth.

TABLE V.

Effect of Shaking of Beef-Heart Infusions with Fullers
tiarth and Norit.

Net Yeast Bac-

No. FULLERS EARTH growth terial

in mm. growth

1. Beef-heart infusion — 1 cc. (equiv. to 2/5

gm. beef -heart) 12.0 -f -f

2. Beef -heart infusion shaken with 50 gras. per

liter, 1 cc 3.0* + -f

2 (a) Beef -heart infusion shaken with 50 gins, per

liter, 2 cc 3.5* -}- -f

2 (b) Baryta extract of fullers earth (from 2)

1 cc 1.0 +

2 (c) Baryta extract of fullers earth (from 2)

2 cc 1.0 -f

3. Beef-heart infusion (nitrate from 2) shaken

with 100 grams per liter, 1 cc 0. —

3 (a) Beef-heart infusion (nitrate from 2) shaken

with 100 grams per liter, 2 cc 0. —

3 (b) Baryta extract of fullers earth (from 3)

1 cc 1.0 +

3 (c) Baryta extract of fullers earth (from 3)

2 cc 1.0 +

NORIT

4. Beef-heart infusion shaken with 20 grams

(2%) per liter, 1 cc 0. -f

4 (a) Acetic acid extract from norit (from 4)

1 cc 7.5* -f +

5. Beef -heart infusion (nitrate from 4) shaken

with 50 grams per liter, 1 cc 0. —

6 (a) Acetic acid extract of norit (from 5) 1 cc 0. —

• Average result of several extraction!.

LOUIS FREEDMAN 15

It is quite possible that the nature and reaction of the media
play an important role in these adsorption experiments. Inhibit-
ing substances may also be present, so that a solution may be
active after adsorption, although inactive before that procedure.
How important this hypothesis is, we cannot at present state, as
we did not fully investigate these solutions before extraction with
the adsorbents.

A summary of the results obtained shows that the substances
which promote the growth of bacteria and yeast (as extracted
from beef-heart, peptone, autolyzed yeast, etc.) belong to the
class of vitamines of the water soluble B type ,but are not identi-
cal with B vitamine; that they are comparable in activity and
show similar properties in that they are easily extracted from
their natural sources by the same adsorbents, and are again
readily recovered from these adsorbents without appreciable loss
in activity. These results point strongly to the conclusion that
they are either identical with vitamine D, or so similar to it in
their physiological behavior, that only by actual isolation, purifi-
cation, and complete chemical analysis will it be possible to differ-
entiate them.

The separation of these vitamine-like substances from the
bulk of impurities with which they are associated in nature, by
adsorbing them in fullers earth and norit and their subsequent
extraction by chemical treatment of the adsorbents, gives us
these unknown substances in a highly concentrated and com-
paratively pure form. This procedure is a distinct advantage
over working with these substances in their natural media.

SUMMARY OF CONCLUSIONS.

1. There are present in beef and beef -heart infusions, pep-
tone and autolyzed brewer's yeast, certain substances which show
a strong growth-stimulating activity on both hemolytic strepto-
cocci and yeast cells.

2. These active substances can be extracted from their na-
tural sources by shaking with certain adsorbents, such as ful-
lers earth and norit charcoal, and can be recovered by extract-
ing the adsorbents with baryta and acetic acid respectively.

3. The properties of these substances show them to be of a
vitamine-like nature ; and they are either identical with the vita-

16 NUTRITION FACTORS IN YEAST AND BACTERIA

mine D described by Funk and Dubin5 or of a similar nature.

4. There is also present in beef and beef-heart infusions
another substance which is necessary for the growth of hemo-
lytic bacteria, and this substance is thought to be associated with
hemoglobin.

BIBLIOGRAPHY TO PART I.

1. Mueller, J. Howard, Proc. Soc. Exp. Biol. Med., 1920, 18, 14.

2. Funk, C., and Dubin, H. E., Jour. Biol. Chem., 1920, 44, 487.

3. Davis, D. J., Jour. Infec. Dis., 1921, 29, 171.

4. Rivers, T. M., and Poole, A. K., Johns Hopkins Hasp. Bull., 1921, 32, 202.

5. Funk, C., and Dubin, H. E., Jour. Biol. Chem., 1921, 48, 437.

6. Seidell, A., U. S. Public Health Report, 1921, 36, 665.

7. Eddy, W. H., Heft, Hattie L., Stevenson, Helen C., and Johnson, Ruth,

Proc. Soc. Exp. Biol. Med., 1921, 18, 38, Jour. Biol. Chem., 1921, 47, 249.

