Differential antibody response of Gambian donors to soluble Plasmodium falciparum antigens

26 TRANSACTKBNS OF THE ROYAL SOCIETYOF TROPICALMEDICINE AND HYGIENE (1991) 85, 26-32

Differential falciparum

antibody antigens

response

of Gambian

donors to soluble

Plasmodium

P. H. Jalcobsen’, E. M. Rilefl, S. J. Allen’, S. 0. Larsen3, S. Bennett4, S. Jepsen5 and B. M. Greenwood* ‘Malaria Research Laboratory, Department of Treponematoses, State Serum Institute, Copenhagen, Denmark; ‘Medical Research Council Laboratories, Fajara, The Gambia; 3Department of Biostatistics, State Serum Institute, Copenhagen, Denmark; 4Tropical Health Epidemiology Unit, London School of Hygiene and Tropical Medicine, London, UK; ‘World Health Organization National Malaria Diagnostic Centre, State Serum Institute, Copenhagen, Denmark

Abstract A seroepidemiological and clinical study was performed in an area of West Africa (The Gambia) where Plasmodium falciparum is endemic with seasonal transmission. Plasma samnles were tested bv intermediate gel immunoelec~ophoresis for antibodies aaainst 7 soluble P. falciparum antiaens. There were marked differences -in ihe age-related pattern of antibody response to the different antigens. Antibodies to 4 of the antigens were acouired slowlv with a maximum prevalent; reached afier 25-35 years of age. Antibodies against the 3 remaining antigens, including the endotoxin-like antigen, Ag7, were acquired earlier with a plateau of maximum prevalence reached after 5-11 years, i.e. at the time when morbidity due to malaria decreased. Children who had not appeared to be infected with malaria during the preceding transmission seasonhad lower levels of antibodies to soluble antigens than did children who had had a documented attack of clinical malaria or parasitaemia. There was no difference in antibody profiles to soluble antigens between children with sickle cell trait and children with normal haemoglobin. Introduction Immunity to malaria is acquired in stages. Initially protection is acquired against severe disease and death. Subsequently there is gradual acquisition of resistance to febrile episodes and parasitaemia. Immunity is rarely complete and periodic episodesof low grade parasitaemia occur throughout adult life. It is probable that different malaria antigens, and different immune mechanisms, are involved at each of these stages.For example, protection from death and severe diseasemay be due to the development of antibodies against parasite-derived endotoxin-like molecules which trigger the release of acute phase reactants, lymphokines and endogenous pyrogens which have been implicated in the nathoaenesis of severemalaria (CLARK; 1987). This ‘has &en termed ‘anti-toxic’ immunitv (PLAYFAIR et al., 1990). Immunitv to the

parasite itself develops later and’ may invoive antibodies which block merozoite invasion into erythrocytes or cell-mediated mechanisms of parasite killing. Soluble malaria antigens, which are released from infected cells at the time of erythrocyte rupture and Author for correspondence: Palle Hey Jakobsen, Malaria ResearchLaboratory, Department of Treponematoses, State Serum Institute, Amager Boulevard 80, DK-2300 Copenhagen S, Denmark.

merozoite release.,and also during merozoite reinvasion, may be derived from the processing of parasite surface antigens or may be functional moleculessuch as enzyme+-secreted by the parasite into the ervthrocvte cvtonlasm. One aroun of soluble antigens, characterized by their &g&c stability after boiling, are the S-antigens (WILSON et al., 1969). These antigens are highly immunogenic but extremely polymorphic and-may play a role in immune evasion bv the narasite (reviewed bv HOWARD. 1986). The soluble antigens used in our studies were ‘a mixture of heat-stable and heat-labile antigens. This mixture of soluble antigens snecificallv induces nroliferation of, and production of interferon-gamma (IFNy) by, lymphocytes from individuals immune to malaria (RILEY et al., 1988; THEANDER et al., 1986). Monoclonal antibodies to one of the antigens (Agl) inhibit the growth of Plasmodium falciparum in vitro (JACKOBSEN et al., 1987a). Another

