We developed a single-step real-time fluorescence resonance energy transfer (FRET) multiplex polymerase chain reaction (PCR) merged with melting curve analysis for the detection of Wuchereria bancrofti and Brugia malayi DNA in blood-fed mosquitoes. Real-time FRET multiplex PCR is based on fluorescence melting curve analysis of a hybrid of amplicons generated from two families of repeated DNA elements: the 188 bp SspI repeated sequence, specific to W. bancrofti, and the 153-bp HhaI repeated sequence, specific to the genus Brugia and two pairs of specific fluorophore-labeled probes. Both W. bancrofti and B. malayi can be differentially detected in infected vectors by this process through their different fluorescence channel and melting temperatures. The assay could distinguish both human filarial DNAs in infected vectors from the DNAs of Dirofilaria immitis- and Plasmodium falciparum-infected human red blood cells and noninfected mosquitoes and human leukocytes. The technique showed 100% sensitivity and specificity and offers a rapid and reliable procedure for differentially identifying lymphatic filariasis. The introduced real-time FRET multiplex PCR can reduce labor time and reagent costs and is not prone to carry over contamination. The test can be used to screen mosquito vectors in endemic areas and therefore should be a useful diagnostic tool for the evaluation of infection rate of the mosquito populations and for xenomonitoring in the community after eradication programs such as the Global Program to Eliminate Lymphatic Filariasis.
Human lymphatic filariasis, a mosquito-borne disease, is caused by the nematodes Wuchereria bancrofti, Brugia malayi, and B. timori (WHO 1992, Weil and Ramzy 2007). It is a major public health problem, particularly in the tropics and subtropics. A principal intention of the Global Program to Eliminate Lymphatic Filariasis (GPELF) is the interruption of filarial transmission (Weil and Ramzy 2007). One important step to achieve this is the identification and appraisal of lymphatic filariasis in endemic communities.
The microscopic detection of filarial larvae in vectors based on dissection has been the gold standard for estimating infection levels in mosquitoes, but it is laborious, tedious, and time consuming and carries a low sensitivity and a need for specially trained microscopists. The method also needs dissection of a lot of mosquitoes caught from the field and the use of morphological criteria to identify these parasites. It cannot differentiate between filarial larvae that infect human and animals, which are sometimes harbored by the same vector. A diagnostic method that is sensitive, rapid, species specific, and provides a high throughput for filarial detection would be an essential tool in the evaluation of disease prevalence and monitoring elimination programs.
A number of conventional polymerase chain reaction (c-PCR) assays (McCarthy et al. 1996, Ramzy et al. 1997, Williams et al. 2002) and a single-step multiplex PCR (Mishra et al. 2005, 2007) have been developed and have been shown to have superior sensitivity to detect filarial DNA in human blood and mosquito vectors. However, these procedures need gel electrophoresis for analyzing the results, which is slow, has a limited throughput, and is prone to carryover contamination. Recently, W. bancrofti and Brugia spp. DNA have been shown in infected blood and in infected mosquito vectors by either a Taqman probe or an Eclipse minor groove binding probe based on real-time PCR (Rao et al. 2006a, b, Fischer et al. 2007). In addition, these filarial larvae have been detected in infected mosquitoes by another assay principle using real-time fluorescence resonance energy transfer (FRET) PCR combined with melting curve analysis (Lulitanond et al. 2004, Thanchomnang et al. 2008). This effective real-time FRET PCR is not only accurate, sensitive, fast, and can quantify specific DNA in biological samples, but it also differentiates the DNA of W. bancrofti (Lulitanond et al. 2004) or B. malayi species (Thanchomnang et al. 2008) from that of other parasites and humans, as well as the vectors, by melting curve analysis. Moreover, this method provides a high throughput because it does not need agarose gel electrophoresis for analysis of the amplicons and has a broad dynamic range.
