Immunopathology of malaria*

Bull. Org. mond.Sant: 1974, 50, 177-186 Bull. Wld Hlth Org. Immunopathology of malaria* ALISTER VOLLER 1 Antibodies with different spectra of reactivity are produced during malarial infections and marked changes in IgG and IgM levels occur. In addition malaria elicits serological changes that are usually associated with connective tissue disease. The excessive anaemia associated with malaria may, in part, be an autoimmune phenomenon. Transient nephritis accompanies many plasmodial infections but chronic malarial nephrotic syndrome is specifically associated with quartan malaria. Malarial infection leads to splenomegaly, the most extreme form of which is idiopathic tropical splenomegaly, which probably represents an aberrant immune response to the infection. Malaria can affect the humoral immune response to unrelated antigens and infectious agents. This may be relevant to the etiology of Burkitt's lymphoma. During pregnancy there is some loss of acquired immunity to P. falciparum and the placenta appears to be an immunologically privileged site for the multiplication of this parasite. In this paper an attempt is made to examine those pathological events that involve the host's immune response to the malaria parasite or its antigens. The direct effects of the parasite on the host are not considered. COURSE OF INFECTION Under natural conditions malaria infections are initiated when an infected mosquito injects the spo- rozoites as it feeds. The sporozoites then circulate for only a few minutes before becoming localized in the liver where subsequent development of the pre- and exo-erythrocytic forms takes place. There seems to be no immunological response to the injected viable sporozoites under natural conditions. This is rather surprising as it has been shown that injected spo- rozoites modified by radiation or chemical treatment can lead to a serological response and to the develop- ment of immunity to subsequent challenge by viable sporozoites (1). Possibly even more surprising is the fact that the pre- and exo-erythrocytic forms of the malaria parasite developing in the liver also seem to provoke no immunological response until they reach maturity. At the time of rupture of the tissue schizont * Presented at the Symposium on Malaria Research, Rabat, Morocco, 1-5 April 1974. 1 Senior Lecturer, Department of Clinical Tropical Medi- cine, London School of Hygiene and Tropical Medicine; and Nufflield Institute of Comparative Medicine, The Zoological Society of London, England. polymorphonuclear leucocytes invade the site, to be followed by other cellular infiltrates, which decline in a short time. However, no serum antibodies are detectable at this time. It is only when the invasion of the blood occurs that the full immunopathological response is ini- tiated. Each asexual cycle of parasitaemia in the blood consists of broods of parasites that begin as small rings and mature to large parasites forming schizonts, each containing many merozoites. These are released on the rupture of the host cell, to infect more red cells. In most malaria infections the theo- retical rate of multiplication is rarely reached even in the early stages of invasion of the blood; this is ascribed to the removal of the parasites and infected cells by normal phagocytic activities, especially in the spleen. Usually the parasitaemia increases up to a certain point and then decreases abruptly. This point where the host's defences have just gained control is known as the crisis and can lead to a very sudden drop in the number of parasites. At this point there is intense reticuloendothelial hyperplasia and reactivity, many of the macrophages being loaded with malarial pigrnent and other products resulting from the phagocytosis of malarial parasites. MALARIAL ANTIBODIES The phagocytosis of the malaria parasites at crisis or thereafter is not just an increase in the usual 3177 - 177 A. VOLLER nonspecific activity of macrophages but is due to a highly specific opsonization process. Tests made in vitro showed that schizont-infected red cells were phagocytosed at a high rate only when sensitized with the variant specific antibody (2). Such highly specific antibody with a capacity for discrimination even between the parasite populations (variants) of different relapses had already been detected by in vitro agglutination tests (3). The agglutination can be shown in vivo by the transfusion of immune serum into animals infected with the appropriate variant. Whether the massive concentrations of P. falciparum trophozoites and schizonts that are found in the deep capillaries are, to some extent, due to immune agglutination is not known at present. The explana- tion is certainly not entirely immunological as in this case the clumps of parasitized red cells are also found in the capillaries in acute infections of non-immune persons. Variant specific antibodies are also effective in the inhibition of parasite growth in vitro (4). It should not, however, be thought that all malarial antibodies are as highly specific as these, as there are others that have a much wider spectrum of reactivity. The practical serological tests such as immunofluores- cence and haemagglutination are based on