y-82B

 

CIBA FOUNDATION SYMPOSIUM

ON

CELLULAR ASPECTS
OF IMMUNITY

Editors for the Ciba Foundation

G. E. W. WOLSTENHOLME, O.B.E., M.A., M.B., M.R.C.P.
and

MAEVE O’CONNOR, B.A.

With 117 Illustrations

 

LONDON

J. & A. CHURCHILL Lr.

104 GLOUCESTER PLACE, W.1
1960
42 Discussion

is also phagocytosis of inert particles. We know that these inert
particles are probably not influenced at all by the presence of com-
plement, so there may be two kinds of process: a fundamental
process of phagocytosis similar to this of inert particles where
complement is not involved, and an immunological process where
complement is certainly implicated.

Robineauz: I tried to demonstrate some morphologically different
processes of phagocytosis. I cannot say whether they are funda-
mentally different, that is to say whether each of them involves a
different mechanism at a molecular level or not. I think, as you do,
that we could give all these phenomena the name of phagocytosis.

As for the second point, it is generally admitted that complement
is not necessary for phagocytosis to occur; however, it enhances the
phagocytosis of particles which have previously fixed a specific anti-
body. The phagocytosis of inert particles does not require comple-
ment but it is known that some of the particles adsorb proteins
which can thus modify their surface properties so that they are made
adherent.

Humphrey: The liver seems to be able to distinguish quite sharply
between native and thermally altered proteins. We found evidence,
both in livers perfused in vitro and in living rats, that there must be
more than one mechanism involved in the uptake and subsequent
catabolism of native and denatured material (I‘reeman, T., Gordon,
A. H., and Humphrey, J. H. [1958]. Brit. J. exp. Path., 39, 459),

Dr. Robineaux, did you see labelled material in the nucleus of the
cells, or was it entirely outside the nucleus?

Robineaux: We never observed fluorescent material in the nuclei.
Labelled material has been seen localized in the Golgi zone, in the
same place as vacuoles are localized in the process of pinocytosis.

Smith: What information have you about the differentiation cells
may make between homologous and heterologous proteins, or
between degraded products of homologous and autologous proteins?

Robineauz: It is possible that there is no difference in the uptake
of homologous and heterologous proteins and that pinocytosis is
simply a physiological function of the cell.

Lederberg: Can you give any quantitative estimate of the rate of
pinocytosis ?

Robineauz : Holter has studied this in the amoeba and Lam hoping
to undertake similar experiments. But so far I am unable to answer
you.

Lederberg: Have you any morphological evidence of the converse
of pinocytosis, a mechanism for ejecting large volumes of fluid,
functionally comparable to a contractile vacuole ?

Robineaux: It seems that the cell takes up water, but there must
Discussion 43

be some mechanism of excretion, otherwise it would swell consider-
ably. There is no evidence of this mechanism of excretion, but by
means of interference microscopy it is possible to demonstrate a pro-
cess of concentration in the vacuoles. When a vacuole disappears, it
proves that the refraction index of its contents has become the same
as that of the entire cytoplasm. For this to occur, water has to go
out of the vacuole. We show in our paper some evidence of this
phenomenon.

Dixon: Are the studies on homologous and heterologous proteins
both done with labelled protein ?

Robineaux: No. I intend to study the penetration into these cells
of homologous and heterologous proteins labelled with different sub-
stances—for example with sulphofluorescein which gives yellow-
green fluorescence, and sulphorhodamine which gives orange
fluorescence—in order to find whether there is a competition between
the two types of protein or not.

Voisin: This question of penetration of autologous, homologous
and heterologous proteins is very important from the theoretical
point of view, but since we cannot see the proteins without labelling
them—that is without denaturing them to some extent—it seems
very likely that there will not be any difference in the morphological
studies between autologous, homologous and heterologous proteins.
This therefore remains a theoretical consideration of great interest
but one which cannot be solved by this method. The best approach
to this question would be to study the penetration of unlabelled
autologous, homologous and heterologous proteins by the Coons
technique, utilizing specific fluorescent antibodies to the phago-
cytized proteins.