The real-time PCR-based method for filarial detection used previously could detect only one infection at a time by targeting either W. bancrofti (Lulitanond et al. 2004, Rao et al. 2006a) or B. malayi (Rao et al. 2006b, Fischer et al. 2007, Thanchomnang et al. 2008). The separate diagnostic procedures for bancroftian and brugian infections are expensive and take time. In this study, we used a simple real-time FRET multiplex PCR for the rapid detection of W. bancrofti and B. malayi in mosquito samples at the same time. The method used two pairs of specific primers amplifying the repetitive sequence SspI repeat of W. bancrofti (GenBank accession number L20344) (Zhong et al. 1996) and the HhaI repeat of B. malayi (GenBank accession number M12691) (McReynolds et al. 1986). This developed real-time FRET PCR using a LightCycler-based PCR system (Roche Applied Science, Mannheim, Germany) was sensitive and genus specific for the differential detection of larval stages of W. bancrofti and B. malayi in mosquitoes. Two pairs of fluorophore-labeled specific hybridization probes were used, and their melting point profiles were compared with those of control DNA. The importance of this method is the simultaneous detection and differentiation of W. bancrofti and B. malayi, especially in countries such as India and Thailand where both parasites overlap incidence.
Materials and Methods
Culex quinquefasciatus, a mosquito species in the urban area of Khon Kaen Province, northeastern Thailand, were artificially infected with W. bancrofti. Aedes togoi, a mosquito species from Koh Nom Sao, Chanthaburi Province, Eastern Thailand (Choochote et al. 1987), was similarly infected with B. malayi. The mosquito larvae were taken from their breeding places and reared in an insectarium.
Infection of Mosquitoes.
Blood containing the nocturnally periodic W. bancrofti microfilariae (density = 44 microfilariae/20 μl) was collected from a male carrier, a Burmese immigrant, at Mae Sot District, Tak Province, northwestern Thailand, whereas blood infected with nocturnally subperiodic B. malayi microfilariae (density = 160 microfilariae/20 μl) was obtained from an infected cat. The worms originated from a Thai woman in Bang Paw District, Narathiwat Province, southern Thailand, and were used to experimentally infect domestic cats. The cats are now kept in the Department of Parasitology, Faculty of Medicine, Chiang Mai University. The Cx. quinquefasciatus and Ae. togoi mosquitoes were obtained as previously described (Lulitanond et al. 2004, Thanchomnang et al. 2008). The mosquitoes were allowed to feed on heparinized blood (both B. malayi- and W. bancrofti-positive blood samples) using an artificial membrane feeding technique. Fourteen days after feeding, the mosquitoes were dissected in normal saline solution, and the number of larvae was counted under a dissecting microscope. Only W. bancrofti- (n = 30) or B. malayi- (n = 30) infected mosquitoes were used for the experiments, and the number of parasites was recorded. After the dissection of an infected mosquito, all filarial larvae and the mosquito body were mixed and placed in a 1.5-ml microcentrifuge tube, labeled, and kept at -20°C for DNA extraction. The range and mean numbers ± SD of W. bancrofti larvae per infected mosquito were 1–18 and 5.36 ± 5.16 larvae, respectively, whereas the numbers of B. malayi per infected mosquito were 1-42 and 7.06 ± 10.35 larvae, respectively. Another set of third-stage larvae (L3) of W. bancrofti or B. malayi dissected from infected mosquitoes were confirmed by microscopic examination and were counted, separately collected, and kept at -20°C for further use.
The study protocol was approved by the Khon Kaen University Ethics Committee of Human Research and the maintenance and care of animals in this study complied with the current Thai laws.
Preparation of Specimens for Real-time FRET Multiplex PCR.
Each specimen, infected and noninfected mosquitoes or artificially inoculated L3 of filarial worms in noninfected mosquitoes, was put in a 1.5-ml microcentrifuge tube, homogenized with disposable polypropylene pestles (Bellco Glass, Vineland, NJ), and extracted using the Nucleospin Tissue kit (Macherey-Nagel, Duren, Germany). The DNAs were eluted in 100 μl of 5 mM Tris-HCl, pH 8.5, of which 5 μl was used in the reaction. For specificity evaluation, Dirofilaria immitis adults from infected dogs (from Khon Kaen Province), noninfected Ae. togoi and Cx. quinquefasciatus, Plasmodium falciparum-infected human red blood cells, and human leukocytes were separately extracted, purified, and eluted in a similar way.
Real-time FRET Multiplex PCR Assay.
The LightCycler PCR and detection system (LightCycler 2.0; Roche Applied Science) was used for amplification and quantification. The reaction was performed in glass capillaries.
