February 20, 195)

The Genetics of Bacteria (C-2157 M & I: formerly £72-C5)

Annual Progress Report submitted to the
National Institutes of Health
Public Health Service
Bethesda 1), Maryland

by

Josta Lederberg, Associate Professor of Gere tics
Department of Genetics, University of Wisconsin

This report covers the period from October 21, 1952 to Jamary 31, 195)

A. Summary

Studies on the mechanism of genetic transduction and its appli-
cation to immmogenetic analysis are detailed, A wide variety of
serotypes can now be artificially synthesized.

In E, coli, single cell studies and the analysis of persistent
diploids have confirmed the picture of sexual recombination in this
species as including the interaction of two intact parental cells
and the migration of an intact nucleus from one parent to the other.

This is not a publication, but may be informally quoted as a private commnicatian.
Genetics of Bacteria
SALMONELLA

During the period covered by this report, our program has enjoyed the close
collaboration of Dr. C. C. Spicer and Dr, A. Bernstein, both on leave from the
British Public Health Laboratery Service, and especially of Dr. P. R. Edwards
of the Enteric Bacteriology Laboratory, Communicable Diseases Center (P,H.S.),
Chamblee, Ca. Some of these cbservations are also directly related to previous
wovk of Dr, Bruce Stocker (presently at the Lister Institute, London) and the
dissertation of Mr, (now Dr.) N. D. Zinder, presently at the Rockefeller Institute.

In previous reports and publications (1,9,10,11) an account has been given
of the discovery of genstic transduction by means of temperate bacteriophage,
which forms the basis of most ef our work in Salmonella genetics at this time.
Briefly, it has been established that temperate bacteriophage particles may
carry small hereditary fragments from the cells on which they have been grown
to new host cells, Provided that the new hosts are not killed outright by the
phage, they may then acquire the genetic traits thus transduced. Since each of
a wide variety of genetic markers is subject to transduction, it is inferred that
they are typical of the entire heredity of the bacterium. However, the fragmenta-
tion of the genotype in transduction dees not necessarily imply a similar disorganiza-
tion jin the normal vegetative history of the bacterium. More likely, the hereditary
content is normally organized in chromosomes or nuclei, and is only broken up (and
transferred to other cells) by the lytic action of the growing phage. The most per-
plexing problem 1s the means by which such fragments can be implanted in the stable
genetic structure of the new host, However, this enigma is shared by students of
related processes of genetic exchange among bacteria, viz. the "type transformations"
of the pneumecoceus and the influenza bacillus, These phenomena betray a fundamental
similarity that justifies their being subsumed as comparable examples of genetic
transduction. The unique feature of transduction in Salmonella is the intervention
of bacteriophage; in transduction in the pneumocnccus, nuclear fragmentation is
achieved by chemical means, but the manner in which the fragments are able to
penetrate the bacterium without the help of a biological vector is not understood

