Impact of red blood cell polymorphisms on the antibody ... - CiteSeerX

Microbes and Infection 8 (2006) 1260e1268 www.elsevier.com/locate/micinf

Original article

Impact of red blood cell polymorphisms on the antibody response to Plasmodium falciparum in Senegal Jean Birame Sarr a, Ste´phane Pelleau a, Ce´cile Toly a, Juliette Guitard a, Lassana Konate´ b, Philippe Deloron c, Andre´ Garcia a, Florence Migot-Nabias a,* a

Institut de Recherche pour le De´veloppement (IRD), Unite´ de Recherche 010: Sante´ de la me`re et de l’enfant en milieu tropical, BP 1386, Dakar, Senegal b Laboratoire d’Ecologie Vectorielle et Parasitaire, De´partement de Biologie Animale, Faculte´ des Sciences et Techniques, Universite´ Cheikh Anta Diop (UCAD), Dakar, Senegal c IRD, UR 010, Laboratoire de Parasitologie, Faculte´ de Pharmacie, 4 avenue de l’Observatoire, 75006 Paris, France Received 23 June 2005; accepted 7 December 2005 Available online 3 February 2006

Abstract The evidence of protection afforded by red blood cell polymorphisms against either clinical malaria or Plasmodium falciparum blood levels varies with the study site and the type of malaria transmission. Nevertheless, no clear implication of an antibody-related effect has yet been established in the protection related to red blood cell polymorphisms. We performed a prospective study, where plasma IgG and IgG subclasses directed to recombinant proteins from the merozoite surface protein 2 (MSP2/3D7 and MSP2/FC27) and the ring-infected erythrocyte surface antigen (RESA) were determined in a cohort of 413 Senegalese children before the annual malaria transmission season. The antibody response was dependent on age, and to a lesser extent, on the village of residence. IgG3 responders to all proteins, IgG responders to RESA and MSP2/ 3D7, as well as IgG2 to RESA and IgG1 responders to MSP2/3D7, presented enhanced mean values of parasite density, as evaluated during an 18-month follow-up. The levels of IgG and IgG3 to MSP2/3D7 were negatively associated with the risk of occurrence of a malaria attack during the following transmission season. Compared to normal children, sickle cell trait carriers presented lower levels of IgG to MSP2/3D7. Similarly, G6PD A girls had lower levels of IgG and IgG3 to MSP2/FC27 than did G6PD normal girls. The impact of these particular genetic polymorphisms on the modulation of the antibody response is discussed. Ó 2006 Elsevier SAS. All rights reserved. Keywords: Plasmodium falciparum malaria; Gene polymorphisms; Immunity; Antibody; Senegal

1. Introduction Many studies have related an influence of human genetic factors affecting red blood cells (sickle-cell trait, glucose 6 phosphate dehydrogenase (G6PD) deficiency, aþ-thalassaemia, ABO blood groups) to the susceptibility/resistance to Abbreviations: aMPD, adjusted mean parasite density; AU, arbitrary unit; ELISA, enzyme-linked immunosorbent assay; G6PD, glucose 6 phosphate dehydrogenase; IgG, immunoglobulin G; MSP2, merozoite surface protein 2; OD, optical density; RESA, ring-infected erythrocyte surface antigen. * Corresponding author. UR 010/IRD, 01 BP 4414 RP, Cotonou, Benin. Tel.: þ229 21 30 98 21; fax: þ229 21 30 88 60. E-mail address: [email protected] (F. Migot-Nabias). 1286-4579/$ - see front matter Ó 2006 Elsevier SAS. All rights reserved. doi:10.1016/j.micinf.2005.12.005

mild and severe forms of Plasmodium falciparum malaria. Mechanisms of this influence are located at both the erythrocyte and the immune levels. With regard to the mutations or deletions of the globin gene, several mechanisms have been proposed, such as an increased sensitivity to oxidizing stress, increasing parasite phagocytosis [1e3], a decrease in the parasite growth in abnormal red blood cells [4,5], or a modification of erythrocyte surface antigens enhancing immune recognition [6,7]. Nevertheless, the increase in antibody-related protection against malarial surface antigens that was established in several studies [6,7] could not be confirmed by others [8e10]. In the case of G6PD deficiency, a selective phagocytosis of oxidantdamaged ring-parasitized deficient erythrocytes has been

