155 Corona Ave
Pelham 65, N. XY.
Friday, April 14

Dear Joshua:

You have prebably been rather eager to wmek hear my further
comments about pleiotropy, and I'm sorry that this is the first
chance I've had to sit down and write to you.

The secondary fermentation affects I mentioned in ny last
letter were in our own mutant strains: tryptophaneless was lace”,
mal- (plus-minus on second day); histidineless was mal~, mannitol
and galactose slow plus-minus; another with unidentified require-
ments was xylose plus-minus, and a cyst&neless £m appeared to be
negative for all fermentations until we found that its metabolism
was somehow interfered with by the EMB medium.(we never solved it
completely). Some were, unfortunately, already lac™, and we didn't
test on other sugars at that time because plates weren't madeup.

Your evidence against pleiotrop/ dg is convincing in the case
of maltose, and if maltose can behave that way, so can the other
sugars. The pleiotropygfi idea would have explained most of the
aberrent behavior of sugarz=x markers insofar as there was an excess
of megatives, because there might have been two or more mutations
which singly, or in combination with other mutations, could prevent

a eteerfermenteationg. The great excess of negatives among all my
recombinants would 2xzk% be nicely explained by this, and it would
seem also to account for the mmequak segregation bias you told me
of this summer, It could account for some degree of “partial
segregation" because a Lac v Mal- might be sezregating for two
different Mal- mutations. To account for more than 75% of Mal-
you'd have to assume 3 or more loci, or else postulate some sort
of repulsion between the two immxx mutations. If Lac v Mal+@0pem ee cece
gemmem-~ tiocugh, pleiotropy could not explain it.

worth considering, I think. He compares it with the mucoid charac-
ter, which is very common £m among recombinants and among some
mutants, especially on selected media. He thinks the mucoid trait
depends on a ohysiological state that can be achie¥ed by many dif-
ferent factors,’ both genetic and environmental, mmixtfx Some of
the non-fermenters, e.g. grlactose in W-677, actually utilize the
sugar very effectively without producing enough acid to be stained,
or so his experiments seem to indicate at this stage, and the
failure to stain might maskix conceivably have a diversity of
causes like the mucoid character, The only difference, between tifis
and pleiotropy, I think, is that it could depend on“ genetic
differences wtween the parent strains, instead of on the known
mutations.

re Bernié's suggestion of physiological side-effects is also

i

For our quadruple stocks we are going to neglect fermentations,
since they would so limit the mumber of mutations we could use,
and since we can probably get plenty of drug-resistant and phage-
resistant markers, in addition to the requirements, to mark the
best-understood part of the chromosome. The crosses I'm doing
now involving the descendants of W-677 should give enough informa-
 

 

April 14 -- 2

tion on the fermentations, that is if they take a turn for the
bétter. Of 6 B tines I have done the experiment, only 2 the

2nd and 3rd yielded complementaries., The main troubhe seems

to be incompleteness of the requirement for By. I have there-
fore decided to take your suggestion on selecting the comple-
mentaries by drug resistance. if I get the Ver mutation to-
morrow, then next week I'll repeat the experiment with the cross

W1234 IaP by (B-)M-P-S' ver
selecting for Az’s™ complementaries.

Of the stocks you sent us, W760 wes the most desirable be-
cause it did not duplicate the requirements of any of the doubles
we were using. But the vial contained no growbh; we tried incu-
bating it further and even washing it out, without success. Why
don't you use slants instead of stabs? I can't see that it makes
very much difference, but Bernie's technicians get quite angry
about it. By the time you could send us another sample of W760
now we'll certainly be working with triples.

You asked if I used dilute inocula for making the crpsses.
In the most successful experiment many of the colonies were picked
from a plate with only 66 colonies on it, and it was my aim to
get the numbers even lowerethah that. However, I have been using
different stocks since then, and, perhaps because of different
densities in the cultures, they seem to pequire different size
inocula to give the same number of recombinants. The result is
that my crosses since that last successful one have all given
too few or two many prototrophs,

The short-cut methods we use for testing markers include a
lot of details that it would be silly to list. But the main points
are the following:

1. Eliminate all handling of tubes.

Zz. a. Bick two colahies at a time into a double-depres-
Sion slide freshly flamed, containing a drop of
water in each depression.

Dd. test for aumgotrophy on plates, each plate of
the series omiting one of the growth factors.

2. Spot the cultures on the plates instead of streaking

them.

3. Mark (preferably permanently) a grid on the back of

each plate, in vertical rows of ten with a heavy line
across the middle, so that the spots can be put on
without looking to see where the last one mma was, and
so that the number of each will ke correspond to its
position on the plate. I spot 60 on one set of plates,
breaking up the two end rows into 4 and 6 (first 5 above,
second five below the heavy line), but if you can always
get plenty ef plates 50 or even 40 might be more effi-
cient.

I mark the plates and lay them out in order, and then spot two
colonies at a time on the whole set (one colony all the way thru,
then the other, just moving my loop back and forth between the
slia€éé ané the plates), after everg two I dip the slide in alco-
hol, then water, then wipe it and flame it, hl if you can get

abt 50 difacein bakes they cold hl be toe toned 4 drind Getler, ford poser.