October 22, 1958

Dr. Mel Cohn

Department of Microbiology
Washington University

St. Louis, Mo.

Dear Mel:

! was really delighted to have a chance to have a good talk with you. Our
interests have converged so much In the past it should not be surprising that
they would make for a parallel approach on questions of antibody formation.
This makes me all the more eager to get going at Stanford and see you at
close hail.

You may be interested to see the enclosed project which | have thermofaxed
from our application to NIH. {| was only waiting to see you before writing up
page 9, and trust this is sufficiently innocuous you can have no objection to
having your workcited in that fashion. | think you will see why # was chuckling
when you read the more fanciful projects.

On reflection, |! am inclined to think that your Immunization procedure may
be the main reason for the discrepancy in our results. For example, by hyperim
munizing to one antigen, you were undoubtedly building up a large cellular popu-
lation (pre-?)adapted to this; then by forcing a maximal response to a second antigen
you may In effect be selected for a second-step ‘mutatéen' in whatever cells were
available. While this may be reasonably fatal to Burnet's'strong' version, it
would not fit too badly with a weaker revision, which allows for further mutation
to establish new specificities. To exclude the chromosomes as the actual sites
of the 'mutatéens' it may be profitable to investigate whether a single cells
can actually make a great many different antibodies: the cross-reaction studies
may be the easiest approach to this. i don't know at what numerical point i would
bive up the chromosomes-- one might get tired of proving the point first, and
among other things the diverse antibodies would have to be demonstrably related
gamma globulins. But if It's not the chromosomes that are hypermutable, there are
tots of microsomes, and 1! look forward to some difficulty in distinguishing be-
tween election among some millions of microsomes per cel! versus Instruction, at
least until an in vitro system is worked out. In some ways, | think the very fact
that cell fragments can make antibody is one of the most exciting prospects, and |
hope we might be able to get Paul Berg Interested in this in a more purposeful
way. Has anybody seriously tried to demonstrate antibody synthesis Jf in disrupted
cell preparations? I assume one should use active, presensitized, secondary cells
at the first trials.

Yours truly,

Joshua Lederberg