Regulation of delayed-type hypersensitivity to pathogens and alloantigens

Immunology Today rot. 3, .Nb. I 1982

18

Regulation of delayed-type hypersensitivity to pathogens and alloantigens F. Y. Liew Department of Experimental Immunobiology, The Wellcome Research Laboratories, Beckenham, Kent BR3 3BS, U.K.

Recent studies reinforce the notion that delayed-type hypersensitivity plays'a key role in the host defence against microbial and inlracellu/ar parasitic infection, and in the rejection ()f skin aIlografts. F. T. Liew reviews these studies and discusses the observation that this T-cell mediated immunity is profoundly regulated by antigenspecific suppressor Tcells, some of which are restricted by products of the I-J subregion of the MHC. The concept of classical delayed-type hypersensitivity (DTH) was originally based on the tuberculin-type skin reaction which develops in sensitized guinea pigs within 4-6 h and becomes fully manifest in 18-24 h (Ref. 1). Today D T H is defined as an immunologically specific inflammatory reaction maximal at 24-48 h and with a characteristic histologic appearance of infiltration with mononuclear cells. The reaction is transferable by T lymphocytes whose induction and manifestation are restricted by products of the major histocompatibility complex (MHC). A subset of antigen-specific precursor T cells, upon stimulation by antigens presented in the context of M H C determinants ori antigen-presenting cells, differentiate and become sensitized T cells which are destined to mediate a D T H response (Fig 1). This subset of T cells, T~ cells, carries the Lyt 1 surface antigen but neither Lyt 2 nor la antigens. Some of these may differentiate separately to become memory cells for DTH. These specifically sensitized T D cells form a long-lived recirculating pool which reacts to antigen at a local site by proliferation and release of soluble mediators (lymphokines) which attract and activate a non-specific population of macrophages. There is evidence that the activation of sensitized T~ cells requires compatibility of I-region gene products between T D cells and cells presenting the antigens. The non-specific bone-marrow-derived macrophages, upon activation by lymphokines, continue to proliferate and infiltrate at the site of reaction and become predominant in the lesions. A major drawback to the study of D T H remains the lack of a precise assay system. The footpad test devised in 1955, by Gray and Jennings 2, is currently most widely used for detecting D T H in rodents. This assay measures the gross manifestation of D T H reactioi~ and has often been criticized for its lack of sensitivity and objectivity. The ear/footpad radioactivity accumulation method 3 and the in-vitro macrophage migration inhibition assay 4 have also been used to some extent. However, there is uncertainty as to the direc~ correlation between these assays and the classic D T H reaction. It is also unclear whether the local © Elsevicr Biomedical Press 1982 0167 4919/82/0000-0000/$275

transter of DTH, by injecting sensitized T cells together with specific antigen intradermally into the mouse footpad, also represents a true form of DTH. This confusion is compounded by the findings that cloned antigen-specific T-helper cells s for antibody production and cytotoxic T cells 6 transfer 24 h footpad swelling locally. These results imply that D T H is mediated by T H cells (Lyt-l+2 , H-2 I restricted) as well as by T c cells (Lyt-l-2 +, K I D restricted). It may be that macrophages are responsive to a wide range of stimuli released by activated T cells and that D T H is a gross immunological phenomenon, reflecting merely the activation of macrophages. O n the other hand, it is equally possible that only limited numbers of constituents in lymphokines released by the so called T~ cells are responsible for the activation of macrophages for the elimination of invading pathogens and foreign antigens, and that the local inflammatory reaction is caused by irritant unrelated to the classical D T H response. Thus, there is a real need for a definitive reference assay for D T H reaction based, perhaps, on the production of macrophage-activating lymphokines by T cells, coupled with classical histological identification of mononuclear cell infiltration of the reacting site.

@

No.-..cJf~o /

Fig. 1. Schematic representation of DTH induction and manifestation. APC = antigen presenting cells, Ag ~ antigen, Cy cyclophosphamide, irr. = X-irradiation. TDp= precursor of T cells mediating DTH (TD), TDm = memory T D cells, Mono = monocytes, M(/) = macrophages.

