Plasmodium vivax has been the most common cause of the human malaria parasite in the Brazilian Amazon region over the last seven years 1. Innate resistance to malaria infections in humans is conferred by different blood group polymorphisms. The association of the Duffy blood group (FY) with P. vivax human malaria has been well-documented, where Duffy-negative individuals are naturally resistant to invasion by this parasite 2.

The Duffy blood group antigen has been identified as a scavenger on the red blood cell (RBC) surface eliminating excesses of circulating toxic chemokines, named Duffy antigen/receptor for chemokines (DARC) 3. The determinant, a glycoprotein that passes through the membrane seven times, includes a 35–43 KDa epitope in the extracellular, N-terminal domain that mediates erythrocyte invasion by P. vivax merozoites 4. The Fy gene has two exons (Fy^a and Fy^b) that are encoded by the co-dominant alleles FYA and FYB, located on chromosome 1 56. The corresponding anti-Fy^a and anti-Fy^b antibodies define four different phenotypes; Fy(a+b+), Fy(a+b-), Fy(a-b+) and Fy(a-b-) 7. The FYA and FYB alleles differ by a point mutation in the major cDNA transcript 8, encoding glycine in Fy^a or aspartic acid in Fy^b at residue 42 of the most important form of the protein, encoded by the exon 2 98. The molecular mechanism that gives rise to the null Duffy phenotype [Fy(a-b-)] has been classically associated with a point mutation in the GATA box of the DARC promoter which silences the gene encoding the Duffy system antigens in the RBCs of these individuals resulting in a FYB allele 9. Recently, the FYA allele has been identified 10 and a new FYB allele was found with three single nucleotide polymorphisms (11). It has been demonstrated in blood donors from Southeast Brazil that Caucasians and African descendents were serologically Fy(b-) with the majority of the African descendents being FYB with the GATA single nucleotide polymorphism (SNP), while the majority of Caucasian typing Fy(b-) had FYB with 265T/298A SNPs 11. The Fy-negative is common on RBCs of Negro individuals from ethnic African groups, but it is rare in many other populations 12.

Previous studies have demonstrated that heterozygous Duffy negative malaria individuals remain susceptible to infections by P. vivax 1314. It was demonstrated in individuals living in a malaria endemic region of Papua New Guinea that Duffy binding protein adherence to RBCs is significantly reduced for erythrocytes of P. vivax malaria patients who carry one Duffy-negative allele 15. In addition, it was observed that in different areas from the Brazilian Amazon region some of the Duffy phenotypes showed significant correlation between blood donors and malaria patients 16. On the other hand, a previous study in the Western Brazilian Amazon region (Rondônia State) did not correlate Fy heterozygosity with protection against P. vivax 17. The present study reports the Duffy blood group genotyping in P. vivax malaria patients from four different Brazilian endemic areas, exploring these significant phenotypic associations between blood group variants, as well as polymorphic regions of RBC receptors, with susceptibility or resistance to malaria.

The genotypic and allelic frequencies of the Duffy blood group in the 330 blood donors and 312 patients infected by P. vivax as determined by PCR/RFLP are summarized in Table 1. The data show a high frequency of the FYA/FYB genotype in 199 individuals (31.0%) of the studied population. This is followed by 117 (18.2%) and 94 (14.6%) of homozygotes for the FYB and FYA alleles, respectively. While the frequencies for heterozygote individuals with FYA/FYB-33 and FYB/FYB-33 were comparable at 14.9% (96). Low frequencies were detected for the FYA/FY^X, FYB/FY^X, FYX/FY^X and FYB-33/FYB-33 genotypes. The negative Duffy genotype (FYB-33/FYB-33) was found in both individuals infected by P. vivax and unaffected blood donors.