Lederberg: Dr. Robineaux, how could pinocytosis, which involves
a mass of fluid, discriminate between autologous and heterologous
solutes ? One could safely assume that there is at least one molecule
of autologous protein and at least one molecule of heterologous pro-
tein in every such droplet.

Robineaux: It is possible, as has been shown in amoeba, that pro-
teins get fixed on the cell surface so that they induce the formation of
pinocytosis droplets; so far, this has not been demonstrated with
macrophages. In amoeba, glucose can penetrate into the cell by
pinocytosis only if proteins are present in the medium. Brandt
recently showed that these proteins were first fixed on the surface of
amoeba, then they penetrated into the cell by pinocytosis, and
finally they were laid out on the vacuole wall. There might be a
selective pinocytosis process in as far as there might be a selective
adsorption onto the cellular surface of proteins which induce the
pinocytosis. This would be an interesting problem to investigate.
EnzyMes AND PROTEIN ANTIGEN METABOLISM 57

LAPRESLE, C., and Suizewicz, P. (1958). Bull. Soc. Chim. biol. (Paris),
40, 1085.

Rows, P. (1925). J. exp. Med., 41, 399.

Soper, H. A., and Perrerson, E. A. (1958). Fed. Proc., 17, 1116.

TALLAN, H. H., Jones, M. E., and Fruton, J. S. (1952). J. biol. Chem.,
194, 793.

WEICHUSELBAUM, T. E. (1946). Amer. J. clin. Path., 10, 40.

- WitistaTrer, R., and BaMann, E. (1929). Hoppe-Seylers Z. physiol.

Chem., 180, 127.

DISCUSSION

Kunkel: Have you tried to separate the antibodies concerned in
the different antigen fractions?

Lapresle: Yes, but so far I have not succeeded. I tried by electro-
phoresis on a column of agar gel but the mobilities were so close that
it was not possible to separate them. We have also tried by chroma-
tography on a cellulose column. It might be possible to separate
them in this way but there was not enough antibody to be detected,
so we shall have to try again.

Fagraeus: We did some transfer experiments, in. Prof. Grabar’s
department, with spleen cells from rabbits injected with antigen from
6 to 48 hours before the transfer to fresh rabbits, and they did not
keep the antigenicity for very long. Have you injected these pro-
ducts into animals and then taken the spleen out after a certain time
to see if you could find the same products as you found in vitro ?

Lapresle: I have not done that.

Lederberg: Do you think that an enzyme which operates at pH
3-5 should have some physiological significance in norma] meta-
bolism; or is it involved in the solubilization of proteins in necrotic
foci where the pH can drop very low ?

Lapresle: This pH does not conflict with a possible physiological
activity, since Rous showed (1925, loc. cit.) that when the phagocytic
cells are active their intracellular pH is about 8-5 (see text).

Humphrey: In the whole spleen does the degradation of the
albumin stop at the stage you described, or is most of it broken down
much further ?

Lapresle: I do not know, but when the degradation is pursued
further you get products which do not precipitate but which retain
the ability to combine with the antibodies.

Medawar: Do these enzymes occur in blood leucocytes, and if so,
do they occur in the polymorph or lymphocyte fraction or in both ?

Lapresle: They were demonstrated in leucocytes by Barnes (1940,
loc. cit.) in a lymphocyte fraction as well as in a polymorph fraction.
58 DIscussiIon

Robineaux: Could it be that there is a correlation between the
enzymic activity and the fixation of the protein on the mitochondria?

Lapresle: I do not know about that.

Halpern: What is the molecular weight of the sub-products ?

Lapresle: I do not know. I studied this with Slizewicz (1958, loc.
cit.) and we found a family of products which were not homogeneous
and which had a mean sedimentation coefficient less than that of
the original molecule.

Lederberg: The apparent heterogeneity of your enzyme may be
due to its complex formation with other proteins as quasi-substrates.
At pH 6 or 7 your protease should be inactive but perhaps still
capable of making such complexes with other components in the
extracts.

