PAST, PRESENT, AND FUTURE CONCEPTS
Nobel Lecture, 8 December, 1980
by
JEAN DAUSSET
University of Paris VII
Institute of Research in Blood Diseases, Paris, France
As George Snell (1) so rightly said, the supergene, the major histocompatibility
complex (MHC), is like a page from the nature book read outside of the context.
Today this context is beginning to be better understood. We would here like to
recall the evolution of concepts regarding these molecular structures found in the-.
membrane of cells. First, attention was centered on the almost botanical description
of their genetic polymorphism. Then the spotlight was turned, for several years,
on their importance in transplantation. More recently, their role in the immune
response has become more and more apparent. This, however, is probably not the
last stage in our search. We, as well as others (2-5), have suggested that the
essential function of these structures resides in self-recognition. These structures
are, in fact, the identity card of the entire organism.
We will discuss these four viewpoints successively. Far from being mutually
exclusive, they are landmarks in the stages of our thought process as we have
gained deeper knowledge of the subject.
First concept: Polymorphism and Linkage Disequilibrium
Polymorphism:Since Landsteinerís discovery of the first genetic polymorphism
in man, knowledge of polymorphic genes has not ceased to increase and will
continue to increase with DNA hybridization techniques. Most of these systems,
however, are pauci-allelic and more often than not have one very frequent allele,
one that is more infrequent, and a few variants. None of these can be compared
with the extreme polymorphism of genes in the MHC of vertebrates, and
particularly in the human lymphocyte antigen (HLA) complex
The definition of this polymorphism began to emerge in three laboratories: in
ours where the first antigen Mac (HLA-A2) was defined (6), in Van Roodís
laboratory (7) with the 4a 4b series (Bw4, Bw6), and Rose Payneís and W.
Bodmerís (8) with the two alleles HLA-A2 and -A3. Then, thanks to an intense
international effort that has spanned more than 15 years and included eight
workshops, the web began to be disentangled. The importance of this international
effort, launched by Amos in 1964 [see (9)], and followed by other workshops
directed byVan Rood (10), Ceppellini (11), Terasaki (12), Dausset (13), Kissmeyer-
Nielsen (14), Bodmer (15), and again Terasaki (16), cannot be over-emphasized
and is to the credit of the whole histocompatibility community. The four presently
welldefined, closely linked loci, HLA-A, HLA-B, HLA-C, and HLA-D/DR, have
each from 8 to 39 codominant alleles, and the number of haplotypical or
genotypical combinations already amounts to several million. It is very likely that
other closely linked polyallelic loci will be discovered, similar, for example, to the
various loci in the I region of the mouse. If one adds to this complexity the
polymorphism of other genes in the HLAregion, coding, for example, for factors
C2, C4S, C4F, and Bf
of complement, one reaches such levels of complexity that
virtually every human has a different gene combination. If one considers all the
genes of the human genome, it can be said that there is not and will never be on
earth, apart from true twins, two identical people: every person is unique.
A question that immediately comes to mind is: Why is the MHC so complex?
It is clear that a particular pressure was exerted on these genes to make them
different and to maintain this differentiation. If it is true that these structures play
a role in self-recognition and that they derive from primitive genes coding for
surface molecules, then one can conceive of this diversity quickly becoming a
necessity when living matter passed from the unicellular stage -- or from a
syncytium of identical cells able to fuse without harmful consequences -- to the
organized multicellular organism whose tissues must coexist and even cooperate
and whose cells therefore cannot even merge with the cells of an organism of the
same species.
Maintenance of this polymorphism is undoubtedly aided by the selective
advantage given to the heterozygotes, possibly through the immune functions
attributed to the MHC molecules in a subsequent stage of evolution.
The HLA system is now known to have two types of products that are very
different from each other (their single nomenclature sometimes obscures this
difference).
The products of the HLA-A, -B, and-C loci [Kleinís class I (17)] are ubiquitous,
being present at the surface of all (or almost all) cells of the organism. This wide
distribution would suggest that they play a very general biological role.
In contrast, the products of the D/DR locus -- and probably of the ìfutureî DR
loci (class II) -- exist only at the surface of certain specialized cells, essentially
immunocompetent cells, a valuable piece of information with respect to their
functions.
We must not neglect another valuable piece of information afforded by the
major similarities between the class I products and immunoglobulins: the light
chain (the b2-microglobulin) as well as one of the domains (a3) of the heavy chain
have significant similarities with the immunoglobulins and therefore suggest the
possibility of a common ancestral gene (18, 19).
