Mechanisms of immunity to malaria*

Bull. Org. mond. Sante 1974, 50, 251-257Bull. Wld Hlth Org. J Mechanisms of immunity to malaria* S. COHEN, G. A. BUTCHER, & G. H. MITCHELL The erythrocytic phase ofmalarial infection provides apotent stimulusfor theproduction ofspecific malarial antibody. Serologic tests do not provide any indication ofimmune status, so that much specific antibody has no protective function. The role of serum antibody in acquired malarial immunity has, however, been established by passive transfer tests, and in the case of P. knowlesi by specific inhibition of the cyclic growth ofparasites in vitro. The inhibitory antibody appears to combine with merozoites and prevents their attachment to red cells, thus interrupting the cyclic proliferation ofthe parasite. The inhibitory antibody response is predominantly variant-specific, but cross-reacting antibody occurs in sufficient amount to suppress proliferation of most other variants of the species. The occurrence of cross-immunity between variants is encouraging from the point of view of vaccination. If it were possible to isolate cross-reacting antigens, these could provide the basis for a malarial vaccine effective against erythrocytic forms of the parasite. The clinical manifestations of malaria are, in many instances, effectively controlled by the specific immune response of the host. Clinical and experi- mental observations and, in particular, in vitro studies on cultured parasites have elucidated the nature of these immune protective mechanisms. However, acquired malarial immunity is often slow to develop and in many species is associated with persistent low-grade parasitaemia, a phenomenon referred to as " premunition ". It is therefore apparent that plasmodia effectively sensitize the host's immune system and yet may survive for long periods. The mechanisms that allow parasites to evade the potentially lethal consequences of im- munization remain a central problem of malarial immunity. In this paper, our present understanding of the nature of immune effector responses is outlined; the nature of malarial immunity and the mechanisms for its evasion by parasites are discussed; and finally, on the basis of these findings, the outlook for malarial vaccination is briefly assessed. IMMUNE EFFECTOR MECHANISMS Adaptive immune responses are initiated by the interaction of an antigen with specific receptors on the surfaces of lymphoid cells. Two major classes of lymphocytes are recognized and these react differ- * From the Department of Chemical Pathology, Guy's Hospital Medical School, London SEI 9RT, England. This paper was presented at the Symposium on Malaria Research, Rabat, Marocco, 1-5 April 1974. ently after induction by the antigen. T-lymphocytes undergo transformation and mitosis to produce a population of cells specifically reactive with the inducing antigen; B-lymphocytes differentiate into plasma cells which secrete humoral antibody, but this reaction usually requires cooperation with T-cells. The effector mechanisms generated by these cellular responses (summarized in Table 1) therefore fall into two major categories. The first is mediated by specific humoral antibody and the second by specifically sensitized T-lymphocytes (cell-mediated immunity). While it is true that antibody action may, in certain circumstances, be independent of cells (reactions 1 and 2, Table 1) and that certain specific cell-mediated reactions are independent of antibody (reaction 6, Table 1), there remain several immune effector mechanisms that require the interaction of specific antibody with cells such as macrophages (reactions 3 and 4, Table 1) or mast cells (reaction 5, Table 1). In addition, specific interaction of antigen with T lymphocytes may liberate cytotoxins and macrophage activating factors and so damage cells or organisms that do not carry the specific antigen (reactions 7 and 8, Table 1). THE IMMUNE RESPONSE TO MALARIA Specific malarial antibody Chronic malarial infection in man provides a most potent stimulus for Ig synthesis (1, 2, 3). The reduced IgG production observed in West Africans main- 3186 --251- 8 S. COHEN ET AL. Table 1. Immune effector mechanisms SPECIFIC MECHANISMS Re- Effector lg Effector cell Effector complex b Biological effectaction 1 all classes - Ant-Ab neutralization