Immunology T o @ vol. 3, 2¢~. 1 1982

Whatever the classification of the T-cell subset mediating D T H , this antigen-specific cellular reaction has been closely correlated with both the host defence against certain infectious diseases and homograft rejection. However, by arming the host macrophages non-specifically the D T H response may produce severe host tissue damage, particularly if infections occur in vital organs. Thus, from the evolutionary stand point, it is critical that D T H reactivity be kept in optimum homeostasis for the success of the infectious agents as well as for the survival of the host. It is therefore not surprising that the D T H reaction has turned out to be tightly regulated by suppressor T cells. In experimental systems, antigen-specific suppressor T cells (Ts) for D T H to picryl chloride were described by Asherson and Zembala 7 and to dinitrofluorobenzeine (DNFB) by Claman el al. 8. Later Lyt1+2- T s for D T H to particulate antigens, horse and sheep red blood cells 9,1°, was reported and the antigen-specific I - J determinant-bearing suppressor factor (TsF) characterized 1~,~2. These T s cells appear to be specific for D T H and have no apparent effect on antibody responses. A detailed regulation scheme for D T H has recently been proposed by Sy et al. ~3 based largely on data obtained from chemically defined haptenic systems. This involves the sequential induction of various sub-populations of T s and TsF, constituting a circuit of modulation reminiscent of the regulation of B-cell responses. Validation of this concept and its generality will no doubt be the subject of intensive investigation in the future. This review, however, will be restricted to some of the recent studies on the regulation and intrinsic importance of D T H to pathogens and alloantigens. D T H in viral infection Mice infected with an aerosol of influenza virus or immunized with purified u.v.-inactivated whole virus or viral subunits developed transient D T H to influenza virus. The level of D T H is greatly enhanced and sustained when the mice are pretreated with 100-150 mg/kg of cyclophosphamide, a treatment known to eliminate T s precursors and augment D T H response. D T H induced by influenza virus infection is type-specific, whereas that induced by bromelaintreated purified haemagglutinin (HA) subunits shows subtype specificity. It thus appears that the effector T cells of D T H do not discriminate between the antigenic variants of influenza virus which can be distinguished by serology ~4. The genetic influence on the manifestation of D T H to viral antigens has been extensively studied. The cells mediating D T H to infectious lymphocytic choriomeningitis virus are restricted only to the K and D subregion of the I I - 2 (Ref. 15), whilst the cells mediating D T H to reovirus ~', infectious influenza or Sendai virus are/f, D and IA region restricted ~7. On the other hand, the D T H response to inactivated influenza or Sendai virus is restricted to the I A subregion only 17. In the case of herpes simplex

19 virus (HSV), D T H induced by the infectious virus, injected subcutaneously, is only I A restricted ~s. Thus, the genetic requirement for the expression and probably the induction of D T H to viral antigens is highly dependent on the replicating nature of the virus which in turn is conditioned by the pathology, route and nature (infectious or inactivated) of the viral antigen presentation by the antigen-presenting cells. Antigen-specific T s cells for D T H , analogous to those induced by erythrocyte antigens and contactsensitizing antigens are generated during influenza virus infection ~9. The T s cells are Lyt-l+2 and therefore unlikely to be T c cells. They first appear 2 weeks after infection and can be detected in the spleen for at least 40 days thereafter. The T s cells are specific for a subtype of HA and do not affect D T H to other antigenic epitopes of the virus. T-helper cells for antibody response to HA are induced concomitantly with the T s cells for DTH. A T s cell system similar to this has also been demonstrated in mice injected i.v. with u.v.inactivated reovirus 2°. In infection of mice with HSV, it was also found that i.v. doses of living avirulent mutant strains or u.v.-inactivated virulent viruses can produce a state of tolerance that is specific for D T H and is readily transferable by splenic T cells 21. In contrast to the suppression above, Lyt-l+2 splenic and lymph node cells fl'om mice primed with matrix protein of influenza A virus, significantly enhance the D T H response to the HA of the same type, upon transfer to syngeneic recipients 22. This helper effect is carrier-specific, and is expressed only when both matrix protein and HA are presented in physical association. More recently, T H cells tor the generation antigen-specific D T H to influenza virus has been demonstrated in vilro 23 and to herpes simplex virus in vivo (Nash, A. A., personal communication). In the influenza virus system, it is important to note the difference between the suppression of D T H during influenza virus infection and the apparent lack of suppression by immunization with inactivated virus. Inactivated viruses when administered to the host may passively attach to the host's antigen-presenting cells via their HA, thus providing equal presentation probability to both the 'suppressor epitope' (IIA) and the 'helper epitope' (matrix protein). In contrast, replicating viruses mainly express viral subunits such as HA on the surface of infected cells, and internal components such as matrix protein are barely detectable on the surface of host cells 24. Thus, there may be presentation advantage for the suppressor epitope during infection-leading to the predominant activation o f T s cells by HA. Whatever the detailed mechanism, it appears that the D T H response in viral infection is readily suppressed. This could well be the consequence of evolutionary pressure, since there is little evidence that D T H is an important arm of immunity against viral infection. For example, in the murine-HSV system, even though D T H may be involved with primary immune defence, once antibody production and T<