The light and heavy chains of the class II products bear no similarity either with
the class I products or with the immunoglobulins. It can be said, therefore, that
the products of the HLA complex are bipolar and are probably derived, by
duplication and successive mutation, from two very distinct genes, the function
and origin of which go far back in the evolution of the species.
At present, at least three variable regions on the heavy chain of the class I (20)
products are known. In the most distal domain (al) is the main variability zone
(between amino acids 60 to 80), which probably corresponds to the serologically
defined allelic epitope (individual antigenic determinant). This same domain
also has (between amino acids 30 to 40) an apparently variable zone that is
responsible for interaction with the influenza virus. In the median domain (a2)
is a third variable zone (between amino acids 105 to 114).
The extreme frequency of cross-reactions between the various allelic molecules
of each HLA locus is well known. Two interpretations that are not mutually
exclusive are possible: the similarity in the structure of the allelic epitopes [see
Colombani et al. (21)] or the existence of determinants common to two molecules
butdifferentfrom the epitopes; determinantswhich I, and Ivanyi (22), have called
ìantigenic factorsî (also known as supertypical antigens or public antigens):
According to our hypothesis, the molecules having an identical epitope-let us
say, for example, A2-are not identical in their composition. Some may have one
or more different antigenic factors. This variability may be found in the same
population but is more often found in different populations (23). . .
This concept suggests that the various parts of an HLA molecule might have
different functions (as is the case for the different portions of the immunoglobulin
molecule). It has recently been shown that the interaction between the HLA
molecule and the influenzavirus does not take place at the level of the serologically
distinguishable determinant since this virus has a different interaction with
molecules, which, nevertheless, have a serologically identical A2 determinant
(24).
The same hypothesis could be applied to the DR molecules and could perhaps
make it possible to solve the problem of the relations between the D and DR series.
In effect, D might be no more than a variable part of the DR molecule having a
stimulating function, since disassociated haplotypes do appear to exist-that is,
where the determinant D is not the determinant usually found with the DR
antigen (16, 25).
A second possibility, which should not be excluded, is that D in itself does not
exist but is defined by an average of allostimulation due to a certain combination
of alleles in the loci of region D, the linkage disequilibrium involving not only the
DR locus but other loci as well, equivalent to IA, IJ, IC, IE of the mouse.
A second series of DR molecules is already in the process of being defined both
by serological and cellular procedures. In particular an allelic SB series, centrometric
in relation to the DR locus, has just been described (26, 27) through the use of
mixed secondary lymphocytic cultures.
With regard to the genetic organization, we cannot yet grasp this in precise
terms, but with the aid of modern DNA hybridization techniques it will not be long
before we do understand it (28).
Linkage disequilibrium. The four loci of the HLA complex are closely linked on
the short arm of chromosome 6 (21p). They are, however, sufficiently distant for
relatively frequent recombinations to occur (0.8 percent between A and B and 1
percent between B and D/DR in man). This special situation seems presently to
be virtually unique to human genetics. It is, moreover, accompanied by a
particular phenomenon, which is the preferential gametic association between
alleles of several loci of the same complex. A linkage disequilibrium is said to exist
between these alleles. The phenomenon has given rise to numerous speculations.
Is this merely reminiscent of ancestral combinations (when populations were
isolated some 2000 to 4000 years ago) that were revealed or increased by human
migration? Could a mixture of populations temporarily set a certain haplotypical
formula that would survive for only so long as was necessary for it to be dispersed
through segregations occurring in successive generations (29)? Or is this linkage
disequilibrium really a preferential association, that is to say, selected for the
biological advantage or advantages it confers in a certain environment?
Of course, these two mechanisms can operate simultaneously.
One might ask whether the feature of linkage disequilibrium is specific to the
MHC or whether it is a very general feature that recurs at other points of the
genome.
HLA polymorphism and its linkage disequilibrium are valuable tools for
anthropologists and epidemiologists. They allow the former to characterize a
population; to discern its origin and draw up its genetic history. They allow the
latter to compare HLA formulas and HLA haplotypes with the particular
susceptibility of a population or groups of populations to certain diseases; and
perhaps in the future they will be able to reconstitute major diseases and epidemics
that have occurred in the past by observing the selections that have operated.
Finally, in formal genetics, the HLA complex is certainly the segment of the
human genome that is best known and is a major example of our relating a human
product to the sequence of the corresponding gene.
Second Concept: Transplantation Antigens
It is in such terms that these membrane structures are most often defined,
because at the same time that genetic polymorphism was being elucidated,
transplantation in humans was assuming great importance in therapeutics.
