Molecular typing reveals substantial Plasmodium vivax infection in asymptomatic adults in a rural area of Cameroon

This study was authorized by the South West Regional Delegate of Public Health and the University of Buea Institutional Review Board. All subjects gave signed informed consent before enrollment into the study. All protocols involving human subjects were approved by the IRB of the University of Notre Dame.

The study was carried out in Bolifamba, a multi-ethnic rural setting [22] 530 m above sea level situated on the east slope of Mount Cameroon in the South West Region. Although malaria is endemic throughout Cameroon [23], the country has very different geographical and epidemiologic levels [24]. The epidemiology of malaria in Bolifamba has been well described [22]. Malaria transmission occurs all year round, with peak transmissions during the peak rainy months (July and August). There are two seasons in Bolifamba: the rainy season that runs from March to October and the dry season from November to February. The prevalence of malaria parasitaemia in this area ranges from 30% in the dry season to 65% in the rainy season [25]. Plasmodium falciparum accounts for up to 96% of malaria infections in this area [26], with Anopheles gambiae s.s. being the dominant vector [27]. Hydrologically, a stream runs through the village and is of prime importance to the villagers and to the epidemiology of the disease. In this forested area of Southern Cameroon, the equatorial climate has been modified by the double influence of the ocean and the mountain. The average humidity is constantly high, between 75-80% or more and temperatures are lower than in the other areas of the southern part of the country; the mean values of the minimum temperatures are 20°C in December and 18°C in August, the mean values of the maximum temperatures are 35°C in August and 30°C in March. Notwithstanding, seasons in Mt. Cameroon region have become quite variable in the past decade, with rains beginning in June and lasting until early November, causing some variability in malaria prevalence and transmission patterns [28]. The mountain starts at an altitude of about 50 m from the coast as a sedimentary plain that extends from Limbe to Mutengene and Tiko. From Mutengene, the terrain gradually elevates to an altitude of 800–1200 m in Buea town [29]. The mountain in Buea, as well as the Limbe Botanic garden and zoo and other recreational sites in Limbe attract tourists of various nationalities. This makes the populations along the Mount Cameroon area highly cosmopolitan. The indigenous inhabitants of Bolifamba are of several ethnic groups [22] and the main occupation is farming. Farm lands extend to the forest areas. By 2005, the population of Bolifamba was assessed to be about 3,500 inhabitants [22].

The study group consisted of apparently healthy adults resident in Bolifamba and aged 18–55 years. The subjects had lived in the area for at least five consecutive years. Subjects with signs and symptoms of any chronic illness were not included in the study.

A door-to-door visit of randomly selected houses was carried out in all the quarters of the Bolifamba village. All participants were non febrile (body temperature below 37.5°C) at time of blood draw. A brief history of malaria infection and management habits was recorded for each subject. Five mL of blood was collected by venipuncture from each of the 269 subjects enrolled into EDTA containing tubes during the (peak) rainy malaria transmission months throughout the study period (July 2008 to October 2009). The blood was separated into plasma and red cell components by spinning in a Beckmann® centrifuge (Beckmann®, State Technology Inc, Bridgeport NJ, USA) at 2500 rpm for 5 min.

Thick and thin blood smears were prepared from a small portion of whole blood of each subject on the same slide according the methods previously described [30‐32]. The thin blood smears were methanol-fixed and both smears stained with Giemsa stain and examined under the oil immersion lens (x100) and x10 eye piece of an Olympus® light microscope. Parasitaemia was evaluated as previously described [31]. Essentially, with each positive smear, the level of parasitaemia was estimated by counting the parasites against at least 200 leucocytes and then using the leucocyte (WBC) count to estimate the number of parasites/μL of whole blood while assuming an average of 8,000 WBCs/μL of whole blood as shown in equation 1.