Harris: Have you tried to use these enzymes at pH 7?

Lapresle: Yes, but it does not work.

Miles: Does this enzymic extract attack the rabbit albumin as
well ?

Lapresle: I have not tried that yet.

Humphrey: I think you know, Dr. Lapresle, that Dr. E. M. Press
is working with Dr. R. R. Porter on spleen enzymes which are per-
haps the same as those whose action you have studied. Their pH
optimum is 3-5, when tested on haemoglobin as substrate, though of
course haemoglobin is broken down much further. The interesting
thing is that in beef spleens they find quite a large number of well-
defined active peaks by ion-exchange chromatography, all with
apparently identical enzymic activity. Each spleen showed a dif-
ferent distribution pattern for these enzymes, and the only two
which were indistinguishable came from identical twin cattle. I do
not think they would put too much emphasis on this, but it may be
significant.

Grabar : I discussed that with Dr. Porter very recently. In general
he has used a stepwise increase or modification of the pH and ionic
strength and each time he got a new peak. Thus he made the same
observation as Dr. Lapresle, but when Drs. Lapresle and Webb had a
gradual modification of their buffer there were less pronounced peaks.
So it is possible that for each of these substances there is stepwise
liberation. It is also possible, as Prof. Lederberg pointed out, that
in every extract the enzyme may be more or less in combination
with one of the other proteins present in the tissue. So when we
undertake fractionation of such an extract we may also have some
of the activity bound in some way with other proteins present in the
extract; this would explain this heterogeneity. Dr. Lapresle wanted
only the degradation of serum albumin, and therefore it was not of
primary importance for him to study this particular case.
118 DIscussIon

extracts; in the extracts both y-globulins and many specific anti-
bodies were found. On the other hand lymphocytes were extracted
from the thoracic duct and, although y-globulin was found in the
extracts, it was not possible to demonstrate that the y-globulin con-
tained specifically anti-ovalbumin antibodies.

Seligmann: Has anyone seen a transfer of soluble material coming
from any other cell to a plasmocyte or a precursor of a plasmocyte ?

Thiéry: I have not observed a direct transfer.

Seligmann: In connexion with what Dr. Bernhard said, I may add
that five years ago in our work on leucocytic antigens with Prof.
Grabar and Jean Bernard, we controlled by immunochemical means
the washing of the different types of cells which we took from human
blood. In the extracts of normal or leukaemic well-washed lympho-
cytes isolated from circulating blood we were never able to detect
any substance having the immunological specificity of serum +-glo-
bulins. So it seems difficult to think that these circulating lympho-
cytes, normal or leukaemic, could produce any globulin having the
antigenic structure of y-globulins.

Lederberg: Is anything known of the origin of the ergastoplasmic
vesicles and lamellae ? Could they be remnants of membranes which
are infolded during pinocytosis—by analogy with the origin of the
myelin sheath from the Schwann cell membrane ?

Bernhard: Several authors, among them Palade, admit that in
macrophages and perhaps also in other cells this phenomenon of
invagination does exist. Membranes coming from outside, which are
originally cellular membranes, can afterwards become ergasto-
plasmic membranes. Personally, I believe that this phenomenon
exists, but it would be erroneous to admit that all intracytoplasmic
membranes are of external origin. What we need in all cases are
lipoprotein layers which may secondarily be transformed into a
specific support for RNP granules.

