INTRODUCTION
Three major malaria parasite species infect individuals of the Raglai community of the Khanh Phu commune, Khanh Hoa Province, Vietnam; Plasmodium falciparum, Plasmodium vivax and Plasmodium knowlesi (Plasmodium malariae is also present at very low prevalence). The latter of these species is predominantly a parasite of non-human primates, and infects macaque monkeys in the surrounding forest region. In order to design appropriate interventions to reduce the malaria burden in this region, it is crucial to understand the transmission dynamics of these three species. Such an understanding may be helped by characterisation of the gametocyte production in human hosts of the species, along with surveys of sporozoite carriage amongst mosquitoes from the same region.
It has been suggested that humans and macaques living in the same region of Borneo may share a P. knowlesi parasite pool, based on the fact that P. knowlesi parasites isolated from humans and monkeys share identical sequences of two commonly used genetic markers, the circumsporozoite protein (csp) gene and mtDNA (mitochondrial DNA) haplotypes (Lee et al. Reference Lee, Divis, Zakaria, Matusop, Julin, Conway, Cox-Singh and Singh2011). More recent work utilizing a population genetics approach, has revealed that the P. knowlesi parasites infecting humans in Malaysia are a mixture of two distinct populations each associated with two separate sylvatic macaque reservoirs (Divis et al. Reference Divis, Singh, Anderios, Hisam, Matusop, Kocken, Assefa, Duffy and Conway2015). Whole-genome sequencing of 48 clinical isolates from Borneo confirmed this population structure, and revealed evidence for reproductive isolation between the sympatric populations (Assefa et al. Reference Assefa, Lim, Preston, Duffy, Nair, Adroub, Kadir, Goldberg, Neafsey, Divis, Clark, Duraisingh, Conway, Pain and Singh2015). It was also shown that genes under strong balancing selection in human malaria parasites, such as ama1, do not appear to be subject to the same degree of selection in P. knowlesi. Conversely, the CSP encoding gene, csp, was shown to be under stronger balancing selection in P. knowlesi than it is in P. falciparum (Assefa et al. Reference Assefa, Lim, Preston, Duffy, Nair, Adroub, Kadir, Goldberg, Neafsey, Divis, Clark, Duraisingh, Conway, Pain and Singh2015).
It is currently unknown whether transmission of P. knowlesi occurs within human populations (human-to-human transmission) or whether human infection is only ever the result of zoonotic transmission from infected monkeys. Establishing whether human to human transmission occurs is of crucial importance when considering both the status of P. knowlesi as a true human malaria parasite, and for the implementation of appropriate control measures. Gametocytes of P. knowlesi have been observed in human infections both by microscopy (Cox-Singh et al. Reference Cox-Singh and Singh2008) and by reverse transcription PCR (RT–PCR) targeting the pks25 gene (Grigg et al. Reference Grigg, William, Menon, Dhanaraj, Barber, Wilkes, von Seidlein, Rajahram, Pasay, McCarthy, Price, Anstey and Yeo2016), suggesting that the parasite has the potential to infect mosquitoes from man, at least in Malaysia. Human-to-human transmission via mosquitoes has previously been demonstrated in the laboratory (Chin et al. Reference Chin, Contacos, Collins, Jeter and Alpert1968), and it seems likely that this would occur under natural conditions. However, P. knowlesi infections of man in Khanh Phu commune, Vietnam are markedly different from those observed on Malaysia, with the parasite exclusively infecting individuals in co-infections with other malaria parasites, and never having been observed by microscopy (Marchand et al. Reference Marchand, Culleton, Maeno, Nguyen and Nakazawa2011). Given this, we were interested in determining whether the P. knowlesi parasites observed amongst the forest fringe dwelling Raglai population were exclusively the result of opportunistic zoonotic infections, or were being transmitted from person to person within the commune.