Immunology Today vol. 3, No. 1 1982

20 responses are established, it is superfluous or even damaging2L In influenza pneumonia in mice, there is good evidence that T D cells contribute significantly to the pathology resulting in an increased early mortality 25. Thus, it is to the advantage of the host and ultimately to the survival of the virus that D T H should be rigorously regulated. D T H to p r o t o z o a l i n f e c t i o n

In principle, infectious organisms whose primary targets are macrophages should be susceptible to macrophage-activating lymphokines produced by sensitized T , ceils. Indeed, there is evidence that acquired resistance to Salmonella typhimurium, Listeria monocytogenes, Mycobacterium Ieprae and Leishmania tropica is correlated with D T H reactivity. Leishmaniasis provides perhaps the clearest example of the protective role played by D T H in protozoal infection. The majority of cases of cutaneous leishmaniasis caused by Leishmania tropica consist of self-healing skin lesions. However, a small minority of patients develop persistent or diffuse forms of the disease. The variation in outcome can be reproduced successfully in inbred mouse strains which vary greatly in their resistance to L. lropica. B A L B / c mice are exceptionally susceptible in that the disease is induced with even minimal infecting doses and is inexorably progressive, terminating invariably in cutaneous and fatal visceral metastasis. Innate resistance to L. tropica is controlled by a single autosomal gene dissimilar to the Lsh gene involved in L. donovani infection and not linked to the H-2 complex, although H-2 genes may be involved late in the infection2% Experiments using reciprocal chimerism between 11-2 compatible pairs of strains which differ widely in their susceptibility to L. lropica have clearly established that the major regulatory gene is expressed intrinsically in some population(s) of haematopoietically derived cells, and is not influenced by the chimeric host's environment 27. The likelihood that the cell implicated is a macrophage, by analogy with L. donovani infection, is currently receiving experimental support (R. M. Gorczynski and S. MacRae, submitted). In L. tropica infection, cell-mediated immunity rather than humoral antibody appears to be the most important protective mechanism. In adoptive transfer experiments in mice, it has been clearly demonstrated that serum is not protective 2s whereas immune T cells of Lyt-f +2 Ia phenotype, analogous to T Dcells confer a significant level of protection 29. The lack of immunity in susceptible mice is due not to a failure to elicit a D T H reaction but to its suppression by the induction of L. tropica-specific T s cells. Indeed, experimental procedures such as cyclophosphamide pretreatment, adult thymectomy 3°, and sublethal doses of irradiation (550 rads) which preferentially impair T s cell precursors, restore the D T H reactivity resulting in healing in susceptible B A L B / c mice to L. tropica infcction3L Conversely, Lyt-l+2 - T s ceils from acutely

