A new malaria vector mosquito in South Africa

South Africa aims to eliminate malaria within its borders by 2018. Despite well-coordinated provincial vector control programmes that are based on indoor residual insecticide spraying, low-level residual malaria transmission continues in the low-altitude border regions of the north-eastern sector of the country. In order to identify the underlying causes of residual transmission, an enhanced vector surveillance system has been implemented at selected sites in the Mpumalanga and KwaZulu-Natal (KZN) provinces. The collection periods for the data presented are March 2015 to April 2016 for Mpumalanga and January 2014 to December 2015 for KZN. The mosquito collection methods used included indoor and outdoor traps based on the use of traditional ceramic pots, modified plastic buckets and exit window traps (KZN only). All Anopheles funestus species group mosquitoes collected were identified to species and all females were screened for the presence of Plasmodium falciparum sporozoites. Two An. vaneedeni females, one from each surveillance site, tested positive for P. falciparum sporozoites. These are the first records of natural populations of An. vaneedeni being infective with P. falciparum. As both specimens were collected from outdoor-placed ceramic pots, these data show that An. vaneedeni likely contributes to residual malaria transmission in South Africa.

Malaria is a parasitic disease caused by Plasmodium protozoa and transmitted by female Anopheles mosquitoes (Diptera: Culicidae) 1 . The populations most at risk live in sub-Saharan Africa which accounts for 80% of cases and 90% of total deaths 2 .
Malaria transmission in South Africa is limited to the low-altitude northern and north-eastern border regions of the country which span the Limpopo, Mpumalanga and KwaZulu-Natal (KZN) provinces 3 . Historically, only the major malaria vector Anopheles funestus was directly implicated in malaria transmission in South Africa 4 . In addition, the malaria vectors An. arabiensis and An. merus were provisionally implicated based on their occurrence in South Africa's malaria affected regions and because they have been directly implicated in transmission in neighbouring southern Mozambique [5][6][7][8] . Recently, several An. arabiensis and one An. merus specimen collected outdoors during 2014-2016 were also found to be infected with P. falciparum sporozoites in KwaZulu-Natal (Dandalo et al., submitted), thus expanding the number of species directly implicated in malaria transmission within South Africa.
Malaria vector control in South Africa's malaria affected provinces is primarily based on indoor spraying of long-lasting residual insecticides 8 . The indoor residual spraying (IRS) method has been the mainstay of malaria vector control in South Africa since the 1940s and has remained effective owing to carefully co-ordinated provincial IRS programmes 9 . Despite this, low-level residual malaria transmission continues and is likely caused by outdoor feeding and resting Anopheles vector mosquitoes that are unaffected by indoor applications of insecticide 10 . Maintaining effective control whilst scaling up control methods to address ongoing residual malaria transmission within South Africa are high priority activities because the country has adopted a malaria elimination agenda and aims to eliminate malaria within its borders by 2018 11 .
Continuing residual malaria transmission and the burgeoning incidence of insecticide resistance in malaria vector populations within South Africa's borders 4,12 have necessitated an intensification of vector surveillance activities in the affected provinces. The principle objectives of these enhanced surveillance activities are to compare and establish optimal methods of collecting adult Anopheles mosquitoes, to establish which Anopheles species are responsible for ongoing residual malaria transmission, to assess the extent of residual malaria transmission within South Africa and to assess the range and geographical extent of insecticide resistance in incriminated vector populations. Within these broad objectives, the aim of this project was to assess whether An. vaneedeni, a member of the An. funestus species group 13 , contributes to residual malaria transmission in South Africa.
A total of 255 and 77 An. funestus group specimens was collected from the KZN and Mpumalanga sites respectively (Table 1). Of these, 45 of the adult females collected from Block A and Vlakbult (Mpumalanga) and 51 of the adult females collected from Mamfene (KZN) were positively identified as An. vaneedeni, first by morphology 10 to An. funestus group followed by PCR 14 to species (Table 1). Two of these An. vaneedeni specimens, one from each province, tested positive for the presence of Plasmodium falciparum sporozoites through two ELISA assays. These data give P. falciparum infectivity rates for An. vaneedeni of 2.44% and 1.96% for the Mpumalanga and KZN sites respectively.
The Anopheles species identification of these two specimens as An. vaneedeni was further confirmed by sequencing their internal transcribed spacer 2 (ITS2) region 14 . There was a 99% sequence identity between the ITS2 region of the KZN and Mpumalanga specimens and the published An. vaneedeni sequence originating from South Africa (GenBank accession number JN994152.1 15 ). The presence of P. falciparum sporozoite DNA was confirmed by nested PCR for the KZN specimen 16 . Sequence analysis of the nested PCR product revealed that there was a 99% identity to the published sequence for P. falciparum (GenBank accession number KT991235.1 17 ).