Our understanding of allogenic response in man has evolved rapidly. When
only the HLA-A and -B antigens were known, allogenice response amounted to
cytotoxicity on targets A and B. Thanks to admirable volunteers, the correlation
between the survival of skin graftsand the number of HLA-A and-B incompatibilities
was clearly demonstrated (30-33). The same correlation was seen in recipients of
related or non-related donor kidneys. This correlation, which was long debated,
is no longer challenged; however, the benefit of compatibility limited to these two
loci is very variable depending on the categories of patients.
When locus D was discovered (34,35), its importance in transplantation was
immediately suspected (36, 37). In fact, it was shown in vitro that lymphocytic
proliferation in allogenic culture was only possible when there was incompatibility
at the D locus (38). Clinically, a correlation has been found between the intensity
of proliferation during the mixed lymphocytic reaction, that is, between the
recipient and the related donor, and the survival of the graft. However, it has not
been possible to apply this observation to the transplantation of nonrelated organs
because of the time required for its elaboration.
In contrast, as soon as DR antigens (14, 15) could be detected by serological
means - and thus rapidly - it became possible to use this new method of selection
successfully. A DR incompatibility is accompanied by a drop in the survival rate of
grafts, both skin (39, 40) and organ transplants (41, 42). In both cases there is a
clear additive effect with those of the A and B incompatibilities (Fig. 1).
Incompatibility is in most cases both DR and D, and thus has two consequences:
1) With DR incompatibility it provides a target for cytotoxic cells. DR antigens
are true targets: they do not behave like minor antigens because they do not need
the HLA-A and -B identity between the killer cell and the target (43).
2) It induces, by the D disparity, the appearance of auxiliary cells, some of which
are helper cells and others suppressor cells. In the normal state, helper cells
dominate suppressor cells. However, it must be made clear that in certain
circumstances, the suppressor cells dominate the helper cells and the scale then
tips in favor of tolerance.
It is possible that the indisputably beneficial effect of preoperative transfusions
is due to the development of suppressor cells or factors in the recipient (44). In
fact, with Sasportes and others (45, 46) we have shown that hyperimmunization
against DR is accompanied by the in vitro appearance of a suppressive factor
capable of producing a specific feedback inhibition on its own cells. The exact
circumstances that cause the suppression invivo to sometimes dominate immunity
are unknown. However, it is known that in the monkey the beneficial effect of
transfusions has been observed onlywhere the animals are also immuno-suppressed
(47). Recipients ofkidneys are, so to speak, alwaysimmunosuppressed due to their
renal insufficiency; this would explain why, in the course of transfusions, the
immune balance leans in favor of suppression.
On the basis of the preceding considerations we have proposed (48) a
theoretical plan, of necessity provisional, for the choice of blood donors for
transfusion and organ donors. Without going into detail, it is based on the
following principles:
1) Before transplantation, a state of tolerance must be developed in the
recipient and at the same time immunization against HLA-A and B-targets must
be avoided. Thus, transfusions should be made with DR incompatible blood that
is A, B compatible. The same DR incompatibility should be used constantly in
order to increase the chances of the appearance of suppressive cells and factors
of the allogenic proliferation.
2) In selecting the organ donor one must avoid providing targets againstwhich
the recipient could be immunized. Thus, priority should be given to HLA-DR
compatibility in patients who have not produced antibodies against HLA-A and
-B antigens in the coarse of transfusions (nonresponder recipients). On the other
hand, priority should be given to HLA-A and -B compatibility in those who have
been sensitized in the course ofpreoperative transfusions (responder recipients).
Indeed, for the former there is little chance of immunization occurring against A
or B antigens, but the appearance of helper cells and a supply of DR targets should
be avoided. Conversely, in the responders, helper cells are already present and
their action must be neutralized by not contributing incompatible A and B targets.
Very precise and detailed treatment protocols will be necessary to verify or
disprove the validity of this plan.
Third Concept: Role in the Immune Response
This third concept is essentially based on our knowledge of animals since
systematic experiments are ethically difficult and thus rare in man(49). Nonetheless,
to date, the parallelwith the H-2 complex is striking. Here again we find the bipolar
division of the functions of the products of the HLA complex:
1) Class I products appear to serve as targets when a cell is either infected by a
virus or covered with a hapten.
2) Class II products appear to serve as a regulator between the various cell
subgroups involved in the immune response.
In both cases, a phenomenon of restriction is most often observed, that is to say,
an identity with class I or II products is apparently necessary between the
cooperating cells.
Thus a phenomenon of restriction exists in the cytotoxicity of a killer T
lymphocyte cell against a cell carrying a virus (50,51), a hapten (52) such as the
DNP, or a normal antigen such as the H-Yantigen (53). The killer cell must have
at least one identity with the HLA-A and -B (class I) antigens of the target cell in
order for the lysis to be effective, or else the killer must have matured in the
presence of the histocompatibility antigens of the target.