As a measure of exposure of the subjects to malaria (through infective bites of mosquitoes), the antibody status of each subject was assessed qualitatively by a rapid immunochromatographic card assay detecting either antibodies to the merozoite surface protein (MSP) or the circumsporozoite protein (CSP) (International Immuno-Diagnostics, CA, USA). The assay was carried out as described by the kit manufacturers (International Immuno-Diagnostics, CA, USA). Briefly, 10 μL of plasma sample was added to a sample well and 3 drops of the supplied assay diluent added immediately to the same sample well and then allowed to migrate by capillary action. Results were interpreted within 10–20 minutes of adding plasma and assay diluent. The appearance of a visible band corresponded to plasma reactivity with either P. falciparum or P. vivax antigens, or both. An additional control band always appeared as an internal control to enable validation of assay results and/or to ascertain that the test card was not faulty in production. When no internal control band was seen, even when the test band(s) appeared, the assay result was discarded and repeated using a fresh card.

Plasmodium speciation was done for all the 269 subjects using the Nested Polymerase Chain Reaction (PCR) method described by Kimura et al.,[33] with some modifications. The Phusion® blood PCR kit was used for the Nest 1 reaction. This enabled the amplification of parasite DNA directly from 0.003% of frozen packed red cells. All PCRs were carried out in a total volume of 20 μL. The reaction mixture had final concentrations of MgCl₂ (3.0 mM), 200 μM of each dNTP, 0.25 pmol/μL of each primer, 2.5 mM EDTA, 5% DMSO, 0.4 U Phusion Blood II DNA Polymerase. The cycling parameters for the Nest 1 reaction using the PTC-100 programmable thermal cycler (MJ. Research Inc., Waltham. MA, USA) were as follows: 1) Lysis of cells/initial denaturation 98°C for 5 minutes; 2) Denaturation at 98°C for 1 sec; 3) Annealing at 72°C for 15 sec; 4) Extension 72°C for 30 sec; 5) Repeat steps 2–4 for a total of 35 times; 6) Final annealing at 72°C for 10 minutes and 7) 4°C indefinitely.

For the Nest 2 reaction the TaKaRa kit (TaKaRa Ex Taq™, Hot Start Version) with the high fidelity HS Taq polymerase was used instead of the Phusion® blood kit. All PCRs were also carried out in a total volume of 20 μL. The reaction mixture had final concentrations of MgCl₂ (2.005 mM), 150 μM of each dNTP, 0.25 pmol/μL of each primer, 0.05 U Taq HS DNA Polymerase. Two micro liters of a 1:50 dilution of the Nest 1 amplicon in MiliQ water (MiliQ system) was used as template in the Nest 2 reaction. The cycling parameters for the Nest 2 reaction using the PTC-100 programmable thermal cycler (MJ. Research Inc., Waltham. MA, USA) were as follows: 1) Initial denaturation 98°C for 5 minutes; 2) Denaturation at 98°C for 1 sec; 3) Annealing at 60°C for 15 sec; 4) Extension 60°C for 30 sec; 5) Repeat steps 2–4 for a total of 18 times; 6) Final annealing at 60°C for 10 minutes and 7) 4°C indefinitely. The sequence of oligonucleotides used in both the NEST 1 and 2 reactions were as previously published [6, 33].

The Nest 2 amplicon was separated on a 2% agarose gel and a gel picture taken under UV light. All buffers and gel preparations were as described by Sambrook et al.[34].

The malaria parasite species bands were extracted from the gel for one subject with triple infection and purified using the QIAquick Gel Extraction Kit essentially as described by the manufacturer (Qiagen®). Five (5) μL of the purified DNA was run on a 2% agarose gel to ascertain purity and for semi-quantification.

A sequencing order for all DNA to be sequenced was submitted (on line) to the sequencing and genotyping facility of the University of Chicago cancer center [35] and the appropriately labeled samples sent to the facility by postal mail service.

The finished sequences were downloaded from the University of Chicago comprehensive cancer center web page [36], cleaned and used in a BLASTn algorithm against the GenBank reference genomic sequence of Plasmodium available at the NCBI web page [37] and the output file showing the hits and similarity profile copied and saved.

