The African malaria mosquito, Anopheles gambiae
Giles (Diptera: Culicidae), is the most efficient vector of malaria in Sub-Saharan Africa [1
], and insecticide treated bed nets (ITNs) and/or indoor residual spraying (IRS) are used to decrease its populations. Reliance on chemical insecticides has resulted in widespread insecticide resistance to at least two insecticide classes [2
] and is a continuous factor impeding the success of malaria elimination efforts [1
]. Approximately 80% of the countries endemic with An. gambiae
report resistance to at least a single class of insecticide, and more than 60% of these countries reported resistance to two or more insecticide classes [1
]. Protection of military personnel from arthropod-vectored diseases is often achieved with the use of insecticide-treated combat and work uniforms [3
], and permethrin, a Type-I pyrethroid insecticide, has long been approved for this use. The U.S. Environmental Protection Agency (EPA) has recently approved the so-called pseudo-pyrethroid etofenprox, which is a non-ester containing pyrethroid, for use on military uniforms [5
According to the Insecticide Resistance Action Committee’s (IRAC’s) mode of action classification system, pyrethrum and pyrethroid insecticides (Group 3) modify the insect voltage-sensitive sodium channel (VSSC) [6
]. VSSCs are transmembrane proteins that function in the movement of sodium ions into the cell, resulting in membrane depolarization during an action potential. Disruption of the VSSC by pyrethroids alters channel function by delaying inactivation; this action has been extensively studied and recently reviewed [7
]. Different classes of pyrethroid insecticides differently affect the VSSC causing depolarization of resting membrane potential, and/or repetitive nerve firing [9
Arthropods have evolved mechanisms to decrease the toxicity of pyrethroid insecticides. Mechanisms of pyrethroid resistance include increased metabolism of the insecticide by cytochrome P450-monooxygenases, general esterases, glutathione S
-transferases, and/or target site modification resulting in reduced sensitivity [2
]. Target site modification is characterized as the genotypic modification to the amino acid sequence resulting in an altered phenotype. For pyrethroid insecticides, genotypic modification of the VSSC results in a knockdown resistant (kdr) phenotype. In An. gambiae
, a leucine to phenylalanine replacement at amino acid position 1014 (L1014F) in the para
-type sodium channel is commonly associated with pyrethroid resistance [15
]. Additional target site mutations have been reported in the VSSC from resistant An. gambiae
populations and other insects [16
The goal of this study was to characterize the toxicity of permethrin and etofenprox and determine the level of resistance in the An. gambiae
(Akdr; MRA-1280) strain, which is known to carry the L1014F mutation (and perhaps other mechanisms) resulting in pyrethroid resistance [18
]. This study is an extension of previous work [2
] with an emphasis on documenting the resistance observed with etofenprox. The characterization of etofenprox is particularly important based on the recent approval of its use on military uniforms [5
2. Materials and Methods
An. gambiae were reared from eggs obtained from pyrethroid susceptible (G3, MRA-112) and pyrethroid-resistant (Akron-kdr (Akdr), MRA-1280) colonies maintained by the Malaria Research and Reference Reagent Resource Center (MR4), part of the Biodefense and Emerging Infections (BEI) Research Resources Repository at the Center for Disease Control and Prevention (CDC), Atlanta, GA, USA. Research with An. gambiae at the University of Florida was approved by the Florida Department of Agriculture and Consumer Services (FDACS permit #10-33), where An. gambiae were contained under BSL3/ACL3 conditions at the Emerging Pathogens Institute. Fourth-instar larvae of the MRA-1280 strain were selected with 1 ppm permethrin for 24-h and survivors used to maintain the resistant colony (personal communication with Paul Howell, 9 September 2015). An. gambiae eggs from either colony were placed into unfiltered tap water and fed a 2% (w/v) Brewer’s yeast suspension (MP Biomedicals, LLC, Santa Ana, CA, USA) for the first 12–24-h. First-instar larval density was between 500–1000 eggs in 500 mL of tap water. Larvae were then split into several pans at a density of 100–200 larvae per 1.2 L of tap water. The remaining larval life stages were fed pulverized beta-fish food (Spectrum Brands Holdings, Inc., Madison, WI, USA). Pupae were collected and placed into waxed paper containers with a mesh lid, where freshly emerged adults were provided a 10% sucrose water (w/v) solution via a soaked cotton ball ad libitum. All life stages were maintained in Percival incubators at 28 ± 2 °C with a relative humidity greater than 70% on a light: dark (12:12 h) photoperiod.