Similarly, when an antigen such as PPD is presented by human macrophages
(54), the presence of a DR identity (class II) is apparently necessary for the
presentation to be effective and for lymphocytic proliferation to occur. This
restriction is not absolute, however, and a certain number of proliferative
reactions that can be explained by cross-reactions between DR antigens has been
observed.
As in the mouse, where soluble factors carry antigens from the I region that
convey a specific message to another T or B cell population, so in man there is
evidence of a certain number of soluble factors of this type (55,56). Undoubtedly,
when we have a better understanding of the various products of region D in man,
numerous specific and aspecific factors, either restricted or unrestricted, will be
described.
The restriction phenomenon is probably the most direct proof of the role of
the products of the HLA complex in the immune response of man.
Indirect proof has been sought in the numerous associations between HLA and
diseases. Based on the murine model, the first study of associations between HLA
and disease was done in our laboratory on acute lymphoblastic leukemia (57). A
slight but definite increase of A2 has now been demonstrated in numerous
worldwide studies (58). Likewise, the Al antigen is slightly increased in Hodgkinís
disease. These two observationswould suggest thatagene acting on hematopoiesis
may exist near locus A (58).
However, most of the diseases indisputably associated with HLA are neither
tumorous nor obviously infectious (59). These are chronic or subacute diseases
having a definite familial though mild character, that are of unknown etiology and
are not included in any of the major classifications. For a good number of these
there is an obvious autoimmune component.
We will give only two examples to illustrate once again the bipolar nature of the
functions of HLA products. In the first example, class I products may still be
considered as possible targets; in the second example, class II products may be
considered as regulators of the immune response.
Articular and more especially sacropelvic disorders that are strongly associated
with B27 seem to require the molecule HLA-B itself to play an essential role. In fact,
the same B27 antigen is found in a series ofdisorders (Reiterís syndrome, ankylosis
spondylarthritis, psoriatic rheumatism) which tend to affect the articulations of
the sacrum and pelvis. Further, this same predisposition is found in all populations
of the globe. However, it has now been clearly demonstrated that the same
pathological manifestations can also affect a small number of individuals who are
B27 negative. The B27 epitope is therefore not indispensable. At least two
hypotheses may be advanced: either the responsible gene in all populations is
strongly linked with the B27 antigen, or molecule B27 has a variable part (an
antigenic factor) that is responsible for susceptibility; this antigenic factor would
not always be present on all B27 molecules and could be found on other HLA
molecules probably having cross-reactions with B27 (in keeping with our concept
explained above).
It seems, moreover, that for these diseases there is a factor that triggers
infection. In fact, it is known that some acute intestinal infections caused by Gramnegative
bacteria such Shigella, Salmonella, and Yersinia are complicated by ankylosing
spondylarthritis mainly in patients who are B27 positive. Recently, a direct
relationship was suggested between the B27 antigen and a type of Klebsiella
(KlebsiellaB43). The antibodies against Klebsiella would be capable of recognizing
specificallyanantigenpresenton the lymphocytes of B27 positivepatientsaEected
by ankylosing spondylarthritis. Further, lysates from this infectious agentwould be
capable of transforming lymphocytes in B27 normal individuals and of making
them sensitive to the antibodies against Klebsiella (60, 61). Although clinically the
link between a Klebsiella infection and ankylosing spondylarthritis is still unclear,
this observation, not yet confirmed, suggests that certain infectious agents would
be capable of modifying HLA antigens and of apparently making them similar to
those in patients. It is not impossible that this type of mechanism may one day
explain the linkages with other microorganisms referred to above. These
microorganisms would be capable of modifying the B27 antigen and perhaps
certain other HLA molecules as well (to take into account B27 negative patients),
and of making them privileged targets for T-immune autolymphocytes. Although
this hypothesis is enticing, it cannot yet explain the very special localization of
lesions, since the B27 antigen, like all HLA antigens, is practically ubiquitous.
If we now consider disorders associated with HLA-D/DR, one is at the outset
struck by the large number of them that are associated with Dw3/DR3 and, more
especially in Caucasians, with the Al, B8, DR3 haplotype. For the most part, there
are diseases with a strong autoimmune component and a low family penetrance,
such as myastheniagravis, Gravesí disease, Addisonís disease, Sj-grenís syndrome,
disseminated lupus erythematosus, and active chronic hepatitis. It appears that
this haplotype, in strong linkage disequilibrium, has a gene or perhaps a series of
genes that are conducive to autoimmunization.