at all.
BACTERIOPHAGE AS THE VECTOR OF TRANSDUCTION

Previous reports (1,10; see also 12) have cited the evidence for the gross
morphological and surface (i.e. sige, reactions with antiserums and bacterial
receptors, behavior in chemical purification and response to inactivating chemicals
and heat)similarities of the Salmonella transducing agent (FA) with phage, However,
these findings would be reconcilable with the release of two categories of particles
by bacteria in which phage has grown: phage, strictly speaking, and superficially
similar particles carrylog FA, to the exclusion of phage nuclei as inhabitants of
phage skins (13). The only contribution to this problem has been the finding of a
statistical association of the incidence of induced lysogenicity among transducti ons
and their unaltered neighbors under conditions where not all the surviving progeny
of infected bacteria are lysogenized, As in E. coli (),1)), infected Salmonella
cells generally give rise to mixed clones with sensitive and lysogenic survivors,
and the observed correlations presumably refer to the coordinate distribution of
lysogenicity and the transduced trait among the progeny of infected individuals,
This point has been much more definitely established with an E, coll system (see
below). The relationship of phage to FA has also been examined with the help of
a virulent mutant (227) of the temperate phage 22. As bacteria lysogenie for 22
are immune to 227, selective survival of 22~infected cells is achieved by exposing
them to 22¥, With a lew ratio of phage 22 to bacteria, 22V was foun (as expected)
to destroy most of the bacteria, but the incidence of transduction was not concure
rently reduced, i.e., transduction is associated with protection against 22V. As
the surviving bacteria are all lysogenic for 22V, it is concluded that the FA
particle contains both a phage 22 nucleus and a bacteriat fragment, That this
association is fortuitous is suggested by the virtual absence of specificity of
transduction, i.@., any single trait is subject to transfer at arate within a50-fold
range in tne neighborhood of one per million phages, In addition, the phage nucleus
may be functionally separated from FA in either of two ways: the use of a recipient
host immune to the phage, which does not impede transuction so long as the phage is
adsorbed by the bacterium, or by treating the phage with ultraeviolet light, which
destroys the lytic and lysogenizing function much more rapidly than transducing
competence, Since very large doses of UV are known to interfere with the adsorbing
power of phage, it is questionable whether any direct action on the enclosed fragment
has yet been demonstrated, however interesting this might be in anticipation of
mutagenesis studies with "free genes",

The phages employed in previous transduction work were all obtained from
lysogenic S, typhimurium, and are probably related to Boyd's Al series (6,15).
The “virulent mutant" 22V has a somewhat reduced transducing competence, provided,
of course, immune (lysogenic or otherwise) recipients are used. Baron (16) has
reported that the S. typhi typing phage Vi IT, in several of its adapted states
is also competent. Through the courtesy of Dr. A. Felix and Dr. E. S, Anderson
it has been possible to test a number of typing phages for 5, paratyphi B and
type-determinant phages of S. typhi (9); of the former, "B.A,0.R." and uf the
latter, k, proved to be capable of transduction, Thus at least four, serologically
unrelated, phages of the Salmonella group will function in transduction, pointing
to the possible generality of thr phenomenon, Unfortunately, the effective host
range of the other phages is generally rather less than that of 22, though the Vi IT
might be expected to function with Vit strains of S. paratyphi C(8). To date,
however, transductions have been Limited to groups A, B and D. Although these groups
share the "XII" somatic antigen complex, it has not been possible to inculpate any
specific component as the unique receptor.

ANALYSIS OF NON-MOTILE VARIANES

The solidification of the principles of transduction analysis now permits
their application to specific genetic problems, The first of these to be considered
was the determination of the flagellar apparatus (6). Previous conclusions on the
existence of three levels of determination=-the formation of flagella, their anti-
genic content, and their locomotor function~<-by distinct sets of non-allelic genes
have been confirmed in further studies, The restoration of motility has proven to
be a useful artifice for the typing of nonemotile Salmonella variants which could
not otherwise be diagnosed as to serotype for epidemiological control.

LINKAGE

& faw exceptions have been noted to the otherwsie well-founded rule that only
a single factor can be transduced at a time (the basic evidence for fragmentation as
the first step of the process), The most clear out of these exceptions involves one
of the factors for motility (Fla,) and the factor controlling the flagellar antigen
of phase 1 (By). Transductions of motility to a nonemotile strain of S. paratyphi B,
SW-666, whose constitution may be given as Fla, - H,”,will result in Fla* , i.e., motile
selections whose flagellar antigen may be either H,5 or that characteristic of the
donor strain, That is, if this is S, typhimurium, Flay 4)° =x Fla, Hb may
engender either Fla* H,® » a9 required by the single transduction of the Fla marker,
but algo in some Flat Ht 3 as would indicate the concurrent transduction of the
marker as well, An extensive series of backerosses and tests with a variety of distinct
donors (6,7) have helped to verify that this is indeed a dual transduction, occurring
mich too frequently to be explained by coincidence, and therefore to justify the
invocation of linkage, It is not unlikely that Linkage is as characteristic of

the genetic makeup of Salmonella as of E, coli, but that it will be obscured by the
mechanics of transduction except for markers unusually closely related to one another.
Similar observations have been made for the transduction systems of other organisms
(reviewed by Zinder, 12).

