https://orcid.org/0000-0003-3117-2457
Figures
Abstract
Culex pipiens complex is an important vector of epizootic and zoonotic pathogens, including West Nile virus. Chicago, Illinois and its suburbs have suffered high incidence of human West Nile virus infections in the past. This makes abatement programs in and around the Chicago area an essential service. The control of Cx. pipiens is often complicated by rapidly evolving resistance to pyrethroids, which are the most widely used chemical class in US mosquito abatement programs. The present study assessed Sumithrin® resistance in Cx. pipiens collected from five locations around Cook County, Illinois, neighboring the city limits of Chicago. According to CDC guidelines, samples from all five locations demonstrated some resistance to Sumithrin®. When assessed with Anvil®, a formulated product made of Sumithrin® synergized with piperonyl butoxide, susceptibility was rescued in mosquitoes from three out of the five locations, suggesting involvement of mixed-function oxidases and/or carboxylesterases in Sumithrin® resistance at these locations. Not all locations had susceptibility rescued by Anvil®, but these locations had relatively low knockdown resistance allele frequencies, suggesting that mechanisms other than knockdown resistance may be involved. Enzyme activities did not reveal any marked trends that could be related back to mortality in the bottle bioassays, which highlights the need for multiple types of assays to infer enzymatic involvement in resistance. Future directions in pyrethroid resistance management in Chicago area Cx. pipiens are discussed.
Citation: Burgess ER IV, Lopez K, Irwin P, Jaeger CP, Estep AS (2022) Assessing pyrethroid resistance status in the Culex pipiens complex (Diptera: Culicidae) from the northwest suburbs of Chicago, Illinois using Cox regression of bottle bioassays and other detection tools. PLoS ONE 17(6): e0268205. https://doi.org/10.1371/journal.pone.0268205
Editor: Ahmed Ibrahim Hasaballah, Al-Azhar University, EGYPT
Received: April 21, 2022; Accepted: June 10, 2022; Published: June 29, 2022
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This manuscript was supported in part by Cooperative Agreement Number U01CK000505, funded by the Centers for Disease Control and Prevention to PI and KL. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The Culex pipiens complex (Cx. pipiens) plays a major role in vectoring several epizootic and zoonotic pathogens significant to birds and humans, including West Nile virus (WNV) and St. Louis encephalitis [1, 2]. WNV is considered the most widespread arbovirus in the United States [3]. In the summer of 2002, cases of WNV in Chicago surged, with 884 cases and 66 deaths, followed by another outbreak in 2003 but with only 53 cases [4]. Chicago resides in Cook County, Illinois, and is home to an estimated 5.1 million people. Factors related to geography, housing, population, and abatement strategies around Chicago and the surrounding county makes this a potential high-risk region for WNV cases [5].
Typically thought to feed primarily on birds, Cx. pipiens from the upper Midwest are aggressively anthropophagic [6]. The Cx. pipiens complex is currently comprised of Cx. australicus, Cx. pipiens, and Cx. quinquefasciatus, with various hybridizations and biotypes among the species, two of which are known as form pallens and molestus [7]. Form molestus is present in Chicago [8] but does not appear to be significantly introgressed into these Cx. pipiens populations [9]. The molestus form is important because it appears to be more strongly associated with anthropophagy, increasing its potential to transmit pathogens such as WNV [10, 11]. Mosquitoes from the Cx. pipiens complex are primarily active at night, which makes them particularly good targets for nightly insecticidal control efforts using ultra-low volume (ULV) fogging [12, 13]. ULV fogging remains the most competent tool available for quickly reducing WNV and other mosquito vector populations during times of arboviral disease outbreaks [14].
ULV products typically contain one of two classes of insecticides, either pyrethroids or rarely organophosphates. Pyrethroids are generally safer than organophosphates to vertebrates, as they break down into safe metabolites relatively quickly and they are cheap. But resistance to pyrethroids has been well-documented in Cx. pipiens throughout the United States (reviewed in [15]). Among the likeliest of the proposed physiological mechanisms that confer this resistance are target site mutations in the para region of the voltage-sensitive sodium channel and metabolic detoxification by a few families of enzymes, including mixed-function oxidases (MFO) known as cytochrome P450s, carboxylesterases (CarE), and glutathione S-transferases (GST). How these mechanisms combine to confer resistance is not well-understood but nevertheless should be monitored by mosquito control operations to better understand their relationship to product failure.
In the present study, CDC bottle bioassays were conducted as an initial screening for resistance to Sumithrin® (d-phenothrin), a Type I pyrethroid and Anvil® (Sumithrin® plus piperonyl butoxide, a synergist) in Cx. pipiens from four sites in the northwest suburbs of Cook County, Illinois. These sites were in Wheeling, Arlington Heights, and two sites in Des Plaines. As a post hoc laboratory analysis, organisms from the bottle bioassays were genotyped for single nucleotide polymorphisms in the voltage-sensitive sodium channel that confers knockdown resistance (kdr). Mosquitoes from these samples also were tested for enzymatic activity. Finally, bottle bioassay data was analyzed using clustered Cox regression using time-dependent covariates to account for non-proportional hazards when they occurred [16]. This approach to bottle bioassay data analysis provides additional information on the rate of mortality between start and endpoint. Using Cox regression in this way also is beneficial because non-proportional hazards due to heterogeneity of resistance factors in recently field-derived strains can be accounted for in this type of model.