Insulindependentjuvenile diabetes (IDD) , itself associated with Al, B8, DR3
and also with B18, DR3, and B15, DR4 is in this respect very instructive (62, 63).
.The viral etiology of IDD is highly suspected: experimental models of the disease
do exist and specific observations in some cases in man incriminate the B4
Coxsackie virus. One can therefore infer that the virus has destroyed a certain
number of islets of Langerhans and triggered a process of cellular autoimmunity
where cytotoxic lymphocytes persist in the organism against antigens modified by
or associated with a virus. The disease would thus be self-sustained. In this
hypothesis the D region products would have been incapable of inducing an
adequate immune reaction against the virus causing the disease. In contrast,
individuals who are DR2 positive (most frequently A3, B7, DR2), who appear to
be ìprotectedî against IDD would have a more effective immune response against
the responsible agent.
This is only a working hypothesis which would have the advantage of applying
to other diseases associated with HLA-DR such as multiple sclerosis (DR2) and
chronic polyarthritis (DR4) orjuvenile rheumatism (DR5).
The reality, however, is certainly far more complex. In fact, in no case is the
association complete with a DR antigen. This is generally explained by a linkage
disequilibrium between a simple susceptability gene for the disease and a DR
allele. Here too, however, one might think there are polymorphic parts to DR
molecules other than the epitopes presently known and that these might have a
particular immune function. Even better, it could be assumed that the interaction
of two (or more than two) genes from the D region would be conducive to an
adequate immune response. This interaction could take place in the &position
(64 ) between genes of the same haplotype (as in the interaction between I-A b and
I-Eb to form an I-E molecule in the mouse) or in the trans (65) position between
two haplotypes (such as I-Akand 1-Ab) complementation in the mouse).
As a corollary to these gene interactions we propose that each HLA haplotype,
and especially those in linkage disequilibrium that are found most frequently in
the numerous diseases associated with HLA, has its own gene configuration that
confers on it a particular capacity for immune response, which may be favorable
in certain environmental conditions and unfavorable in others (for example the -
A3, B7, DR2) haplotype gives a susceptibility to multiple sclerosis and protects
against IDD) Thus, each HLA complex would be composed of a set of genes that
have subtle interactions among one another (such as gene C2 with the two C4
genes) thereby giving them a specific identity in immunological terms.
Likewise, every individual possessing two HLA haplotypes has his or her own
immunological capacitywhich is conferred on him by the two particular haplotypical
formulas inherited from his two parents, but which is also the result of the genetic
interaction or complementation between the these two complexes. Thus each
individual has a personal immune response that makes him either susceptible or
resistant to certain diseases. Here again each haplotypical combination may be
beneficial or harmful depending on the type of challenge to which the individual
is subjected.
In terms of the population, we can thus conceive that certain individuals are or
were more exposed and thus are or were more easily eliminated than other
individuals more resistant to past and present epidemic or endemic diseases. But
it is not the same individuals who are susceptible or resistant to the different
attacks; this is what makes the survival of a population possible, and thus the
perpetuation of the human species.
MHC products are distributed on the surface of cells. Those of class I are
virtually ubiquitous. Class II products may be found on immunocompetent cells
but also on endothelial and other specialized cells. Their location suggests that
they play a role in the social organization of cells of the same organism.
This assumption is strongly supported by the following observation: it appears
to be necessary for MHC products to share an identity in order for cooperation to
be etablished between two populations of cells in the same organism or in two
different organisms. This is the restriction phenomenon which we discussed
above, and which is valid for the two classes of products. Results confirming this
apparent need for identity are accumulating very rapidly in both animals and in
man and are no longer limited solely to immunocompetent cells. The same is true,
for example, in the adhesion phenomenon between fibroblasts and especially in
the ìhomingîphenomenon in ganglia. Degos and others (66) have observed that
splenocytes injected intravenously must share an identity with the cells of the
capillary endothelium with regard to class I products so that homing can occur.
The identity of class II products does not intervene (Fig. 2).
Fig. 2. Homing in lymph nodes according to identities (a) or differences (0) Labeled lymphocytes
were injected intravenously into mice under different genetic conditions: allogenic (first line),
syngenic (second line), congenic where the difference involves only one or several genes of the H-
2 complex (all other lines). The average value of r (percentage of horning from which the control
value has been deducted) is high wherever there isidentity (0) with H-2D or H-2K; H-21 identity has
no influence (66).
We are thus faced with a very general phenomenon with such a consistent
record that it is difficult to escape the conclusion that the two types of molecules
have a common function. They could serve as a recognition signal (recognizers)
among cells of the same organism; a signal necessary but probably insufficient to
permit effective cooperation between the two subpopulations of cells because it
would be the same, at least at class I, for all cells of the organism (2-5).