As all the P. knowlesi infections of humans in this region occur in co-infections with the human malaria parasites P. falciparum and/or P. vivax, we assayed the malaria parasite species composition in the salivary glands of sporozoites collected both at the forest fringe, and deeper within the forest, in order to determine if P. knowlesi is transmitted to humans independently of the human malaria parasites, which would suggest zoonotic transmission rather than a human–human route.
We attempted to detect gametocyte-specific mRNA of P. falciparum, P. vivax, P. malariae and P. knowlesi by RT–PCR using dried blood spots collected from malaria parasite-infected individuals, and to match their allelic types to those of sporozoites from infected mosquitoes captured in the same region. For the detection of P. knowlesi gametocytes, we used a RT–PCR targeting Pks25, the P. knowlesi orthologue of Pvs25.
MATERIALS AND METHODS
Parasites
Positive control RNA and genomic DNA (gDNA) for P. falciparum was obtained from 3D7-9A in vitro cultured parasites, whilst P. vivax and P. malariae gDNA was extracted from Giemsa's solution-stained blood films (Maeno et al. Reference Maeno, Nakazawa, Dao, Yamamoto, Giang, Hanh, Thuan and Taniguchi2008). Plasmodium knowlesi H strain (ATCC No. 30158) gDNA and RNA, obtained from an experimentally infected Japanese macaque (Macaca fuscata) (Kawai et al. Reference Kawai, Hirai, Haruki, Tanabe and Chigusa2009) was used as a positive control. Twenty-five μL aliquots of clone 3D7-9A or P. knowlesi H strain-infected blood was placed in a micro-reaction tube and kept at −80 °C until PCR or RT–PCR analysis. Twenty-five μL of ten times serially diluted (104, 103, 102, 101, 100, 10−1, and 10−2) gametocytes μL−1 P. knowlesi H strain-infected blood were prepared for sensitivity tests for the determination of the detection limit of P. knowlesi gametocytes by RT–PCR. For these sensitivity tests, cDNA was synthesized from whole blood twice, independently, and serial dilutions of these carried out in duplicate with PCR analyses carried out four times, independently. Experimental infection was approved by Animal experiment Committee, Dokkyo University Medical School, Japan (permit number: 0656).
Patients
People living in Khanh Phu commune, Khanh Vinh district, Khanh Hoa Province, Vietnam and diagnosed with malaria both by microscopic examination and by PCR, were recruited for blood sample donation. Participants who frequently work in the forest were enrolled from March 2009 to March 2010. The study area and these samples were previously described (Marchand et al. Reference Marchand, Culleton, Maeno, Nguyen and Nakazawa2011). One hundred and twenty-five people, aged 1–67 (mean ± s.d.: 23 ± 16 years) were enrolled in this study. There were 82 male and 43 female patients. Treatment was performed according to the guidelines of the Vietnamese Ministry of Health. Twenty healthy Vietnamese (with a past history of malaria), who were negative by both microscopy and PCR, and 20 healthy Japanese (never exposed to Plasmodium species) volunteers were also enrolled as controls. The study was reviewed and approved by the National Institute of Malariology, Parasitology and Entomology (Hanoi) and by the ethics committees of Institute of Tropical Medicine, Nagasaki University. All adult volunteers provided informed consent and for children, consent was obtained from close relatives. Blood was collected by finger-prick; thick and thin blood films were made for microscopy, and blood was applied to filter paper for molecular analyses. Each blood-spotted filter paper was immediately air dried, placed in a sealed plastic bag and stored at room temperature. Blood samples from participants were previously examined by PCR for the presence and species of malaria parasites. The number of single and mixed Plasmodium species infections was 71 and 54, respectively (Table 1). There were a total of 95 samples containing P. falciparum parasites, 69 with P. vivax, 6 with P. malariae and 32 with P. knowlesi (Table 1). Plasmodium ovale was not detected in this study. All 32 P. knowlesi infections were found in co-infections with at least one of the human malaria parasite species, these results were previously published elsewhere (Marchand et al. Reference Marchand, Culleton, Maeno, Nguyen and Nakazawa2011).
