Molecular Evidence of Plasmodium vivax Mono and Mixed Malaria Parasite Infections in Duffy-Negative Native Cameroonians

The malaria parasite Plasmodium vivax is known to be majorly endemic to Asian and Latin American countries with no or very few reports of Africans infected with this parasite. Since the human Duffy antigens act as receptors for P. vivax to invade human RBCs and Africans are generally Duffy-negative, non-endemicity of P. vivax in Africa has been attributed to this fact. However, recent reports describing P. vivax infections in Duffy-negative Africans from West and Central parts of Africa have been surfaced including a recent report on P. vivax infection in native Cameroonians. In order to know if Cameroonians living in the southern regions are also susceptible to P. vivax infection, we collected finger-prick blood samples from 485 malarial symptomatic patients in five locations and followed PCR diagnostic assays with DNA sequencing of the 18S ribosomal RNA gene. Out of the 201 malaria positive cases detected, 193 were pure P. falciparum, six pure P. vivax and two mixed parasite infections (P. falciparum + P. vivax). The eight P. vivax infected samples (six single + two mixed) were further subjected to DNA sequencing of the P. vivax multidrug resistance 1 (pvmdr1) and the P.vivax circumsporozoite (pvcsp) genes. Alignment of the eight Cameroonian pvmdr1 sequences with the reference sequence showed high sequence similarities, reconfirming P. vivax infection in all the eight patients. DNA sequencing of the pvcsp gene indicated all the eight P. vivax to be of VK247 type. Interestingly, DNA sequencing of a part of the human Duffy gene covering the promoter region in the eight P. vivax-infected Cameroonians to identify the T-33C mutation revealed all these patients as Duffy-negative. The results provide evidence of single P. vivax as well as mixed malaria parasite infection in native Cameroonians and add knowledge to the growing evidences of P. vivax infection in Duffy-negative Africans.


Introduction
Malaria is a highly infectious vector-borne disease of the tropical and sub-tropical countries of the globe. Almost all the African countries are endemic to malaria, contributing about 90% of the total global malaria death incidences [1]. Apart from global effort to control malaria, this disease remains as a frontline infectious disease in Africa and in other malaria endemic regions of the globe. Relief and management of malaria majorly rely on treatment of infected patients by chemotherapy. Since as many as five different species of the genus Plasmodium (Plasmodium falciparum, P. vivax, P. ovale, P. malariae and P. knowlesi) are known to infect humans [2,3] either singly or as mixed parasitic infections, proper diagnosis of specific malaria parasitic infection holds the key for effective treatment and management of malaria. Although microscopy is traditionally considered as gold standard for malaria diagnosis, in recent years, molecular diagnostic approaches by PCR (Polymerase Chain Reaction) assays have evolved as the most sensitive method to accurately diagnose single as well as mixed malaria parasite infections [4,5,6]. Several field epidemiological studies in different malaria endemic countries of the globe (including some African countries) [6,7,8,9,10,11] have adapted this highly sensitive malaria diagnostic approach.
In recent years, malaria diagnosis by PCR assays have not only confirmed high incidences of malaria due to P. falciparum infections in Africa, but instances of P. vivax infections in Duffynegative individuals have surfaced in many African countries, e.g., Madagascar [9], Mauritania [12], Angola [11], Equatorial Guinea [11], Ethiopia [13]. Historically, P. vivax is known to be majorly prevalent in Asia and Latin America but largely absent in west and central African countries. This situation is explained by the fact that (i) the human Duffy Antigen Receptor for Chemokines (DARC) is used by P. vivax merozoites to invade the human RBCs [14], (ii) a mutation (T-33C) located in the promoter region of the human Duffy gene (otherwise known as Duffy-negative individuals) blocks the red blood cell invasion of P. vivax, and (iii) the T-33C mutation (causing Duffy-negativity) is nearly fixed in Africans [15,16]. Since the Duffy-negative Africans are naturally protected from P. vivax infection, it has been proposed that the T-33C mutation has been selected in humans [17]. Considering the recent hypothesis on the African origin of P. vivax [18] it is quite possible that the fixation of the T-33C mutation might have been propelled by long exposure to P. vivax infection in Africans.
