Background
The
Anopheles gambiae sensu lato (s.l.) species complex contains the most important mosquito vectors of malaria in sub-Saharan Africa. It comprises seven morphologically indistinguishable sibling species up to four of which may be sympatric [
1‐
4]. The principal malaria vectors in the complex are
Anopheles gambiae sensu stricto (s.s.) and
Anopheles arabiensis. Of the remaining members
Anopheles quadriannulatus species A, which is widespread in southern Africa, and
Anopheles quadriannulatus species B, found in Ethiopia, are considered to be zoophilic non-malaria vectors [
1,
5].
Anopheles melas and
Anopheles merus are both salt water breeding and consequently only important vectors in costal regions [
6,
7]. The final named member of the complex
Anopheles bwambae is restricted to a region close to the Buranga hot springs in Uganda [
8].
The marked differences in the vectorial efficiency of the species within the complex mean that rapid identification of species is vital for focussed effort in malaria control programmes. To this end a large number of methods have been developed for identification including cross-mating techniques [
9], polytene chromosome analysis [
2], allozyme electrophoresis [
10,
11], and gas chromatography of cuticular lipids [
12]. However, the most widely adopted method for species identification is based on PCR amplification of ribosomal DNA (rDNA) sequences [
13]. The original report of this approach described by Scott
et al is a multiplex PCR method using species specific primers to amplify products which are of a diagnostic size when visualized by agarose gel electrophoresis. Several modifications or additions to the original method, or similar methods based on the same rDNA sequences have also been reported [
3,
14‐
17]. Although the Scott method is now considered to be the 'gold standard' for species identification in this complex recent reports have described non-specific amplification and high rates of failures leading to the need for repeat PCR of the same samples [
17]. Additional drawbacks of this method are the time required to process samples when applied to large scale screening of mosquito populations and the safety hazard that ethidium bromide staining of agarose gels entails.
Recently, a high-throughput method for identification of the principal vectors in the complex
An. gambiae s.s. and
An. arabiensis has been described [
18]. The method is based on TaqMan single nucleotide polymorphism genotyping and represents the first description of a 'closed tube' approach that requires a single step to identify a mosquito DNA sample. The method was tested in a large study of field collected
An. gambiae s.s. and
An. arabiensis from western Kenya and was found to have a specificity and sensitivity comparable to standard PCR. However, this assay was not tested using
An. gambiae s.s. or
An. arabiensis from other regions in Africa, nor was it tested for its ability to distinguish these two species from the non-vector
An. quadriannulatus. The current study describes an extensive blind trial of field collected mosquitoes from a range of sites across sub-Saharan Africa comparing the Walker TaqMan SNP genotyping and standard PCR methods with a newly developed TaqMan assay which distinguishes between the principal vector and non-vector species of the complex.
Methods
Mosquito samples and preparation of plates for species identification trial
For the initial optimization of each assay, field-caught mosquito specimens from Burkina Faso, Ghana, Kenya, Cameroon and Malawi were used in addition to samples obtained from two laboratory colonies, Kisumu and RSP. These samples included several samples of each of An. gambiae s.s, An. arabiensis, An. quadriannulatus species A, An. melas and An. merus. Samples of the Ethiopian An. quadriannulatus species B and An. bwambae were not tested in this study because of their more limited distribution.
The blind species identification trials were performed using five 96 well test plates containing 466 samples. Samples were field collected from Cameroon, Ghana, Kenya, South Africa, Malawi, Sao Tome, La Reunion, Tanzania, Sudan, Angola, Burkina Faso, Gabon, and Mozambique. These samples had been initially identified to species at the time of collection using the standard PCR method and included 169
An. gambiae s.s., 173
An. arabiensis, 66
An. quadriannulatus, and 21 samples of
An. melas and
An. merus, the remaining samples were either undetermined or negative controls. This information was withheld from the persons who carried out the testing of each assay to ensure no bias occurred in the scoring of results. For all samples DNA was extracted from single mosquitoes using either the Livak or Ballinger Crabtree methods [
19,
20]; DNAzol reagent (Molecular Research Centre, Inc) at one-fifth the recommended reagent volume for each extraction or using a crude boil and centrifugation protocol. This was a variant of the STE method of O'Neill
et al. [
21] in which samples were ground in STE buffer (100 mM NaCl, 10 mM Tris-HCl, pH8.0, 1 mM EDTA, pH8.0), heated to 95°C for 5 mins, centrifuged for 3 mins at 13,000 rpm and the supernatant used directly as PCR template. The DNAs were resuspended in either TE buffer or sterile water at volumes between 100 and 200 μl. To determine the sensitivity of the three identification methods a dilution series of the DNA from each of the five main species in the
Anopheles gambiae complex was included in the trial. For this, DNA preparations were diluted to 20 ng/ul (as determined by absorption at 260 nm using a NanoDrop spectrophotometer, NanoDrop Technologies). The samples were then serially diluted down to a 1 in 1 × 0
6 dilution.