Methods
Mosquito sources and pyrethroid exposure histories
Five study sites were selected within the Northwest Mosquito Abatement District, located in Cook County, IL, USA (Fig 1 and Table 1). Each site is approximately 2.59 km2 and located 0.8 km to 8 km from other sites. Within each site, four homeowners allowed mosquito collections with one CDC gravid trap (John W. Hock Company, Gainesville, FL). Gravid traps were baited with an infusion of 75 g Timothy Complete rabbit food (Kaytee Products Inc., Chilton, WI), 58 g Kentucky bluegrass (Poa pratensis) collected from the district’s property, 2.5 g lactalbumin whey protein (TGS Nutrition, Las Vegas, NV), 0.22 g Altosid® pellets (Central Life Sciences, Schaumburg, IL), and 9.46 L water. Gravid infusion was stored outside for at least one week before use. Cx. pipiens egg rafts were collected from gravid trap basins every other week between the months of July and August in 2020. To ensure the correct species was collected, egg rafts were hatched and reared individually until species could be identified at second instar [17]. Mosquitoes from the same site and collection day were pooled together and reared at 27°C and in a 16:8 hour light:dark cycle. Larvae were fed with ground TetraMin® tropical fish flakes (Spectrum Pet Brands LLC, Blacksburg, VA) and adults were fed a 10% sucrose solution.
A = WHE (Wheeling), B = AHB (Arlington Heights South Catch Basin), C = DPN (Des Plaines North), D = DPS and DPSB (Des Plaines South and Des Plaines South Catch Basin).
Sprays were conducted once a week for 5 weeks in Wheeling, Arlington Heights North, and Des Plaines South in 2019 and 2020, starting in July and ending in August. In 2019, Zenivex® E20 (20% etofenprox) in a 1:1 mix with mineral oil (10% etofenprox) was used and in 2020, Anvil® 10+10 was sprayed at 0.0036 Lb per acre of active ingredient and piperonyl butoxide (PBO). From 2013–2018 Northwest Mosquito Abatement District averaged 1 spray event (etofenprox) per year in these areas. The sites where the mosquitoes were obtained are in residential neighborhoods with numerous parks, schools, and manicured lawns. Numerous residential mosquito/pest control companies operate in these residential areas. Pyrethroids are commonly used by homeowners, park districts, schools, and golf courses [18].
Bottle bioassays
CDC bottle bioassays were completed using three different insecticide solutions on adult Cx. pipiens aged 3–6 days [19] (Table 1). Most bottle bioassays were completed with technical grade Sumithrin® (provided in the CDC bottle bioassay kit) diluted with acetone to 20 μg/bottle, the diagnostic dose used in the bottle bioassay guidelines. The diagnostic time for Sumithrin® in Cx. pipiens adults is 30 minutes. This time and concentration represent empirically determined parameters that are specific for this species to display 100% mortality when fully susceptible ([19], section 3.2). The second solution consisted of Anvil® 10+10 (Clarke® Mosquito Control Products, Inc., Roselle, IL) diluted to 22.2 μg Sumithrin® and 22.2 μg PBO per bottle [20]. This dose of Sumithrin® is similar to the diagnostic dose in the guidelines. The second Anvil® 10+10 solution was made with the same protocol, instead using the 1:1 mineral oil tank mix (abbreviated AnvilTM) diluted with acetone to 11.1 μg Sumithrin® and 11.1 μg PBO per bottle. 250 mL glass Wheaton bottles were treated with 1 mL insecticide solution according to the CDC procedure and allowed to dry for at least four hours. Insecticide solutions were stored at 4°C. Control bottles were treated with 1 mL acetone and allowed to dry for four hours. 15–25 mixed sex Cx. pipiens were aspirated into each bottle and knockdown was recorded every 5 min for 45 min, then every 15 min until 120 min total. At the completion of the bioassay, mosquitoes were killed and stored at –80°C.
Genotyping for knockdown resistance alleles
Genotyping for the 982L and 982F alleles, canonically known as 1014L and 1014F alleles, was conducted using a PCR-based melt curve assay modelled on the assay of Saavedra-Rodriquez et al. [21] but with primers designed for the Cx. pipiens complex (Table 2). Controls for the LL, FF, and the LF genotypes, from mosquitoes that had been previously genotyped by Sanger sequencing, were included to ensure that the assay reliably detected all three genotypes common to the US. No template (negative) controls were also included. Individual organisms collected from the CDC bottle bioassays were loaded into 96-well plates (Omni International, Kennesaw, GA) with 400 μL of nuclease free water and cubic zirconium beads (BioSpec Products, Bartlesville, OK). Plates were sealed with Teflon™ sealing mats and homogenized for 60 seconds at 30 hertz (Omni International, Kennesaw, GA). Samples were centrifuged for 1 min at 805g and then maintained on ice until assay setup. PCR master mix was prepared in sufficient quantity for 400 10 μL reactions (2,000 μL SYBR Select, 1,161.2 μL nuclease free water, 2.4 μL of Cxq_1014L primer, 13.2 μL of Cxq_1014F primer, 10.0 μL of Cxq_1014S primer, and 13.2 μL of Cxq_1014_3’ primer). Eight microliters of mastermix was added to each well of a 384-well plate (Thermo Fisher Scientific, Waltham, MA) followed by 2 μL of centrifuged homogenate using an epMotion 5075 liquid handling system (Eppendorf, Hamburg, Germany) with filtered tips. Amplification and melt curve data were collected on an Applied Biosystems QS6 (Thermo Fisher Scientific, Waltham, MA) using default fast cycling conditions. Determination of alleles present in a sample was assessed by examination of the derivative melt curve for temperature peaks (Tm) as in [21] (Fig 2). In this assay, an LL genotype is characterized by a distinct Tm of 85.3 ± 0.5°C, an FF by a Tm of 82.6 ± 0.5°C, and an LF heterozygote has peaks at both Tms (Fig 2). Samples that did not amplify or that amplified with a cycle threshold greater than 35 were excluded from analysis.