antigen in the

complex (Ag7) has endotoxin-like properties (JAKOBSEN et al., 1988, 1990). Thus it is possible that immunity to such antigens plays a role both in anti-toxic and anti-parasitic immunity. We have previously examined the humoral response to a number of soluble P. falciparum antigens using the technique of crossed immunoelectrophoresis (CIE) (JAKOBSEN et al., 1987b; JEPSEN & AXELSON, 1980). Using this technique, each serum or plasma sample can be tested for reactivity with several distinct antigens on a single gel. The antibody titre can be judged in a semi-quantitative manner from the position of the precipitate. Each antigen possessesa distinct CIE nrofile which allows unambiaous identification, in contrast to immunoblotting, where wide variation in the molecular mass of soluble antigens from different parasite isolates has been reported (HOWARD

et al., 1986).

This naner describes a sero-enidemioloaical and clinical study performed in an area of W&t Africa (The Gambia) where falcinarum malaria is endemic with seasonal transmission. Materials and Methods Study area and subjects

The study was performed in the Kataba villages near the town of Farafenni on the north bank of the Gambia river. Malaria transmission in this area is seasonal,with most transmission occurring from July to November; 90% of infections are caused by P. falciparum and 10% by P. malariae. The epidemiology of malaria in the area has been described in detail by GREENWOOD et al. (1987) and MARSH et al. (1989).

27 The population of the Kataba villages is approximateIv 50% Mandinka. 40% Wolof and 10% Fula. In Mav f988 (after several months of low malaria transmission) 120 children resident in these villages were recruited into the study, and 82 were re-examined in November 1988 at the end of the wet season (peak malaria transmission). At this time 163 additional randomly selected subjects, aged 9-70 years, and 15 subjects aged l-2 years, were recruited. A clinical examination, including palpation of liver and spleen, was performed and a venous blood sample was collected. Thick and thin blood 6hns were stained with Giemsa’s stain and examined for malaria parasites and the haematocrit was measured. Haemoglobin genotype was determined by electrophoresis and plasma was saved for serology. Morbidity

surveillance

At the beginning of the rainy season all parasitaemic children received a curative dose of chloroquine together with Maloprim@ (pyrimethamine plus dapsone). This combination of drugs was used to overcome low levels of chloroouine and nvrimethamine resistance in this area (MENON et -il., 1987; PANTONet al.. 1985). All children were then visited once a week by a field worker. A health questionnaire was completed and the child’s axillary temperature was measured. A linger prick blood sample was taken from all children with a temperature of 37.X or more. An episode of clinical malaria was recorded in children with a raised temperature and a positive blood film with parasitaemia ~5OOO/nl.Children with fever and low parasitaemia (<5OOO/ul) were categorized as intermediate malaria infections. All children with parasitaemia were treated with chloroquine. All children were re-examined in November 1988. Anv child with a positive blood f&n, or who had acquired splenomegaly during the rainy seasonbut in whom an episode of fever had not been recorded, was classified as having had an asymptomatic malaria infection. Acquired splenomegaly was defined as the presenceof a palpable spleen in November which had not been palpable in May, or an increase in spleen size of at least 2 cm during this period. Children in whom no clinical episode of malaria had been recorded, and in whom neither parasitaemia nor acquired splenomegaly was detected, were regarded as having no evidence of malaria infection.

column (260 ml gel) (Pharmacia Fine Chemicals, Uppsala, Sweden) containing as ligand a pool of immunoglobulin G (IgG) (2.5 mg/ml gel) with high titres of precipitating antibodies against soluble P. falciparum antigens. Bound antigens were eluted with 3 M KSCN in 0.02 M Tris-Verona1buffer at pH 9.0. The pooled fractions containing eluted antigens were dialysed, concentrated x250 and tested for their content of soluble antigens by CIE. Intermediate gel immunoelectrophoresis