The passive or negative discrimination of self implies that the cells ìignoreî one
another. This seems improbable in view of the cohesion of tissues and of their
interaction. Without self-recognition, each specialized cell and each tissue would
be isolated and incapable of surviving. These considerations thus suggest that selfrecognition
is an active phenomenon.
The subjacent mechanism of self-recognition is still unknown. At least three
possibilities come to mind. The most orthodox is the complementarity between
two different molecules. But one cannot exclude recognition by identity, whether
that recognition takes place between two identical molecules or through a ligand.
This fascinating problem has been discussed in detail elsewhere (67). Suffice it to
mention here just two remarks in relation to complementarity:
lIf a second molecule (receptor) existed with the same immunogenicity as the
HLA determinant, a second allelic system as complex as the former would have
been found. To date no such system has been found. However, one should remain
aware of the weak immunogenicity of the idiotypes which could represent these
receptors.
lIf the receptor and the determinant were coded by two different genets, any
mutation and selection of one should correspond to the mutation and selection -
of the other. This is unlikely. Here again, however, one might envisage, according
to Jerneís theory, that each individual has all possible receptors, the appropriate
receptor being selected in early life, perhaps in the thymus. This is probably what
occurs, for example, in the growth of allophenic mice in which cells carrying
different H-2 antigens coexist.
Whatever the mechanism, the fact remains that self-recognition is a general
and active phenomenon of any cell that is at least partially, linked to the MHC. We
suggest that class I products are responsible for individuality, for integrity, and
perhaps for the general cohesion of the being and that class II products are an
example of cellular cooperation, thanks to self-recognition at the level of the
differentiated cells of the immune system, allowing the immune system to
function harmoniously.
Future Prospects
Thus the increasingly deep understanding of the MHC in man opens up
exhilarating prospects both in public health and in basic science.
With regard to organ transplantation, we do not feel that the choice of the most
compatible donor will be the last word. On the contrary, our aim must be to
provoke in the recipient a specific tolerance to incompatible donor antigens
without at the same time diminishing his or her immunological defenses. It seems
that with preoperative transfusions the way has been opened to this type of
preparation. We must now attempt to unravel its detailed mechanism so that the
method can be used more generally. This will be the objective of the years to come,
and we have no doubt that it will be achieved.
Although organ and bone marrow transplantations mark a milestone and have
already brought help to numerous patients, they should not be considered an end
in themselves. Etiological treatment should progressively replace them.
The discovery of more than 50 diseases associated with or linked to HLA is
perhaps still more promising, and although the diagnostic or prognostic benefits
to practising physicians are still limited, physicians recognize the validity of this
approach. They know that correction of the abnormalitywhich provokes a disease
is close at hand when the gene responsible has been located and its function
defined. Thanks to the astounding possibilities offered by genetic engineering it
will henceforth be possible to know the exact DNA sequence in the vicinity of the
HLA genes. The latter will serve as markers and will make it possible to discern
anomalies of susceptibility genes. We must emphasize the possibility that in some
cases no anomaly will be found because a gene (or combination of genes) may be
perfectly active in the defense against certain antigens but totally inactive in the
defense against others. Thus an inventory of the immunological capacities of each
individual will need to be drawn up. This inventory will show the weaknesses
(susceptibility), the excesses (autoimmunization), and the good capacities
(protection) afforded by each type of gene combination. In this way, preventive
medicine of high precision will be possible; a personalized medicine that will be
more efficient and less burdensome for the community than the present mass
system.
At the same time, researchers now have the means with which to approach the
crucial problem represented by the subtle organization of manís immune system.
The cascade of interrelationships, and the language, between different
immunocompetent cells will be clarified; the place of the ìspecificî and the
ìnonspecificî will be recognized; the role of HLA products in messages between
welldefined cells will be determined. This deeper understanding of manís
immune response will quickly have major repercussions in pathology. It will
perhaps provide the key to the irritating problem of the treatment of cancer, and
may also provide a simple means of inducing graft tolerance at will. It will also
perhaps lead to an immunological treatment of the major parasitic diseases that
still afflict such a large part of mankind.
Finally, the discovery of the primary function of the molecules of the MHC
found at the surface of all, or almost all cells of the organism will be a decisive step
in our understanding of the differentiation and social organization of cells.
The way already trod is but a simple introduction. They are still many
marvellous pages to be written. . .
References and Notes
1. G. D. Snell, Havey Lect, 74, 49 (1979).
2. , Folio Biol. (Prague) 14, 335 (1968).