Representative melt curve assay results for a 1014L homozygote (LL), a 1014F homozygote (FF), a 1014LF heterozygote (LF), and a nuclease-free water blank (NFW) showing distinct melting temperatures for the L (85.3°C) and F (82.6°C) alleles.
Enzyme activity assays
The content of mixed-function oxidases (MFO), and the activity of α- and β-carboxylesterases (α-, β-CarE), and glutathione S-transferases (GST) were determined using modified methods from [22] using only freeze-killed control mosquitoes from the bottle bioassays (i.e., mosquitoes that were not exposed to any toxicant). Where multiple bottle bioassays were done on the same strain within a week (i.e., the same generation), enzyme content and activities were assessed on only one set of controls from that week. Because sample size for the controls varied between nine and thirty individuals, eight individual mosquitoes were used from each strain. To produce the enzyme source, mosquitoes were placed into 2.0 mL screw cap microcentrifuge tubes loaded with two 2.0 mm zirconium oxide beads and 400 μL of 100 mM sodium phosphate buffer (pH 7.4) with no chelating agents or protease inhibitors. Mosquitoes were homogenized with a bead mill and then centrifuged at 10,000 x g at 4°C for 10 min. The resulting supernatant was used as the enzyme source for all subsequent activity assays as well as for soluble protein determination.
The MFO content in samples was determined using the 3,3’,5,5’-tetramethylbenzidine (TMB) method originally devised by [23]. This assay does not measure the activity of MFOs toward a model substrate. Instead, it measures the heme content in the sample, the majority of which is thought to belong to MFOs [24]. A 0.2% TMB solution was freshly prepared in methanol and then further diluted to 0.05% in 250 mM sodium acetate buffer (pH 5.0). A 20 μL volume of supernatant was added to a 96-well plate in duplicate followed by 200 μL of 0.05% TMB solution. The reaction was started by adding 25 μL of 3% H2O2 to each well and then incubating the plate at room temperature for 10 min inside a dark cabinet. Results were compared to a standard curve generated from cytochrome C using the same reagents and volumes as the samples (R2 = 0.997). After the incubation period, wells were read at 620 nm in a BioTek Epoch 2 spectrophotometer (BioTek, Santa Clara, CA). Units of activity were thus reported as cytochrome equivalents.
For α- and β-CarE activity, 15 μL of supernatant was incubated with 135 μL of freshly prepared 0.3 mM α- or β-naphthyl acetate (final reaction concentration 0.27 mM in wells) in duplicate in a 96-well plate covered with a lid. Incubation lasted for 15 min at room temperature inside a dark cabinet. The reaction was stopped by adding 50 μL of freshly prepared 0.3% Fast Blue B in 5.0% sodium dodecyl sulphate solution. Color was allowed to develop for 5 min at room temperature in the dark cabinet. The α-naphthyl acetate (α-CarE) samples were read at 600 nm and the β-naphthyl acetate (β-CarE) samples were read at 550 nm. Standard curves of α- and β-naphthol were run in triplicate (both R2 = 0.999) and used to quantify the enzymatic conversion of α- and β-naphthyl acetate to α- and β-naphthol, respectively.
For GST activity, 20 μL of supernatant was added in duplicate to wells of a 96-well plate. The substrate mixture consisted of 1 mL of 21 mM 1-chloro-2,4-dinitrobenzene (CDNB) in methanol added to 10 mL of freshly prepared 10 mM of reduced glutathione (GSH) in 100 mM sodium phosphate buffer (pH 6.5) immediately prior to being mixed with the supernatant. A pH of 6.5 was used for the substrate mixture to minimize auto conjugation of GSH with CDNB. 180 μL of substrate mixture was added to each well of supernatant for a final concentration of 0.9 mM CDNB and 0.86 mM GSH. The plate was read at 340 nm in 1 min intervals for 5 min. An experimentally derived extinction coefficient of 0.00580 μM-1 cm-1 was used that accounted for the path length through the 200 μL total volume in each well.
Total protein was measured with the Bradford method using bovine serum albumin as a standard (R2 = 0.978) [25]. Units for all four enzyme assays were MFO content as μg cytochrome c equivalents/mg protein, and CarE and GST specific activities as nmol substrate/min/mg protein.
Statistical analyses
All analyses were conducted in R version 4.1.1 [26]. Enzyme activities were compared by Kruskal-Wallis tests due to this test being robust to smaller sample sizes. When applicable, a Mann-Whitney-U test was used for pairwise comparisons between the strains. Two analyses were done, one that included DPN1 and another that included DPN2. This was done to simplify the analyses because DPN1 and DPN2 are repeated measures of the same strain, thus not independent, and all other strains were independent. Each enzyme activity assay also was regressed against percent mortality at both 30 min and at 120 min and the slopes were used to determine an effect of enzyme activity on percent mortality. This was done by building general linear models, which were visually assessed for heteroscedasticity and normality of residuals.
