|Home | About | Journals | Submit | Contact Us | Français|
Anopheles longipalpis is morphologically similar to the major African malaria vector Anopheles funestus at the adult stage although it is very different at the larval stage. Despite the development of the species-specific multiplex PCR assay for the An. funestus group, the genomic DNA of Anopheles longipalpis type C specimens can be amplified with the Anopheles vaneedeni and Anopheles parensis primers from this assay. The standard, species-specific An. funestus group PCR, results in the amplification of two fragments when An. longipalpis type C specimens are included in the analysis. This result can easily be misinterpreted as being a hybrid between An. vaneedeni and An. parensis. Anopheles longipalpis type C can be identified using a species-specific PCR assay but this assay is not reliable if other members of the An. funestus group, such as An. funestus, An. funestus-like and An. parensis, are included. The present study provides a multiplex assay that will identify An. longipalpis along with other common members of the African An. funestus group, including Anopheles leesoni.
A total of 70 specimens from six species (An. funestus, An. funestus-like, An. parensis, Anopheles rivulorum, An. vaneedeni and An. leesoni) in the An. funestus group and An. longipalpis type C from Malawi, Mozambique, South Africa and Zambia were used for the study. A restriction fragment length polymorphism (RFLP) assay was designed based on the DNA sequence information in the GenBank database.
The enzyme, EcoRI digested only An. longipalpis type C and An. funestus-like after the species-specific An. funestus group PCR assay. The An. longipalpis and An. funestus-like digestion profiles were characterized by three fragments, 376 bp, 252 bp and 211 bp for An. longipalpis type C and two fragments, 375 bp and 15 bp for An. funestus-like.
An RFLP method for the group was developed that is more accurate and efficient than those used before. Hence, this assay would be useful for field-collected adult specimens to be identified routinely in malaria vector research and control studies.
The Anopheles funestus group originally consisted of nine species [1,2]: the major African malaria vector An. funestus and eight minor or non-vectors, Anopheles aruni, Anopheles parensis, Anopheles vaneedeni, Anopheles confusus, Anopheles rivulorum, Anopheles leesoni, Anopheles brucei and Anopheles fuscivenosus. Subsequent studies on the systematics of the group have resulted in a reclassification of the group with An. funestus, An. aruni, An. parensis, An. confusus and An. vaneedeni being grouped together as members of the "An. funestus subgroup", An. rivulorum, An. rivulorum-like, An. brucei and An. fuscivenosus forming their own subgroup, and An. leesoni being classified with the Asian Anopheles minimus subgroup . Recently, a new species has been described from Malawi, An. funestus-like , and this species falls within the An. funestus subgroup.
In addition to the 10 species mentioned above, there are closely related species that are not included in the group because of morphological differences in the adult females - Anopheles longipalpis being one of these . These differences are very subtle and the probability of confusing this species with the vector An. funestus is very high. To further complicate the issue, molecular studies on this taxon have revealed at least two cryptic species, one from South Africa (Type A) and the other from Zambia (Type C) .
The adult biology of An. longipalpis has not been well studied despite the fact that it is widely distributed in eastern and southern Africa, from Sudan to Angola and South Africa [1,2,6]. Smith [7,8] reported human feeding behaviour in An. longipalpis from indoor and outdoor collections on Ukara Island in Lake Victoria and outdoors in Tanzania. Adugna and Petros  also found An. longipalpis specimens containing human blood meals from their collections in Ethiopia. Recently, Kent et al  reported that although the species is found in large numbers resting indoors in sympatry with An. funestus and Anopheles arabiensis in Zambia, it is predominantly zoophilic with only a small number feeding on humans. The species has never been implicated as a vector of malaria or involved in malaria transmission [1,2,10,11]. However, it may be a secondary vector of malaria in areas of sufficiently high endemicity and densities of this species .
Although adult An. longipalpis are morphologically characterized by maxillary palpal length and pale basal and apical bands spanning the hind tarsal joints [1,2], this species can be misidentified as An. funestus due to similarities of wing venation, palpal banding patterns and small body size. Currently, there are two identification methods for field-collected specimens of An. longipalpis. The first method, suggested by Kent et al , is to identify An. longipalpis using the species-specific PCR assay for the An. funestus group from the internal transcribed spacer (ITS) region of the rDNA . The assay produces two diagnostic fragments from An. longipalpis type C which correlates with An. parensis and An. vaneedeni due to sequence similarity among these species. However, this analysis would lead to misidentification with hybrids of An. parensis and An. vaneedeni, although these have never been found in nature. The second method is a species-specific PCR assay for An. longipalpis from the ITS2 region developed by Koekemoer et al , but this is not specific when other members of the An. funestus group are included such as An. funestus, An. funestus-like and An. parensis. Hence, this study proposes a new method for the molecular identification of An. longipalpis type C and six members of the An. funestus group using a Restriction Fragment Length Polymorphism (RFLP) that is more accurate and efficient than currently used methods.