NUTRITIONAL FACTORS IN THE GROWTH OF CERTAIN
YEASTS AND BACTERIA.1

II. PROTEIN HYDROLYSATES.

BY LOUIS FREEDMAN

Biochemical Laboratory of Columbia University at the

College of Physicians and Surgeons, New York City,
and the Research Laboratory of H. A. Metz, New York City.

INTRODUCTION.

The use of proteins and protein hydrolysates for enriching
culture media for the .growing of bacteria is well known; and
we find proteins to be a definite constituent of many culture
media. Bainbridge1 showed that certain bacteria required pro-
teins for growth, while Robinson and Rettger2 used a casein
hydrolysate to enrich bacterial media. However, the cases in
which it has been definitely shown that the growth of the bac-
terial organisms was due to the presence of the protein alone or
to any of its hydrolytic cleavage products, are extremely rare.
In most cases it has been shown that vitamines were present as
an added constituent or as an adhering impurity.

The possible presence in proteins of a hitherto unknown
substance which is now thought to be of biological importance
in nutrition, has of late been suggested by several investigators.
In this connection, Mueller3 thought that the growth-stimulating
substances for hemolytic streptococci, which he obtained from
certain proteins, were definite constituents of the protein mole-
cule, and he described the isolation of two such substances from
a preparation of enzyme-digested milk proteins called "ami-
noids". After his final precipitation of these substances, how-
ever, he found them, to be inactive. Goy4 also described the iso-
lation, from certain fungi, of a nitrogen-free acid which he

1 Read before the Society of Experimental Biology and Medicine. New York City,
February 15, 1922.

18 NUTRITION FACTORS IN YEAST AND BACTERIA

claimed stimulated the growth of bacteria and yeast, but he also
found the finally purified substance to be inactive.

In a previous communication5 we have shown that we can
obtain from beef and beef-heart infusions, peptone and autolyzed
brewer's yeast, by adsorption with fullers earth and norit char-
coal, certain substances which show a strong growth-stimulating
activity on yeast cells and hemolytic streptococci. We have also
shown that these substances can be extracted from their com-
bination with fullers earth and charcoal by treating these ad-
sorbents with baryta and glacial acetic acid respectively. The
properties of these substances were found to be similar to those
of the vitamine D described by Funk and Dubin6.

In the following work, the authors have tried to show the
relationship between the substances found in proteins, and vita-
mine D; and that these substances are not present in purified
proteins in general but are found associated only with certain
proteins.

EXPERIMENTAL PROCEDURE.

For this work we have prepared a series of well known pro-
teins, consisting of twelve animal and ten vegetable proteins.
The proteins of animal origin varied to a great extent, and in-
cluded the milk proteins, casein and lactalbumin, three ox-blood
proteins, three egg proteins, muscle protein, liver albumin and
globulin, and gelatin.

For the proteins of vegetable origin, we have prepared
edestin (from hemp seed) , the wheat and corn proteins, the three
proteins from the pea, oryzenin (from rice), hordein (from
barley) , and yeast protein. These proteins were prepared accord-
ing to the standard published methods, particular care being
taken to have the proteins, in so far as it was possible, free from
vitamines.

HYDROLYSIS OF THE PROTEINS.

All of the above proteins were hydrolysed by boiling them
with seven parts of concentrated hydrochloric acid (specific
gravity 1.19) for eight to ten hours. After cooling, the humin

LOUIS FREEDMAN 19

substance was filtered off and the filtrate evaporated to dry-
ness in vacuum. The residue was taken up in water and the
solution again evaporated to dryness. This operation was re-
peated twice, making four evaporations, to remove the bulk of
hydrochloric acid. The remaining acid was then neutralized by
addition of normal sodium hydroxide ; the solution was made up
to a volume corresponding to a 10% solution of the protein, ad-
justed to a PH of about 7.3 and sterilized at ten pounds pressure
for ten minutes.