Table 1. Frequency of Plasmodium species in samples of human blood collected from Khanh Phu determined by microscopy and PCR (Marchand et al. Reference Marchand, Culleton, Maeno, Nguyen and Nakazawa2011) and the number of samples analysed for gametocyte markers
Pf, Plasmodium falciparum; Pv, Plasmodium vivax; Pm, Plasmodium malariae; Pk, Plasmodium knowlesi.
Mosquito collection and dissection of salivary glands
Mosquitoes were collected from both the forest fringe, and within the forest in Khanh Phu from January 2008 to February 2010, as previously described (Marchand et al. Reference Marchand, Culleton, Maeno, Nguyen and Nakazawa2011; Maeno et al. Reference Maeno, Quang, Culleton, Kawai, Masuda, Nakazawa and Marchand2015). Female anopheline mosquitoes were dissected for salivary glands, midguts and ovaries and these were examined by microscopy for sporozoites, oocysts and parity (tracheoles). Sporozoite-infected salivary glands were applied to filter paper and dried in an ambient atmosphere, before storage in closed vials at 4–6 °C.
gDNA extraction and PCR detection
Extraction of gDNA from dried human blood and sporozoite-infected salivary glands on filter paper and subsequent PCR analysis was carried out as previously described (Maeno et al. Reference Maeno, Nakazawa, Dao, Yamamoto, Giang, Hanh, Thuan and Taniguchi2008; Nakazawa et al. Reference Nakazawa, Marchand, Quang, Culleton, Manh and Maeno2009). Briefly, gDNA was extracted using QIAamp DNA micro kit (Qiagen, Tokyo, Japan). Plasmodium species-specific nested-PCR assays to identify human malaria parasites were performed as described by Singh et al. (Reference Singh, Bobogare, Cox-Singh, Snounou, Abdullah and Rahman1999). The first PCR (nest 1) was performed as described by Singh et al. (Reference Singh, Bobogare, Cox-Singh, Snounou, Abdullah and Rahman1999). For the detection of the P. knowlesi 18S rRNA gene, the primers Pmk8 and Pmk9 were used in the second PCR (nest 2) (Singh et al. Reference Singh, Kim Sung, Matusop, Radhakrishnan, Shamsul, Cox-Singh, Thomas and Conway2004). For nest 2, 2 µL of nest 1 amplification product was used as template in the reaction mixtures (25 µL). For authenticating P. knowlesi infection, detection of the csp gene of P. knowlesi from samples was carried out as previously described (Vythilingam et al. Reference Vythilingam, NoorAzian, Huat, Jiram, Yusri, Azahari, NorParina, Noorrain and Lokmanhakim2008). For the detection of gDNA for region 3 of Pfg377, pvs25 and pks25 in mosquito samples, reaction mixtures (25 µL) comprised 1μL of gDNA as a template, 0·5 µ m of each primer, 200 µ m dNTP, 1 units of KOD-Plus-Neo DNA polymerase (Toyobo, Osaka, Japan), 1·5 mm MgCl2 and 1× PCR buffer. Target gDNA was amplified using a PCR protocol consisting of a denaturation step (98 °C, 2 min) followed by 35 cycles (98 °C, 10 s; [pfg377: 55 °C, pvs25 and pks25: 58 °C], 30 s; and 68 °C, 30 s). A 2720 Thermal cycler (ABI, Foster city, CA, USA) was used for all PCRs. PCR products were separated by electrophoresis on 1·5% agarose gels and stained with ethidium bromide. DNA bands were analysed with Lane & Spot Analyzer software (Atto, Tokyo, Japan). No amplification was observed with human gDNA controls. Primer sequences for 18SrRNA of human Plasmodium species (Singh et al. Reference Singh, Bobogare, Cox-Singh, Snounou, Abdullah and Rahman1999), and 18SrRNA of P. knowlesi (Singh et al. Reference Singh, Kim Sung, Matusop, Radhakrishnan, Shamsul, Cox-Singh, Thomas and Conway2004) and the csp gene of P. knowlesi (Vythilingam et al. Reference Vythilingam, NoorAzian, Huat, Jiram, Yusri, Azahari, NorParina, Noorrain and Lokmanhakim2008) were as previously described. The primer sequences for region 3 of pfg377 gDNA were those used by Menegon et al. (Reference Menegon, Severini, Sannella, Paglia, Sangare, Abdel-Wahab, Abdel-Muhsin, Babiker, Walliker and Alano2000). Primer sequences for pvs25 gDNA and pks25 gDNA were 5′-CCC TAG GCA AAG CAT GCG GGG AA-3′, 5′-TTG CCA ATA GCA CAT GAG CAA CC-3′ and 5′-ACC TAG GTA AAA CAT GTG GAG AC-3′, 5′-TTG CCA ATA GAG CAT GAA CAT GC-3′, respectively. These are designed based on GenBank sequences XM_001608410·1 of P. vivax Sal-1 and XM_002261792·1 of P. knowlesi strain H (PKH_061530).
