Genetic factors in malaria*

Bull. Org. mond. Sante 1 1974, 50, 195-202 Bull. Wid Hlth Org. ' Genetic factors in malaria* L. LUZZATTO1 Some of the available information on the genetics ofPlasmodium is reviewed, andsome of its peculiarities are emphasized. Genetic factors in the human host that may affect susceptibility to malaria are critically evaluated. Most of the studies thus far have been concerned with the genetics of host erythrocytes but there is recent evidence that genes affecting immune processes may also be involved. At least two genes affecting red cells confer relative resistance to P. falciparum : the autosomal gene for haemoglobin S (Hb S) and the sex-linked genefor the glucose-6-phosphate dehydrogenase (G6PD) variant known as A-. Whereas malaria selection can be regarded as established for these genes, it still remains a hypothesis for some other polymorphic traits ofred cells. Differential susceptibil- ity to P. falciparum ofred cells with differentgenotypes has been tested by in vitro cultures, in which the invasion ofnew cells and intracellular development of the parasite can be followed by parasite counts and by 14C-isoleucine uptake. A model that relates genetic factors in Plasmodium and in man and that may account for certain features of host-parasite interactions is presented. In malaria, as in any other parasitic disease, the ultimate fate of the infection depends on the global interaction between host and parasite, which involves the genomes of both of them as well as the environ- ment. No matter how prominent the role of the environment, the primary importance of genetic factors is best demonstrated by the species specificity of plasmodial infections, which is often extremely marked. The capacity of a Plasmodium species X to infect a certain mammal and not another must be due to genetic differences between the two hosts; but the fact that a second Plasmodium species Y can infect the host resistant to X must be related to genetic differences between Plasmodium species X and Y. The complexity of host-parasite interactions is so great that it is usually not easy to analyse the nature of such genetic differences. However, specific cases can provide specific evidence. For example, the susceptibility of splenectomized but not of normal chimpanzees to P. falciparum clearly establishes the spleen as the major site of resistance in this pri- mate to this particular Plasmodium of man, a finding that has obvious implications for human patho- logy as well. While the genetic basis of parasite * Presented at the Symposium on Malaria Research, Rabat, Morocco, 1-5 April 1974. 1 Professor of Haematology, Department of Haemato- logy, University College Hospital, Ibadan, Nigeria. infectivity and of host resistance is one of general biological interest, this communication will be espe- cially concerned with the control of human malaria. GENETIC FACTORS IN PLASMODIUM Genetic analysis of malaria parasites has been hindered by the difficulty of developing in vitro cultures owing to the very complex natural life cycle of Plasmodium. Excellent reviews have covered previous work (1, 2), and it is intended here only to raise some of the salient questions that remain un- answered. Organization of the genome of the parasite While Plasmodium can be classified as a eukaryote a on the basis of a clear nucleo-cytoplasmic demarca- tion (4, 5), extensive ultrastructural studies still leave doubts as to whether its chromosomes are structural- ly similar to those found in metazoa. The electron- dense, DNAse-sensitive material found at the equa- torial plate during parasite schizogony (6) does not lend itself to a chromosome count; but from kine- tochore counts Sinden & Canning (7) have recently a This is corroborated by the demonstration that the ribosomes of P. berghei are of the 80 S variety, which is considered characteristic of eukaryotes, although their bio- synthesis appears to differ considerably from that of host cell ribosomes (3). 