activation 2 IgG3 IgGl IgG2 - Ant + Ab + C complement- 1gM dependent cytolysis 3 IgGI IgG3 IgM macrophage Ant + Ab(+C) combined immune adherence SMAF a (phagocytic) with macrophage endocytosis 4 Ig K-cell Ant + Ab combined with Ab-mediated(nonphagocytic) K cell (nonphagocytic) cytolysis 5 IgE mast-cell Ant + Ab combined immediate hyper- with mast cell sensitivity 6 - T-cell Ant + T-cell cell-mediated cytotoxicity NONSPECIFIC MECHANISMS action Cooperating cell Effector cell Effector complex Biological effect 7 T-cell Ant + T-cell-- nonspecific cytotoxin lymphocyte- mediated cytotoxicity 8 T-cell macrophage Ant + T-cell--o nonspecific Mediators- macrophage- activated macrophage mediated cytotoxicity a SMAF = specific macrophage activating factor, a cytophilic product of T cells with antigen specificity. b Ant = antigen; Ab = antibody; C = complement. tained on malarial prophylactic therapy (3) suggests that in subjects exposed to chronic infection about a third of the circulating immunoglobulin might consist of malarial antibody. Specific antibody can be demonstrated in immune sera by several serologic tests including precipitation, agglutination, opsoniza- tion, antibody fluorescence, and complement fixa- tion. The fact that the serologic cross-reactions between malarial species revealed by these tests cannot be correlated with cross-immunity indicates that much of the specific antibody formed during infection does not have a protective function. Protective malarial antibody Acquired immunity in malaria is directed mainly against the asexual parasite cycle in the blood. Circulating gametocytes of P. falciparum are appa- rently unaffected by immune serum (3), although in some monkey malarias immunity is associated with a loss of gametocyte viability in the mosquito host. Immunity to the erythrocytic stage of infection does not modify the exoerythrocytic development of malaria parasites in man, chimpanzee, or monkey (4), but suppresses the subsequent phase of erythrocytic development. The role of serum antibody in acquired malarial immunity has been established by passive transfer tests in monkey and human infections, and, to a lesser degree, in rodent malaria (5). These studies suggested that protective antibody acts against either mature schizonts or extracellular merozoites, and provided some information about the classes of Ig associated with immune protection (3). However, passive transfer tests do not provide a suitable basis for detailed investigations on the mechanism of malarial immunity. Serum from rhesus monkeys immune to P. knowlesi has been shown to inhibit the cyclic pro- liferation of the parasite maintained in vitro (6). Parasite growth was assessed by incorporation of 3H-leucine into parasite protein using cultures giving average multiplication rates of at least sixfold in 24 hours. Immune serum had no effect upon the growth of intracellular parasites but inhibited the 252 MECHANISMS OF IMMUNITY TO MALARIA cycle of development that followed schizogony. This effect was species-specific, as was shown by the failure of serum from a monkey immune to P. cynomolgi bastianellii to inhibit P. knowlesi. The immune serum appears to act directly on the para- sites (Table 1, reaction 1), and its effect is not complement-dependent (Table 1, reaction 2) being unaffected either by heating at 56°C for 3 h or by the addition of fresh normal monkey serum as a source of complement; moreover, the F(ab'), fragment of immune IgG is actively antiparasitic. The degree of parasite inhibition is dose-dependent. In the sera studied, protective antibody was associ- ated with both IgG and IgM, but not with IgA or IgE (7). The failure to demonstrate activity in monkey IgE is of interest because levels of IgE are high in human and monkey parasitic infections, and this, together with earlier observations on the common occurrence of immediate type sensitivity, had suggested a possible protective role for IgE in parasitic diseases (Table 1, reaction 5). There has long been a need in malarial research for a dependable in vitro technique for the detection of protective malarial antibody. The schizont- infected red cell agglutination test (SICA) (9) and the assay of inhibitory antibody described above have been proposed for this purpose (6). Comparison of these