RNA extraction and RT–PCR conditions
Extraction of total RNA and reverse transcription for first-strand synthesis of cDNA from each dried blood sample were carried out as previously described (Maeno et al. Reference Maeno, Nakazawa, Nagashima, Sasaki, Higo and Taniguchi2003). Briefly, total RNA was extracted using ISOGEN (Nippon Gene, Tokyo, Japan) according to the manufacturer's instructions. The extracted total RNA was treated with DNase (Takara Bio, Shiga, Japan), and subjected to PCR without reverse transcription in order to check for gDNA contamination of RNA. Extraction of total RNA and RT-PCR for the positive controls and sensitivity tests samples performed in the same way as for the field samples. PCR designed to amplify mRNA for the region 3 of Pfg377, pvs25, pmg or pks25 was performed on the cDNA. Reaction mixtures (25 µL) comprised 1 µL of cDNA as a template, 0·5 µ m of each primer, 200 µ m dNTP, 1 unit of KOD-Plus-Neo DNA polymerase (Toyobo, Osaka, Japan), 1·5 mm MgCl2 and 1× PCR buffer. Target cDNA was amplified using a PCR protocol consisting of a denaturation step (98 °C, 2 min) followed by 35 cycles [98 °C, 10 s; (pfg377: 55 °C, pvs25, pmg and pks25: 58 °C), 30 s; and 68 °C, 30 s]. The primer sequences for region 3 of pfg377 mRNA were as previously described (Menegon et al. Reference Menegon, Severini, Sannella, Paglia, Sangare, Abdel-Wahab, Abdel-Muhsin, Babiker, Walliker and Alano2000). Primer sequences for pvs25 mRNA were the same as those used for detecting gDNA sequences. Primer sequences for the first PCR (nest 1) of pks25 mRNA and pmg mRNA for gametocyte of P. malariae were 5′-TAT GTA ACG ACG GGC TGG TG-3′, 5′-ATG GCG TAC GAA AGG ACT AGC-3′, 5′-GAA ACA GCT ACC GAA GGC GA-3′, 5′-GGT CGG TTT GCC CAA TTT CA-3′, respectively. Primer sequences for the second PCR (nest 2) of pks25 mRNA were the same as those used for detecting gDNA sequences. These were designed based on GenBank sequences XM_002261792·1 of P. knowlesi strain H (PKH_061530). Primers sequences for pmg mRNA were designed based on GenBank sequences LT594635·1 of P. malariae genome assembly, chromosome 14. A 2720 Thermal cycler (ABI) was used for all PCRs. PCR products were separated by electrophoresis on 1·5% agarose gels and stained with ethidium bromide. The gels were then photographed, and the densities of the DNA bands were analysed with Lane & Spot Analyzer software (Atto, Japan).