malarial antibody reactions that are at best species specific and that often cross-react to a considerable extent with other related malaria parasites. These tests are considered in detail elsewhere in this symposium. The complexity of the antigen-antibody reactions in malaria is now becoming apparent from the results of studies on P. falciparum in The Gambia (5) where gel diffusion analysis revealed that soluble malarial antigens were present in the plasma of infected individuals and that antibodies to them, and to antigens prepared from infected erythrocytes, could be detected in many people resident in malarious areas. There are several classes of antigens with different properties and in some of the classes twenty or more individual antigens have been identified. It is possible that future study will indicate that particular antigens are associated with specific immunopatho- logical events. IMMUNOGLOBULINS Malaria infections are accompanied by marked changes in serum immunoglobulin levels. Induced malaria in nonimmune adults leads to increases in the IgG and IgM fractions (6) and these in part consist of malarial antibodies. Field studies consis- tently show that IgG and IgM levels are especially elevated in tropical areas where malaria in en- demic (7) and there is frequently a strong correlation between total IgM levels and malaria antibody levels, suggesting that malaria is implicated in the excessive IgM response (8, 9). At this point it is appropriate to stress that immunopathological observations made on malaria infections, whether accidental or induced, in non- immune adults are not necessarily applicable to individuals who have been exposed since birth in zones where malaria is endemic. Immunoglobulin and malaria antibody levels tend to increase with age in malarious populations. However, the non-immune adult may develop extremely high levels of immuno- globulins and malarial antibodies after even a single short initial infection. The notion of an age influence in the human immunological response does not generally meet with approval although there is clear evidence from animal studies that the young are more susceptible to malaria than are older animals of the same species. AUTOIMMUNE SEROLOGY AND MALARIA Malaria often elicits serological responses, which-in temperate zones-are usually associated with connective-tissue disease. This may be due to (a) the release of cross-reacting antigen (i.e., parasite material that shares some antigens with the host) or (b) the alteration ofhost tissue, by infection, rendering it antigenic. In either case this could lead to the production of antibodies that react with the sub- strates used in the auto-immune serology. Virtually all the sera of adults in malarious areas contain heterophile agglutinins of the IgM class (10) that react with foreign erythrocytes in vitro or even with trypsinized human erythrocytes (11). Rheu- matoid factor, an antiglobulin, usually to lgG, is also common in malarious areas (12) and the titres in adults often correlate with their malarial antibody levels. Greenwood et al. (12) have even shown that the rheumatoid factor will react in vitro with IgG malarial antibody that has been complexed to malar- ial antigen on slides. Wells (13) put forward a reason- able hypothesis that the heterophile agglutinins are produced in response to parasitized red cells, which have been altered owing to their parasitization or because they are affected by antibody. Similarly he suggests that the antiglobulin response is initiated by 178 IMMUNOPATHOLOGY OF MALARIA the immune globulins rendered antigenic in the malaria antigen-antibody circulating complexes. It could also be argued that antinuclear factor (ANF) (14), which is common in malarious areas, particularly in older individuals, may represent a cross-reacting antibody induced by malarial nuclear material. From the diagnostic point of view these abnormal serological findings are of some relevance as they mean that the usual indicators of auto-immune disease must be interpreted with great care in malari- ous areas. AUTOIMMUNITY AND MALARIAL ANAEMIA Zuckerman (15) has repeatedly stressed that the erythrocyte destruction in many malarial infections is greatly in excess of what could be accounted for by direct parasite rupture of infected cells. She sug- gested that an autoimmune haemolysis or opsoniza- tion of uninfected red cells occurs. If any autoimmune mechanism were involved in malarial anaemia then autoantibodies might be detectable by the usual antiglobulin Coombs tests. Such antiglobulins were not however found in the plasma of malaria-infected animals or people. The difficulties earlier workers encountered in demon- strating the presence of anti-erythrocyte antibodies may have been due to the types of antiglobulin reagents they used, since Rosenberg et al. (16) have recently shown that IgM class anti-erythrocyte anti- bodies occur in P. falciparum infections and can be visualized by immunofluorescence techniques. Topley et al. (17) also reported finding complement on the surfaces of uninfected erythrocytes in rodents with P. berghei infections. Synthesizing this informa- tion Woodruff (18) has suggested that a malarial antibody-antigen reaction may