EFFECT OF THE PROTEIN HYDROLYSATES ON THE
GROWTH OF STREPTOCOCCI.

The procedure for testing the effect of the hydrolysates on
bacterial growth was the same as described in our previous
paper5. One c.c. of the hydrolysate was added to the sterile
D.I.-G.S. medium and the whole, after suitable inoculation, incu-
bated at 37° C. for 24 hours. Of the animal proteins tested, only
two, casein and commercial gelatin, gave a definite positive stimu-
lation to the growth of streptococci, while fibrin and lactalbumin
gave only a slight activity. Sulphuric acid hydrolysates of casein
were also strongly active. Of the hydrolysates of the proteins of
vegetable origin, we found only two to support the growth of
streptococci, — namely, edestin and yeast protein.

TABLE I.

Effect of Protein Hydrolysates on Growth of
Streptococci and Yeast.

Net yeast Bac-

NO- Hydrolysates of Animal Proteins. 1 cc. used. growth in terial

mm. growth

1. Casein (Purified) HC1 hydrolysate 1.0 -f

2. " " H8SO« " 0 +

3. " (Technical) HC1 « 1.0 +

4. Lactalbumin 0.6 ±

5. Fibrin

6. Serum globulin

7. Serum albumin

8. Beef -heart (muscle) protein

9. Egg globulin 0

10. Egg albumin

20 NUTRITION FACTORS IN YEAST AND BACTERIA

Net yeast Bac-

No. Hydrolysates of Animal Proteins. 1 cc. used. growth in terjri

11. Vitellin (egg yolk) ....... .....................

12. Liver globulin (beef) ..........................

13. Liver albumin " ..........................

14. Gelatin (commercial— "Silver Label" brand) ..... —

15. Gelatin (prepared from bones and marrow) ...... —

16. Gelatin (prepared from bones without marrow) . . —

Hydrolysates of Vegetable Proteins. 1 cc. used.

17. Edestin (hemp seed) .......................... —

18. Gliadin (wheat) .............................. —

19. Zein (corn) ................................... —

20. Glutelin (corn) ...............................

21. Legumin (peas) ...............................

22. Vicilin (peas) .................................

23. Legumelin (peas) .............................

24. Oryzenin (rice ................................

25. Hordein (barley) .............................. 0

26. Yeast protein ........................ ......... 1.0 +

-j- denotes growth.

± denotes slight growth.

— (under Bacterial column) denotes no growth.

0 (under Yeast column) denotes that growth was same as control.

— ( " " " ) " inhibition.

Quantitative Results : We attempted to devise a method for
the determination of the amount of stimulation of the growth
of streptococci by centrifuging the bacteria, as is done in the
method for yeast described by Funk and Dubin.7 This pro-
cedure, however, was found to be impractical, due to the dif-
ficulty of stopping the growth of bacteria, and also because of
the occasional precipitation of substances which were not due
to bacterial formation.

A better method for comparative results was devised, based
on the change in hydrogen ion concentration due to the fer-
mentation of the glucose in the D.I.-G.S. medium by the strepto-
cocci. This change in hydrogen ion concentration was deter-
mined by means of the Sorenson8 indicator method, using the
Clark and Lubs9 indicator set. Cole and Jordan10 used a similar
method for diagnosis, by fermentation tests, of the gonococcus,
meningococcus and other pathogenic organisms.

The variation in hydrogen ion concentration, in all of our
tests, covered a range from a PH of 7.4 to a PH of 4.0. For this
test, we prepared standard solutions consisting of mixtures of

LOUIS FREEDMAN 21

Na2HP04 and NaH2P04 for the alkaline range, and solutions
consisting of mixtures of acetic acid and sodium acetate, ac-
cording to Walpole11, for the acid range. With these standard
solutions, we used the following indicators :

Phenol red, for PH 8.0 — 6.8
Brom thymol blue, for PH 7.6—6.0
Brom cresol purple, for PH 6.9 — 5.2
Methyl red, for PH 6.0—4.4

Three-tenths cc. of the required indicator solution was added
to 10 cc. of the standard solution. The colored solutions were
then put into uniform test tubes and the tubes sealed. The
tests were made as follows : Duplicate tubes were prepared, each
containing 9 cc. of the D.I.-G.S. medium and 1 cc. of the protein
hydrolysate. One tube was inoculated with the streptococci and
the other kept sterile. Both tubes were incubated together. After
twenty-four hours, 0.3 cc. of the indicator necessary, which was
determined roughly by the amount of bacterial growth formed,
was added to each tube and the resulting color compared with the
standard tubes.