Sequencing and analysis of pfg377 pvs25 and pks25 mRNA
For nucleotide sequencing, human blood and mosquito samples were randomly chosen from pfg377, pvs25 or pks25 mRNA-positive samples. Specific products resulting from PCR amplification of pfg377, pvs25 or pks25 were cleaned using the Wizard SV Gel and PCR Clean-up System (Promega, Tokyo, Japan) according to the manufacturer's instructions, and were then sequenced with the BigDye Terminator v3.1 Cycle Sequencing Premix Kit (ABI). The reaction products for sequencing were separated with an ABI/Hitachi 3130 × 1 Genetic Analyzer (ABI) and the resulting nucleotide sequences were compiled using Genetyx (Genetyx Corporation, Tokyo, Japan).
Statistical analyses
Statistical evaluation was performed using chi-square (χ 2) test, Mann–Whitney test (two-tailed) and regression analysis (two-tailed). For the statistical analysis, expression data of target mRNAs of pfg377, pvs25 and pks25 were standardized by quantification of β-actin mRNA as an internal control. All analyses were performed using SPSS software (SPSS Japan, Tokyo, Japan), a value of P < 0·05 considered significant.
RESULTS
Detection of mRNA of the gametocyte-specific proteins pfg377, pvs25 and pks25
In order to assess the relative rates of gametocyte prevalence for each of the malaria parasite species infecting humans in Khanh Phu, we assayed the presence of species-specific gametocyte derived mRNA in blood samples from malaria-infected individuals. RNA samples derived from dried human blood did not yield any RT–PCR products without the addition of reverse transcriptase (Fig. 1A), confirming the absence of DNA in these samples. Each species-specific primer pair only amplified the species for which they were designed, and there was no cross-reactivity between species (Fig. 1B). The specific RT–PCR products for pks25 mRNA were not detected for the asexual stages of P. knowlesi (data not shown).
Fig. 1. Agarose gel electrophoresis of RT–PCR products for pvs25, pks25 and pfg377. (A) PCR with or without RT step for detection of pvs25, pks25 and pfg377 mRNA. (B) Detection of gametocytes-specific mRNA, pvs25, pks25 and pfg377, in dried positive control blood. (C) RT–PCR and PCR products for Plasmodium knowlesi gametocytes, pks25 mRNA in serial dilution of dried P. knowlesi H strain infected blood. M, 100-bp DNA ladder markers; ‘+’ and ‘−’, PCR with or without the RT step; Pv, P. vivax; Pk, P. knowlesi; Pf, P. falciparum; 104, 103, 102, 101, 100, 10−1 and 10−2 number of gametocytes μL−1 P. knowlesi H strain-infected blood.
Detection of gametocytes was performed by PCR targeting genes only expressed by gametocytes from cDNA derived from mRNA by RT–PCR for pfg377, pvs25 and pmg, or by nested RT–PCR for pks25. The sensitivity of detection of pks25 mRNA by nested RT–PCR was found to be 1 gametocyte µL−1 using a serial dilution of gametocytes derived from the blood of a P. knowlesi infected monkey (Fig. 1C). The detection limit of the pfg377 assay was also shown to be 1 gametocyte µL−1 (Maeno et al. Reference Maeno, Nakazawa, Dao, Yamamoto, Giang, Hanh, Thuan and Taniguchi2008; Nakazawa et al. Reference Nakazawa, Marchand, Quang, Culleton, Manh and Maeno2009). The sensitivity of the pvs25 and pmg assays were not examined in the present study, but were assumed to be comparable to that of the previous two assays.
Expression of mRNAs for pks25, pfg377 and pvs25 was detected in 15 out of 32 (47%), 35 of 95 (37%) and 42 of 69 (61%) of samples infected with P. knowlesi, P. falciparum or P. vivax, respectively. The detection rate of gametocytes by microscopy was 0 (0%), 15 (15%) and 26 (37%) in the same samples (Fig. 2). Expression of mRNA for pmg was not detected in this study. Expression of mRNA for pks25, pfg377 or pvs25 was not detected in any blood samples from non-malaria-infected Vietnamese and Japanese volunteer controls. There was no statistically significant association between the presence of mRNA of pfg377, pvs25 and pks25 in P. falciparum, P. vivax and P. knowlesi malaria patients and gender, age or symptoms.