occur on the surface of red cells and that, owing to the involvement of complement, this may lead to haemolysis of the cells. In support of this theory one could cite the fact that there are indeed reduced levels of complement in persons with acute malaria (16), but there are a variety of possible non-immunological explanations. It could even be that the widespread disseminated intravascular coagulation associated with some ma- laria infections could lead to much reduced comple- ment levels. To summarize, the excessive anaemia could be caused by (a) autoantibodies to the erythrocytes or (b) immune adherence of circulating malaria anti- gen-antibody complexes to uninfected erythrocytes. THE KIDNEY There are many early reports of proteinuria being linked with malaria infections in man, and quartan malaria was often mentioned as being particularly involved. It is now realized that there are at least two types of renal involvement of probable immunologi- cal origin: acute transient nephritis and chronic malaria nephrotic syndrome. Transient nephritis This has been reported in P. falciparum infections in man (19) and in P. cynomolgi infections in mon- keys (20) as well as in rodent malaria infections (21, 22). In our own studies (23) it has been shown to be the usual sequel even to quartan malaria infections in Aotus monkeys. Usually a mild to severe proteinuria develops a week or two after infection and renal biopsies taken at this time will show glomerular immunoglobulin, usually IgM, deposits sometimes associated with complement and malarial antigen. After a few weeks the proteinuria resolves and immunological abnormalities are no longer detectable. Quartan malaria nephrotic syndrome The second type of renal disease is the much more important chronic malarial nephrotic syndrome. This is specifically associated with P. malariae infec- tions. Irrefutable proof of this association came from the studies of Gilles & Hendrickse (24), who found that almost all of their West African children with nephrosis also had demonstrable P. malariae infec- tion. In contrast, in only a small minority of the non- nephrotic children could quartan malaria be de- tected. In both groups, however, the incidence of P. falciparum was the same. Since that work the association of P. malariae and the nephrotic syn- drome has been established in many different geo- graphic areas (25). Hendrickse & Gilles (26) first advanced the suggestion that the nephrotic syndrome might be due to glomerular damage caused by the deposition of immune complexes. There is now much evidence to support this point of view. Ward & Conran (27) and Allison et al. (28), working indepen- dently, found immunoglobulin deposits on the glomerular basement membranes in renal biopsies from patients with nephrotic syndrome. IgM pre- dominated but other immunoglobulins and comple- ment were also present in some instances. In a series of extensive immunofluorescence studies Houba et al. (29) describe the immunological aspects of the 179 A. VOLLER syndrome in great detail. They showed that immuno- globulin deposited in a coarse granular fashion usually consisted of IgG, IgM, and complement whereas those deposits of a more diffuse character usually contained IgG alone. Immunofluorescence with specific antisera to P. malariae often detected antigens but P. falciparum antisera did not. Similarly P. malariae antigen could be eluted from the kidney material but P. falciparum antigens could not be detected (Houba & Lambert, personal communica- tion). Nephrotic syndrome has also been reported in a monkey experimentally infected with P. malariae (30) and the histological and immunofluorescence findings on the kidney were similar to those seen in human nephrotic syndrome. Again, like the human condition, there was no improvement follow- ing antimalarial therapy. Electron-microscopic studies on human quartan malaria nephrotic syndrome (28) have shown that the glomerular basement membrane is often thick- ened and that the foot processes of the epithelial cells are frequently fused. These findings all support the hypothesis that quartan malaria nephrotic syndrome is one of the diseases associated with the deposition of circulating immune complexes in the kidneys. It seems likely that the disease is initiated by specific malarial antigen-antibody complexes but is then perpetuated by an auto-immune reaction to damaged host tissue or to the initial antigen-antibody complex. We do not know why some individuals develop the chronic nephrotic syndrome during P. malariae infections whereas the majority have only a transient spontane- ously resolving nephritis. It may be that some indi- viduals are inherently more susceptible and react in an aberrant manner to quartan infection. Alterna- tively, extrinsic factors, such as the age at which the infection is encountered or the presence of other concomitant infections, might be involved. It is known (31) that low-affinity antibody is more likely to lead to immune complex formation, so it may be relevant that recently Steward & Voller (32) have shown that malaria infection can lower the affinity of antibody produced against unrelated anti- gens. Paradoxically induced malaria has been re- ported as being