JOURNAL OF BACTERIOLOGY, Dec. 1974, p. 1078-1084 Vol. 120, No.

3
Copyright ( 1974 American Society for Microbiology Printed in U.S.A.

Genetic Basis of Nutritional Requirements in

Lactobacillus casei
T. MORISHITA, T. FUKADA, M. SHIROTA, AND T. YUIRA
Yakult Institute for Microbiological Research, Kyoto, and Institute for Virus Research, Kyoto University,
Kyoto, Japan
Received for publication 24 September 1974

In a study of the genetic basis of multiple nutritional requirements in

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest

Lactobacillus casei, systematic attempts were made to isolate mutants that can
grow in the absence of' a specif'ic nutrient required by the parental organism.
Such mutants have successfully been isolated with respect to seven of twelve
amino acids (aspartic acid, leucine, isoleucine, lysine, methionine, serine, and
threonine) and three of four vitamins (pantothenic acid, nicotinic acid, and
pyridoxal) tested, after extensive screenings employing various mutagens.
Mutants that can grow without tryptophan were not isolated, but those that can
grow on anthranilate or indole as well as on tryptophan were obtained at a
frequency expected for single-step mutations. Activity of' tryptophan synthetase
was demonstrated in extracts of these anthranilate-utilizing mutants, but not in
the parental strain. These results suggest that the multiple nutritional require-
ments of L. casei are often, if not always, due to one or a few small lesions such as
base substitution mutations rather than large deletions affecting the genes
involved in each biosynthetic pathway. The data would also imply that many of
the biosynthetic pathways that are not fully functional in L. casei were active at
one time and became nonfunctional during evolution of the present species.

One of the striking characteristics of the mutations were found to result in the appear-
genus Lactobacillus is the well-known multiple ance of' a specif'ic enzyme activity responsible
nutritional requirements that these bacteria for the biosynthesis of a specific amino acid.
require for their normal growth. They require a These results indicate that most, if' not all, of
number of amino acids, vitamins, purines, and the genes essential for biosynthesis of these
pyrimidines when grown in synthetic media. It nutrients are present in L. casei, but they do not
has generally been thought that this character- produce active proteins due to one or a few
istic may be the result of' adaptation of these deleterious mutations that occurred in the genes
organisms to the natural environments to which for each biosynthetic pathway.
they have been exposed. Thus, genetic and
biochemical analysis of the mechanisms under- MATERIALS AND METHODS
lying the growth requirements in lactobacilli
should contribute not only to the general prob- Bacterial strains. The wild-type L. casei used as
the parental organism in this study was strain S1 from
lem of nutritional requirements among microor- The Yakult Institute for Microbiological Research,
ganisms found in nature but to ecological or Kyoto, Japan.
evolutionary problems as well. Media. Natural medium consisted of (per liter): 10
Although a number of mechanisms can be g of polypeptone (Wako Drug Co.), 10 g of yeast
contemplated to explain growth requirements extract (Difco), 10 g of sodium acetate, and 20 g of
for a given nutrient, a specific question was D-glucose. Basal (synthetic) medium was a glucose-
asked as to whether these requirements are salts medium supplemented with all the nutrients
brought about by the lack of gene(s) responsible required by the wild-type L. casei, strain S1, for its
for its biosynthesis. Thus we made a systematic maximal growth (Table 1). Solid media contained
attempt to isolate mutants of Lactobacillus 1.5% agar (Hakko Agar Co.). The pH of all media was
casei that had lost the growth requirement for a adjusted to 7.2.
Isolation of mutants. Mutants that had lost a
specific nutrient. As will be shown below, such requirement for a given nutrient were isolated after
mutants could indeed be isolated for a majority mutagenic treatment of cells by N-methyl-N'-
of nutrients required by the parental organism. nitro-N-nitrosoguanidine (NG), 2-aminopurine (AP),
Moreover, at least in one case studied, certain or ultraviolet light (UV).
1078
VOL. 120, 1974 NUTRITIONAL REQUIREMENTS IN L. CASEI 1079
TABLE 1. Composition of basal mediuma
Compound Concn (mg/ml) Compound Concn (mg/ml)
D-Glucose .......................... 10 L-Lysine ........................... 0.1
Sodium acetate ............. 30 L-Methionine 0.1
Ammonium chloride ............. 3 L-Phenylalanine 0.1
KH2PO1 ........... .. 2 L-Serine 0.1
K2HPO4.. . . . 2 L-Threonine 1.......
0.1
Tween 80 ................................ 1 L-Tryptophan 0.1
Sodium chloride .......... 0.02 L-Tyrosine 0.1
MgSO4.7H20 .......... 0.2 L-Valine ... 0.1
MnSO4.7H20 .......... 0.04 Riboflavine ... 0.001
FeSO4 7H20 .......... 0.02 Pantothenic acid ... 0.001