Fig. 2. Gametocyte positivity rate in patients infected with Plasmodium knowlesi, Plasmodium falciparum and Plasmodium vivax. pks25, pvs25 and pfg377 mRNA expression by RT–PCR.
Allelic diversity of pfg377, pvs25 and pks25
To investigate the relative strain diversity of the species infecting humans in this region, we genotyped the gametocyte-specific genes of each species to determine the number of distinct alleles per strain, and their relative proportions in both humans and mosquitoes. Two distinct alleles of pfg377 region 3 were detected in the 35 samples selected for sequencing. Amino-acid sequence analysis revealed that these were identical to alleles A and C previously observed in Vietnam (Fig. 3A) (Maeno et al. Reference Maeno, Nakazawa, Dao, Yamamoto, Giang, Hanh, Thuan and Taniguchi2008). Thirty-three of the 35 pfg377 mRNA-positive samples showed monotypic allele infections; the remaining two contained both alleles. Among the monotypic allele infections, 19 were of type A and 14 of type C.
Fig. 3. Deduced amino-acid sequences of region 3 of Pfg377 (A), Pvs25 (B) and Pks25 (C). 3D7, control sample of known amino-acid sequences of region 3 of Pfg377 of Plasmodium falciparum 3D7 (GenBank sequences, XM_001350849); Sal-1, control sample of known amino-acid sequences of Pvs25 of Plasmodium vivax Sal-1 strain (GenBank sequences, XM_001608410·1); Pk H, control sample of known amino-acid sequences of ookinete surface protein Pks25 of Plasmodium knowlesi strain H (PKH_061530) (GenBank sequences, XM_002261792·1).
Nucleotide sequencing was performed on the RT–PCR amplified portion of the pvs25 mRNA (nucleotide positions 656–816) for 21 samples out of the 42 pvs25 mRNA-positive samples for which there was sufficient material for sequencing. Three non-synonymous nucleotide mutations were identified with respect to the P. vivax Sal-1 strain (GenBank sequence, XM_001608410·1) (Fig. 3B). The non-Sal-1 type allelic variants were classified into two types; type 1, amino acid substitution at E97Q and I130T; and type 2, amino acid substitution at I130T and Q131K (Fig. 3B). Monotypic allele infections of types 1 and 2 were identified in four and 10 samples, respectively. Seven samples were mixed infections of the two allelic types. All of the pks25 mRNA (nucleotide positions 203–422) positive samples had nucleotide sequences identical to that of ookinete surface protein pks25 of P. knowlesi GenBank sequence XM_002261792·1 (Fig. 3C).
The presence of particular allelic types of either pfg377, pvs25 or pks25 was not associated with either age or sex of the patient.
Allelic polymorphism of pfg377, pvs25 and pks25 in human and mosquito samples
In order to determine whether the malaria parasite species and strains observed infecting individuals in Khanh Phu were transmitted in co-infections or singly, and whether separate mosquito populations were responsible for the transmission of different strains and species, we assayed the species and strain (based on Pvs25, Pks25 and Pfg377 alleles) composition in mosquitoes collected from the forest, and the forest fringe areas. We analysed 11 sporozoite-positive mosquitoes captured in the forest fringe and 59 captured in the forest (Supplementary Table S1). PCR analysis on sporozoite-positive salivary glands revealed the presence of P. vivax, P. falciparum, P. malariae and P. knowlesi (Table 2). Of the 70 sporozoite-positive mosquitoes, 46 were examined for pfg377, pvs25 and Pks25, and eight from the forest fringe and thirty-eight from the forest. pfg377 was detected in 33 samples, pvs25 in 22 and Pks25 in 10. Seven samples were positive for all three genes, one was a mixed infection composed of Pfg377 and Pks25, one was positive for both Pfg377 and Pks25, and nine were positive for Pks25 and Pvs25 (Table 2).
Table 2. Number of salivary glands of Anopheles dirus mosquitoes infected with parasites collected in different sites in the forest near Khanh Phu, Vietnam
Pf, Plasmodium falciparum; Pv, Plasmodium vivax; Pm, Plasmodium malariae; Pk, Plasmodium knowlesi.