For W. bancrofti detection, specific primers, WB-F (5′-CGT GAT GGC ATC AAA GTA G-3′) and WB-R (5′-CCC TCA CTT ACC ATA AGA CA-3′) (Sigma-Proligo, Singapore), and specific probes, one labeled at the 5′ end with the LightCycler Red 640 fluorophore (5′-Red 640-GAA AAT ATT AAA AAA ACA ATT CCC TTA C-phosphate-3′; WBLC640 probe) and the other labeled at the 3′ end with fluorescein (5′-AAA ATT GCT AAA AAT TCC ATT CAT ACT-Fluo-3′; WBFL530 probe) (Tib Molbiol, Berlin, Germany) were used. The primers and probes were designed to bind to the SspI repetitive sequence of the W. bancrofti genome as described before (GenBank accession number L20344) (Zhong et al. 1996).
For B. malayi detection, the specific primer pair, BM-F (5′-TCA TTA GAC AAG GAT ATT GGT TC-3′) and BM-R (5′-TTT AAA CTA TAA AAT GAC AAC ACA-3′) (Sigma-Proligo), as well as a pair of adjacent oligoprobes: one labeled at the 5′ end with the LightCycler Red 705 fluorophore (5′-Red 705-TGT ACC AGT GCT GGT CGT GTA-Phosphate-3′; BMLC705 probe) and the other at the 3′ end with 530 fluorescein (5′-AAA TTA ATT GAC TAT GTT ACG TGA A-Flou 530–3′; BMFL530 probe) (Tib Molbiol). The primers and probes were designed to bind to the HhaI repetitive sequence of the B. malayi genome as previously described (GenBank accession number M12691) (McReynolds et al. 1986). All primers and probes were designed by the LC probe design software (Roche Applied Science). The schematic diagram of the hybridization analysis used in the test is shown in Fig. 1.
The LightCycler FastStart DNA Master HybProbe Kit (Roche Applied Science) was used for amplification detection according to the kit’s protocol. In short, a pair of adjacent oligoprobes was hybridized to an internal genus-specific repetitive sequence of homologous parasite DNA (W. bancrofti or B. malayi). When the probes were hybridized to the same DNA strand internal to the PCR primers, the probes came in close proximity and produced a FRET. During FRET, the 530 fluorescein was excited by the light source of the LightCycler instrument. The excitation energy was transferred to the acceptor fluorophore of the LightCycler Red 640 (for W. bancrofti detection) or 705 (for B. malayi detection) only when it was positioned in close vicinity to the former and the emitted fluorescence was measured after annealing by the instrumental photohybrids. After a complete PCR reaction, it was possible to construct a melting point analysis, during which the temperature was lowered below the annealing temperature of the probes and increased slowly. The fluorescence signal decreased when the probe melted off its target.
The PCR mixture contained LightCycler Faststart DNA Master HybProbe (Roche Applied Science), 3 mM MgCl2, 0.5 μM WB-F primer, 0.5 μM WB-R primer, 0.4 μM WBLC640 probe, and 0.2 μM WBFL530 probe, as well as 0.5 μM BM-F primer, 0.5 μM BM-R primer, 0.4 μM BMLC705 probe, and 0.2 μM BMFL530 probe. The total reaction volume was 20 μl. Samples were run by performing 45 cycles of repeated denaturation (5 s at 95°C), annealing (15 s at 45°C), and extension (8 s at 72°C). The temperature transition rate was 20°C/s. After amplification, a melting curve was produced by heating the product at 20°C/s to 95°C, cooling it to 40°C, keeping it at 40°C for 30 s, and heating it slowly at 0.1°C/s to 85°C. The fluorescence intensity change was measured throughout the slow heating phase. To determine the specificity of the oligonucleotide hybridization based on the FRET technique, DNA extracted from D. immitis, human leukocytes, noninfected Ae. togoi and Cx. quinquefasciatus, and P. falciparum-infected human red blood cells were separately analyzed. Each run contained at least one negative control consisting of 5 μl distilled water.
For improved visualization of the melting temperatures (Tm), melting peaks were derived as previously described (Lulitanond et al. 2004). Melting curves were applied to determine the specific amplified products, which were verified by conventional agarose gel electrophoresis.
Positive Control Plasmids.