3179 - 195- 196 L. LUZZATTO estimated 4-8 chromosomes in P. berghei, a number somewhat higher than previously estimated from phase-contrast observations (8, 9). Cytogenetic data are thus certainly inadequate. However, one impor- tant point seems clear. There is no evidence of meiosis during gametocyte development, whereas some reductive process has been described as occur- ring during oocyst development and probably takes place immediately after zygote formation (10, 11). Thus, during all of its cycle in the vertebrate host, and during most of its cycle in the invertebrate host, the parasite is haploid. DNA content Compared with other microorganisms, there ap- pears to be a very large amount of DNA in the malaria parasite (12, 13), suggesting that, as in metazoa, a considerable fraction of it may not be transcribed into specific gene products. The DNA base composition has been reported for a number of Plasmodium spp., and in some species a remarkably low guanine-cytosine content has been observed (14). Purification of plasmodial DNA from mammalian erythrocytes is facilitated by the absence of host DNA, provided white cells have been fully eliminated (since each leucocyte contains approximately 100 times more DNA than a trophozoite). Genetic analysis Until very recently, only drug-resistance markers were available in Plasmodium, and among these fewer than a handful, notably the marker for pyri- methamine resistance, appeared stable enough to demonstrate convincingly that they had a genetic and not an adaptive basis. However, over the past three years, elegant and systematic work by the Edinburgh group has revealed a number of enzyme poly- morphisms in P. berghei (15) and has enabled them for the first time to isolate well-characterized doubly marked strains that can be crossed by feeding mosquitos on mice infected with them, thus produ- cing gametocytes from both strains. Genetic recom- bination has been observed at high frequency (16), indicating that the genes involved are not closely linked. Resistance of previously immunized animals to challenge by a different but related strain has proved to have a genetic basis as well (17). Thus, it should be possible in due course to carry out conventional genetic mapping in Plasmodium- " three point crosses " have already been reported (18)-and to establish the number of linkage groups n its genome, so as to complement the scanty cytogenetic data. A remarkable experiment suggest- ing rare recombination in the asexual erythrocytic stage was reported by Yoeli et al. (19); they referred to this phenomenon as " synfolia ". Until recombina- tion of external markers is clearly demonstrated,a it is not possible to exclude some alternative interpreta- tion of this experiment, notably a new mutation. However, the phenomenon is highly reminiscent of recombination events in another intracellular para- site, the bacteriophage. The rarity of recombination in this experiment could be due at least in part to the fact that the two markers involved happen to be closely linked. At any rate, it is clear that such recombinational events, taking place between tro- phozoites,b depend on simultaneous infection of the same red cell by two genetically different parasites and would be expected to be very rare, as compared with the likelihood of recombination when the zygote is formed. GENETIC FACTORS IN MAN On theoretical grounds, biological qualities affect- ing resistance to Plasmodium spp. might come into play at any stage of the parasite's life cycle in its hu- man host. Specifically, the following genetically deter- mined mechanisms might be involved. (a) Variation in the production of humoral antibodies against sporozoites or merozoites (antibodies against game- tocytes will affect the epidemiological characteristics of malaria but not the course of the infection in the individual); here the genetic mechanism can be visualized as operating at the level of the quantity or quality of antibody production. (b) Variation in the suitability of internal tissues for supporting the pre- erythrocytic development of the parasite; here the genetic mechanism could operate at the level, for instance, of specific functions in the liver. (c) Varia- tion in the suitability of the erythrocyte for penetra- tion by, and for supporting the development of, the asexual forms of the parasite; here the genetic mechanism might operate at the level of structure and function of the red cell itself. (d) Variation in the ability of the reticuloendothelial system (RES) to remove parasitized red cells and to interrupt the parasite cycle; here the genetic mechanism could be visualized as operating within the RES itself, and a Ferone et