Thus far, innate resistance to malaria infections in humans has been attributed to blood group polymorphisms. Duffy blood group polymorphisms are important in areas where P. vivax predominates, because this molecule acts as a receptor for this protozoan (but not for the other human malaria parasites) on the surface of RBCs 19. Field observations from West Africa and Ethiopia have indeed established a strong correlation between absence or low endemicity of P. vivax malaria and the high prevalence of the Duffy negative allele 2021. Little is known on the frequency of RBC polymorphisms that confer either partial or complete resistance against malaria. The data obtained in the present study emphasize the importance of the evaluation of Duffy blood group genotypes in malaria endemic areas of the Brazilian Amazon region.

The Brazilian population has a highly heterogeneous ethnic composition, a result of the hybridization of the numerous native indigenous populations and immigrants from Europe, Africa and Asia. The immigration flow was not uniform in the different regions of the country 222324. The differential distribution of Duffy antigenic determinants among ethnic groups is an aspect characteristic of this blood system. Hence, this has been used as a marker for the ethnic composition as well as an indicator in the populational evolution. In the current study, significant differences in the allelic frequencies of the FY gene were identified when compared with previous studies of patients infected with human plasmodium in Brazilian Amazon regions 17252627. However, the FYA, FYB and FYB-33 alleles showed differential distributions compared to the population of the southeastern region of Brazil 11. The genotypic compositions obtained in this study demonstrated significant variations compared to previous studies performed in the states of Pará and Rio Grande do Sul 2829 and in the cities of São Paulo 30, Campinas 11 and Ribeirão Preto 31 all of which are in the state of São Paulo, Southeastern region of Brazil.

The currently obtained results showed that the FYA/FYB genotype was the most common, followed by the homozygotes for the FYB and FYA alleles and by the heterozygotes FYA/FYB-33 and FYB/FYB-33 (Table 1). The FYB-33/FYB-33 genotype, which is classically seen in individuals unaffected by P. vivax infection, was identified in 3.2% of the general population and in 5.5% of blood donors. These numbers differ from previous reports on a group of malaria patients uninfected by P. vivax in the state of Rondônia 17, where the frequency of this genotype was 12% and also for a mixed sample from the Northern and Southeastern regions of Brazil 27. Some populations had higher genotypic frequencies for FYA/FYB and FYB/FYB-33 than for a caucasian-like and negroid population from the Southeastern of Brazil 11. On the other hand, the frequency of the FYB-33/FYB-33 genotype was lower. In fact, with the exception of negroid ethnic groups, this genotype is extremely rare 12303233.

The FYA allele is common in European and Oriental peoples but it is rare in African Negroes. Additionally, the FYB is more common in populations classified as white than in asiatic and negroid populations 3435. The frequency of the FYA and FYB alleles in the Brazilian Amazon population was higher than those in the Southeastern region 11 probably due to the massive influence of Portuguese colonization in the North region of Brazil as well as the presence of Amerindians 36, as recent molecular analysis has corroborated that Amerindians have an Asian origin 37. Indeed, in the North region of Brazil, studies carried out with different tribes of Amerindians have shown that the FYA allele is the most common 3138. As expected, a lower frequency of the FYB-33 allele was observed in the North region (P = 0.0001). Although a high prevalence of this allele was demonstrated in isolated communities in the states of Amapá and Pará, North region of Brazil 39, the introduction of a negroid ethnic component in the Amazon region is recent 4041. In respect to the FY^X allele, there does not seem to be a differential ethnic distribution as, the frequency detected here is similar to those described in other Caucasian-like populations in Brazil 11 and also in Europe 42.

The different FY allelic frequencies in individuals from North compared to Southeastern Brazil, may be due to the contribution of the three major ethnic groups (Europeans, in particular Portuguese; Blacks and Amerindians) in the formation of both populations. The Amerindian contribution was higher in the north of the country whereas European migration took place more in the South and Southeastern regions 434445. In spite of the current knowledge of the relationship between structure/function and tissue location of DARC, the functional significance of each of the alleles and the different genotypic combinations require further elucidation. The variation in the ethnic composition of the urban and rural populations of the Brazilian Amazon region and of distinct regions in Brazil 2646, may influence the allelic and genotypic distributions reflecting in matrixes of genetic mechanisms favorable to the susceptibility to infectious and parasitic diseases, in particular to malaria.