antibody activities with the clinical immune status of rhesus monkeys (18) reveals that SICA antibody titres of appropriate specificity may be high in susceptible animals or undetectable in immune animals; on the other hand, inhibitory antibody levels correlate with immune status in all situations that have been tested (Table 2). These observations indicate that the inhibitory antibodies assayed during in vitro culture initiate specific protective immunity in vivo. This antibody is probably directed against merozoites and its action in vitro is independent of complement or cells and is analogous to viral neutralizing antibody (Table 1, reaction 1). Inhibitory antibodies are predominantly IgG and, since this class is cytophilic for macrophages and K-cells, it seems certain that cell-mediated killing of parasites must also occur in the living animal (Table 1, reactions 3 and 4). The phagocytic activity of macrophages has long been recognized in malaria, and the role of specific antibody in promoting macrophage ingestion of parasites has been demonstrated in vitro and in vivo. Synergism between cells and antibody is suggested by the finding that immune splenic cells confer greater protection than serum when passively transferred to rats challenged with P. berghei (14). Similarly, the antimalarial action of passively transferred immune serum in rats is greatly dimi- nished by previous splenectomy of the normal reci- pients (22). Role of cell-mediated immunity in malaria The proven role of serum antibody, outlined above, does not exclude the possibility that cell- mediated immunity, dependent upon actively sen- sitized lymphocytes of thymic origin (Table 1, reaction 6), may play a part in specific acquired resistance to malaria. Attempts to demonstrate this have rested upon the following findings: (a) Thymectomized rats are more susceptible than control animals to subsequent infection with P. berghei (10). Table 2. Clinical immunity to P. knowlesi variants in relation to SICA and inhibitory antibody titres in rhesus monkeys [from Butcher & Cohen (18)] Antibody to challenge Immunization a Challenge variant Clinicalvariant immunity SICA Inhibitory repeated infection with WI strain Wi ++ ++ + repeated infection with Wl strain W2 0 + + repeated infection with G strain Wi 0 ++ + one infection with Wl Wi ++ + or 0 + or 0 a P. knowlesi strains: W, supplied by Walter Reed Army Institute; G, supplied by London School of Hygiene and Tropical Medicine. 253 S. COHEN ET AL. (b) Anti-rat thymocyte serum (ATS) raised in rabbits reduces the resistance of rats to subsequent infection with P. berghei (11). The interpretation of these findings solely in terms of cell-mediated responses is complicated by the fact that the majority of antigens are now known to require cooperation between thymus (T) and marrow (B) cells for induction of specific serum antibody. It follows that thymectomy or ATS may render animals susceptible to malarial infection by reducing the serum antibody response. (c) Lymphoid cells transferred from immune rats to inbred nonimmune animals confer resistance to P. berghei (12). This finding cannot, however, be taken as evidence for cell-mediated immunity alone since such cells are capable of producing specific antimalarial antibody (13). Since the published studies (14) and our experi- ments show that immune lymphocytes are not anti- parasitic when tested in vitro, it must be concluded that cell-mediated immunity has no defined role in specific malarial resistance. SURVIVAL OF PARASITES IN THE IMMUNIZED HOST The mechanisms that lead to a state of " pre- munition" in malaria, i.e., clinical immunity asso- ciated with continued low-grade infection, have long been a subject of speculation. Interest in this phe- nomenon is increased by recent data revealing the potent immunogenicity of the parasite, the successful passive transfer of immunity (3), and the effective in vitro inhibition of cyclic plasmodial growth by specific antibody (6). Under natural conditions, the apparent survival of the organism may in fact be related to reinfection with distinct strains of the species or represent a relapse derived from persistent preerythrocytic stages of the plasmodium. In addition, however, laboratory studies have shown that recrudescences do arise from preexisting