al. (20) have reported changes in dihydrofolate dehydrogenase in the " recombinant " parasites. b In the absence of any observation on the physical basis of this type of recombination it is not possible to rule out, from the experiment as reported (19), that recombination might have taken place, instead, between extracellular merozoites. GENETIC FACTORS IN MALARIA possibly at the level of cellular immune processes. Naturally, each of these stages can be a site of action of not only genetic but acquired mechanisms. In view of the multiplicity and complexity of these host-parasite interactions (and the list could be made more exhaustive), it is not surprising if definite evidence about genetic variation in susceptibility to malaria has been obtained in only a few cases. The concept of such variation was originally introduced in 1949 by J. B. S. Haldane with respect to thalassae- mia; later it was argued by Allison in 1954 with respect to Hb S, and by Allison & Motulsky in 1960 with regard to G6PD deficiency. Since that time, the possibility that a genetic variation in human red cells can affect susceptibility to malaria has been referred to as the " malaria hypothesis ", and has been raised also with respect to Hb C, Hb E, ABO blood groups, elliptocytosis, and the intraerythrocytic level of ade- nosine triphosphate (ATP). The now extensive litera- ture on this subject, mainly but not exclusively devoted to the red cell, has been ably reviewed by F. B. Livingstone (21). Rather than analyse it again here, it may be useful to consider some of the outstanding problems and prospects for the future. Relative versus absolute protection Over the years, professional malariologists have often been reluctant to accept the hypothesis that specific genetic host factors may affect susceptibility to malaria. The main reason for this reluctance has probably been that the effects involved are always relative rather than absolute. Thus far, no one has been able to identify a gene that will prevent a human being, devoid of acquired immunity, from developing malaria when bitten by an anopheline mosquito carrying a mouthful of sporozoites of P. falciparum. However, the severity of the attack and the extent to which it threatens life will depend on the genetic complement of the person attacked. Since this is true of infectious diseases in general, it is not surprising that it can apply to malaria. The main arguments that have been advanced against the malaria hypothesis are as follows. 1. Not all genes presumed to affect susceptibility to malaria are present at high frequency in all malaria-endemic areas. This fact is easily explained on the basis of population genetics. A mutant gene cannot increase in frequency in any area unless it is present there to begin with, whether it reaches the area by migration from elsewhere or arises by mutation on the spot. Furthermore, when- ever a favourable gene is already established in a particular population, the frequency of another fa- vourable gene may not be allowed to become as high, depending on the relative fitness values of the many genotypes that coexist. 2. The overall prevalence of malaria parasitaemia in subjects having a presumed protective genotype is often no lower than that of control groups. This evidently means that the genotype confers no protec- tion against malaria infection per se, and perhaps not even against malaria morbidity. It does not mean, however, that the genotype has no effect on mortality or fertility-and these are the only relevant para- meters from the point of view of natural selection. Thus, the finding of parasites in the blood of such subjects does not constitute evidence against the malaria hypothesis so long as biological fitness, as defined by Fisher (22), is increased. 3. Occasionally, malaria of life-threatening severity is observed in a person with a presumably pro- tective genotype. This is not surprising on statisti- cal grounds, since selection acts on the global gene pool of a particular population so as to increase the average biological fitness. This is not necessarily reflected in each individual case. For example, if a person with a " favourable " genotype at a particular locus has an " unfavourable " genotype at another locus or an acquired condition lowering his resistance (e.g., if he is splenectomized), he might develop severe malaria. By analogy, one would not consider immunization against poliomyelitis to be ineffective because of a single case of the disease arising in a vaccinated person. 