Optimized Pan-species and Speciation Duplex Real-time PCR Assays for Plasmodium Parasites Detection in Malaria Vectors

Background An accurate method for detecting malaria parasites in the mosquito’s vector remains an essential component in the vector control. The Enzyme linked immunosorbent assay specific for circumsporozoite protein (ELISA-CSP) is the gold standard method for the detection of malaria parasites in the vector even if it presents some limitations. Here, we optimized multiplex real-time PCR assays to accurately detect minor populations in mixed infection with multiple Plasmodium species in the African malaria vectors Anopheles gambiae and Anopheles funestus. Methods Complementary TaqMan-based real-time PCR assays that detect Plasmodium species using specific primers and probes were first evaluated on artificial mixtures of different targets inserted in plasmid constructs. The assays were further validated in comparison with the ELISA-CSP on 200 field caught Anopheles gambiae and Anopheles funestus mosquitoes collected in two localities in southern Benin. Results The validation of the duplex real-time PCR assays on the plasmid mixtures demonstrated robust specificity and sensitivity for detecting distinct targets. Using a panel of mosquito specimen, the real-time PCR showed a relatively high sensitivity (88.6%) and specificity (98%), compared to ELISA-CSP as the referent standard. The agreement between both methods was “excellent” (κ = 0.8, P<0.05). The relative quantification of Plasmodium DNA between the two Anopheles species analyzed showed no significant difference (P = 0, 2). All infected mosquito samples contained Plasmodium falciparum DNA and mixed infections with P. malariae and/or P. ovale were observed in 18.6% and 13.6% of An. gambiae and An. funestus respectively. Plasmodium vivax was found in none of the mosquito samples analyzed. Conclusion This study presents an optimized method for detecting the four Plasmodium species in the African malaria vectors. The study highlights substantial discordance with traditional ELISA-CSP pointing out the utility of employing an accurate molecular diagnostic tool for detecting malaria parasites in field mosquito populations.


Introduction
Malaria remains the most prevalent parasitic disease worldwide. In 2010, an estimated 216 million malaria episodes with an estimated 655,000 deaths were reported of which more than 90% occurred in Africa [1]. Five species of the malaria parasite cause human disease. This includes Plasmodium falciparum, Plasmodium malariae, Plasmodium ovale, Plasmodium vivax, and Plasmodium knowlesi, which is gaining widespread recognition as a human pathogen [2]. The transmission of these malaria-causing parasites to humans is exclusively caused by Anopheles mosquitoes of which five species (An. gambiae s.s., An. funestus, An. arabiensis, An. moucheti and An. nili) have been identified as the major malaria vectors in Africa. In southern Benin, a western African country, An. gambiae s.s. and An. funestus are the main Plasmodium vectors; An. funestus being responsible for the prolonged period of malaria transmission during the dry season [3].
Malaria in Benin is still of primary health concern among children under five and pregnant women, and motivates up to 40% of outpatient visits and 30% of hospitalizations [4]. The Malaria Control Strategy currently recommended by the WHO [5] relies on the use of the artemisinin-based combination therapy (ACT), intermittent preventive treatment during pregnancy (IPTp) and the universal distribution of Long Lasting Insecticidal Nets (LLINs). The search for an effective malaria vaccine as a supplement to the disease control strategy, remains a major aspect that holds much hope [6]. However, the success of such a vaccine, whose efforts are currently focused on P. falciparum malaria, raises the question of the management of mixed infections by multiple species of Plasmodium spp. [7].
In malaria patients, mixed species infections are common and generally under reported. A cohort study conducted on 764 children in southern Benin (Tori-Bossito) using microscopy as diagnostic tool showed the predominance of P. falciparum in the analyzed samples (91%), with co-infections rates involving P. malariae and P. ovale of 3% and 2%, respectively. Different patterns of mixed infections (P. falciparum/P. malariae, P. falciparum/P. ovale and P. falciparum/P. ovale/P. malariae) were reported in the proportions of 1.17%, 2.35%, and 0.48%, respectively [8]. As the operating characteristics of microscopy in many malaria endemic settings are known to be poor, substantial proportions of mixed-species infections can frequently be missed even by welltrained microscopists. This justifies the need for reliable alternative tool for the accurate diagnosis of malaria infection [9,10]. In mosquito vectors, the infectious status is usually assessed by the presence/absence of Plasmodium sporozoites in the salivary glands. This was initially achieved by microscopic assessment of glands after the mosquito dissection. But this technique is time consuming and requires skilled staff and does not allow identification of sibling Plasmodium species in the sample. The development of more rapid immunological and molecular approaches such as the circumsporozoite protein enzyme linked immune-sorbent assay (ELISA-CSP) [11,12] and PCR-based techniques rapidly got widely adopted, [13,14]. Although ELISA-CSP seems to be relatively robust and cheap, there are potential drawbacks in using this approach. A lack of specificity has been raised as an important issue because this method does not only detect the sporozoites in the salivary glands, but can also detect CSP from other mosquito tissues [14]. An overestimation of true salivary gland infection could also result from measuring circulating CSP as this could originate from sporozoites migrating through the mosquito [15,16]. Moreover, the ELISA-CSP technique is also subjected to an underestimation of the vector's actual level of infection because it does not target all infecting Plasmodium spp in a given mosquito species [17].
In the context of the deployment of global effort towards malaria control and elimination, it is of primary importance to develop sensitive and reliable diagnostic techniques for detecting Plasmodium spp in both humans and mosquitoes. Recently, highthroughput assays based on real-time PCR have been developed for detecting malaria parasites in humans. These methods allow some rapid and simultaneous detection, and a quantification of several target DNAs through the use of the specific fluorophorelabeled probes [7,18,19]. The benefits of these methods come from very low contamination risks and high sensitivity that reaches 100 fold over the ELISA technique [14]. The use of more sensitive and effective diagnostic technique for the detection of parasites in the vectors can ensure better estimation of transmission intensity in different malaria settings.
The aim of this study was to optimize a sensitive PCR-based method that can accurately estimate mixed infection rates of Plasmodium species in An. gambiae and An. funestus, the main malaria vectors in Africa.

