Introduction

Acetylcholinesterase (AChE, EC 3.1.1.7) is an important enzyme that terminates the neurotransmission by rapidly hydrolyzing the neurotransmitter acetylcholine at cholinergic synapses in all animals1. It is also involved in many cellular processes in eukaryotes, including apoptosis, modulation of cellular interactions, cell adhesion and synaptogenesis in vertebrates2,3,4,5. In insects, AChE has been extensively studied because it functions in neurotransmission, serves as a major target for anticholinesterase insecticides (organophosphates and carbamates) and constitutes a common mechanism of insecticide resistance through its reduced sensitivity to the insecticides6,7,8.

AChE has long been known to be encoded by a single gene (Ace) in vertebrates1. In insects, the first Ace was cloned from the fruit fly (Drosophila melanogaster) in 19869 and subsequently other insects were thought to possess a single Ace. However, many studies identified insecticide resistance associated with reduced sensitivity of AChE to anticholinesterase insecticides, but were unable to identify any amino acid substitutions of deduced AChE sequences from the resistant strains6. This conflict was resolved by the first report of an Ace gene in the greenbug (Schizaphis graminum) that is paralogous to that of D. melanogaster in 200210 and subsequent reports of paralogous Ace in the cotton aphid (Aphis gossypii)11 and the African malaria mosquito (Anopheles gambiae)12.

It is now known that most insect species have two AChEs8,14 except for the species in the suborder Cyclorrapha of Diptera including D. melanogaster as confirmed by its genome sequence13, the house fly (Musca domestica) and the Australian sheep fly (Lucilia cuprina). One AChE (AChE1 or AP-AChE) is encoded by a paralogous Ace (Ace1) and the other AChE (AChE2 or AO-AChE) is encoded by an orthologous Ace (Ace2)8,14. It is also known that the amino acid substitutions conferring insecticide resistance are associated with AChE1 in the species possessing both Ace1 and Ace2 and with AChE2 in the Cyclorrapha species such as Drosophila that have a single AChE gene (i.e., Ace2)8.

To date, cDNAs encoding AChEs have been isolated from at least 43 insect species. Of these species, 27 have been reported for both Ace1 and Ace215. However, the functions of the two gene products (AChE1 and AChE2) have hitherto been unclear. There has been scant information on which of the two AChEs is responsible for cholinergic neurotransmission and therefore a target of anticholinesterase insecticides. We have recently studied both AChE1 and AChE2 genes in the red flour beetle (Tribolium castaneum), a globally distributed major pest of stored grain products and an emerging model organism16 and found that both of the deduced AChE sequences contain all the conserved sequence motifs including a choline-binding site, a catalytic triad and an acyl pocket15. However, the T. castaneum AChE1 gene (TcAce1) resides on chromosome 5, whereas the T. castaneum AChE2 gene (TcAce2) is on chromosome 2. Our large-scale protein simulation studies suggest that T. castaneum AChE1 is a robust acetylcholine hydrolase, whereas AChE2 is not a catalytically efficient acetylcholine hydrolase15.

In this context, we compared transcript abundances of TcAce1 and TcAce2 and investigated their biological functions using RNA interference (RNAi) in T. castaneum. The results of these studies have provided for the first time crucial evidence with regard to which of the two AChEs is responsible for cholinergic neurotransmission and therefore a target of anticholinesterase insecticides and which of the two is responsible for non-cholinergic functions.

Results

Comparison of mRNA levels of TcAce1 and TcAce2

We determined the abundances of TcAce1 and TcAce2 transcripts of T. castaneum at different developmental stages and the abundances in the brains dissected from late pupae by using quantitative PCR (qPCR) (Table 1). The transcript levels of TcAce1 in 3-day eggs, 5 and 20-day larvae, 6-day pupae, 2-day adults and late pupal brains were about 8.7-, 5.4-, 1.2-, 3.8-, 3.0- and 5.2-folds higher than those of TcAce2, respectively (P≤0.05). Apparently, TcAce1 mRNA was more abundant than TcAce2 mRNA at all the stages and in the pupal brains. Because the transcript levels of TcAce1 and TcAce2 were nearly identical in 20-day larvae, we chose this stage for our RNAi experiments to evaluate the silence specificity and to explore functional differences of the two genes.