A positive control plasmid of W. bancrofti was constructed by cloning a PCR product of the W. bancrofti SspI repeat (Zhong et al. 1996) into the pGEM-T (Promega, Madison, WI) vector according to the manufacturer’s instructions. The PCR products were obtained by c-PCR using the primers WB-F and WB-R. A positive control plasmid of B. malayi was constructed by cloning a PCR product of the B. malayi HhaI repeat (McReynolds et al. 1986) into the pCR4-TOPO vector (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. The PCR products were obtained by c-PCR using the primers BM-F and BM-R. The plasmid was propagated in Escherichia coli, the nucleotide sequence of the inserted gene was sequenced in both directions, and the resulting sequences were matched using the multalin program (http://prodes.toulouse.inra.fr/multalin/multalin.html), which showed a homologous structure to the gene sequence from which the primers were designed.
Standardization of the Real-time FRET Multiplex PCR.
Five microliters of serial dilutions of either W. bancrofti-positive control (6 ×103-6 × 1010 copies) or of B. malayi-positive control plasmids (3 × 103-3 × 1010 copies) in water were used to assess the sensitivity of the real-time multiplex FRET PCR. At a cut-off limit for detection of 35 PCR cycle numbers, the limit of detection for the SspI repeat target DNA sequence was ≈6 × 104 copies of positive control (Fig. 2A), whereas for the HhaI repeat target DNA sequence, the limit was ≈3 × 104 copies of positive control plasmid (Fig. 2B). No fluorescence indicator was shown when purified DNA from D. immitis, noninfected Ae. togoi and Cx. quinquefasciatus, P. falciparum-infected human red blood cells, or human leukocytes was used.
To determine the capability of detection of the real-time FRET multiplex PCR, pools of 5, 10, and 30 noninfected Cx. quinquefasciatus adult mosquitoes each inoculated with one W. bancrofti L3 as well as pools of 5, 10, and 30 noninfected Ae. togoi adult mosquitoes each inoculated with one B. malayi L3 were used. These samples underwent DNA extraction and were examined by the assay. The assay could ascertain the filarial DNA presence of as little as one W. bancrofti L3 inoculated in a pool of 30 noninfected Cx. quinquefasciatus and one B. malayi L3 inoculated in a pool of 30 noninfected Ae. togoi (data not shown).
To observe the assay’s capability of detecting mixed infections of Brugia and Wuchereria DNA, a rebuilding test was done by combining varying numbers of filarial L3 (both Wuchereria and Brugia) with 1, 5, 10, and 30 noninfected adult mosquitoes ranging from four to one larvae per sample. The dissected L3 were mixed with each noninfected adult mosquito samples, followed by DNA extractions and examinations. The method was found to detect both filarial worms in different admixtures ranging from four to one in the order of decreasing intensity of the fluorescence signal (Fig. 3). The test could detect one L3 of W. bancrofti in a pool of four L3 of B. malayi and vice versa inoculated in a pool of 10 noninfected vectors. For determination in a pool of 30 noninfected vectors, the assay was positive only for W. bancrofti DNA detection.
Real-time FRET Multiplex PCR to Detect Filarial Worms in Mosquitoes.
We applied the W. bancrofti-specific DNA sequence SspI repeat and the Brugia-specific DNA sequence HhaI repeat target DNA sequences to detect either W. bancrofti or B. malayi in infected mosquitoes using the real-time FRET multiplex PCR assay joined with melting curve analysis of the amplicon product. The melting curve analyses are shown in Fig. 4. All of the 30 W. bancrofti-infected Cx. quinquefasciatus and the 30 noninfected Cx. quinquefasciatus, as well as the 30 B. malayi-infected Ae. togoi and the 30 noninfected Ae. togoi, were separately examined. When using W. bancrofti-specific primers and probes, the mean Tm value of W. bancrofti was 58.28 ± 0.30 (SD; n = 30), whereas when using B. malayi-specific primers and probes, the mean Tm value of B. malayi was 57.04 ± 0.06 (n = 30). Neither Tm value was shown in noninfected mosquito groups or other control DNA. The 30 total W. bancrofti-infected Cx. quinquefasciatus and 30 B. malayi-infected Ae. togoi were positive, whereas all 30 noninfected Ae. togoi and 30 noninfected Cx. quinquefasciatus were negative by melting curve analysis. The real-time FRET multiplex PCR could measure as little as one larva in a single mosquito. The diagnostic values had 100% sensitivity and specificity.