Study Area
The Anopheles gambiae mosquitoes tested in this study were collected by the team of the Centre de Recherches Entomologiques de Cotonou (CREC) Research under the framework of the President's Malaria Initiative (PMI) program of the USAID. Mosquitoes were collected in five districts (Adjarra, Adjohoun, Dangbo, Missérété, and Sèmè) in the Ouémé department (6u349711E -2u319358N) in Southern Benin.
The temperatures in these areas vary between 25uC and 30uC with an annual rainfall ranging from 900 mm to 1500 mm.

Mosquito Collection and Sample Processing
Indoor and outdoor mosquito collections were conducted in two sites per village using the human landing catch technique (HLC). Collectors were hourly rotated along collection sites and/or position (indoor/outdoor). At each position, all mosquitoes caught were kept in individual tubes and in hourly bags. The collection period took place at the night between 21:00 and 05:00 AM.
Mosquitoes were also captured by using window traps placed in different houses in each village. The houses were selected according to the number of the people sleeping there. Traps were placed on the outside windows in each selected house from 6 PM up to 6 AM. Mosquitoes were then transferred in the cups, using a vacuum for the identification of anopheline species.

Identification of Sibling Species and Infection Rates
All collected mosquitoes were first identified through morphological identification keys [20,21,22]. Female mosquitoes identified as An. gambiae sensu lato (Diptera: Culicidae) and An.funestus group were taken to CREC laboratory and stored at 220uC in Eppendorf tubes with silica gel for subsequent analyses. Heads and thoraces of An. funestus and An. gambiae s.l. were processed for detection of P. falciparum circumsporozoite protein (CSP) using ELISA technique as described [11,12]. Abdomen and legs were used for DNA extraction destined to molecular identification of sibling species using polymerase chain reaction (PCR) as described previously [23,24].

Plasmodium Genomic DNA Samples, Plasmid Clones and DNA Standards
Mosquito's homogenates of the head-thorax obtained from the preparation meant for ELISA-CSP (100 Anopheles gambiae and 100 Anopheles funestus) and stored at 220uC was later used for DNA extraction. Genomic DNA was extracted from the homogenates using the DNeasyH Blood & Tissue kit (Qiagen) as recommended by the manufacturer. The DNA was eluted in 100 mL and stored at 220uC. Plasmodium genomic DNAs of P. vivax, P. malariae or P. ovale and plasmids containing insert of the 18S gene of each of those species were kindly provided by Dr Stephanie Yanow at the Provincial Laboratory for Public Health, Edmonton, Alberta, Canada. For P.falciparum the 18S gene was amplified from 3D7 gDNA (MR4) using outer primers of the Nested PCR established by Snounou et al. [14,25], and cloned into the pGEM-T vector (Promega). The insert quality was verified by sequencing. In plasmid-mixing experiments where 1.10 2 , 1.10 5 , and 1.10 7 copies of one plasmid were mixed with similar copy numbers of the second plasmid, or 1.10 2 copies of one plasmid were mixed with 1.10 3 , 1.10 4 , and 1.10 5 copy numbers of the second plasmid and used as the template for the real-time PCR. Cycle threshold (CT) values were based on duplicate samples. Plasmid copy number quantification was performed by the spectrophotometric analysis. For normalization purpose, specific primers were selected and the mosquito RS7 (ribosomal protein S7) gene was amplified from gDNA of Anopheles gambiae ss. The purified PCR product (396 bp) was quantified by spectrophotometric analysis and used in serial dilution standard.

