November 9, 1949.

Dr. L. Cavalli,

Dept. Genetics,
University of Cambridge,
Cambridge, England.

Dear Dr. Cavalli:

I would like very mch to have E. coli 123, andi would appseciate your
sending 1% to ms. Usanwhile, we have found that another coli strain which
was isolated from chickens, and which showed evidence of recombination
within substrains, also seems to recombine with K-12 mtants. Thus, very
surprisingly, the three sexually competent coli strains so far discovered
are part of the same mating system! Like 123, the new W-1113 differs fron
K-12 in many respects— resistance to T phages, sucrose fermenting... Un-
fortunately, for unexplained reasons, one can select prototrophs occasionally
in a single step from the double mtant I have used. I am sending A transfers
of some W~1113 derivatives under separate cover.

Concerning the mechanism of Hfr, the medium used was Pennassay broth (for—
mila given in papers) in which the pH does go down ultimately. This may well
account for the differences. I have the feeling that a nutritional selection
is not very satisfactory for quantitative studies on recombination owing to
the fact that fusions occur at indeterminate times after plating. I tried an
experiment following Newcombe's Nature paver, to iry to establish the time
at which the recombinant prototrophs begin to proliferate. There was a con-
siderable variance, but respreading the minimal plates after as much as 5
hours gave no augmentation of the number of prototrophs which ultimatsly
cane up. I should reveat the exneriment with longer tins intervals. Rahker
than use the nutritional selection, I would carry out crosses in liquid medium
in which the concentrations of the reacting cells are under some control. I
have recently finished some work on the application of drug resistance to the
selection of recombinants (started about 3 years ago), and finally found strepto-
mycin x azide to be a satiafactory though not an optimal combination. Strepto-
mycin is a very excellent selector, owing to the complete inhibition of sensitives ;
and the very low spontaneous mitation rate to resistance (10-10), Azide is
usable but not completely satisfactory. In an experimdnaat using 58-161/Az x ~
677/sr 93/100 of the double resistants selected were recombinations for fermen-
tative, nutritional or. phage reactions. The remaining 7 were presumably mostly
or all mitations. The azide was used in M/500 concentration. It should be used
in a buffered medium to prevent loss of HN, in acid, or lack of ppnatration
in elkahaiae alkaline media. Streptomycin Wags used at 100u/ml. The advantage of
this system is that mixing the parents in the plate gives very poor yields com
pared to growing them together. I think I have already mentioned an im-
pression that Mg is essential for recombination. Bernard D. Davis tells me that
he has gotten very poor yields by mixing parents on his minimal medium which
contains substantial citrate.
Concerning het, I had been planning for some time to carry out the
same kind of experiments that you probably have in mind. The diploids are
probably too unstable to give very much information on the simple determi-
nation of killimg rates. Also, unfortunately, they seem to carry already a
lethal deficiency which confuses matters. I have hoped, however, the to
compare the rate at which lethal mutations (recessive) are accumulated with
that at which haploid cells are killed, hoping to find some correction for
multinucleate cells. I am sending you W477, a T-L-B)-V,* het stock, however, *also

if you would like to make an attempt at quantitative studies of radiation Lac)-
killing of diploid cells. Lac+ ptototrophs from W-477 x 58-161 should be

picked from EMS lactose, and streaked out on EMB lactose. Suspensions corres-
ponding to Lacy streaks can then be propagated by streaking out on EMS lactose,
etc. After a few initial disappointments, you should have no trouble finding

and maintaining the diploids, if you attend to them. May I assume that our

studies in this direction will not interfere? If you really have a vital interest
in this particular direction, however, I will withdraw. K. C. Atwood, at Columbia
University has been carrying out comparable experiments on Neurospora conidia
(heterokaryotic for a variety of markers) and concluded [according to an abstract
in the Proceedings of the forthcoming Genetics Society meeting] that lethal mtations
are not accumlated as rapidly as killings. so that they account for only part

of radio~lethal effects.

Our cytological comparisons of diploid and haploid cells are progressing
slowly. partly owing to the fact that we have not yet received apochromatic optics
ordered some time ago. However, it seems quite clear that the typica} diploid cell
is quite distinct from the typical haploid, especially insofar as the nuclear
structure can often be resolved into a number of smaller separate? bodies, which
cannot usually be done with the haploids. Since there seems to be only a single
chromosome however, I don't know yet how we will be able to interpret these
bodies,

In summary, I am sending under separate cover the following stocks:

W477 [K-12] T-EOB,- Lacl-nV," het

W-1113  prototrophic, fertile coli from chickens, and its derivatives:
W-1205 threonine, histidine, lactose-, (produces prototrophs; very rough)

W-1189 valine; arginine+glutamic, maltose~

P.S. I don't think that het stocks are especially useful for proving
allelism. If two Lac- factors are so closely linked that crossovers are very
rare, then crosses on Lactose EMS will give occasional spontaneous heterozygous
diploids anyhow. Crosses on EMS Lac are the best for Lac allelism. For phage tests,
I would cross likewise, but test numerous colonies without bothering to purifying
them from contaminating parents, instead making the tests on minimal or EMS medium.

Yours sincerely,

Joshua Lederberg