Electrophoresis was performed on glass plates (3.5~5 cm) in 1% agarose gel (Litex@! Glostrup, Denmark? type HSA) in Tris-barbital buffer (pH 8.6, IO& strength O-02). 10 ul of the affinitypurified soluble antigens (described above) were run in the lirst dimension gel at 10-15 V/cm until a parallel blue albumin marker had migrated 1.5 cm. The second dimension electrophoresis was run perpendicularly to the first dimension gel at 2 V/cm for 18 h into a gel containina 21 t&cm2 reference immune serum Plasma from individual Gambian donors were tested by intermediate gel immunoelectrophoresis (AXELSEN, 1983)incorporating 48 $/cm2 plasma into the intermediate gel. The plates were washed and pressed3 times and stained with Coomassie brilliant blue R250. After electrophoresis each gel was scored for reactivity with each soluble antigen and given a symbol depending on the reactivity: - , no reactivity; +, weak reactivity (extensions of the precipitate were deflected into the intermediate gel); and + +, strong reactivity, (the whole precipitate was positioned in the intermedate gel).

Immwwfluorescence antibody tests (IFAT)

Serum from each donor was tested for antibodies by indirect imagainst whole P. falciparum munofluorescence. using as antigen acetone-fixed. air-dried preparations 07 schizon&nfected cultured erythrocytes. The method has been described in detail elsewhere (~URSH et al., 1989).

continuous culture in a modified Trager and Jensen system (TRAGER& JENSEN,1976) as described by JEPSEN& ANDERSEN(1981).

Results At the beginning of the study in May 1988, 116 of the 120 children had antibody titres to P. falciparum, measured by immunofluorescence, greater than 1:1280. In November 1988, 248 of the 255 subjects had immunofluorescence titres greater than 1:1280. Two of the children without detectable infection during the rainy season showed a 3-fold increase in IFAT from Mav to November, indicating that at least some of these children had in fact -experienced infection during the rainy season.

Reference immune serum

Prevalence of antibodies in different age groups

Pflrasite culture ‘I’he P. falciparum isolate F32iTanxania was kept in

Pooled serum from clinically malaria-immune (Liberian) adult blood donors was used as a reference standard. This standard had a high titre of precipitating antibodies directed against soluble P. falciparum antigens. Isolation of soluble P. falciparum antigens

Supernatants from the daily change of exhausted medium from parasite cultures were pooled, dialysed against column buffer (Tris-barbital pH 8.6, O-5 M NaCl, 15 mu NaN3) and added to a CNBr-sepharose

The prevalence of nrecinitatina antibodies to the various soluble antigens at t‘heend-of the transmission season,as measured in the November cross-sectional survey; was markedly age-dependent and differed between antiaens (Fie. 1). The antieens could be separatedbroadly &&two groups. 1n”me first group (Agl, Ag2, Ag3 and Ag4) antibody prevalence increased onlv slowly with age and maximal nrevalence was reached inearly adult life (25 to 35 years of age). Very few children under the age of 5 years were seropositive for these antigens and even among adults

28

Antigen

1

Antigen

I pre”ale”ce

2

70 6.6 50 40

40

30

30

20

20

10

10 0

0 10

20

30

Antigen

% prevalence

40

10

60 50 AS. Of donors

3

20

% PreY*le"Ce

30

40

Antigen

4

30

40

50

60

Age

oi donors

50

60

100 so

100 SO

60 70

60

60

60

70

50

50

40

40

30

30

20

20

10

10 0

0 10

.20

30

Antigen

40

60 50 AS. 01 donors

5

10

20

Aat) of donors % rJrwamc.

Antigen

6

30

40

100 t so 60. 70 60 SO 40 30 20 10 0 10

20

Agm of dormrs

Antigen

50 80 Age 01 dOnOr,

?