Real-time PCR Assay for the Detection of Plasmodium spp in Mosquitoes
Genus-specific and species-specific primers and probes for the gene encoding the small subunit (18S) of Plasmodium rRNA as reported by Shokoples et al [7] with modification reported by Diallo et al [26] (Table 1 shows all oligonucleotide sequences used).
(i) Monoplex real-time PCR. A mosquito housekeeping gene (ribosomal protein S7) was amplified as an internal control to ensure that the DNA from the sample was successfully extracted and to later allow normalization when comparing different samples. PCRs were run in a final volume of 20 ml, consisting of 2 ml of DNA, 10 ml of SensiMix DNA kit (Quantace), and 300 nM of each primer. The protocol described by Dana et al. [27] allowed systematic and efficient amplification of the S7 gene in both mosquito. Reactions were run on a Rotor-Gene 6000 TM (Corbett Research) using the cycling conditions of: 10 minutes at 95uC followed by 40 cycles of 95uC for 15 seconds and 60uC for 30 seconds and 72uC for 30 seconds. All amplification products were quantified using the standard range achieved with the S7 gDNA purified PCR product.
(ii) Duplex real-time PCR assays for Plasmodium Spp detection. Each sample was analyzed in two separate reaction tubes, containing either the genus-specific and P. falciparum primers and probes detection system (Plasmo/Pf) as described by Diallo et al [26] or the P. malariae and P. ovale primers/probes system (Pm/Po). Each reaction mixture contained 5 ml of DNA, 10 ml of PerfeCTa qPCR FastMix, UNG (Quanta Biosciences), 300 nM of each primer, and 100 nM of each probe in a final volume of 20 ml. Reactions underwent 40 cycles under conditions (95uC for 5 s, 60uC for 1 min). As P. vivax is traditionally believed to be virtually absent in West and Central Africa, the search for P. vivax was achieved from pooled samples. Samples were pooled into groups of 10 samples with the same amount of S7 gDNA. Five microliters of each of the pooled samples were amplified in 20 ml reaction mixtures under the same condition described above.
In order to compare parasite densities between individual samples, relative ratio was calculated by dividing the amount of Plasmodium DNA obtained by absolute quantification by the amount of the housekeeping DNA (S7) determined in the same sample.

Statistical Analysis
The Cohen's kappa coefficient (k) was used to measure interrater agreement between the referent ELISA-CSP and the novel real-time PCR [28]. Categorical variables were compared using Fisher's exact test, while continuous variables were compared by the Kruskal-Wallis test. Differences were considered statistically significant when p-values were less than 0.05.

Ethical Statements
This study was approved by the National Research Ethics Committee of Benin and the Center for Entomological Research of Cotonou (IRB00006860). All necessary permits were obtained for the described field studies. No mosquito collection was done without the approval of the head of the village, the owner and occupants of the collection house. Mosquito collectors gave their written informed consent and were treated free of charge for malaria presumed illness throughout the study.

Mosquito Collection and Plasmodium Infection rates by ELISA-CSP
A total of 1.830 An. gambiae s.l. and 1.234 members of the An. funestus group were collected in the field and analyzed by ELISA-CSP. Fifty An. gambiae s.s. and twenty An. funestus mosquitoes were found positive to P. falciparum, corresponding to a prevalence of infection of 2.7% and 1.62% respectively.
In this study, 200 hundred specimens (100 An. gambiae s.s. and 100 An. funestus) were used in the evaluation of the real-time PCR assays. For An. gambiae, 50 Plasmodium infected and 50 randomly selected uninfected mosquitoes were included according to ELISA-CSP. Due to lower prevalence of P. falciparum infection among An. funestus mosquitoes, only 20 infected mosquitoes and 80 randomly selected negative were included.

