ANNUAL REPORT 1947-48

Preject Ho. 742. The nature and action of the gene in bacteria. Lederberg
(Mrs, Esther Lederberg, Bational Cancer Institute, Junior
Research Fellew, Nerton D. Zinder, Research Assistant,
Donald A. Gordon, Research Assistent), State funde,
W.A.R.F., Rockefeller Foundation, National Institute
of Health.

The research undertaken in this field ean de reviewed under the

following headings:

1, Genetic Aspects of the life cycle of Escherighia soll.
2. Genetic control of fermentation enzymes in E. goli.
3. Genetic control of gene mutability. (Mrs. Lederberg)

4, Gene recombination in Salmonella, (Mr. Zinder)
mPa

1. Genetic aspects of the life cycle of Escherichia coll.

The discovery by Tatum and Lederberg (1) of a sexual cycle in strains of
the common bactertum Hecherichia gold opens up the field ef bacteria to genetic
analysis in a way that wae not formerly possible. In this earlier work, con-
dusted at Yale University in 1946-47, genetic recombination was demonstrated
by the use of nutritional mutants, obtained with i-ray or ultraviolet light.
These mutants have mtritionsal deficiencies which prevent them from growing
on synthetic mediua unless their particular growth factors are supplied. It
wae found that when different mutants were mixed together and inoculated into
synthetic ager thas a very small proportion of colonies did grow, and that
these consisted of cells which could grow on the synthetio medium (i.e.
prototrophe). By using mutante which differed in several other characters,
it could be shown that these characters vere redistributed to the prototrophs
in all possible combinations, suggesting that the proper explanation of the
occurrence of prototrophs is gene recombination. If one nutritional mutant
is represented as A-B+ and the other as A+5-, one would get by recombination
A~B® x AtB- --> AtB+ and A-B-. In this case, only the AtBt+ would be recovered,
ae it wlone could grow on the synthetic medium which is used to sieve it out
from the two parents, whose celle preponderate, However, other markers can
segrégate freely into the prototrophs, From a study of the relative proportions
of different recombination types, it was concluded that the genes of Escherichia
Soli were linked in linear order, as in higher forms, and there was probably
but a single linkage group.

From thie earlier work, it was concluded that &@) cell fusion occurs only
very rerely, about one zygote appearing per million cells; and b) the sygote
does not proliferate, mt undergoes reduction division immediately. This is
Based on the finding that individual prototrophs were uniform, each colony

oonsisting usually of bet a single recombination type which would be different

from colony to colony. If the diploid sygote were to mltiply, one would
-3j-

expect to find several recombination types in a single prototroph colony, each
derived from the segregation of a different zygote cell. These facts vere dis-
coureging to any hopes of seeing the sygote cytologically, or of determining
such questions as dominance for which the diploid fe necessary. We could put
in a certain mixture of types, and get them out again in various recombinations,
but the intermediate process wae not directly accessible to analysis and had to
be inferred. Yor the further development of genetical work on bacteria, it
would be very important, if not essential, to clarify this intermediate

process.

In the course of other work, and quite accidentally, an exception was
noted te conclusion bd); that is to say, a heterozsygotio culture was found which
could be maintained in the diploid condition, although it has a marked propensity
to undergo reduction. ‘The exceptional culture was a prototroph obtained in a
cross between tvo parents which differed in nutritional factors (3, M, T, L, By.)
and in lactose fermentation (Lac-/Lact). The indicator medium, EBosin-Methylene
Blue agar is used to test fermentative characteriatics (also see 5). When this
culture originally thought to be Lact was streaked out and inoubuted for two or
three days on the EME medium, the colonies were observed to be mosaics of Lac-
and Lac* cells. Furthermore, when these segregants were tested, they proved to
show all possible recombinations of the nutritional and other factors, including
certain recombinations, corresponding to the A-B- above, which could not other-
wise be secured. Evidently, this culture was prototrovhic not because it was a
recombination of the type A+B+, but because it was a diploid heterozygote Gt .
Porther study of thie culture and of ite segregants suggests that its exceptional
behavior may de the result of a mtation which occurred spontaneously in one of
the parents, for outcrossing segregants to standard stocks results in prototrophs
a substantial proportion of which are diploid heterozygotes.

