Background

Before the 1960s, the Republic of Korea (ROK) was a Plasmodium vivax malaria-endemic country, and malaria prevailed throughout the country. In the late 1970s, malaria was eliminated through eradication efforts by the World Health Organization and the government of the ROK. However, since 1993, malaria has re-emerged in the north-western region of Gyeonggi-do [1], and over 1,000 cases of malaria have been diagnosed every year with a peak of 4,142 cases diagnosed in 2000 [2]. Although the malaria-vector mosquitoes, the Anopheles species, are prevalent throughout the Korean Peninsula, most of the incident cases in the ROK have been reported in the northern part of Gyeonggi-do and the north-western part of Gangwon-do near the Demilitarized Zone (DMZ) between North Korea and the ROK [3]. An area of high transmission is found in North Korea along the DMZ as well, which suggests a parallel outbreak occurred in both countries [4].

In the ROK, eight Anopheles species, including the two recently reported species of Anopheles belenrae and Anopheles kleini, have been recorded so far [5, 6]. These eight species belong to three groups (Hyrcanus, Barbirostris, and Lindesayi) in the subgenus Anopheles. The Hyrcanus Group consists of 27 species, six in the Lesteri subgroup, four in the Nigerrimus subgroup, and the rest 17 unplaced species within the group [7]. The following six of these species have been reported in the ROK: Anopheles sinensis sensu stricto, Anopheles lesteri, Anopheles pullus, Anopheles sineroides, An. kleini and An. belenrae [5, 6, 8, 9]. These six species comprise a species complex called An. sinensis sensu lato. Because they occur in sympatry and it is hard to distinguish by morphology An. belenrae and An. kleini from An. pullus and An. sinensis s.s. [1014]. Thus, molecular methods have been used for species identification [9, 13, 15, 16]. The other two non-Hyrcanus species, Anopheles koreicus and Anopheles lindesayi, belong to the Barbirostris and Lindesayi Groups, respectively [6].

The primary vector species for malaria in Korea has long been declared An. sinensis s.s. [3, 13], but their P. vivax malaria susceptibilities and vectorial capacities have been controversial. Recent studies have suggested that An. pullus An. kleini and An. lesteri are the primary vector species and that An. sinensis s.s. does not play a primary role [2, 15, 17]. In addition, An. sineroides and An. belenrae have been reported as P. vivax-positive based on enzyme-linked immunosorbent assays and polymerase chain reactions (PCR) [18, 19].

Because all six species of An. sinensis s.l. in the ROK transmit the malaria parasite P. vivax, studying their discriminative ecology, such as their blood feeding and resting behaviours, their larval habitats, and their responses to insecticides, is necessary for malaria control. Hybridization between related species of the anopheline species, which rarely takes place in natural environments, might complicate malaria vector control [2023]. Evidence of introgressive hybridization has been reported for the Anopheles gambiae complex [2026]. In addition, recent studies have reported the presence of natural hybrids between some anopheline species in the ROK [27].

Until now, the use of insecticides has been the most effective and economical vector control method for malaria and other vector-borne diseases, such as dengue and filariasis [28]. Since the first introduction of insecticides to the ROK in the 1970s, pyrethroids and organophosphates have been used throughout the country in order to control medically and agriculturally important arthropod pests, including mosquitoes. The national malaria eradication programme that was based mostly on the chemical control of vector mosquitoes was successful in the 1990s in the ROK. However, overdoses of insecticide have quickly led to the presence and spread of insecticide-resistant mosquitoes [13], which have caused serious problems for malaria-controlling interventions.

Resistance to pyrethroids and DDT, which is known as knockdown resistance (kdr), is caused by a single mutation in the S6 transmembrane segment of domain II in the voltage-gated sodium channel (VGSC) gene [29]. Several mutations at codon 1014, such as L1014F (Leu-to-Phe), L1014S (Leu-to-Ser), and L1014C (Leu-to-Cys) have been reported in many Anopheles species, including An. gambiae, Anopheles arabiensis, Anopheles culicifacies, Anopheles stephensi, An. sinensis complex, Anopheles sacharovi, Anopheles subpictus, Anopheles sundaicus, Anopheles aconitus, and Anopheles vagus [3037]. A positive correlation between the kdr genotype and the resistance phenotype to pyrethroids and DDT in the Anopheles species was well documented by articles [34, 38, 39].

A previous study reported that the frequencies of the kdr allele of An. sinensis in the ROK ranged from 25.0 to 96.6%, which suggested that pyrethroid resistance was already widespread in natural populations of the ROK [34]. A standard World Health Organization insecticide susceptibility test was conducted on the Anopheles species in the ROK [4042], and the results showed the development and widespread distribution of pyrethroid-resistant phenotypes. In order to monitor knockdown mutations of pyrethroid resistance in An. sinensis, real-time PCR amplification of a specific allele (rtPASA) has been developed. The results suggested that L1014F mutation was a major allele that showed a high allele frequency, whereas L1014C mutation was a minor allele that showed a low allele frequency within the An. sinensis populations in the ROK [34]. However, these previous studies conducted chemical and molecular assays on the An. sinensis group and species-specific assays have never been conducted because the sibling species are morphologically indistinguishable, and, unlike An. sinensis s.s., the other species have relatively low population densities. Because the mechanisms and degree of insecticide resistance vary among species and populations [4346], the present study aimed to explore the species-specific distribution of kdr resistance alleles in all members of the An. sinensis complex in the ROK.

