PIOUS) PRENDS Te OTCA: RUS cARCHY [DISCUSSION APSR BURST

Jot, Miticbore

MEDAWAR: In opening the discussion of Burnet's characteristically
fascinating paper, I shall say nothing about his avowedly speculative
referees to the poasible role of somatic mutation in ageing or in the
inception of tumours. As Burnet implies, it is antibody-formation that

is likely to b the testing ground for theories of somatic mtation, 60
what I should like to do is simply to explain (as I see it) the train of
thought which has led to the formulation of the theory of somatic mutation
in this particular context. My indebtedness to Lederberg and Monod will
become very clear in the course of the argument.

Consider a cell which, as a result of some stimulus impinging upon it
from the outside, has come to indulge in some new synthetic activity. ‘The
relation between atimulus ap Yesponse may be of several different kinds.
When a cell is infected with virus (or, better, ss Schramm has just
explained, with virus RNA), or when a pneumococcus is “infected” with
exogenous DNA, the stimulant iteelf provides the exact instructions
in accordanée with which the cell carries out its new synthetic activity.
The relationship between stimulus and response may therefore, following
Lederberg, be described as “instructive". That is one possible kind of
relationship. But when the stimulus is an enzymic substrate and the
responding cell, a bacterium, comes to manufacture a so-called "adaptive
enzyme"; or when the stimulus is an embryonic induger and, ae a remit of
ite action, a hitherto uncommitted host cell follows one pathway of
differentiation rather than another; ——— in such cases as these it is
most unlikely that the relationship between stimulus and response is
instructive. All that the etimulue seems to do is to call forth or bring
out a potentiality latent in the responding cell. Lederberg describes such
a stimulus as "elective"; Waddington has used the term “evocative"™ in essenti-~-
ally the same sense.

How are we to classify the relationship between stimulus and reaponse
when the stimulant is an antigen and the new synthetic activity is the
manufacture of a specific antibody? Is it instructive or elective? Pauling
at one time suggested that gamma globulin acquired its specific complementary
pattern under the direct impress of antigen, and this would be classified as

an inatructive theory of antibody formation. But modern opinion is herdening
ue

in favour of an elective theory, and if the history of the theory of
adaptive enzyme formation in bacteria is anything to go by, we should be very
rash to dismise it.
But now the problem arises: can a single vertebrate lymphoid cell
contain enough genetic information to underwrite the formation of any one
of the almost prodigious variety of antibodies which we know a vertebrate
animal can produce? ‘fy feeling still is that the ahswer is Yes; the sygote,
after all, presumably contains within itself the far greater store of genetic
information that is needed to subsidize the development of an adult organiem
of multitudinous complexity—<not forgetting all that is entailed by the
inheritance of differences of behavioural pattern. But Burnet, Lederberg and
Monod (who discussed the problem at a very recent meeting organized by the
Ciba Foundation in Paris) are inclined to think that the answer is No. If
the true answer is No, and if antibody formation is indeed an elective process,
then we must suppose that new genetic information arises within the lineage
of celle that descends from the zygote. But this is equivalent to adopting a
somatic mutation theory, fer mtation is by definition the proceas by which
new genetio information arises. Burnet sees no reason at present to regard
these mutataions as other than genio, and he suggests that the body contains
genetically distinct clones of lympioid cells each with a single and distinct
immunolozical capability. It was this particular variant of the somatic
mutat¢ion or clonal theory that led to Lederberg and XNosaal's test of the
"one cellione antigens one antibody” hypothesis. But, as Monod has pointed
out, this is by no means the only possible form of the somatic mutateion
theory. The matataxons mig-t be "ribosomal": we are not obliged to believe
in a one celleone antibody relationship: all we are obliged to believe in
is that the genetic information intecrated over the lymphoid population of
the body as 4 whole is greater than that which was originally present in
the zygote. This does not logically entail a one celleone antibody relationship
The point I am trying to make is that the key to the argument is
whether or not we suppose that one lymp'ioid cell can contain enough information
to underwrite the entire repertoire of antibody formation. If the answer is
yes, the somatic mutation theory is supererogatory; if it is no,then the
aoceptance of a somatic mata tpion theory in one form or another, not necessaril

Burnet's, is logically entailed.