2.2. Chemicals and Chemical Preparation
Technical grade permethrin was purchased from Chem Services Inc. (West Chester, PA, USA). Etofenprox (99%, Pestanal® analytical standard), piperonyl butoxide (PBO; 99%, Pestanal® analytical standard), and diethyl maleate (DEM; >96%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Ethanol (>99%) was purchased from Fisher Scientific (Hampton, NH, USA). All test compounds and synergists were initially diluted into >99% ethanol (Fisher Scientific) before being serially diluted to the desired concentration (all in ethanol).
2.3. Larval Toxicity Assays
First-instar mortality bioassays with intact larvae were performed as previously described with minor modifications in Reference [19
]. Control mortality or paralysis of all larval assays at all life stages was 10% or less. If greater than 10% control mortality was observed, the results were not used for analysis. Ten first-instar larvae (12–24-h post-hatch) were added to each well of a 24-well plate with 995 μL of tap water (containing 0.04% yeast suspension), followed by the addition of the test compound (5 μL) dissolved in ethanol (final ethanol concentration was 0.5%). To ensure even mixing of the treatments, the Petri dishes were swirled in clockwise and counterclockwise motions, along with front-and-back and side-to-side motions ten times. Wells containing ethanol only (0.5%) served as the negative control. Larval moribundity was determined 24-h post-treatment. Larvae that showed no movement after manual disturbance of the water by a pipette tip were scored as “dead”.
Fourth-instar (intact) An. gambiae larval bioassays were performed in a similar fashion as the first-instar bioassays, but used a 35-mm Petri dish containing 4.75 mL of tap water prior to the addition of 25 μL of the test compound dissolved in ethanol. The final concentration of ethanol in all treatments and the negative control was 0.5%. Contents of the wells were mixed as previously described. Larval mortality was determined 24-h after the addition of test compounds or solvent control. Larvae that did not move when manually disturbed were recorded as dead.
The paralytic effect of the test compounds was examined using a previously described headless larval assay [20
] that was adapted for An. gambiae.
Briefly, the heads of fourth-instar larvae were removed using forceps before being transferred to a 35-mm Petri dish containing 4.75 mL of mosquito larval physiological saline, as in References [20
]. Test compounds, dissolved in ethanol (25 μL), were added to the Petri dish and mixed as previously described. Five-hours post-treatment, the paralytic effect of the test compounds and control was measured. Paralysis was scored if there was no movement or slight twitching after larvae were probed with an insect pin.
2.4. Adult Topical Biological Assays
Non-blood fed adult female An. gambiae mosquitoes (1–5 days post-emergent) were anesthetized on ice, and treatments were applied (volume of 0.2 μL) to the mosquito’s thorax using a Hamilton ten-microliter gas-tight syringe equipped with a repeating dispenser (Hamilton Company, Reno, NV, USA). Ethanol (>99%) alone was used as the negative control. Synergism experiments were performed in a similar manner, where mosquitoes were dosed with piperonyl butoxide (PBO) or diethyl maleate (DEM) at 100 ng/mosquito; mosquito mortality at these synergist doses was less than 20%. Mosquitoes that survived the synergistic treatment were again anesthetized on ice before being topically treated with the pyrethroid insecticide or solvent control four-hours after the synergist treatment. The 1-h knockdown and 24-h moribundity were then determined. Knockdown (KD) was defined as mosquitoes that have erratic and uncontrollable flight or do not maintain proper posture when at rest; moribund mosquitoes were also included in this measurement at 1-h. Mosquito moribundity/mortality at 24-h was defined as no movement or only slight leg movements when the container was agitated.