beneficial to patients with nephro- sis (33), and in experimental animal studies it has been shown to delay the onset of the spontaneous nephrosis that some strains of mice develop (34). These observations could mean that in these diseases antibody is usually produced that has the affinity for giving rise to immune complexes but that malaria infection changes that affinity so that the complexes develop less readily. THE SPLEEN An increase in the size of the spleen usually accompanies malaria infections. At first this increase is due to circulatory, congestive changes, but after a short time the immunologically important lymphoid and reticuloendothelial hyperplasia occurs (35). As Brown (36) has pointed out this is the normal re- sponse to intravenous stimulation with a particulate antigen. The spleen is likely to remain enlarged during chronic and repeated infections, but if the malaria infection is terminated there is a fairly rapid return to normal spleen size. The spleen plays a crucial role in the immune response to malaria. Splenectomy almost invariably leads to recrudescences of subpatent, latent infec- tions. Sometimes splenectomy can be used to render innately immune animals susceptible to malaria in- fections. It is not clear whether the loss of immunity following splenectomy is due to the simple decrease in the numbers of cells able to carry out erythro- phagocytosis or whether it is due to the loss of specifically activated or antibody producing cells. Other lymphoid tissues can take over these functions as immunity can be reestablished in splenectomized subjects. Malaria plays the dominant role in determining the spleen rate (the percentage of individuals with enlarged spleens) in many tropical areas and the spleen rate in particular age groups is used as an indicator of malarial endemicity. In areas where there is massive and sustained malaria exposure most of the children will have enlarged spleens but the spleen rates fall with increasing age. This correlates with an increasing degree of effective immunity. In contrast when the malaria challenge is less intense and is more sporadic the state of a stable immunity is not achieved and spleen rates remain high even in the older age groups. It should be recognized that where antimalarial measures have been instituted there can be a considerable effect on the normal immune response to malaria in the affected populations and the usual spleen rate classifications of endemicity may not be applicable. Of particular interest in the immunopathological sense is what a variety of authors have considered to be an identifiable entity and have termed the " tropi- cal splenomegaly syndrome" or " big spleen dis- 180 IMMUNOPATHOLOGY OF MALARIA ease ", and which they consider to be associated with malaria. An apparently identical disease has been reported from areas of East and West Africa, India, and New Guinea where malaria is endemic (37). It is claimed that this is a discrete syndrome, the characteristics of which include massive chronic splenomegaly asso- ciated with lymphocytic infiltration of the hepatic sinusoids and high serum levels of IgM and malarial antibodies. The affected individuals, although often living in areas of highly endemic malaria, do not have higher than usual parasite rates or densities, but long-term antimalarial therapy does lead to a reduction in their spleen size (38, 39). It has been suggested that the patients with tropical splenome- galy syndrome have an aberrant although effective immune response to malaria (40). As mentioned earlier the sera ofthese patients contain large amounts of IgM and in addition they also have high levels of antiglobulins and haemagglutinins. Wells (13) sug- gested that the extreme antiglobulin responses are induced by altered immunoglobulin G, possibly ma- larial antibody complexed to malarial antigen, and the haemagglutinins by erythrocytes rendered anti- genic owing to their infection with malaria parasites. Again, as in the case of nephrotic syndrome, we do not know if tropical splenomegaly syndrome is due to the inherent susceptibility of some individuals or to a combination of rather uncommon extrinsic factors. In either case it appears that malarial experi- ence is a prerequisite for the development of the disease. THE EFFECT OF MALARIA ON IMMUNE RESPONSES The first hint that malaria might affect the immune response to unrelated antigens was provided by McGregor & Barr (41), who showed that malaria- infected persons responded less well to tetanus toxoid injections. Greenwood et al. (42) pursued this and found that some humoral but not cell-mediated responses were depressed and the depression was greatest in those individuals with high parasitaemia. This was in accordance with experimental animal studies, which had already shown that the immuno- depression was maximum at the time of peak para- sitaemia (43). Later studies (44) indicated that the malaria influence was due to altered capacity of the macrophages to process antigen. Terry et al. (per- sonal communication) have put forward a theory that trypanosome infections may interfere with the cooperation of T and B lymphocytes and so lead to excessive production of partially nonspecific IgM and to immunodepression