Mutations Linked to Insecticide Resistance Not Detected in Ace-1 or VGSC Genes in Nyssorhynchus darlingi from Multiple Localities in Amazonian Brazil and Peru

1. Introduction

Indoor residual spray (IRS) and more recently, long-lasting insecticidal nets (LLINs), have been used as the primary malaria vector control interventions due to their cost-effectiveness and protection against vectors that feed and rest mainly indoors [1,2]. In contrast, in Latin America, many vectors feed and rest primarily outdoors (exophagy and exophily, respectively). This behavior, together with the widespread existence of houses with incomplete walls, communities where unprotected travel-related occupations are common [3,4,5], and temporally irregular application of IRS in many communities diminish the effectiveness of insecticide-based interventions throughout Latin America in relation to some other malaria endemic regions [6,7].

In 2007, the World Health Organization (WHO) recommendations for LLIN distribution were broadened to include all individuals in malaria endemic areas, rather than solely pregnant women and children under the age of 5 [8]. Simultaneously, WHO guidelines called for malaria endemic nations to adopt triennial LLIN mass distribution campaigns which would allocate one LLIN for every two people per household in a given region, aspiring to universal coverage [9]. Following these updates, mass distribution campaigns became more prevalent in Latin America, particularly in the Amazon region. In Peru, the Project for Malaria Control in Andean Border Areas (PAMAFRO) was mainly responsible for a 70% decrease in cases from 2005-2011 [9]. Since the program was discontinued, malaria cases have increased fairly steadily due to a decline in international and domestic funding [10]. The sharpest increase in cases throughout the Latin American region occurred between 2014-2017, when case incidence nearly doubled [11]; ironically, incidence has apparently decreased with the recent COVID-19 pandemic [12]. An encouraging sign is that the Ministry Health, Peru, has adopted a new plan as of January 2022 to eliminate malaria by 2030 [13].

Between 2010-2019, an estimated 28 malaria endemic countries (of 82) have detected insecticide resistance to all four of the most commonly used insecticides (pyrethroids, organochlorines, carbamates, organophosphates), and nearly all (73/82) have reported resistance to at least one class [14,15]. Worldwide, the exposure of mosquitoes to any of these classes of insecticides [16], whether for public health or agricultural use, has the potential to be a strong selective force that favors the survival of resistant populations [17]. In Latin America there is relatively little entomological surveillance and few published studies on IR, except for Nyssorhynchus albimanus [18]. The primary malaria vector in the Amazon, Ny. darlingi, is generally susceptible to insecticides [17,19] and exhibits exo- and endophagic behavior, depending on local circumstances [20]. Nevertheless, resistance based on bioassays using WHO paper bioassays [21] or CDC bottle bioassays [22,23] has been reported for Ny. darlingi to pyrethroids and carbamates (Bolivia), pyrethroids (Brazil), pyrethroids and organochlorines (Colombia), and pyrethroids, carbamates and organophosphates (Peru) [24].

Insecticide resistance in mosquitoes can be derived from several mechanisms including behavioral (plasticity) [25], metabolic (enzymes that metabolize (detoxify) or sequester insecticides) [26,27], and target-site (genotype) [28,29]. Of these, only target-site – considered to be the most accurate indicator of resistance – can be identified using molecular assays [30]. These assays detect amino acid substitutions that give rise to non-synonymous amino acid changes in insecticide targets, ultimately preventing the insecticide from binding, leading to resistance [31]. Among the genes of interest are voltage-gated sodium channel (VGSC) para and acetylcholinesterase-1 (Ace-1), which encode for target binding of pyrethroids/organochlorines and carbamates/organophosphates, respectively [29,31]. Mutations linked most frequently with IR are L995F and L995S in VGSC [32] and G280S in Ace-1 [33]. Other mutations include L995C, L995W, V991L, and V994S in VGSC [33] and, in Ace-1, A221T and S216T [18]. Codon 995 in VGSC and codon 280 in Ace-1 were referred to formerly as 1014 and 119, respectively [18]. Despite the accuracy of molecular assays, susceptibility (or phenotypic) bioassays are advantageous because hundreds of mosquitoes can be tested simultaneously with relatively simple equipment [24,30]. Having data from both molecular assays and phenotypic bioassays is ideal for assessing resistance frequency [31].