Table 1 Transcript copy numbers of TcAce1 and TcAce2 in each egg and insect of different developmental stages and in each brain of late pupa of T. castaneum*

Double-stranded RNA-mediated depletion of TcAce transcripts and effects on AChE activity

To confirm target specificity of RNAi, we prepared double-stranded RNA (dsRNA) from each gene and injected 20-day larvae with an individual dsRNA (dsTcAce1 or dsTcAce2) or their mixed dsRNAs (dsTcAce1+2) at the same amounts. qPCR analyses using primer pairs that did not overlap with the dsRNA regions showed that both dsTcAce1 and dsTcAce2 dramatically reduced their respective transcript levels without significantly affecting the non-target mRNA levels on day 4 (Fig. 1A). Specifically, the injections of dsTcAce1 and dsTcAce2 suppressed TcAce1 and TcAce2 transcripts by 92.3 and 95.2%, respectively, as compared with the control larvae injected with the buffer only (see our rationale in the method section). No significant impact on the transcript levels of the non-target gene was observed. Specifically, the injection of dsTcAce1 did not significantly reduce the transcript level of TcAce2, whereas the injection of dsTcAce2 did not significantly reduce the transcript level of TcAce1 in the larvae. As expected, the injection of dsTcAce1+2 suppressed both TcAce1 and TcAce2 transcripts to a similar extent.

Figure 1
figure 1

The dsTcAce1, dsTcAce2 and dsTcAce1+2 mediated suppressions of TcAce1 and TcAce2 transcripts as determined by qPCR (A) and AChE activity as determined by enzyme assays (B) and non-denaturing polyacrylamide gel electrophoresis (PAGE) (C) on day 4 after 20-day larvae of T. castaneum were injected with buffer (control), dsTcAce1, dsTcAce2, or dsTcAce1+2.

The results are presented as the mean and standard errors of three replicates (each was performed with a RNA sample prepared from four insects). Different letters above the standard error bars indicate significant differences based on the one-way ANOVA followed by Fisher's LSD multiple comparison test (P ≤ 0.05). For PAGE analysis, 55 μg of total protein were loaded in each well. The gels were stained for AChE activity by using ATC as substrate.

We further evaluated the effect of RNAi on the enzyme levels on day 4 by measuring total AChE activity (Fig. 1B) and performing non-denaturing polyacrylamide gel electrophoresis followed by staining for the enzyme activity (Fig. 1C). Both methods demonstrated significant reductions in the enzyme levels. However, because TcAce1 and TcAce2 encode two different AChEs with similar predicted molecular weights15, we were not able to separate the contribution of RNAi for each gene to the reduced AChE activities. Nevertheless, our study clearly showed that injections of dsTcAce1, dsTcAce2, or dsTcAce1+2 dramatically decreased the levels of respective transcripts which, in turn, led to the reduced enzyme levels four days after the treatments.

Effects of RNAi of TcAce1 and TcAce2 on pupation and emergence

The injections of dsTcAce1 and dsTcAce2 to 20-day larvae significantly delayed the pupation and emergence of T. castaneum. For the larvae injected with buffer alone (control), 100% pupation were observed nine days after the injection (Fig. 2A). In contrast, only 87.6, 72.2 and 65.0% of pupation rates were observed when larvae were injected with dsTcAce1, dsTcAce2, and dsTcAce1+2, respectively. For the adult eclosion, 100% eclosion was achieved 16 days after 20-day larvae were injected with buffer alone (Fig. 2B). In comparison, only 81.2, 71.4 and 8.0% of the normal eclosion rates were observed when the larvae were injected with dsTcAce1, dsTcAce2 and dsTcAce1+2, respectively. There was an enhanced effect on the adult eclosion when both TcAce1 and TcAce2 were simultaneously silenced.

Figure 2
figure 2

The effect of gene-silencing mediated by RNAi on the pupation (A), eclosion (B) and mortality (C) after 20-day larvae of T. castaneum were injected with buffer (control), dsTcAce1, dsTcAce2, or dsTcAce1+2.

The results are presented as the mean and standard errors of three replicates (each replicate was performed with at least 30 insects).