J.B. Sarr et al. / Microbes and Infection 8 (2006) 1260e1268

identified [11], and a reduced expression of the parasite G6PD could also be related to impaired growth of parasites in G6PDdeficient hosts [12]. The ABO blood group antigens may interfere with severe malaria by facilitating (group A) or by limiting (group O) the erythrocyte rosetting [13]. If these erythrocyte genetic factors affect susceptibility to infection, the modifications of the immune response that they generate remain unclear, and need new immuno-genetic studies on malaria in order to be better understood. To elucidate the interactions between selected host defence mechanisms, a longitudinal survey was conducted in Senegal in an area of seasonal malaria transmission. A parasitological follow-up was performed on a cohort of 464 children during 18 months, and a clinical survey was assured among part of this cohort during the malaria transmission season. Several red blood cell polymorphisms were determined and the following prevalence rates were found: sickle-cell trait (13%), G6PD A-deficient allele (harboured by 12% of males and 20% of females), a3.7 type of aþ-thalassaemia (24%) and blood groups O (52%), A (27%), B (18%), and AB (3%). The sickle cell trait or the presence of non-O blood groups was associated with protection against clinical malaria attacks, while aþ-thalassaemic children were protected against high parasite densities (Migot-Nabias, unpublished). In the present study, plasma IgG and IgG subclasses directed to recombinant proteins from the P. falciparum asexual blood stage antigens merozoite surface protein 2 (MSP2) and ring-infected erythrocyte surface antigen (RESA) were evaluated by enzyme-linked immunosorbent assay (ELISA). The choice of MSP2 and RESA antigens relied on the fact that they are the targets of protective immunoglobulins and are promising candidates for a sub-unit vaccine [14]. Indeed, they both induce protective immunity in experimental models [15,16] and are recognized by naturally acquired antibodies [17,18]. The influence of age on the presence of antibody reactivity to MSP2 and RESA recombinant proteins has already been demonstrated, and the levels of specific antibodies to MSP2 and RESA have also been related to reduced malaria morbidity [19e21]. It was of interest to dissociate the IgG response into its subclass components in order to evaluate the fine impact of red blood cell polymorphisms. The response to defined P. falciparum asexual-stage antigens is dominated by the cytophilic IgG1 and IgG3 isotypes [22]. As cytophilic antibodies mediate the in vitro phagocytosis of P. falciparum infected erythrocytes [23], they could therefore be associated with the development of protection. In fact, immuno-epidemiological studies have reported a predominance of IgG3 directed against MSP2 [18,24], as well as their strong relation to protection [25,26]. For RESA, controversial results afforded either a negative [27] or a positive [28] relation of IgG1 to protection, as well as no [27] or a positive [28] relation of IgG3 to protection. These discrepancies could be attributed to the epidemic (in the highlands of Madagascar) or highly endemic (in Papua New Guinea) characteristics of malaria in the two study areas. Concerning the non-cytophilic IgG2 and IgG4, if a blocking role is recognized for IgG4, a protective role was attributed to the non cytophilic IgG2 directed to RESA and MSP2 epitopes [29].

1261

In the present study, our objectives were (i) to determine the fine antibody responses to MSP2 and RESA recombinant proteins in a cohort of Senegalese children, (ii) to relate antibody responses to red blood cell polymorphisms by taking into account the effects of risk factors and (iii) to relate antibody responses to parasitological and clinical data. Results are discussed in the context of the impact of red blood cell polymorphisms on the modulation of the immune response against P. falciparum. 2. Materials and methods 2.1. Subjects The study took place in two villages (Diohine and Toucar) in the Niakhar area, 115 km south-east of Dakar. Malaria is endemic in the area and its transmission is seasonal, from September to December, with an average of 9e12 infective bites per person per year [30]. A cohort of 464 unrelated children aged from 2 to 10 years, living permanently in the area, was constituted from a larger cohort of children grouped into siblings and already enrolled into a malaria genetic epidemiology programme [31]. Our study group included the 169 youngest children from each nuclear family of the associated programme, for whom an active clinical survey aimed at detecting any malaria attack was performed. It was completed to 464 by recruiting unrelated children from neighbouring households. Informed consent was given parents of all children, and the protocol was approved by the Ethics Committee of the Health Ministry of Senegal. 2.2. Field study As described (Migot-Nabias, unpublished), ten parasitological measurements were performed from thick blood smears obtained by finger-prick in June, September, October and November 2002 as well as in January, April, June, September, October and December 2003. After staining the smears with Giemsa, malaria parasites were counted and the parasite density was defined as the number of P. falciparum parasites per 100 leucocytes. The 22 children who totalized less than two thick blood smears were excluded from the analysis, which was therefore performed with 442 children. A unique quantitative variable accounting for the level of infection was attributed to each child by using a linear mixed model (‘‘PROC MIXED’’ SASÒ procedure), which took into account the repeated parasitological measurements and fixed variables such as age, sex, village of residence, period of malaria transmission, year of the follow-up, presence of chloroquinuria and of concomitant P. malariae infection. From this model, predictions (best linear unbiased predictors) of the random variable were computed for each child and were referred to in the text as adjusted mean parasite density (aMPD). An active survey of malaria attacks was conducted during the transmission period of 2002 for the sub-group of 169 children. Axillary temperature was measured twice a week, and a thick blood smear was done in the case of fever (axillary

1262


temperature higher than 37.5 C) or history of fever. A malaria attack was defined as the association of fever with a parasite density higher than 2500/ml without any other apparent cause of fever. Morbidity data generated a qualitative variable according to the absence or the presence of at least one malaria attack during the transmission season. At initiation of the study in June 2002, before the annual malaria transmission season, venous blood was collected into heparinized VacutainerÒ tubes (Becton Dickinson, Meylan, France). ABO blood groups were determined by serology, and the other red blood cell polymorphisms were detected using genomic DNA. Plasma was collected for 413 children. The other cases corresponded either to children who refused to have blood drawn or to samples that lacked blood. 2.3. Antibody measurements ELISA was used to determine plasma antibodies directed to the three recombinant proteins MSP2/3D7, MSP2/FC27 and RESA from the asexual blood stages of P. falciparum. All were expressed in Escherichia coli (gift from Robin F Anders, La Trobe University, Victoria, Australia), and were stocked at 1 mg/ml in sterile distilled water before being used at a final concentration of 0.5 mg/ml in a carbonate buffered (pH 9.6) solution. Plasma samples were diluted (IgG2 and IgG4: 1/ 10; IgG1 and IgG3: 1/50; total IgG: 1/200) in 5% bovine serum albumin, 1% Tween-20 and 0.02% sodium azide in phosphate buffered saline and tested in duplicate. These plasma dilutions were chosen in order to obtain a sufficient range of optical density (OD) values from the negative to the positive reference pooled control plasmas included in each plate. The monoclonal antibodies used for determination of the immunoglobulin isotypes were purchased from Tebu (Le Perrayen-Yvelines, France) and were mouse anti-IgG1 (clone HP 6069) at 1 mg/ml, anti-IgG3 (clone HP 6047) at 0.5 mg/ml, anti-IgG2 (clone HP 6014) and anti-IgG4 (clone HP 6023) at 0.25 mg/ml. Two polyclonal antibodies conjugated to alkaline phosphatase were used: an anti-human IgG (Fc specific) (Sigma, Saint-Quentin-Fallavier, France) diluted 1/2000 for IgG, and a goat anti-mouse IgG (H þ L) (Tebu) diluted 1/ 1000 for IgG2 and IgG4, 1/2000 for IgG1 and 1/4000 for IgG3. Bound enzyme was detected with p-nitrophenylphosphate and the OD was read at 405 nm (reference filter 620 nm). Reference positive and negative pooled control plasmas were included in each plate and results were expressed in arbitrary units (AU) calculated from the formula 100 [ln(OD test plasma) ln(OD Pool)]/[ln(OD Pool þ ) ln(OD Pool)] [32]. The thresholds for positivity were determined from the mean reactivities þ 2 SD of 30 plasma samples from non-immune subjects. 2.4. Statistical analysis Differences in proportions were analysed using the Chi square test. Differences in means were tested by the nonparametric ManneWhitney U-test or KruskaleWallis test (for more than 2 groups to be compared), except for age,