For CDC bottle bioassays, a clustered Cox regression was generated on pairwise comparisons of Sumithrin® and Anvil® or Sumithrin® and AnvilTM at the 30 min diagnostic time point used by the CDC to assess resistance [19] using the ‘survival’ package [16]. The clustering effect was assigned to the multiple bottles for each of the treatments. Prior to statistical testing, the Cox models were assessed for proportionality using the ‘cox.zph()’ function. Models that violated the assumption of proportionality had a time-dependent coefficient added to them (i.e., an interaction term between time in minutes and treatment), and these hazard ratios are reported with a time factor change in hazard rate of mortality. For ties in times to death, the Efron approximation was used. A Wald test was used to test the null hypothesis that the beta coefficients = 0 at α = 0.05 and results are reported as fold change in the hazard rate of mortality in the Anvil® or AnvilTM (both Sumithrin® + PBO) treatment compared to the Sumithrin® treatment. Bottle bioassays were analyzed only if at least three replicates of each treatment were done per location. Thus, analyses of AHB, WHE, and DPN were included.
Results
Bottle bioassays
The distribution of observed mortality over 30 min differed in several strains when comparing Sumithrin® technical (20 μg/bottle) against either Anvil® at 22.17 μg Sumithrin® and 22.17 μg PBO or an Anvil® tank mix (AnvilTM) equivalent with 11.09 μg Sumithrin® and 11.09 μg PBO in 1:1 mineral oil (Table 3). Mortality reached 100% in the Anvil® treatment by the diagnostic time of 30 min but was 70% against the technical Sumithrin®. There was a significant difference between Sumithrin® and Anvil® in AHB mosquitoes (Fig 3; HR = 3.85 (2.496–5.949 95% CI), P < 0.001). There was a 3.9-fold change in hazard rate of mortality when mosquitoes were exposed to the Anvil® treatment over the course of 30 min.
Top panel: common mortality curve. Bottom panel: Kaplan-Meier survival curves and 95% confidence intervals (shaded area). Vertical dotted line at 30 minutes denotes the diagnostic time for Sumithrin®.
Mortality also reached 100% in Anvil® at the diagnostic time in WHE. The hazard ratio was not proportional over time by a factor of 0.91 (Fig 4; HR = 0.91 (0.841–0.992), P = 0.032), and there was a significant difference between Sumithrin® and Anvil® (HR = 36.11 (8.465–154.007 95% CI), P < 0.001). Thus, there was a 36.1-fold change in hazard risk of mortality in the Anvil® treatment at zero min that changed by 0.91 every minute (i.e., 36.1 x 0.91n, where n = minutes). At 30 min, the hazard rate of mortality in the Anvil® treatment was about a 2.1-fold compared to Sumithrin® alone. The hazard also was not proportional over time between Sumithrin® and AnvilTM at 30 min (HR = 0.91 (0.858–0.970), P = 0.003). The hazard ratio significantly differed between Sumithrin® and AnvilTM (HR = 17.33 (4.707–63.817), P < 0.001). There was a 17.3-fold change in hazard rate of mortality in the AnvilTM treatment starting at zero min and changed by 0.91 every minute for 30 min. By 30 min, the hazard rate of mortality was about double in the Anvil® treatment.
Top panel: common mortality curve. Bottom panel: Kaplan-Meier survival curves and 95% confidence intervals (shaded area). Vertical dotted line at 30 minutes denotes the diagnostic time for Sumithrin.
There was no difference in hazard rate of mortality between Sumithrin® and Anvil® in the DPN samples (Fig 5; HR = 1.26 (0.605–2.629), P = 0.537).
Top panel: common mortality curve. Bottom panel: Kaplan-Meier survival curves and 95% confidence intervals (shaded area). Vertical dotted line at 30 minutes denotes the diagnostic time for Sumithrin.
Genotyping for knockdown resistance alleles
All five strains tested were positive to varying degrees for the 1014F allele (Table 4). The DPS strain had the highest percent homozygosity for the kdr genotype (FF), as well as the highest frequency for the 1014F allele. No homozygotes for the susceptible genotype (LL) were identified among the tested DPS organisms. The DPN strain that was genotyped was from a collection earlier than those used in the bottle bioassay and enzyme activity assays and had the lowest percent homozygosity of the FF genotype, as well as the lowest 1014F allele frequency. Heterozygosity (genotype LF) was generally high compared to both homozygous genotypes in all five strains.
Enzyme activity assays
The analysis was first done with all strains including DPN1 but not DPN2 and then all strains including DPN2 but not DPN1. Finally, DPN1 and DPN2 were compared. This was done because DPN1 and DPN2 were repeated measures of the DPN strain and thus not independent. Activity of GSTs were only significantly different when DPN2 was included in the group (Fig 6; χ2 = 10.59, df = 3, P = 0.014). AHB was significantly different from DPS (W = 54, P = 0.021) and DPN2 (W = 55, P = 0.015) and WHE was significantly different from DPN2 (W = 10, P = 0.021). No other enzyme activities or MFO content were significantly different among the strains. There also was no difference in any of the enzyme activities or MFO content between DPN1 and DPN2 (all P > 0.130). The coefficient of variation for α-CarE and β-CarE in the AHB strain were the two highest among all tested strains and enzyme activities (Table 5). There was no effect of enzyme activity or MFO content that explained the differences in percent mortality at either 30 min or at 120 min (i.e., all regression slopes were statistically equal to zero).
The lower and upper hinge of boxes represent the 25th and 75th quartiles, respectively. The middle line represents the median. The whiskers represent the minimum and maximum range of the data.
Larger numbers indicate greater variability relative to their mean.