As we used a control check for each test, we found this
method to be very convenient and fairly accurate for compara-
tive results. We have for convenience tabulated the results ob-
tained with our protein hydrolysates in their stimulating action
on the growth of streptococci, in tables Ila and b.

TABLE II (a)

Quantitative Effect of Protein Hydrolysates on the
Growth of Streptococci.

PH of standard D.I.-G.S. solution = 7.3

PH

No. 'Hydrolysates of Animal Proteins. Growth after

(1 cc. used in each test) incubation

1. Casein (purified) HC1 hydrolysate + 5.8

1 (a). (sterile control) — 7.3

2. Casein (purified) H2SO< hydrolysate + 5.3

2 (a). (sterile control) — 7.0

3. Casein (technical) HC1 hydrolysate + 5.8

3 (a). (sterile control) — 7.3

4. Lactalbumin ± 7.0

4 (a). (sterile control) — 7.3

5. Filbrin — 7.3

5 (a). " (sterile control) — 7.3

22 NUTRITION FACTORS IN YEAST AND BACTERIA

TABLE II. (a) (continued)

PH

No. Hydrolysates of Animal Proteins. Growth after

(1 cc. used in each test) incubation

6. Serum globulin — 7.3

6 (a). " (sterile control) — 7.3

7. Serum albumin — 7.3

7 (a). " (sterile control) — 7.3

8. Beef-heart (muscle) protein — 7.2

8 (a). " (sterile control).. — 7.3

9. Egg globulin ± 7.0

9 (a). " (sterile control) — 7.3

10. Egg albumin — 7.3

10 (a). " (sterile control) — 7.3

11. Vitellin (egg yolk) — 7.3

11 (a). " " (sterile control) — 7.3

12. Liver globulin — 7.2

12 (a). " (sterile control) — 7.3

13. Liver albumin — 7.3

13 (a). " (sterile control) — 7.3

14. Gelatin (commercial) + 6.0

14 (a). (sterile control) — 7.2

15. Gelatin (prepared from bone and marrow) .... — 7.2

15 (a). " " " " " " (ster-

ile control) — 7.3

16. Gelatin (prepared from bone without mar-

row) — 7.2

16 (a). Gelatin (prepared from bone without marrow)

(sterile control) — 7.3

TABLE II (b)

PH after

No. Hydrolysates of Vegetable Proteins. Growth incubation

17. Edestin + 6.5

17 (a). (sterile control) — 7.3

18. Gliadin — 7.0

18 (a). (sterile control) — 7.1

19. Zein — 7.3

19 (a). " (sterile control) — 7.3

20. Glutelin (corn) — 7.0

20 (a). (sterile control) — 7.2

21. Legumin — 7.0

21 (a). (sterile control) — 7.2

22. Vicilin _ 7.0

22 (a). " (sterile control) — 7.2

LOUIS FREEDMAN 23

TABLE II. (b) (continued)

PH after
No. Hydrolysates of Vegetable Proteins. Growth incubation

23. Legumelin — 7.3

23 (a). (sterile control) — 7.3

24. Oryzenin — 7.0

24 (a). (sterile control) — 7.2

25. Hordein ± 7.0

25 (a). (sterile control) — 7.3

26. Yeast protein -f 4.9

26 (a). " (sterile control) — 7.3

These quantitative results confirm our first results, namely,
that only two of our animal and two of our vegetable proteins
contain the growth stimulating vitamine for streptococci in
appreciable amounts. Of the proteins whose hydrolysates con-
tain the active substances, yeast protein is by far the richest,
with casein second, commercial gelatin closely following, and
edestin containing the smallest amount of the active substance.

EFFECT OF PROTEIN HYDROLYSATES ON GROWTH OF YEAST.