Although it was formerly believed that there were no exceptionsto rule
wlpe

%) in crosses of standard stocks, current studies using better selective means
now indicate that pereistent heterozygotes may occur here too, although at a
lower rate than among cutcrosses of the above-mentioned exception. In this
test, parents are used which carry different, closely linked, recessive Lac-
in the "repulsion phase, 1.e. A~ Lac)-Lag,+B+ x A+ Lac,+Laq,- B-. When
such a pair is croseed, most of the A+B+ recombinant prototrophe are either
Laq~Laq,* or Lac, tLacy—,» Decause of the very close linkage of the two Lac
lool. These are all lactose-negative: much less than 1% of the prototrophs
ebtained here are lactose-positive, and these can be detected visually by
conducting the cross on a synthetic EMB medium. Even when “normal” stocks
are used, about half the lactose~posttive prototrophs are not A+B+Lac)tlac,+,

tut are heterozygous diploids 4-Laci-Lac)t Bt as shown by their subsequent

segregation. These diploids may differ from those previously oBtained in
being more stable, but this ie not definitely eatablished,

Detailed studies of the segregations from tke exceptional heterozygote
showed deviations from random distribution that oan be best accounted for by
aseuming that one or both chromosones of thie diplold carry several deficient
regions, These deficiencies, which would make inviable a segregaat carrying
them alone, may be the inciting factor for the persistence of this heterozygote,
and detract somewhat from the usefulness of this type in genetic analysis.
Whether the same situation prevails in the "normal" heterosygotes rensias
te be determined. |

Studies are underway to determine the dominance relationships of a nusber
of genes. The normal alleles of the nutritional mtations all seem to be
completely dominant, and the functioning, "*", alleles of the several fer-
mentation factors studied are nearly completely so. Im addition, sensitivity

to bacteriophage Tl fe dominant to resistance, a fact of considerable importance
San
in interpreting studies on the induction of this mutation by radiations and
cheaioale (See 2). We have under way further experinents to determine the
dominance of mtations affecting bacterial resistance to antiblotic aad
antibacterial agents.

2. Genetic gontre) of fermentation anaynes in E. goll.

Investigations on Neurospora have led many vorkere to the conclusion
that ‘eingle genes determine the specifiaity of single ensynes." (See 3.)
Since bacteria are very favorable material for enxyme research, and we have
now the aapacity for genetical work on then, it has seemed desirable to
atudy the preblem of gene-ensyme relationships ia Escherichia coli, to deter-
mine in the first instance whether the "one-to-one" theory quoted above
could be verified,

The ensyne selected for study first is the lactase (bdeta~galactosidase)
of B. goli. It is relatively ensy to isolate mtants which have lost the
capacity to produce this enzyme, by the use of the EMB indicator mediuns
referred to earlier. Sy the examination of some aillions of colonies of
ultra-violet treated bacteria on this medium, severe] hundred, independently
produced, lactose-negative mutants have been isolated. These mutants have
been analysed by croseing them with each other to determine whether they are
genetically identical. [If two lactose-negative mutants are crossed, carrying
the same Lue- mutation, then obviously there will be no lactose-positive recon
Dinante, On the other hand, if two mutants carry mutations at different
genetic loci, then they can be expected to give occasional lactose-posltive
recombinante with a frequency depending upon their linkage relationships.
That is, Lac,-lec,+ x Lac, tiac,~ can give Lac,+hac,+. With these tests,
the mutants eo far crossed can be placed into seven distinct groups, each
earrying a distinct mutant gene which interferes with the production of

lactase. Two of these mutant types aleo show enzymatic effects in addition

to lactase. One of them, Lac.~ is unable to ferment glucose or maltose;
another, Lacg~ is unable to ferment gluconate or maltose. These observations
are net in accord with the simple "one-to-one® theory, but imply that the
relationships between gene and enzyme are much more complex.