As there is no means of selecting for the other combination, Fla H+ » this
is not an ideal system for linkage studies. However, other Fla” markers linked to
Hy have been found, e.g. in an S. dublin (recognizable as such after the trans-
duction of motility!) from Guatemala. Attempts were made to map these two Fla
factors and Hy, to determine if they would fit a unique linear sequence, but the
results were ambiguous owing to the looseness of one of the linkages. (This
program is being extended by Dr, Stocker with more comprehensive materials).
However, significant progress in mapping studies will probably ensue only after
additional linked pairs are discovered, The negative results of previous efforts
portend that the projected search will be tedious,

ABORTIVE TRANSDUCTIONS

With most markers, it is possible to score only those genetic transfers that
result in stable, hereditary changes insofar as the necessary selection and identi-
fication procedures require the handling of very large clones. However, motility
ean be scored directly or indirectly on single cells, so that it has been possible
to detect abortive transductions of motility to nonemotile variants. These were
firet observed (6) as trails of small colonies that have worked themselves deep into
soft agar, but do not swarm out as motile clones, The trails were concluded to
represent abortive transductions, i.e., the consequence of the introduction of a
genetic factor (Flat for motility) into a non-motile recipient in such a fashion
that the factor can function but can not reproduce, The motile cell thus deposits
a spoor of non-motile progeny at each fission, This interpretation has been verified
by direct micromanipulatory studies on the motile transductions (Stocker, private
communication, and confirmed, Lederberg, unpublished) which also promise to give
more intimate knowledge of the early steps of the transduction process. The
possibility that abortive transductions result from the transmission of fragments of
a size or configuration unsuitable for their incorporetion in the stable chromosome
set is a leading hypothesis that hag, as yet, no particular evidence, It would be
difficult to imagine that the transduced fragment is directly incorporated so as
to displace actively the previous homologue; instead, it may be incorporated (either
directly, or as a copy) at a subsequent replication of the genetic material,

An alternative hypothesis for the trails would have assumed that they reflect
the persistence of an unincorporated residue jneluding Flat after part of a fragment
had been incorporated by crossing over. Hewever, tests of trails from transductjons
with linked markers have failed to uncover crossover types such as the me
referred to umler Linkages, The antigenic character of the motile bacterium in
case would be of interest, but has not yet been settled, As trails are inhibited
by either serum it may be bti, 1.e., reflect a "double heterozygous" condition for
the segment, Flat |/-~ 1 Hi Pannm. If this is generally the case, it would
suggest that linked”factors are separated by some sort of cross-over process in
addition to the primary fragmentation {as has been shown in the E. coli system).
IMMUNOGENETICS

The Salmonella group is especially appropriate for immunological studies,
both from the point of view of the fund of consolidated information, and the im-
portance of systematic findings on the serological and immunogenetic structure of
a medically important bacterial group. The flagellar, or H, antigens are especially
handy, Previous accounts (7,10) relate the results of systematic serotypic recombin-
ations, This report will emphasize otherwise unrecorded details, They support the
thesis that the mosaic structure of the group can be most readily explained by
recombination processes, But it must be reemphasized that the genetic possibilities
are only the initial raw material for evolutionary forces, The typhoid fever
syndrome is still associated with a bacterium with certain biochemical peculiarities
and, i.a., the d flagellar antigen, though it is dubious whether these are alt
directly concerned with pathogenetic properties.

Although S. pes aes and "S, pullorum" are serotypically indistinguishable
the former ean be shown to carry a covert Hy antigenic factor pm, identical with
that of S, enteritidis (reciprocal absorptions by P. R. Edwards, priv. comm.), as
well as certain "la* factors, On the other hand, S. pullorum has given completely
negative results in similar experiments, supporting the differentiation of these
species (though identical serotypes) which cause distinct diseases in the fowl.