Standard PCR
Standard PCR was carried out as described previously [
13].
Walker TaqMan Assay
The design of a TaqMan assay to distinguish
An. gambiae s.s. from
An. arabiensis has been described previously [
18]. The PCR conditions used in the study by Walker
et al are not given in the original manuscript but were obtained from Edward Walker (personal communication). These were modified slightly and were as follows: PCR reactions (25 μl) contained 1 μl of genomic DNA, 12.5 μl of SensiMix DNA kit (Quantace), 900 nM of each primer and 200 nM of each probe. Samples were run on a Rotor-Gene 6000™ (Corbett Research) using the temperature cycling conditions of: 10 minutes at 95°C followed by 40 cycles of 95°C for 20 seconds and 60°C for 45 seconds. The increase in VIC and FAM fluorescence was monitored in real time by acquiring each cycle on the yellow (530 nm excitation and 555 nm emission) and green channel (470 nm excitation and 510 emission) of the Rotor-Gene respectively.
Novel TaqMan Assay
A novel TaqMan assay that would distinguish the main malaria vectors
An. arabiensis and
An. gambiae s.s. from the non-vector
An. quadriannulatus was designed after it was found the Walker TaqMan assay incorrectly identified
An. quadriannulatus, An. merus and
An. melas as
An. gambiae s.s. An alternative region within the rRNA gene at the 5' end of the intergenic spacer (bases 475 to 487) was selected where the nucleotide sequence in
An. gambiae s.s. and
An. arabiensis is GCTCGTCTTGGTC. In
An. quadriannulatus,
An. melas and
An. merus the same region of sequence shows a SNP (GC
G CGTCTTGGTC). Flanking this SNP was an area of conserved sequence between the species which allowed for the design of forward, comF (5'-GCTTGGTGGTTTGTCCG-3'), and reverse, comR (5'-CTGTGTCGACGTGGTCCC-3'), primers. Antisense minor groove binding (MGB) probes (Applied Biosystems) that bind over the SNP site were designed using the Primer Express™ Software Version 2.0. The probe AG/AA (5'-GACCAAGACGAGC-3') was labelled with 6-FAM at the 5' end for the detection of
An. gambiae s.s. and
An arabiensis and the probe AQ/AM (5'-GACCAAGACGCGC-3') was labelled with VIC for detection of
An. quadriannulatus,
An. melas and
An. merus. Each probe also carried a 3' nonfluorescent quencher and a minor groove binder at the 3' end. The minor groove binder provides more accurate allelic discrimination by increasing the T
M between matched and mis-matched probes [
22]. This new TaqMan assay was used in the blind genotyping trial to score samples into a group containing the principal malaria vectors in the complex
An. gambiae s.s. and
An. arabiensis and a second group containing
An. quadriannulatus, An melas and
An. merus. Samples that had scored as 'vectors' were then further identified to species using the Walker TaqMan assay.
PCR reactions (25 μl) contained 1 μl of genomic DNA, 12.5 μl of SensiMix DNA kit (Quantace), 800 nM of each primer and 200 nM of each probe. Samples were run on a Rotor-Gene 6000™ (Corbett Research) using the temperature cycling conditions of: 10 minutes at 95°C followed by 45 cycles of 95°C for 15 seconds, 50°C for 20 seconds and 72°C for 20 seconds. The increase in VIC and FAM fluorescence was monitored in real time by acquiring each cycle on the yellow (530 nm excitation and 555 nm emission) and green channel (470 nm excitation and 510 emission) of the Rotor-Gene respectively. PCR reactions were also carried out using whole mosquitoes or single legs instead of genomic DNA, in this case the leg or body was simply placed in the PCR tube and covered with the PCR mastermix.
Discussion
Identification of the species in the Anopheles gambiae complex is important for ecological research studies interested in the geographic distribution, abundance and behaviour of the different vector species. In addition as the complex comprises species with different efficiencies as vectors, and includes non-vector species accurate identification is also paramount for vector control programmes. In this blind species identification trial of three assays using a range of mosquito specimens collected from a wide variety of geographic areas the most widely used multiplex PCR method was compared with two TaqMan assays, one which has been described previously and one which is described here for the first time.