4. A particular gene may be protective in a certain genotypic combination and not in another. Probably the clearest example of this is the Hb S gene, which protects when paired with the Hb A gene but not in the homozygous SS combination (see Adeloye et al. (23) and Table 1). This is not surprising either, since the pathophysiological characteristics of the two conditions are entirely different. What matters, again at the population level, is that since the frequency of the AS genotype in a population is always much higher than that of the SS-roughly by a square- power factor-the protective effect for AS hetero- zygotes more than offsets the disadvantage of SS homozygotes. 5. The mechanism whereby a particular gene in- creases resistance to malaria is not clear, and some of the mechanisms proposed are now considered to be unlikely or even untenable. This, of course, is not an 197 L. LUZZATTO Table 1. Severe malaria in sickle-cell anaemia a Nubr Number with % withNuambner obvious heavyexamined parasitaemia b parasitaemia b homozygous S 50 5 10 homozygous A 780 15 1.9 a Data from a survey of unselected children, sent from the General or from the Children Out-patient Department to the Haematology Department, University College Hospital, Ibadan, for routine blood examination. The specimens (thin blood films) were examined by Dr L V. Pozzi and Mr E. A. Usanga. b Obvious parasitaemia is defined as the presence of at least one parasite in every other 450 x power microscope field on a thin blood film. In slides with obvious parasitaemia, parasite rates were determined by counting at least 2 000 red blood cells, and were found to range from 0.2 % to 1.2 % in the S group and from 0.1 % to 4.5 % in the A group. argument at all. The precise mechanism of action of digitalis is still not elucidated and some earlier hypotheses on the subject have been disproved, yet no one would deny that it is an effective drug in the treatment of heart failure. Specific mechanisms of genetically determined resis- tance On the basis of the evidence currently available, which has been summarized elsewhere (21, 24), and with the above objections having been answered, it can now be asserted that the selection by malaria of specific genes affecting the red cell is no longer an hypothesis but an established fact, at least with respect to the Hb S gene and to some of the genes associated with G6PD deficiency, notably the A- gene (Table 2). It is important to determine the mechanism of such protective effects not only in order to improve our understanding of host-parasite interaction in these cases but also because such information may assist the development of new approaches to malaria control. Basically, an intrinsic change in the red cell could improve host resistance in one or more of three ways. (a) The red cell may be more difficult for the parasite to penetrate. (b) The red cell may be less suitable for the parasite to develop in after infection. (c) The red cell may be more susceptible to destruction in the body at some stage of parasite development, with Table 2. Evidence for increased resistance to malaria conferred on heterozygotes by specific genes a Thalassaemia deficiency Hb S worldwide geographic distribution b + +++ + + 'micro-mapping - c ++ + ++ age stratification ++ d increase in live offspring + e + e segregation ratiof + e decreased average spleen index + 9 decreased mortality from malaria + + lower parasite counts + ++ mechanism of resistance partially elucidated + + a This table is modified from Table 2 of Luzzatto (24) all pertinent references not cited below can be found in the original publication. The number of + signs represents a rough assessment of the strength of the corresponding line of evidence in favour of malaria selection for each of the three traits heading the table columns. b See also Luzzatto (25). c This term is used to indicate a correlation between malaria endemicity and frequency of polymorphic genes shown within a restricted geographical area (for references see Livingstone (21) and Luzzatto (24)). d See also Fleming et al. (26). e Stamatoyannopoulos et al. (27). f Ratio of heterozygotes for the abnormal gene to normal homozygotes among the offspring of heterozygote x homozygote matings. g Boyo (28). 