Discussion
According to CDC guidelines [19], all five strains in the present study may be resistant to Sumithrin®, and the formulated product Anvil® reclaimed efficacy in three out of five strains. This is not surprising given the frequency of ULV applications in these areas during the time the mosquito collections were made. Previous studies have demonstrated resistance in both adult [27] and larval [28, 29] Cx. pipiens in the suburbs of Chicago. With historically high incidence of WNV in the Chicago area when compared to the rest of the upper Midwest and northeast US where Culex pipiens and Culex restuans are the main vector. [30], monitoring resistance status in Cx. pipiens is of public health importance. Knowing what mechanisms these strains rely on to survive chemical control will help to inform mosquito abatement districts on application timing and product rotation.
Numerous resistance mechanisms have been implicated in Cx. pipiens (reviewed in [15]). Although Culex spp. are known to have a leucine-to-phenylalanine or leucine-to-serine single nucleotide polymorphism (SNP) at the 982nd amino acid in the para region of the voltage-sensitive sodium channel (structurally the same location as the L1014F/H mutation in house flies), their roles in resistance to pyrethroids is acknowledged but not well-resolved (e.g., [31, 32]). In the case of the leucine-to-serine mutation, there appears to be stronger resistance to DDT than to pyrethroids [33]. The house fly L1014F mutation is only known to confer moderate levels of resistance by itself [34] compared to bi- and tri-allelic kdr combinations [35]. In the case of the Cx. pipiens complex, hybridization rate may also play a role in expression of kdr [31]. In the present study, kdr alone does not appear to explain the relatively large differences in mortality to Sumithrin® at 30 and 120 min but does appear to have some involvement. The DPS strain had the lowest mortality (10.5%) at the diagnostic time of 30 min and had the greatest resistant phenylalanine (F) allele frequency, but WHE had similar allele frequency and resulted in over 40% greater mortality (51.4%) over 30 min. WHE had 100% mortality with Anvil® treatment, suggesting significant involvement of MFOs and/or CarEs.
Tests utilizing synergists in the CDC bottle bioassay can suggest enzymatic detoxification playing a role in resistance [36, 37], especially in the presence of high L982F kdr allele frequencies [38]. That we saw considerable restoration of susceptibility when these strains were tested with Anvil®, a Sumithrin® product that is synergized with PBO, is consistent with previous findings (e.g., [38]). Traditionally, PBO is thought to inhibit cytochrome P450s (aka MFOs), a key enzyme in many oxidative processes, including insecticide metabolism ([39]). But PBO also has been shown to inhibit carboxylesterases [40]. A correlative relationship with enzyme activity and resistance to deltamethrin, permethrin, and DDT in Cx. quinquefasciatus, a close relative to Cx. pipiens, has been previously shown [41]. Xu et al. [42] found over 1,000-fold difference in LC50 in two resistant strains of Cx. quinquefasciatus that were selected with permethrin. PBO contributed significant reduction in LC50 values, down to around 100-fold, of the resistant strains compared to the susceptible strain. Although methodology is different in the present study, the results of the bottle bioassays are consistent with PBO contributing significant restoration of Sumithrin® efficacy, suggesting that either cytochrome P450s and/or carboxylesterases play a role in the resistance phenotype among these strains. Similar restoration of susceptibility was recently demonstrated in populations of Cx. quinquefasciatus from Florida and California [36, 37].
In the case of the MFO activity assay, 3’,3,5’,5’-TMB is a substrate used to measure heme content in a sample, of which most is thought to be attributed to cytochrome P450s [23]. This assay is considered a surrogate assay and does not measure activity directly as in other types of activity assays, such as the CarE and GST assays in the present study. Although regressions of bottle bioassay mortality did not suggest a relationship based on MFO quantity, CarE, or GST activities in the control mosquitoes, it does not rule out altered enzyme activity toward Sumithrin®. It also does not rule out inducible overexpression of these enzymes, especially MFOs, which have been documented in resistant Cx. quinquefasciatus [43]. This is a plausible explanation for the discrepancy between enzyme activities and PBO synergism because induction usually happens shortly after initiation of exposure to a toxicant, and enzyme activities conducted in the present study were done on control mosquitoes (i.e., not exposed to Sumithrin®). Regardless of the dynamics of cytochrome P450 expression, PBO tends to reduce pyrethroid LD50s even in lab-reared, fully susceptible Culex spp. (e.g., [37]), and Musca domestica (e.g., [44]).
Although limited in what they can tell us about enzymatic involvement in resistance, an application of enzyme activity assays that may be useful in describing the dynamics of the enzymatic role in resistance is in an induction experiment. In an induction experiment, the researcher exposes the animals to a sublethal dose of the insecticide and collects individuals over a time series. The researcher then runs activity assays on the specimens and relates the activities back to toxicological response data. The assay in the present study only had the potential to detect constitutively expressed detoxification enzymes. The goal of a future study in Cx. pipiens complex adults from the Chicago area should include an induction experiment with Sumithrin®. That the synergized formulation of Sumithrin® and PBO in Anvil® restored susceptibility in multiple strains suggests that expression of MFOs and/or CarEs may be induced at high levels upon exposure to a toxicant like Sumithrin®, or perhaps even have a target site mutation that confers an advantage in processing Sumithrin®. Another explanation could be that the model substrates used in the present study simply do not interact with the isozymes that may play a role in Cx. pipiens complex resistance to Sumithrin®. This could also be true of any induction experiment undertaken in future studies.