2.5. Data Analysis
The half-maximal response for knockdown (KD50), lethality (LD50 or LC50), and paralysis (PC50) was calculated using the PROC PROBIT procedure in SAS 9.4 (SAS Institute, Carey, NC, USA). A minimum of five concentrations were tested per compound to generate a KD50, LD50, LC50, or PC50 and a minimum of three replicates were performed. A replicate was defined as treated mosquitoes obtained from different rearing cohorts. Statistical differences in knockdown, toxicity, or paralysis was performed by using individual determinants (LD50, LC50, or PC50) using Graphpad Prism 7.03 (La Jolla, CA, USA). First, data underwent logarithmic transformation then were examined to ensure that they fit a Gaussian distribution (Shapiro-Wilk normality test), before performing an unpaired t-test (α = 0.05). Resistance ratios (RR50) were calculated by dividing the KD50, LD50, LC50, or PC50 from the MRA-1280 strain by the corresponding values generated by the susceptible G3 (MRA-112) strain. Synergistic ratios (SR50) were calculated by dividing the LD50 obtained without synergist by the LD50 obtained with a synergist pre-treatment.
The objective of this study was to examine the toxicity and the level of resistance of two pyrethroid insecticides against two laboratory strains of An. gambiae
at three life stages (first-instar larvae, fourth-instar larvae, and adult female). Previous characterization of the An. gambiae
WHO Akron strain (MRA-913) was performed using permethrin, deltamethrin, etofenprox, and DDT. Resistance was observed for all of these VSSC-acting insecticides except etofenprox, which had a RR50
of 1.4 [2
]; these data were perplexing and needed further examination. It’s important to note the differences between the WHO Akron (MRA-913) and Akdr
(MRA-1280) strains. These strains were isolated in the Akron District of Porto Novo, Benin (Africa). MRA-913 resistance to carbamate insecticides (phenotype) is a result of a genotype modification in the acetylcholinesterase enzyme (ACE-1 mutation) [2
] and is selected in the laboratory with bendiocarb at the adult stage (personal communication with Paul Howell, 9 September 2015). Whereas, the MRA-1280 is selected with permethrin in the larval stage (previously stated in Materials and Methods). The selection in the laboratory is the only difference between these two An. gambiae
strains; both strains display carbamate and pyrethroid resistance.
The first-instar An. gambiae
larval bioassay was developed to evaluate the toxicity of insecticides [19
]. The benefit of this assay is that it has relatively high-throughput, eliminating the need to rear mosquitoes to older larvae or adults. Additionally, due to the small size of first-instar larvae a smaller water volume was used; therefore, less test compound was needed to perform the assay. This is an important advantage in an insecticide discovery program, where the amount of the test compounds may be limited. The caveat to this first-instar bioassay was that these small larvae were more susceptible to xenobiotics and may not provide an accurate prediction of mortality for later instars or adults. This effect was recently highlighted by comparing the toxicity of fluralaner to Aedes aegypti
mosquito larvae [23
]. Fluralaner’s toxicity to Ae. aegypti
fourth-instar larvae was found to be 1.8 ppb [23
]; however, a previous study reported greater than 90% mortality to first instar Ae. Aegypti
larvae at 1.2 ppt [24
], a greater than 1500-fold difference in toxicity between the two life stages. While we did not observe such a drastic difference between life stages for permethrin or etofenprox (Table 1
and Table 2
), we did see differences between the larval life stages. For instance, in the G3 strain, permethrin was 31-fold more toxic to first instar larvae and etofenprox 8.4-fold more toxic to first instar larvae, when compared to fourth-instar (intact) larvae. Larger differences between the larval life stages were observed in the Akdr
strain, where permethrin and etofenprox had a 667-fold and an 86-fold difference between first-instar and fourth-instar larvae, respectively. These differences in toxicity need to be taken into consideration when using the first-instar larval assay.