The protein hydrolysates as described above were tested on
yeast; and, with the exception of casein and yeast protein, all
the hydrolysates either failed to stimulate, or actually inhibi-
ted, the growth of yeast cells. Even the action of casein and
yeast protein hydrolysates were, however, practically negative,
the net increase in growth of yeast cells due to the vitamine
activity of the hydrolysates amounting to one mm. for each
of these proteins.

These results are not at all surprising, as Vansteenberge12
found that leucine, tyrosine and asparagine inhibited the
growth of yeast cells but not of the lactic bacteria. All of our
hydrolysed proteins contain large amounts of leucine, while most
of them contain varying amounts of tyrosine and aspartic acid.

A close analysis of the results obtained with these proteins
together with their method of purification and their physical
structure will, if anything, tend to confirm our view that the
growth-stimulating action is due, not to any constituent part
of the protein molecule, but to the tendency of the protein to
adsorb vitamine and hold on to it in spite of all attempts at
purification. Thus, yeast protein, the hydrolysate of which gave
considerable stimulation to the growth of streptococci, was ob-

24 NUTRITION FACTORS IN YEAST AND BACTERIA

tained from a medium which we know to be very rich in vitamine,
particularly the newly isolated vitamine D. This protein was
prepared by heating the filtered autolyzed yeast until the protein
had coagulated. As this protein separates in a form similar to
milk curd, we see the possibility of great adsorption of vitamines.
Although this protein underwent considerable washing, and was
suspended in a dialyzing bag for several days, it still retained
most of the vitamine which it had originally adsorbed.

The same is true of casein. This protein, which is precipi-
tated by means of dilute acid and is purified by repeatedly
redissolving and reprecipitating, separates at the start in the
form of a coagulum, and, as other investigators have shown,
has great adsorbing powers.* Consequently, no amount of re-
dissolving and reprecipitating will entirely free it from vita-
mines, as its physical structure and adsorbing properties remain
the same.

With gelatin, we have a somewhat analogous situation.
Gelatin, which we prepared in the laboratory, from bones with
and without the marrow, was entirely freed from vitamines by
the process of purification described by Van Name.14 This con-
sists, essentially, in precipitating the gelatin in absolute alcohol,
extracting with ether, redissolving the protein in water, concen-
trating the solution to a jelly, reprecipitating in absolute alcohol
and again extracting with ether. Commercial gelatin, ("Silver
Label" brand) however, contains enough of the active substance
to stimulate the .growth of streptoccoci, as we have shown in
tables I and II. It seems apparent therefore that the present
commercial process for purifying gelatin is not adequate for
freeing this protein entirely from the vitamines contained in
the bones and parts of the hide that are used. In this connection
it might be mentioned that Boyer15 used an hydrochloric acid
extract of finely divided bones for the cultivation of streptococci.

With edestin, we have a protein which was prepared in a
somewhat different manner. We prepared this protein by
extracting the ground hempseed with five percent sodium

The following far-reaching statement was made by Prof. F. G. Hopkins** in his
Chandler Lecture given at Columbia University, 1921. "It is remarkable what a considerable
portion of the vitamines present in milk is adsorbed by precipitated casein. A failure to
recognize this has often obscured the results of feeding trials."

LOUIS FREEDMAN 25

chloride solution at 60 °C. The solution was allowed to cool
gradually when the edestin separated as a crystalline preci-
pitate. It was purified by redissolving in warm saline solution
and again cooling the solution. Apparently, edestin also has
great adsorbing properties. Osborne, Wakeman and Ferry1 r>
were unable to free edestin entirely from vitamine.

Regarding the proteins, the hydrolysates of which showed
no growth stimulating activity, a review of their methods of
preparation and purification may throw some light on this point.
Thus the prolamines, e.g., wheat gliadin, zein and hordein from
cereals, were all extracted by means of hot alcohol, the alcoholic
solutions evaporated to small volumes and the protein precipi-
tated either in water or very dilute salt solutions. As the water
soluble vitamines are also soluble in alcohol, we can readily see
how all or nearly all of the vitamines would remain in solution.
Those proteins, such as the albumins and globulins, which were
salted out either by half saturation or complete saturation with
ammonium sulphate were all purified by dialysis, and we know
that vitamines are dialysable. Thus it is possible that these
albumins and globulins have very little adsorbing properties, and
therefore can be freed from vitamines by thorough purification.
The same apparently holds true for the vegetable globulins with
the exception of edestin.