To this respect, Cameroon, a West-Central African country is inhabited mostly by Duffy-negative humans (95-99%) [19], with high prevalence of P. falciparum malaria (up to 100%) [20,21]. Till 2013, incidences of P. vivax infection reported in Cameroon were restricted to non-native Cameroonians [22]. Very recently, it has been found that native Cameroonians (Duffy positive as well as Duffy negative individuals) were also able to be infected by the malaria parasite P. vivax [23]. Considering Cameroon as ''Africa in miniature'' with respect to malaria epidemiology [21], the recent report on native Cameroonians being infected with P. vivax [23] and the fact that Cameroon borders with malaria endemic countries (such as Equatorial Guinea) reporting P. vivax infection in Duffy-negative individuals, it is therefore important not only to survey the extent of malaria due to infection of different species of malaria parasites in native Cameroonians, but also to explore other regions of the country for the possible detection of P. vivax malaria infection. Since it has been hypothesized that the Duffy-positive Africans serve as reservoirs for P. vivax transmission [9,11] and Duffy-positive native Cameroonians found to be infected by P. vivax [23], in case we find P. vivax infections in native Cameroonians, it will be of further interest to know the Duffy-status of the P. vivax infected patients.

Ethics Statement
The present study has been approved by the Ethical Committee of Cameroon (No003/CNE/SE/2012) and written informed consents were obtained from all adult patients and the guardians of the minor patients.

Sample collection and DNA isolation
We have collected finger-prick blood samples as 4-5 spots (each spot contains about 100 microliters of blood) on Whatman filter papers from 485 malaria symptomatic patients attending hospitals located in five different areas in southern Cameroon ( Figure 1). The spots were dried and brought to the laboratory in New Delhi, India. In the laboratory, routine protocols were followed to isolate malaria parasite DNA from the blood spots using QIAamp mini DNA kit (Qiagen, Germany). Only one blood spot for each sample was used to isolate DNA at a time. The genomic DNA was eluted in a total volume of 80 ml of AE buffer (Qiagen, Germany).

PCR-amplification of different genes
Since the conclusions drawn in the present study greatly relies on the molecular assays of different genetic fragments, PCR reactions were performed in four different genes (three genes of Plasmodium species and one of human). Details of the PCR conditions followed for specific genes are provided in the Tables S1 and S2. Whereas we used the published primers for amplification of the 18S rRNA gene of both P. falciparum and P. vivax, pvmdr1, pfmdr1 and human Duffy genes, the primer sequences of the pvcsp gene were freshly designed in the present study. Although the general PCR conditions for all the genes were almost similar, different annealing temperatures were required for proper amplification of each of the genes (Tables S1 and S2).
For diagnosis of malaria parasites through PCR assay, we followed the nested PCR technique targeting the 18S rRNA gene of malaria parasites [4,5]. This sensitive technique has been used in a number of studies [6,11] and is suitable for differential identification of the four principal Plasmodium species associated with human malaria; i.e., P. falciparum, P. vivax, P. malariae and P. ovale as single or mixed infections. Following the published literature [4,5], two sets of nested PCR reactions were performed for each sample; one for identification of P. falciparum, P. vivax, P. malariae and the other for P. ovale (Table S1). The DNA isolation (from a separate blood spot) and PCR amplification processes were repeated two more times for reconfirmation in samples presenting infections of P. vivax following similar protocol as described above.
In order to verify the PCR detection of P. vivax infection in Cameroonians based on the 18S rRNA gene, we have additionally PCR amplified two other genes of P. vivax; pvmdr1 and pvcsp. Whereas the pvmdr1 gene is associated with resistance to multiple antimalarial drugs (pvmdr1), the pvcsp is a highly polymorphic gene, commonly used in the characterization of P. vivax populations for different strain variants (VK210, VK247 and P. vivax-like) circulating around the world [24,25]. The published primers [26] of pvmdr1 gene amplify a 543 bp DNA fragment and the primers for pvcsp gene amplify a 507 bp product. The details of the PCR conditions of both the genes are presented in the Table  S2. Similarly, in order to verify the presence of P. falciparum in the mixed infections (P. vivax + P. falciparum) as detected with PCR diagnostic assay with the 18S rRNA gene ( [4,5] see above), we PCR amplified a 510 bp DNA fragment of the P. falciparum multi-drug resistance gene 1 (pfmdr1) by nested PCR using the published primers [27,28] (Table S2).