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest

L-Arginine ............................... 0.1 Pyridoxal . 0.001
L-Aspartic acid ... ....... 0.4 Nicotinic acid ..... ....... 0.001
L-Cystine .......... 0.1 Folic acid ............ 0.0001
L-Glutamic acid ... ....... 0.4 Adenine ... ......... 0.02
L-Isoleucine .......... 0.1 Uracil ............ 0.02
L-Leucine .......... 0.1 Xanthine ... ......... 0.02

a
This medium is based on the synthetic medium routinely used at the Yakult Institute (H. Endo, personal
communication) and contains all the nutrients required by the wild-type L. casei, strain Si, for its maximal
growth. It includes some compounds that are not absolutely required but do exhibit strong stimulatory effects
(see Table 2).
(i) Mutagenesis with NG. Cells of the wild-type tris(hydroxymethyl)aminomethane (Tris)-hydrochlo-
strain grown in natural medium overnight at 37 C ride buffer (pH 7.8). For preparation of crude ex-
were collected, washed in saline, and treated with NG tracts, 5 mM 2-mercaptoethanol and 10% glycerol
(700 ,g/ml) in 0.2 M acetate buffer (pH 5.0) at 37 C were added to the cell suspension, which was then
for 3 h. Cells were collected, washed twice in saline, disrupted in a Raytheon 10-kc sonic oscillator and
and plated onto appropriate selective agar media. In centrifuged at 10,000 x g for 20 min. The resulting
some experiments, NG mutagenesis was performed by supernatant fluids were used as crude extracts.
placing crystals of NG at the center of a petri dish Tryptophan synthetase activity was assayed by the
containing selective agar media seeded with a lawn of standard procedure used for the enzyme of Esche-
washed wild-type cells. richia coli (7). The reaction mixture (0.5 ml) con-
(ii) Mutagenesis with AP. Cells were grown in tained the following (micromoles): Tris-hydrochloride
natural medium containing AP (50 ,g/ml) overnight (pH 7.8), 35; indole, 0.2; D,L-serine, 40; pyridoxal
at 37 C. The cells were harvested, washed twice in phosphate, 0.04; 0.015 ml of saturated NaCl solution;
saline, and plated on selective agar medium directly, and 0.25 ml of cell suspension or crude extract. After
or, in some experiments, after incubation in liquid 30 min of incubation at 37 C, the reaction was stopped
selective medium for 2 days. by adding 0.1 ml of 1 N NaOH to each tube.
(iii) UV irradiation. UV irradiation was carried Disappearance of indole during incubation was esti-
out by exposing a saline suspension of cells to a mated by the colorimetric precedure (7).
Toshiba germicidal lamp (15 watt) for 1 min at a Protein was determined by the method of Lowry et
distance of 60 cm. Irradiated cells were incubated at al. (3), and optical density was measured with a
37 C overnight in natural medium collected, washed, Klett-Summerson colorimeter with a no. 54 filter.
and plated on selective agar media.
All agar plates were incubated at 37 C, and colonies RESULTS
that appeared after 4 to 5 days were purified by single Mutants that lost a specific nutritional
colony isolation on appropriate selective agar plates. requirement. We first determined the nutri-
Presumptive mutants thus obtained were then tested
for their ability to grow in liquid selective media tional requirements of the parental L. casei in
lacking a given nutrient. All liquid cultures were liquid culture (Table 2). Several compounds not
incubated without aeration, and optical densities present in the basal medium were also included
were determined with a Klett-Summerson colorime- to observe their possible stimulatory effects on
ter using a no. 54 filter. growth. It can be seen that this strain required
Assay for tryptophan synthetase. Either cell 12 amino acids and 4 vitamins for normal
suspension or crude extract was used for assay of growth. In addition, several other nutrients,
tryptophan synthetase (EC 4.2.1.20). Cultures were including amino acid, vitamin, purine, and
grown with limiting amounts of L-tryptophan (1 to 2 pyrimidine, were found to stimulate growth of
4g/ml) or indole (2 ,g/ml) in the otherwise standard the organism appreciably under the conditions
basal medium. After standing at 37 C for 3 days, cells
were harvested, washed, and suspended in 0.05 M employed. These results generally confirmed
1080 MORISHITA ET AL. J . B ACTERIOL .
TABLE 2. Nutritional requirements of the wild-type L. caseia
Nutrient omitted Growth Nutrient omitted Growth
Amino acids Vitamins
Alanine + Thiamine +
Arginine - Riboflavine
Aspartic acid - Pyridoxal
Cystine ± Biotin +
Glutamic acid - Pantothenic acid
Glycine + Nicotinic acid
Histidine + Folic acid +
Isoleucine - p-Aminobenzoic acid +
Leucine
Lysine - Purines and pyrimidines