The glutelins probably were free from vitamines even be-
fore extraction with dilute alkali, as they were obtained from
the cereal residues after the extraction of the gliadins by means
of alcohol. At any rate, the vitamines, if present would prob-
ably be destroyed by the continued action of the alkali.

The active proteins, from published analyses, show nothing
in common that cannot be shown to apply to the other, inactive,
proteins, so that we can safely rule out the question of amino
acids or nitrogen content.*

* McLeod and Wyon1T, in a recently published report on the "Supposed importance of
vitamines in promoting bacterial growth," used among other extracts, alcoholic extracts of the
kidneys of the guinea pig, and they believed that the growth was due to vitamine B, although
they stated that the amino acid content of the extracts was also a factor. They also
obtained better results with hydrolysed casein than with casein. They also found that
"marmite," which is supposedly rich in vitamines, was inactive for the pneumococcus and
meningococcus, and that charcoal, devoid of vitamines, when suspended in their bouillon
cultures, stimulated the growth of these two organisms. The part of the work of these authors
dealing with charcoal and "marmite" will no doubt need confirmation.

26 NUTRITION FACTORS IN YEAST AND BACTERIA

In summarizing the above results, three points stand out in
explanation of the growth stimulating power of the hydrolysates
of proteins.

1) The source of the protein; whether it is obtained from
a medium rich in water-soluble vitamine of the B type such as
Vitamine D.

2) The physical structure of the protein and its ability to
adsorb vitamine.

3) The method of purification.

EXPERIMENTS WITH NEUTRAL SODIUM CASEINATE.

If these active substances present in casein and three of the
other proteins, as described, are nothing more than adsorbed
vitamines, and if the protein could be brought into solution with-
out subjecting it to hydrolysis, then the vitamine would also go
into solution and could be removed by adsorbing it with fullers
earth or some other adsorbing agent.

To test this theory, we prepared a solution of neutral
sodium caseinate by suspending our purified casein in water, and
adding five percent sodium hydroxide solution until the casein
had practically all dissolved. A small amount of insoluble de-
natured casein, formed by the action of the alkali, was filtered
off, and the solution was then made neutral to litmus. This
neutral sodium caseinate solution (containing about 5% casein)
was shaken for three hours with ten grams (50 grams per liter)
of fullers earth.

After shaking, the fullers earth was separated from the solu-
tion by centrifugation, as filtration was very slow, due to the
formation of an emulsion, and the supernatant liquid was then
filtered. The fullers earth was repeatedly washed with water
and centrifuged to remove small amounts of sodium caseinate
adhering to it, pressed out on a suction filter, washed with
alcohol and dried. This fullers earth was then extracted with
baryta and worked up as described in paper I.5 The resulting
solution was made up to a volume corresponding to a ten percent
solution of the original casein, adjusted to a PH of 7.4 and ster-
ilized. This solution, when tested, was found to have a strong
stimulating effect on the growth of streptococci, but had no
action on the growth of yeast cells.

The neutral sodium caseinate solution, after shaking with

LOUIS FREEDMAN 27

fullers earth, was subjected to hydrolysis by adding concen-
trated sulphuric acid until a 30% solution was obtained and
boiling for 36 hours. This hydrolysate failed to give any growth
stimulating activity on either streptococci or yeast.

It is therefore apparent that most or all of the vitamine was
removed from the casein as described. This shows that the
active substance in the protein is not necessarily a constituent
part of the protein molecule, and does not have to be isolated by
breaking up the protein.*

SUMMARY OF CONCLUSIONS.

1. We can obtain from purified casein, commercial gelatin,
yeast protein and edestin, by hydrolysis of these proteins, cer-
tain substances which show marked growth stimulating activity
on hemolytic streptococci. Hydrolysates of purified egg globulin,
lactalbumin and hordein, show traces, whereas hydrolysates of
the other proteins examined are devoid of these activating sub-
stances.