Design and Validation of the Real-time PCR Assays
The quality of the DNA extracted was checked by spectrophotometric analysis at 260 nm and 280 nm respectively, and the absence of PCR inhibitors in the preparation was confirmed by amplification of the ribosomal protein S7 in all samples; means Ct values were 24.7262.62 for An. gambiae and 24.8461.5 for An. funestus.
As part of the study, we first evaluated each detection system on a 10-fold dilution range of the corresponding standard (from 1.10 0 to 1.10 10 copies). However, the optimal linear range of the external standard curves was observed from 1.10 1 to 1.10 10 copies with PCR efficiencies all above 90%. Amplification of non-specific targets guided the combination of duplex assays. Evaluation of duplex assays for the simultaneous detection of Plasmodium spp in the same reaction was performed in plasmid mixing experiments. Results presented in Table 2 and figure 1 highlight substantial specificity and a lack of competition between the mixed oligonucleotides.

Prevalence and Co-infection Rates with Plasmodium spp in Mosquitoes
The infection rates with Plasmodium spp are shown in Figure 2. The speciation by real-time PCR of the 43 PCR positive An. gambiae s.s. and 22 PCR positive An. funestus revealed the presence of P. falciparum in all samples (100%). While mono infections with only P. falciparum were found in 81.4% and 86.4% of infected An. gambiae and An. funestus respectively, mixed infections with multiple Plasmodium species were detected in 18.6% and 13.6% of the

Absolute and Relative Quantification of Plasmodium spp DNA in Mosquitoes
Absolute quantification of all positives specimen was done using the standard curve generated on Plasmodium falciparum-specific18S DNA plasmid dilutions amplified in each run with the Plasmo/Pf detection system. The standard curve was generated on the 10-fold dilution of Plasmodium falciparum-specific plasmid range (1.10 1 to 1.10 6 copies in 5 mL reaction). Samples in which the target was detected at a late Ct value beyond the linearity range (36,Ct,40) were considered positive but not quantifiable. Three ranges (1.10 1 , 1.10 3 and 1.10 6 ) of P. malaria/P. ovale plasmid mixture were amplified consistently with the detection system Po/Pm in each  The PCR amplification of ribosomal protein S7 gene (E = 99%) allowed the estimation of the amount of mosquito DNA in each reaction. Normalization of the amount of Plasmodium DNA on the amount of mosquito DNA showed no difference in parasite load observed between the infected An. gambiae and An. funestus studied (Kruskall-Wallis test, P = 0.2197) (Figure 3).