No criticism is implied of the serological scheme nor of the application of binomial
epithets to serotypes} To be consistent, however, a serotype should include all
serolopically similar taxons (e.g. S. paratyphi C and S, choleraeesuis) and nothing
else, and not be confused with species, the delineation of which has wisely been
held in abeyance by international agreement.

Experiments like those reported previously on the mechanism of diphasic
flagellar variation (7,10) have been contimed and amplified, supporting but not
proving the same tentative hypothesis: that the two alternative phases reflect the
allelic states of two distinct loci, resembling e.g., the Rh:CDE and the ABO
systems of human blood groups. In Salmonella, however, only one phase of a cell
is expressed at a time, and almost all of the progeny of any cell remain in the
same phase, The chief uncertainty concerns the mechanism of this alternation
and its heredity, but the data still suggest a local fluctuation from active to
inactive states, and vice versa, at the two loci. More decisive experimental
material and methods are being sought.

A POSSIBLE GENETIC DUPLICATION

In a systematic examination of genetic homologies, the "specific" or phase-1
factors of a wide variety of serotypes were found to be mutually homologous (i.@e,
would displace one another in transduction experiments), and the same for the "group"
or phasee? factors, However, a strain of S. paratyphi B, var. java (CDC~157) (7) was
found to be exceptional. This was first examined as an apparent "monophasic" 1,2..
type, but its 1,2... factor was found to be homologous with typical phase~l (rather
than phase? as expected for this antigen), The parent strain, showing b:1,2 (as
expected for this serotype) was then studied, It showed a quite sluggish phase
variation from b to 1,2, whose reversal was progressively more difficult. Nevertheless,
(by transductions in both directions) it was possible to show that both the b and the
1,2 factors of this strain are Hy homolo e8y I.E., the immunogenetic constitution
of this strain must be represented as H,” H, 92 » though as in typical diphasics,

H,> Hots 2 3 only one locus is patent at any time, This constitution is required to

explain not only the homology of 1,2 with specific phases in transductions from this

strain, but the synthesis of such types as bra and then c:a (otherwise unheard of)
from transductions to the strain from typical stocks, Though both H, factors are
linked to Fla,, atteapts to demonstrate their linked transduction were frustated

ty the progressive diminution of phase variability required to demonstrate the rull
antigenic potentiality, Monophasic variants of Salmonella species are well known for
their usefulness as antigenic reagents. Three types have been distinguishad by
genetic analysis, 1) S. typhi: no locus homologous to Hp can be demonstrated; if
ever present it has prssunebiy been deleted; 2) Some monophasic S, paratyphi B
strains: an H, lecus is demonstrable by transductions to the strain leading to the
diphasic condition, but not otherwise. These can be understood as carrying a "null"
allele at the Ho locus. 3) S. abortus-equi: an allele is present at the unexpressed
locus, The gonetic background is presumably unfavorable to phasic variation in any
direction, but has not been satisfactorily dissected, When transduced to other
stocks, the componente show normal behaviov in at least some cases,

For technical reasons, the somatic antigens (0, Vi) have not proved so amenable
to genetic study. Some antigenic variations have been noticed incidentally, e.g0,
the acquisition of the V factor (by a S. abortuseequi in the course of transductions
involving its monophasic character, Howevex, neither this example, nor alterations
in the I component of S. paratyphi A are convincing examples (compare 18), as they
have not been consistently related to the occurrence of a comparable factor in the
donor cell, In facet, just these antigens are known to undergo "Form Variation" in
at least some stocks, and it is possible that circumstances of the experiments.
apart from the transduction itself played a part in the selection of the antigenic
variants, These experiments are being extended in several other laboratories
(Washington, London, Copenhagen).