The frequency of the allele with the FY GATA mutation (FYB-33) was greater in blood donors than in patients infected with malaria (P = 0.0033), suggesting that there is a reduction in the infection rate of carriers of the FYB-33 allele. These results were recently reinforced by data from of individuals infected by P. falciparum when compared to those infected by P. vivax in Brazilian Amazon regions 27. The FYB-33 allele is a variant of the FYB allele resulting from a T → C point mutation in the gene promoter region (neucleotide-33), which abolishes its expression 9. The same occurs in the FYA allele, determining the FYA^null allele 10. The presence of the FYB-33 allele results in a 50% reduction in the Duffy protein expression on the erythrocyte surface 1547. This process demonstrates the action of the dose-related effect of the gene 1548, which may limit the invasion process of red blood cells by the parasite, although susceptibility to P. vivax may occur in heterozygotic Duffy-negative individuals 1314. Thus the FY/FYB-33 genotypic combination, with either Fy(a+b-) or Fy(b+b-), seems to convey a reduction in the susceptibility to malaria. In vitro studies 15 support this hypothesis as RBCs that express these phenotypes have a significant reduction in cytoadherence of the parasite when compared to RBCs that express Fy(b+b+). Recently, in Papua New Guinea, a significant reduction in infection by P. vivax was observed in Duffy negative individuals heterozygotic for the FYA^null allele 49.

In this study a low frequency of the negative Duffy genotype (FYB-33/FYB-33) was detected among uninfected subjects. As has previously been demonstrated, the absence of the Fy antigen in many ethnic Negro groups and their descendents, does not seem to exert any deleterious effect, however it does bestow a natural resistance against infection by P. vivax 2914355051. These individuals are homozygotes for the FYB mutation in the GATA, that completely abolishes the Fy expression in erythrocytes but not in cells of other tissues 9. Recent reports have provided evidence on the transmission among Duffy-negative patients in Africa 52 and in Brazil 53. These data suggest the possibility that P. vivax is utilizing alternative receptors, apart from Fy, for binding in the erythrocyte invasion process. Whether FYB carrying the GATA box mutation is the primordial gene that encodes the Duffy system antigens or whether the GATA box mutation has evolved to escape malaria infection per se is controversial. However, the presence of this allele (FYB-33) may be a demonstration of the selection advantage by mutation in the population. Thus, in the Brazilian Amazon, where P. vivax predominates, the frequency of the FYB-33 allele is higher than expected given the ethnic population, both in terms of heterozygotes and homozygotes. This might, over time, cause an increase in the number of Duffy-negative individuals in the population and, as a consequence, a reduction in the rate of infection by P. vivax in this region.

In the group of patients infected by P. vivax, different allelic combinations with FY^X were detected, including five (1.6%)FYA/FY^X individuals, four (1.3%) FYB/FY^X and two (0.6%) FY^X/FY^X (Table 2). The frequency of this allele among patients and blood donors did not show statistical differences (P = 0.3873). Studies carried out in blood samples from European and American Caucasians and Afro-American individuals, demonstrated that the C265T (FY^X) mutation altered the Duffy protein expression on the RBC surface differently, depending on the ethnic group, using at least two mechanisms. One mechanism involves silent transcription in one of the FYB alleles and the other affects the translation and/or the stability of the protein 4754. These studies showed that heterozygote individuals for the FY^X (FY^WK) allele have 50% less of the Duffy protein on the surface of the membrane of the erythrocyte. On the other hand, individuals homozygotic for the FY^X allele have about one tenth of the antigen expression in the erythrocytes 47. However, the information shown by the present study suggest that these polymorphisms do not seem to be associated with susceptibility to malaria.