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest

Methionine - Adenine +
Phenylalanine i Cytosine +
Proline + Guanine +
Serine - Thymine +
Threonine - Uracil +
Tryptophan - Xanthine ±
Tyrosine
Valine -

a Cells of the wild-type strain grown in natural medium overnight at 37 C were collected, washed twice in
saline, and inoculated into a series of liquid synthetic media containing all but one of the nutrients listed. These
cultures were incubated at 37 C for 3 days, during which optical density was measured at intervals of 24 h. At
least duplicate tubes were employed for each test. Symbols: +, normal growth; -, partial growth; -, no growth.

those obtained previously by H. Endo (personal TABLE 3. Frequencies of mutations leading to the loss
communication). of a specific growth requirement
Systematic attempts were then made to iso-
late mutants that no longer required a specific Nutrient Spontaneous Induced mutation0
nutrient for growth. Thus, cells of the wild-type omitteda mutation Mutagen Frequencv
L. casei were treated with various mutagens and
plated on appropriate media to select for possi- Aspartic acid <6 x 10-10 NG 10-5
ble mutants. Such mutants were successfully Leucine <5 x 10- 10 NG 10-6
isolated with respect to seven of twelve amino Isoleucine <5 x 10 '10 AP 10-8
acids, and three of four vitamins required by the UV 10-8
Lysine <6 x 10-10 NG 10-7
parental strain (Table 3). These mutants were Methionine <6 x 10- 10 AP 10-7
obtained only after treatment by mutagens; no Serine <6 x 10 10 NG 10-5
mutants have so far been obtained spontane- Tyrosine <5 x 10-10 NG 10-7
ously. The frequencies of mutants shown in Pantothenic acid <7 x 10-10 NG 10-7
Table 3, however, should be considered only as Nicotinic acid <7 x 10- 10 AP 10-7
approximations, in view of the procedures em- Pyridoxal <7 x 10-10 NG 10-8
ployed for selecting these mutants.
A number of mutants obtained from each aEach of these nutrients was omitted from the
selection were purified by single colony isola- basal medium used for selection of the mutants.
hMutagenesis was carried out as described in
tion, and their nutritional properties were con- Materials and Methods. Frequency of mutants shown
firmed by streaking tests on agar media with gives an approximate order of magnitude estimated
appropriate supplements. The rate of growth of from several experiments.
these mutants on selective agar varied widely
among different mutants, and some of the
faster-growing mutants were further examined changed, however, indicating that the effect of
in liquid media. Typical-growth curves obtained such mutations is strictly limited to a specific
for some of the mutants in the corresponding biosynthetic pathway. It was also found that the
selective medium are shown in Fig. 1. It can be maximal growth of these mutants attained in
seen that each of the mutants tested grew in the natural medium was generally somewhat re-
medium lacking one of the amino acids required duced as compared to that of the wild-type
by the parental organism. Other nutritional strain.
characters of these mutants remained un- Mutants similar to those described here have
VOL. 120, 1974 NUTRITIONAL REQUIREMENTS IN L. CASEI 1081
that can grow on certain precursors of a nutrient
that cannot be utilized by the parental orga-
nism. The tryptophan biosynthetic pathway
was chosen to investigate this possibility.
It was first shown that the tryptophan re-
quirement of the wild-type L. casei was not
(I)
satisfied by anthranilate or indole, possible
z/
intermediates on the biosynthetic pathway. Ex-
periments were then carried out to find possible
10 1 mutants that can grow on indole or anthranilate
as well as on tryptophan in otherwise standard
uLJ basal medium. Wild-type cells grown in the
C D presence of 2-aminopurine were harvested and