IMMUNOCHEMICAL DIFFERENTIATION OF FLAGELLAR PHASES

An earlier, and generally overlooked report (19) on the differentiation of
phase-l1 and phase»? has been confirmed and amplified, In tube tests, after 2);
hours at 37 C., acriflavine will agglutinate smooth, flagellar phase=2 cultures,
but not phase~l, The differentiation has been found be general throughout the
group (except that Hts? is agglutinated. However, Hp“”, S, wien, is agglutinated;
dj)", S, dar-es- salaam is not.) The same reaction holds for the separated
flagella, The agglutination is remarkably similar in appearance, and in timing
and reaction to heat to that of a serum H agglutination, O-~forms are unresponsive,
the reaction is specific from pH 6-8; salt is not required, Several basic amino-
acridines are equally effective, but 1,2,3,«tetra-hydro-5-aminoacridine (generously
furnished by Prof. A. Albert) was not, in surprising agreement with the bactericidal
specificities of these compounds, Experiments with purified flagella have suggested
that both phases will combine with the dye. Thus there may be a subtle difference
in the physical structure of the flagella in the two phases, corresponding to the
determination by two loci.

ESCHERICHIA COLT

These studies enjoy the active participation of Dr. E, M. Lederberg, Dr. T. C,
Nelson and Mr, M, L. Morse. The sexual recombination mechanism of E, coli (5,20, 21)
is to be sharply distinguished from transduction as it operates in Salmonella, e.g.
the interaction occurs exclusively between intact cells of the two parents, and the
interchange of genetic factors occurs in large blocs approaching or equivalent to
the entire genotype. In Salmonella, interchange generally involves just one marker
at a time (very exceptionally two) and is mediated by filtrable bacteriophage
particles, More recent observations (22) have indicated a formerly unsuspected
polarity in the sexual behavior which is best explained by the assumption that one
parent transmits only a nucleus, the other both nucleus and cytoplasm to the zygote.
~6-

But despite contrary speculations, (23), there remains no evidence to justify

the implication of an extracellular, virus-like agent as the vector of the

migratory mucleus, all experiments designed to detect recombination via extra=
cellular elements having been categorically negative. The working hypothesis

that still fits all present factual knowledge is that the zygote proceeds from a
conjugation (contra copulation) of two cells in the course of which nuclear migration
takes place, However, while this hypothesis has the support of the negative evidence
just quoted, it has not yet been affirmatively substantiated by direct morphological
study, for which there are considerable technical obstacles, ©

The discovery of the differentiation of F+/F- states has been recounted
previously, (2,3,10). Fe cells appear to be able to function only as the stationary
parent; F+ as either migratory or stationary, possibly on the basis of a physiological
heterogeneity of the cells in a culture. Thus F- x F= crosses are sterile, while
other combinations are fertile, The remarkable, and least umlerstood fact is that,
i.a., the state of a culture depends on the presence of a hypothetical "F+" agent,
That is, F- cells can be converted to the F+ state (which persists as a stable
hereditary potentiality) by mixed culture with other F+ cells, The simplest
interpretation of this finding is to assume an infective, hereditary particle,
the "F+ agent", But despite an extraordinary efficiency of transfer of the F+
state in mixed cultures, all experiments with cell-free preparations have been
entirely negative, For example, actively exchanging populations have been separated
through membrane filters directly into a competent recipient, in a matter of a few
seconds, The contagious properties of the F+ state would make this a very sensitive
test, but it gave a negative result. It can only be concluded that F transfer requires
the direct, superficial contact of two intact cells: it ia not necessary to assume
cytoplasmic connections, Experiments to duplicate the experiments under microscopic
control are in process, If the intact cell is the normal biological vector of the
¥ status in transmission experiments, there may thus be a sense in which one could
refer to the F agent as the vector of recombination, i,e., this is the cell itself,
This reconciliation of apparent antitheses makes it likely that any resolution will
have to follow the physical separation of the F agent as an extracellular entity;
any functional separation, such as already recorded, will be equivocal, However,
insofar as the presence of an infective F agent has been only one of several
genetic and physiological factors controlling incompatibility, and in particular
several’, other lines of E., coli seem to function in recombination as well with as
without a transmissible F agent, it would seem premature to place the primary
burden of genetic exchange on this obscure principle.

