Introduction

Top Summary Introduction Materials and Methods Results Discussion References

The prevalence of drug and pesticide use in modern society has led to numerous instances of genetic resistance in target organisms. This poses significant risks and challenges for control of infectious disease and agricultural pests (Collins and Paskewitz, 1995; Taylor and Feyereisen, 1996). Resistance development is an evolutionary process, arising through selection of low-frequency genes in the population that confer survival advantage. Of major importance are loci that promote decreased sensitivity to the drug/toxicant at the molecular site of action (Taylor and Feyereisen, 1996). Understanding the molecular bases for resistance can assist in devising effective strategies for mitigation and in the choice of suitable alternative chemicals.

Resistance to synthetic organic insecticides emerged soon after their introduction in the 1940s. However, elucidation of underlying molecular mechanisms has occurred only recently, aided by the revolution in molecular biology. Most insecticides are neurotoxins that modify the properties of ion channels, receptors, or enzymes involved in signaling within the nervous system. Variability in structural or regulatory genes for these targets provides a basis for selection of resistant individuals with target-site insensitivity (Taylor and Feyereisen, 1996). For example, resistance to cyclodiene insecticides arose from a single amino acid substitution in the neuronal -aminobutyric acid receptor channel pore (ffrench-Constant et al., 1993). Organophosphate resistance can result either from overproduction of acetylcholinesterase or from selection of structural isoforms with altered catalytic properties (Fournier et al., 1993). One form of pyrethroid resistance in temperature-sensitive Drosophila melanogaster mutants has been traced to the nap locus, a regulatory gene that controls sodium channel density (Kasbeker and Hall, 1988).

In one of earliest reports of insecticide resistance, insects previously exposed to dichlorodiphenyltrichloroethane (DDT) also exhibited resistance to pyrethroids (Busvine, 1951). The signature of this surprising cross-resistance was delayed paralysis attributed to nervous system insensitivity (Miller et al., 1979). The trait came to be known as "knockdown resistance", and two alleles were isolated: kdr and superkdr (Farnham, 1977; Sawicki, 1978). Both forms of resistance are linked to the para-homologous sodium-channel locus in houseflies (Williamson et al., 1993), and cross-resistance traits with similar linkage patterns occur in other insect species (Dong and Scott, 1994), including the tobacco budworm Heliothis virescens (Payne et al., 1988; Taylor et al., 1993). Sodium channels homologous to para of several resistant strains have been cloned, and a pattern of amino acid substitutions has emerged in two different S6 transmembrane segments and a cytoplasmic loop (Dong and Scott, 1994; Williamson et al., 1996; Park et al., 1997; Park and Taylor, 1997). These findings clearly indicate that the mechanism of knockdown resistance results from target-site insensitivity, specifically changes in sodium channel structure. They also confirm that sodium channels are a primary site of action for pyrethroids.

At least five mutations have now been associated with knockdown resistance originating in field populations of insects, but no information is yet available on the properties of sodium channels in neurons of resistant insects. In a preliminary report (Lee et al., 1996), we demonstrated changes in both pharmacological and biophysical properties of sodium channels in pyrethroid-resistant (Pyr-R) H. virescens. Since then, the resistance trait has been correlated with the amino acid substitution V421M in the I-S6 transmembrane segment of the hscp sodium channel (H. virescens sodium channel para homolog) (Park et al., 1997). This is the first association of a I-S6 mutation with altered sodium channel properties and pyrethroid resistance. Interestingly, this region of the channel also is at or near binding domains for site 2 and site 3 neurotoxins known to be allosterically coupled to pyrethroid binding (Trainer et al., 1996, 1997). Our evidence shows that sodium channels of Pyr-R insects have reduced sensitivity to pyrethroids as well as altered voltage-dependent gating properties, providing a compelling rationale for reduced nervous system sensitivity to pyrethroids.