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest

washed, and samples were plated on basal agar
>-100- medium containing anthranilate or indole in-
stead of tryptophan. After incubation at 37 C
C) for 5 days, mutant colonies appeared on the
m agar plates at a frequency of about 10-6. These
mutants grew on anthranilate or indole, as well
as on tryptophan, in contrast to the parental
/- strain, whose requirement for tryptophan was
not satisfied by these compounds (Table 4).
This was true whether the selection was made
0 48 72 0 24 48 72
on the medium containing indole or anthrani-
TIME AT 370 ( HR) late.
FIG. 1. Typical growth curves of the mutants that
These results suggest that anthranilate and
have lost a specific nutritional requirement. (A) indole are probably involved in tryptophan
Aspartate-independent mutant, (B) serine-independ- biosynthesis in L. casei as in E. coli, and that
ent mutant, (C) lysine-independent mutant, and (D) the wild-type strain carries genetic blocks at the
leucine-independent mutant. Washed cell suspension stages of both anthranilate synthesis and the
of each mutant was used to inoculate the basal
medium (A) and a selective medium lacking a specific TABLE 4. Growth of mutants that can utilize
amino acid (A) at time zero. Cultures were incubated precursors of tryptophan
at 37 C without aeration, and optical density was mea-
sured at the times indicated. The parental strain was Growth (optical density)a
grown in the basal medium (0) as well as in the
selective medium (0) as a control for each mutant Strain B sal -Tryp- -Tryp-
tested. Bsl - Tryp- tophan tophan
medium tophan ±Anthra- + Indole
nilate
also been isolated from another species, L.
acidophilus, at least with respect to several Wild type 207 7 9 11
amino acids, including aspartic acid, tyrosine, Mutant 2b 162 5 151 143
and isoleucine. This suggests the general occur- Mutant 4b 167 2 149 150
rence of such mutants among various lactoba- Mutant 8b 174 2 156 160
cilli. Mutant 9b 158 0 159 147
Mutant 10b 189 3 178 175
Mutants capable of utilizing precursors of Mutant 11 155 0 141 132
tryptophan. The above results strongly suggest Mutant 12C 181 1 143 155
that all the genes determining the structure of Mutant 15c 176 2 159 162
enzymes involved in most biosynthetic path- Mutant 17c 163 6 156 147
ways are present in the wild-type L. casei and Mutant 18C 158 0 155 144
that one or a few mutations can lead to the
normal functioning of each pathway. With re-
a
Washed cell suspensions were used to inoculate
spect to the nutrients for which no such mu- the media as indicated and were incubated at 37 C for
48 h. Values represent optical density as expressed in
tants could be isolated by one-step selection, Klett units.
however, it is not clear whether the defect is due "
Mutants selected for their ability to grow on
to an "irreparable" damage such as large dele- indole.
tion or to a number of reparable damages. In the c Mutants selected for their ability to grow on
latter case, we might expect to find mutants anthranilate.
1082 MORISHITA ET AL. J. BACTERIOL.

conversion of indole (or indole glycerol phos- therefore, that the lack of enzyme activity in the
phate) to tryptophan (Fig. 2). wild-type extract is due to the presence of an
Tryptophan synthetase activity in an- enzyme inhibitor. Nor does it seem likely that
thranilate-utilizing mutants. To find out the appearance of enzyme activity in the mu-
whether the utilization of anthranilate or indole tant is due to the production of an enzyme
by these mutants was due to an increased activator. It may tentatively be concluded that
activity of tryptophan synthetase or to other these mutations affected one of the genes deter-
factors such as increased permeability to these mining the structure or amount of tryptophan
compounds, tryptophan synthetase activity was synthetase.
assayed in the mutant and in the wild-type Attempts to isolate tryptophan-independent
strains. Cultures were grown in basal medium mutants starting from the anthranilate-utiliz-
with limiting concentrations of L-tryptophan or ing mutants obtained above have so far failed,
indole, and washed cell suspensions were used despite many efforts that were made, employing