The analysis of crossing data in E, coli K-12 has been obscured by discrepancies
from mendelian segregation for which no rationale was in sight prior to the discovery
of F polarity, While, in 197, one had no basis to doubt the single linear linkage
grouping of the then available markers (20; see also 25) the discovery of aberrant
heterozygotes in 1918 indicated that "the linkage map of E. coli Kel2 will have to
be systematically reexamined, with the use of several unrelated sets of stocks"

(2h), a warning that has since been reechoed (5,11, A). The qualitative anomaly of
these stocks was hemizygosity (i.e. deficiency in one of two homologous chromosomes)
for certain markers, particularly Mal and S (controlling maltose fermentation and
streptomycin-resistance, respectively). The possibility was considered, at that
time, that the deficiencies reflected defects already in the parental gametes, but
had to be rejected for reasons similar to the argument below,

Not the least difficulty in rationalizing the segmental elimination that
was postulated to explain the hemizygous defect was its non-random involvement
of the contributions from the two parents. It was later realized that these parents
were polarized in regard to F, and that it was the contribution of the F+ parent that
was usually, but not always eliminated, In all these studies, heterozygosity for Mal
-7-

hed never been observed (barring a single primary nondisjunction of different
origin), including tests of nearly a thousand diploids in whose parentage Mal+ was
either an "unselected" marker, or in which Mal+ ( having come from the F+ parent)
was selected on synthetic maltose mediun,. ‘The Mal--S region thus clearly follows
a different rule from other factors such as Lac and Mtl which are regularly hetero-
gygous wheather selected or not.

An explicit study of the role of F polarity on the character of diploid exe
ceptions has therefore been undertaken, and recently concluded for the Mal--5
region (26), To simplify the account, markers of the Fe parent will be referred
to as orthotypic, from the F+ as paratypics when both are represented in part, we
have metatypic. It was first verified that regardless of parental couplings of
P+/P< elimination in the Mal~-S region was usually paratypic, i.e., about 85% of
the diploids (selected as Lac +/+ prototrophs) retained the orthotype Mal--5 markers.
However, 13% were orthotype eliminations, retaining the paratype segment, and 2%
were metatype, retaining ons of the markers from the F+, one from the F-, The
latter two classes, and the complete absence of heterozygotes for this region are
inconsigtent with a prezygotic defect of the F+ gamete, as proposed elsewhere (23).
The metatypic hemizygotes exculpate possible reversals of polarity, by means of
twansmission of F on the cross plates, a possible error also excluded by other
evidence. It is therefore concluded that segmental elimination occurs after a
complete zygote has formed from the union of two intact gametic nuclei. This
elimination will usually involve the paratypic segment. Where it does not, one
may assume that there has been a preceding cross-over between the markers and
specific site of chromosome separation (see also 30), A superimposed pre-zygotic
elimination in some cases is not excluded, but supererogatory. For most analyses,
it is immaterial whether the elimination is gametic or zygotic; as a rule only
experiments permitting the observation of the unreduced diploid will allow a
distinction.

Microscopic studies of E. coli recombination have beon initiated to define
more closely the cellular events of the process, "Hfr" strains are available
which allow of up to about 10% recombination after about ) hours growth in broth
and tests of colonies from dilute platings on unselective media, (Other criteria
of the incidence of recombination are inadmissible without detailed kinetic analysis
owing to residual growth of heavier inocula even on selective media.) These crosses
show a polarization favoring orthotypic markers of the Mal and Gal regions to an
extreme degree, while Lac, Vi and Ara are much less distorted. At first sight, it
would be tempting to ascribe such a polarization to a failure of paratypic markers
to enter the zygote, but single cell analyses have proven otherwise,

