MASSACHUSETTS INSTITUTE OF TECHNOLOGY
DEPARTMENT OF BIOLOGY 16-417
CAMBRIDGE 39, MASSACHUSETTS

August 1, 1962

Dr. Joshua Lederberg

Dept. of Genetics

Stanford Univ. Medical Center
Palo Alto, Calif.

Dear Dr. Lederberg,

Thank you for your reply to my letter of June 25.

Up to now, I've done two types of work, both of which involved the
control of specific protein synthesis. I am essentially interested in
antibody formation, but feel that the search for adequate models to
describe the control of antibody formation may still be usefully pursued
from genetic and biochemical studies with microorganisms. I am interested
in knowing how control may be exerted on genetic materials in different
states, i. e. integrated or non-integrated, chromosomal or episomal; and
how one genetic entity may interact with others in the same cell. Certain
intriguing puzzles, such as the reason why lac genes within an entering
phage element are not subject to the same control as chromosomal jac, lead
one to realize that gur present molecular models of control are hardly
exhaustive. *

In Dr. Coons' laboratory, using the techniques of immunofluorescence,
tissue culture and serology, I asked the question--can antibody-producing
cells be found in unresponsive mice (whose paralysis had been induced with
repeated injections of antigens starting either in adulthood or at birth)?
The answer was no. This was true for the adult series, to bovine serum
albumin and pneumococcal polysaccharide SII; and for the neonatal series, to
BSA or ovalbumin. The neonatal series was inhibited with daily injections
of as little as 0.25 ye/g, or regularly at 2.5 yg/g and above; whereas the
adults were not inhibited at 2.5 pg/g, but were at 500 yg/g (only two doses
used). No positive cells were visible by fluorescence; no AG-AB complexes
could be detected (Farr assays-in vitro and in vivo addition of I -BSA);
transfer of unresponsive tissue to normal recipients with or without stim-
ulation at transfer led to no antibody response in the recipients. Perhaps
this was all to be expected. Finally, it was shown that the mice recovered
"spontaneously" from their unresponsiveness without further stimulation.
After recovery, they were in a state of heightened reactivity: an injection
of ovalbumin, for example, boosted their antibody level to a point tenfold
-2.

above the normal secondary response level of their littermates, who had
received no repeated ovalbumin injections.

The concept that these data suggest to me is extremely close to AD:
the onset of the antibody response within any cell may be preceded by an
unresponsive first stage, in certain ways analogous to repression of enzyme
synthesis in bacteria. Accepting as possible Al-A4 (it seems to me that
Al is much more likely now that people think the code is degenerate: the
gamma-globulin gene just adds or subtracts a base very readily during repli-
cation), we may add the following suggestions:

Zi--The complete antibody response requires two contacts with antigen.
42--The first of these leads to the production of several stereospecific
peptide chains which become enmeshed in the periphery of the cell

(L-chains of Edelman et al.), converting the cell to a prepared" cell!
Z3--Contact with the complete antigen molecule by the prepared cell
induces a reaction leading to the production of the rest of the
gamma-globulin molecule which presumably could be identical for all
antibodies (H-chains).
Z4--Contact with antigen also stimulates replication and differentiation
down the plasma cell path to a dead end in the plasmacyte (mature )
_which produces its antibody and is then finished. (similar to AT)
25--The first phase of the immime response is inhibited by excess antigen.
Z6--The repressor sensitive site is located along the genome. (A ribosomal
site is unlikely if messenger RNA is a short-lived, continuously
synthesized entity in mammals, unless there is some other way of
shutting off messenger synthesis; also too many inhibitory molecules
of antigen would seem to be necessary, at least one for each ribosome).
47--The unresponsive, tolerant cell when released from inhibition is in
the prepared state as in Z2. (In a previous publication, Coons and
I called this prepared or memory cell a "Y" cell in an X-Y-Z scheme).
Z8--The apparently greater susceptibility of neonatal animals to induction
of paralysis can be explained by special ad hoc hypotheses. (E.g., the
development of an intrinsic adjuvant with age, or the increased
accessibility of neonatal cells to penetration by. antigen.

A further implication of Z7 is that the cell escaping from paralysis
will not produce circulating antibody unless restimulated from without by
antigen. The fact that Avrion Mitchison didn*t find "spontaneous escape"
after paralysis with chick cell antigens is reassuring in this regard,
since cellular antigens (in contrast to protein antigens) would be expected
to disappear rapidly and not be able to serve as a low level external stimulus
to spark plasmacyte differentiation (Zl). Other aspects of this formilation
can be tested. For example, does stimulation actually speed escape? What is
the course of events immediately upon massive antigen injection; maybe by
immunofluorescence or single-cell techniques, pre-paralytic antibody production
would be shown. Can one install paralysis in an immine animal? According
to your published ideas (unless you include the possibility of repression in
mature cells) it should be extremely difficult to install paralysis in an
adult because with the mixed population of cells are some which are stabilized
to proliferate, etc. and which in this way would be difficult to inhibit unless
they recycled through an antigen-sensitive phase. Also, the escape from
paralysis should be studied to see whether the first paratype produced corres-
ponds to the epitype in smallest goncentration. This should also be expected
in the normal antibody response to rather large doses of antigen.
With luigi Gorini, my work was essentially concerned with the action
of small molecules in controlling rates of protein and repressor synthesis.
I learned many of the techniques of bacterial physiology and genetics, and
of continuous culture. We were able to show, we think, that exogenous
arginine has a greater affinity or accessibility for the repressor-forming
machinery than the arginine formed endogenously. In Dr. Luria's lab
during the next year I hope to learn to deal with a different level of
organization, that of macromolecular and supramolecular units (e.g. episomes,
transduced genetic material, etc.)

You asked for my specific interests and qualifications, but Im not
sure that this question has been answered. The impression I gave in paragraph
three was of preferring epigenetics to genetics, microbial systems to
mammalian. Actually, the difference in both these cases appears to be one in
the ease of thinking about, and carrying out experiments: it's easier to
understand Vivaldi than Webern. But the Weberns can be listened to with
enjoyment.

I would appreciate your comments on 25-7, if you have the time.
Sincerely yours,

Eli Serearz