The PCR was done with primer pairs previously published [6] with a slight modification of the PCR constituents. The 224 bp fragment was amplified directly from diluted packed red cells or neat plasma using the Direct PCR Blood Phusion® kit (Thermo Scientific) according to the manufacturer’s instructions. All reactions were carried out in a total volume of 20 μL. The reaction mixture had final concentrations of MgCl₂ (3.0 mM), 200 μM of each dNTP, 0.25 pmol/μL of each primer, 2.5 mM EDTA, 3% DMSO, 0.4 U Phusion Blood II DNA Polymerase and 0.5 μL of 0.001% packed red cells. The cycling parameters using the PTC-100 programmable thermal cycler (MJ. Research Inc., Waltham. MA, USA) were as follows; Lyses of cells/initial denaturation 98°C for 5 minutes, Denaturation at 98°C for 1 sec, Annealing at 59°C for 15 sec, Extension 72°C for 30 sec, Repeat steps 2–4 for a total of 35 times, final annealing at 72°C for 10 minutes and 4°C forever.

The amplicons were separated on a 2% agarose gel alongside a 100 bp DNA ladder and the gel photographed under UV light. All buffers and gel preparations were as described by Sambrook et al.[34].

The 224 bp band corresponding to the DARC gene promoter fragment was extracted from the gel and purified using the QIAquick Gel Extraction Kit essentially as described by the manufacturer (Qiagen®). Five μL of the purified DNA was run on a 2% agarose gel to ascertain purity and for semi-quantification.

The sequences were downloaded from the University of Chicago comprehensive cancer center web page [36], cleaned and used in a multiple sequence alignment of the 12 subjects’ DARC gene promoter sequences on the NCBI web site against the DARC gene promoter Duffy positive consensus sequence downloaded from the NCBI web page. The output file showing the alignment profile was copied and saved.

In order to assess by another means whether the fragments amplified from the various subjects contained the sought T → C point mutation that confers the Duffy negative phenotype, a portion of the gel purified DNA was digested with the enzyme StyI, (New England Biolabs, NEB labs). This enzyme will cut the 224 bp fragment at 2 recognition sites to produce 3 restriction fragments of sizes 82, 77 and 64 bp, for Duffy positive genotypes or will cut at 3 recognition sites, producing 4 restriction fragments of sizes 82, 65, 64 and 12 bp, if the subject is Duffy negative. In Duffy negative subjects, the point mutation in the DARC promoter gene segment creates an additional restriction site for StyI in the 77 bp fragment, breaking it into the 65 and 12 bp fragments.

The restriction reaction was preformed according to the kit manufacturer’s instructions. Briefly, 10 μL of each purified DNA was used in the restriction mixture comprising 5 μL of miliQ water, 2 μL of 10X Bovine serum albumin (BSA), 2 μL of 10X NEBuffer 3, and 1 μL of StyI enzyme. The 20 μL total restriction volume in a 1.5 mL eppendorf tube was incubated in pre-heated water in the wells of a heating block at 37°C for at least 1 hour. The entire 20 μL tube content for each subject was mixed with 5 μL of 6x Orange G DNA loading dye (Sigma) and loaded onto a 20% DNA polyacrylamide gel (prepared as described by Sambrook et al.[34]), along with a 100 bp DNA ladder (Promega). The gel was run at 50 V for 3 hours. The bands were visualized under UV light on a Trans-illuminator.

Given that the wet lab separation of the restriction product of the 20% native DNA PAG could be very dubious, the full DARC promoter sequence of each subject was subjected to in silico restriction digestion with the same enzyme (StyI) that was used in the wet lab analysis in order to confirm or refute the more convincing sequence alignment profile relative to the restriction electrophoregraph. This was done using NEBcutter 2 software freely available on the web. Each subject’s full length amplified DARC DNA sequence was interrogated with in silico StyI restriction enzyme and the restriction map showing number of restriction sites copied.

The data were analysed using the Statistical Package for Social Sciences (SPSS) version 17.0 (Chicago, IL USA), and the Microsoft Excel 2010 (Microsoft Corporation). The Chi-Square test and the Pearson’s correlation were used to find correlation between categorical and numeric variables respectively. Statistical significance was set at p ≤ 0.05. Multiple sequence alignments were done using freely available web-based ClustalW software. The free web based NEBCutter 2 software was also available for in silico enzymatic restriction digestion. Sequences for analyses were either reversed or reverse complemented using free web-based softwares. Reference sequences were downloaded from respective public databases.