where the Student unpaired t-test was employed as age was normally distributed. Statview 5.0 (SAS Institute Inc., Cary, NC) was used for these calculations. The associations between specific immune responses and covariates (village, age, P. falciparum carriage at blood drawing, genetic factors) found to be significant in the univariate analysis were investigated by multiple linear regression analysis using STATA (StataCorp. 1999, Release 6.0). For all tests, P values of less than 0.05 were considered significant. 3. Results Characteristics of the children (age, parasitological data at enrolment and genetic factors) did not differ according to the groups considered for collection of parasitological, immunological or clinical data (Table 1). At the end of the parasitological follow-up, 82% of the children retained for analysis totalized at least 8 thick blood smears, including 78 children (18%), 183 children (41%) and 102 children (23%) with 8, 9 and 10 thick blood smears, respectively. At enrolment, among the 413 children for whom antibody responses were investigated, 121 (30%) harboured P. falciparum parasites and 17 (4%) presented mixed P. falciparum and P. malariae infections. Only pure P. falciparum infections were considered in analyses involving parasite carriage at enrolment. All children but one were asymptomatic for malaria. 3.1. Anti-MSP2 and anti-RESA antibody responses For each protein, IgG responses predominated, with prevalence rates ranging from 81% to 95%. The most prevalent subclass was IgG3 against both MSP2 proteins, and IgG1 against RESA, although the highest median values of responders were recorded for IgG3 for all proteins. The lowest prevalence was consistently observed with IgG2 (Table 2). 3.1.1. Univariate analysis of anti-MSP2 antibody responses 3.1.1.1. Effects of risk factors. The antibody responses to MSP2 (MSP2/3D7 and MSP2/FC27) were influenced by several factors. IgG2 and IgG4 responses to both proteins were higher in children from Diohine than from Toucar (all P < 0.008), children from Diohine being older (5.0 1.8 years vs. 4.6 1.2 years, P ¼ 0.008). IgG, IgG1 and IgG3 responders to MSP2 were older than non responders, and similarly IgG4 responders to MSP2/3D7 were older (all P < 0.01). Children harbouring P. falciparum parasites at blood drawing had higher IgG, IgG1 and IgG3 levels of responses to MSP2 than others (all P < 0.0001). 3.1.1.2. Effects of host genetic factors. Sex, ABO blood group and aþ-thalassaemia were not related to the antibody response. Children with normal haemoglobin (AA) exhibited higher levels of IgG, IgG1 and IgG3 to MSP2 than sickle cell trait carriers (AS) (all P < 0.04; Fig. 1). G6PD normal girls had higher levels of IgG and IgG3 to MSP2 than G6PD A heterozygous and homozygous girls (all P < 0.02;


1263

Table 1 Characteristics of the children at enrolment

No. of children Sex ratio (M/F) Age (years): mean SD Parasitological datad P. falciparum infection P. malariae infection P. falciparum and P. malariae infection Host genetic factors Blood group O non O Haemoglobin AA AS aþ-Thalassaemia Normal Mutatione G6PD in females Normal G6PD A variante G6PD in males Normal G6PD A variantf

Parasitological groupa

Immunological sub-groupb

Clinical sub-groupc

442 0.97 (218/224) 4.9 1.6

413 1.03 (210/203) 4.9 1.6

169 0.99 (84/85) 5.2 1.2

128/427 (30%) 7/427 (2%) 20/427 (5%)

121/407 (30%) 7/407 (2%) 17/407 (4%)

42/167 (25%) 1/167 (1%) 9/167 (5%)

226 (52%) 208 (48%)

212 (51%) 200 (49%)

88 (53%) 79 (47%)

374 (87%) 58 (13%)

354 (86%) 57 (14%)

143 (86%) 23 (14%)

275 (76%) 89 (24%)

262 (75%) 86 (25%)

104 (74%) 37 (26%)

174 (80%) 43 (20%)

163 (81%) 38 (19%)

71 (86%) 12 (14%)

183 (88%) 25 (12%)

181 (88%) 24 (12%)

71 (88%) 10 (12%)

Missing values for blood group (n ¼ 8), haemoglobin (n ¼ 10), aþ-thalassaemia (n ¼ 78) and G6PD (n ¼ 17) corresponding to samples that lacked blood or DNA, or for which technical problems were encountered for specific determinations. b Missing values for blood group (n ¼ 1), haemoglobin (n ¼ 2), aþ-thalassaemia (n ¼ 65) and G6PD (n ¼ 7). c Missing values for blood group (n ¼ 2), haemoglobin (n ¼ 3), aþ-thalassaemia (n ¼ 28) and G6PD (n ¼ 5). d Thick blood smears obtained for 427, 407 and 167 children from the respective parasitological, immunological and clinical groups. e Mutation indicated in both heterozygous and homozygous states. f Mutation corresponds to G6PD A hemizygous genotype. a

Fig. 2). Interestingly, we observed a gradual decrease of the IgG, IgG1 and IgG3 levels to MSP2/3D7 (P ¼ 0.0006, P ¼ 0.05 and P ¼ 0.0003, respectively) and MSP2/FC27 (P ¼ 0.006, P ¼ 0.007 and P ¼ 0.01) from normal children to those carrying either sickle cell trait or G6PD A variant (n ¼ 93) and to those carrying both abnormalities (n ¼ 12).