infected susceptible mice inhibit the in-vivo induction as well as the expression of D T H to L. tropica and reverse the resolution of lesions in 550 rads irradiated B A L B / c mice 3°,31 (F. Y. Liew, C. Hale, and J. G. Howard, submitted). Thus, the following sequence of events may explain this genetically determined susceptibility to cutaneous leishmaniasis. The primary cell target for invading promastigotes is the mononuclear phagocytic cell, which, in the majority of cases, is capable of controlling proliferation of the parasite in its amastigote stage. Processed antigens stimulate T~ cell precursors to T D effector cells, which upon encountering parasitic antigens, 'activate' macrophages via lymphokines to eliminate amastigotes even more effectively. A single major regulatory gene determining L. tropica susceptibility and involving a primary macrophage 'defect' could lead to rapid amastigote accumulation (Fig 2). Evidence has now been obtained that such a high antigen load leads preferentially to the generation of T s cells (R. M. Gorczynski and S. M a c R a e , submitted) which profoundly impair the induction and expression of potentially curative ceil-mediated immunity. Despite the Lyt-1 ~2- phenotype, these T s cells are unlikely to be analogous to the T H cells for humoral response. Sublethally irradiated and infected BALB/c mice injected with such cells did not produce antibody quantitatively or qualitatively different from unreconstituted mice during the critical first 45 days after infection, despite the fact that the resolution of lesions in these irradiated mice was completely reversed by the cell reconstitution (F. Y. Liew, C. Hale and J. G. Howard, submitted). Thus, the outcome of cutaneous leishmania infection depends largely on the delicate balance between T s cells and T D cells determined by macrophage

AMASTIGOTE

,.o..,.,oo.,

//"

Fig. 2 Regulatory interaction of T D and T s cells leading to

susceptibility of BALB/c mice to L. tropiea. Interupted arrows rcpresent inhibition. TSp = prccuvsor of TS, TSm - memory ceiL, TDp - precursor of TI).

Immunology TodayvoL 3, .No. I 7982 activity in the primary infection which in turn is controlled by an intrinsic regulatory gene. In contrast to L. tropica, mice infected with Trypanosoma cruzi (causing Chagas' disease in humans) rarely develop a significant level of specific D T H , although mice immunized with killed T. cruzi antigen produce a classical D T H response. Indeed, antigenspecific T s cells which inhibit D T H to T. cruzi can easily be seen in mice chronically infected with T. cruzi 32. Specific antibodies appear to play a major role in protective immunity against acute T. cruzi infection, a feature characteristic of most haematoparasites. However, a unique feature of Chagas' disease is the prolonged chronic infection. The acute form of human Chagas' disease is rare and usually subsides in 2-4 months, after which the disease enters its latent phase with persisting low levels of parasitaemia, resulting in chronic Chagas' cardiomyopathy. Pathological tissue lesions of chronic Chagas' disease are currently considered to result from cell-mediated immune attack by T. cruzi-sensitized T cells against tissues exhibiting T. cruzi cross-reacting antigens 33. This is supported by the fact that in Chagas' patients pathology is associated with strong D T H reactivity 34,35. Thus, infection-induced suppression of D T H to T. cruzi may have the following important implications. Firstly, the persistent low level of tissue parasites are more susceptible to cell-mediated immunity and the generation of specific T s cells could contribute towards the avoidance of immune destruction by the intracellular parasites. Secondly, the suppression of cell-mediated immunity would also minimise host mortality through reduced pathology and thereby further ensure parasite survivaP 2.