Methods

Mosquito collections and species identification

Mosquitoes were collected from 22 locations (Figure 1, Table 1) in the ROK with a CDC Miniature Light Trap (John W Hock Company, Gainesville, FL, USA) and preserved on site using dry ice. Because it is morphologically hard to distinguish the species within the An. sinensis complex, the mosquito specimens were isolated based on their morphological characteristics that were determined first under a stereomicroscope in the laboratory compared to the other insects, and then the specific species of all of the Anopheles specimens were identified using multiplex assays [47]. DNA was extracted from the entire body or legs by following a standard phenol extraction protocol or DNeasy Blood and Tissue kit (QIAGEN, USA).

Figure 1
figure 1

The geographical locations of the sampling sites in a map of the Republic of Korea (ROK).

Table 1 Summary of regional information and the species composition of the An. sinensis complex

DNA sequencing of VGSC

In order to detect the kdr mutation, part of the IIS6 domain of the VGSC was PCR amplified and directly sequenced using 5′ASIIS56 and 3′ASIIS56 intron primers [34]. The sequences were determined using an ABI 3730xl DNA analyser (Applied Biosystems, USA) and visually confirmed using Sequence Navigator 1.1 software (Applied Biosystems, USA). The obtained sequences were aligned with Clustal X version 2.0 [48]. The deduced amino acid comparisons were conducted by MEGA version 5 [49].

Statistical analyses

The maximum likely frequency (y) of an allele present or absent in a sample of a given size (x) was obtained from the upper of 95% confidence limit of binomial distribution, given by y = 1-0.051/x, following the example of Post and Millest [50].

Results

Species composition and distribution

A total of 755 An. sinensis complex mosquitoes were collected from 22 sites in the ROK, as briefly described in Figure 1 and Table 1. As a result of the molecular identifications that were conducted by multiplex assays (Table 1), most of the specimens were An. sinensis s.s. (665 of 755, 88.08%), which is known as the dominant species in the ROK, and these were followed by An. kleini (30 of 755, 3.97%), An. pullus (26 of 755, 3.44%), An. sineroides (19 of 755, 2.52%), An. lesteri (13 of 755, 1.72%), and An. belenrae (two of 755, 0.26%). Hybrid individuals were not detected. An. sinensis s.s. was found at most of the sites (21 of 22), whereas the other species were found at a few sites (one to five of 22) in the northern part of the ROK, and these results were in concordance with previous studies showing topoclinal distributions.

kdr mutations

In order to examine the mutations at codon 1014, 343-bp sequences of part of the IIS6 domain of the VGSC gene were obtained from a total of 177 specimens that consisted of An. sinensis s.s. (n = 87), An. kleini (n = 30), An. pullus (n = 26), An. belenrae (n = 2), An. lesteri (n = 13), and An. sineroides (n = 19) from 10 sites in the ROK (Table 2 and 3). The DNA sequences were different among the species. However, the deduced amino acid sequences showed no non-synonymous mutations within the VGSC-coding regions in all six species, except for at codon 1014. Pyrethroid- and DDT-resistant kdr mutations were detected in An. sinensis s.s. only, and the other five species had the wild-type kdr allele L1014 only (Figure 2, Table 2 and 3). Interestingly, An. sineroides had a different amino-acid coding sequence (TTA). But the same amino acid (leucine) occurred at codon 1014. Met-Thr mutation at codon 918 (M918T), or super-kdr, which enhances the resistance in combination with L1014F within the same genetic region was not detected in any of the six species.

Table 2 Species-specific genotype frequencies of the kdr allele in each study site
Table 3 Species-specific allele frequencies of the kdr allele (and 95% confidence intervals, Cls) in each study site
Figure 2
figure 2

Alignment of multiple DNA sequences at codon 918 and codon 1014 within the voltage-gated sodium channel gene in the An. sinensis complex in the ROK. The blue box indicates codon 918 and the red box indicates codon 1014.

An. sinensis s.s.

A total of four types, which consisted of one wild type (TTG L1014) and three mutant types (TTT L1014F, TTC L1014F, and TGT L1014C), of kdr alleles were detected in An. sinensis s.s.. TTC L1014F was found for the first time. TTT L1014F was the dominant mutation in all of the regions, whereas TTC L1014F, which was distributed in the eastern part of the ROK alongside the Taebaek Mountain (YY, AD, and GR in Figure 1 and Table 1 and 2), existed only in the heterozygotic state (Table 2). The allele frequencies of L1014 ranged from 7.7% to 75.0%, while that of L1014F (TTT) ranged from 0% to 70.6%, that of L1014F (TTC) ranged from 0% to 14.7%, and that of L1014C ranged from 0% to 23.1% (Table 3).

