September 26, 1955

Dr. Ellsworth C. Dougherty
Department of Physiology
University of California
Berkeley 4, California

Dear Dr, Dougherty:
Thank you for your reprints and your letter of the 21st.

The experiment that you propose has been considered, and, to a degree,
attempted. There are some difficulties both in executio: and interpretation,
but the results do suggest the occurrence of intracellular recombination.

I do not believe that there are any pairs of auxotrophic mutants for
which the experiment would be feasible, owing to the relatively low rates
of induced reversions. Mutants that were inherently revertible enough to
be useful would probably never have been characterized as auxotrophs in
the first place. If they were available, however, they would be best handled
by using replica plating (rather than relying unon syntrophic mainéenance)
along the lines of the following experiment:

Both Witkin (Cold Spr. Harbor Symp.) 1951, p. 357) and Newcombe (1952,
Genetics, 383134) have studied UV-induced mtations from Lact to Lac-, and

from v,° to V,"» and their coincidence, in E. coli strain B. Their technique

was to select for V," mtant clones, by means of phage Tl, on an FSB lactose
agar on which coincidental Lac- mtants could also be detected. Both authors
found an appreciable fraction of vi clones that contained both Lac+ and Lac-

components, and the question is how to explain them. Neither ‘Jitkin nor Newcombe
appear to have favored the suggestion (see the CSH paper discussion, p.372)

that they may arise by recombination between Lac+¥}* and Lac-V 8 nuclei arising
by mutation in an originally Lac+¥78 cell, or by the other possible "cross"

of a double mitant by wild type. If this possibility is barred, one probably
has to accept some sort of delayed or fractional effect of the UV, so that it
would be a matter of some interest to settle it. The techniqyue that was used
precluded the isolation of v8 components in these clones: if it could be shown
that many of the clones already known to have Lac+V,™ and Lac-V," also contained

Lae-V, 9 individuals, the suggestion of recombination would be verified. Actually,
it should not be too difficult to look for such cells by replica plating: UV-
treated bacteria would be plated to give about 1000 survivors per plate, and the
colonies replicated (while still small) to FMB lactose+ phage. If any Lac-V r
show upon the replicas, Lac-V,° can be rather easily looked for in inocula from
the homologous sites on the ortginal. The same design can be used, of course, for
other markers. If the coincidences were too infrequent, the experiment could
still be done, using larger numbers of the surviving bacteria, but the definition
of single clones would be less reliable,

When this work was first dene, we all believed that F. coli B was an
infertile strain; it has since been found that B is an F- strain, that can
be crossed with other F+ strains, or made F+ (by contagion) and self—comp&tible.

I must add that we have had other tests for recombination in transitory hetere
karyons (e.g., immediately after crossing) and that the results, generally, do
not suport its occurrence; it would not surprise me at ail if uv treatment had
something to do with it.

Unfortunatelg, both Witkin and Newcombe are working on new problems, and this
particular problem is still left hanging. Could you be interested in going into
it? Technically it ought not bo be emnecially difficult or tedious, but would of
course take some time. If you wanted to visit us, say for a month, that oughtvto
wrap it up-- or if I could be af any assistance to your doing it at Berkkby,
please let m know.

It is easy to at least guess what you're driving at in this matter- whether
this process represents a stage in the evolution of sexuality. But as the B story
lilustrates, it will be impossible to decide whether intragsallular recombimation
developed first, or whether strains that show this tc the exclusion of intercellular
sexuality are simply degenerate or defective in their ability to mate. You mst
have faced the same problem in evaluating Pontecorvo's "parasexual" recombination
in fungi.

,(By the way, I fear that Pontecorvo has muddied the field, and the impact of
his own brilliant studies, by his terminology of "parasexuality". If it were pos-
sible togstart again, I would have thought "parasexual' to be an excellent term
to describe the external paraphernalia of sexuality, reserving this for the funda-
mental process of karyogamy. Then "aparasexual" would describerbuuktatimxcresomix
'vegetative” fusion and recombination, etc. 43 matters stand, sexual vs. parasexual
is an artificial cleavage, separating some processes of karyogamy on the one hand,
from other karyogamies plus ammassortment of quite different mechanisms of gem tic
recombination on the other. I would have thought that any logical classification
would have separated karyogamy as a unique category. What do you think about this?
Can anything practical be done’? How did you handle it in your review— or would
a terminological appendix be out of the question? There are still a lot of prefixes
not yet preemoted, but I am sorry about “para".)

Yours sincerely,

Joshua Lederberg
Professor of Genetics