	Materials and Methods

Top Summary Introduction Materials and Methods Results Discussion References

Neurotoxins

The synthetic pyrethroid permethrin (79% cis and 21% trans isomers; Chem Service Inc., West Chester, PA) was dissolved in dimethyl sulfoxide to make stock solutions (which were kept frozen) at concentrations of 10 and 100 mM. An aliquot of the stock solution was diluted in external recording solution (see below) to give secondary stock solutions of 10 or 100 µM. Working solutions were prepared by dilution of secondary stocks in external recording solution. Dimethyl sulfoxide was present in control and test solution at concentrations lower than 0.1% (v/v), which had no effect on sodium current. Tetrodotoxin (TTX) was purchased from Sigma Chemical Company (St. Louis, MO). LqhIT was kindly supplied by Dr. Eliahu Zlotkin (Hebrew University of Jerusalem, Jerusalem, Israel) and Dr. Michael Gurevitz (Tel Aviv University, Tel Aviv, Israel).

Experimental Animals

The Pyr-R H. virescens strain, maintained at the Department of Entomology, Clemson University, originated from ICI-82 and is descended from the resistant HSB strain, in which pyrethroid resistance was inherited as a partial dominant trait (Payne et al., 1988). Pyr-R was founded in 1987 by outcrossing HSB to the pyrethroid-susceptible strain Woodrow83 (Brown et al., 1996) and selecting hybrid larvae with permethrin applied topically at 14 µg/g. This was followed by six additional rounds of outcrossing Pyr-R to Woodrow83 and selecting F2 progeny with permethrin at 143 µg/g, a discriminating dose that sacrificed most hybrids. Pyr-R larvae were selected with permethrin at 714 µg/g in generations 24, 25, 29, 33, 34, 38, 39, 41, 44, 50, 56, 59, 65, 72, 78, 83, and 90. The average mortality in the 17 selections was 43 ± 13%. The higher selecting dose was sufficient to kill >99% of Pyr-R × Woodrow83 hybrids. In its current form, Pyr-R exhibits pyrethroid resistance linked to hscp in a Woodrow83 background (Payne et al., 1988; Taylor et al., 1993). Pupae of generations 82 through 88 and 90 were used for sodium channel analysis. In this study, we used adult moths (2-9 days after emergence) for all experiments.

Two control strains of H. virescens were used for comparison to Pyr-R: UCR-susceptible (UCR-S) and acetylcholinesterase-resistant (Ace-R). UCR-S is a pyrethroid-susceptible tobacco budworm H. virescens (UCR-S) that has been maintained at the University of California-Riverside for the past 15 years. The Woodrow83 strain could not be used for comparison because it was lost before this study, but in some experiments, Pyr-R was compared with the pyrethroid-susceptible strain Ace-R, which was produced by crossing Woodrow83 with Florence87 for seven generations (Gilbert et al., 1996).

Bioassay of Permethrin

The vial technique was used to quantify the in vivo toxicity of permethrin against adult (3-9 days postemergence) H. virescens. Permethrin was dissolved in pure acetone and deposited uniformly on the inner walls of 20-ml glass scintillation vials. Two insects were tested per vial. Susceptible (UCR-S and Ace-R) H. virescens were exposed to six dosages (0.06, 0.2, 0.6, 2, 6, and 10 µg/vial. Pyr-R moths were exposed to six dosages (2, 6, 10, 20, 60, and 200 µg/vial); 22 to 24 moths were tested at four dosages (6, 10, 20, and 60 µg/vial), and 10 moths were tested at the lowest (2 µg/vial) and highest (200 µg/vial) dosages. After a 24-h holding period at room temperature, percentage mortality was determined and corrected for control mortality. Animals were scored as positive for "mortality" if uncontrolled movement and/or loss of righting response were observed. Probit analysis was performed as described previously (Raymond, 1985).