3.1.2.2. Effects of host genetic factors. AA children exhibited higher levels of IgG, IgG1 and IgG3 than AS children (all P < 0.03; Fig. 1). A gradual decrease of these antibodies operated from normal children to those carrying either sickle cell trait or G6PD A variant (n ¼ 93) and to those carrying both abnormalities (n ¼ 12) (P ¼ 0.01, P ¼ 0.06 and P ¼ 0.04 for IgG, IgG1 and IgG3 respectively).

3.1.2. Univariate analysis of anti-RESA antibody responses 3.1.2.1. Effects of risk factors. The antibody response to RESA also varied with several factors. The IgG1 response was lower in Diohine than in Toucar (P ¼ 0.0003). Children harbouring P. falciparum parasites at blood drawing had higher levels of IgG and IgG subclasses than others (all P < 0.0002).

3.1.3. Multivariate analysis of anti-MSP2 and anti-RESA antibody responses 3.1.3.1. Effects of risk factors. A multivariate analysis confirmed higher IgG2 and IgG4 levels to respectively MSP2/ FC27 and MSP2/3D7 proteins in Diohine than in Toucar, as

Table 2 Thresholds for positivity, prevalence rates and median values of antibody responses to MSP2/3D7, MSP2/FC27 and RESA recombinant proteins MSP2/3D7 (n ¼ 413)

IgG IgG1 IgG2 IgG3 IgG4 a

MSP2/FC27 (n ¼ 410)

RESA (n ¼ 408)

Positivity threshold (AU)

No. of responders (%)

Median values (AU)a



Median values (AU)



Median values (AU)

30.9 55.1 34.0 42.7 18.8

334 300 116 332 230

77.9 77.4 45.2 93.2 42.0

18.2 45.7 72.0 30.4 47.9

388 280 37 362 147

82.3 62.5 82.4 88.7 58.2

10.5 10.9 42.4 11.1 53.1

342 226 133 185 194

47.1 40.7 58.6 74.4 70.7

(81) (73) (28) (80) (56)

(57.6e92.2) (63.6e92.7) (38.7e53.5) (77.7e99.8) (30.3e54.2)

(95) (68) (9) (88) (36)

(46.0e103.4) (54.3e74.3) (78.1e90.7) (61.8e101.0) (52.8e66.7)

Median values (25the75th percentiles), including responders only, and expressed in arbitrary units (AU).

(84) (55) (32) (45) (47)

(29.2e68.5) (25.4e62.3) (49.8e71.1) (37.2e109.6) (62.0e82.2)

1264


Fig. 2. Gradation of the antibody levels to MSP2/3D7 and MSP2/FC27 according to the allelic carriage of the G6PD A variant among girls. P value of the KruskaleWallis test. Box-whisker plots illustrate medians with 25th and 75th percentiles, and whiskers for 10th and 90th percentiles.

on the IgG3 levels to MSP2/3D7 was observed in girls carrying both abnormalities. Fig. 1. Differences in the levels of specific antibodies between AA and AS children. P value of the ManneWhitney U-test. Box-whisker plots illustrate medians with 25th and 75th percentiles, and whiskers for 10th and 90th percentiles.

well as a lower level of IgG1 to RESA in Diohine (all P < 0.0001). P. falciparum carriers at blood drawing remained with higher levels of IgG, IgG1 and IgG3 to all proteins than non carriers (all P < 0.0001). Age-related increase of the levels of IgG, IgG1 and IgG3 to MSP2 and of IgG4 to MSP2/3D7 was maintained (all P < 0.02). 3.1.3.2. Effects of host genetic factors. The multivariate analysis confirmed that higher levels of IgG and IgG3 to MSP2/ 3D7 and of IgG1 to RESA were present in AA than AS children (P ¼ 0.01, P ¼ 0.06 and P ¼ 0.06, respectively). Similarly, higher levels of IgG3 to MSP2/3D7 and of IgG and IgG3 to MSP2/FC27 were observed in normal than in mutated G6PD girls (P ¼ 0.06, P ¼ 0.004 and P ¼ 0.008 respectively). No cumulative effect of sickle cell trait and G6PD A variant

3.1.4. Stratification of the antibody response according to age The age-linked effects of sickle-cell trait (in all children) and G6PD deficiency (in girls) on the antibody responses were verified when both age and these genetic factors were introduced in the model, and they were manifest for young children, particularly those aged from 4 to 5 years (Table 3). 3.2. Immunological determinants in relation to prospective parasitological and clinical data Children presenting with IgG3 against RESA and/or MSP2 had higher parasite densities adjusted on age and other risk factors (aMPD) during the subsequent follow-up (Table 4). Similarly, IgG responders to RESA and MSP2/3D7, as well as IgG2 and IgG1 responders to RESA and MSP2/3D7 respectively, presented higher aMPD values than non responders. When considering the antibody responses related to both


1265

Table 3 Influence of age on antibody responses of AS children and G6PD A girls Pa

2e3 years AA/AS IgG to MSP2/3D7 AA AS IgG3 to MSP2/3D7 AA AS

n ¼ 106/11

G6PD nl/G6PD A IgG to MSP2/FC27 nl A

n ¼ 42/13

a b

4e5 years n ¼ 175/31

6e7 years

P

8e10 years

n ¼ 56/9

P

n ¼ 13/5

71.0 (29.4; 86.4)b 23.0 (11.1; 71.5)

0.03

71.3 (43.9; 89.9) 54.2 (28.6; 74.5)

0.007

79.7 (49.3; 93.9) 65.9 (28.8; 68.4)