We note that in the AHB strain, there was a high coefficient of variation in the CarE activities relative to the other strains. The inclusion of PBO with Sumithrin® appeared to provide satisfactory control in this strain at the time of data collection and should be considered in future regional control efforts. WHE had notable resistance to Sumithrin® but similarly the addition of PBO in Anvil® provided 100% mortality at the diagnostic time of 30 min. AHB and WHE had the greatest difference in mortality at 30 min diagnostic period between Sumithrin® and Anvil® (30% and 48.6%, respectively). The DPS strain showed the highest resistance according to CDC diagnostic dose and time for Sumithrin®. DPS also showed the highest F allele frequency and the highest percentage of resistant FF homozygotes. DPS was not screened with Anvil® and there was no indication of enzymatic involvement based on the activity assays. A follow up study should focus on the DPS strain and include an Anvil® bottle bioassay to measure an effect, if any, that MFOs and/or CarEs have on its resistance. Interestingly, neither DPSB nor two of the DPN resamples reverted to 100% mortality when treated with Anvil®. By the second resampling (DPN2), resistance to Sumithrin® increased by 11%.
Examined from the perspective of hazard ratios, in the case of AHB and WHE, we saw that treatment with Anvil® resulted in a 3.9-fold and 2.1-fold change in the risk of mortality at 30 min compared to Sumithrin®. In DPN, the raw bottle bioassay mortality data shows only a small increase in mortality in the Anvil® treatment, which statistically conferred no greater risk of mortality than the Sumithrin® treatment. With the inability to calculate informative numbers such as resistance and synergist ratios (e.g., [45, 46]) using diagnostic doses in bottle bioassays, hazard ratios from Cox regression offer the potential to make ratio-based quantitative comparisons among treatments. In the present study, the choice to analyze bottle bioassay data using Cox regression was made after bottle bioassays had already been run and some potentially important factors such as sex and genotypes were not collected. The inclusion of these types of factors would help to explain results like these in future studies. We recommend the use of clustered Cox regression with time-dependent covariates when describing how much risk multiple variables contribute to the rate of mortality across a given diagnostic time period. Typical analyses of bottle bioassays, including binomial generalized linear models and repeated measures ANOVA, violate several important assumptions of these statistical tests, including the independence of observations [47], which seldom allow for the correct grouping variables to account for this fact. This can be accounted for using a clustering effect in Cox regression, which is akin to a random factor in a mixed model. Quite often, studies that utilize Kaplan-Meier survival analysis are not assessed for the assumption of proportional hazards, a key assumption that must be met for model estimates to be deemed accurate [16]. This is evident by the often-seen crossing of survival curves in a figure (e.g., Fig 4). We have provided code for readers to use on future analyses of bottle bioassay data using Cox regression. Cox regression is a benefit to the interpretation of bottle bioassay data because it allows multiple factors to be assessed simultaneously across numerous time points, not just a single endpoint. We liken these two statistical comparisons to taking a picture (end-point analysis) versus watching a movie (Cox regression analysis). For instance, in Fig 3 (WHB) mortality in the Anvil® and AnvilTM treatments were in excess of 75% within the first 10 and 15 min, respectively, at which point an inflection in the rate of mortality can be seen. With downstream molecular and biochemical analyses, this point of inflection could be characterized, with Cox regression as the statistical method to infer risk of mortality due to those types of factors. Future studies on Cx. pipiens resistance in field strains, where topical application of pesticides is not feasible, should incorporate this type of analysis to make clearer distinctions among the numerous resistance mechanisms and their relative impact on operational success.
Supporting information
S1 File. R-script for Cox regression survival analysis of bioassay data.
https://doi.org/10.1371/journal.pone.0268205.s001
(PDF)
Acknowledgments
The authors thank Skyler Finucane, Natalia Szklaruk, Jim Downing, Colin Murphy, and Jack Ponterelli for assistance with mosquito rearing and bottle bioassays.
Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services or the USDA. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity provider and employer.
References
- 1. Gottlieb S. West Nile virus detected in mosquitoes in Central Park. B World Health Organ. 2000;78(9):1168-. PubMed PMID: WOS:000089263900021.
- 2. Tsai TF, Mitchell CJ. St. louis encephalitis. The arboviruses: epidemiology and ecology. 1989;4:113–43.
- 3. Kuehn BM. Shifting Trends in West Nile and Other Arboviral Diseases. JAMA. 2021;326(12):1140. pmid:34581727.
- 4.
Illinois Department of Public Health. WNV Surveillance 2021. Available from: https://dph.illinois.gov/topics-services/diseases-and-conditions/west-nile-virus/surveillance.html].
- 5. Ruiz MO, Tedesco C, McTighe TJ, Austin C, Kitron U. Environmental and social determinants of human risk during a West Nile virus outbreak in the greater Chicago area, 2002. Int J Health Geogr. 2004;3(1):8. Epub 20040420. pmid:15099399; PubMed Central PMCID: PMC420251.
- 6. Hamer GL, Kitron UD, Brawn JD, Loss SR, Ruiz MO, Goldberg TL, et al. Culex pipiens (Diptera: Culicidae): a bridge vector of West Nile virus to humans. J Med Entomol. 2008;45(1):125–8. pmid:18283952.
- 7. Harbach RE. Culex pipiens: species versus species complex taxonomic history and perspective. J Am Mosq Control Assoc. 2012;28(4 Suppl):10–23. pmid:23401941.