The injection of either dsTcAce1 alone or dsTcAce1+2 in 20-day larvae resulted in 100% mortality within two weeks after adult eclosion or 19 days after the injection (Fig. 2C). Many pupae treated with dsTcAce1+2 in the larval stage died even before adult emergence. In contrast, the injection of buffer (control) or dsTcAce2 alone did not lead to any significant mortality even though the remaining TcAce2 transcript level was only 4.8% of the control (Fig. 1). In addition, complete adult mortality was observed 27 days after the injection of dsTcAce1+2. This was four days earlier than the adults whose larvae were injected with dsTcAce1 alone (Fig. 2C).

Effects of RNAi of TcAce2 on egg-laying, egg hatching and offspring development

Since the injection of dsTcAce2 in 20-day larvae had a negligible effect on insect survival in the same generation, we further investigated: 1) whether the injection of dsTcAce2 in 20-day larvae would affect reproduction of adult females, 2) whether the RNAi effect would be carried over to the next generation and 3) which gender does that if so (Fig. 3A). Although the injection of dsTcAce2 in 20-day larvae did not lead to the mortality of the insects, we observed dramatic decreases in the egg-laying (Fig. 3B), egg hatching (Fig. 3C) and larval body weight of the following generation (Fig. 3D) when a female was from the larva injected with dsTcAce2. In contrast, such effects were not associated with any injections of dsTcAce2 in the males. Thus, our observed effects of RNAi for TcAce2 were clearly carried through the female rather than the male. As consequences of these RNAi effects, we observed relatively lower numbers of pupae and adults in the offspring when a female was from a larva injected with dsTcAce2 than those when a female was from a larva injected with buffer alone (Fig. 3E).

Figure 3
figure 3

The gender-dependent effect of RNAi for TcAce2 (A) on female egg laying (B) and egg hatching (C) and larval body weight (D) and distribution of different stages of the offspring (D) after 20-day larvae of T. castaneum were injected with dsTcAce2.

B-M+B-F: a male injected with buffer mating with a female injected with buffer; R-M+R-F: a male injected with dsTcAce2 mating with a female injected with dsTcAce2; R-M+B-F: a male injected with dsTcAce2 mating with a female injected with buffer; B-M+R-F: a male injected with buffer mating with female injected with dsTcAce2. An aliquot of 400 ng of dsTcAce2 was used in injection in all the treatment. The results are presented as the mean and standard errors of three replicates (each was performed with eight pairs of the female and male). Different letters above the standard error bars indicate significant differences based on the one-way ANOVA followed by Fisher's LSD multiple comparison test (P ≤ 0.05).

Effects of RNAi of TcAce1 and TcAce2 on insect susceptibility to insecticides

To assess whether the reductions of TcAce1 and TcAce2 transcript levels and total AChE activity by RNAi can lead to a change of insect response to anticholinesterase insecticides, we injected 20-day larvae with dsTcAce1, dsTcAce2, or dsTcAce1+2 and performed the bioassay with each of four anticholinesterase insecticides including two organophosphates (dichlorvos and malathion) and two cabamates (carbaryl and carbofuran) on day 4 after the injection (Fig. 4). Although the accumulative larval mortalities increased in all the insects injected with buffer, dsTcAce1, dsTcAce2 and dsTcAce1+2 from 24 to 72 h due to insecticide exposures, significant increased mortalities, as compared with the controls, were observed in larvae injected with dsTcAce1 and dsTcAce1+2. Generally, we did not find significantly increased mortalities in the larvae injected with dsTcAce2 alone as compared with the larvae injected with the buffer when all the larvae were later treated with each of the four insecticides. Slightly increased mortalities in the dsTcAce2-injected larvae were only found at 72 h for carbofuran and at 24 and 72 h for dichlorvos (Fig. 4).

Figure 4
figure 4

The effect of gene-silencing mediated by RNAi on the susceptibility to each of four anticholinesterase insecticides: carbaryl (A) and carbofuran (B), dichlorvos (C) and malathion (D) after 20-day larvae of T. castaneum were injected with buffer (control), dsTcAce1, dsTcAce2, or dsTcAce1+2.

After the larvae were injected with buffer or dsRNA, they were treated with each of the four insecticides on day 4. The larval mortality was observed at 24, 48 and 72 h after the insecticide treatment. The results are presented as the mean and standard errors of three replicates (each was performed with about 15 larvae). Different letters above the standard error bars indicate significant differences among the treatment of the same time point based on the one-way ANOVA followed by Fisher's LSD multiple comparison test (P ≤ 0.05).