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest

for enzyme assay. Tryptophan synthetase activ- several different mutagens.
ities in terms of serine-dependent disappear-
ance of indole were demonstrated in all the DISCUSSION
mutants tested, whereas little or no activity was Mutants of lactobacilli that have lost the
detected in the wild-type strain (Table 5). growth requirement for a specific nutrient have
These results were also confirmed by enzyme been reported previously in several different
assays performed with crude extracts (Table 6). species. For example, a thiamine-independent
The enzyme in the mutant extract required mutant of L. fermenti (1), tryptophan-
serine for maximal activity. Furthermore, mix-
ing of equal amounts of wild-type and mutant
extracts resulted in activity about equal to that TABLE 6. Tryptophan synthetase activity in crude
extracts of an anthranilate-utilizing mutant
of the mutant extract alone. It seems unlikely,
Disappear-
Extracts Reaction ance of Specific
A. WILD TYPE system indole activitya
(mg of protein)
ndo e
($mol)
DI] C* Anthraonilic c
glycerol I Tryptophan
acid phosphate Wild (0.58) Complete <0.01 <0.17
tindole, Wild (0.29) Complete <0.01 <0.37
|TRYPTOPHAN Mutant 2 (0.61) Complete 0.097 1.59
B. ANTHRANILATE-UTILIZING MUTANT SYNTHETASE Mutant 2 (0.31) Complete 0.063 2.03
Mutant 2 (0.61) -Serine 0.015 0.25
ndole Wild (0.29) + Complete 0.053 0.88
o Anthranilic
= glycerol > Tryptophan
Mutant 2 (0.31)
acid phosphate
(,dae) a Expressed as units per milligram of protein. One
FIG. 2. Location of genetic blocks on the postu- unit is defined as that amount of enzyme which
lated pathway of tryptophan biosynthesis in L. casei, causes disappearance of 0.1 Mmol of indole in 30 min
strain Sl. under standard conditions.

TABLE 5. Tryptophan synthetase activity in cells of anthranilate-utilizing mutantsa

Addition to Cell suspensions Disappearance of av

Expt strain the medium used" indole (Amol) Specific activityc

(,Mg/mi)
I Wild type Tryptophan 1 1.00 <0.01 <0.01
Wild type Tryptophan 2 1.00 <0.01 <0.01
Mutant 2 Tryptophan 1 0.72 0.06 0.083
Mutant 2 Tryptophan 2 1.69 0.08 0.047
II Wild type Tryptophan 2 1.00 <0.01 <0.01
Mutant 8 Indole 2 1.04 0.077 0.074
Mutant 9 Indole 2 0.75 0.047 0.063
a Cells were grown in basal medium containing limiting concentrations of L-tryptophan or indole, as
indicated, to stationary phase, and washed cell suspensions were used for assay of tryptophan synthetase
activity.
b Expressed as optical density, which was measured and represented in Klett units divided by 100.
c Expressed as micromoles/optical density.
V OL . 120, 1974 NUTRITIONAL REQUIREMENTS IN L. CASEI 1083
independent mutants of L. arabinosus and L. that of the wild-type strain when cultured in
casei (6), and a mutant capable of growing in natural medium. Furthermore, we have occa-
the absence of nucleotides and folate in L. sionally observed that these mutants can revert
fermenti (2, 5) have been described. Mutants to the original phenotype with respect to the
with altered nutritional requirements were also nutritional requirement after successive trans-
found among those selected for resistance to fers in natural medium. It seems as though the
certain antibiotics in various species of lactoba- wild-type strain of L. casei represents the most
cilli (4). However, in all these instances, the well-adapted form in the natural environments,
kinds of mutants obtained are limited, and and any deviations from this genetic constitu-
mutants altered with respect to only a few tion by single-step mutations are likely to be
nutrients were obtained from any specific pa- deleterious to the whole organism.
rental organism employed. This notion seems to be further supported by
In our study, extensive search for such mu- the following experiments. Starting with one of