Since P. vivax is reported to infect Duffy-negative Africans [7,11,12,13], with an interest to know if the P. vivax-infected Cameroonians are also Duffy-negative, we used the isolated DNA from samples presenting P. vivax infection (which also contain human genomic DNA) and used the published primers [29] to PCR amplify a fragment of ,516 bp region covering the -33 rd nucleotide position (located in the promoter region) of the human Duffy gene [29]. We have followed this approach because, it is well known that DNA sequencing of the promoter region of the human Duffy gene with respect to the -33 rd nucleotide mutation can adeptly determine the Duffy-status of humans [11,17,29,30]. This approach relies on the fact that finding the C nucleotide in a single peak in the DNA chromatogram at the -33 rd position specifies complete Duffy-negative status (also known as FY*O homozygote genotype), meaning no expression of the Duffy protein on the erythrocytes [31]. However, detection of both the nucleotides C and T (in double peak) at the same position depicts the heterozygote genotype and the T nucleotide in a single peak indicates the absence of FY*O genotype [29]. The details of the PCR protocols are provided in the Table S2. Alike the 18S rRNA gene, the PCR amplification processes for pvmdr1, pfmdr1, pvcsp and the human Duffy genes were repeated twice for the accurateness of identification/characterization of P. vivax species and determination of the Duffy status.

Purification of PCR products, DNA sequencing and multiple DNA sequence alignments
In order to sequence and determine sequence identity/ variations in the respective genes of the malaria parasites and Cameroonians with the respective reference sequences, we first purified the respective PCR products with Exonuclease-I and Shrimp Alkaline Phosphatase (Fermentas, Life Sciences) using the PCR thermal cycler (37uC for 1 hour and then 85uC for 15 minutes). The purified PCR products were then processed for DNA sequencing reactions with Big Dye Terminator as per standard protocol of the Applied Biosystems (ABI). The products were then run in the ABI 3730 XL DNA analyzer (in-house facilities of NIMR, New Delhi). For each DNA fragment, sequencing was performed from both the 39 and 59 directions (2X coverage). For each individual (human and parasites alike), separate contigs were formed with the two sequences from both the directions (39 and 59) using the SeqMan module of the DNASTAR (Madison, USA) computer program. Multiple sequences of each homologous gene of the malaria parasites (18S rRNA, pvmdr1, pvcsp and pfmdr1) and of the Cameroonians (Duffy) were aligned using the MegAlign module of the DNASTAR (Madison, USA) computer program to ascertain similarities and differences with the respective reference sequences. For example, the respective 18S rRNA gene sequences of P. vivax and P. falciparum were aligned with the reference sequences of P. vivax 18S rRNA gene of the SAL-1 strain (accession number U03079.1) and with the reference sequence of P. falciparum 18S rRNA gene of the 3D7 isolate (accession number AL844501), independently. Similarly, the pvmdr1 and pfmdr1 genes were aligned with the respective reference strains of P. vivax (SAL-1 strain, accession number XM_001613678) and P. falciparum (3D7 strain, accession number AL844504). Further, the human Duffy gene and pvcsp sequences generated in the present study were aligned with the respective reference sequences [accession numbers NG_011626.1 (Duffy) and GU339059.1 (pvcsp SAL-1 strain)]. All the references sequences (18S rRNA genes of P. vivax and P. falciparum, pvmdr1 gene, pfmdr1 gene, pvcsp and Duffy gene) were retrieved from the NCBI website (http://www.ncbi. nlm.nih.gov/) using the BLAST search.