D T H to alloantigens D T H to histocompatibility antigens has been reported since the early days of transplantation immunology. It was originally suggested by Medawar and his colleagues that D T H might be the effector mechanism in graft rejection 36. Recent studies on D T H to the male (H-Y) antigens strongly suggest that this may indeed be so. A complete correlation exists between D T H to H-Y and the ability to reject syngeneic or histoeompatible male skin in all strains of mice studied so far and, in addition, the b-gene control of D T H to H-Y appears to be indistinguishable from that of skin graft rejection 37,38. Furthermore, in an adoptive transfer system, it was clearly shown that Lyt-l+2 cells, closely correlated to those mediating D T H , rather than Lyt-l+2 + T c cells, are responsible for the rejection of skin allografts in lethally irradiated recipients > . Although indefinite survival of skin allografts on neonatally tolerized animals was originally attributed to clonal deletion of relevant antigen-reactive lymphocytes, there is now mounting evidence that this is not the only mechanism whereby self-tolerance is achieved or maintained. Allograft tolerance has been adoptively transferred with living lymphoid cells from tolerant

21 animals suggesting that suppressor cells contribute to the induction and maintenance of transplantation tolerance. Indeed, there is evidence that antigenspecific T s cells are only involved in profound and prolonged tolerance to skin allografts 4°. However, the precise mechanism by which these T s cells manifest their specific inhibitory activity leading to the acceptance of allografts and the prevention of incipient autoimmune diseases is so far unknown. If D T H plays a role in skin allograft rejection and T s cells are involved in tolerance to the grafts, then it would be of considerable interest that antigen-specific T s cells for D T H to histocompatibility antigens are inducible in the same manner as T s cells for D T H to other antigens or pathogens. When mice are injected i.v. with a high dose of Xirradiated allogeneic lymphoid cells, they not only fail to develop D T H to the allogeneic cells, but their ability to respond to an immunogenic challenge of the alloantigens is also significantly depressed 41,42. Such treatment has been shown previously by several groups to significantly prolong the survival of specific skin allografts (e.g. Ref. 43). The suppression of D T H is adoptively transferable by specific T s cells but not by immune serum. The primary T s cells appear to be Lyt-l+2 -, Ia , whereas the secondary T s cells appearing after boosting injection are Lyt-1 +2+ and Ia (Ref. 42). These T s cells localize in the lymphoid organs shortly after their induction and are largely absent from the spleen or lymph nodes one month later. However ~suppressor memory' can be recalled by an immunogenic dose of alloantigens which would normally induce D T H effector cells rather than T s cells in naive mice. When the T s cells are cultured in viLro for 48 h the supernatant contains specific-suppressive activity. It appears likely that the manifestation of the T s function is via soluble antigen-specific suppressor factor(s) the production of which is dependent upon viable T cells 42. Using the same protocol as above, suppression of D T H directed against H-2 subregion products could also be induced, provided that the H-2 incompatibility between the cells used for the induction of suppression and the recipients, includes the I-J subregion genes. Thus, the I-J subregion difference is both necessary and sufficient for the induction of suppression for D T H to the whole or part of the H-2 gene products 44. The suppression appears to be mediated by antigenspecific suppressor T cells which recognise the allo-I-J molecules and are able to suppress the D T H response to other H-2 subregion gene products in an associative recognition manner. T cells from (B10 x BALB/c)F 1 mice suppressed against B10.A(5R) cells (directed against JkEk antigens), when adoptively transferred to normal syngeneic recipients, were capable of suppressing the hosts ~ D T H response to B10.A(4R) cells (KkA k antigens), when the recipients were challenged with (B10.A(4R) x B10.A(5R)) F, cells (KkA k x JkEk). The recipients express normal DTtt[ reactivity to B10.A(4R) cells when challenged with a

22

mixture of B10.A(4R) and B10.A(5R) cells (KkA k + results provide direct evidence that when functioning as alloantigens, the l - j determinants preferentially induce suppressor T cells which specifically impair the immune response to the /-J molecules as well as other t1-2 gene products if they are physically associated with the 1-7 determinants.