::Ir4dlinrlr70

: weak

antibody

reaction

00

: strong

50

antibody

reaction

40 30 20 10 0 10

20

30

40

50

60

Fig. 1. Prevalence of parasites (line) and precipitating antibodies (CO~UIIUB) to soluble P. falcipamm antigens in different age groups, measured in the November 1988 cross-sectional survey, for the following age groups: 1-2, 3-& 5-67-8, P-11, 12-14, 15-24,25-34, 35-U, 45-54 and >55 years.

29

Antigen

1

Antigen

?i Pr*YaI.“C* * 100.

2

90. a0 70. 60 30 40 30 20 10 0

x DrI”.l*“C*

Antigen

3

A 100,

May NO”

May NO”

Ma” NO”

May NO”

NM

AM

IM

CM

Antigen

?4Ore”al.“Ca

4

100

90

90

a0 70. 60 30. 40

80 70 60 30 i 10 30 20 10 n MW NW

Anttgen

5

May NW

t.4.y NO”

AM Antigen

IM 6

CM

AM

IM

CM

t4.y No”

NM % Dr.Yamme 100 90

a0

NM

0

: weak : strong

antibody antibody

reaction reaction

Fig. 2. Seasonal variation io prevalence of preciptiating antibodies to soluble P. fa&ipmum antigens in four groups of children 3-S years old, 14 uninfected (NM), 35 with asymptomatic infection (AM), 21 with fever plus low pmsitaemia (IM) and 50 with fever plus high parasitemia (CM) collected in May 1988 (dry season) and 13 NM, 23 AM, 14 IM and 32 CM collected io November 1988 (west season).

30

the prevalence of high-titre precipitating antibodies to these antigens was very often less than 60%. In the second group (Ag5 and Ag7) antibody prevalence increased much more rapidly and a plateau prevalence was reached in childhood. More than 60% of 5- to 8-year-olds had precipitating antibodies to both Ag5 and Ag7, and these levels increased throughout adult life. Ag6 fell somewhere between the 2 groups. About l&30% of very young children (<4 years) were serooositive and a olateau of 90% orevalence was reached by the age-of 9 to 11 years. Considerina onlv the 3-8 vear old children there was a markedage-dependent antibody response to all the soluble antigens (except Ag3) in the dry season (data not shown). However in the wet season an age-dependent response of these children was seen only for Ag7 as evaluated by the x2 test for trend (x2=12*97, 1 degree of freedom (dfi, P=O*O5). .._ Antibody responses to the different- antigens ‘were hiahlv correlated within individual samolesin the drv se&oh, as evaluated by x2 tests in- 2X2 tables. Responsesto Agl correlated with responses to Ag3 (P
The 120 children for whom complete morbidity data were available were divided into 4 groups according to their clinical experience of malaria during the 1988 rainy season, i.e. July-November. Patterns of malaria morbiditv are described in detail elsewhere (RILEY et al., 1990): There were 14 children in whom neither symptomatic nor asymptomatic infections were detected, 35 children had asymptomatic malaria (parasitaenua or acquired splenomegaly without fever). 21 children had fever and low parasitaemia, and 50 children had fever and high parasitaemia (clinical malaria). There was no correlation between the antibody response to most of the soluble antigens as measured in May and the subsequent pattern of malaria exnosure and morbidity (Fig. 2). However the prevalence of high titre antibodies to Ag4 was significantly higher in the group without infection than in the groups with infection even after age stratification (Mantel-Haenszel, x2=8-25, 1 df, P
Antibody profiles were measured at the end of the dry seasonand again at the end of the rainy seasonin 82 children aged 3-8 years (Fig. 2). Overall, antibody responses were both more prevalent and of greater intensity in May than in November. Comparison of antibody titres in individual children showed a fall in the titre of antibodies to all antigens from the dry to the wet season in children who did not appear to become infected. This difference was significant for Ag7 (McNemar test, P=O*O16). Among children with symptomatic and asymptomatic infections, levels of antibodies to Ag5 and Ag7 increased during the rainy seasonbut levels of antibodies to antigens 1,2,3 and 4 declined.