The large difference in life stage susceptibility to permethrin and etofenprox indicates that physiological factors (size and weight) play a role; this was previously reported in Culex quinquefasciatus,
where the larval instar correlated with susceptibility to permethrin toxicity [25
]. One physiological factor that likely plays a role in the susceptibility to xenobiotics between life stages is the development of the cuticle. The importance of the cuticular barrier is demonstrated with the headless larval assay, which allows a direct diffusion pathway of test compounds; thereby facilitating penetration of the toxicant to exert its toxicodynamic effect without the need to cross the cuticular barrier. Curiously, the headless fourth-instar larvae had 5-h PC50
values (Table 3
) similar to the 24-h LC50
values obtained with first-instar larvae (Table 1
). These values were dramatically different from intact fourth-instar larvae (Table 2
). It is likely that there is thickening of the cuticle between life stages, but cuticular composition may also differ between susceptible and resistant strains. In adult An. gambiae
, changes in the expression of two P450 enzymes changed the cuticular hydrocarbon production on the cuticle of resistant adult mosquitoes [26
]. As a result, there was a decrease in the penetration of pyrethroid insecticides. However, it is not yet clear what form of cuticular changes might occur in An. gambiae
larvae, if any.
Topical application of permethrin and etofenprox was performed in adult female An. gambiae
mosquitoes. Permethrin’s toxicity aligned with previously reported topical data in the same species, with minor differences that were likely related to the health-status or rearing conditions of the colony [2
]. When An. gambiae
Akron mosquitoes are selected with permethrin, instead of bendiocarb, there appears to be slightly more resistance to pyrethroid insecticides [2
], which is not surprising. To-date, only the L1014F mutation has been characterized in either the WHO-Akron (MRA-913) or Akdr
(MRA-1280) strains [2
]. However, other mutations have been reported in the VSSC of field-collected An. gambiae
]. Intense selective pressure with insecticides on An. gambiae
mosquitoes will ultimately result in the development of multiple types of insecticide resistance [11
], including the potential for further mutations in the VSSC.
Previously we reported an increase in the general esterase and cytochrome P450 O
-deethylation activities of WHO Akron An. gambiae
(MRA-913), compared to G3. While these biochemical assays have yet to be performed in the Akdr
strain (MRA-1280), we did conduct toxicity assays using PBO (an inhibitor of P450 monooxygenases) and DEM (an inhibitor of glutathione S
-transferases; GSTs). Esterase inhibitor (e.g., S
-tributyl phosphorotrithioate) studies were not performed because esterase metabolism of etofenprox is unlikely, since it is a non-ester containing pyrethroid. Permethrin was synergized by PBO in the G3 strain and by PBO and DEM in the Akdr
strain. These results indicated that metabolism of permethrin by cytochrome P450s and GSTs were likely mechanisms involved in resistance. In G3 mosquitoes, PBO and DEM were able to synergize etofenprox significantly, but the effect was not significant in the Akdr strain. These results suggested that there was an increase in phase I and II metabolism in the wild-type strain that was not present in the resistant (Akdr
) strain. These results are baffling, since it would be expected that higher metabolic activities would be found in the resistant strain rather than the susceptible strain. Furthermore, PBO should synergize etofenprox toxicity, and this was previously reported in field-caught An. funestus
]. It has been reported that the lipophilic nature of the chemical synergist PBO enhances penetration of deltamethrin allowing it to reach its target site before being metabolized [29
]. However, the lack of PBO synergism with etofenprox in the Akdr
strain makes this mechanism unlikely. Other factors that could affect etofenprox’s cuticular penetration, such as changes in cuticular components or transporters (ABCs), may be involved and have been reported in resistant insects [30
]. Previously, the ZANDS An. gambiae
strain, which possesses organochloride resistance via elevated GST activity, was tested against etofenprox. Etofenprox did not display any resistance (RR50
0.89) in this mosquito strain, supporting the lack of DEM synergistic activity that we observed in the Akdr
strain. The WHO-Akron (MRA-913) strain displayed resistance to DDT which was attributed to target site insensitivity, since there was not an increase in GST activity in biochemical assays [2
]. Collectively, the known target site modification (L1014F) along with the studies performed with synergists and the headless larval assay indicate that multiple mechanisms (kdr
, metabolism, and cuticle thickness) are likely involved in pyrethroid resistance of the Akdr
strain; a similar conclusion made with the WHO Akron strain [2