Genetic diversity of Plasmodium falciparum isolates from Baka Pygmies and their Bantu neighbours in the north of Gabon

Study area and studied populations

The study was carried out in Woleu Ntem Province, located in the north of Gabon (Fig. 1). Three countries surround this province: Cameroon, Equatorial Guinea and the Republic
of Congo. The project focused on Baka Pygmies and Bantus living in rural and forest
environments.

Fig. 1. Geographical location of the study sites in the departments of Woleu and High Ntem
in Woleu Ntem Province. Ten sites were sampled: (1) green isolated Pygmy villages; (2) red mixed villages (Bantus and Pygmies) and (3) blue Bantu populations in forestry villages

The study was conducted following two stages: (1) a sensitization stage in May 2013
during which persons were informed of the aim of the study, and, (2) blood sampling
from June to July 2013. The sampling was conducted just after the rainy season (occurring
from February to May) when malaria transmission is the highest.

Participants were recruited from ten villages (Fig. 1; Table 1): three in the Woleu region and seven in the region of Ntem Minvoul. Two villages
(Bitouga and Etho) corresponded to what will be hereafter be referred as ‘isolated
Pygmy villages’ (IPV). These villages were located in isolated forest zones, with
no road access and were only composed by Pygmies. Five villages were ‘mixed villages’
(MV). Located along roads, these villages are mostly populated by Bantu peoples and
a minority of Pygmy peoples. In these villages, both Bantus and Pygmies were sampled
when possible (Table 1). Finally, three villages were villages built by forestry companies for their workers
[hereafter named ‘forestry villages’ (FV)]. Only Bantu peoples lived in these villages,
accessible by road. Distances between villages ranged from 1 to 136 km (Fig. 1). In mixed villages and other sites, participants were subjected to a questionnaire
in which they would indicate their ethnic origin and the origin of their parents.
Participants born from inter-ethnic marriages were not included in this study.

Table 1. Number of blood samples collected per village and ethnic group (Pygmies and Bantus)
as well as prevalence (%) of Plasmodium falciparum

In each village, all volunteers were included in the study. Volunteers were recruited
with their informed consent and in the case of children the consent of their parents.
The translation of consent for participants was done by members of the NGO AGAFI (Gabonese
Association of Assistance to Indigenous Women and Indigent). Before blood collection,
a medical doctor examined each volunteer. Age and sex were retrieved for each patient.
Blood was collected in EDTA tubes using venipuncture. Each sample was aliquoted and
stored at ?20 °C until used.

A total of 342 blood samples were obtained: 170 from Baka Pygmies (121 from MV and
49 from IPV) and 172 from Bantus (38 in MV and 134 in FV). This study was authorized
by the Ministry of Health and the agreement of the National Ethics Committee on Research
(authorization number: PROT No 0030/2013/SG/CNE).

Determination of Plasmodium falciparum infections

For each blood sample, total DNA was extracted from 200 µl of blood using the DNeasy
blood and tissue kit Qiagen (QIAamp
^®
DNA Blood Mini Kit) according to manufacturer’s instructions.

For each sample, Plasmodium infections were detected using the method based on the amplification of a portion
of the Plasmodium cytochrome b (cyt–b) gene by a nested PCR as described in Additional file 123]. The PCR amplified products were sequenced by Eurofins (MWG, France). Then, a logistic
regression (binomial family) was used to test for variations of P. falciparum prevalence between Pygmies and Bantus (fixed effect) or between village types (fixed
effect). In both cases, the factor ‘village of collection’ was considered as a random
effect. Tests were realized using the R software 24] and the library lme 4 25], 26].

Molecular genotyping of Plasmodium falciparum isolates

Seven microsatellite markers as well as four candidate genes coding for antigenic
proteins were genotyped.

The microsatellite markers amplified by semi-nested PCR are shown in Additional file
2. The PCR primer sets and amplification conditions were those described by 9], 27]. Fluorescence-labelled PCR products were sized on ABI Prism 310 genetic analyzer
(Applied Biosystems), with a Genescan 500 LIZ internal size standard.