Molecular Analysis of hscp Sodium Channel

Each of four individual specimens from the Pyr-R and UCR-S strains were treated with 5 µg of permethrin (Chem Service) using the vial technique. All susceptible specimens were sacrificed, whereas none of the resistant individuals were affected. DNA from each individual was extracted by the standard phenol-chloroform method. Two independent polymerase chain reactions (PCR) were performed for each individual specimen to obtain the nucleotide sequence in the region where mutations associated with pyrethroid resistance were identified (Park et al., 1997; Park and Taylor, 1997) (Fig. 2A). The PCR reactions were conducted with primers Nhp1375+ (5'-CCGAACCCTAACTACGGNTA-3') and IS6r2 (5'-CTGTTCCTCTTCTTCGGCTT-3') for determination of the polymorphism at V421M in I-S6, and primers Nhp3304+ (5' ATGTGGGACTGIATGTTGGT-3') and Nhp3448- (5'-CTGTTGAAGGCCTCTGCTAT-3') for determination of the polymorphism at L1029H in II-S6 according to reaction conditions described previously (Park et al., 1997; Park and Taylor, 1997). Controls for homozygotes, His/His, Leu/Leu, Met/Met, and Val/Val shown in Fig. 1 are RcaI or Hsp92II restrictions of the PCR product using a template of cloned DNA for respective sequences that were published elsewhere (Park et al. 1997). Controls for heterozygotes, His/Leu and Met/Val, are restrictions of PCR product using a template of gDNA of heterozygous individuals that were determined previously (Y. Park, unpublished observations). The hscp sequence was aligned with the D. melanogaster para sodium channel sequence, GenBank accession number M32078 (Loughney et al., 1989); numbering of amino acid positions followed the scheme used for the para channel.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   A, diagram showing the predicted hscp sodium channel structure. Sequence comparisons made between the Pyr-R and UCR-S strains (except for dashed areas at the amino and carboxyl termini) showed only one sequence polymorphism: V421M (arrow, B) in the I-S6 transmembrane segment (Park et al., 1997). Analysis of a different H. virescens collected from Louisiana revealed the L1029H mutation (arrow, C) in the II-S6 segment (Park and Taylor, 1997). B, PCR-RFLP using the restriction enzyme RcaI that cuts a methionine motif. Controls (Ctrls) for Met/Met homozygotes, Met/Val heterozygotes, and Val/Val homozygotes, and each of four individual specimens from Pyr-R and UCR-susceptible (UCR-S) strains are shown. C, PCR-RFLP using the restriction enzyme Hsp92II that cuts a histidine motif. Controls (Ctrls) for His/His homozygotes, His/Leu heterozygotes, and Leu/Leu homozygotes, and each four individuals of Pyr-R and UCR-S strains are shown.

The sequence for the Pyr-R strain was obtained in both directions from one individual specimen for the L1029H site and two specimens for the V421M site. The sequence for UCR-S was obtained as described previously (Park et al., 1997; Park and Taylor, 1997). PCRs coupled with restriction fragment length polymorphism analysis (RFLP) were conducted with Hsp92II and RcaI for L1029H and V421 M sites, respectively, as described previously (Park et al., 1997; Park and Taylor, 1997). For the V421M site, RcaI cut at the TCATGA motif was counted as methionine and uncut as valine (Fig. 2B). For the L1029H site, Hsp92II cut at the motif CATG was counted as histidine and uncut as leucine (Fig. 2C).

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Alignment of sodium channel sequences in the IS6 transmembrane segment. The V421M mutation associated with pyrethroid resistance in the Pyr-R strain and comparable positions in other sodium channels are shown in bold. GenBank accession numbers are listed in the right column.

Preparation of Central Neurons

Central neurons of adult H. virescens were prepared from dissociated thoracic and abdominal ganglia and cultured for 1 to 3 days. Ganglia were removed and desheathed in sterile, ice-cold Ca²⁺-free insect saline containing 100 mM NaCl, 4 mM KCl, 10 mM HEPES, 5 mM glucose, and 137 mM mannitol, pH 7.0 (modified from Hayashi and Levine, 1992). Desheathed ganglia were treated for 5 to 7 min (~37°C) with a mixture of 0.5 mg/ml collagenase (Type IA; Sigma) and 2 mg/ml dispase (Boehringer Mannheim, Indianapolis, IN) in Ca⁺⁺-free saline. After the ganglia were washed and transferred to a sterile hood, they were gently triturated in L-15 Leibovitz culture medium (GIBCO, Grand Island, NY) supplemented with 700 mg/liter glucose, 400 mg/liter fructose, 60 mg/liter succinate, 3000 mg/liter TC yeastolate, 2800 mg/liter lactalbumin hydrolysate, 60 mg/liter imidazole, 100 µg/ml streptomycin, 100 units/ml penicillin, and 1 µg/ml 20-hydroxyecdysone (modified from Hayashi and Levine, 1992). Dissociated cell bodies were plated onto poly-D-lysine-coated coverslips and incubated for 1 to 3 days at room temperature. Although peak sodium currents decreased over a period of 3 days in culture, no significant changes in channel-gating properties (voltage-dependent activation, steady-state inactivation, and kinetics of fast inactivation) were observed.