Discussion

The canonical biological function of AChE is to terminate impulse transmission at cholinergic synapses by rapidly hydrolyzing the neurotransmitter acetylcholine in animals1. However, the discovery of a paralogous gene (Ace1) in many insect species prompted us to determine which of the two AChE genes is responsible for the cholinergic function and insecticide resistance as well as which of the two is for non-cholinergic functions.

Two Ace genes reportedly display significant differences in tissue-specific expressions and molecular properties in the German cockroach (Blattella germanica)17,18,19, the silkworm (Bombyx mori)20, the cat flea (Ctenocephalides felis)21 and T. castaneum15. Significant differences of biochemical properties between the two Ace products heterologously expressed in the baculovirous-infected cells have also been reported in the African malaria mosquito (Anopheles gambiae)22, the wild silkworm (Bombyx mandarina)23 and C. felis21. Recent studies suggested that AChE is involved in larval growth and development as observed by RNAi-mediated down-regulation of Ace transcript levels in the cotton bollworm (Helicoverpa armigera)24, B. germanica25 and the Asiatic rice borer (Chilo suppressalis)26. Functional differences of the two genes have not been adequately analyzed, although, based on selective and irreversible inhibition studies of aphid AChEs, it has been suggested that AChE2 does not contribute significantly to the overall AChE activity in aphids27.

In this study, we first determined the transcript abundances of TcAce1 and TcAce2 in T. castaneum at different developmental stages by using qPCR in 3-day eggs, 5-day larvae, 20-day larvae, 6-day pupae, 2-day adults and in brains from the late pupae (Table 1). The transcript levels of TcAce1 at these stages and in the brain were 1.2- to 8.7- fold higher than those of TcAce2 (P≤0.05), indicating that TcAce1 transcript is more abundant than TcAce2 at all the developmental stages and in the brains. Our data are consistent with those found in other insect species, including B. germanica18, the diamondback moth (Plutella xylostella)28, the oriental tobacco budworm (Helicoverpa assulta)29 and the mosquito (Culex pipiens)30. Nevertheless, the relative transcript abundances of the two genes vary significantly among the insect species and different tissues of the same species, ranging from 13- to 250-fold in the different tissues of P. xylostella28 and 3-fold in the nerve cord of B. germanica18. The mRNA level of Ace1 is generally more abundant than that of Ace2 in insects.

Because T. castaneum has robust RNAi responses at all developmental stages and in different cell types31,32, we took advantage of RNAi to systematically investigate cholinergic and non-cholinergic functions of each AChE gene. We observed 100% mortality within about two weeks after adult eclosion when 20-day larvae were injected with dsTcAce1 or dsTcAce1+2. In contrast, the injection of buffer (control) or dsTcAce2 alone did not lead to any significant mortality even though the TcAce2 transcript level was reduced to only 4.8% of the control (Fig. 2C). These results clearly indicate that AChE1 is essential for insect survival, presumably by regulating cholinergic neurotransmission, whereas AChE2 does not seem to play a major role in the neurotransmission in T. castaneum. The respective cholinergic and non-cholinergic functions of AChE1 and AChE2 are consistent with our large-scale protein simulation studies suggesting that T. castaneum AChE1 is a robust ACh hydrolase, whereas T. castaneum AChE2 is not a catalytically efficient ACh hydrolase, because the entrance of the active site of the AChE2 model refined by multiple molecular dynamics simulations appeared to be reversed relative to that of the corresponding AChE1 model and because ACh does not adopt the fully extended conformation and its carbonyl oxygen atom is not placed in the oxyanion hole in the AChE2 model15.

Interestingly, the injection of dsTcAce2 in 20-day larvae led to significantly delayed development of the insect, reduced egg-laying, egg hatching and growth of the offspring. Further examination of the ovaries in the adult females indicated relatively fewer mature ovarioles in the dsTcAce2-treated insects than the control (data not shown). This phenomenon appears to be similar to the incomplete ovary development and subsequent premature regression that correlates with changes in activity of the “acetylcholinesterase cells” of the pars distalis in the bird (Zonotrichia leucophrys gambelii)33. These results indicate that T. castaneum AChE2 plays an important, non-cholinergic role in insect embryonic development, growth and reproduction and are consistent with the observations that the expression of Ace in insects, including T. castaneum, was detected during early embryonic developmental stages long before the nervous system starts functioning15. These results are also consistent with the reports that Ace2 may confer non-cholinergic activities in most insect species possessing both Ace1 and Ace234,35 and that, although still in debate36, AChE may have non-cholinergic functions such as regulation of cell differentiation and neural formation in vertebrates and D. melanogaster3,4,5.