Representative examples of thin films positive by Giemsa staining confirmed malarial infection suggestive of P. falciparum as well as other species (see Figure 1). Unfortunately, some of the morphologies shown in Figure 1 do not show hallmark characteristics of any species. Figure 1 B and C are not likely to be P. falciparum because they are trophozoite stage parasites and schizont stages seen in peripheral circulation. Figure 1 D does show a characteristic P. falciparum ring, but convincing speciation required additional analyses.

To determine whether P. vivax was present in our samples, at first pass they were examined using antigens that will bind to both P. vivax and P. falciparum specific antibodies. This was done using qualitative anti-MSP/CSP antibody immunochromatographic card tests that distinguish between antibodies to both the MSP/CSP of P. falciparum and P. vivax. Based on this anti MSP/CSP antibody card assay, the overall prevalence of malaria exposure was 72.1% (194/269). The prevalence of single and mixed Plasmodium antibodies is shown on Table 1. The vast majority (66%) of samples indicated exposure to P. falciparum alone, ~1% exposure to P. vivax and about 4% exposure to both.

Since P. malariae can also be found in Cameroon but not detected by the antigen kit, PCR analysis was used for definitive analysis on speciation (Figure 2). Of 269 samples collected, 66 were positive for P. falciparum (Table 2). Notably nine were positive for P vivax, four were positive for P. malariae and none contained P. ovale alone. Three were double positive P. vivax and P. falciparum another four showed both P. malariae and falciparum. One was triply positive for P. malariae, P. falciparum and P. vivax and here the primary data are shown in Figure 2. These bands were isolated and sequenced to confirm the simultaneous presence of all three species based on similarity scores (Figure 3). Although parasite morphology on the P. vivax positive smears (Figure 1, Panels A-C) did not all show landmark features of P. vivax-infected erythrocytes, the fact that they were read from slides that were solely positive for P. vivax by PCR was a confirmation that the parasites seen were indeed P. vivax. In Figure 1, Panel D shows a typical Pf ring from a sample that was PCR positive for just P. falciparum.

BLASTn analyses of the DNA sequences generated from the sample with triple bands on Figure 2 confirmed that the P. falciparum, P. vivax and P. malariae bands matched the Genbank P. falciparum isolate SF4 18S ribosomal RNA gene partial sequence ID: gb│JQ627152.1│, P. vivax isolate SV6 18S ribosomal RNA gene partial sequence ID: gb│JQ627158_1│, and P. malariae type 2 small subunit ribosomal RNA gene complete sequence ID: gb│AF488000.1│ by 100%, 97%, and 100% respectively (Figure 3).

Overall there were as many as 87 infected of a total of 269 participants, suggesting as many as a third of asymptomatic patients at this site may contain parasite reservoirs. There were many cases where the subjects were positive for parasite by PCR but negative for anti-MSP/CSP antibodies (9.3%, 25/269). This may be due to very low levels of both P. falciparum and P. vivax infection and suggests that almost a third of the P. vivax infections are truly asymptomatic in absence of an antibody response.

Having recoded a number of P. vivax positive cases among the study subjects by Nested PCR and confirmed by microscopy and sequenced data, the Duffy genotype of the 13 subjects with either P. vivax single or mixed infections was investigated. The 224 bp DARC gene promoter fragment was successfully amplified in all the samples. Each subjects’ DARC promoter DNA sequence interrogated with in siliso StyI restriction enzyme showed that StyI cut the Duffy positive promoter DNA sequence at two sites, producing three fragments of 82, 77 and 64 bp (Figure 4) while the Duffy negative promoter DNA sequence was cut additionally at a site created by the T → C point mutation in the 77 bp fragment, thereby producing 4 fragments of 82, 12, 65, and 64 bp (Figure 4). Multiple sequence alignment of the sequenced DARC promoter gene fragments for 12 of the 13 assessed subjects; alongside a Duffy positive consensus sequence showed the T → C point mutation which confers the Duffy negative phenotype (Figure 5). This established the prevalence of Duffy positive individuals in the present investigation at 50% (6/12). One case was lost and could not be analysed.