Whole Cell Recording

Sodium channel currents were measured using the whole cell configuration of the patch clamp technique. Patch pipettes were made of borosilicate glass tubing (Drummond, Broomall, PA) using a Zeitz Universal micropipette puller (Augsburg, Germany), coated with Sylgard, and fire polished. Pipettes filled with the internal solution described below had tip resistances of 1 to 2 M.

Voltage Clamp. Sodium currents were recorded with patch pipettes filled with an internal recording solution containing 100 mM CsF, 40 mM CsCl, 3 mM MgCl₂, 10 mM EGTA, and 5 mM HEPES, pH 7.0. The extracellular recording solution contained 100 mM NaCl, 50 mM choline-Cl, 4 mM KCl, 2 mM CaCl₂, 30 mM tetraethyl ammonium-Cl, 1 mM 4-aminopyridine, 10 mM HEPES, and 10 mM glucose, pH 7.0. Currents were recorded using an Axopatch 200A amplifier (Axon Instruments, Foster City, CA) and filtered at 2 kHz.

Current Clamp. Pipettes were filled with 140 mM KCl, 5 mM NaCl, 1 mM MgCl₂, 5 mM EGTA, 10 mM HEPES, and 0.5 mM CaCl₂, pH 7.0. The external solution contained 140 mM NaCl, 4 mM KCl, 10 mM HEPES, 10 mM glucose, 2 mM CaCl₂, and 2 mM MgCl₂, pH 7.0. Neurons were maintained a resting potential of about 72 mV by injection of direct current. Square current pulses were injected through the patch pipette to determine the threshold of action potential generation.

	Results

Top Summary Introduction Materials and Methods Results Discussion References

Resistance to Permethrin in Pyr-R H. virescens. The Pyr-R line of resistant H. virescens resulted from repeated outcrossing of the resistant HSB strain (Payne et al., 1988) to the pyrethroid-susceptible Woodrow83 line and selection of hybrids with lethal doses of permethrin. This placed the Pyr-R gene in a Woodrow83 background and resulted in a resistance ratio of 22-fold using cypermethrin (Pimprale et al., 1997). Because the Woodrow83 line was unavailable for this study, we used two pyrethroid-susceptible strains for comparison to Pyr-R: UCR-S and Ace-R; the latter strain contained a substantial contribution from Woodrow 83 (see Materials and Methods). Insects were held in permethrin-coated glass vials or a vehicle control and scored after a 24-h exposure. The median effective concentration (EC₅₀) values for the UCR-S and Ace-R strains were 1.1 and 1.0 µg/vial, respectively, whereas the LC₅₀ value for the Pyr-R strain was 25.1 µg/vial (Table 1). The resistance ratio comparing Pyr-R to either susceptible stain was thus ~23- to 25-fold, which is virtually identical with the value of 22-fold obtained from previous comparisons of Pyr-R with Woodrow83.

A Single Amino Acid Substitution in hscp Sodium Channel. Recently, sequence analysis of the hscp sodium channel identified two coding polymorphisms associated with pyrethroid resistance in different resistant strains of H. virescens (Park et al., 1997; Park and Taylor, 1997). One polymorphism, detected in individual specimens ancestral to Pyr-R (RR; see Park et al., 1997), involves a deduced amino acid change from valine to methionine at position 421 (V421M) located in the IS-6 transmembrane segment (I-S6 nomenclature of Noda et al., 1986; see Fig. 1 for further details). The second polymorphism is a leucine-to-methionine change at amino acid 1029 (L1029H) in transmembrane segment II-S6 (Fig. 1A), found in Louisiana populations of H. virescens (Park and Taylor, 1997)