198 GENETIC FACTORS IN MALARIA consequent destruction of the parasite itself. These three mechanisms we have referred to elsewhere (41) as failure of infection, abortive infection, and suicidal infection respectively. While it is not easy to distin- guish among these mechanisms in vivo, they can be separated to some extent in vitro. For instance, the finding of increased sickling of parasitized cells under anaerobic conditions (29) favours the model of suicidal infection in the case of AS red cells. It now seems possible to approach the problem in an even more direct way by preparing cultures of P. falcipa- rum. Raper (30) had already shown that the develop- ment of this parasite was not significantly different in previously infected AA erythrocytes as compared with AS erythrocytes. Recently, since new infective cycles by P. falciparum have been obtained through in vitro cultivation by several workers (31-34), we have attempted to test directly in collaboration with V. C. N. Okoye and Akintunde I. 0. Williams (35) whether red cells of different haemoglobin and G6PD types differ in their susceptibility to the parasite in culture. Thus far, erythrocytes of several genotypic combinations have been infected in vitro (Table 3). This rules out the existence of any absolute obstacle to parasite penetration. Preliminary quanti- tative data, however, suggest a lower than normal rate of successful infection in A- cells, although not in AS cells. It is hoped that, as data accumulate, it will be possible to draw more definitive conclusions on the first two mechanisms of protection mentioned above, namely failure of infection and abortive infection. Also, this approach may permit testing of the susceptibility of red cells with a variety of other genotypes, including thalassaemia. Extraerythrocyte factors The intrinsic qualities of red cells mentioned earlier are perhaps the easiest to investigate but are not the only genetic factors that can markedly affect host resistance to malaria. Impressive differences in sus- ceptibility of mice to P. berghei have been produced by inbreeding individual animals with low and high resistance (36). In man, it has long been maintained that individuals of African ancestry are highly resis- tant to P. vivax-although, again, this is not an absolute resistance (18)-and recently Hall & Can- field (37) have suggested that people of European descent are less resistant to P. falciparum. These Table 3. Infection in vitro of different types of erythrocyte a Experi- Donor of parasites b Recipient of parasites Parasites/1000 RBC ment Hb type G6PD type Hb type G6PD type at 24 h at 48 h 1 AA B AA B 24 AS B 20 AC B 28 AC A- 16 2 AA AB AA AB 10 6 AS A- 6 3 AA B 6 10 AA B 13 3 3 AC A AA B 55 15 AS A 30 20 AA B 20 25 AS B 25 20 a Experiments by L. Luzzatto, V. C. N. Okoye, and A. I. 0. Williams (unpublished data). The culture technique was similar to that of Siddiqui (33). Donor and recipient red cells were separated by differential agglutination with blood group isoantibodies having serological specificities appropriate to each individual case. b Blood infected with P. falciparum was obtained from children with acute malaria in experiments 2 and 3 (initial parasitaemia of 12 % and 20 % respectively) and from a freshly delivered placenta in experiment 1 (initial parasitaemia of 13 %). 199 200 L. LUZZATTO findings do not reveal anything about the specific basis of differential susceptibility, but it does not appear likely that such differences can be accounted for entirely by those red cell genes that have different frequencies in the populations concerned. Other possible sites of genetic resistance have been men- tioned above. In view of the importance of immuno- logical phenomena in acquired resistance to malaria, it can be safely assumed that genetic differences in the capacity for mounting an immune response, whether humoral or cellular, will be found to be of major importance. In this connexion, it is highly relevant that Ceppellini and his group (38) found in Sardinia a higher value of Wahlund's variance for the second locus (Four) of the HL-A system, as compared with other neutral loci. While particular alleles at the HL-A locus may be related in their own right to