Simonsen: I think you rightly stressed the impossibility of deciding,
by morphological methods alone, whether there is any such thing as
a transition between the plasmocytic and lymphocytic series. Dr.
M. Holub reported some findings which I think are very pertinent
(1960, In Mechanisms of Antibody Formation, Proceedings of
Symposium, Prague). He put thoracic duct lymph into a diffusion
chamber which was then implanted intraperitoneally into newborn
rabbits. He added to these lymphocytes a heterologous antigen and
found not only that antibody formation occurred but also that,
when the cellular content of the tissue chamber was examined
histologically, there was very slight mitotic activity but strong
evidence of development of pyroninophilic cells. He concluded that
most of these had developed from the lymphocytes. Could these
182 DIscussIOoN

in the spleen and also in the thymus, which I believe you examined
too. I thought I could see in the microscope that the centre of this
necrosis very often started as an accumulation of polymorphs, which
I suppose to have been derived from the host, and that the reaction
around it was perhaps rather the reaction of the grafted cells, which
is exactly the opposite of your interpretation. Irrespective of which
interpretation is right, I think we both agree that what brings about
this necrosis, as well as the spleen enlargement, is the graft versus
host reaction. My answer to your final remark lies in the experiment
I described in my paper, in which grafting from adult F, hybrids to
newborns of the parental strain failed to produce splenomegaly; I
think it should have done so if your interpretation was true.

Burnet: I was concerned rather with the continual passage from
one chicken to another. If Isaacson’s interpretation was correct, the
possibility would be that your immunologically competent cells in
the next passage were derived from the recipient rather than from
the donor cells which had multiplied in the recipient.

Medawar: Some of Loutit’s experiments on radiation chimeras
clarify an essentially analogous situation. The cell which is respons-
ible for secondary radiation sickness (which is generally interpreted
as a graft versus host reaction) can also be serially passaged, and
there is no ambiguity about the provenance of the cell concerned
because it is identified by Ford’s chromosomal marker. So I do not
think we need doubt Dr. Simonsen’s prima facie belief that he is
serially propagating the descendants of the cell which he originally
inoculated into the first embryo.

Gorer: By putting an antigenic tag on the different cells one can
show fairly easily, by a cytotoxic test, which sort of cell is being
propagated, at least in mice; and there again it is a donor cell.

Mitchison: Another unambiguous reaction takes place when cells
of adult turkey origin are transplanted into chick embryos. The
cells react against their host, and can continue to do so in serial
passage. The cells of the graft are certainly responsible for the reac-
tion, since turkey globulin can be detected on the surface of the host
erythrocytes.

. Lederberg: You indicated that a dose as small as 30,000 lympho-
cytes could give selective proliferation. Were those lymphocytes
from unsensitized adult donors ?

Simonsen: Yes.

Lederberg: If you use sensitized donors or passage material, will
smaller numbers of cells then set off the reaction ? And what would
the limit be in that case ?

Simonsen: I have not pinned that down in chickens yet, but there
is strong evidence on this point from experiments on mice. If F,
THEORIES OF IMMUNOLOGICAL TOLERANCE 141

are summarized in Table J. They appear to divide the un-
responsive states into two clearly distinct classes: the “‘toler-
ant”’ states, states of essential non-reactivity, in which there
has been a central failure of the immunological response,
namely: (A), (B), and (C); and the states represented by (D)
and (E) in which there is some reason to believe that im-
munologicalactivity has been thwarted rather than suppressed.
However, we are not yet in a position to classify (D) and (E)
with certainty. The case for separating (D) from (E) seems to
rest mainly upon the interpretation of the “‘non-immune”
rate of decay of antigen in protein-overloading paralysis, but
if it turned out that some of this antigen was in the form of a
circulating antigen-antibody complex the distinction of
principle would disappear.

Possible Theories of Immunological
Tolerance ,

In this section an attempt will be made to answer the
question: what kinds of theories of immunological tolerance
is it now possible to devise? The answer clearly depends upon
the choice we make between various alternative possibilities
that are at present entirely open. To begin with, however, we
shall make two assumptions, (A) and (B).

(A) The maintenance of the tolerant state depends upon the
persistence of antigen: tolerance disappears when, or shortly,
after, antigen disappears. The evidence that justifies this ©
assumption has been given above. For reasons that will be
given later, it is important to notice that the “tolerance” it
refers to is tolerance conceived as a property of the whole ©
organism and not as a property of a cell or cell-lineage.