The combination of resurgent cases of malaria since 2014 and the widespread increase in the use of insecticides, yet scarce reporting of resistant vectors, suggest a potential knowledge gap in IR throughout the Latin American region. IR detection requires strategic selection of localities in which mosquito samples are collected and analyzed. Commonly used selection criteria from previous reports include accessibility by land or water, history of insecticide use, high malaria prevalence, and a sample size sufficient even for a population with low frequency of resistant vectors [34,35]. This study aims to help close the knowledge gap in Ny. darlingi by using molecular assays for a relatively large sample size covering multiple localities, uncovering novel codon mutations, and providing insight that may assist in improving malaria vector control interventions.

2. Materials and Methods

2.1. Sample Collection

Samples of Ny. darlingi were collected in several rural and riverine localities throughout Amazonian Brazil and Peru in 2014 – 2021 (Table 1; Figure 1). Sample locations were based on information from local health officials about exposure to IRS or LLINs. In Cruzeiro do Sul, Acre, Brazil, local public health personnel (Brazil Ministry of Health and Instituto Evandro Chagas) demonstrated phenotypic resistance via discriminating concentration bioassays to deltamethrin and cypermethrin from 2012-2014 [36]. In the villages of Gamitanacocha and Zungarococha, Peru, local health authorities (Laboratorio Referencial de Salud Pública Loreto) demonstrated phenotypic resistance and possible resistance, respectively, to pyrethroids in 2018 [36]. Furthermore, Zungarococha and Cahuide are both along the highway where extensive IRS was used during a major malaria outbreak in 2012 [10]. Samples used for genetic analysis were selected randomly from the available specimens in storage using a random number generator [37]. As year-round samples from Cahuide were available, we selected specimens from both the rainy and dry seasons [38].

Mosquito collections were performed indoors, peridomestically, or in forest edge settings during various time periods throughout the evening using human landing catch (HLC), barrier screen sampling (BSS), or Shannon traps (ST) [38,39,40]. Details on forest cover level [39], field collection protocols [38,40], and P. vivax malaria incidence [40] have been reported previously. Genomic DNA was extracted from whole mosquitoes using the Qiagen DNeasy Blood & Tissue Kit (Qiagen, Germantown, MD).

2.2. Molecular Analysis

Amplification of a 228 bp fragment of the kdr region of the VGSC gene, between exons 20 and 21, was performed in a 20 𝜇L PCR mixture containing 1.0 – 15 ng/𝜇L DNA template, 1x AllTaq Master Mix (Qiagen), and 0.5 𝜇M of each primer: AAKDRF2 and AAKDRR2, with the same cycling conditions as in [16]. Amplification of a 456 bp fragment of Exon 2 in the Ace-1 gene was performed in a 25 𝜇L PCR mixture containing 1.0 – 15 ng/𝜇L DNA template [40], 2 U of Taq DNA Polymerase (Qiagen), 10x PCR Buffer (Qiagen) containing 1.5 𝜇M MgCl₂, a supplemental 0.5 𝜇M MgCl₂ totaling 2.0 𝜇M of MgCl₂, 0.2 𝜇M dNTPs, and 0.4 𝜇M of each primer ACE1DAF and ACE1DAR, with the same cycling conditions as in [18].

Samples were Sanger sequenced in forward and reverse directions at the Wadsworth Center Advanced Genomic Technologies Core (New York State Department of Health). Chromatograms of each sample were cleaned, converted into consensus sequences, translated, and exported to FASTA files using Geneious Prime Version 2020.2 [41]. Consensus sequences were aligned using ClustalW in MegaX Version 10.1.7, then analyzed in comparison to existing VGSC and Ace-1 sequences of Ny. darlingi from GenBank [16]. Unique sequences for VGSC were deposited on GenBank under accession numbers: OR260704 - OR260712 and Ace-1 under accession numbers: OR260713 - OR260857. Sequences for both genes were examined for non-synonymous (amino acid change) and synonymous (no amino acid change) point mutations, with special focus on documented codons known to convey insecticide resistance within these gene regions.