3. N. K. Jerne, 1 Eur.J. Immunol. 1, 1 (1971).
4. J. Dausset, A. Lebrun, M. Sasportes, C. R. Acad. Sci. Paris 275, 2279 (1972).
5. W. F. Bodmer, Nature (London) 237, 139 (1972).
6. J. Dausset, Acta Haematol. 20, 156 (1958).
7. J. J. Van Rood and A. Van Leeuwen, J. Clin. Invest. 42, 1382 (1963).
8. R. Payne, M. Tripp, J. Wiegle, W. Bodmer, J. Bodmer, Cold Spring Harbor Symp. Quant. Biol. 29,
285 (1964)
640 Physiology or Medicine 1980
9. Ií. S. Russell, H.,J. Winn, D. B. Amos, Eds., Histcompatibility Testing (Publ. No.
11229, National Academy of Sciences, Washington, D. C., 1965).
10. H. Balner, F. J. Cleton, J. G. Eernisse, Eds., Histocompatihzli~ Testing 1965(Munksgaard,
Copenhagen, 1965).
11. E. S. Curtoni, P. L. Mattiuz, R. M. Tosi, Eds., Histocompatibility Testing 1967
(Munksgaard, Copenhagen, 1967).
12. P. I. Terasaki, Ed., Histocompatibility Testing 1970 (Munksgaard, Copenhagen, 1970).
13. J.Daussetand J. Colombani, Eds., Histocompatibility Testing 1972(Munksgaard, Copenhagen, 1973).
14. F. Kissmeyer-Nielsen, Ed., Histocompatibility Testing 1975 (Munksgaard, Copenhagen, 1975).
15. W. F. Bodmer, J. R. Bachelor, J. G. Bodmer, H. Festeinstein, P. J. Morris, Eds., Histocompatibility
Testing 1977 (Munksgaard, Copenhagen, 1978).
16. P. I. Terasaki, Ed., Histocompatibility Testing 1980 (University of California, Typing Laboratory,
Los Angeles, 1980).
17. J. Klein, Science 203, 516 (1979).
18. H. T. Orr, D. Lancet, R. J. Robb, J. A. Loper de Castro, J. I.. Strominger, Nature (London) 282, 266
(1979).
19. J. L. Strominger, Immunology 80, 541 (1980); M. Fougereau and J. Dausset, Eds., Fourth
International Congress on Immunology, Paris, 1980 (Academic Press, London, 1980).
20. H. T. Orr, J. A. Lopez de Castro, P. Pat-ham, H. Ploegh, J. L. Strominger, Proc. Natl. Acad. Sci.
U.S.A. 76, 4395 (1979).
21. J. Colombani, M. Colombani, J. Dausset, in (12), pp. 79-92.
22. P. Ivanyi and J. Dausset, Vox Sang. 11, 326 (1966).
23. J. Dausset, Transplant. Proc. 3, 1139 (1971).
24. W. E. Biddison, M. S. Krangel, J. L. Strominger, F. E. Ward, G. M. Shearer, Shaw, HumImmunol
1, 225 (1980).
25. M. Sasportes, A. Nunez-Roldan, D. Fradelizi III, Immunogenetics, 6, 55 (1978).
26. C. Mawas, D. Chat-mot, M. Sivy, P. Mercier, M. M. Tongio, G. Hauptmann, J. Immunologenet.
5,383 (1978).
27. S. Shaw, M. S. Pollack, S. M. Payne, A. H. Johnson, Hum. Immunol. 1, 177 (1980).
28. H. L. Ploegh, H. T. Orr, J. L. Strominger, Proc. Natl. Acad. Sci. U.S.A. 77, 608l (1980).
29. L. Degos and J. Dausset, Immunogenetics 1, 195 (1974)
30. J. Dausset, F. T. Rapaport, P. Ivanyi, J. Colombani, in (10), pp. 63-72.
31. J. J. Van Rood, A. Van Leeuwen, A. Shippers, M. J. Vooys, E. Frederiks, H. Balner, J. G. Eernisse,
in (10), pp. 35-50.
32. R. Ceppellini, E. S. Curtoni, P. L. Mattiuz, G. Leighes, M. Visetti, A. Colombi, Ann. N. Y. Acad.
Sci. 129,421 (1966).