of other immune re- sponses that require the T and B cell cooperation. It is likely that malarial immunosuppression is basically due to the same mechanism as that postu- lated for trypanosomiasis. Malarial immunodepression has received much attention recently in the context of what it might mean in terms of other diseases. Greenwood (45) speculated that the apparent rarity of some auto- immune diseases in the tropics might be related to immunological disturbances produced by malaria. Support for this point of view has come from experimental work in which it was showed that some diseases, thought to be of auto-immune origin, which occur spontaneously in some strains of mice, could be delayed by infecting the animals with malaria (46). Burkitt (47) pointed out that malaria and Burkitt's lymphoma have a very similar distribution and he suggested that the immunological effect of malaria might allow an otherwise latent or benign viral infection to give rise to the lymphoproliferative disease. Epstein-Barr virus is a well qualified candi- date for this role since it is associated with infectious mononucleosis in temperate lands and is present in a large proportion of patients with Burkitt's lympho- ma (48). Wedderburn (49) has convincingly shown that malaria renders adult mice, which are otherwise resistant, susceptible to tumour induction by Moloney virus. It is probable that more immunopathological sequels of malaria combined with other agents will be identified in the future. Already Strickland et al. (50) have shown that joint infections of malaria and toxoplasmosis give rise to particularly severe disease in mice. It is well known too that accidental contamination of animal malaria parasite strains with viruses, Eperythrozoon, Haemobartonella, or Mycoplasma can easily occur and the course of a malaria infection and its pathology can then be altered. MALARIA AND PREGNANCY In areas where falciparum malaria is highly en- demic the adults usually suffer little frank disease. Many of these adults will be infected with malaria but their parasite densities usually remain at a low level. However, during pregnancy there is a dramatic 181 182 A. VOLLER increase in both parasite rates and densities (51, 52). The resultant anaemias are one of the major hazards to pregnant women in these areas. As there seems to be little direct correlation between the haemolytic episodes and the density of the falciparum infections, it is thought that the effect might be immunologically mediated. However there is no doubt that malaria is implicated, since antimalarial treatment leads to rapid resolution of the anaemia. Hamilton et al. (53) have shown that pregnant women with the tropical splenomegaly syndrome, which is also thought to have an immunological basis, are particularly likely to undergo severe haemolytic episodes. The mechanisms involved in the loss of immunity to falciparum malaria during pregnancy are not known. Studies on malarial antibody levels (52) indi- cate hardly any change although Cohen & Mc- Gregor (54) found that turnover rates of lgG were much lower in pregnant women than in other adult Gambians in a malarious zone. One of the most remarkable features is the intense accumulation of P. falciparum in the placenta. In these heavy placental infections all the asexual devel- opmental stages of P. falciparum, including tropho- zoites and schizonts, are seen. The very high para- sitaemias seen in placental blood are not reflected in blood smears made from the mother (52). It is well known that schizogony of P. falciparum normally occurs in the deep capillaries and not in the periph- eral blood; however, the degree of parasitization of the placenta is greatly in excess of that resulting from the normal retreat from the peripheral blood at schizogony. It seems that in spite of the intense phagocytic activity of the macrophages in the pla- centa, this is still an immunologically protected site in so far as P. falciparum is concerned. RESUME IMMUNOPATHOLOGIE DU PALUDISME Dans le paludisme, la principale reponse immunitaire est suscitee par les stades erythrocytaires du parasite et est dirigee contre eux. Outre la production d'anticorps specifiques, on constate une augmentation des immuno- globulines G et M et I'apparition de diverses anti- globulines et hemagglutinines. La tres forte anemie du paludisme peut etre due a une hemolyse par auto- immunisation ou 'a l'opsonisation d'erythrocytes non infectes. On peut observer au cours des infections palu- deennes deux types d'atteinte renale d'origine immunolo- gique. Le premier est represente par la nephrite transitoire qui se manifeste peu de temps apres le stade aigu dans beaucoup de cas. Elle s'accompagne souvent de dep6ts d'immunoglobulines M dans les glomerules. Le second type, le syndrome nephrotique, est une affection chro- nique specifiquement associee a l'infection par Plasmo- dium malariae. On decel parfois dans ce cas la presence d'immunoglobulines M et G, de complement et d'anti- gene plasmodique dans les glom6rules., Le syndrome nephrotique est probablement un exemple de maladie due a des immuncomplexes, provoquee au debut par la presence de complexes specifiques antigenes-anticorps puis acquerant le caractere d'une