Effect of carriage of sickle cell trait on antibody response to soluble antigens

Thirty-two of the 120 children aged 3-8 years were carriers of the sickle cell trait (haemoglobin genotype AS). Eoisodes of fever (intermediate or clinical malaria) were significantly less common in AS children than in AA children (12/32 vs 58/88, x2=7*79, df= 1, P=O.OOS).Although for most antigens (except Ag3) AA individuals tended to have stronger and more prevalent antibody responses than did AS individuals, these differences were not statistically significant -when evaluated by the x2 test (data not shown).

Discussion

In the rural area of The Gambia where this study was performed, mortality due to malaria decreases markedly after the age of 5 years and clinical episodes are most freauent in children aeed Z-6 vears (GREENWOOD et al. ,- 1987). We have measured the antibody responses to soluble malaria antigens in a crosssection of this population and compared them with malariometric variables. We have also investigated the relationship between humoral immunity to soluble antigens and resistance to symptomatic and asymptomatic malaria infection. Antibodv resoonsesto all the soluble antigens were strongly age-dipendent, but 2 different age profiles were seen. For antigens 1,2,3 and 4 the prevalence of antibody was very low in children under the age of 7 years but increased slowly to a plateau after thcage of 25-35 vears. These antibodies are thus unlikelv to play a- role in anti-toxic immunity but might be involved in anti-parasite immunity. This pattern has been described previously for antibodies to other, defined, malaria antigens such as the circumsporozoite motein (MARSH et al.. 1988) and the merozoite antigen Pfl55RESA (PETERSEN kt al., 1989). Various explanations have been offered for the extremely slow development of the antibody response to these poor imantigens, including the intrinsically munogenicity of the antigen and parasite-induced immunosuppression during acute malaria infection. A significant proportion of adults did not have detectable precipitating antibodies to these antigens. Studies in mice have suggestedthat immune responsesto some malaria antigens are geneticallv restricted to certain major hist&ompatibil& (MHC) loci (GOOD et al.. 1988: LEW et al.. 1989). but there is. as vet. no direct evidence that such restriction is important in man. However, antigenic polymorphism has been well documented for soluble antigens and may explain apparent non-reactivity. The pattern of response to antigens 5 and 7 was in contrast to that described above. The prevalence of these antibodies increased very rapidly, reaching peak prevalence by the ageof 5-8 years and remaining high throughout life. This suggests a possible role for antibodies to Ag5 and Ag7 in antitoxic immunity. The high prevalence in young children of antibodies to An7 is narticularlv interesting since it has previously b& &own that Ag7 is pyrogenic and stimulates the release of both interleukin-1 (JAKOBSEN et al., 1990) and tumour necrosis factor (TAVERNE et al., 1990) from cultured human mononuclear cells. The role of Ag7 in the induction of malaria fever, and the