H. virescens Sodium Channels Exhibit Fast Kinetics and TTX Sensitivity. Because sodium channels in central neurons of H. virescens have not been characterized previously, we examined their activation and inactivation kinetics and sensitivity to TTX. Short-term cultured H. virescens central neurons show robust voltage-activated sodium currents that display TTX sensitivity and rapid gating kinetics. Low concentrations (0.1-60 nM) of TTX completely abolished sodium currents of both UCR-S and Pyr-R neurons with K_d ~2 nM (Fig. 3, A-C). No difference in TTX sensitivity was observed for the two strains.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   TTX sensitivity and kinetics of H. virescens sodium channel gating. A, concentration-response curves for UCR-S and Pyr-R sodium channels show that TTX completely abolishes voltage-activated currents in the concentration range 0.1 to 10 nM, with Kd ~2 nM. B and C, sample current families recorded from UCR-S and Pyr-R neurons, respectively, after application of 0.6, 2, 6, and 60 nM TTX. Largest current in each family is the untreated control. D, kinetics of sodium channel activation in UCR-S and Pyr-R neurons. Time to the peak current from the onset of the depolarization step was measured and plotted as a function of test potential. E and F, fast decay of the sodium current, fitted with double exponentials. 1 values associated with typical sodium currents (normalized) are shown.

Sodium Channels of Pyr-R Neurons Have Altered Gating Properties. We observed marked differences in current-voltage relationships of sodium currents in UCR-S and Pyr-R neurons. Voltage-dependent activation of UCR-S sodium channels began at approximately 48 mV and reached peak values at approximately 23 mV (Fig. 4A). In contrast, Pyr-R channels activated at approximately 38 mV and reached peak near 10 mV (Fig. 4B). Differences in average sodium conductance as a function of test potential were analyzed from a sample population of 18 UCR-S and 13 Pyr-R neurons (Fig. 4C). For UCR-S channels, 50% activation was achieved at 29.1 ± 0.4 mV, whereas for Pyr-R channels, 50% activation occurred at 16.1 ± 0.3 mV. The slope factors describing voltage-dependent activation for UCR-S and Pyr-R were very similar: 4.4 ± 0.2 and 4.5 ± 0.1 for susceptible and resistant neurons, respectively. Our findings thus indicate that voltage-dependent activation of Pyr-R sodium channels is shifted ~13 mV in the positive direction.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Current-voltage relationships for sodium channels of UCR-S (A) and Pyr-R (B) central neurons. Families of peak currents shown at left were evoked by test potentials every 5 mV from 48 to 3 mV for UCR-S and from 38 to 7 mV for Pyr-R. Typical current/voltage curves for a wider range of voltage steps are shown at right. C, average sodium conductance as a function of test potential for UCR-S and Pyr-R strains. Each value for sodium conductance (gNa) was calculated using the equation, gNa = INa/(VT  Erev) and normalized to maximum sodium conductance. INa was peak current resulting from applied test potentials (VT) and Erev was the reversal potential of sodium current. Curves were fitted with Boltzmann distribution; error bars = S.E.M.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Steady-state inactivation (A) and recovery from inactivation (B) of sodium channels in UCR-S and Pyr-R neurons. A, each peak current was evoked by a test potential (VT) after a 50-ms prepulse as shown in the upper right. The peak current resulting from the test pulse was normalized to the maximum peak current and plotted as a function of the prepulse potential. Steady-state inactivation (h) curves were fitted with the Boltzman equation: h = {(1  C)/{1 + exp[(VP  V1/2)/k]}} + C, where V1/2 is the voltage at half-inactivation; VP and k are the prepulse potential and slope factor, respectively. C, the noninactivated fraction of current if any remains. B, to measure recovery from inactivation, peak currents were evoked by VT2 applied at various time intervals after VT1 (inset). Peak currents (Peak INa) were normalized to the maximum peak sodium current (Max. peak INa) and plotted as a function of pulse intervals (bar = S.E.M.).

Pyr-R Neurons Show Reduced Excitability. We have demonstrated that voltage-dependent activation of Pyr-R sodium channels is shifted ~13 mV in the positive direction relative to UCR-S channels. It follows that neurons expressing such channels would be less excitable, requiring more positive potentials to reach threshold for action potential generation. Also noted was relatively sluggish behavior of Pyr-R individuals in laboratory culture compared with the UCR-S strain. Therefore, we tested the hypothesis that Pyr-R neurons may have altered excitability by quantifying action potential thresholds.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Comparison of action potential threshold (VTH) for UCR-S and Pyr-R neurons measured under current clamp conditions. A, current pulses (Iinjected; 10-ms duration) of increasing magnitude elicited progressively more positive voltage responses (Vm) until the active membrane response led to an action potential in UCR-S (left) and Pyr-R (right) neurons. The relationship between injected current pulses and membrane potential response was linear until VTH was reached (B). As shown in B, VTH is about 11 mV more positive in the Pyr-R cell compared with UCR-S cell.