0.05

93.6 (49.3; 103.0) 53.4 (45.8; 98.7)

0.52

86.5 (39.1; 96.2) 34.4 (23.9; 85.9)

0.08

90.1 (65.2; 98.8) 72.2 (33.6; 92.1)

0.002

93.7 (61.7; 99.9) 86.2 (35.6; 92.6)

0.11

99.8 (74.8; 101.7) 81.8 (63.9; 102.7)

0.96

n ¼ 80/20

69.5 (32.8; 96.8) 35.5 (28.7; 70.4)

0.10

86.3 (60.9; 107.0) 61.6 (29.2; 84.3)

n ¼ 28/5 0.01

94.6 (59.7; 106.7) 60.4 (39.7; 112.5)

n ¼ 7/0 0.88

95.9 (64.5; 109.5) e

e

P value of the ManneWhitney U-test. Median values (25th; 75th percentiles) in arbitrary units (AU).

genetic factors and aMPD, we observed higher aMPD among responders (IgG and IgG3 to MSP2/3D7 and IgG3 to MSP2/ FC27) in the lonely group of AA children (all P < 0.004), but not among AS children. A similar observation of higher aMPD values was made for IgG3 responders to MSP2/FC27 versus non responders among G6PD A normal girls (P ¼ 0.02), but not among G6PD A girls. In parallel, children who did not suffer from a malaria attack during the following transmission season produced higher levels of IgG and IgG3 to MSP2/3D7 than others, the IgG values being adjusted on risk factors (age, sex, village of residence and presence of P. falciparum at blood drawing, Table 5). When considering the antibody responses related to both genetic factors and to presence of a malaria attack, we found higher antibody levels (IgG and IgG3 to MSP2/3D7) Table 4 Differences in prospective aMPD values according to antibody response to recombinant proteins Responders

RESA IgG IgG2 IgG3

n

aMPD

0.06 (0.31; 0.23) 0.02 (0.21; 0.27) 0.00 (0.24; 0.31)

339

0.21 (0.38; 0.03) 0.16 (0.36; 0.15) 0.15 (0.36; 0.12)

MSP2/FC27 IgG3 0.07 (0.29; 0.21)

Pa

Non responders

aMPDb

MSP2/3D7 IgG 0.04 (0.29; 0.23) IgG1 0.05 (0.29; 0.25) IgG3 0.03 (0.29; 0.25)

a

P

130 183

331 298 328

358

0.27 (0.39; 0.03) 0.22 (0.38; 0.07) 0.28 (0.39; 0.04) 0.13 (0.38; 0.16)

n 65

0.008

273

< 0.0001

221

0.0004

78

< 0.0001

111

0.0005

81

< 0.0001

in asymptomatic children only in the group of AA (both P < 0.02) but not of AS individuals.

4. Discussion Our main result is the lower levels of IgG and IgG3 to MSP2 in AS children and in G6PD A girls than among their normal counterparts. These differences were highest in the 4e 5-year-old children, illustrating an age-linked effect of these red blood cell abnormalities on malaria-specific immune responses [33]. Concerning G6PD A, the antibody response decreased gradually from normal to heterozygous and to homozygous girls. Almost all IgG responses were increased in children presenting with parasites at time of blood sampling. This general enhancement of the antibody responses could result from a booster effect of the presence of long-lasting parasites. IgG and IgG3 levels to MSP2/3D7 were negatively associated to the risk to present a malaria attack. This could be demonstrated among the whole group of children under clinical survey, and more particularly among AA but not AS children. We reported elsewhere that AS children from the same cohort suffered from less malaria attacks than AA children, but controlled similarly their parasite density (MigotNabias, unpublished). Brought together, these results demonstrate that AS children are protected against mild malaria Table 5 Differences in antibody levels to MSP2/3D7 in children having presented (n ¼ 74) or not (n ¼ 90) a malaria attack during the following transmission season MSP2/3D7

48

0.004

P value of the ManneWhitney U-test. b Median value (25the75th percentiles) of the Mean Parasite Density adjusted on fixed variables (age, sex, village of residence, period of malaria transmission, year of the follow-up, presence of chloroquinuria and of concomitant P. malariae infection) and named aMPD.

IgG No malaria attack At least one malaria attack IgG3 No malaria attack At least one malaria attack

Adjusted AUa

Pb

0.08 (0.29; 0.20) 0.10 (0.55; 0.04)

0.009

0.13 (0.08; 0.23) 0.03 (0.15; 0.16)

0.01

a Median values (25the75th percentiles) of antibody levels expressed in arbitrary units (AU) and adjusted on risk factors (age, sex, village of residence, presence of P. falciparum at blood drawing). b P value of the ManneWhitney U-test.