31 potential for blocking this effect with specific antibodies, is currently being investigated. In terms of estimating protective immunitv against disease it is appropriate-to compare those -&&en who were infected but did not become ill with those children who experienced at least one febrile episode associated with infection. Studies conducted in the samearea have shown that weekly morbidity surveillance may miss up to 25% of malaria episodesand thus some children were probably classified incorrectly (SNOWer al., 1989). Similarly, some asymptomatic, but short-lived, infections may have been missed. Nevertheless, this method of morbidity surveillance is able to detect the clinically protective effect of carriage of the AS haemoglobin genotype, indicating that, despite its imperfections, this epidemiological approach is capable of detecting factors which confer protection against clinical malaria (MARSH et al 1989). When antibody responsesto the various antigens were compared in these 2 groups no significant association between antibody response before the rainy season (May) and clinical protection was seen. It is possible that antibodies to the remaining antigens have no protective function in sivo and that their appearance in immune or partially-immune individuals simply reflects cumulative exposure to infection. Malaria-specific antibodies measured by immunofluorescence and enzyme-linked immunoassay have previously been shown to be related to exposure rather than to infection (MARSH et al., 1989). Also considerable antigenic variation occurs between soluble antigens isolated from different parasite strains and it may be necessary for an individual to acquire immunity to all the isolates circulating in the local environment before protection is achieved, as has be-en shown for the erythrocyte surface neoantigens (MARSH & HOWARD, 1986). This orocess mav be further comolicated bv the raoid generation bf variant parasitis in &JO* (HOW&L, 1985). Cumulative exposure to a range of different parasites may, in itself, explain the gradual development of anti-parasite immunity with age. Finallv, the class of antibodv induced mav be importalii for conferring protect&. Antibodies ieutralizing the stimulation of tumor necrosis factor production by soluble P. yoelii antigens were shown to be IgM and IgG3 (BATE et al., 1990). The CIE assay however does not differentiate between immunoglobulin classes or subclasses. The importance of repeated exposure to malaria parasites for maintaining antibody levels is clearly shown in this study. The prevalence of antibodies in children who showed no evidence of infection during the rainy seasonwas consistently lower in November than in May. This could indicate that these children did not experience a blood-stage infection and sug gests that they did not receive a sufficient number of bites from infected mosquitoes to establish an infection. This is quite possible in this area, where it has been estimated that children receive between 1 and 5 infective bites per year (SNOWet al., 1989). However, 2 of the children showed a 3-fold increase in IFAT. indicating that they experienced infection during th; rainy season. Interestingly, the children without evidence of infection had a higher prevalence of high-titre antibodies to Ag4 than children with infection.

Although antibody prevalence did not increaseafter recent infection, responses were high in comparison with children who had not recently been infected, suggesting that boosting had in fact occurred. However, re-exposure to antigen usually induces a secondary antibody response which is significantly higher than the pre&i&g response, as &en for -&ple with total antimalarial antibody levels and antibodies to the circumsporozoite protein-(MARSH et al., 1988, 1989). The fact that such secondary responsesare not seento soluble antigens suggests that these antigens may stimulate antibody production by mechanisms independent of T cells (ROITT et uf., 1985). Carriage of the sickle cell tnut (AS) is generally believed to reduce the severity of malaria infection, and this was confirmed in the present study. The prevalence of antibodies was generally higher in homozygous (AA) children, again suggesting that recent exposure to soluble antigens is an important determinant of antibody status, It is extremely difficult, with uooulation studies of this kind, to sliow any ciear reia6onship between a particular immune response and protection from disease, and any apparent correlation must be interpreted with caution. Such studies can, however, identifv reswnses which mav be imwrtant and which warran: fu&er investigation. The almost universal recotition of antigens 5 and 7. with hir& titres of anti&dy develop&g at the t&e of p&k malaria morbidity, suggests that these antigens may be involved in the pathogenesisof acute malaria and that responses to them may be the first step in the development of clinical immunity. Further investigations of the role of these soluble antigens are needed. Acknowledgemcnta

Jette Severinsen,Helle Andersen, Angela O’Donnell, Id&a Sambou,Lang Bayo, Lamin Kuyateh, Olalekan Daramola, Simon Correa, Chris Grumnn‘tt and Hamoro Camaraare thanked for excellent technical assistance. The work wassupportedby grant 104.Dan.81445 from the Danish InternationalDevelopmentAgencyand grant TS-20057DKfrom the Commissionof the EuropeanCommunity. This investigationalso receivedfinancialiupport from the UNDP/World Bank/WHOSue&l Proaramme for Research and Training in Tropical Ijiseases. References

