November 13, 1950.

Dre Plerre Frederica,
Institute of Becteriolocy,
University of Liege,
Relctun.

Dear Or. Frederica:

Thank you very much for sending the set of reprints, which arrived
several days ago. I have been studying thom diligently, and profitably,
Bince receivine ther.

My previous impression, that there was an unlimited series of distinct
colicins, seene to be inoorrect in view of your classification of 17
main types. In view of this, I wonder if 1¢ would not be simpler for you
to send a representative culture of each cclicin type, and your basic
seneitive strain, such as CA @i.

I well appreciate that uy request for tramsshipping 18 cultures might
put you to some considerable trouble and thet you might well wish same jus-
tification for euch a request. “y primary interest in colicins 4s to use
them as genetio testers, since the mutations for recistance to different
colicine appear to be eerentially independent. Thus, 17 colicins would,
in principle add 17 additional senetic markers for use in our recombination
etulles. Iam particularly interested in application to a new development,
namely crosses between distinct strains of E. coli, which have recently
been shown to be feasible. The cenetic differences existing between strains
as isolated (a8 epposed to their experimental induction in the laboratory?
ic a matter of some interest to the geneticist as well ac the taxonomict,
and differences in colicin responnes promise to have some use in thic direction.

On page 84 of your “These d'agregation", I note a reference to colicin-
sensitivity of S. typhirurium. One of my students hes been studying recom
dination in this species, with resultewwhich barely constitute definitive
evidence for recombination between a few strains. His major difficulty is
a paucity of simple genetic markers. Have you encountered specific antiblesic
actions against thie species, aside from the colicins already recomized?

Your expressed interest in the genetic aspects of variation in respect
to colicins leads me to make the following, perhaps presumptive sugvestion,
which I hope yeu will forgive. A number of the problems discussed in Part 3
of your review would seem te be eepecialip amenable to recombination study,
particularly the question of "mutations associees". The methodology of recor-
bination is net difficult, and can readily be applied to such questions as
the genetic identity (4.e. alleliem) of different mutations, or whether two
“associated” changes are based on a single genetic change. Ao a partial con-
pensation for the aesistance you have promired, I would be delighted to send
you suitable, intercrossable cultures of E. coli K-12, and the very cimple
precedure for their use. Sincerely,

Joshua Lederberg