systemic infection. Such recrudescence could be accounted for by soluble immune complexes or free antigen blocking the action of antibody in a manner analogous to the blocking associated with neoplasms. Alternatively, the immunosuppressive effect of malarial infection or the action of protozoal factors that specifically inhibit host defence mechanisms (15) might lead to recrudescent infection. One significant mechanism for parasite survival is connected with the ability of the organism to undergo antigenic variation. The occurrence of antigenic variants during the course of the asexual cycle (5) has been described in P. berghei infections of mice, in the monkey malaria parasites P. knowlesi and P. cynomolgi bastianellii, in the avian malaria parasites P. lophurae and P. gallinaceum, and in the human malaria parasite P. falciparum. In P. knowlesi, a wide spectrum of antigenic variants is recognizable on the basis of the schizont agglutination (SICA) test (9). Schizont agglutinins are also present in P. gallinaceum and P. falciparum infections, but cannot be demonstrated in many other species in which the occurrence of serologic variation is deduced from infectivity tests. Antigenic variation could account for many features of premunition mentioned above, since each relapse during a chronic P. knowlesi infection is associated with a distinct variant (9). Recent experiments indicate that SICA antibodies, which as mentioned above are not protective, induce antigenic variation of P. knowlesi (Table 1, inductive reactions) in a manner analogous to that observed in free-living ciliates (16). These antibodies may therefore promote the survival of plasmodia in the immunized host. However, other mechanisms must also be operative since SICA antibodies cannot be detected in many malarial species that show antigenic variation. The importance of antigenic variation as a cause of relapse must be assessed in the light of the follow- ing facts. (a) Spontaneous relapses in P. knowlesi infections are usually well controlled by the immunized host and parasitaemia rarely rises above 1 %; nevertheless, the relapse variant produces a rapidly fatal infection if inoculated into an unimmunized recipient. (b) After repeated challenge with a single variant of P. knowlesi, rhesus monkeys are equally resistant to the same and several other variants (17, 18). For example, monkeys repeatedly challenged with a single variant (Wl) of P. knowlesi were immune to WI and several serologically distinct variants (e.g. W2, W3). Sera from these animals showed in- hibitory antibody directed predominantly against the specific variants to which the animal had been exposed, but also contained lower levels of cross- reacting antibody directed against variants that had never been patent. Animals were therefore sensitized for a rapid protective antibody response when in- fected with new variants. The presence of this cross-reacting antibody explains why P. knowlesi parasites, which arise by antigenic variation during chronic infections, produce mild parasitaemia in the host and yet are fully virulent in normal monkeys. 254 MECHANISMS OF IMMUNITY TO MALARIA 255 PROSPECTS FOR VACCINATION AGAINST ERYTHROCYTIC FORMS OF MALARIA An ability to evade the lethal consequences of the host's immune response constitutes an essential requirement for obligate parasitism (15). As shown above, parasites sensitize the host's immune system and their survival depends upon evasion of effector immune mechanisms. Our knowledge of how this is achieved in malaria is incomplete, but in the case of P. knowlesi the importance of antigenic variation is established. Other possible mechanisms, including the presence of serum blocking factors, have not been investigated. The circumstances that favour parasite survival in natural infections may be equally operative in determining the outcome of artificial vaccination and it is not surprising that attempts to induce immunity have achieved only limited success. The occurrence of cross-immunization between variants of P. knowlesi is however encouraging from the point of view of vaccine production. A vaccine containing cross-sensitizing antigens may be expected to induce a degree of clinical immunity similar to that observed after chronic infection. Our own search for such antigens