Mosquitoes were collected from Malawi, Mozambique, South Africa and Zambia (Table (Table1).1). There were eleven An. longipalpis type C specimen available. Initially, specimens were identified morphologically using the keys of Gillies and Coetzee . All DNA samples were extracted from either single mosquitoes or available parts of mosquitoes using the Ballinger-Crabtree protocol  except An. longipalpis where the method of Collins et al  was used. The DNA templates were resuspended in TE buffer at volumes between 50 μL and 100 μL. All DNA templates were identified using the method of Spillings et al  for An. funestus-like and the method of Koekemoer et al  for the rest of the species.
Although the method of Spillings et al  can be carried out simultaneously with the method of Koekemoer et al , they recommend that these assays be carried out separately because the An. funestus-like 390 bp fragment of Spillings et al  is close to the An. rivulorum 411 bp fragment of Koekemoer et al . The procedure in the present study simultaneously carried out both these methods [4,11] with minor modifications. The G^AATTC restriction site for EcoRI enzyme lies within the An. funestus-like and An. longipalpis type C fragments amplified by the methods of Spillings et al  and Koekemoer et al  respectively. The restriction enzyme digested only the fragments for An. funestus-like and An. longipalpis type C directly. A total volume of 25 μL for each PCR reaction contained 1 μL of the genomic DNA of an individual mosquito, 1 × PCR Buffer, 0.2 mM of each dNTP, 0.26 pM (0.2 μM of MalaFB primer) of each primer, and 1 unit of Taq DNA polymerase. The PCR cycling conditions were as follows: a 2 minute 94°C followed by 35 cycles of 30 seconds at 94°C, 30 seconds at 50°C and 40 seconds at 72°C; there was a final extension step of 10 minutes at 72°C. After amplification, 1 unit of EcoRI in 1 × buffer EcoRI (Roche Diagnostic) was added to the PCR reactions and digestion carried out at 37°C for a minimum of three hours. Digested fragments were electrophoresed through an ethidium bromide 2.5% agarose gel and photographed under ultraviolet light illumination using a gel imaging system.
The PCR-RFLP assay for the six members of the An. funestus group and An. longipalpis type C, resulted in different sizes of DNA fragments as recorded in Table Table2.2. Two fragments (252 bp and 587 bp) of An. longipalpis type C were amplified after the species-specific An. funestus group PCR. The restriction enzyme digested only the fragments for An. funestus-like and An. longipalpis type C. There were restriction sites at position 375 in An. funestus-like and at position 376 in the large fragment of An. longipalpis type C (Figure (Figure1).1). Although the restriction site was identical to the DNA sequence of An. parensis, the specific fragment was not involved in the primer design for the method of Koekemoer et al . The digestion profiles for An. funestus-like and An. longipalpis type C were characterized by two fragments, 375 bp and 15 bp and three fragments, 211 bp, 252 bp and 376 bp respectively. However, the length of fragment for An. funestus-like after PCR amplification was not accurate due to approximate length for the species in Spillings et al  and not available for the full alignment of DNA sequences in GenBank. Figure Figure22 shows DNA bands for the six members of the An. funestus group and An. longipalpis type C produced after PCR amplification and restriction enzyme digestion.
The assay for the An. funestus group and An. longipalpis type C used in this study requires one PCR and one RFLP step instead of the three PCR steps currently needed. This will save time and reagents when routinely identifying collections of wild mosquitoes belonging to this important group. All seven species were easily separated on a 2.5% ethidium bromide agarose gel even though the digested amplicons of An. longipalpis type C (252 bp and 211 bp), An. funestus-like (375 bp) and An. rivulorum (411 bp) are close to each other. However, individuals that do not amplify at all may need additional processing using the An. longipalpis type A primers .
Koekemoer et al  reported that sequence analysis of the ITS2 region in An. longipalpis type C is similar to the sequences of An. parensis and An. vaneedeni in the An. funestus subgroup. These three species are almost identical at the adult stage with only a few minor differences [1,2]. However, An. longipalpis was not included in the An. funestus group because of its very different larval morphology [1,3]. Koekemoer et al  indicated that it might be due to extensive evolutionary divergence and the sequence differentiation for An. longipalpis type A, which is close to Anopheles pampanai and Anopheles varuna in the Asian An. minimus group based on the DNA analysis of the ITS2 region. They suggested that An. longipalpis type C should be placed in the An. funestus group supported by sequence similarity to both An. parensis and An. vaneedeni. Furthermore, Kent et al.  also reported that the fragments diagnostic for An. longipalpis type C were the same as the diagnostic amplicons for a hypothetical An. parensis/An. vaneedeni hybrid although no such hybrids have been recorded in nature. However, further investigation of An. longipalpis is still required with more field specimens.
The application of the method described is expected to greatly improve the efficiency of large-scale analysis of field-collected samples of the An. funestus group and An. longipalpis type C.
The authors declare that they have no competing interests.
KSC designed the study, developed the new RFLP assay and drafted the manuscript. MC and LLK assisted with analysis of the data and helped draft the manuscript.
All authors have read and approved the final manuscript.
This work was supported by the Department of Science and Technology, National Research Foundation Research Chair Initiative grant to MC. We thank the following for assistance with field collections or supply of specimens: A. Mabuza (Department of Health, Mpumalanga Province, South Africa); K. Hargreaves (Department of Health, KwaZulu/Natal Province, South Africa); T. Ledwaba (Department of Health, Limpopo Province, South Africa) and Prof R.H. Hunt (University of the Witwatersrand).