The standard PCR method was found to be very specific with a low rate of incorrect scores (<1%), however, when compared to the TaqMan assays it showed a significantly higher rate of failed reactions (15% compared to 1.25% and 2.96%). This was noticed to occur more frequently using DNA templates that had been extracted by the STE buffer protocol which, in the authors' experience, is quick to carry out but yields DNA that is of lower quality and more likely to degrade with time or multiple freeze-thawing. The chief disadvantage of this method is the requirement for the post-PCR processing of samples by agarose gel electrophoresis. This is time consuming, restricts throughput, requires the use of the safety hazard ethidium bromide, and variation in the quality of the agarose gels can lead to difficulties in interpreting results.
The TaqMan method, in contrast to standard PCR, does not require post-PCR processing due to the real-time detection of the specific binding of fluorescently labelled probes during PCR. This 'closed-tube' approach makes the method high-throughput and simple to carry out. Initially, a recently developed TaqMan method that identifies the two main mosquito vector species of the complex
An. gambiae s.s. and
An. arabiensis was trialled [
18]. During optimization with mosquito samples of known species it became clear that this assay could not be used to differentiate these two members of the complex from the others as
An. melas, An. merus and the non-vector species
An. quadriannulatus were all incorrectly identified as
An. gambiae s.s. Closer examination of the 5' intergenic spacer region of rDNA to which the probes bind reveals that while
An. gambiae s.s. and
An. arabiensis differ at seven positions
An. melas and
An. merus differ from
An. gambiae s.s. at only two positions and
An. quadriannulatus at only one which may explain the non-specific amplification exhibited. If this method was able to accurately identify
An. gambiae s.s. and
An. arabiensis from the other members of the complex one additional potential disadvantage of the assay design is that an
An. gambiae s.s. or
An. arabiensis specimen that failed in the PCR reaction might be incorrectly scored as
An. melas/
An. merus/
An. quadriannulatus as these would also score as 'no templates/fails'.
The TaqMan assay developed in this study is designed to distinguish between the main malaria vectors
An. gambiae s.s. and
An. arabiensis as one group and
An. quadriannulatus, An. melas or
An. merus as a second group. The rationale behind this design is that although
An. melas and
An. merus have been shown to be malaria vectors their distribution is limited to coastal regions as a consequence of their requirement for brackish water to breed. In contrast,
An. quadriannulatus species A which is a non-vector species is widespread in southern Africa and is sympatric in many areas with
An. gambiae s.s. and
An. arabiensis. Therefore, there is a considerable need in many vector control programs to rapidly distinguish between the two main vector species and the non-vector species. The new TaqMan addresses this requirement and has application as a vector/non-vector identification test for large parts of sub-Saharan Africa. An alternative application for this assay is to distinguish
An. melas from
An. gambiae s.s./An. arabiensis where they are sympatric such as regions along the west coast of Africa. This is viable as the former does not occur sympatrically with
An. merus or
An. quadriannulatus [
5]. It should, however, be used with caution in regions close to the east coast of Africa where
An. quadriannulatus has been found to occur sympatrically with
An. merus [
5]. In contrast to the previous TaqMan assay this method will also correctly identify a failed reaction as all members of the complex (except for
An. bwambae which was not tested due to its rarity) are detected by one of the two fluorescently labelled probes. Used alone the new TaqMan assay will not distinguish between
An. gambiae s.s. and
An. arabiensis which may be significant for certain studies, such as those examining the spread of resistance genes in mosquito populations. However if the identification of
An. gambiae s.s. from
An. arabiensis is required, then the two TaqMan assays described here can be used sequentially. In this case the new TaqMan is run first having the advantage that it can be used on a single leg from a silica-dried mosquito without the need to first extract DNA. Samples that are scored as
An. gambiae s.s./
An. arabiensis can then be further identified to species using the Walker TaqMan. This approach was used in the blind species identification trial and proved to be successful, taking less time to genotype all samples than the standard PCR method. It may be possible in future to increase throughput and reduce consumable costs by combining the two TaqMan assays so that
An. gambiae s.s, An. arabiensis and
An. quadriannulatus can be detected in a single tube using probes labelled with fluorophores with distinct emission and excitation spectra. An additional advantage of the TaqMan assays seen in the species identification trial was the very low rate of incorrectly identified samples (0 and 0.54%) and failed reactions (1.25% and 2.96%) which again increases throughput over the standard PCR as it alleviates the need to spend time repeating reactions.
Authors' contributions
CB designed the study, developed the new TaqMan assay, optimized the standard PCR method and the previously described TaqMan method and drafted the manuscript. CSW and MJD collected mosquito specimens, extracted DNA and helped draft the manuscript. MSW helped design the study, prepared the plates of mosquito DNAs and helped draft the manuscript. LMF helped design the study and helped draft the manuscript. All authors read and approved the final manuscript.