In order to suppert this conclusion, the enzyme has deen extracted from
the celle and studied in solution with the ald of an artificial substrate, o-
aitrophenol begalactoside, which releases a colored substance, o-nitrophencl
when it is split by the enzyme lactase, ‘This permits the reaction to be studied
conveniently with a spectrophotometer. The ensyme hes been partially purified
with aumoniun ealfate precipitation, These preparations are active in the
absence of phosphate, pointing to a simple hydrolysis for the enzyme action.
An interesting effect of alkali metele has been noted: the enzyme is strongly
stimulated by sodium ions in fairly high concentration (M/50), and is inhibited
by rubidium. The inhibition by rubidium cen be competitively reversed either
with eedium, or with potassium, maggesting that all of the alkali metals compete
for s position on the enzyme, certain ¢onbinations of metal-enzyme having a
higher efficiency than others. I: is presumed that H+ is also displaced hy
hicher salt concentrations, but that while du-enzyce is more active than
Reensywe, K-enzyme has tho same activity, it has been observed, further,
thet most of the difference can be exprested in terme of the dissociation
constant for the enzyme and ite substrate so that these tone may be regarded
as facilitating the absorotion of the substrate to the eusyze. Tthylene~
diammonium and other substituted amaonium ions behave in much tho same way
as rubidium, These observations are strikingly parallel to the effeots which
have deen noted by Snell on the growth of bucteria.

The onsyme lactase is strictly adaptive, f.6., 1¢ eannoct be demonstrated
in cells which have not been exposed to lactose for at least 2-3 hours. It is

volieved that the various mutations affecting lactase production do se via the
adaptation mechanism, which is now being studied in detail. While the all or
none effect of most of the mutante makes it difficult to analyse their effects,
an allele of Lac3~ hes been found which ie responsive to temperature, and may
be of great help. This mutant 1s wild type at 30°C., but like Lae;- at 40°.
At different intermediate temperatures, different enzymes can be formed. It
hae been pomsible to show that the temperature-sensitivity is not « reflection
of this property of the enzymes themselves, once formed, Imt on the activity
of the adaptation mechaaisus.—

3- Genetic control of gene mutability.

Rhoades hes deseribed "dotted" stocks of sorn in which the status of one
gene affecte the mtability of another. In the presence of the dt allele, the
gene @ is cuite stable and rarely if ever mitates to A (shown ae color in the
aleurene). However, with increasing numbers of Dt alleles, ga shows rapidly
increasing numbers of such mutations, Although this phenonencn is of the
greatest genetice) interest, it is very difficult to study the possible
ehemical pathways because ef obvions enutomical linitations.

fhe observation thet various Lac- mutants of EB. goli, secured as already
deseribed, nuotute at different rates buck to Lact, sugcsested that hers might
be excellent material for a parallel type of study. The mutability of « Lac-~
ateck ia readily datermined by inenbating its seolonies on EMB agar for 23 days.
At this time, Lact mufantese are seen as papillate, dark ontgrowths in the white
or pink colonies.

Matants at the Lac, locus only were studied so far. As obtained by
f{rradiation of the wild type, some of these mtants are quite stable, whereas
others show many papillae in aach colony. Genetic tests on such stocks have
shown that the differences are due to different allelic states of the same

gene, showing again that mtations which may be indistinguishable in all other

respects can be distinguished in this way (i.e. are ise-alleles). The mutations
to Lact in the mitable strains have also been studied, and proven to be true
reverse mitations, restoring the Lac- gone which was originally impaired by
ultra-violet light. Occasionsl Lact mutations in the more stable strains

have proven, however, to be due to matations at other loci, such mutations having
the effect of bypassing or “suppreseing" the phenotypic manifestations of the
original Lac- mitation. Such “suppreseor” mutations have aleo been observed

in the other Lac- mitanta mentioned under Heading 2.

Attempts were then made to pick up effects of mutations of other lool on
the mutability of Lac-, Stable strains were irradiated, and matables looked
for, without success se far, Conversely, mutable strains were irradiated,
and derived stable atrains have been recovered from them, Also, although
the degree of mutability is cuite a constant feature of a particular stock,

a very few spontaneously occurring stuble derivatives have been noticed.