1266


through another mechanism than the cytophilic antibody response to MSP2/3D7. Although it is unusual that amounts of IgG3 exceed those of IgG1 in response to a protein antigen [34], the predominance of IgG3 to MSP2, and that of IgG1 to RESA are now well known. The higher level of IgG and IgG3 to MSP2/ 3D7 in children who did not experience a malaria attack is in full agreement with observations from Papua New Guinea [21] and The Gambia [25,26]. Moreover, individuals infected with parasites from the 3D7 allelic family of msp2 are less susceptible to clinical malaria than those carrying parasites from the FC27 allelic family [35]. The IgG2 response was low for all proteins and was not related to red blood cell polymorphisms or clinical and parasitological data. This suggests that no many children under study carried the mutated allele H131 encoding the FcgRIIA receptor that was found critical in another African population to ensure the binding of IgG2 to FcgRIIA and the following activation of effector cells detrimental for P. falciparum [29]. The dry season is characterized in the Niakhar area by a very low level or even the absence of malaria transmission during an average of 8 months [30]. One third of the children presented with parasites at the end of the dry season, reflecting long-lasting infections. Parasite infection at enrolment was positively associated with both anti-RESA and anti-MSP2 antibodies, and with mean parasite density, adjusted on risk factors during subsequent follow-up. This indicates that those antibodies were rather able to limit parasite density than to clear parasites, and allow therefore to avoid the evolution towards a malaria attack, as shown by the protective IgG and IgG3 response to MSP2/3D7. Similar data were obtained by Metzger et al., [26], who suggested that the antibodies measured at the end of the dry season correlated with the presence of memory B cells of the corresponding specificities that were the real actors of clinical protection. Strikingly, the antibody response was decreased in AS individuals. Diatta et al., [36] reported that AS individuals produced higher levels of IgG against live infected red blood cells, and suggested an increased immune phagocytosis of AS erythrocytes through the efficient presence of the cytophilic IgG1 and IgG3 [23]. This observation was limited to 20 AS individuals from a Senegalese village, and was not confirmed in the 12 AS individuals from another village of the same area. The carriage of the sickle-cell trait was associated to enhanced IgG responses, particularly IgG1 and IgG4, to parasite-derived variant surface antigens [37]. These antigens correspond to the ‘‘neo-antigens’’ of the infected erythrocyte surface evoked by Marsh et al. [6] in The Gambia, where higher titres of antibodies were reported in AS than in AA children. We may suggest that these two mechanisms, observed either with a conserved parasite strain or variant surface antigens, may be additive for protection of AS individuals against clinical malaria. Hitherto, investigations of antibody responses with specificity for other P. falciparum asexual stage antigens have failed to detect differences in the prevalence or the magnitude of such responses in AS and AA children. It was notably the case for antibodies to infected erythrocytes, schizont extract, circum-

sporozoite protein, RESA and MSP1 in The Gambia [6,8], Sudan [38], and Gabon [39]. A single study reported lower prevalence rates of IgG to RESA in AS than in AA young Cameroonian children [10]. To date, a single study compared antibodies to MSP2 according to the haemoglobin type in Gabon. Similar prevalence rates of IgG and IgG subclasses of MSP2 family-specific antibodies were observed in AA and AS haemoglobin groups, except for IgG2 being more prevalent in AS children [40]. It was proposed that IgG2 in AS plasmas led to a lower efficiency of the cytophilic IgG3 to reduce the parasite burden, explaining the higher complexity of P. falciparum infections in AS isolates [41]. Some explanations have been provided for similar and even lower specific antibody levels in AS than in AA subjects. A predominance of B-cell activation and a marked reduction in T-cell mediated responses to malarial antigens in AA individuals, as compared to AS ones, was interpreted as an effect rather than a cause of protection: the red cell disorders, by bringing down parasite density either physiologically at the red cell level or through better clearance by splenic macrophages, might contribute to present different proportions of malarial antigens, at reduced doses, to the immune system of AS and AA individuals [42]. The lower antibody levels in both AS vs. AA, and G6PD A vs. normal individuals corroborated this idea of an effect rather than a cause, due to different mechanisms operating at the red cell level in these two abnormalities. In a similar way, the seasonality of malaria transmission in the Niakhar area leads to a limited contact between host and parasite, and to a low pyrogenic threshold of P. falciparum parasites. The lower parasite densities in aþ-thalassaemic children may prevent them from reaching this pyrogenic threshold. In conclusion, a better knowledge of the immune mechanisms induced by the presence of red blood cell polymorphisms found associated with protection against infection by P. falciparum is crucial in order i) to reproduce these mechanisms by vaccination and ii) to get a picture of the particularities of the immune responses of the carriers of these polymorphisms when evaluating the efficiency of a vaccine trial using MSP2 and/or RESA antigens.

Acknowledgements We thank all the inhabitants of Diohine and Toucar who took part in the surveys as well as the health care agents of these two villages. We are grateful to C. Chevillard (INSERM U399, Marseilles, France) and to R. Nabias (Hoˆpital Principal, Dakar, Senegal) for laboratory activities. We are indebted to R.F. Anders (La Trobe University, Victoria, Australia) for the gift of recombinant proteins. We also thank the UR 010 team of Dakar for field activities and microscopic examinations: L. Barboza, R. Ehemba, P. Niokhor, P. Senghor, R. Senghor, S. Senghor. This work was supported by the French Research Ministry Programme PALþ (2001), the Institut de Me´decine et d’Epide´miologie Africaines (IMEA), Paris, France, and the Institut de Recherche pour le De´veloppement.