JkEk). These

The role of I-J determinants on regulation of DTH Unlike /-A or I-E subregion coded molecules, [-J subregion products have yet to be isolated and characterized. In spite of this, there is now a large body of experimental evidence suggesting that I-Jsubregion products are preferentially associated with immunosuppression./-J determinants, as functionally defined by anti-l-J sera, have been found to be expressed on T s cells and on suppressor factors for humoral and cell-mediated immunities 45. /-.7 subregion genes have also been found to restrict interaction between suppressor T cells and B cells in the humoral response to K L H 4s. 1-,7+ T s cells are implicated in the suppression of immune surveillance against u.v.-induced tumours 46. In mice chronically intected with Schistosoma mansoni, granuloma formation (which is directly related to the D T H response) is reduced by adoptive transfer of T s cells. Effective transference requires homology at the I-J locus between the donor and the recipient. Furthermore, injection of antiserum.specific for the recipients' 1-,7 determinants prevents adoptive transfer of suppression in vivo47. Thus, it appears that I-J subregion genes are intimately associated with immune modulation to exogenous antigens in that the T s cells and TsF bear /-J determinants, and that the manifestation of these suppressive signals requires the recognition of these determinants. When functioning as alloantigens, a situation regarded by many to be an exception to the rules of immune response, 1-J determinants are also involved with immune suppression. Streilein and his colleagues have shown that mice rendered neonatally tolerant of cells bearing a combination of foreign lay and D determinants reject skin grafts which are foreign for the same H-2D alone, but accept the foreign D if that graft also bears the same 1-,7 difference 48. Zinkernage149 found that graft-versus-host reaction against a pure Ij difference significantly suppresses the bactericidal activity of macrophages to Listeria monocylogenes. In an in-vitro system Czitrom el aL 5° showed that I-J gene products suppress the proliferative response to a Dregion difference in a mixed lymphocyte reaction. If lymphocyte transfbrmation is an in-vitro correlate of invivo cutaneous hypersensitivity as is evidenced in bacterial antigen systems, and D T H is involved with the rejection of skin allografts and immunological defence against intracellular parasites, then these results provide strong evidence for the following hypothesis: when suppression involves I-J molecules as foreign alloantigens, the immunological impairment is mediated by suppressor T cells directed

Immunology Today vol. 3, ,No. I 1982

against the /-J determinants. The subsequent suppression against other H-2 gene products is effected in a classical 'carrier-hapten' manner, involving associative recognition through physical linkage between the suppressive epitopes (/-7 molecules) and other H-2 gene products. Thus, I-J subregion genes may serve a dual function: (a) When the host encounters non-H-2 antigens, including pathogens and probably minor histocompatibility antigens, the 1-7 molecules function as guiding signals for immune suppression, in a similar manner as K / D antigens guide T c cells and I-A + I-E molecules TH and TD cells; (b) On the other hand, w h e n / - J determinants are recognised as foreign alloantigens in an unnatural environment such as tissue grafting, they preferentially induce I - J specific T s cells leading to the suppression of immune response to other Ig-2 antigens. It is at present unclear whether this dual function represents the different manifestations of the same gene product or reflects heterogeneity of the 1-,7 subregion gene itself. While the pervasiveness of 1-J determinants in suppression, tolerance and self/non-self recognition is becoming apparent, there are, however, some notable exceptions to this rule which should be highlighted. For example, in the in-vitro systems, GAT-specific helper factors were found to bear I-J determinants5% and some helper T cells to carry I-J molecules s2. Niederhuber and his colIeagues found that /-J subregion compatibility between macrophages and lymphocytes is required for an effective primary antibody response s3. Furthermore, Holan and Hasek 54 failed to obtain any evidence of l-J involvement in neonatal tolerance to skin allografts. These discrepancies are at present unresolved.

Conclusion Recently, there has been a resurgence of interest in the study of DTH, despite some poorly understood aspects of the phenomenon and the unsophisticated nature of available assay systems, as it is now held to play an important role in defence against certain infectious agents and in transplantation immunity. Since these two immune processes, defence against pathogens and rejection of homografts, are directly opposite in their effect on the host, it is vital that D T H should be rigorously regulated in order to provide effective homeostasis. D T H is apparently of little or no use to the host against viral infection yet can induce pathological lesions in vital organs. It is therefore effectively suppressed in most viral infections. In contrast, D T H appears to be a major protective mechanism in most intracellular parasitic and microbial infections. In L. troDica, a regulatory gene defect leading to the generation of T s cells and the subsequent suppression of a potentially curative cellmediated immunity invariably results in fatality in BALB/c mice. In a blood-borne protozoal infection, T. cruzi, where D T H has little or no protective role, yet may be responsible for infection-induced auto-