Regarding the candidate genes, merozoite surface protein 1 (msp1—block 2), merozoite surface protein 2 (msp2—block 3), glutamate-rich protein (glurp-region 2), and erythrocyte binding protein-175 (eba–175—region III) genes were genotyped as described in 21], 22], 28]. Msp1 is a single copy gene located on chromosome 9. It encodes a merozoite surface antigen
of approximately 190 kDa. The C-terminal part (MSP1
₁₉
) is immunogenic and contains Epidermal Growth Factor (EGF)-like domains. MSP1
₁₉
is very rich in cysteine and is involved in the invasion of red blood cells by merozoites
29], 30]. MSP1 is composed of 17 blocks. Block 2, which is the first variable block, located
at the N-terminal region is the most polymorphic. Block 2 of msp1 was genotyped in this study. MSP-2 is located on chromosome 2 and encodes a 46–53 kDa
antigenic protein located at the surface of the merozoite 31]. It consists of three regions: two semi-conserved (at the N-and C-terminal regions)
and one polymorphic region composed of tandemly repetitive sequences 32]–34]. The central polymorphic region (block 3) of msp2 was genotyped in this study. The EBA-175 antigen is a molecule expressed at the surface
of the parasite, which binds to red blood cells. It is involved in the invasion of
the red blood cells 35]–37]. It is a dimorphic molecule composed of two alleles: CAMP (C) and FCR3 (F). Dimorphism
is related to two segments of region III. This is this region that has been genotyped
in this study. Finally, GLURP is a 220 kDa protein expressed in both pre-erythrocytic
and erythrocytic stages of P. falciparum, as well as on the surface of newly released merozoites in human host 38]. It consists of a N-terminal region of limited diversity named R0 followed by two
polymorphic repeat regions named R1 and R2 39]. The region 2 (R2) of GLURP has been genotyped in this study.

The specific primers used for genotyping each gene are given in Additional file 1. For msp1 gene, three allelic families (K1, MAD20 and RO33) were determined. The two allelic
families 3D7 and FC27 of the msp2 gene were considered and the F (FCR3) and C (CAMP) alleles for the gene eba–175. For glurp, alleles were defined based on the amplicon size. PCR products were all visualized
on a 1.5 % agarose gel electrophoresis.

Plasmodium falciparum population genetic analyses

Allele frequency was computed as the proportion of the total of all alleles detected
among the isolates examined (see Additional file 3).

All population genetic analyses were performed using FSTAT version 2.9.4 40]. Genetic diversity within populations was assessed by the unbiased expected heterozygosity
(He) 41]. Comparison of genetic diversity was performed either between Pygmy and Bantu populations
or between IPVs against the others village types using a permutation test (1000 permutations).

Population genetic differentiation was estimated using the estimator (?) of the Wright’s
F-statistics F_ST42], 43]. This index measures the genetic differentiation (i.e., the differences of allelic
frequencies) between populations of interest. Deviation of ? from 0 was tested using
a permutation test (1000 permutations). Mantel tests were used to test for patterns
of isolation by distance (IBD). A partial Mantel test was performed to test for an
effect of the ethnic group on P. falciparum genetic differentiation while controlling for geographic distance (1000 permutations).
For this test, the first matrix was the matrix of F_ST
computed between all pairs of populations. The second matrix was the matrix of pairwise
geographic distances computed between populations and the third one was a matrix containing
1 when the populations belonged to the same ethnicity (Pygmy/Pygmy and Bantu/Bantu)
and 2 when they belonged to different ethnic groups (Pygmy/Bantu).

An additional partial Mantel test was performed with two new groups of populations:
one including only the IPVs that are strictly inhabited by Pygmies and the other including
the other villages that are mostly inhabited by Bantus (MV + FV). MV isolates were
indeed considered more likely to be representative of Bantus because, inside MVs,
pygmies represent only a minority of individuals (even if they were sometimes sampled
in majority) and also because no genetic differentiation was observed between P. falciparum populations collected from Pygmies and Bantus within these villages (see “Results”).
For this test, the first matrix was the matrix of F _ST
between populations, the second matrix corresponded to the geographic distances between
populations and the third one contained 1 when the populations belonged to the same
village type (IPV or MV + FV) and 2 when they belonged to two village types (IPV/MV + FV).