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest

tants using several different mutagens led to the the isoleucine-independent mutants listed in
successful isolation of mutants for the majority Table 3, tyrosine-independent mutants were
(10 of 16) of nutrients required by the parental selected after nitrosoguanidine mutagenesis. As
strain of L. casei. These mutants most probably expected, the double mutants obtained required
represent single-step or at most two-step muta- neither isoleucine nor tyrosine for growth. When
tions. In the case of tryptophan, tryptophan- grown in natural medium, they appeared to
independent mutants could not be isolated by grow even more slowly than their immediate
single-step mutations, but anthranilate-utiliz- parental strain. By repeating the same selection
ing mutants were obtained as a result of muta- procedure, we finally obtained quintuple mu-
tion affecting the activity of tryptophan synthe- tants that had lost the growth requirements for
tase (Fig. 2). No attempts were made, however, serine, aspartic acid, and leucine, in addition to
to see if an inactive tryptophan synthetase is isoleucine and tyrosine. The comparison of
produced by the parental strain. Therefore, the growth rates among these mutants in natural
question of whether the present mutations af- medium revealed the inverse relationship be-
fected one of the structural genes for the enzyme tween growth rate and the number of muta-
rather than a regulatory gene that controls the tional steps involved to obtain the mutant, with
amount of enzyme synthesized remains open for the quintuple mutants being the slowest (T.
future investigation. Morishita, unpublished data). It remains possi-
In view of the results obtained for the trypto- ble, however, that some unrelated deleterious
phan pathway, it is possible that similar situa- mutations accumulated during the repeated
tions may be found with other biosynthetic treatments with nitrosoguanidine.
pathways for which single-step mutants able to As to the mechanisms of nutritional require-
grow without the respective end product have ments in wild-type L. casei and other related
not so far been isolated, i.e., arginine, gluta- organisms, the present status of our knowledge
mate, threonine, valine, and riboflavine. In any is too limited to permit meaningful discussion
event, it may be concluded that growth require- at this time. It should be stated, however, that a
ments for specific nutrients in lactobacilli are genetic as well as a biochemical approach would
usually, if not always, due to one or a few prove most fruitful in solving this long-lasting
minor defects in the bacterial genome which problem. The importance of discovering genetic
can easily be eliminated or phenotypically sup- mechanisms such as those found in Escherichia
pressed by a single or a small number of coli cannot be overemphasized for future devel-
mutations. The defects presumably involve opment in this field.
point mutations rather than large deletions.
A wild-type strain of L. casei whose trypto- LITERATURE CITED
phan requirement can be satisfied by anthrani- 1. Cheldelin, V. H., M. J. Bennett, and H. A. Kornberg. 1946.
late or indole has been described by Snell (8). Modifications in the Lactobacillus fermenti 36 assay
Although the precise nature of difference(s) for thiamine. J. Biol. Chem. 166:779-780.
between the two strains of L. casei remains 2. Cheldelin, V. H., and A. P. Nygaard. 1951. Adaptations in
growth factor requirements for lactic acid bacteria. J.
obscure at present, strain Si used in our labora- Bacteriol. 61:489-495.
tory appears to carry an additional defect in the 3. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.
conversion of indole to tryptophan, and the Randall. 1951. Protein measurement with the Folin
latter defect can be overcome by single muta- phenol reagent. J. Biol. Chem. 193:265-275.
4. Miyazawa, S. 1962. Studies on the nutritional require-
tional events. ments of lactic acid bacteria in the presence of antibiot-
The maximal growth of most of the mutant ics. Part XI. The vitamin and amino acid requirements
strains obtained was significantly lower than of antibiotics resistant strains in lactic acid bacteria.
1084 MORISHITA ET AL. J. BACTERIOL.
Nippon Nogei Kagaku Kaishi 36:514-520. 7. Smith, 0. H., and C. Yanofsky. 1962. Enzymes involved in
5. Nygaard, A. P., and V. H. Cheldelin. 1951. Nutrition the biosynthesis of tryptophan, p. 794-806. In S. P.
studies of two variants of Lactobacillus gayoni. J. Colowick and N. 0. Kaplan (ed.), Methods in enzymol-
Bacteriol. 61:497-505. ogy, vol. 5. Academic Press Inc., New York.
6. Sbarra, A. J., and M. M. Hardin. 1951. Attempts to 8. Snell, E. E. 1943. Growth promotion of tryptophan-defi-
develop strains of Lactobacillus arabinosus 17-5 and cient media by o-aminobenzoic acid and its attempted
Lactobacillus casei ATCC 7649 independent of certain reversal with orthanilamide. Arch. Biochem.
growth factors. J. Bacteriol. 61:99-100. 2:389-394.

Downloaded from http://jb.asm.org/ on April 5, 2019 by guest