Axelsen, N. H. (1983). Intermediate gel immunoelectrophoresis.ScandinavianJournal of Immunology, 17, supplement10, 141-149. Bate,C. A. W., Taveme,J., DavC,A. & Playfair, J. H. L. (1990).Malaria e.xoant&ens induce T&de ndent antibody that blockstheir ability to induceTN r . Zmmunology, -70, 315-320. Clark, I. A. (1987).Cell-mediated*unity in protection a3;: pathologyof malarm.Paraatology Today, 3, 300Good, M. F., Miller, L. H., Kumar, S., Quakyi, J. A., Keister, D., Adams,J. H., Moss,B., Berzofsky,J. A. & Carter,R. (1?88).L+nited &munolo~@cal recqgnitionof ;hnm5y7nalarlavaccmecan&dateanugens.Scaence, 242, Greenwood,B. M., Bradley, A. K., Greenwood,A. M., Byass,P., Jammeh,K., Tulloch, S., Oldfield, F. S. J. & Hayes,R. (1987).Mortality and morbidity from malaria amongchildren in a rural areaof The Gambia,West Africa. Transactions of the Royal Societyof Tropical Medicine and Hygiene, 81, 478-486. Hommel, M. (1985).Antigenic variation in malaria parasites. lmmwwlogy Today, 6, 28-33.

32

Howard, R. J. (1986). Malaria: antigens and host-parasite interactions. In: Toward New Strategies for Vaccines, Pearson, T. W.(editor). New York & Basel: Marcel Dekker, pp. 111-165. Howard, R. J., Panton, L. J., Marsh, K., Winchell, E. J. & Wilson, R. J. M. (1986). Antigenic diversity and size diversity of Plasmodium falciparutn antigens in isolates from Gambian patients. 1: S-antigens. Parasite Zrnmunology, 8, 51-67.Jakobsen, P. H., Jepsen, S. & Agger, R. (1987a). Inhibitory monoclonal antibodies to soluble Plasmodium falciparum antigens. Parasitology Research, 73, 518-523: * Jakobsen, P. H., Theander, T. G., Jensen, J. B., Molbak, K. & Jepsen, S. (1987b). Soluble Plasmodium falciparum antigens contain carbohydrate moieties important for immune reactivity. Journal of Clinical Microbiology, 25, 2075-2079. Jakobsen, P. H., Back, L. & Jepsen,,S. (1988). Demonstration of soluble Plasmodium falcsparum antigens reactive with Limulus amoebocyte lysate and polymyxin B. Parasite Zmmunology, 10, 593-606.

Jakobsen, P. H., Huiid, L., Theander, T. G., Riley, E. M., Grellier, P., Bruun, L., Dalsgaard, K., Schrevel, J. & Jepsen, S. (1990). Isolation and characterization of a soluble antigen complex of Plasmodium falciparum with pyrogenic properties. Acta Pathologica Microbiologica et ZmmufwZogicaScandimavica, section C, in press. Jepsen, S. & Andersen B. J. (1981). Immunosorbent isolation of antigens from the culture medium of in vitro cultivated Plasmodium falciparum. Acta Pathologica Microbiologica et Zmmunologica Scandinavica, section C, 89, 99-103. Jepsen, S. & Axelsen, N. H. (1980). Antigens and antibodies in Plasmodium falciparum malaria studied by immunoelectrophoretic methods. Acta Pathologica Microbiologica et Zmmunologica Scandinavica, section C, 88,

-36L2713 -- - -.