maladie auto-immune. La splenomegalie du paludisme est d'origine immuno- logique et s'accompagne d'une hyperplasie lymphoide. La rate joue un r6le essentiel dans la defense immunitaire contre les infections paludeennes, mais ses fonctions peuvent etre assumees par d'autres tissus reticulo- endotheliaux. Le syndrome de splenomegalie tropicale correspond probablement a une reponse anormale au paludisme due A des facteurs innes ou extrinseques. Caracterise par une splenomegalie massive chez l'adulte, une infiltration des sinus hepatiques par des lymphocytes et une forte teneur serique en IgM et en anticorps plas- modiques, il regresse sous l'influence d'un traitement antipaludique prolonge. Le paludisme a souvent pour effet d'affaiblir les reactions humorales A d'autres antigenes non apparentes, probablement en diminuant l'aptitude des macrophages a combattre l'antigene ou en alterant l'action conjointe des lymphocytes T et B. Les modifications de la reponse immunitaire dues au paludisme jouent peut-etre un r6le dans certains etats pathologiques comme le lymphome de Burkitt. On constate une diminution de l'immunite a l'egard de P. falciparum pendant la grossesse et le placenta semble etre un site privilegie, du point de vue immunologique, pour le developpement du parasite. REFERENCES 1. NUSSENZWEIG, R. S. ET AL. American journal of tropical medicine and hygiene, 21: 722 (1972). 2. BROWN, K. N. ET AL. Experimental parasitology, 28: 318 (1970). IMMUNOPATHOLOGY OF MALARIA 183 3. BROWN, I. N. ET AL. Immunology, 14: 127 (1968). 4. COHEN, S. & BUTCHER, G. A. American journal of tropical medicine and hygiene, 21: 713 (1972). 5. McGREGOR, I. A. & WILSoN, R. J. M. Transactions of the Royal Society of Tropical Medicine and Hy- giene, 65: 136 (1971). 6. TOBIE, J. E. ET AL. Journal of immunology, 97: 498 (1966). 7. ROWE, D. S. ET AL. Clinical and experimental immu- nology, 3: 63 (1968). 8. TARGETr, G. A. T. Clinical and experimental immuno- logy, 7: 501 (1970). 9. VOLLER, A. ET AL. Journal of tropical medicine and hygiene, 74: 45 (1971). 10. GREENWOOD, B. M. Clinical and experimental immu- nology, 6: 197 (1970). 11. KANO, K. ET AL. Proceedings of the Society of Experimental Biology and Medicine, 129: 849 (1968). 12. GREENWOOD, B. M. ET AL. Clinical and experimental immunology, 8: 161 (1971). 13. WELLS, J. V. Transactions of the Royal Society of Tropical Medicine and Hygiene, 64: 531 (1970). 14. GREENWOOD, B. M. ET AL. Clinical and experimental immunology, 7: 75 (1970). 15. ZUCKERMAN, A. Experimental parasitology, 15: 138 (1964). 16. ROSENBERG, C. B. ET AL. American Journal of Tropi- cal Medicine and Hygiene, 22: 146 (1973). 17. TOPLEY, E. ET AL. Transactions ofthe Royal Society of Tropical Medicine and Hygiene, 67: 51 (1973). 18. WOODRUFF, A. W. Transactions of the Royal Society of Tropical Medicine and Hygiene, 67: 313 (1973). 19. BERGER, M. ET AL. Annals of internal medicine, 62: 1163 (1967). 20. WARD, P. A. & CONRAN, P. B. Military medicine, 131, Supplement: 1225 (1966). 21. EHRICH, J. H. H. & VOLLER, A. Zeitschrift fur Tropenmedizin und Parasitologie, 23: 147 (1972). 22. BOONPUCKNAVIG, S. ET AL. Transactions of the Royal Society of Tropical Medicine and Hygiene, 67: 410 (1973). 23. VOLLER, A. ET AL. British journal of experimental pathology, in press. 24. GILLES, H. M. & HENDRICKSE, R. G. British medical journal, 2: 27 (1963). 25. KIBUKAMUSOKE, J. W. The nephrotic syndrome of quartan malaria. London, Amold, 1973, p. 1. 26. HENDRICKSE, R. G. & GILLES, H. M. East African medical journal, 40: 186 (1963). 27. WARD, P. A. & CONRAN, P. B. Military medicine, 134: 10 (1969). 28. ALLISON, A. C. ET AL. Lancet, 1: 1232 (1969). 29. HOUBA, V. ET AL. Clinical and experimental immuno- logy, 8: 761 (1971). 30. VOLLER, A. ET AL. British medical journal, 4: 208 (1971). 31. SOOTHILL, J. F. & STEWARD, M. W. Clinical and experimental immunology, 9: 193 (1971). 32. STEWARD, M. W. & VOLLER, A. British journal of experimental pathology, 54: 198 (1973). 33. GARDNER, G. & SHUTE, P. G. Lancet, 2: 946 (1965). 34. GREENWOOD, B. M. & VOLLER, A. Clinical and experimental immunology, 7: 793 (1970). 35. TALIAFERRO, W. H. & MULLIGAN, H. W. Indian medical research memoranda, 29: 1 (1937). 36. BROWN, I. N. Advances in immunology, 11: 289 (1969). 37. PrrNEY, W. R. Transactions of the Royal Society of Tropical Medicine and Hygiene, 62: 717 (1968). 38. STUIVER, P. C. ET AL. British medical journal, 1: 426 (1971). 39. SAGOE, A. British medicaljournal, 3: 378 (1970). 40. EDITORIAL. British medicaljournal, 4: 4 (1969). 41. McGREGOR, I. A. & BARR, M. Transactions of the Royal Society of Tropical Medicine & Hygiene, 56: 364 (1962). 42. GREENWOOD, B. M. ET AL. Lancet, 1: 169 (1972). 43. SALAMAN, M. H. ET AL. Journal of general micro- biology, 59: 383 (1969). 44. GREENWOOD, B. M. ET AL. Clinical and experimental immunology, 8: 467 (1971). 45. GREENWOOD, B. M. Lancet, 2: 380 (1968). 46. GREENWOOD, B. M. & VOLLER, A. Clinical and experimental immunology, 8: 805 (1970). 47. BURKITT, D. Journal ofthe National Cancer Institute, 42: 19 (1968). 48. KLEIN, G. The Epstein-Barr virus (EBV). In: Kaplan, A., ed. The herpesviruses. New York, Academic Press, in press. 49. WEDDERBURN, N. Lancet, 2: 1114 (1970). 