Under conditions permitting microscopic control, the maximum incidence of
recombinants is somewhat reduced, 301 single cells have been separated from
crossing cultures and allowed to form clones, later examined for their content
of genotypes, 28), of these clones were pure cultures of one of the two parents .
Of the remaining 17 single-cell clones, all engendered progeny typical of the F-
parent, but in addition, 7 also contained cells carrying the full genetic complement
of the F+ parent, These clones also contained numerous recombinants, but these
were generally orthotypic. Thus in 7 single cells, the full genetic content of
the two parents was represented transiently, but the recombinant output (some
possibly from subsequent crossing) was still restricted to the orthotypes. The
renaining 10 clones contained only orthotypic recombinants and parents, Tims,
the frequent exclusion of paratypic markers from the recombinant progeny cannot
be aseribed to their exclusion from the zygote, but is assigned to a later elin~
inaticn process as already suggested by the study of more persistent diploids,
These experiments are designed ultimately to define the formation as well as
the fate amd content of the zygote. The efficiency of mating is still too low to
make the solution of the problem as direct as may seme There has been a possible
correlation of zygote formation with clumps of the two parent cells, but the age
gregates have been difficult to analyse. The zygote cells appear to be slightly
larger than the parents, No evidence of copulatory fusion or of specialized sex
sivuctuyves or spores has been founds on account of trivial clumping and transient
associations, a passing conjugation would not be so readily delineated. These
studies are continuing coordinately with cytological studies of fixed and stained
preparations.

LYSOGENIZATION AND TRANSDUCTION

Studies are continuing of the genetic role of bacteriophage in E, coli K-12.
Tt has been possible to confirm previous findings on the localization of the deter-
minant of lysogenicity, Lp*, (prophage?) on the bacterial chromosome in the vicinity
of a set of determinants for galactose fermentation, (Gal), These findings have
also been confirmed, extended, and subjected to attempted reinterpretation by various
other workers (27,28,29). While crosses of lysogenic x sensitive give progeny
generally orthotypic for this trait and for Gal, para» and meta-typics are also
found, so that Lp* behaves like any other marker in its transmissibility by either
parent, But the most decisive evidence for chromosomal localization stems from
diploids, obtained by rather involved sequences of crosses, that are heterozygous
for Lp* Gal+/Lp® Gal”. ‘Such diploids, which segregate these markers usually, bat
not always, in the parental couplings, are also obtained by the infection of
LpS/Lp®, homozygous diploids, giving rise to segregation for Lpt/Lp® such as has
not been observed from the infection of haploids.

A naclear basis for the fixation of symbiotic bacteriophage was algo suge
gested by the segregation of lysogenicity among the progeny of presumed single
infected cells (5; also the subject of a more extensive study, 1), This point
remains to be decisively verified by single cell methods, whose application is
currently planned,

A limited system of transduction by this phage, confined to the markers
closely linked to Lp, is the subject of the dissertation work of Mr. M. L. Morse.

1 Many of the foregoing problems have been ths subject of a collaborative corres-
pondence with Dr, L. L. Cavalli of Milan, Italy. We are anticipating a visit from
Dr. Cavalli during the term of the present grant which will allow an even more direct
association.
Ae

B,

-9-

REFERENCES

Publications from this laboratory that have appeared during the period of this
report.

1)

2)

3)

\)

5)

6)

7)

8)

Zinder, No D. and J. Lederberg. 1952. Genetic Exchange in Salmonella.
J. Bact. 64: 679~699.

Lederberg, Ji, L. L. Cavalli, and E, M. Lederberg, 1952. Sex
compatibility in Escherichia coli, Genetics 37: 720-730.

Cavalli, L. L., J. Lederberg, and E, M. Lederberg. 1953. An infective
factor controlling sex compatibility in Bacterium coli. J. General Microbiol.

8 : 89-103,

Lederberg, E, M. and J. Lederberg, 1953. Genetic studies of lysogenicity
in Escherichia coli. Genetics 38: 51-6).

Lederberg, J, and E. L. Tatum, 1953. Sex in bacteria: genetic studies,

-1945~1952, Science 118: 169-175.