involves (a) the isolation of soluble plasmodial fractions that effectively absorb out inhibitory antibody of several variant specificities from immune sera (20) and (b) the production of extracellular merozoites in high yields from malaria cultures using appropriate antisera or lectins to precipitate red cells and intracellular parasites (21). The immunizing properties of these protective antigen fractions and pure merozoite preparations are at present being tested in this laboratory. ACKNOWLEDGEMENTS This work was supported by grants from the Medical Research Council, London, and the World Health Organization. RtSUMt MECANISMES DE L'IMMUNITIt DANS LE PALUDISME La phase erythrocytaire de l'infection palud&enne est un puissant stimulant de la production d'anticorps antipaludiques specifiques. Les epreuves serologiques ne fournissent aucune indication sur le degre d'immunit6, ce qui montre qu'une grande partie des anticorps ne jouent aucun r6le protecteur. La contribution des anticorps seriques a l'immunite acquise a cependant 'te mise en evidence par des 6preuves de transfert passif et, dans le cas de Plasmodium knowlesi, par l'inhibition sp6cifique de 1'evolution cyclique du parasite in vitro. En se com- binant avec les m6rozoItes, les anticorps inhibiteurs les empechent d'adherer aux erythrocytes et interrompent la multiplication cyclique du parasite. La reponse en anti- corps inhibiteurs est essentiellement specifique de variant, mais la production d'anticorps a reactivit6 crois&e est suffisante pour inhiber la prolif6ration de la plupart des autres variants d'une espece. L'existence d'une telle immunite crois6e est un facteur encourageant en ce qui regarde la vaccination. Si l'on parvient a isoler des antigenes a antigenicit6 croisee, on disposera des elements de base pour la mise au point d'un vaccin actif contre les formes erythrocytaires du parasite. REFERENCES 1. TOBE, J. E. ET AL. J. Immunol., 97: 498-505 (1966). 2. RowE, D. S. ET AL. Clin. exp. Immunol., 3: 63-79 (1968). 3. COHEN, S. ET AL. Nature, 192: 733-37 (1961). 4. GARNHAM, P. C. C. Primate malaria. In: Jackson, G. J., Herman, R. & Singer, I. ed. Immunity to parasitic animals, Vol. 2, New York, Appleton- Century-Crofts, 1970. 5. BROWN, I. N. Adv. Immunol., 11: 267-349 (1969). 6. COHEN, S. ET AL. Nature, 223: 368-71 (1969)., 7. COHEN, S. & BUTCHER, G. A. Immunology 19: 369-83 (1970). 8. EATON, M. D. J. exp. Med., 67: 857-69 (1938). 9. BROWN, K. N. & BROWN, I. N. Nature, 208: 1286 (1965). 10. BROWN, I. N. ET AL. Nature, 219: 292-93 (1968). 256 S. COHEN ET AL. 11. SPIRA, D. T. ET AL. Immunology, 19: 759-66 (1970). 12. STECHSCHULTE, D. J. Milit. Med., 134 (Suppl.): 1147-52 (1969). 13. PHILLIPS, R. S. Exp. Parasitol., 27: 479-95 (1970). 14. PHILLIPS, R. S. ET AL. Exp. Parasitol., 28: 339-55 (1970). 15. COHEN, S. Immunoprophylaxis of protozoal diseases. In, Gell, P. G. H., Coombs, R. R. A. & Lachman, P. Clinical aspects of immunology, 1974 (in press). Oxford, Blackwell. 16. BROWN, K. N. Nature, 242: 49-50 (1973). 17. VOLLER, A. & ROSSAN, R. N. Trans. roy. Soc. trop. Med. Hyg., 63: 507-23 (1969). 18. BUTCHER, G. A. & COHEN, S. Immunology, 23: 503-21 (1972). 19. BUTCHER, G. A. & COHEN, S. Trans. roy. Soc. trop. Med. Hyg., 64: 470 (1971). 20. COHEN, S. ET AL. Proc. Helmint. Soc., Washington, 39: 231-37 (1972). 21. MrrCHELL, G. H. ET AL. Int. J. Parasit., 3: 443-45 (1973). 22. ZUCKERMAN, A., GOLENSER, Y. & SPA, D. T. Proc. 9th Int. Cong. trop. Med., p. 276 (1973). DISCUSSION MOLINEAUX: With respect to the persistence of low parasite densities among immune persons in endemic areas, I would argue that the circulating population of parasites must be equal to the ratio between the number added per unit time and their death rate. It would therefore be zero only if the rate of addition became zero or if the death rate became infinite; if immunity effects neither, that by itself could explain persistence of a small parasite population. COHEN: I doubt whether sucn a rigorous but simple mathematical model could be applicable to such a complex problem. BRUCE-CHWATr: What is the explanation of the fact that the injection of heterologous antilymphocyte serum into Balb/C mice infected with P. berghei yoelii (which normally controls the infection) in- creases parasitaemia and lengthens the course of the