Discussion
The perennial occurrence of several An. funestus species group members-An. vaneedeni, An. rivulorum, An. leesoni and An. parensis-at the Mpumalanga and KwaZulu-Natal field sites warrants investigation into their possible contribution to malaria transmission in these regions, especially given that An. rivulorum has been implicated in malaria transmission in Tanzania 18 and Kenya 19 , and a previous survey has shown that An. parensis will readily rest indoors in the Mamfene region 20 . The evident absence of An. funestus sensu stricto at these sites can be attributed to ongoing annual IRS based control activities, particularly the use of DDT which has effectively eradicated this species from South Africa. This is because An. funestus in South Africa is highly susceptible to DDT 4,8 and has a strong tendency to rest indoors, making this species especially susceptible to IRS programmes that utilize DDT 7,9 . The data summarised here represent the first record of wild-caught P. falciparum sporozoite positive An. vaneedeni females, directly implicating this species in malaria transmission in South Africa. Although this species is considered to be to be primarily zoophilic 10 , it will readily feed on humans outdoors and has previously been experimentally infected with P. falciparum under laboratory conditions 21 . The outdoor-resting and feeding traits of this species are reinforced by the fact that most of the An. vaneedeni specimens collected in these surveys, including the two that tested sporozoite positive, were found in outdoor-placed ceramic pots (Fig. 2) deployed at randomly selected households at the two sites.
The geographical range of An. vaneedeni primarily includes the north-eastern low-altitude regions of South Africa, and likely extends into southern and eastern Zimbabwe and southern Mozambique 22 . However, there are collection records of An. vaneedeni in the western highlands of Kenya 23 , suggesting that its range may be substantially more extensive.
The collection of sporozoite-positive, outdoor-resting An. vaneedeni supports the hypothesis of ongoing outdoor residual malaria transmission in South Africa, as first proposed by De Meillon et al. in ref. 21, and tentatively suggests that this species may also be contributing to malaria transmission in other malaria-endemic countries in which it occurs. This information highlights the need to intensify malaria vector control in South Africa by including methods designed to target outdoor feeding vector populations without compromising the efficacy of the IRS programme.

Materials and Methods
Ethical statement. Informed consent was obtained from all household owners involved in this study.
Ethical clearance for the collection of mosquito specimens was obtained from the University of the Witwatersrand (M141023 & W-CJ-150520-2) and the KwaZulu-Natal Department of Health (HRKM337/14).

Mosquito collections.
Adult Anopheles mosquitoes were collected using traditional ceramic pots and modified plastic buckets (Fig. 2). These were placed both inside and outside selected households in Vlakbult and Block A in Mpumalanga (March 2015-April 2016), and only outside at households in Mamfene, KZN (January 2014-December 2015). Traps were not deployed indoors at Mamfene because homeowners consent did not include this provision and because exit window traps were instead used to collect indoor resting mosquitoes at this site (Fig. 3). The traps were cleared at sunrise weekly by the malaria vector surveillance teams based near the collection sites in each province.
Anopheles species identification and vector incrimination. All Anopheles specimens collected were preserved on silica and initially identified by external morphology using dichotomous keys 10 . Those identified as belonging to the An. funestus group were subsequently identified to species level by PCR 14 . All females were screened for the presence of Plasmodium falciparum sporozoites by ELISA 24 . Plasmodium falciparum infectivity, where indicated by ELISA, was confirmed with a nested Plasmodium PCR 16 . Sequence analysis of mosquito ITS2 and Plasmodium falciparum ssRNA. In order to confirm Anopheles species identity, the internal transcribed spacer region 2 (ITS2) region of the rDNA of each An. vaneedeni sporozoite positive sample was amplified using the following primers: ITS2A: 5′-TGTGAACTGCAGGACA-CAT-3′ ; and ITS2B: 5′-TATGCTTAAATTCAGGGGGT-3′. PCR conditions were the same as those used in the An. funestus species-specific PCR 14 . In order to confirm Plasmodium species identity, Plasmodium ssRNA from the KwaZulu-Natal An. vaneedeni sporozoite positive sample was amplified using the following primers: rPLU5: 5′-CCTGTTGTTGCCTTAAACTTC-3′; rPLU6: 5′-TTAAAATTGTTGCAGTTAAAACG-3 for the first amplification step, and rFAL1: 5′-TTAAACTGGTTTGGGAAAACCAAATATATT-3′; and rFAL2: 5′-ACACAATGAACTCAATCATGACTACCCGTC-3′ for the second amplification step and sequencing. PCR conditions were the same as those previously described 16 . All amplicons were electrophoresed on 2.5% agarose gels stained with ethidium bromide and product sizes were confirmed using a molecular weight marker (Thermo Scientific O'GeneRuler 100 bp DNA Ladder; 0.1 μ g/μ l concentration, supplied with 1Ml 6x Orange DNA Loading Dye). ITS2 PCR products were purified and sequenced by Macrogen. The sequences were manually edited by BioEdit version 7.2.5 25 . Subsequently, the sequences were aligned with sequences stored in GenBank using nucleotide BLAST (BLASTn) (http://blast.ncbi.nlm.nih.gov/Blast.cgi).