The frequency of pfg377 allelic type A was higher than that of type C in the forest, while that of A and C was equal at the forest fringe. Pvs25 allelic type 2 was detected more than type 1 in the forest, while only type 2 was detected at forest fringe (Fig. 4). The differences in allele frequencies in samples collected in the forest fringe and those collected in the forest are not statistically significant, but the frequencies of the pfg377 and pvs25 allelic types identified in sporozoite-infected mosquito salivary glands collected in the forest were similar to those observed in blood samples collected from patients (Fig. 4).
Fig. 4. Allelic types for Pfg377 (A) and Pvs25 (B) in human blood samples and mosquito samples. Mosquito samples were collected in the forest and the forest fringe. Allelic types of pfg377 and pvs25 in gDNA were determined by PCR. Those in mRNA were by RT–PCR.
There was no polymorphism at the pks25 gene; all the alleles sequenced from both human and mosquito infections were identical.
A phylogenetic tree showing the relationships of the pvs25 and pks25 alleles detected in this study compared with sequences available in GenBank is given in Supplementary Fig. S1.
DISCUSSION
Molecular detection of gametocytes of P. knowlesi, P. falciparum and P. vivax revealed variations in gametocyte production between species. Expression of mRNA for pks25, pfg377 and pvs25 was detected in 47% (15/32), 37% (35/95) and 61% (42/69) of samples infected with P. knowlesi, P. falciparum or P. vivax, respectively. Although not statistically significant, the difference in gametocyte detection between P. falciparum and P. vivax is consistent with that reported from previous epidemiological studies; gametocyte carriage in P. vivax infections is more frequent than in P. falciparum infections (Mckenzie et al. Reference Mckenzie, Wongsrichanalai, Magill, Forney, Permpanich, Lucas, Erhart, O'Meara, Smith, Sirichaisinthop and Gasser2006; Kuamsab et al. Reference Kuamsab, Putaporntip, Pattanawaong and Joungwuiwes2012), although differences in methodologies and epidemiological settings between the present report and these previous reports makes direct comparison difficult.
There has been relatively little work on gametocyte production in naturally occurring P. knowlesi infections of man. Gametocytes were observed by microscopy in the seminal report from Malaysian Borneo describing the infection of large numbers of people with the parasite (Singh et al. Reference Singh, Kim Sung, Matusop, Radhakrishnan, Shamsul, Cox-Singh, Thomas and Conway2004). A recent report utilizing a similar molecular approach to gametocyte detection, also based on detection of pks25 mRNA, revealed that 85% of symptomatic P. knowlesi patients in Malaysia with microscopically detectable parasitaemia carried gametocytes at the time of sampling. This rate of gametocyte carriage is higher than that reported here for P. knowlesi infections (47%). Many factors could explain this difference, including parasite strain differences, host genotype differences, the fact that the Vietnamese infections were all sub-microscopic infections, and possible differences in assay sensitivity.
We were able to identify the same allelic types of pfg377 and pvs25 both in infected human blood and in the local mosquito population, which demonstrates their usefulness for the study of transmission dynamics in this area. There was only one allelic type of pks25 detected in this study, and it was found in both human and mosquito infections. The relatively low genetic diversity, in both P. falciparum and P. vivax marked by a restriction to their most predominant allelic types, is as expected in a region that experienced a strong, albeit gradual, transmission reduction.
In Khan Phu, all P. knowlesi infections occur in co-infections with either P. falciparum, P. vivax or both parasites. We were interested to know whether the same mosquitoes were responsible for transmitting all species, or if the P. knowlesi infections were acquired independently from the other species, via the bites of mosquitoes that had previously fed on monkeys. If P. knowlesi is predominantly infecting humans via mosquitoes that have previously fed on monkeys, then most P. knowlesi infections in mosquitoes might be expected to occur in the absence of co-infections with human malaria parasites. Conversely, a large percentage of P. knowlesi co-infections with human malaria parasites in mosquitoes would suggest that the parasite is capable of transmitting from human to human in this setting. We found that 70% of mosquitoes that had P. knowlesi in their salivary glands also carried human malaria parasites, suggesting that mosquitoes are infected with P. knowlesi from human infections. The alternative possibility, that mosquitoes acquire mixed infections of P. knowlesi and human malaria parasites sequentially, following feeding on monkeys and humans is also possible, although less parsimonious.