Sodium Channels of Pyr-R Strain Are Less Sensitive to Permethrin. Pyrethroids modify sodium channels through inhibition of inactivation and deactivation, the latter evident as an extreme prolongation of the tail current (Narahashi, 1996). We found that exposure to permethrin prolonged sodium channel tail currents (I_Na-tail) of both UCR-S and Pyr-R central neurons (Fig. 7). Notably, compared with UCR-S channels, Pyr-R sodium channels were less sensitive to permethrin.

View larger version (34K): [in this window] [in a new window]   Fig. 7.   Differential sensitivity of sodium currents in UCR-S and Pyr-R H. virescens neurons to permethrin. A, example of prolonged tail currents caused by sequential applications of 0.3, 0.6, and 1 µM permethrin to a UCR-S neuron. Depolarizing pulses (VT = 18 mV) to achieve maximal channel activation were applied at ~10-s intervals; the holding potential (VH) was 108 mV. Control sodium current (Con) completely deactivated upon returning to VH, but each application of progressively higher permethrin concentrations produced rapid increases of tail current to new steady-state values within ~3 to 5 min. B, tail currents recorded from a UCR-S neuron following exposure increasing concentrations of permethrin (VT = 18 mV). C, tail currents recorded from Pyr-R neurons following permethrin exposure (VT = 8 mV). D, percentage of modified channels plotted as a function of permethrin concentration. Each INa-tail measured 12 ms after the end of VT (*) was used to calculate % modified channels; see text for details (error bars = S.E.M.).

Enhanced Sensitivity of Pyr-R Sodium Channels to -Scorpion Toxin. Sodium channels are modified by a variety of natural and synthetic agents. As many as six nonoverlapping binding sites have been defined, many of which interact allosterically (Lombet et al., 1988). Given the differences in voltage-dependent gating and pyrethroid sensitivity observed here for Pyr-R sodium channels, we examined whether they might exhibit altered sensitivity to other classes of modifiers. In particular, we compared the sensitivities of UCR-S and Pyr-R sodium channels to the -scorpion toxin LqhIT (Eitan et al., 1990). LqhIT dramatically slowed inactivation of both UCR-S and Pyr-R sodium channels in a concentration-dependent manner (Fig. 8 insets). To quantify modification of sodium currents by the toxin, elevation of steady-state sodium current (I_Na(s-s)) was normalized to maximum amplitude and plotted as a function of LqhIT concentration (Fig. 8). The median effective dose (ED₅₀) for UCR-S and Pyr-R channels was 0.71 ± 0.03 and 1.84 ± 0.06 nM, respectively. Slope factors, 2.12 ± 0.1 for UCR-S and 2.15 ± 0.1 for Pyr-R, were not significantly different. These results show that Pyr-R sodium channels were about 2.6 times more sensitive than their UCR-S counterparts to LqhIT.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Differential effects of the -scorpion toxin LqhIT on sodium currents in UCR-S and Pyr-R neurons. Inset (lower right), inhibition of sodium-channel inactivation in a UCR-S neuron caused by exposure of 0.6, 2, and 10 nM LqhIT. Inset (upper left), effects of 0.6, 1, and 6 nM LqhIT on a Pyr-R neuron. All steady-state currents modified were normalized to a maximum steady-state sodium current induced by this toxin and plotted as a function of LqhIT concentration. Error bars indicate S.E.M. Sodium channels in Pyr-R neurons are about 2.6 times more sensitive to the toxin.