3. Results

3.1. Detection of Voltage-Gated Sodium Channel (VGSC) Mutations

A total of 495 wild-caught Ny. darlingi, 296 from Peru and 199 from Brazil, were successfully sequenced for the kdr target-site resistance region of the VGSC gene (Table 1). Sequences were aligned with the aid of Ny. darlingi VGSC sequences available on Genbank [16]. We detected nine unique genotypes (denoted as V1 – V9, corresponding to Genbank IDs OR260704 - OR260712), with only the susceptible genotype TTA (leucine) observed at codon position 995 (Figure 2). The only point mutations detected were in the intron and primer regions.

3.2. Detection of Acetylcholinesterase-1 (Ace-1) Mutations

We successfully sequenced 493 Ny. darlingi, 295 from Peru and 198 from Brazil, for the Ace-1 gene (Table 1) and detected 145 unique genotypes (named AC1-AC145, corresponding to Genbank IDs OR260713 - OR260857). The susceptible genotype GGG/GGG or GGG/GGC (glycine) was observed at codon 280 across all samples. Three non-synonymous mutations were detected in samples from three populations (Humaitá and Mâncio Lima, Brazil; Cahuide, Peru) within other regions of the Ace-1 gene not known to convey insecticide resistance (Figure 3).

3.3. Data Management

Unique sequences of Ny. darlingi for VGSC were deposited on GenBank under accession numbers: OR260704 - OR260712 and for Ace-1 under accession numbers: OR260713 - OR260857. All sample results will be deposited in VectorBase pending publication.

4. Discussion

Molecular analysis of a ~228 bp fragment that encodes for the kdr target-site resistance region of the VGSC gene and a ~456 bp fragment of the Ace-1 gene did not detect any non-synonymous mutations in the specimens of Ny. darlingi from endemic malaria areas of Brazil and Peru in the current study, similar to results of a recent analysis of Ny. darlingi from locations in Brazil (Manaus, Unini River, Jau River in Amazonas State and Porto Velho, Rondônia state) and Colombia (Tagachi, Chocó Department) [42]. An investigation of specimens of Ny. darlingi from Chocó Department, Colombia, that had been demonstrated to be phenotypically susceptible and resistant, sequenced for the same region of the VGSC gene, also did not reveal any kdr mutations [16]. However, the classic L1014F kdr mutation has been detected in other important anopheline malaria vectors i.e., Ny. albimanus [16,43,44], and Ny. albitarsis s.s. [45]. Other species of Latin American malaria vectors evaluated with these molecular assays include An. vestitipennis and An. pseudopunctipennis, both of which exhibited genotypic susceptibility [16]. The lack of genotypic evidence of IR in Ny. darlingi could be a reflection of limited regional data rather than the absence of resistance [46]. On the other hand, Floch [47] suggested that frequent reintroduction of wild susceptible populations of Ny. darlingi from forest into village populations could reduce selection for insecticide resistance. This hypothesis received some support from a study in the Porto Velho area in Rondônia state, Brazil that detected seasonal gene flow between forested and urban populations of Ny. darlingi [48].

Several previous studies of Anopheles malaria vectors attribute a proportion of recent malaria case resurgence to increased outdoor biting and insecticide resistance (IR) following mass distribution campaigns [49,50,51,52], although there is no evidence in support of this latter trend in Ny. darlingi across Latin America (scale-up of LLIN distribution has been limited compared with Africa) or after completion of the intensive PAMAFRO project in Peru. Even though daytime biting behavior in members of the An. gambiae complex has been hypothesized to increase transmission [53], in Latin America there has been scant investigation of this phenomenon except for observations of Ny. darlingi biting during the day in forested French Guiana malaria hotspots [54].