33. D. B. Amos, H. F. Siegler, J. G. Southworth, F. E. Ward, Transplant Proc. 1, 342 (1969).
34. F. H. Path and D. B. Amos, Science 156, 1506 (1967).
35. E. J. Yunis and D. B. Amos, Proc. Natl. Acad. Sci. U.SA. 68, 3031 (1971).
36. J. Hamburger, J. Crosnier, B. Descamps, D. Rowinska, Transplant Proc. 3,260 (1971).
37. K. C. Cochrum, H. A. Perkins, R. 0. Payne, S. L. Kountz, F. O. Belzer, ibid. 5, 391 (1973).
38. V. P. Eijsvoogel, J. J. Van Rood, E. D. DuToit, P. H. A. Schellekens, Eur. J. Immunol. 2,413 (1972)
39. J. Dausset, L. Contu, L. Legrand, A. Marcelli-Barge, T. Meo, F. T. Rapaport,J. Clin. Invest. 63,
893 (1979).
40. M. Jonker, J. Hoogeboom, A. Van Leeuwen, C. T. Koch, D. B. Van Oud Alblas, J. J. Van Rood,
Transplantation 27, 91 (1979)
41. A. Ting and P. J. Morris, Lancet 19781, 575 (1978).
42. T. Moen, D. Albrechtsen, A. Flatmark, A. Jakobsen, J. Jervell, S. Halvorsen, B. G. Solheim,
E. Thorsby, N. Engl. Med. 303,850 (1980).
43. C. F. Feighery and P. Stastny, Immunogenetics 10, 39 (1980).
44. G. Opelz, M. R. Mickey, P. I. Terasaki, Lancet 1972-1, 868 (1972).
45. M. Sasportes, D. Fradelizi, J. Dausset, Nature (London) 276, 502 (1978)
46. M. Sasportes et al., J. Exp.. Med. 152 (No. 2), 270s (1980).
47. A. A. Van Es and H. Balner, in Immunogenetic and Transplantation Studies in the Rhesus
Monkey,A. A. Van Es, Ed. (J. H. Drukkerij, B. V. Pasmanís, síGravenhage, 1980), p. 117.
48. J. Dausset and L. Contu, Transplant. Proc. 13, 895 (1981).
49. , Immunology 80, 513 (1980); M. Fougereau and J. Dausset, Fourth
InternationalCongress on Immunology, Paris, 1980 (Academic Press, London, 1980)
Thr Major Histocompatibility Complex in Man. 641
50. A. J. McMichael, A. Ting, H. J. Zweerink, B. A. Askonas, Nature (London) 270, 524 (1977).
51. S. Shaw, G. M. Shearer, W. E. Biddison, J. Exp. Med. 151, 235 (1980).
52. E. Dickmeiss, B. Soebrrg, A. Svejgaard, Nature (London) 270, 526 (1977).
53. E. Goulmy, J. D. Hamilton, B. A. Bradley, J. Exp. Med. 149, 545 (1979).
54. H. Hirschberg, O. J. Bergh, E. Thorsby, ibid, 150, 1271 (1979).
55. F. B. Mudawwar, E. J, Yunis, R. S. Geha, ibid. 148, 1032 (1978).
56. S. M. Friedman, 0. H. Irigoyen, D. Gay, L. Chess,,/ Immunol. 124, 2930 (1980).
57. F. M. Kourilsky, J. Dausset, N. Feingold,,J. M. Dupuy, J. Bernard, in Advances in Transplantation,
First International Corgress of the Transplanation Society, Paris, 1967 (Munksgaard, Copenhagen
1967), pp. 515-522.
58. L. P. Ryder, E. Andersen, A. Svejgaard, Eds., HLA and Disease Registry, Third Report (Munksgaard,
Copenhagen, 1979).
59. J. Dausset and A. Swejgaard, Eds., HLA and Disease (Williams & Wilkins, Baltimore, 1977).
60. A. F. Geczy, K. Alexander, H. V. Bashir, J. Edmonds, Nature (London) 238, 782 (1980).
61. C. Druery, H. Bashir, A. F. Geczy, K. Alexander, J. Edmonds, Hum. Immunol. 1, 151 (1980).
62. A. Svejgaard, P. Platz, L. P. Ryder, in (10), pp. 638-656.
63. I. Deschamps, H. Lestradet, F. Clerget, C. Bonaiti, M. Schmid, M. Busson, A. Benajam,
A. Marcelli Barge, J. Dausset, J. Hors, Diabetologica, 19, 189 (1980).
64. P. P. Jones, D. B. Murphy, H. O. McDevitt, J. Exp. Med. 148, 925 (1978).
65. W. P. Lafuse, J. F. McCormick, C. S. David, ibid. 151, 1709 (1980).
66. L. Degos, M. Pla, J. Colombani, Eur. J. Immunol. 9, 808 (1979).
67. J. Dausset and L. Contu, Hum. Immunol. 1, 5 (1980)