Sodium Channel Density Is Unchanged in Pyr-R Neurons. Low levels of pyrethroid resistance in D. melanogaster mutants have been associated with a reduction in sodium channel density and lower excitability of neuronal membranes (Kasbeker and Hall, 1988). We compared sodium-channel density in UCR-S and Pyr-R neurons by normalizing the maximum peak sodium current to relative cell size, expressed as a function of whole cell capacitance (C_m). We found that, after the initial 3-day incubation period, sodium-channel density gradually decreased in culture, but that relative channel densities of UCR-S and Pyr-R neurons were not significantly different (Fig. 9).

View larger version (24K): [in this window] [in a new window]   Fig. 9.   Relative sodium channel densities in UCR-S and Pyr-R neurons calculated as maximum peak sodium current divided by whole cell capacitance (Cm). Relative density decreases as a function of time in culture, but no significant difference is evident between UCR-S and Pyr-R strains.

	Discussion

Top Summary Introduction Materials and Methods Results Discussion References

Our examination of sodium channels in central neurons of Pyr-R insects has yielded evidence for structural change resulting from the mutation V421M reported previously (Park et al., 1997). These data provide a clear mechanistic basis for pyrethroid resistancereduced sensitivity at the molecular target site. The involvement of sodium channel regulatory elements, implicated in other instances of pyrethroid resistance (Kasbeker and Hall, 1988), is not indicated, because relative channel density is unchanged. Our study has been restricted to characterization of sodium channels in central neurons, and therefore other genetic factors (e.g., metabolic processes) could also contribute to the overall magnitude of resistance. Nevertheless, our findings together with those of previous reports on the Pyr-R strain strongly support the hypothesis that structural change in the sodium channel is a major factor in the resistance observed. This evidence includes nervous system insensitivity to pyrethroids (Payne, 1987), cross-resistance to pyrethroids and DDT (Brown et al., 1996), genetic linkage between resistance and the hscp sodium channel (Taylor et al., 1993), and the mutation V421M in the I-S6 transmembrane segment of the channel (Park et al., 1997).

The V421M substitution fits an emerging pattern of S6 mutations implicated in kdr-like pyrethroid resistance. The most common mutation (Leu-to-Phe) was first identified by Williamson et al. in the II-S6 transmembrane segment of kdr houseflies and is now documented in numerous species of Pyr-R insects (Williamson et al., 1996; Dong, 1997; Park et al., 1997; Park and Taylor, 1997). This resistance factor occurred in the original housefly strain reported to possess pyrethroid/DDT cross-resistance (Busvine, 1951) and seems to be a particularly stable allele. A mutation at the homologous position, but involving a Leu-to-His mutation instead of a Leu-to-Phe mutation, was detected in a Pyr-R strain of H. virescens different from Pyr-R (Park and Taylor, 1997). Finally, a III-S6 mutation is associated with pyrethroid resistance in the temperature-sensitive para⁷⁴ D. melanogaster mutant (Pittendrigh et al., 1997). Based on the prevalence of S6 mutations in Pyr-R insects, the V421M substitution is well placed as a causal factor in the resistance observed in the Pyr-R strain.

Does the V421M substitution account for the differences in sodium channel properties we have observed in this study? This seems likely, based on extensive sequencing of the hscp gene and the nature of the alterations we have observed. V421M is the only sequence polymorphism found in hscp sequence comparisons between UCR-S and Pyr-R specimens encompassing 80% of the channel sequence (Park et al., 1997). Regions of the hscp channel remaining to be defined lie only at the extreme 5' and 3' ends of the molecule (see Fig. 1), areas unlikely to account for the changes reported here. More to the point, the altered gating properties we have observed correspond to those obtained independently by others examining mutations at the same location. A virtually copositional mutation of the rat µ1 sodium channel (N434A; corresponds to 422 in the hscp channel; see Fig. 2) shifts voltage- dependent activation ~23 mV positive and fast inactivation ~12 mV positive (Wang and Wang, 1997). These shifts are similar to those we have associated with the V421M substitution. Position 421 also lies in the putative batrachotoxin binding site [site 2; (Trainer et al., 1996)], which is allosterically coupled to pyrethroid binding. Mutations N434K and I433K (the latter copositional with 421 in the hscp channel) in the rat µ1 sodium channel abolish sensitivity to batrachotoxin (Wang and Wang, 1998). Based on this corroborative evidence, the V421M substitution seems well situated to account for reduced pyrethroid sensitivity and shifts in voltage-dependent gating properties we have found.