qualitative differences in immunological resistance to malaria, the authors consider it more likely that selection operates on the possibly closely linked " immune response " genes, which exert a quantitative effect on antibody production. That differences in levels of bulk serum immunoglobulins may have a partly genetic basis was suggested some time ago (39, 40). RELATIONSHIPS BETWEEN GENETIC FACTORS IN PARASITE AND HOST Intracellular parasitism brings about an extreme degree of contact between two completely different genomes. It is therefore not surprising if unusual features sometimes arise. A case in point is the infection by malaria parasites of G6PD-normal and G6PD-deficient red cells. It has been found that, in heterozygous females whose blood contains a mix- ture of these two cell types, the G6PD-normal cells are infected preferentially (41); these persons enjoy relative protection against P. falciparum (42). How- ever, no protection is apparent in hemizygous males having only G6PD-deficient cells. These findings have led to the suggestion that adaptation of the parasite to a particular cell type may place it at a disadvantage when it is confronted in turn with two different types of erythrocyte (24). While adapta- tion might be merely a phenotypic phenomenon, it may well be associated with loss of individual genes. Until some of the many gaps in our knowledge of genome organization and replication in Plasmodium are filled (see above), it cannot be ruled out that this organism can tolerate relatively common accidental losses ofDNA in the course of schizogony, especially since this method of multiplication is otherwise very efficient and very rapid. Perhaps the explosive char- acter of the growth of Plasmodium was overlooked when the suggestion was made that a " classic haploid organism cannot overcome the vicissitudes " of its complex life cycle (1). Indeed, as mentioned earlier, Plasmodium is haploid in the vertebrate host, and the relatively high incidence of drug-resistant organisms (2) appears consistent with this view: if the parasite were diploid, excessively high mutation frequencies would be required in order to obtain the same number of resistant phenotypes (unless most drug-resistance mutations were dominant). Thus, we can visualize four operationally different types of merozoite arising during the asexual cycle of P. falci- parum: (a) merozoites that have lost some indis- pensable gene function a and die before or soon after infecting a new erythrocyte; (b) merozoites that have lost functionally dispensable gene functions, and can still invade some types of (" normal ") red cell but not others (e.g., G6PD-deficient red cells); (c) mero- zoites that can invade any type of red cell but have lost some gene functions required for gametogenesis; and (d) merozoites with intact genome function that can become gametocytes. This situation may not be comparable for all Plasmodium spp. At any rate, the many known effects of the method of parasite transmission, of host physiology, and of host immu- nity on gametogenesis (43), and especially the com- plete loss of capacity to produce gametocytes after repeated blood passages observed in certain cases, are consistent with the above model. Some striking phenomena of adaptation, for example of P. lophurae to mice (44), or of P. berghei to the gerbil (45), might have a similar basis. The value of any model lies in its capacity to make predictions that can be verified. Verification is not easy in this case, given the nature of the system, but at least one experimental test can be envisaged. It has been suggested (24) that parasites developing in G6PD-deficient erythrocytes may be forced to ex- press their own G6PD gene(s), whereas parasites developing in G6PD-normal erythrocytes may not need to do so. Indeed, Theakston & Fletcher found no G6PD activity, as assessed by an electron cyto- chemical method, in P. falciparum developing in G6PD-normal red cells of Aotus trivirgatus (46) or of a " Loss of gene function " might be associated with physical loss of a gene (deletion), in which case it would be irreversible; or it may be associated with a point mutation, or with inability to express the gene because of changes in regulatory processes, in which case the loss might be reversible. GENETIC FACTORS IN MALARIA 201 man (personal