2. We have found that these active substances are not a
constituent part of the protein molecule, and that the amount
of the substances present in the protein depends on the physical
and adsorptive properties of the protein and the method and de-
gree of purification.

3. The properties of these active substances, as obtained
from the proteins described above, show that they are probably
related to if not identical with the water-soluble vitamines ob-
tained from brewers' yeast, particularly vitamine D.

4. We have described a method for determining the amount
of these active substances by the use of Sorensen's indicator
method, which is based on the change in hydrogen ion concentra-
tion due to the fermentative action of the bacteria. This method
is advocated for comparative results only, and not as a direct
quantitative method.

* The experiments with neutral sodium caseinate, just described, may be of practical
importance in that they may lead to a method of preparing a vitamine-free casein. The
casein thus obtained gives positive tests for all the color reactions known for the various
»mino acids, while its nitrogen content is the same as that of the original casein from which
it was prepared. Further work is now in progress on this problem.

\

28 NUTRITION FACTORS IN YEAST AND BACTERIA

BIBLIOGRAPHY TO PART II.

1. Bainbridge, F. A., J. Hyff., 1911, 11, 341.

2. Robinson, H. C., and Rettger, L. F., J. Bad., 1918, 3, 209.

3. Mueller, J. Howard, Proc. Soc. Exp. Biol. Med., 1921, 18, 225.

4. Goy, Pierre C., rend. Acad. Sci., 1921, 172, 242.

5. Freedman, L., and Funk, C., J. Metabolic Research, 1922, 1, 457.

6. Funk, C., and Dubin, H. E., J. Biol. Chem., 1921, 48, 437.

7. Funk, C., and Dubin, H. E., J. Biol. Chem., 1920, 44, 487.

8. Sorensen, Bioc/iem. Ztsc/i., 1909, 21, 253.

9. Clark and Lubs, J. Bad., 1917, 2, 109 and 191.

10. Cole and Jordan, J. Path, and Bact., June, 1917.

11. Walpole, G. S., Biochem. J., 1914, 8, 636.

12. Vansteenberge, P., Ann. Inst. Past., 1917, 31, 601.

13. Hopkins, F. G., J. Ind. Eng. Chem., 1922, 14, 67.

14. Van Name, W. G., J. Exp. Med., 1897, 2, 117.

15. Boyer, Louis, C. r. soc. de biol., 1918, 81, 229.

16. Osborne, T. B., Wakeman, A. J., and Ferry, E. A., J, Biol. Chtm., 1919,

39, 35.

17. McLeod, J. W., and Wyon, G. A., J. Path, and Bact., 1921, 24, 205.

LOUIS FREEDMAN 29

ACKNOWLEDGMENT.

I wish to express my sincere appreciation to Professors
W. J. Gies and E. G. Miller, Jr., of the Department of Biological
Chemistry, for their valuable advice and criticism in the prep-
aration of this dissertation. I also wish to express my gratitude
to Dr. Casimir Funk, of the Department of Biological Chemistry,
who suggested the problem and under whose painstaking guid-
ance this work was done.

I also take this opportunity to express my acknowledgment
and thanks to the Hon. Herman A. Metz, who extended many
facilities to me in this work and gave me the opportunity of com-
pleting my dissertation while engaged in research in his
laboratory.

30 NUTRITION FACTORS IN YEAST AND BACTERIA

VITA.

Louis Freedman was born in New Haven, Conn., January
28, 1894. He graduated from the New Haven High School in
1912 and entered the Sheffield Scientific School of \ale University
the same year. During freshman year, he was the recipient of a
Connecticut High School scholarship. He received the degree of
Ph. B. in 1915, after which he spent two years in the Yale Gradu-
ate School, from which he received the degree of Master of
Science. During his graduate course, he was the holder of both
a graduate and an industrial fellowship. From 1917 to 1919 he
was employed by the Calco Chemical Company as research chem-
ist in Pharmaceuticals. During the war, he was engaged in
special research work on high explosives for the United States
Government. He matriculated in Columbia University in Sep-
tember, 1920, and is at present associated with the Research
Laboratory of H. A. Metz, New York City.

502082 f

.'iOLOGY

LIBRARY

0

UNIVERSITY OF CALIFORNIA LIBRARY