(B) Any one antibody-forming cell (and its derived lineage)
-responds to only one antigen at any one time. In other words the
antibody response is singular and not plural—an idea that
seems to have arisen out of the beautiful work of A. H. Coons
(see Coons, 1958) and which is supported, or at all events not
contradicted, by the work of Nossal and Lederberg (1958) and
142 P. B. MeEpsawar

White (1958). (More recent evidence of E. S. Lennox and M.
Cohn suggests that this is not the invariable rule, but I shall
make the assumption nevertheless, if only to emphasize that
its acceptance does not oblige us to believe in the “‘ predesti-
nation” theory that forms the subject of the first of the three
pairs of alternative possibilities discussed below.)

Fortified by these two assumptions, one of them dubious,
we may now turn to consider three sets of alternative possi-
bilities within and between which a firm choice must be made
before there can be any question of propounding a theory of
tolerance.

Each antibody-forming cell and its derived lineage responds to
only one antigen...

(C1) ...and this is the only antigen to which it could have
responded (“‘predestination”’ theory).

(C2) ... but tt could have responded to a different. antigen if
that other antigen had engaged it first (“‘ pre-emption” theory).

Alternative C1 amounts to the declaration that antibody-
forming cells are a heterogeneous population, the individual
members of which are in some way predetermined, preadapted
or predisposed to react upon some one antigen; its conceptual
history has been traced back to Ehrlich, though we associate
it more concretely with the names of Jerne (1955), Talmage
(1957), Burnet (1957, 1959), and Lederberg (1958). It pro-
vides an a priori explanation of the “‘one cell: one antibody”
relationship (and, conversely, must be false, or in need of
drastic modification, if that relationship does not hold good);
it leads to a consistent theory of immunological tolerance (see
below), and explains the fact that an organism is not swamped
by one antigen to the prejudice of its ability to react upon any
other. But how do individual antibody-forming cells come to
be allotted their particular capacities to react ? Burnet (1959)
now believes that pre-adapted variants arise as ‘‘part of the
general process of differentiation during embryonic and post-
natal development”. But if that is so, why should we not
regard exposure to antigen itself as the specific inductive
stimulus which commits an (until then) pluripotent cell to a
144 P. B. MEDAWAR

perhaps as a result of hypersensitivity to antigen at some
stage of its maturation. This possibility has been suggested
by Burnet (1957) and by Lederberg (1958).

(E2) The inception of tolerance involves some change in the
immature antibody-forming cell other than its death and elimina-
tion. According to hypothesis E1 there is no such thing as a
tolerant cell: tolerance is a state that can be enjoyed only by a
whole organism. It was Woodruff (1959) who first called our
attention to the fact that no-one has yet demonstrated the
existence of a tolerant cell. Much evidence superficially sug-
gests that tolerant cells do exist, but on closer enquiry it
turns out to be impossible, at present, to distinguish between
the presence of a tolerant cell and the absence of a non-
tolerant cell.

Before discussing the theories of tolerance that might be
arrived at by making various choices between the alternatives
set out above, it should be pointed out that only a limited
number of combinations is in fact possible. If we accept El
(that the inception of tolerance involves the destruction of a
cell) then we must also accept C1 (the predestination hypothe-
sis), for otherwise all the future antibody-forming cells of an
organism would be killed when a foetus is exposed to antigen.
Again, the choice of the combination E1, Cl and A makes it
virtually obligatory to accept D2; El, C1 and A jointly de-
clare that persistence of antigen is needed to maintain a state
of tolerance that has been brought about by the destruction
of certain specifically preadapted immature antibody-forming
cells; but why should there be any need for continuous ex-
posure to antigen unless to subvert the newly-differentiating
stem cells envisaged by alternative D2? This, as I understand
it, is the argument of Lederberg (1958). It should also be
noted that if D2 and E2 are true—i.e. if antibody-forming
cells mature from a quasi-embryonic reservoir throughout life,
and if the inception of tolerance does not involve their destruc-
tion—then we simply do not know whether persistence of
antigen (A) is needed to maintain tolerance in cells and cell
lineages which have been made tolerant at some earlier stage
THEORIES OF IMMUNOLOGICAL TOLERANCE 149

FELTON, L. D., KAUFFMANN, G., Prescort, B., and OTTINGER, B. (1955).
J. Immunol., 74, 17.