communication, 1973). Similar deter- minations may disclose G6PD activity in parasites developing in or emerging from G6PD-deficient red cells. In addition, it should be possible to test in culture in vitro whether the growth of the parasite is affected by providing, as host cells, mixtures of G6PD-normal and G6PD-deficient erythrocytes in varying ratios. ACKNOWLEDGEMENTS Thanks are due to Dr Laura V. Pozzi, Mr E. A. Usanga, Dr V. C. N. Okoye, and Dr A. I. 0. Williams for carrying out work reflected in Tables 1 and 3, and to the staff of the General Out-patient Department, University College Hospital, and of the Adeoyo Hospital, Ibadan, for providing samples of blood. The author is also grateful to Dr G. M. Edington, Dr H. M. Gilles, and Dr A. 0. Lucas for stimulating much of the early research carried out on malaria in the Haematology Department, University College Hospital, and to all colleagues in the Department for their help and advice. Financial support for portions of this work was received from the World Health Organization, through support to the Regional Reference Centre for Glucose-6-Phosphate Dehydro- genase in Ibadan, and from United States Public Health Service Grant GM 17261. RISUMt FACTEURS GE-NETIQUES DANS LE PALUDISME L'auteur examine un certain nombre de donnees actuel- lement acquises concernant la genetique de Plasmodium et met l'accent sur quelques aspects particuliers. Les facteurs genetiques propres a F'h6te humain qui peuvent modifier la receptivit6 au paludisme font l'objet d'une evaluation critique. Jusqu'A present, la plupart des recherches ont porte sur la genetique des erythrocytes de l'h6te, mais il est avere depuis peu que les genes intervenant dans le processus immunitaire jouent aussi un r6le. Deux genes, au moins, agissant sur les erythrocytes conferent une resistance relative vis-a-vis de P. falciparum: le gene autosomique de l'hemoglobine S (HbS) et le gene lie au sexe responsable de la carence en glucose-6-phosphate deshydrogenase (G6PD), due au mutant A-. Si le rapport entre ces genes et la receptivite au paludisme est bien etabli, il n'en va pas de meme de certains autres ( traits)> des erythrocytes dont I'action est encore hypothetique. Les diff6rences de receptivite d'erythrocytes de divers genotypes a 1'egard de P. falciparum ont ete etudiees in vitro grace a des cultures ou l'on peut suivre 1'envahisse- ment des cellules et le developpement du parasite par des numerations parasitaires et par la mesure de la captation de 14C-isoleucine. L'auteur expose sa conception d'un modele etablissant un rapport entre les facteurs genetiques chez Plasmodium et chez 1'homme et capable de rendre compte de certains aspects des interactions h6te-parasite. REFERENCES 1. WALKER, P. J. Investigational problems and the mechanism of inheritance in blood protozoa. In: D. Weinman & G. M. Ristic, ed. Infectious blood diseases of man and animals. New York, Academic Press, 1968, vol. 1, pp. 367-391. 2. PETERs, W. Chemotherapy and drug resistance in malaria. London, Academic Press, 1970. 3. TOKUYASU, K. ET AL. Military medicine, 134: 1032- 1038 (1969). 4. RUDZiNsKA, M. A. & VICKENNAM, K. The fine structure. In: D. Weinman & M. Ristic, ed. Infec- tious blood diseases of man and animals. New York, Academic Press, 1968, vol. 1, pp. 217-306. 5. AIKAWA, M. Experimental parasitology, 30: 284-320 (1971). 6. AIKAWA, M. ET AL. Proceedings ofthe Helminthologi- cal Society of Washington, 39: 174-194 (1972). 7. SINDEN, R. E. & CANNING, E. U. Journal ofprotozoo- logy, 20: 528 (1973). 8. WOLCOTT, G. B. Journal of heredity, 46: 53-57 (1955). 9. WOLCOTT, G. B. Journal of protozoology, 4: 48-51 (1957). 10. CANNING, E. U. & SINDEN, R. E. Parasitology, 67: 29-40 (1973). 11. BANO, L. Parasitology, 49: 559-585 (1959). 12. GUTTERIDGE, W. E. & TRUiG, P. I. Journal of protozoology, 17: 89-97 (1970). 13. BAHR, G. F. & MIKEL, U. Proceedings of the Helminthological Society of Washington, 39: 361-372 (1972). 14. GUITrERIDGE, W. E. ET AL. Parasitology, 62: 209-220 (1971). 15. CARTER, R. Parasitology, 66: 297-307 (1973). 202 L. LUZZATIO 16. WALLIKER, D. ET AL. Parasitology, 66: 309-320 (1973). 17. OxBROw, A. I. Parasitology, 67: 17-27 (1973). 18. WALLIKER, D. & CARTER, R. Journal ofprotozoology, 20: 529 (1973). 19. YOELI, M. ET AL. Parasitology, 59: 429-447 (1969). 20. FERONE, R. ET AL. Science, 167: 1263-1264 (1970). 