Discussion
In the context of malaria elimination (eradication) policy, it is essential to develop reliable diagnostic techniques for detecting Plasmodium spp infections in humans and in the vector. The aim of this study was to optimize a high-throughput sensitive and specific real-time PCR assay to detect and quantify Plasmodium infections in malaria vectors in Benin. Here, we used the same region of DNA encoding the small subunit of the18S rRNA to redefine the optimal multiplex PCR assays based on allele-specific primers/ probe systems previously reported by Shokoples et al. [7]. Minor sequence modifications and labeling were made on the probes used by Shokoples et al. [7] to increase specificity and adapt the method to a duplex-based detection system [26]. The genes encoding the small subunit of the18S rRNA exist in several copies (7) in the Plasmodium genome. One of the major advantages of the method previously reported by Shokoples et al. [7] over other approaches [13,18,19], is that the primers designed target all copies thus increasing the sensitivity of the reaction. The use of species-specific oligonucleotides that could accurately detect all four malaria-causing Plasmodium species (Pf, Pm, Po and Pv) without significant competition between the oligonucleotides designed for the different templates was one of the major advantages of this approach. Here, the multiplexing of the reaction was optimized for the simultaneous detection of the four Plasmodium species at a time in two reaction tubes. This method was tested on plasmid preparations and showed good amplification efficiencies (E.90%). We also noticed a good sensitivity with the ability of detecting and quantifying down to 10 copies of Plasmodium 18S rDNA in 5 mL DNA used per reaction, meaning that at least 200 copies, approximately 30 sporozoites, are necessary in DNA preparation for a positive reaction to be quantified. Targets copy number detected below this threshold were still considered positive but unquantifiable as this falls outside the linearity range of external standards. The specificity of the real-time PCR was demonstrated by the absence of cross-reactivity between different primer-probe systems on artificial mixtures of plasmid preparations. The analytical sensitivity of the assays for P. malariae, P. ovale and P. vivax as the minor species in cases of mixed infection with P. falciparum showed, like in the data reported by Shokoples et al. [7] that we could reproducibly detect minors populations at a greater fold down to 1:1000 ratio. This performance was optimized by the formulation of the multiplexing that we have defined (Plasmo/Pf and Pm/Po). With these modifications, we implemented this assay as a confirmatory test for malaria species identification in anopheline vectors (An. gambiae and An. funestus). Plasmodium DNA was consistently amplified from frozen mosquito homogenates initially prepared for ELISA. This suggests that parasite target DNA will likely remain detectable by PCR in mosquito homogenates for longer periods, from the time they are stored at 220uC. This preservation condition of fieldcollections for subsequent target PCR-detection of Plasmodium DNA is highly amenable to field work and does not seem to promote biological degradation processes which are favored by the release of nucleases after grinding. In comparison with traditional ELISA-CSP; the real-time PCR assay was more useful for the identification of Plasmodium species in the vectors. From the 70 positive mosquitoes for P. falciparum by ELISA-CSP the presence of Plasmodium could be PCR-confirmed in 62 samples. Of important diagnostic significance, 11 samples were misdiagnosed by ELISA-CSP. Among these, 8 samples that were positive by ELISA-CSP were not confirmed by real-time PCR. The absence of Plasmodium DNA was further ascertained in those samples by using the conventional nested PCR described by Snounou et al [14]. These results are concordant with the hypothesis that ELISA-CSP may be compromised by overdiagnosis and poor specificity due to circulating antigens originating from the rupture of oocysts [15] or from other non-malaria antigens [29]. This phenomenon has been highlighted in other studies when testing the plasma fractions of pig and bovine blood in Thailand [30]. In Senegal, false positive results were also associated with bovine and/or sheep blood meals when testing An. gambiae s.l. for the presence of P. malariae and P. ovale [31]. In most studies where false positive ELISA results were reported in mosquitoes [31,32,33,34], only head and thoraces were used for ELISA assays hence excluding the possible contamination with oocyst-sporozoites present in the abdomen.
In this study, the real-time PCR identified 3 new cases of positive mosquitoes that have been missed by the ELISA-CSP. The apparent discrepancy between the two diagnostic methods on the positives samples may be due to the detection limit of ELISA-CSP. Actually ELISA-CSP has a detection limit of 250 sporozoites in 50 ml [35], while the detection limit for PCR in salivary glands was previously estimated at 10 sporozoites [13], this being lower with the real-time PCR used in this study.
The analysis of all Plasmodium positive samples of An. gambiae s.s. and An. funestus by real-time PCR revealed the presence of P. falciparum in all the samples. This is not surprising because P. falciparum is the most prevalent malaria parasite in west and central Africa. The relative quantification applied to normalize the copy number of Plasmodium target detected to that of the amount of mosquito DNA showed that parasite densities could vary between individual mosquitoes but were on average similar between infected specimens of either An. gambiae or An. funestus. These findings confirm the trend that An. funestus is an important vector of malaria parasites in Benin [3].
Following the high specificity of real-time PCR technique demonstrated in target mixing experiments, the assay allowed us to identify other Plasmodium species alongside the dominant species Plasmodium falciparum. These species occurring mainly as minor populations in cases of coinfection. The rate of mixed infections of 18.6% and 13.6% in An. gambiae and An. funestus respectively appeared greater than what is described in human populations living in southern Benin. Actually the proportion of co-infection with P. ovale and P. malariae in humans was only estimated to 5% [8]. Though this difference is potentially due to the methods used in the two studies (microscopy and real-time PCR), a recent analyses of human specimen in Southern Benin with the same real-time PCR technique revealed less frequency of co-infections rate than what we observed in mosquitoes (Tuikue Ndam et al., Unpublished data). This suggests the existence of other factors, including the immunity that may affect the diversity of Plasmodium infection in humans and mosquito. Plasmodium vivax was detected by real-time PCR in none of the specimens studied. This was not surprising as this species is known to be very rare in west and central Africa.

Conclusion
This study reports the analytical validation of new real-time PCR assays for the detection and identification of Plasmodium species in the mosquito hosts. Combined with efficient DNA extraction methods, the assay has demonstrated a good analytical sensitivity in detecting mixed infections with distinct malariacausing Plasmodium among the two main malaria vectors in Benin. The results suggest that the method described here is appropriate for the detection of malaria parasites in field-collected mosquito specimen. The application of this highly specific multiplex realtime PCR assay in larger and prospective studies will allow a better appreciation of the intensity of transmission and the knowledge of the movement of different Plasmodium species in the malaria vectors in Africa. This technique could facilitate the implementation of more basic research to address the fitness cost associated with Plasmodium spp infections on several life history traits including survival, behavioral and reproductive capacity of mosquito vectors.