Discussion

Knockdown resistance (kdr) is caused by mutations at codon 1014 of the VGSC, which is the target of pyrethroids and DDT. In this study, molecular analysis of the VGSC in the An. sinensis complex in the ROK revealed that mutations at codon 1014 existed only in An. sinensis s.s., whereas no kdr mutations were observed in the other five species, including An. pullus, An. kleini, An. sineroides, An. lesteri and An. belenrae. For wild-type L1014, a previous comparative study of the VGSC of various taxa showed different usages of coding triplets at L1014 [51], while in the present results, the amino-acid coding sequences at L1014 differed between the species. The codon of L1014 was TTC in five species (An. sinensis s.s., An. pullus, An. kleini, An. lesteri and An. belenrae), whereas for An. sineroides it was TTA, which is observed in a number of species, including An. gambiae An. arabiensis and An. vagus [30, 36, 37].

Pyrethroids have been widely used in large quantities for the control of agricultural pests and medically important arthropod pests through thermal fogging, residual spraying, impregnated clothing, and mosquito nets since their introduction to the ROK in the 1970s. Inevitably, a variety of mosquitoes, including malaria vectors, might have developed resistance to pyrethroids and jeopardized the successful malaria control programmes in the ROK [34, 40, 42]. A number of recent studies have reviewed the numerous cases that were examined and whether there were correlations between mutations in the VGSC and the resistance phenotype in diverse taxa [38, 51, 52]. Almost all of the results of the previous studies have shown a strong causal relationship between the kdr genotype and insecticide susceptibility to pyrethroids and DDT. However, the absence of a kdr mutation is not a causal factor of total susceptibility to the insecticides.

Although the detailed research on fitness costs of kdr mutations, additional biochemical assays and specific gene expression studies are needed for future study, the highly polymorphic kdr gene that was observed only in An. sinensis s.s. and the absence of a kdr mutation in the other five species suggest two tentative hypotheses. First, because the anopheline species other than An. sinensis s.s. are geographically distributed in the northern part of the ROK near the DMZ, the absence of kdr mutations in these species may have resulted from migration of these kdr-free mosquitoes from North Korea or the DMZ to the ROK to seek for blood sources, such as cows or pigs. Along the DMZ and in North Korea, the environment was unsuitable for the mosquitoes due to the lack of blood sources. However, the amount of insecticide used in North Korea and along the DMZ was much less than the amount used in the ROK, which reduces the insecticide selection pressure [53]. This hypothesis is in concordance with the renowned hypothesis that the re-emergence of P. vivax in the ROK arose from North Korea through sporozoite-infected mosquitoes that dispersed from North Korea to the ROK across the DMZ [3, 5357]. Second, the highly polymorphic kdr gene in An. sinensis s.s. (the dominant Anopheles mosquito in East Asia) in the ROK may be explained by their large population size and wide species range. The level of genetic variation within a species is generated by mutation and eliminated by genetic drift due to finite population size [58]. Large population size and wide species range are precondition to polymorphism, as a large effective population size will have an increased genetic variability [59, 60] and a higher rate of mutation [61].

For An. sinensis s.s., the geographical distribution of kdr alleles is supported by a previous study that examined the population genetic structure of An. sinensis using mitochondrial control regions, and the results suggested distinct subdivisions in the Northern Group (NG) and Southern Group (SG) in the ROK [62]. The Great Mountains, such as the Sobaek and Taebaek Mountains that cross the ROK from the northeast to southwest may play a major role as potent genetic barriers. These two groups have different genetic properties, such that the SG is genetically more diverse and has a larger number of private alleles and effective population size than the NG. A total of four kdr alleles including one wild type and three mutant types were observed in this study. The three types of alleles (TTG L1014, TTT L1014F, and TGT L1014C) were observed in both the NG and the SG. Whereas the TTC L1014F allele, which was the rarest allele and which was detected for the first time in An. sinensis s.s. in this study, was only detected in the SG (AD, GR, and YY in Figure 1 and Table 2 and 3).

In summary, the different allele statuses of kdr genes in the An. sinensis complex suggest the importance of monitoring insecticide susceptibilities and the resistance of target vectors in control programmes. Effective resistance monitoring that is based on species-specific insecticide bioassay tests, molecular studies of allele diversity, origins of insecticide resistance, and minor resistance mechanisms (behavioural and cuticular resistance) will be crucial for building successful malaria vector control programmes that can explain and predict the development and spread of insecticide resistance traits.

Conclusions

This is the first report that has explored the presence and absence of 1014 codon mutations in the VGSC of the six malaria-vector species in the ROK. The highly polymorphic 1014 mutations were only observed in An. sinensis s.s., while the other five species showed no mutations. In An. sinensis s.s., the rarest allele, L1014F (TTC), was detected for the first time in this study, and the distribution of this allele was restricted to the southeast of the ROK. The present study is the first step in exploring species-specific insecticide resistance mechanisms. Further in-depth experiments are needed to prove the current epidemiological and evolutionary dynamics of malaria vectors in the ROK. In addition, these efforts will be one of the prerequisite steps needed to establish effective long-term vector control strategies in the ROK.