Immunology To@ no/. 3, Nr). / 1982

i m m u n e disease, effective s u p p r e s s i o n of D T H b y T s cells w o u l d be beneficial to t h e h o s t as well as t h e l o n g t e r m survival of t h e p a r a s i t e . I n the i m m u n e r e s p o n s e to allografts, t h e I-,7 s u b r e g i o n c o d e d d e t e r m i n a n t s a p p e a r to p l a y a pivotal role in its s u p p r e s s i o n . A n u n d e r s t a n d i n g of the m e c h a n i s m b y w h i c h a l l o - I - J m o l e c u l e s m a n i f e s t t h e i r r e g u l a t o r y effect will p o s sibly o p e n the w a y for t h e c o n t r o l of u n t o w a r d i m m u n o l o g i c a l r e a c t i o n s in t r a n s p l a n t a t i o n and autoi m m u n e d i s e a s e or, conversely, for s t r e n g t h e n i n g the h o s t ' s defenee a g a i n s t p a t h o g e n s .

References 1 Zinsser, It. (1921),7" Exp. Med. 34, 495-524 2 Gray, D. F. and Jennings, P. A. (1955) Am. Rev. Tuberc. 72, 171-195 3 Sabolovic, D., Bcugnot, M-C., Dumont, F. A. and Bujadoux, M. (1972) Eur. J. Immunol. 2, 604-606 4 George, M. and Vaughan, J. H. (1962) P~oc. Sac. Exp. Biol. Med. 111,514-521 5 Bianchi, A. T. J., Hoo{ikaas , H., Benner, R., Tecs, R., Nordin, A. A. and Schreier, M. H. (1981).Nature (London) 290, 62-63 6 Lin Yun Lu and Askonas, B. A. (1981).7' Exp. Med. 154, 225-234 7 Asherson, G. L. and Zembala, M. (1974) Proc. Roy. Sac. Land. Ser. B 187, 329-348 8 Phanuphak, P., Moorhead, J. w. and Claman, H. N. (1974) J. Immunol. 113, 1230-1236 9 Ramshaw, I. A., Bretscher, P. A. and Parish, C. R. (1976) Eur. J. Irnmunol. 6, 674-679 10 Lie,v, F. Y. (i977) Eur. J. Immurw/. 7,714-718 11 Greene, M. I., Pierres, A., Doff, M. E. and Benacerraf, B. (1977). 7. Exp. Med. 146,293-296 12 Liew, F. Y., Sia, D. Y., Parish, C. R. and McKcnzie, 1. F. C. (1980) Eur. J. lmmnnol. 10, 305-309 13 Sy, M-S., Nisonoff, A., Germain R. N., Benacerraf, B. and Greene, M. 1. (1981 ) J. Exp. Med. 153, 1415-1425 14 Liew, F. Y., Russell, S. M. and Brand, C. M. (1979) Eur. J. lmmunol. 9,783 790 15 Zinkernagel, R. M. (1976).)*. Evp. Med. 144, 776-787 16 Weiner, H. L., C-reene, M. I. and Fields, B. N. (1980).7. lmmunol. 125,278-282 17 Leung, K. N., Ada, G. L. and MeKenzie, I. F. C. (1980).7. F.xp. Med. 151,815-826 18 Nash, A. A., Phelan, .J. and Wildy, P. (1981).7. lmmurw/. 126, 1260-1262 19 Liew, F. Y. and Russell, S. M. (1980).7. E.~;b. Med. 151, 799-814 20 Greene, M. I. and Weiner, H. L. (1980) J. Iramurwl. 125, 283 287 21 Nash, A. A. and Gell, P. G. 11. (1981) lmmunol. Today 2, 162-165

