December 7, 1951

Dr. A. D. Hershey

Dept. Genetics

Carnegie Institution

Cold Spring Harbor, L.I,,N.Y.

Dear Als

Are you acquainted at all with the Vi-phage-~typing diagnosis of ©, typhi?
Having read only the rather unsatisfactory papers of Craigie and Yen
(Can. P. H. J., 29: 448 and 484) I never had a very clear picture about it.
But a student from Reed's laboratory at Queens College, Kingston, Ont., is
spending a year here and gave a seminar about it that raised some very in-
teresting features.

As best as I could tell, the phages for Vi group II are all obtained from
a pure line phage. This phage has a high e.o.p. only for the Vi-type on which
it was propagated, a much lower eop (10-4 ?) for any other type. The residual
plaques found on heterd@ldgous indicators always Bive rise to an adapted stock
which now has a high eop for its host, and a low eop on all the cthers.

At first sight one can argue simply for a complex cf host-range mutations
selected from the phage stocks. The host-range patterns are, however, quite
unique, since activity (at high eop) for all the other pk&agesaative hosts is
lost upon adaptation to any given one. There are a dozen or more distinct
Vi types.

Your work on phenotypic mixing is very relevant to the adaptation, in oy
opinion. It seems th me possible that there is a directed host adantation
which is not a matation, in the same sense as the modification of the pheno-—
types in the output of T2-T4 mixed infections. The residual plaques are not
mutants, but those phage particles (with a random, non-$nhsrited character)
which accidentally succeed in initiating infection on a heterclogous host.
Having done so, the cutput phenotype is directed by the host in which they
have grown. The decisive question as to whether this is a mutation or not
id whether a particular specificity is inherited in a haterclogous host.

On the mutation hypothesis, selection prevents an answer. I do think, however,
that the problem is soluble, either by statistical studies on the numbers

of "mutant" phages in different cultures (a la Luria and Delbruck) especially
if differential absorption can be used to altervthe proportions of cifferent
types, or by the indirect selection (by replica plating) approach.

We are not going to do these experiments, but would be interested to hear
what you think of the idea. The Canadian student conceivably might want to,
but almost certainly could not do as good a job as somecne with more axperience
and insight (namely A.D.H.)

Yours sincerely,

Joshua Lederberg