Results
We have successfully isolated genomic DNA of malaria parasites for all the 485 blood samples from malaria symptomatic patients. Out of these 485 samples, only 201 were found to be infected with malaria parasites following PCR diagnostic assays of the 18S rRNA gene ( Table 1). As expected, majorities of the infections (193, 96%) were due to P. falciparum. Interestingly, six patients (3%) were found to be infected by P. vivax only (Figure 2), and two cases (1%) of mixed parasitic infections (P. falciparum and P. vivax) were also detected ( Figure 2). However, no single or mixed infections due to P. malariae or P. ovale could be identified in the present study. All the eight cases of P. vivax infections (six single and two mixed) were found in four different places in the southern Cameroon (Table 1). In order to confirm the results from the PCR analyses, different other genes of P. vivax and P. falciparum (pvmdr1, pfmdr1 and pvcsp) were also PCR amplified and sequenced (see below).
In order to substantiate the PCR assays results (as above) on the six incidences of single P. vivax and two cases of mixed infections (due to co-infection of P. vivax and P. falciparum), we have performed DNA sequencing of all the eight samples for different genes specific to P. vivax and P. falciparum. The sequencing of the eight 18S rRNA genes in P. vivax (six single and two mixed) followed by multiple sequences alignment of the eight 18S rRNA genes of P. vivax infected samples (six single and two mixed) and the reference sequence of the 18S rRNA gene (total nine sequences, Figure 3) and similar alignment of the two sequences of P. falciparum (mixed infection with P. vivax) and the P. falciparum reference sequence of the 18S rRNA (total three sequences, Figure 4) indicate perfect homology (98-100% similarities) between the DNA sequences generated in the present study with the respective reference sequences (Figures 3 and 4) for both P. vivax and P. falciparum. The newly generated sequences of the 18S rRNA genes of P. vivax have been deposited in the EMBL-Bank (accession numbers HF945436 to HF945443), and of P. falciparum in the GenBank (accession numbers KC428741 to KC428742).
With a view to corroborate the results on the detection of single P. vivax and the mixed parasitic infection as detected by PCR and DNA sequencing of the 18S rRNA genes, we have additionally sequenced the pvmdr1 gene in the same eight samples (six single P. vivax infection and two mixed infection) and the pfmdr1 gene in the two mixed infected samples. Apart from the presence of four nucleotide substitutions (three non-synonymous and one synony- mous) ( Figure S1), multiple sequence alignments for the pvmdr1 genes with the respective reference sequences revealed perfect sequence homology between the reference pvmdr1 gene of the SAL-1 strain and the eight pvmdr1 sequences of Cameroonian P. vivax. For the pfmdr1 gene too, apart from a single nonsynonymous substitution (at the 86 th amino acid position) in one of the two Cameroonian P. falciparum isolates from the mixed parasitic infections with P. vivax, perfect sequence homology was observed when compared with the reference 3D7 P. falciparum strain. The results are presented in the Figures S1 and S2. The eight newly generated sequences of the pvmdr1 and the two sequences of the pfmdr1 gene have been deposited in the GenBank    with accession numbers from KJ534638-KJ534645 and KJ534646-KJ534647, respectively. With some evidences of P. vivax infection in Cameroonians using PCR diagnostic approaches followed by DNA sequencing and sequence alignments of the 18S rRNA and the pvmdr1 genes, we were interested not only to reconfirm P. vivax infection, but also to know the type of P. vivax strain present in these eight Cameroonians. For this, we have sequenced a portion of the pvcsp gene covering the majority of the repeat regions in all the eight P. vivax isolates. Multiple sequence alignments with the reference sequence from the SAL-1 strain retrieved from the GenBank (accession number GU339059.1) showed that all the sequences from Cameroon display the non-apeptide repeats ANGA(G/ D)(N/D)QPG, characteristic of the VK247 variant ( Figure S3). These results therefore not only revalidated the observation of P. vivax infection in eight native Cameroonians, but also indicated that all the eight P. vivax are of VK247 type. The newly generated sequences of the pvcsp gene have been deposited in the GenBank (accession numbers KM099676 to KM099683).