Humeurey, J. H. (1956). Proc. roy. Soc. B, 146, 34.

JERNE, N. K. (1955). Proc. nat. Acad. Sci. (Wash.), 41, 849.

LEDERBERG, J. (1958). J. cell. comp. Physiol., 52, suppl. 1, 383.

LENNOX, E. S., and Coun, M. Unpublished data.

LoutitT, J. F’. (1956). Jn Lectures on the Scientific Basis of Medicine, 5,
439. London: Athlone Press.

Main, J. M., and PREnN, R. T. (1955). J. nat. Cancer Inst., 15, 1023.

MeEpawakg, P. B. (1956). Proc. roy. Soc. B, 146, 1.

Mircuison, N. A. (1959). Jn Biological Problems of Grafting, a Sym-
posium, p. 239, ed. Albert, F., and Medawar, P. B. Oxford: Black-
well.

Monon, J. (1959). Jn Cellular and Humoral Aspects of the Hyper-
sensitive States, p. 628, ed. Lawrence, H. S. New York: Hoeber-
Harper.

Nossa, G. J. V., and LEDERBERG, J. (1958). Nature (Lond.), 181, 1419.

OvELL, T. T., TauscHe, F. G., LINpsLEy, D. L., and OwEN, R. D.
(1957). Ann. N.Y. Acad. Sci., 64, 811.

Situ, R. T. Personal communication.

Situ, R. T., and BRIpGEs, R. A. (1958). J. exp. Med., 108, 227.

Stark, O. K. (1955). J. Immunol., 74, 180. .

SULZBERGER, M. B. (1939). Arch. Derm. Syph. (Chicago), 20, 669.

TaLMaGE, D. W. (1957). Ann. Rev. Med., 8, 239.

TRENTIN, J. J. (1958). dnn. N.Y. Acad. Sci., 73, 799.

WEIGLE, W. Q., and Dixon, F. J. (1958). Proc. Soc. exp. Biol. (N.Y.),
99, 226.

Wuitre, R. G. (1958). Nature (Lond.), 182, 1383.

Woonprurr, M. F. A. (1959). In Biological Problems of Grafting, a
Symposium, p. 258, ed. Albert, F., and Medawar, P. B. Oxford:
Blackwell.

DISCUSSION

Dizon: You asked whether the antigen could be circulating in the
form of a complex. This would have to be a very fortuitous situation
in which at all times all complexes would have to be in extreme anti-
gen excess, and even then the rate of elimination of such complexes
would be somewhat faster than the rate of elimination of the antigen
itself; so that I think we can rule out the idea of the non-immune
elimination of the antigen being the result of complex formation.

Good: Gitlin (Gitlin, D., Monckeberg, F., and Craig, J. M. [1958].
A.M.A. J. Dis. Child., 96, 496) has presented convincing morpho-
logical evidence which indicates quite strongly that antibody pro-
duction is not going on in the Felton system. So the classical im-
munological paralysis may not be as different from immunological
unresponsiveness to protein antigen and classical immunological
tolerance as now seems to be the case.
150 DISCUSSION

Lederberg: Prof. Medawar has referred to the apparent tolerance to
homografts which is obtained when adult mice given lethal doses of
X-radiation are rescued with homologous bone marrow, as J. M.
Main and R. T. Prehn (1955. J. nat. Cancer Inst., 15, 1023) and J. J.
Trentin (1958. Ann. N.Y. Acad. Sci., 73, 799) have shown. How-
ever, these authors have also questioned whether they had induced
tolerance in the same sense as is supposed for perinatal inoculations,
namely the modification of the host lymphoid system, or whether
the damaged lymphoid system of the host has merely been sup-
planted by that of the graft.