The hot, rainy climate of the Amazon basin is optimal for mosquito habitats [55]; however, anthropogenic landscape changes – namely, forest fragmentation and increased ecotone density – suggest that vector behavior (mainly exophagy and exophily in Ny. darlingi) and distribution (i.e., along ecotones for Ny. darlingi) in one location may not be generalizable to an entire region [56,57]. For example, during a recent malaria outbreak in French Guiana, Ny. darlingi was the only anopheline collected both outdoors and indoors, and its abundance was exceptionally high, possibly attributed to regional deforestation, and/or the higher than average rainy and dry seasons in 2017 [58]. Nyssorhynchus darlingi has also been collected biting during the day in French Guiana forest [59,60] and along the Maroni River, a former malaria hotspot, in Suriname [61]. This appears to be a focal phenomenon in Ny. darlingi, perhaps a behavioural avoidance response to IRS or LLINs.

The majority of IR reports in Latin America are based on bioassay data from the Amazon Basin or Central America [24,31]. However, recent genotypic reporting of several Brazilian samples of Ny. albitarsis s.s. showed heterozygous L995F mutations in VGSC [45], and a sample of Guatemalan Ny. albimanus had a heterozygous G280S mutation in Ace-1 [62]. Both these reports infer that agricultural insecticide use was the driver of IR, and a recent review of the contribution of agricultural insecticides and increasing insecticide resistance in malaria vectors found a strong association across Africa, that could be affected by crop type (especially cotton and vegetables), urban development, and strategies undertaken for farm pest management [63]. Questionnaires and insecticide susceptibility bioassays utilized in a field study in two South Côte d’Ivoire communities determined that local mosquito vectors were resistant to three of four insecticides tested and the authors highlighted the need for collaboration between the public health and agricultural sectors to develop interventions that would benefit both [64]. Resistance in the important regional malaria vector, Ny. albimanus, has been detected in Central America, Panama and northwestern coastal Peru, linked mainly to agriculture in general and rice cultivation in coastal Peru in particular [17].

Public health insecticide use can exert comparable selective pressure on malaria vectors [65] including, for example, the organophosphate malathion used in Brazilian public health for the arboviral vectors Aedes aegypti and Aedes albopictus to reduce transmission of viruses such as dengue, chikungunya and Zika [66]. Resistance in the vector Culex quinquefasciatus in Brazil has been detected for organophosphates, carbamate, DDT, pyrethroids and biolarvicides; the concern for such resistance to arise in malaria vectors in Brazil where they co-occur with Culex quinquefasciatus is limited to Fortaleza, Ceará state and parts of Mato Grosso state [67].

For control of adult mosquitoes, IRS on interior house walls will kill resting mosquitoes; some also repel mosquitoes such that they modify their behavior and rest outdoors [17]. Based on an evaluation of the residual effects of four insecticides (deltamethrin, pyrethroids, lambda-cyhalothrin, and etofenprox) on a range of wall materials in Amazonian Brazil [68], the Brazilian National Malaria Control Plan has been consistently using etofenprox PM 20% for residual spray in houses since 2013 [19], although a similar study of six insecticides in Amapá, Brazil by Correa et al. [69] found that deltamethrin WG at 0.025 gm/m² had the highest residual effects. In Peru, the pyrethroid deltamethrin 5% is the most commonly applied insecticide for IRS [38].

Other approaches to tackling insecticide resistance consider biological control in general [70] or the replacement of synthetic compounds with plant-based compounds formulated as bioinsecticides, reviewed in Demirak and Canpolat [15]. The classes of compounds described and discussed are phytochemicals, pheromones, microbial pesticides, and plant-incorporated protectants (PIPs); selected candidate compounds demonstrate larvicidal, adulticidal and repellent properties. In general, their advantages are, compared with synthetic compounds, lower toxicity, target-specificity, highly effectivity in small quantities, and they are biodegradable. Despite considerable promise, these products remain in various stages of development, and have not yet been field-tested for use against malaria vectors. As many target insects have evolved successful resistance mechanisms to most classes of insecticide, the evolution of different modes of action against plant-based insecticides could temper the early enthusiasm for such novel products [26].