Plasmodium knowlesi is, somewhat controversially, sometimes considered the ‘fifth human malaria parasite’. The controversy stems from the fact that it is unknown whether this parasite can be transmitted from human to human. If human infection is only the result of ‘spill-over’ from transmission between monkeys, then it must be considered a zoonotic disease rather than a human one. The first step in establishing whether infections of P. knowlesi in humans are infectious to mosquitoes is to determine whether such infections are capable of producing gametocytes. Previous authors have reported the presence of P. knowlesi gametocytes, identified by microscopy, in blood from infected patients (Singh et al. Reference Singh, Bobogare, Cox-Singh, Snounou, Abdullah and Rahman1999; Lee et al. Reference Lee, Cox-Singh and Singh2009; Ta et al. Reference Ta, Salas, Ali-Tammam, Martínez Mdel, Lanza, Arroyo and Rubio2010). We were able to detect transcripts of pks25 in dried blood spots on filter papers taken from P. knowlesi infected individuals in Khanh Phu, indicating that P. knowlesi infected humans in this region may very well be infectious to mosquitoes. We were unable, however, to observe either gametocytes or the asexual forms of P. knowlesi by microscopy. The detection limit for P. knowlesi gametocytes by nested RT–PCR was as low as 1 gametocyte µL−1, comparable with that for P. falciparum using pfg377 [1 gametocyte µL−1 by conventional RT–PCR (Maeno et al. Reference Maeno, Nakazawa, Dao, Yamamoto, Giang, Hanh, Thuan and Taniguchi2008; Nakazawa et al. Reference Nakazawa, Marchand, Quang, Culleton, Manh and Maeno2009)]. All 15 transcripts of P. knowlesi Pks25 were identical at the nucleic acid level, in contrast to those of Pfg377 and Pvs25. This could be due to either limited genetic diversity amongst P. knowlesi parasites circulating in this region, supported by the fact that P. knowlesi 18S rRNA sequences from these samples were also non-polymorphic, or to a lack of polymorphism in this particular sequence between strains.
In summary, molecular typing of gametocyte-specific mRNA from malaria parasite-infected individuals in the Khanh Phu region of Vietnam revealed that 37, 61 and 47% of P. falciparum, P. vivax and P. knowlesi infections produce gametocytes detectable by RT–PCR. The same allelic types at similar frequencies were observed in both humans and mosquitoes inhabiting the forest area.
SUPPLEMENTARY MATERIAL
The supplementary material for this article can be found at https://doi.org/10.1017/S0031182016002110.
ACKNOWLEDGEMENTS
We thank Nguyen Son Hai, Phan Chau Do, Dang Duy Vu and Nguyen Le Dung for their technical support for the field collections. We also thank the Provincial Health Service and Malaria Control Centre of Khanh Hoa Province as well as the authorities of Khanh Vinh district and Khanh Phu municipality for their administrative support.
FINANCIAL SUPPORT
This study was supported in part by the following four grants: (i) the Japan Society for the Promotion of Science (JSPS) Asia Africa Science Platform Program; (ii) JSPS KAKENHI Grant Number 23406022 and 2630029; (iii) the Joint Usage/Research Center on Tropical Disease, Institute of Tropical Medicine, Nagasaki University (2011-Ippan-7, 2012-Ippan-5); and (iv) a Grant-in-Aid for Scientific Research from Fujita Health University. The Khanh Phu Malaria Research team and the field research costs were financially supported through private donations to the Medical Committee Netherlands-Vietnam (MCNV).