21. LIVINGSTONE, F. B. Annual review of genetics, 5: 33- 64 (1971). 22. FISHER, R. The genetical theory of natural selection, 2nd rev. ed. New York, Dover, 1958. 23. ADELOYE, A. ET AL. British medical journal, 2: 445- 446 (1971). 24. LuzzATro, L. Proceedings of the Helminthological Society of Washington, 39: 100-106 (1972). 25. LUZZATrO, L. Israel journal of medical sciences, 9: 1181-1194 (1973). 26. FLEMING, A. F. ET AL. Haemoglobin electrophoretic patterns and malaria parasite densities in a rural population in Kano State. In: Second Meeting of the Nigerian Society of Haematology and Blood Trans- fusion, Zaria, 6-8 December 1973, Abstracts, No. 30. 27. STAMATOYANNOPOULOS, G. ET AL. Hb S, glucose 6- phosphate dehydrogenase deficiency and thalas- saemia in Greece before and after malaria eradication. In: IX International Congress on Tropical Medicine and Malaria, Athens, 14-21 October 1973, Abstracts of invited papers, p. 266. 28. Boyo, A. E. American journal of tropical medicine and Hygiene, 21: 863-869 (1972). 29. LUZZATro, L. ET AL. Lancet, 1: 319-321 (1970). 30. RAPER, A. B. Transactions of the Royal Society of Tropical Medicine and Hygiene, 53: 110-117 (1959). 31. TRIGG, P. I. Nature, 213: 1019-1020 (1967). 32. DiGGs, C. ET AL. Journal ofparasitology, 57: 187-188 (1971). 33. SIDDIQUI, W. A. In: IX International Congress on Tropical Medicine and Malaria, Athens, 14-21 October 1973, Abstracts of communications, p. 184. 34. PHLLPS, R. S. ET AL. Transactions of the Royal Society of Tropical Medicine and Hygiene, 66: 523- 524 (1972). 35. LUZZATTO, L. ET AL. In: IX International Congress on Tropical Medicine and Malaria, Athens, 14-21 October 1973, Abstracts of invited papers, p. 265. 36. ZUCKERMAN, A. Basis of host cell-parasite specificity. In: D. Weinman & M. Ristic, ed. Infectious blood diseases of man and animals. New York, Academic Press, 1968, vol. 1, pp. 23-36. 37. HALL, A. P. & CANFIELD, C. J. Proceedings of the Helminthological Society of Washington, 39: 66-70 (1972). 38. CEPPELLINI, R. Specific immune response genes and defence against malaria. In: IX International Con- gress on Tropical Medicine and Malaria, Athens, 14- 21 October 1973, Abstracts of invited papers, p. 267. 39. RAWNSLEY, H. M. ET AL. Science, 123: 991-992 (1956). 40. EDOZEN, J. C. Journal of clinical pathology, 10: 276- 279 (1957). 41. LUZZATTO, L. ET AL. Science, 164: 839-842 (1969). 42. BIENZLE, U. ET AL. Lancet, 1: 107-110 (1972). 43. BisHoP, A. Parasitology, 45: 163-185 (1955). 44. MCGHEE, R. B. Journal of protozoology, 3: 122-126 (1956). 45. WEiss, M. L. In: IX International Congress of Tropical Medicine and Malaria, Athens, 14-21 October 1973, Abstracts of communications, p. 184. 46. THEAKSTON, R. D. G. & FLETCHER, K. A. Life sciences, 10: 701-711 (1971). DISCUSSION BRUCE-CHWA¶T: One single count of the density of parasitaemia cannot be used as a reliable parameter of numerical relationship. It is not clear whether racial differences in the response to infection are due to an acquired factor or to a permanent genetic factor of relatively recent development. LuZZATrO: Parasite counts are used only for groups of individuals with high or low counts. As regards racial differences in response, 1 cannot believe that, in persons not previously infected, there could be an acquired factor. It is more likely that the genes have been selected in the original populations from which the American Negroes descended. MOLINFAUX: Epidemiological data from a WHO field research project in Kano State, Nigeria, with respect to abnormal haemoglobins (Fleming, per- sonal communication) confirm that the AS genotype is partly protected against falciparum parasitaemias. IgM data show, with increasing age, an increase in the average value and in individual variation but a decrease in variation in repeated examinations of the same individual. This indicates that, with the local high level of antigenic stimulation, these people are approaching the individual maximum value of their 1gM level, which is probably genetically de- termined (Cornille-Br0gger, personal communica- tion). MICHEL: It has been observed in a longitudinal study in Senegal that low parasite densities always occur in children with variant haemoglobins, whereas with normal individuals there are fluctuating densities some of which are very high.