infection without changing the immunofluorescent antibody pattern of immune response? COHEN: Antilymphocyte serum acts against T-lym- phocytes. Antibody response requires the coopera- tion of T-cells-hence if T-cell populations are des- troyed no further antibody response of that particu- lar type occurs. But, even after the injection of antilymphocyte serum, the fluorescent antibody re- sponse would persist for some time because the serum has no effect on plasma cells already induced to produce antibody. NuSSENZWEIG: How can one explain the higher inhibitory antibody titres against the heterologous P. knowlesi strain in the laboratory-bred natural host and the absence of heterologous reactions in rhesus ? Does the rhesus monkey respond to a more restricted population of antigenic determinants? COHEN: The levels of antibody in the rhesus dur- ing the first week of infection probably do not reach significant levels. Those in the natural host are significant and of broad specificity. NUSSENZWEIG: Are merozoites inactivated more rapidly by immune serum than the time taken for them to invade a new erythrocyte and is the activa- tion complete or can invasion occur in immune serum? COHEN: Merozoites attach themselves to erythro- cytes in 10-15 minutes. Agglutination is faster and prevents attachment. LuzZATrO: How is it determined whether merozoites penetrate erythrocytes and is there early destruction of the parasites after penetration? COHEN: In in vitro culture, penetration of new cells does not occur in immune serum. Some penetration did occur, with bush-baby erythrocytes, but the parasites did not develop. BRAY: The erythrocyte is not the only area of susceptibility. First there is the tissue barrier-i.e., the hepatic cells in mammalian malaria. Then there is penetration of erythrocytes but no development un- less the spleen is removed, as is the case with P. falciparum in the chimpanzee. Is there any evi- dence of acquired resistance to invasion by an erythrocyte already invaded? COHEN: No, the erythrocyte-unlike the fertilized ovum-can be invaded again. BRAY: Is there any correlation between McGregor's and Wilson's analysis of placental antigen and an antigen giving rise to Cohen's antimerozoite glob- ulin? MECHANISMS OF IMMUNITY TO MALARIA MCGREGOR: There is a group of antibodies that pass across the placenta and react with heat-labile anti- gens. These would be our choice for further investi- gation in this respect. VAN DER KAAY: Is there any evidence of the prolon- gation of the asexual cycle in immune serum? COHEN: We have not found any. MEUWISSEN: Why are so many schizonts found in the placenta of pregnant women who are immune? Are there conditions in which a merozoite from one erythrocyte can pass to another erythrocyte through immune serum, for instance in the clogged capillary of the brain in falciparum malaria? COHEN: The question of the immune response in pregnancy is very difficult. Some responses may be suppressed. Whether a travelling merozoite would effectively combine with an erythrocyte or with antibody is a question of relative concentrations and affinities of reactants and cannot be answered in simple terms. BRAY: Does Professor Cohen mean that cross-im- munity between variants resides in the merozoites and their variant common antigens, whereas Neil Brown ascribes the cross-immunity to T-cell memory and to T-B-cell interaction? COHEN: It depends on what antibody one thinks is protective. We think antimerozoite antibody is effec- tive. Neil Brown thinks schizont agglutinating anti- body is effective and has produced the T-cell theory to explain observed clinical cross-immunity. GRAMICCIA: If parasitization of red blood cells is due to the presence of available receptors, why does splenectomy cause susceptibility in some normally nonsusceptible species of host and why does a single sojourn in the splenectomized host then give rise to a strain of parasites that will infect the intact host? Does this mean that there can be changes in the merozoite receptors after experiencing a new host? COHEN: One can only say that splenectomy must eliminate the early antibody responses, both natural and acquired. There may be a range of affinities on the surface of merozoites and selection for high affinity receptors may occur during infection. 257