- 8. Mutebi JP, Savage HM. Discovery of Culex pipiens pipiens form molestus in Chicago. J Am Mosq Control Assoc. 2009;25(4):500–3. pmid:20099597.
- 9. Kothera L, Mutebi JP, Kenney JL, Saxton-Shaw K, Ward MP, Savage HM. Bloodmeal, Host Selection, and Genetic Admixture Analyses of Culex pipiens Complex (Diptera: Culicidae) Mosquitoes in Chicago, IL. J Med Entomol. 2020;57(1):78–87. pmid:31576405.
- 10. Huang S, Hamer GL, Molaei G, Walker ED, Goldberg TL, Kitron UD, et al. Genetic variation associated with mammalian feeding in Culex pipiens from a West Nile virus epidemic region in Chicago, Illinois. Vector Borne Zoonotic Dis. 2009;9(6):637–42. pmid:19281434.
- 11. Kilpatrick AM, Kramer LD, Jones MJ, Marra PP, Daszak P, Fonseca DM. Genetic influences on mosquito feeding behavior and the emergence of zoonotic pathogens. Am J Trop Med Hyg. 2007;77(4):667–71. pmid:17978068.
- 12. Hickner PV, Mori A, Rund SSC, Sheppard AD, Cunningham JM, Chadee DD, et al. QTL Determining Diel Flight Activity in Male Culex pipiens Mosquitoes. J Hered. 2019;110(3):310–20. pmid:30668763; PubMed Central PMCID: PMC6503456.
- 13. Veronesi R, Gentile G, Carrieri M, Maccagnani B, Stermieri L, Bellini R. Seasonal pattern of daily activity of Aedes caspius, Aedes detritus, Culex modestus, and Culex pipiens in the Po Delta of northern Italy and significance for vector-borne disease risk assessment. J Vector Ecol. 2012;37(1):49–61. pmid:22548536.
- 14.
American Mosquito Control Association. Best practices for integrated mosquito management. Sacramento, California, USA. 2021.
- 15. Scott JG, Yoshimizu MH, Kasai S. Pyrethroid resistance in Culex pipiens mosquitoes. Pestic Biochem Physiol. 2015;120:68–76. Epub 20141219. pmid:25987223.
- 16.
Therneau T. A package for survival analysis in R-_. R package version 3.2–11 2021. Available from: https://CRAN.R-project.org/package=survival].
- 17.
Darsie RF, Ward RA, Chang CC. Identification and geographical distribution of the mosquitoes of North America, north of Mexico. Fresno, Calif.: American Mosquito Control Association; 1981. 313 p. p.
- 18. Kuivila KM, Hladik ML, Ingersoll CG, Kemble NE, Moran PW, Calhoun DL, et al. Occurrence and potential sources of pyrethroid insecticides in stream sediments from seven US metropolitan areas. Environmental science & technology. 2012;46(8):4297–303.
- 19.
McAllister J, Scott M, Control CfD, Prevention. CONUS manual for evaluating insecticide resistance in mosquitoes using the CDC bottle bioassay kit. CDC, Atlanta, GA. 2020.
- 20.
Petersen J. Measuring Insecticide Resistance by the Bottle Bioassay. Florida Mosquito Control Handbook. 2004:1–10.
- 21. Saavedra-Rodriguez K, Urdaneta-Marquez L, Rajatileka S, Moulton M, Flores AE, Fernandez-Salas I, et al. A mutation in the voltage-gated sodium channel gene associated with pyrethroid resistance in Latin American Aedes aegypti. Insect Mol Biol. 2007;16(6):785–98. pmid:18093007.
- 22.
Valle D, Montella I, Ribeiro R, Viana-Medeiros P, Martins Jr A, Lima J. Quantification methodology for enzyme activity related to insecticide resistance in Aedes aegypti. Fundação Oswaldo Cruz and Secretaria de Vigilância em Saúde, Ministério da Saúde, Rio de Janeiro and Distrito Federal, Brazil. 2006.
- 23. Brogdon WG, McAllister JC, Vulule J. Heme peroxidase activity measured in single mosquitoes identifies individuals expressing an elevated oxidase for insecticide resistance. J Am Mosq Control Assoc. 1997;13(3):233–7. pmid:9383763.
- 24. Hemingway J, Karunaratne SH. Mosquito carboxylesterases: a review of the molecular biology and biochemistry of a major insecticide resistance mechanism. Med Vet Entomol. 1998;12(1):1–12. pmid:9513933.
- 25. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;72:248–54. pmid:942051.
- 26.
R Core Team. R: A language and environment for statistical computing 2021 [cited Vienna, Austria]. Available from: https://www.R-project.org].
- 27. Clifton ME, Xamplas CP, Nasci RS, Harbison J. Gravid Culex pipiens Exhibit A Reduced Susceptibility to Ultra-Low Volume Adult Control Treatments Under Field Conditions. J Am Mosq Control Assoc. 2019;35(4):267–78. pmid:31922942.
- 28. Harbison JE, Layden JE, Xamplas C, Zazra D, Henry M, Ruiz MO. Observed loss and ineffectiveness of mosquito larvicides applied to catch basins in the northern suburbs of chicago IL, 2014. Environ Health Insights. 2015;9:1–5. Epub 20150421. pmid:25987841; PubMed Central PMCID: PMC4406279.
- 29. Harbison JE, Sinacore JM, Henry M, Xamplas C, Dugas LR, Ruiz MO. Identification of larvicide-resistant catch basins from three years of larvicide trials in a suburb of chicago, IL. Environ Health Insights. 2014;8(Suppl 2):1–7. Epub 20141013. pmid:25392699; PubMed Central PMCID: PMC4216650.