23 22 Licw, F. Y. (1980) in Immunological Recognition and AfJ~clor Mechanism in Ir!feclious Diseares Chapter 1, pp. 1-12, Schwabe and Co. A. G., Basel 23 Leung, K. N. and Ada, G. L. (1981) 7. Exp. Med. 153, 1029-1043 24 Haekett, C. J., Askonas, B. A., Webster, R. G. and Van Wyke, K. (1980)J. Exp. Med. 151, 1014-1025 25 Leung, K. N. and Ada, G. L. (1980) &'and..7. lnzrnunol. 12, 393-400 26 Howard, J. G., Hale, C. and Chan-Liew, W. l.. (1980) Parasite Immunol. 2, 303-314 27 Howard, J. G., Hale, C. and Liew, F. Y. (1980))¢bZure (London) 288, 161-162 28 Preston, P. M. and Dumonde, D. C. (1976) C[in. Exp. Immured. 23, 126-138 29 Mitchell, G. F., Curtis, J. M., Handman, E. and McKenzie, I. F. C. (I980) aurl.y.E.~-p. Biol. Med. ,Di. 58, 521-532 30 Howard, J. G., tIale, C. and Liew, F. Y. (1980)J. t:'xp. Med. 152, 594-607 31 Howard, J. G., Hale, C. and Liew, F. Y. (1981).7. Erp. Med. 153,557-568 32 Scott, M. T. (1981)bnmu~wlogy 44, 409-417 33 Santos-Buch, C. A. (1979) [r~l. Rev. Exp. Pathol. 19, 63-100 34 Teixeira, A. R. L., Teixeira, G., Macedo, V. and Prata, A. (1978) J. Chr~. Invest. 62, 1132 1141 35 Teledo Barros, M. A. M., Neto, V. A., Mendes, E. And Mota, I. (1979) Clin. Exp. lmmunol. 38,376-380 36 Brent, L., Brown, J. and Medawar, P. B. (1958) Lamer 13, 561-564 37 Liew, F. Y. and Simpson, E. (1980) Immum~genetics 11,255-266 38 Greene, M. 1., Benacerraf, B. and Dorf, M. E. (1980) lmmorwgenetics 11,267-275 39 Loveland, B. E., Hogarth, P. M., Cevedig, R. H. and MeKenzie, I. F. C. (1981).7. Exp. Med. 153, 1044-1057 40 Smith, R. N. and Howard, J. C. (1980) J. lmmzm¢d. 125, 2289-2294 41 Van der Kwast, T. I1., Bianchi, A. T. J., Bril, H and Benner. R. (1981) Transplantation 31, 79-85 42 Liew, F. Y. Tran~]~lanlation (in press) 43 Hilgert, I. (1979) [rnmunol. Rev. 46, 27-53 44 Liew, F. Y. Eur. J. hnmunol. (in press) 45 Tada, T. and Okumura, K. (1979) Adv. lntmunol. 28, 1 87 46 Greene, M. 1., Perry, L. I.., Sy, M. S. and Bromherg, j. s. (t981) Immunol. 7bday 2, 23-25 47 Grcen, W. F. and Colley, D. G. (1981) Pn~c..Vat/ A~ad..b'~t {LS./I. 78, 1152-1156 48 Streilein, J. W. (1979) hnmurw/. Rev. 46, 125-146 49 Zinkernagel, R. M. (1980) bnrnunogenetics 10,373-382 50 Czitrom, A. A., Sunshine, G. H. and iMitchison, N. A. (1980) lmmnnogeneticr 11, 97 102 5t Howie, S., Parish, C. R., David, C. S., McKenzie, 1 F. C , Maurer, P. lI. and Feldmann, M. (I979) Ear..7. lmmtmo/. 9. 501-506 52 Tada, T., Takemori, T., Okumura, K., Nonaka, M and Tokuhisa, T. (1978).7- l~2xP. Med. 147,446 458 53 Niederhuber, J. E. (1978) Imma~wl. Rev. 40, 28 52 54 Holan, V. and Hasek, M. (1981 ) [mmnrwgeneti~ 12, 46t)-472