Three methods, namely Giemsa staining, antibodies and PCR, were used to detect parasites in asymptomatic individuals. The overall prevalence of anti-parasite antibodies was relatively high (72.1%) but this largely reflected the fact that subjects were constantly exposed to infective bites of mosquitoes or sustained an immune response after active infection. Nonetheless this test and Giemsa gave the first clue that the study subjects may be infected with P. vivax parasites. However PCR clearly established that about a third of the study population was infected with malaria parasites. This suggests an unexpectedly high level of asymptomatic infection. Further a substantial portions of infections (~15%) contained P. vivax. Nonetheless this was lower than the 33.3% prevalence of microscopically diagnosed “P. vivax-like” infection described by Kimbi et al. [20]. It was surprising that P. ovale failed to be detected, since other reports in the area that have documented the presence of P.ovale[20, 22, 38]. As follow-up of the findings in the present study, the wider existence of P. vivax in Bolifamba and Cameroon should be further ascertained.

The present report documents the first substantiated evidence of P. vivax infection in Cameroonian Bolifamba residents, in the SWR of Cameroon, raising the number of infecting human Plasmodium species in Cameroonians to four. There have been other reports of P. vivax elsewhere on the African continent where P. vivax was initially thought to be non-existent or that the population was refractory to infection. These countries include Angola [6], Equatorial Guinea [6, 7], Madagascar [8], Mauritania [9], Sudan [10], Kenya [11], Sao Tomé and Principe [12‐15], Congo [14], Mali [16], and Ethiopia [17]. Although the exact prevalence data for P. vivax in Africa is not known, the parasites seem to be more prevalent in countries where many Duffy positive people are permanent residents [8, 9, 11]. However, the Duffy negative status has been plausibly shown not to be a barrier to P. vivax infection as previously believed [8] and confirmed in other studies [6, 9], as well as the current one. The increasing prevalence of P. vivax on the African continent and its discovery in Cameroonians could be inversely correlating to the decreasing prevalence of P. falciparum as a result of targeted control of the latter relative to the other malaria parasite species. A similar phenomenon has been documented in Malaysia, where an increasing incidence of P. knowlesi is being linked to the control of P. falciparum and P. vivax[39].

With the notion that the Duffy antigen is necessary for the invasion of P. vivax into red cells, the Duffy genotype of all the subjects who were positive for P. vivax was assessed. Six (6) of the P. vivax infected subjects were Duffy positive (50%) and the other 6, Duffy negative. Following the three investigative molecular approaches that were used to determine the Duffy genotype of the P. vivax-infected individuals, two were very evident, while one was compromised by the low power of resolution of restriction fragments on a 20% native DNA PAG. The report of Duffy positive Cameroonians is a new finding. No previous study has been conducted in Cameroon to identify and determine the prevalence of the Duffy positive phenotype or genotype because of the general belief that Africans (Blacks) were largely or entirely Duffy negative and resistant to P. vivax infection. The findings in the present study corroborate a series of previous reports of the Duffy positive genotype on the African continent where it was thought to be non-existent and therefore a natural protecting factor for Africans against P. vivax infection [39, 40]. Recently, Mendes and colleagues [8] showed unequivocally that the Duffy negative genotype was no longer a barrier to P. vivax infection, with the parasite infecting both Duffy positive and negative people, while others have implicated P. vivax as the causative agent in cases of severe malaria in African children [10, 16].

The south-west region of Cameroon is home to Mount Cameroon and there are several recreational and touristic sites that attract individuals of various nationalities. With the movement of people across geographic regions, it is likely that P. vivax was introduced in Cameroon a long time ago. The malaria vectors and parasites are equally changing their host range and/or adapting to new invasion strategies [29]. This is evident from the recent observation that P. knowlesi, originally a simian malaria species, can now infect humans [5]. A similar adaptation of the P. vivax parasite could be occurring, accounting for its increasing prevalence on the African continent. It is equally likely the P. vivax species may have been there but technology did not allow their detection. The observed prevalence of 50% for the Duffy positive antigen is relatively high and therefore cannot be ignored despite the small sample size of P. vivax positive subjects. The finding that both Duffy positive and negative malaria asymptomatic Bolifamba residents habored P. vivax parasitaemia suggests that the parasite may have evolved and could be using novel receptors to gain access into erythrocytes as proposed by some research groups [6, 11, 41]. This idea was most conclusively proven correct by Ménard and colleagues [8].