This may be an academic point, but what is the explicit evidence
that even prenatally inoculated mice have retained the host’s anti-
body-forming tissues ? One might conceive that the graft had again
displaced the host cells in these tissues, rather than modified their
reactivity.

Medawar: It" is very difficult to answer that question, but
one fact that is relevant is that if an animal, before it is irradi-
ated, is sensitized against the cells which are subsequently going to
be used to recolonize it, then the bone marrow graft is rejected. That
is not very good evidence, but it does suggest that one has not com-
pletely eliminated all the antibody-forming cells of the host by radia-
tion—something remains, some residue of the previous sensitization.
The whole problem raised by Prof. Lederberg has been discussed by
D. W. Barnes and J. F. Loutit (1959. Proc. roy. Soc. B, 150, 181).

Monod: It is very difficult to draw a valid parallel between induc-
tion of antibody and induction of enzyme; however, I agree that it
might be possible to find something akin to enzyme induction in the
induction of a system which destroys a particular kind of antigen.
The difficulty there will be the same as one always has in comparing
the two phenomena—namely that the number of enzymes which a
given cell or organism is capable of making is limited, while the
number of antigens against which an organism can be made tolerant
is, L understand, just as unlimited asthe potential numberof antigens.

Turning now to the one cell-one antibody hypothesis, I would like
to refer to some recent results of Cohn and Lennox which seem com-
pletely to eliminate this possibility in the case they have studied.
They studied the synthesis of antibodies to three different phages
(which are not cross-reacting) in singlecells and plasmocytes obtained
from animals which had been immunized with two different phages.
The outcome is not only that many cells are capable of making two
antibodies against T2 and T5, but actually that the number of cells
making more than one antibody is much greater than the combined
probability. So that rather than having to conclude that there is
mutual exclusion between the synthesis of one and another antibody,
154 Discussion

antigen-antibody complexes remain longer in the organism than the
antigen alone is the presence of haematoxylin bodies in systemic
lupus erythematosus. German, from the Rockefeller Institute, put
forward this hypothesis as an explanation for the persistence of these
bodies in the tissues of patients suffering from this disease. In
normal human tissues, such free nuclear material cannot be detected,
so that it looks as if the nuclear catabolism is inhibited by the anti-
nuclear antibody.

Lederberg: One point of the hypercatabolism formulations makes
me uncomfortable; every immunologically competent cell must
vigilantly maintain this accentuated catabolism of the corresponding
antigen.

Burnet: If every cell is made actively tolerant, haven’t you still
got to postulate that all those cells can still suffer pre-emption by an
antigen ?

Medawar: The difficulty is that if you do inject an animal locally
with antigen, it appears that only a few cells, for example in the
regional lymph nodes, react. I think we may have to introduce an
automatic correction when we interpret that phenomenon. We tend
to assume that an animal which has not been deliberately injected
with a particular antigen is a non-immune animal. Actually, all
normal animals are immune; they are immune to a great many anti-
gens which we do not know about, so we must take it that the so-
called normal rabbit is a rabbit which is making antibodies against a
great many antigens, though not as the result of any intervention of
our own.

Lederberg: I might have thought the same, but Coons persuaded
me otherwise. What is the quantitative increase in plasma cells that
one can observe in a local lymph node after a local inoculation ?

Medawar: I am sure you know the answer to that question far
better than I do!

Humphrey: I have made some attempts to test the last hypothesis
you discussed, Prof. Medawar, namely that whether a material acts as
an antigen or not depends upon the rate and extent of its intracellular
catabolism. Even suppose the hypothesis were correct, the dif-
ficulty is to decide in what cells to look for a significant difference in
rate of catabolism between that in an immunologically unresponsive
animal and a normal, or between that of a homologous and a hetero-
logous protein. At first sight the hypothesis receives some encourage-
ment from the results of Coons’s fluorescent antibody technique,
applied to localizing the antigen in cells after parenteral administra-
tion to mice or rabbits of soluble foreign proteins such as human
serum albumin. Coons found antigen in quite a wide variety of cells
such as parenchymal and reticuloendothelial cells of liver, reticular