50. STRICKLAND, G. T. ET AL. American journal of tropical medicine and hygiene, 126: 54 (1972). 51. GILLES, H. M. ET AL. Annals of tropical medicine and parasitology, 63: 245 (1969). 52. KORTMANN, H. F. Malaria and pregnancy. Utrecht, Elinkwijk, 1971, p. 31. 53. HAMILTON, P. J. S. ET AL. British medical journal, 2: 548 (1966). 54. COHEN, S. & MCGREGOR, I. A. Gamma-globulin and acquired immunity to malaria. In: Garnham, P. C. C. et al., ed. Symposium on immunity to protozoa. London, Blackwell, p. 123. 184 A. VOLLER DISCUSSION BRUCE-CHWATT: According to the hypothesis put forward by Greenwood,a the increase of IgM levels may be related to the decrease of B-lymphocytes. In connexion with the question of immunosuppression in malaria, vaccination against measles would have offered a good opportunity to assess whether the results in malaria free and malaria-infested areas would be comparable. It is well known that the adult spleen in malarious areas-although often not de- tectable by physical examination-is about double the size of the normal spleen. CORRADETr: The spleen tissue is more or less re- placed by retracting fibrose tissue in such cases. McGREioR: The numerous immunopathological syndromes described in association with malaria suggest that malaria may well exert considerable indirect pathogenic effects. As yet, however, there is little factual evidence of this and only the nephrosis associated with quartan malaria seems to be well established. Measles vaccination in the Gambia is entirely successful even though malaria is prevalent in children. Immunosuppression by malaria may therefore be selective against certain antigens. MICHEL: Adult spleen size may not depend only on malaria endemicity but also on racial differ- ences-for instance, in the savannah regions in Senegal, where malaria transmission is high, the rate of splenomegaly in adults is low. On the other hand, in the forest regions, where transmission is lower, the splenomegaly rate is higher. In areas with equal transmission rates, the Fulani population has high rates of splenomegaly, whereas among the Negroid peoples the rate is low. CORRADETTI: Responses to normal viable sporo- zoites in the human body and to X-irradiated sporo- zoites differ, and this difference may be related to the difference in the persistence of viable and irradiated sporozoites in the blood. VOLLER: The incidence of the nephrotic syndrome in areas where quartan malaria occurs depends largely on the extent to which it is sought. NuSSENZWEIG: Alterations of the immune system in malaria in man and in several experimental models are brought about by the pathological processes initiated by the disease. The renal involvement re- aGREENWOOD, B. M. Lancet, 1: 435 (1974). sponsible for the nephrotic syndrome in quartan malaria is one example. The presence of glomerular deposits of IgM, IgG, complement, and-in some instances-malaria- specific antigen in human beings, simians, and ro- dents, suggests that circulating immune complexes exist during malaria infections. In addition, soluble malarial antigens and antibodies against these anti- gens have been demonstrated in the plasma of Gambian patients as well as in patients from other areas of endemicity. Complement probably plays a role in the patholog- ical processes, since it is deposited on the glomerular basement membrane in malaria nephrosis. In addi- tion, the coating of erythrocytes with complement components may help to explain the severe anaemia associated with malaria. Hypocomplementaemia has been described in simian and rodent malaria. Another immunological alteration observed in ma- laria is a depressed antibody response to several unrelated antigens. It has been suggested that such immunosuppression results in the lower incidence of autoimmune diseases in certain areas where malaria is endemic. This finding is paralleled by observations in strains of mice (NZB/NZW) that spontaneously develop glomerulonephritis and autoimmune haemo- lytic anaemia. If these mice are inoculated with P. berghei yoelii early in life, the onset of their otherwise lethal autoimmune disease is delayed and its severity is reduced.b Although, as is known, malaria profoundly alters various lymphoid organs, the histopathological basis for this immunosup- pression has not been elucidated. Alterations in the cellular composition of lymphoid organs of P. berghei-infected mice have been de- scribed recently by Krettli & Nussenzweig.C Cell- surface markers, i.e., the 9-isoantigen for the thymus-dependent T-cells,d and the complement re- ceptor for the thymus-independent lymphocytes (B- cells), as described by Lay & Nussenzweig,e have been used to characterize the two different categories of lymphocytes. The main findings were: (a) a progressive, profound reduction of the weight, total b GREENWOOD, B. M. ET AL. Nature, 226: 266 (1970). C KRETTLI, A. U. & NUSSENZWEIG, R. Cell. Immun., in press (1974). d RAFF, M. C. Nature, 224: 378 (1969). e LAY, W. H. & NUSSENZWEIG, V. J. exp. Med., 128: 991(1968). IMMUNOPATHOLOGY OF MALARIA 185 cell number, and percentage of T-cells in the thymus, and (b) a severe decrease in the percentage of lymph- node complement receptor lymphocytes and T-cells with a simultaneous increase in the percentage of a " null cell " population that lacks both e-antigen and complement receptors. The mechanism by which the lymphocyte populations are depleted has been discussed by Krettli & Nussenzweig,c who think that the decreased number of CRL may be due to in vivo removal of lymphocytes coated with circulating im- mune complexes of parasitic or erythrocytic origin. The possibility that immune complexes of erythro- cytic origin participate in coating lymphocytes and in the nephritic deposits cannot be excluded. In fact, high titres of IgM against normal red blood cells have been described in malarious patients by Rosenberg et al.f Depletion of the lymphocyte population is probably only one among several other malaria-induced al- terations leading to immunosuppression. Other alter- ations, such as proliferation and hyperactivation of the reticuloendothelial system, may result in im- munosuppression owing to a defective macrophage processing of antigen in malarious animals. Levels of complement in mice with P. berghei infec- tion have been determined by a recently described technique g based on the finding that normal serum mediates release of immune complexes from in vitro pre-coated lymphocytes upon incubation at 37°C. The release is complement-dependent and is mediated by the activation of the alternate pathway. The main findings were as follows: rates of release were signifi- cantly increased during the initial 3 days after intra- venous inoculation of infected red blood cells, then decreased proportionately to the number of such cells inoculated. No release activity was detectable after the second week of infection. During the infections initiated by the injection of gradient- purified sporozoites," the decrease in the release activity of malarious serum was related to the para- sitaemia, being noticeable only after the second week of infection. Alterations in the complement system of malarious mice observed by the release technique have been confirmed by radial immunodiffusion (Mancini tech- nique) with an anti-C3 antiserum. In vitro incubation of normal mouse serum with infected erythrocytes resulted in inhibition of the f ROSENBERG, E. B. ET AL. Amer J. trop. Med. Hyg., 22: 146 (1973). g MILIER, G. W. ET AL. J. exp. Med., 138: 495 (1973). h KRETTLI, A. U. ET AL. J. Protozool., 20: 662 (1973). release activity as compared with normal serum incubated with noninfected red cells. The inhibition seems to be dose-dependent, but how much of the complement consumption is due to the presence of infected erythrocytes and/or to their lysis in vitro remains to be clarified. Inhibition of the release activity of normal serum has also been produced by malarious sera from mice at terminal stages of infection. Only 43% of malarious sera inhibited normal sera, and the degree of this inhibition was variable. The inhibition apparently operates via the alternate pathway of complement activation, since it occurred in C4-deficient guinea- pig serum and also in the absence of Ca++ions. One can only speculate on the mechanisms that lead to hypocomplementaemia in malaria. Increased cata- bolism of complement components is one of them. The possibility that a defect in synthesis may result from the alterations of the reticuloendothelial sys- tem, an important site of complement production, cannot be eliminated. Increased utilization of com- plement and its deposition in glomerular membranes and perhaps in otber organs may also contribute to this depletion. COHEN: It is difficult to reconcile these results with findings in West Africa, where in endemic areas immune adult Africans synthesize in response to malaria an amount of immunoglobulin about twice the total production in normal Europeans. Is the method used for the assay of B-cells reliable? By this technique, mechanical processes may produce arti- facts, only a proportion of B-cells may be detected, and macrophages may be included. BRAY: Turk has shown that, in lepromatous leprosy, there is a massive invasion of macrophages into the lymph nodes. I believe that this contributes to the loss of T-cells capable of reacting with Mycobacte- rium leprae. Have the " null cells " been identified definitely as small lymphocytes or could they be macrophages and/or fibroblasts progressively invad- ing the lymph nodes? NuSSENZWEIG: The test for CRL covers one sub- population and not all B-lymphocytes; it does not detect the actual antibody-producing plasma cells either. The recovery of cells in the test is easily reproducible, and the total cell numbers recovered from lymph nodes start to decrease onlyvery late in the infection. Recent studies in NZB mice have shown that the "null cells " appear-by all criteria-to be lymphocytes, although they have none of the surface markers that characterize T-cells or B-cells. 186 A. VOLLER MEUWISSEN: Can one discriminate between the de- at present, but studies are under way to achieve such crease of the number of cells and a redistribution of a differentiation. An acute fatal P. berghei infection cells in the reticuloendothelial system? does not reflect the response occurring in more chronic infections. This point is now being investi- NUSSENZWEIG: This question can hardly be answered gated, using P. berghei yoelii in mice.