As illustrated in Table 2, a higher frequency of the FYA/FYB genotype in patients compared to blood donors (P = 0.0404) was detected. In principle, this seems to indicate that the condition of heterozygosis, resulting from the expression of the FYA and FYB genes, favors infection by P. vivax. In fact, as has already been demonstrated by us, based on a phenotyping study, Fy(a+b+) individuals may be more susceptible to infection by this species of Plasmodium 16. Although no qualitative or quantitative measurements of the Duffy glycoprotein expressions were made in this study, we saw a larger number of malaria episodes among patients with the heterozygote genotype than the homozygote genotype. The basis of this observation has not been determined yet, but the phenotypic 16 and genotypic data presented from the same Brazilian Amazon areas associated with pertinent publications, point to the possibility that heterozygote individuals modulate the susceptibility to malaria by P. vivax by means of quantitative and/or qualitative variations that affect the Duffy antigen expression on erythrocytes. In the first aspect, in vitro studies demonstrate differences in the levels of the expression of FY glycoprotein on the surface of the reticulocytes of Caucasian-like and Afro-American individuals with the Fy(a+b+) erythrocytic phenotype 48. These authors verified that individuals with FYA and FYB alleles expressed a lower quantity of DARC than heterozygotes. Hence, it is possible that heterozygote individuals have a greater repertory of receptors for the possible variations that occur in the parasite protein binder to the human erythrocytes.

The recent demonstration of polymorphisms in the Duffy binding protein in isolates of P. vivax from the Brazilian Amazon 55 seems to support the observed results. An association, between human receptor polymorphisms and variations in the parasite binders of Plasmodium falciparum that modulate susceptibility to malaria, was also demonstrated 195657585960. Hence, apart from the different levels of the expression, the specific conformation of the Fya and Fyb antigens may determine differences in the susceptibility to infection. Nevertheless, one of the possible consequences of differential susceptibility to vivax malaria could be modifications in allelic frequencies of FYA and FYB in populations exposed to P. vivax, the most prevalent species in the Brazilian Amazon region. If in fact this process occurs as a consequence of infection, it seems to be necessary, but not sufficient to eliminate heterozygote individuals, even if they are affected by repetitive episodes of malaria by P. vivax, as the infection is rarely severe and occurs in the entire age range with different frequencies for different ages and regions of the Brazilian Amazon without lethality 6162.

Based on the significant associations described herein, our data differ from previous studies 17 carried out in the western Brazilian Amazon region as well as recent studies of patients from the Brazilian Amazon as a whole 27. In respect to the first studies, these differences may be a result of the small sample sizes initially evaluated; the control group of the current study was composed of blood donors who reported never having experienced malaria against samples from individuals who had P. falciparum previously evaluated and the various ethnic groups in the different regions studied today (Eastern and Western Amazon). In relation to the recent studies, different to the findings observed in this study Brazilian Amazon population, the authors detected a significant association among infections by P. vivax and the FYB/FYB genotype, also conflicting with results of Wooley et al 48 in North American individuals (Caucasians and Afro-Americans). These divergences probably occurred due to the smaller sample size involved, and the fact that the control group consisted patients infected by P. falciparum.

The Brazilian population presents an elevated degree of miscegenation which is implicated in variations in the allelic distribution of FY, which was also found in this study. Future analyses that consider haplotypic frequencies of the alleles and quantification of the DARC expression in populational groups in endemic and non-endemic areas, increasing the knowledge on the dynamics of this gene and its possible contribution as a modulator in the susceptibility to malaria. The data obtained in the present study supports the hypothesis that individuals with the FYA/FYB genotype have higher susceptibility to malaria. The presence of the FYB-33allele may be a selective advantage in the population, reducing the rate of infection by P. vivax in this region. However, the populational and longitudinal studies must be amplified accompanying groups of individuals with the FYB-33 allele and the FYA/FYB genotype, including clinical parameters, as well as the evaluation of their expressions which may contribute to better elucidate the physiopathologic differences in this parasite/host relationship in regions endemic for P. vivax malaria, in particular the Brazilian Amazon.

CEC carried out all genotype assay. AADC, VSDC and CEC collected samples from P. vivax malaria individuals and blood donors. CEC, LCM, CRBD and LC participated in the design of the study and drafted the manuscript. RLM and ARBR conceived of the study and participated in all aspects of its design, acquisition of funding, execution, coordination and manuscript preparation. All authors read and approved the final manuscript.