August 2, 1989.

Dr. Le Cavalli,
Dept. Genetics,
U. Cambridge, England.

Dear Cavalli:

I am sending this note in company with some cultures derived mostly from
your 5@~-161Hfr (my W-1033) which I hope will reach you in time to be of use
directly on your return to Cambridge.

The cultures include:

W-1059 Lac, - { a UV mutant from W-1033]
W-1073  Maly-{Lacy-) { a UV mtant from 81059]

NmB1L4 Lac+ Mal,~ etc..... | a Lact reversion from ¥-677 ]

<These cultures are especially useful in demonstrating recombination in
complete medium. For example, after 3-4 days of growth togethsr in complete
liquid medium, W-1059 + W-814 was plated out on EMB-'Maltose + Lactose = Malac]
Ag many as 20% or more of the colonies may be Malac-, and show various
new combinations of the other markers! I cannot say to what extent selection
may distort the true amount of recombination, but it must be very high indeed!
Comparable experinents using Y~87 { also Belin Lacy —) instead of W-1059 gave
no Malac- at all, so that there is little question now of the uniqueness
of your Hfr stock,

It is also possible to plate out somewhat younger mixtures of marked
stocks, e.g. W-1073 + W-909, on « single sugar medium, like FMB Lac, and to
find mosaic colonies which might represent segregating zygotes. These may
constitute 1 - 2 of the mixed cultures, although some of the mosaic colonte
may, of course, represent accidental conjunctions of independent colonies.
The mosaics are, as one might expect, much simpler in structure than those
from complex heterozygotes, and usually consist of large, nearly semicircr
sectors of + and ~ with a common radius in a single colony. Sometimes, on
of the components is but a small wedge, or even a central pdece which do
extend to the periphery. It 1s very likely that the segregating diploid
be picked out with some precision after a little practice, because mos’
them contain a recognizable recombinant.

The segregation data are very peculiar, but support my previous s
that the Mal locus in particular is not segregated at random, but »
partly eliminated.
For example, 30 Lac V colnmnies were taken from W-1073 x W-909. 28 of these
consisted of one parenbal combination, LeMG-, and the one recombination, L-M+G 1]
One colony contained only the two parentals, and may have been fortuitous. One
contained both parentals as well as the same recombinant, L-M+G-. Here again,

the second parent might have been fortuitous.

I have gotten quite compezable results with other crosses between L033 derivatives
and marked stocks from Y-10. In all cases, the pattern points to the elimination

of a chromosome or segment from the BM (Hfr) parent, which is in agreement with

the patterns of the persistent heterozygote. It seems unlikely that an entire chromo-
some is eliminated because of the evidence of linkage to a lethal deficiciency

in the Het stocks. Conceivably, however, this 1s unrelated, and we are dealing with
something compurable to Auerbach's unstable centromere in Drosophila (Genetics,

Jan. '47

My experience is a little ambiguous in regard to the oppositional character of
Hfr. I have been developing the mutants from 1033 in order to use as closely related
stocks as possible in experiments on the source of the biased elimination. So far,

I have not found recombinants in W-1033 x W-1073, but believe to have gotten a few

in W-1084 (a Lac+ reversion from W-1073) x W-1059. But these experiments are still
incomplete.

The success of these experiments bodes well, I think, for your forthcoming cytological
studies. If, in suitable crosses, the zygote frequency reaches 1% it may even be feasible
to look in wjorkkakimcumstiom liquid complete medium for the fusion process. However,
the 1% figure may represent a delay in "germination" of the zygotes which accumulates
them somewhat. As I my have mentioned, I have a student coming next month to work on
the cytology of the persistent heterozygotes, and the work that we are doing should fit
together fery well. I have hopes that we may learn Wow to recognize the zygotes cyto-
logically from cur studies, which in turn may help to verify your descriptions.

Sincerely,

Joshua Lederberg

fs. if ge 4