Some of these derived stables, when tested geneticully, proved to be
more stable alleles of the originu] Lan-. At least one such stable, however,
when crossed with wild type gave rise in addition to the parental types:

Lact and Lac-stable, the recombinetion class Lac-mutable. This shows thst
the avparent stability of this derived stock is due to a mutation at another
locus. ‘This new mutation also turned out to be associated with s nutritional
deficiency, which is at least partially relieved by (autoclaved) cosymase
(@tphosphopyridine nucleotide). It is possible that this is not a true case
of gene control of mutability, vat that the second mutation interfores with
the fermentation cf lactose even when Lac,- reverts to +, This point fe now

ander study. A number of additional stable stocks remain to be studied.

4, Genetics of Salmonella.
Esgherichia cold, strain K-12, is, s0 far, the only bacterium in which

® sexunl phase has been demonstrated by the genetic methods detailed above.
oJ

It is naturally of interest to determine whether other bacteria will behave
iu the same way. Previous experiments on two other strains of B. golj have
given negative results, but rather than pursue further strains of this species,
we have been examining various Salmonella strains. Sslmonella is much better
understood serologically, and, furthermore, ie an advantageous group with
which, eventually, to pursue studies on pathogenicity.
Three strains of Salmonella tyohimurium have been examined to date:
8Y-20, S¥~+21 and S¥-23. Biochemical mutants have been isolated in each of
these with the help of a new methed using penicillin which we have just
developed (4%). Peniolllin hae a permanent bactericidal effect only on grow~
ing celle, If a mixture of wild type and mutant cells ie inoculated into a
synthetic mediua to whieh penicillin is added, the wild type cells are killed at
a mach higher rete than the sutante, This property can be used to advantage te
facilitate the isolation of biochemical mutants from irradiated populations in
which they are far outnumbered hy the original, non-mutated wild type cells,
Ho evidence of genetic recozbination hae heen found so far in aixtures
of mutants fn the following combinations: SY¥-20 x SY-20; SY¥-20 x ST~21;
8Y-21 x SY~21; SY¥-23 x S¥~23; SY-20 x S¥-23, In several experiments, how-
ever, prototrophs have appeared in mixtures of mutants derived from SY¥-21
and SY-23, The situation has been complicated by the fact that ST-23 ie
lysogenic, {.6., carries withont detriment to itself a bacteriophage
which is active on S¥Y-21. Ina addition, S¥-21 is lysogenic, one of its
phages {s inactive on §. §yphimariug streins, and is revealed only when
S. gallinarum is used as a sensitive indicator; the other can attack 5I-
23. Lysogentoity is itself an imperfectly understood phenomenon, and in
the present instance, it has not yet been possible te clarify its re-
lationship to the recombination which may be occurring in SY-21 x SY-23.
The protetrechs which can be recovered from sugh "croseus" are usually
ridden with phage, and it is likely that the very low yield with which they
have been recovered may be secounted for by the destruction of many much prote-
trophs ty these phages.

Yor this reason, ve have tried to determine whether a lysogenic bacteriun
ean be “Aisinfected" of its phage. Experiments with the 8I-21 - 9. geliinarug
aystem have given rather discouraging results. All of the bacteria in euch
cultures seem to carry the phage, as determined hy plating them out and testing
individual colonies, and it has not been possible te eliminate the phage with
muh chemicals as Phosphine GHE and potassium arsenite which are reported te
inhibit phage multiplication. Other agente will have to be tested.

This work has been supported by a grant from the Eational Institute of
Health.
1.

2.
3.

4,

5.

oll.

References

fatun, B. Lb. and Lederberg, d. Gene recombination in the tacteriua
Eegherighia gold. J. Bact. 531673-684. (197)

Lederberg, J. Gene recombination and linked segregations in Escherichia
Mold, Genetics 321505-525, (1947)

Lederberg, ¢. Problems in alerodial genetics. Heredity 22245-198. (1948)

See Bonner, D. Genes and Cytoplasm, Genes as determiners of cellular
Diechemistry. Selence 108:579. (1948)

Lederberg, J. and Zinder, N. Coneentration of blochemioal mutante of
Dagteria with penicillin. J. Amer, Ghes. Soc. In Presse, (1948)

lederberg, J. Deteetion of feruentative variants with tetrasoliun.
J. Bact. $6:695. (1948)

December 1948