References [1] Y. Yuthavong, P. Wilairat, Protection against malaria by thalassaemia and haemoglobin variants, Parasitol. Today 9 (1993) 241e245. [2] A.C. Senok, K. Li, E.A. Nelson, M. Arumanayagam, C.K. Li, Flow cytometric assessment of oxidant stress in age-fractionated thalassaemic trait erythrocytes and its relationship to in vitro growth of Plasmodium falciparum, Parasitology 116 (1998) 1e6. [3] K. Ayi, F. Turrini, A. Piga, P. Arese, Enhanced phagocytosis of ringparasitized mutant erythrocytes: a common mechanism that may explain protection against falciparum malaria in sickle trait and beta-thalassemia trait, Blood 104 (2004) 3364e3371. [4] G. Pasvol, D.J. Weatherall, R.J. Wilson, Cellular mechanism for the protective effect of haemoglobin S against P. falciparum malaria, Nature 274 (1978) 701e703. [5] K. Pattanapanyasat, K. Yongvanitchit, P. Tongtawe, K. Tachavanich, W. Wanachiwanawin, S. Fucharoen, D.S. Walsh, Impairment of Plasmodium falciparum growth in thalassemic red blood cells: further evidence by using biotin labeling and flow cytometry, Blood 93 (1999) 3116e3119. [6] K. Marsh, L. Otoo, R.J. Hayes, D.C. Carson, B.M. Greenwood, Antibodies to blood stage antigens of Plasmodium falciparum in rural Gambians and their relation to protection against infection, Trans. R. Soc. Trop. Med. Hyg. 83 (1989) 293e303. [7] G.A. Luzzi, A.H. Merry, C.I. Newbold, K. Marsh, G. Pasvol, D.J. Weatherall, Surface antigen expression on Plasmodium falciparum-infected erythrocytes is modified in alpha- and beta-thalassemia, J. Exp. Med. 173 (1991) 785e791. [8] S.J. Allen, S. Bennett, E.M. Riley, P.A. Rowe, P.H. Jakobsen, A. O’Donnell, B.M. Greenwood, Morbidity from malaria and immune responses to defined Plasmodium falciparum antigens in children with sickle cell trait in The Gambia, Trans. R. Soc. Trop. Med. Hyg. 86 (1992) 494e498. [9] S.J. Allen, P. Rowe, C.E. Allsopp, E.M. Riley, P.H. Jakobsen, A.V. Hill, B.M. Greenwood, A prospective study of the influence of alpha thalassaemia on morbidity from malaria and immune responses to defined Plasmodium falciparum antigens in Gambian children, Trans. R. Soc. Trop. Med. Hyg. 87 (1993) 282e285. [10] J.Y. Le Hesran, I. Personne, P. Personne, N. Fievet, B. Dubois, M. Beyeme, C. Boudin, M. Cot, P. Deloron, Longitudinal study of Plasmodium falciparum infection and immune responses in infants with or without the sickle cell trait, Int. J. Epidemiol. 28 (1999) 793e798. [11] M. Cappadoro, G. Giribaldi, E. O’Brien, F. Turrini, F. Mannu, D. Ulliers, G. Simula, L. Luzzatto, P. Arese, Early phagocytosis of glucose-6-phosphate dehydrogenase (G6PD)-deficient erythrocytes parasitized by Plasmodium falciparum may explain malaria protection in G6PD deficiency, Blood 92 (1998) 2527e2534. [12] O. Sodeinde, J.L. Clarke, T.J. Vulliamy, L. Luzzatto, P.J. Mason, Expression of Plasmodium falciparum G6PD-6PGL in laboratory parasites and in patient isolates in G6PD-deficient and normal Nigerian children, Br. J. Haematol. 122 (2003) 662e668. [13] A. Barragan, P.G. Kremsner, M. Wahlgren, J. Carlson, Blood group A antigen is a coreceptor in Plasmodium falciparum rosetting, Infect. Immun. 68 (2000) 2971e2975. [14] V.S. Moorthy, M.F. Good, A.V. Hill, Malaria vaccine developments, Lancet 363 (2004) 150e156. [15] W.E. Collins, R.F. Anders, M. Pappaioanou, G.H. Campbell, G.V. Brown, D.J. Kemp, R.L. Coppel, J.C. Skinner, P.M. Andrysiak, J.M. Favaloro, L.M. Corcoran, J.R. Broderson, G.F. Mitchell, C.C. Campbell, Immunization of Aotus monkeys with recombinant proteins of an erythrocyte surface antigen of Plasmodium falciparum, Nature 323 (1986) 259e262. [16] A. Saul, R. Lord, G.L. Jones, L. Spencer, Protective immunization with invariant peptides of the Plasmodium falciparum antigen MSA2, J. Immunol. 148 (1992) 208e211. [17] H. Perlmann, P. Perlmann, K. Berzins, B. Wahlin, M. Troye-Blomberg, M. Hagstedt, I. Andersson, B. Hogh, E. Petersen, A. Bjorkman, Dissection of the human antibody response to the malaria antigen Pf155/RESA into epitope specific components, Immunol. Rev. 112 (1989) 115e132.