Since many African countries in the recent years report P. vivax malaria infection in Duffy-negative Africans, we tested this hypothesis by sequencing a region of the human Duffy gene covering the promoter region in the eight patients infected with P. vivax. Interestingly, we found the -33C mutation in all the eight patients, signifying the fact that all the eight native Cameroonians are Duffy-negative. In order to further discern if these eight Cameroonians are homozygous to the C mutation at the -33 rd position of the promoter region, we have carefully visualized and inspected the occurrence of single/double peaks at the T-33C nucleotide position in each of the two sequence chromatograms of each single Cameroonian (2X coverage, see above). Interestingly, for all the eight Cameroonians, we found single peak of C nucleotide at the -33 rd position of the promoter region of the human Duffy gene in both the sequences from the forward and reverse directions (Figure 5), signifying the fact that all the eight Cameroonians infected with either single P. vivax or mixed infection with P. falciparum are homozygous Duffy-negative. The eight newly generated sequences of the promoter region of the Duffy gene have been submitted at the GenBank (accession numbers from KJ534648-KJ534655).

Discussion
Accurate diagnosis of malaria parasites holds the key for successful intervention through chemotherapy. To this respect, in recent years PCR-based malaria diagnosis has emerged as a gold standard technique for accurate identification of single as well as mixed malaria parasitic infections [4,5,6]. Identification of specific malaria parasite infection through PCR often is followed by DNA sequencing of the 18S rRNA gene for confirmatory purpose [6]. In many recent studies, some malaria parasite species-specific genes, e.g. pvcsp, pvdhfr, etc. have often been sequenced as additional markers for re-confirmatory purpose in different malaria endemic settings [10,11].
In the present study, PCR diagnostic assays with the 18S rRNA gene revealed the presence of 193 P. falciparum, six P. vivax and two mixed infections due to P. falciparum and P. vivax in Cameroon. The results of DNA sequencing and sequence alignments of the 18S rRNA gene therefore re-confirm the ability of the molecular diagnostic approach in detecting the single infections of P. vivax as well as mixed infection due to these two species by PCR amplification [4,5,6]. Although the DNA sequences of the 18S rRNA gene of Cameroonian P. falciparum yielded 100% similarity with the reference sequence, for P. vivax, when compared with the reference sequence of the SAL-1 strain, we found a novel single nucleotide polymorphism (SNP) at the 72 nd nucleotide position in one of the mixed-infected patients ( Figure 6). Very similar results following analogous protocols (PCR diagnostic assays followed by DNA sequencing) could determine high incidences of mixed malaria parasite infection in India [6]. Taking together the results of the present study with the study from India [6], it could therefore be highlighted that nested PCR amplification of the malaria parasite 18S rRNA gene followed by DNA sequencing and multiple sequence alignment with the reference sequences of the respective 18S rRNA genes of malaria parasites could serve as a valuable aid for molecular diagnosis of malaria infection.
DNA sequencing and multiple sequence alignments of the pvmdr1, pvcsp and pfmdr1 genes in the Cameroonian P. vivax and P. falciparum isolates could validate the results obtained with PCR diagnostic assay. While the four nucleotide substitutions in the pvmdr1 gene in Cameroonian P. vivax might be populationspecific in nature, the observed sequence homology of the pvmdr1 gene in the present study with the reference P. vivax SAL-1 strain reconfirms the presence of P. vivax infection in native Cameroonians. The results on the whole on the DNA sequencing and sequence comparison with reference strains of the pvmdr1 and pfmdr1 genes therefore are in good agreement with the findings with the PCR diagnostic approach followed by DNA sequencing of the 18S rRNA gene (see above). It is thus recommended that associated DNA sequencing and sequence alignments of additional malaria parasite species-specific genetic markers should be conducted to confirm the observations with PCR diagnostic assay targeting the 18S region of the rRNA gene.
Genotyping of the pvcsp gene encompassing the repeats region is often considered to differentiate among three different strains circulating in the global P. vivax populations [32]. To this context, the P. vivax VK210 strain type (also called classic type) is known to be distributed in higher frequency in comparison to the VK247 strain in almost all P. vivax endemic regions [32]. However, in some countries (including African countries) where genotyping of the pvcsp gene has been performed, such as India, Afghanistan, Angola, Equatorial Guinea, both the pvcsp variants (VK210 and VK247) have been found to be circulating in different localities of the same country [11,33,34]. Finding of P. vivax strains of only VK247 type in Cameroon in the present study is therefore interesting. Considering the limited number of P. vivax samples in the present study, such a conclusion (on distribution of only VK247 strain in Cameroon) should be dealt with caution; genotyping more number of isolates will ascertain the distribution of specific strain types in Cameroonian P. vivax.