We also have observed altered sensitivity of Pyr-R channels to the -scorpion toxin LqhIT, a site 3 ligand. Biochemical studies have shown that site 3 is allosterically coupled to pyrethroid binding (Trainer et al., 1997). In addition, a portion of the site 3 receptor lies in the I-S5-S6 extracellular loop, directly adjacent to the I-S6 transmembrane segment (Thomsen and Catterall, 1989). Modification of site 3 affinity by the V421M mutation thus has a reasonable biochemical and structural basis.

What is the relationship between reduced pyrethroid sensitivity and altered voltage-dependent gating in Pyr-R sodium channels? The V421M mutation might alter the pyrethroid binding site directly, or allosterically by modifying voltage-dependent conformational states of the channel. Interestingly, examination of the kdr mutation L1014F occurring in the II-S6 transmembrane segment of the housefly sodium channel (Smith et al., 1997) showed reduction of pyrethroid sensitivity without obvious changes in gating properties. This suggests that reduced pyrethroid sensitivity is not necessarily a function of gating-property alteration. However, the same mutation in the rat IIA sodium channel produced a clear shift in the voltage-dependent activation curve (Vais et al., 1997). This discrepancy serves to emphasize that each channel may respond differently to particular point mutations.

As a first step in the analysis of site-insensitive pyrethroid resistance, we have characterized sodium channels of H. virescens in their native environment, central neurons. Further mutational analysis will be facilitated by expression of sodium channels in heterologous systems. Heterologous expression should provide additional evidence for a causal relationship between the V421M substitution and altered channel properties and should allow us to distinguish between direct effects of mutations on the pyrethroid binding site and indirect effects occurring through altered voltage-dependent conformational states of the channel. However, clear answers to these questions may require extensive mutagenesis experiments. Recent work of Wang and Wang (1997) on the rat muscle µ1 sodium channel has implications for this issue. They found that the N434A mutation produced a large effect on voltage-dependent gating, whereas N434K did not. Nevertheless, both mutations abolished batrachotoxin sensitivity. It is possible that heterologous expression of the channel may introduce other modulatory influences on the channel arising from differences in postranslational processing, second messenger modulation, and possible contributions of accessory subunits. Such changes will be possible to recognize given the data on neuronal properties reported here.

The stability of insecticide resistance genes in insect populations is likely to be related to their fitness costs (Taylor and Feyereisen, 1996). In this study, we showed that altered voltage-dependent activation of Pyr-R channels is associated with decreased cellular and behavioral excitability. Specifically, Pyr-R neurons exhibit a higher threshold (~11 mV) for action potential generation and relatively sluggish behavior compared with their counterparts in the UCR-S strain. This suggests a possible fitness cost associated with the V421M substitution. Although comparative measures of fitness among the sodium-channel genotypes in isogenic lines are complex and not yet quantified, some reduction in fitness of Pyr-R H. virescens (Campanhola et al., 1991) may result from decreased neuronal excitability associated with the V421M mutation. Indeed, population genetic analysis of pyrethroid resistance mutations revealed that the frequency of the V421M allele in field populations of H. virescens decreased during the interval 1990 to 1996/1997, whereas the incidence of the L1029H mutation increased during the same period (Park, 1998). As mentioned above, examination of the homologous Leu-to-Phe mutation in housefly sodium channels showed no significant effect on gating characteristics of sodium channels (Smith et al., 1997). This may be a consequence of negligible fitness costs associated with this mutation and thus may help to explain the stability of this mutation in diverse populations of resistant insects.

It has been almost 50 years since the first report of cross-resistance to pyrethroids and DDT, subsequently termed knockdown resistance (Busvine, 1951; Farnham, 1977). Our findings provide some of the first detailed molecular and physiological correlates of this type of resistance, and support the hypothesis that the V421M mutation modifies either the pyrethroid binding site directly or allosterically. Further studies are under way to elucidate the precise molecular mechanisms leading to reduced pyrethroid sensitivity. With the identification of associated sodium channel mutations, hypotheses testing should lead to a greater understanding of the molecular mechanisms underlying resistance, as well as insights into the mechanism by which pyrethroids modify sodium channels.