1267

[18] R.R. Taylor, D.B. Smith, V.J. Robinson, J.S. McBride, E.M. Riley, Human antibody response to Plasmodium falciparum merozoite surface protein 2 is serogroup specific and predominantly of the immunoglobulin G3 subclass, Infect. Immun. 63 (1995) 4382e4388. [19] E.M. Riley, S.J. Allen, M. Troye-Blomberg, S. Bennett, H. Perlmann, G. Andersson, L. Smedman, P. Perlmann, B.M. Greenwood, Association between immune recognition of the malaria vaccine candidate antigen Pf155/RESA and resistance to clinical disease: a prospective study in a malaria-endemic region of west Africa, Trans. R. Soc. Trop. Med. Hyg. 85 (1991) 436e443. [20] P. Astagneau, C. Chougnet, J.P. Lepers, M.D. Andriamangatiana-Rason, P. Deloron, Antibodies to the 4-mer repeat of the ring-infected erythrocyte surface antigen (Pf155/RESA) protect against Plasmodium falciparum malaria, Int. J. Epidemiol. 23 (1994) 169e175. [21] F. al-Yaman, B. Genton, R. Anders, J. Taraika, M. Ginny, S. Mellor, M.P. Alpers, Assessment of the role of the humoral response to Plasmodium falciparum MSP2 compared to RESA and SPf66 in protecting Papua New Guinean children from clinical malaria, Parasite Immunol. 17 (1995) 493e501. [22] G. Aribot, C. Rogier, J.L. Sarthou, J.F. Trape, A.T. Balde, P. Druilhe, C. Roussilhon, Pattern of immunoglobulin isotype response to Plasmodium falciparum blood-stage antigens in individuals living in a holoendemic area of Senegal (Dielmo, west Africa), Am. J. Trop. Med. Hyg. 54 (1996) 449e457. [23] H. Groux, J. Gysin, Opsonization as an effector mechanism in human protection against asexual blood stages of Plasmodium falciparum: functional role of IgG subclasses, Res. Immunol. 141 (1990) 529e542. [24] C.M. Rzepczyk, K. Hale, N. Woodroffe, A. Bobogare, P. Csurhes, A. Ishii, A. Ferrante, Humoral immune responses of Solomon Islanders to the merozoite surface antigen 2 of Plasmodium falciparum show pronounced skewing towards antibodies of the immunoglobulin G3 subclass, Infect. Immun. 65 (1997) 1098e1100. [25] R.R. Taylor, S.J. Allen, B.M. Greenwood, E.M. Riley, IgG3 antibodies to Plasmodium falciparum merozoite surface protein 2 (MSP2): increasing prevalence with age and association with clinical immunity to malaria, Am. J. Trop. Med. Hyg. 58 (1998) 406e413. [26] W.G. Metzger, D.M. Okenu, D.R. Cavanagh, J.V. Robinson, K.A. Bojang, H.A. Weiss, J.S. McBride, B.M. Greenwood, D.J. Conway, Serum IgG3 to the Plasmodium falciparum merozoite surface protein 2 is strongly associated with a reduced prospective risk of malaria, Parasite Immunol. 25 (2003) 307e312. [27] B. Dubois, P. Deloron, P. Astagneau, C. Chougnet, J.P. Lepers, Isotypic analysis of Plasmodium falciparum-specific antibodies and their relation to protection in Madagascar, Infect. Immun. 61 (1993) 4498e4500. [28] H.P. Beck, I. Felger, B. Genton, N. Alexander, F. al-Yaman, R.F. Anders, M. Alpers, Humoral and cell-mediated immunity to the Plasmodium falciparum ring-infected erythrocyte surface antigen in an adult population exposed to highly endemic malaria, Infect. Immun. 63 (1995) 596e600. [29] C. Aucan, Y. Traore, F. Tall, B. Nacro, T. Traore-Leroux, F. Fumoux, P. Rihet, High immunoglobulin G2 (IgG2) and low IgG4 levels are associated with human resistance to Plasmodium falciparum malaria, Infect. Immun. 68 (2000) 1252e1258. [30] V. Robert, H. Dieng, L. Lochouarn, S.F. Traore, J.F. Trape, F. Simondon, D. Fontenille, La transmission du paludisme dans la zone de Niakhar, Se´ne´gal, Trop. Med. Int. Health 3 (1998) 667e677. [31] A. Garcia, A.B. Dieng, F. Rouget, F. Migot-Nabias, J.Y. Le Hesran, O. Gaye, Role of environment and behaviour in familial resemblances of Plasmodium falciparum infection in a population of Senegalese children, Microbes Infect. 6 (2004) 68e75. [32] F.N. Rasheed, J.N. Bulmer, A. De Francisco, M.F. Jawla, P.H. Jakobsen, A. Jepson, B.M. Greenwood, Relationships between maternal malaria and malarial immune responses in mothers and neonates, Parasite Immunol. 17 (1995) 1e10. [33] T.N. Williams, T.W. Mwangi, D.J. Roberts, N.D. Alexander, D.J. Weatherall, S. Wambua, M. Kortok, R.W. Snow, K. Marsh, An immune basis for malaria protection by the sickle cell trait, PLoS Med. 2 (2005) e128.

1268


[34] S.I. Lee, D.C. Heiner, D. Wara, Development of serum IgG subclass levels in children, Monogr. Allergy 19 (1986) 108e121. [35] F. Engelbrecht, I. Felger, B. Genton, M. Alpers, H.P. Beck, Plasmodium falciparum: malaria morbidity is associated with specific merozoite surface antigen 2 genotypes, Exp. Parasitol. 81 (1995) 90e96. [36] A.M. Diatta, L. Marrama, A. Tall, J.F. Trape, A. Dieye, O. Garraud, O. Mercereau-Puijalon, R. Perraut, Relationship of binding of immunoglobulin G to Plasmodium falciparum-infected erythrocytes with parasite endemicity and antibody responses to conserved antigen in immune individuals, Clin. Diagn. Lab. Immunol. 11 (2004) 6e11. [37] G. Cabrera, M. Cot, F. Migot-Nabias, P.G. Kremsner, P. Deloron, A.J.F. Luty, Sickle cell trait is independently associated with enhanced IgG antibody responses to Plasmodium falciparum variant surface antigens, J. Infect. Dis. 191 (2005) 1631e1638. [38] R.A. Bayoumi, Y.A. Abu-Zeid, N.H. Abdulhadi, B.O. Saeed, T.G. Theander, L. Hviid, H.W. Ghalib, A.H. Nugud, S. Jepsen, J.B. Jensen, Cell-mediated immune responses to Plasmodium falciparum purified soluble antigens in sickle-cell trait subjects, Immunol. Lett. 25 (1990) 243e249.

[39] A.J. Luty, S. Ulbert, B. Lell, L. Lehman, R. Schmidt-Ott, D. Luckner, B. Greve, P. Matousek, D. Schmid, K. Herbich, B. Dubois, P. Deloron, P.G. Kremsner, Antibody responses to Plasmodium falciparum: evolution according to the severity of a prior clinical episode and association with subsequent reinfection, Am. J. Trop. Med. Hyg. 62 (2000) 566e572. [40] F. Ntoumi, M.T. Ekala, M. Makuwa, F. Lekoulou, O. MercereauPuijalon, P. Deloron, Sickle cell trait carriage: imbalanced distribution of IgG subclass antibodies reactive to Plasmodium falciparum familyspecific MSP2 peptides in serum samples from Gabonese children, Immunol. Lett. 84 (2002) 9e16. [41] F. Ntoumi, O. Mercereau-Puijalon, S. Ossari, A. Luty, J. Reltien, A. Georges, P. Millet, Plasmodium falciparum: sickle-cell trait is associated with higher prevalence of multiple infections in Gabonese children with asymptomatic infections, Exp. Parasitol. 87 (1997) 39e46. [42] R.A. Bayoumi, Does the mechanism of protection from falciparum malaria by red cell genetic disorders involve a switch to a balanced TH1/TH2 cytokine production mode? Med. Hypotheses 48 (1997) 11e17.