The size of the amplified products, 188 bp for W. bancrofti and 153 bp for B. malayi, are indicated in Fig. 5. All W. bancrofti- (Fig. 5, lanes 1-3) and B. malayi-infected mosquitoes (Fig. 5, lanes 4-6) were amplified by real-time FRET multiplex PCR, whereas DNA from noninfected Ae. togoi (Fig. 5, lane 7), noninfected Cx. quinquefasciatus (Fig. 5, lane 8), D. immitis (Fig. 5, lane 9), P. falciparum-infected human red blood cells (Fig. 5, lane 10), and human leukocytes (Fig. 5, lane 11) did not amplify these two specific bands.
One of the diagnostic tools of the GPELF is molecular xenomonitoring of mosquito vectors (Weil and Ramzy 2007). It can be used for showing the presence of lymphatic filariasis in endemic regions and selecting areas for inclusion into the program. Regardless of the fact that the detection of filarial parasites in mosquitoes by the classical method is a relative low in cost, it requires well-trained persons for the identification of worms in dissected mosquitoes. When the prevalence of parasitic infection in the vector population decreases or vector infection is reduced after mass drug management, the dissection method becomes more inefficient and impractical (Williams et al. 2002, Weil and Ramzy 2007). Previously, c-PCR has been reported as a promising tool to monitor the progress of eradicating lymphatic filariasis (Williams et al. 2002). Recently, real-time PCR has proven to be even more efficient because it is highly specific and sensitive, allows a high throughput, and can be used on very small samples (Lulitanond et al. 2004, Rao et al. 2006a, b, Fischer et al. 2007, Thanchomnang et al. 2008). Our studies showed, for the first time, that real-time FRET multiplex PCR can be used for the detection of both W. bancrofti and B. malayi in a single PCR reaction of infected mosquitoes. Two pairs of primers were used to produce different genus-specific amplicons, which were subsequently shown by their combined melting peak profiles with two pairs of hybridization fluorophore-labeled probes. Both parasites can be differentially detected by real-time FRET multiplex PCR by their different fluorescence channels and melting temperatures. This assay could also differentiate W. bancrofti and B. malayi DNA in infected vectors from DNA of D. immitis and P. falciparum-infected human red blood cells, as well as noninfected mosquitoes and human leukocytes. However, the assay could detect DNA of B. pahangi in infected cat blood with a different Tm from B. malayi. The Tm value of B. pahangi detection was 52.75 (data not shown).
This method is useful for the detection of both W. bancrofti and B. malayi in areas where both parasites are prevalent. Moreover, the assay could also detect mixed infections that were missed by microscopists because its detection limit was a single L3 of one parasite species in a pool of four L3 of the other species inoculated into noninfected mosquitoes. Moreover, the method can detect as little as one filarial larva inoculated in a pool of 30 noninfected mosquitoes. The results showed a high sensitivity, which makes the assay potentially useful for field surveys.
For epidemiological surveys and elimination programs, verification tests for detecting filarial DNA in human populations at risk and transmission monitoring of vectors are essential. Regardless of the fact that the detection of filarial DNA is only an indirect method of measuring transmission, it gives estimates of lymphatic filarial transmission or the potential of transmission.
In summary, this study showed that real-time FRET multiplex PCR and melting curve analysis are sensitive and specific means for the detection of both Wuchereria and Brugia DNA in a single assay. The above-described technique showed 100% sensitivity and specificity and offered a rapid and reliable procedure for differentially identifying lymphatic filariasis. The whole procedure starting with DNA extraction from samples to real-time FRET multiplex PCR can be completely reported within 5 h. In addition, the introduced real-time FRET multiplex PCR can reduce labor time and reagents costs, and it is not prone to carryover contamination. The assay can be further used to screen mosquito vectors of endemic areas or for the diagnosis of human blood specimens. The method should be useful for evaluating the infection rate of mosquito populations and for xenomonitoring in the community after mass drug dispensations and as such should be an efficient diagnostic method for the GPELF.
The authors thank W. Choochote for technical and B. malayi support, S. Chungpivat for B. pahangi support, and M. Roselieb for assistance in manuscript preparation. This study was supported by the Khon Kaen University grants.