The Duffy-status of eight Cameroonians infected with P. vivax could be ascertained by direct sequencing after PCR amplification of the part of the human Duffy gene covering the promoter region with the -33 rd nucleotide position. Very similar approach has been followed in studies in Indians [29] and Ethiopians [13] in successfully determining the Duffy status of humans. Finding of all the eight Cameroonians homozygous for the mutated C nucleotide (-33C) justifies that all the eight Cameroonians infected with P. vivax are homozygous Duffy-negative, and that P. vivax can infect Duffy-negative Cameroonians [23] and to other Africans [7,9,11,12,13]. The present observation on the P. vivax infection in Duffy-negative Cameroonians together with similar findings in other African countries therefore indicate that P. vivax might have evolved other, hitherto unknown mechanism(s), to infect Duffy-negative humans. Considering the emerging hypothesis on the African origin of P. vivax [18], it seems highly likely that Africans in general have been selected for the Duffy-negative mutation due to long exposure to P. vivax infection. However, since the current rate of transmission of P. vivax is fairly low in Africa in general, and Cameroon in particular, other host-genetic factor(s) might also be playing significant role in P. vivax infection in Duffy-negative host genetic background. Furthermore, two out of the five principal malaria vectors of Africa (Anopheles gambiae and An. arabiensis) responsible for majority of malaria transmission in Cameroon have been reported to have high vectorial capacity to transmit P. vivax malaria parasite [7,35]. Since P. vivax infection in asymptomatic native Cameroonians has been reported in the south-western part of the country [23], and the fact that the southern part of Cameroon borders Equatorial Guinea and incidences of P. vivax infection in indigenous human populations have already been reported in Equatorial Guinea [11], it seems imperative that P. vivax parasites are capable of successful transmission in multiple African countries. Although the exact mechanism leading to successful infection of P. vivax to Duffy-negative Africans has not yet been established, it seems for now that P. vivax infection might have been mediated independently (or in combination) of factors like, (i) evolution and spread of specific P. vivax strain capable of infecting the Duffy-negative humans, (ii) host genetic susceptibility and (iii) vectorial competence. Considering the simple genomic architecture, high mutation rate, less generation time (in comparison to humans and mosquitoes), it seems highly likely that the P. vivax strains circulating in Sub-Saharan Africa might have evolved to the extent for being able to infect the Duffy-negative Africans. However, whether such P. vivax strains are specific to Africa or introduced from other P. vivax malaria endemic countries is not known. More in-depth sampling of malaria parasites from Sub-Saharan Africa and following population genomic studies with multiple putatively neutral SNPs already developed in this malaria parasite [36,37] could unravel the evolutionary history of African P. vivax.
How important is the report on P. vivax infection in Duffynegative humans for the malaria control programme of Cameroon? Cameroon, like other African countries is already struggling to control malaria due to P. falciparum; additional burden of P. vivax mono-infection as well as P. vivax-P. falciparum mixed infection is therefore daunting. Furthermore, P. vivax infections were found to be spread in four out of five places of collection in the southern parts of Cameroon, which are already reeling under high malaria mortality coupled with chloroquine resistant P. falciparum [21]. This is exemplified by the recent finding on the occurrence of multiple pfcrt haplotypes conferring chloroquineresistance in Cameroonian P. falciparum [38]. The situation is especially griming considering the infection capabilities of P. vivax in Duffy-negative Cameroonians and other Africans [7,9,11,12,13]. As recommended by the World Health Organization, the national treatment policy in Cameroon has established artemisinin-based combination therapies (ACTs; artesunate + amodiaquine [AS+AQ] and artemeter + lumefantrine [AL]) in 2006 as the first-line treatment for uncomplicated malaria [39,40]. Clearly, the current malaria drug policy in Cameroon does not take consideration of either single P. vivax infection or mixed parasite infections [1].