Stocker, B.A.D.S., Zinder, N. D, and Lederberg, J. 1953. Transduction
of flagellar characters in Salmonella, J. Gen. Microbiol. 9: 10-433.

Lederberg, J, ani P, R. Edwards, 1953. Serotypic recombination in Salmonella.
Jour, Immunol. 71: 232=2)0.

Cavalli-Sforza, L. L. and J, Lederberg. 1953. Genetics of resistance to
bacterial inhibitors. Symposium on growth inhibition and chemotherapy.
International Congress of Microbiology, Rome. Pp. 108-12.

Previous reports.

9)
10)

E72-C3, submitted February 11, 1952.
E72-Ch, submitted October 25, 1952.

Other references.

11)

12)

13)

Uy)

15)

16)

Lederberg, J., E. M. ‘Lederberg, N. 0. Zinder, and E, R, Lively. Recombination
analysis of bacterial heredity. 1951. Cold Spring Harbor Symp. 16: 413-433,

Zinder, N. D. 1953. Infective heredity in bacteria. Cold Spring Harbor
Symp. 18: 261-269.

Levinthal, C. and H. W. Fisher. 1953. Maturation of phage and the evidence
of phage precursors, Cold Spring Harbor Symp. 18: 29-33.

Lieb, M. 1953. Studies on lysogenization in Escherichia coli. Cold Spring
Harbor Symp. 18: 7173.

Boyd, J. S. K, 1950. The symbiotic bacteriophages of Salmonella typhimurium,
J, Path, Bact, 62: 501.

Baron, L. Ss 1953. Genetic transfer by means of Vi phage lysates, Cold
Spring Harbor Symp. 18: 271-272.
17)

18)

19)

20)

21)

22)

23)

2h)

25)

26)

27)

28)

29)
30)

~ 10 =

Anderson, E, S, and A, Felix, 1953. The Vi type-determining phages
carried by Salmonella typhi, Jour. of Gen. Microbiol. 9: 65-88.

Kauffmann, F, 1953. On the transduction of serological properties in
the Salmonella group. Acta. Path, Micr. Scand. 33: 09-20.

Sertic, V. and Boulgakov, N. A. 1936. Ltagglutination par la trypaflavine
des Salmonella de structure antigenique flagellaire nonespecifique., C. R.
Soc, Biol, 123: 951-952,

Lederberg, J. 1947. Gene recombination and linked segregations in
Escherichia coli, Genetics 32: 505-525,

Lederberg, J. 1951. Genetic studies with bacteria, In: Cenetics in the
20th Century, MacMillan: New York. Pp. 263-289.

Hayes, ™. 1952. Recombination in Bact, coli K-12: Unidirectional transfer
of genetic material. Nature 169: 116-119,

Hayes, W. 1953. The mechanism of genetic recombination in Escherichia coli.
Cold Spring Harbor Symp, 18: 75=93,

Lederberg, J. 1949. Aberrant Heterozygotes in Escherichia coli. Proc, Nat.
Acad, Sci, U.S. 35: 178+-18h,

Rothfels, K. H. 1952, Gene linearity and negative interference in crosses
of Escherichia coli, Genetics 37: 297-311.

Nelson, T. C, and J, Lederberg -- mamuscript in preparation, Post~-zygotic
elimination in Escherichia coli.

Wollman, E, L. 1953. Sur le déterminisme génétique de la lysogenie, Ann.
Inst. Pasteur 8h: 282-293.

Appleyard, R. K. 1953. Segregation of lambda lysogenicity during bacterial
recombination in E. coli K-12, Cold Spring Harbor Symp, 18: 95-97.

Lwoff, A. 1953. Lysogeny. Bact. Rev. 17: 269337.
Cavalli-Sforza, L. L, and J, L. Jinks, 1953. Observations on the genetic

and mating system of E. coli K-12. Abstr. 9th International Congress of
Genetics, Bellagio.