August 26, 1946.

Dear Dr. Edwards:

It was most gratifying to hear from you, «nd to find your interest
in our work, It is particularly fortunate,I think, that we do differ in
our formalations of the Salmonella story, re-lizihg that I have had to
approach the matter on an entirely a priori basis with some genetic ideas
transported from other organisms, while you have ahd all the practical
and experimental contact with these organisms.

The cultures ahd serums which you sent have been received in good
condition, aid we are set to go to work with them immediately. However,

a new tack has suggested itself which modifies somewhat the approach I
suggested previously. There seems to be already available in many Salmonella
types a diversity of nutritional requirements, which should be as usable

a8’ induced nutritionel mitents in detecting gene recombinations. Therefore,
instead of selecting, at this mak atage of the game a particular antigenic
type/, a large series gef Salmonellas will be workdd over for their
nutritional requirements, and those selected «hich will be on that basis
most. convenient. I have been able to obtain a few varieties alread from
the Columbia and Yale Laboratories, and have enlisted Dr. ‘theeler!s
cooperation in extending that lest.

There seems to be a most strikigg analogy between phase variation
in Salmonella and some features in the inheritance of mating type in
Paramecium bursaria as described by Jennings (Proc. Amer.Philos. Soc.,
85:25, 1941), in particular the phenomenon of 'self-differentiation',
Jennings finds that a clone of a given mating type (of which there are

four alternatives; may give rise to organisms of a different type; dif-
ferent clones of a given mating type characteristically 'self~differentiate'
to a given alternative, which may be different for different clones. The

clone ootained in this way may in turn 'self-differentiate', always to

tne original type. Although Fara:ecium is a sexual grganism, the genebic

basis of this phenomenogi is stiil quite obscure, and hybridization experiments
so far Rave shed no great light on the matter. I do not know whether any
attempt has been made to characterize mating types serologically.

It will be some time before the survey of nutritional requirements in
Salmonella will be completed, but we shall, of course, keep:you informed in
detail of our progress. While Dr. ‘heeler's collection should be adecuate,
we would appreciate the opportunity to ascertain the nutritional requirements
of any strains in which you may be particular interested,

Re~reading my original letter to you, there seem to be some considerations
that I did not express quite clearly. In order that there may be no misunderak-
standing, the situation should be reviewed:

a) The & col4 work damonstrated that there are genes in bacterla which
caetermine such characters as nutritional requirements, and phage suscep=
tibility.

b) In E. coli, different substrins can be hybridized giving rise to new
types, characterizable as recombinations of the different characters
as the parents.

Se have no evidence as yet as to the inheritance of antigens, nor as

to the occurrence of gene recombination in, or mkbetween Salmonella
strains. In E. coli, the recombination of nutritional requirements can
be used as a basis for selecting vout of a large number of celis those
few in which recoubination hus occurred. In such cells, the segregation
of ovher genes can be determined.

In Salmonella, recombination of nutritional requirements must first be
determined. If this occurs, we are in a position to study the inheritance

of other characters (such as antigens).
Only if a system of inheritance of antigens can be formated on the basis of
‘crosses! can the problem of phase variathon be studied. It probably will be
best to use a form which is normally almost entirely monophasic, but in which
the altersative phase ean be detected ana selected for by well-known methods.

Parentheticelly, the expression ‘induced Variation’ has a rather different

connotation in genetics from that in bacteriology. It is taken to imply a direct
action of the inducing agent on the hereditary materials of the orgeniam., Unless
there is direct evidence to the contrary, cases of apparent induced variation
are formiated as instances of selection in favor of a small proportion of
sponteneausly occurring variants. Your own work seems to bear this out in °almonella,
In the bacteria, only the pneumockccus transformations have been accepted by
most geneticists as bona fide examples of ‘Induced Varaition'. There may of course
be others.

It would be most convenient to have a set of specific Salmonella phages, Can
you suggest where whey might be obtained, and what references there may be in the
literature? |

Finally, I hope that I shahl have an opportunity to visit your laboratory
sometime in the next year, but there is nothing definite in sight as yet.

Very sincerely yours,

Joshue Lederberg.