Lew, A. M., Langford, C. J., Pye, D., Edwards, S., Corcoran, L. & Anders, R. F. (1989). Class II restriction in mice io the malaria candidate vaccine ring infected erythrocyte surface antigen (RESA) as synthetic peptides or as expressed in recombinant vaccinia. 3ournul of Zmmurwlogy, 142, 4012-4016. Marsh, K. & Howard, R. J. (1986). Antigens induced on erythrocytes by P. falciparum: expression of diverse and conserved determinants. Science, 231, 15&153. Marsh, K., Hayes, R. H., Carson, D. C., Otoo, L., Shenton, F., Byass, P:, Zavala, F. & Greenwood, B. M. (1988). Anti-sporozoite antibodies and immunity to malaria in a rural Gambian population. Transactions of the ~3~~~30ciety

of Tropical Medicine

and Hygiene,

66,

Marsh, K.,’ Otoo, L., Hayes, R. J., Carson, D. C. & Greenwood, B. M. (1989). Antibodies to blood stage antigens of Plasmodium falciparum in rural Gambians and their relation to protection against infection. Transactions

of the Royal Society of Tropical Medicine and Hygiene, 83,

293-303. Menon, A., Snow, R. W., Otoo, L. N. & Greenwood, B. M. (1987). Decline in sensitivity of Plasmodium falcipanrm to chloroquine in the Gambia. Lancet, i, 1029-1030. Panton, C. J., Tulloch, S., Bradley, A. K. & Greenwood, B. M. (1985). Susceptibility of Plasmodium falcipancm in The Gambia to pyrimethamine, Maloprim and chloroquine. Transactions of the Royal Society of Tropical Medicine and Hygiene, 79, 484-490.

Petersen, E., Hogh, B., Perlmann, H., Kabilan, L., Troye-Blomberg, M., Marbiah, N. T., Hanson, A. P., Bjorkman, A. & Perhnann, P. (1989). An epidemiological study of humoral and cell-mediated immune res onse to the Plasmodium falciparum antigen Pfl55lRE ! A in adult Liberians. American Journal of Tropical Medicine and Hygiene, 41, 394-402.

Playfair, J. H. L., Taverne, J., Bate, C. A. W. & de Souza, J. B. (1990). The malaria vaccine: anti-parasite or anti-disease? Zmmunology Today, 11, 25-27. Riley, E. M., Jepsen, S., Andersson, G., Otoo, L. N. & Greenwood. B. M. (1988). Cell-mediated immune responses to- Plasmodium fat’ciparum antigens in adult Gambians. Clinical and Experimental Zmmunology, 71, 377-382.

Riley, E. M., Allen, S. J., Bennett, S., Thomas, P. J., O’Donnell, A., Lindsay, S. W., Good, M. F. & Greenwood, B. M. (1990). Recognition of dominant T cell-stimulating epitopes from the circumsporozoite protein of Plasmodium falcipancm and relationship to malaria morbidity in Gambian children. Transactions of the Royal Society of Tropical Medicine and Hygiene, 84, in press. Roitt, I., Brostoff, J. & Male, D. (1985). Zmmunology, chapter 8. London: Gower Medical Publishing. Snow, R. W., Menon, A. & Greenwood, B. M. (1989). Measuring morbidity from malaria. Annals of Tropical Medicine and Parasitology, 83, 321-323. Taverne, J., Bate, C. A. W., Kwiatowski, D., Jakobsen, P. H. & Playfair, J. H. L. (1990). Two soluble antigens of Plasmodium falciparum induce tumor necrosis factor releasefrom macrophages.Infection and Zmmuniry,58, in Thel%& T G. Bygbjerg I C. Jepsen S Svenson M Khara&. A?. Larsen.‘P: Bf & Benhtkk K. (lb86j: Proliferation induced Gy Plasmodium falciparum antigen and interleukin-2 production by lymphocytes isolated from malaria-immune individuals. Infection and Zmmunity, 53, 221-225. Trager, W. & Jensen, J. B. (1976). Human malaria parasites in continuous culture. Science, 193, 673-675. Wilson, R. M., McGregor, I. A., Hall, P., Williams, K. & Bartholomew, R. (1969). Antigens associatedwith Plasmedium falciparum infections in man. Lancet, ii, 201-205. Received 26 March 1990; revised 19 June accepted for publication 19 June 1990

1990;