- 30.
Centers for Disease Control and Prevention (CDC). Average annual incidence of West Nile virus neuroinvasive disease reported to CDC by county, 1999–2020 2020. Available from: https://www.cdc.gov/westnile/statsmaps/cumMapsData.html#five].
- 31. Bkhache M, Tmimi FZ, Charafeddine O, Faraj C, Failloux AB, Sarih M. First report of L1014F-kdr mutation in Culex pipiens complex from Morocco. Parasit Vectors. 2016;9(1):644. Epub 20161216. pmid:27986090; PubMed Central PMCID: PMC5159952.
- 32. Zhou L, Lawrence GG, Vineis JH, McAllister JC, Wirtz RA, Brogdon WG. Detection of broadly distributed sodium channel alleles characteristic of insect pyrethroid resistance in West Nile virus vector Culex pipiens complex mosquitoes in the United States. J Med Entomol. 2009;46(2):321–7. pmid:19351083.
- 33. Martinez-Torres D, Foster SP, Field LM, Devonshire AL, Williamson MS. A sodium channel point mutation is associated with resistance to DDT and pyrethroid insecticides in the peach-potato aphid, Myzus persicae (Sulzer) (Hemiptera: Aphididae). Insect Mol Biol. 1999;8(3):339–46. pmid:10469251.
- 34. Sun H, Tong KP, Kasai S, Scott JG. Overcoming super-knock down resistance (super-kdr) mediated resistance: multi-halogenated benzyl pyrethroids are more toxic to super-kdr than kdr house flies. Insect Mol Biol. 2016;25(2):126–37. Epub 20151222. pmid:26691197.
- 35. Sun H, Kasai S, Scott JG. Two novel house fly Vssc mutations, D600N and T929I, give rise to new insecticide resistance alleles. Pestic Biochem Physiol. 2017;143:116–21. Epub 20170824. pmid:29183579.
- 36. Lucas KJ, Bales RB, McCoy K, Weldon C. Oxidase, Esterase, and KDR-Associated Pyrethroid Resistance in Culex quinquefasciatus Field Collections of Collier County, Florida. J Am Mosq Control Assoc. 2020;36(1):22–32. pmid:32497474.
- 37. McAbee RD, Kang KD, Stanich MA, Christiansen JA, Wheelock CE, Inman AD, et al. Pyrethroid tolerance in Culex pipiens pipiens var molestus from Marin County, California. Pest Manag Sci. 2004;60(4):359–68. pmid:15119598.
- 38. Lee HJ, Longnecker M, Calkins TL, Renfro AD, Fredregill CL, Debboun M, et al. Detection of the Nav channel kdr-like mutation and modeling of factors affecting survivorship of Culex quinquefasciatus mosquitoes from six areas of Harris County (Houston), Texas, after permethrin field-cage tests. PLoS Negl Trop Dis. 2020;14(11):e0008860. Epub 20201119. pmid:33211688; PubMed Central PMCID: PMC7714350.
- 39. Sun Y-P, Johnson E. Synergistic and antagonistic actions of insecticide-synergist combinations and their mode of action. Journal of Agricultural and Food Chemistry. 1960;8(4):261–6.
- 40. Khot AC, Bingham G, Field LM, Moores GD. A novel assay reveals the blockade of esterases by piperonyl butoxide. Pest Manag Sci. 2008;64(11):1139–42. pmid:18481337.
- 41. Sarkar M, Bhattacharyya IK, Borkotoki A, Goswami D, Rabha B, Baruah I, et al. Insecticide resistance and detoxifying enzyme activity in the principal bancroftian filariasis vector, Culex quinquefasciatus, in northeastern India. Med Vet Entomol. 2009;23(2):122–31. pmid:19493193.
- 42. Xu Q, Liu H, Zhang L, Liu N. Resistance in the mosquito, Culex quinquefasciatus, and possible mechanisms for resistance. Pest Manag Sci. 2005;61(11):1096–102. pmid:16032654.
- 43. Gong Y, Li T, Zhang L, Gao X, Liu N. Permethrin induction of multiple cytochrome P450 genes in insecticide resistant mosquitoes, Culex quinquefasciatus. Int J Biol Sci. 2013;9(9):863–71. Epub 20130905. pmid:24155662; PubMed Central PMCID: PMC3805894.
- 44. Kasai S, Scott JG. Overexpression of cytochrome P450 CYP6D1 is associated with monooxygenase-mediated pyrethroid resistance in house flies from Georgia. Pesticide Biochemistry and Physiology. 2000;68(1):34–41.
- 45. Ahmed MA, Vogel CF. Synergistic action of octopamine receptor agonists on the activity of selected novel insecticides for control of dengue vector Aedes aegypti (Diptera: Culicidae) mosquito. Pestic Biochem Physiol. 2015;120:51–6. Epub 20150122. pmid:25987220.
- 46. Estep AS, Sanscrainte ND, Waits CM, Louton JE, Becnel JJ. Resistance Status and Resistance Mechanisms in a Strain of Aedes aegypti (Diptera: Culicidae) From Puerto Rico. J Med Entomol. 2017;54(6):1643–8. Epub 2017/10/06. pmid:28981681.
- 47.
Zar JH. Biostatistical analysis. 5th ed. Upper Saddle River, N.J.: Prentice-Hall/Pearson; 2010. xiii, 944 p. p.