To provide further evidence that AChE1 is the enzyme responsible for cholinergic neurotransmission and the target of anticholinesterase insecticides, we performed RNAi for both TcAce1 and TcAce2 in 20-day larvae followed by insecticide bioassay. Our bioassay results showed significantly increases in larval susceptibility to all four insecticides at 24, 48 and 72 h after the larvae were injected with dsTcAce1 or dsTcAce1+2. In contrast, no major changes were observed when the larvae were injected with buffer (control) or dsTcAce2. Because RNAi for TcAce1 caused the depletion of its transcript, which ultimately led to a reduced translation of AChE1 enzyme, the outcomes of TcAce1 RNAi were expected to be similar to those caused by anticholinesterase insecticides that inhibit AChE activity. Thus, the depletion of TcAce1 transcript will make the larvae more susceptible to any insecticides if AChE1 encoded by TcAce1 is a target of the insecticides as examined in this study. Our results further support that T. castaneum AChE1 is a key enzyme involved in cholinergic neurotransmission. Decrease in TcAce1 transcript level can ultimately lead to insect mortality as was observed in insecticide bioassay. Furthermore, since RNAi for TcAce2 did not significantly increased susceptibility of the larvae to the insecticides, T. castaneum AChE1 is hence a target of the anticholinesterase insecticides.

The present work has not only delineated the functional differences of the two AChE genes in T. castaneum but also offered insight into developing environmentally-safe insecticides for insect pest control. In view of the cholinergic and non-cholinergic functions of two AChEs in T. castaneum, we suggest that any insect-specific anticholinesterase insecticides designed for insect pest control should target the AChE encoded by Ace1 rather than Ace2 unless the insect species possessing only Ace2. This notion is promoted by our findings that AChE1 in T. castaneum and many other insect species possesses a cysteine residue at the opening of the AChE active site but this cysteine residue is absent in fish and mammalian AChEs and insect AChE215,27,37,38. This insect AChE1-specific cysteine residue allows design of new chemicals that irreversibly inhibit AChE1 for insect pest control by conjugation of the chemicals to the cysteine residue27. Indeed, several recent studies have demonstrated promising potencies of some synthetic chemicals in irreversibly inhibiting the total AChE activity of insects including S. graminum and A. gambiae but virtually no or limited inhibition to AChE from humans27,39. Such chemicals could potentially lead to the development of novel and environmentally-safe insecticides that are toxic to most insect species possessing Ace1 that is now known to be responsible for cholinergic neurotransmission40.

In summary, our studies have for the first time delineated cholinergic and non-cholinergic functions of two AChE genes (TcAce1 and TcAce2) in T. castaneum. AChE1 is an essential enzyme involved in cholinergic neurotransmission and is the target of anticholinesterase insecticides. As such, insecticide resistance conferred by reduced sensitivity of AChE to insecticides would be expected to be due to genetic modifications of Ace1 rather than Ace2 in the insect species possessing both Ace1 and Ace2. In this context, we propose that AChE1 should be used as a target for designing new anticholinesterase insecticides for insect control. In contrast, AChE2 plays important but non-cholinergic roles in insect growth, female reproduction and embryo development. Such non-cholinergic functions of TcAce2 are carried through the female rather than the male.

Methods

Insects

The Georgia-1 (GA-1) strain of T. castaneum was reared on whole-wheat flour containing 5% (w/w) of brewers' yeast at 30°C and 65% RH under standard conditions41.

Analysis of TcAce1 and TcAce2 transcript abundances

Total RNA was isolated from 3-day eggs, 5-day larvae, 20-day larvae, 6-day pupae, 2-day adults or the brains dissected from late pupae by using TRIzol reagent (Invitrogen). First strand cDNA synthesis and qPCR were carried out using gene-specific primers15. Plasmid DNA containing a TcAce fragment was used to generate a standard curve and the transcript abundance of each Ace gene was converted to the copy number as previously described42.

RNAi of TcAce1, TcAce2 and TcAce1+2

dsRNA was synthesized based on the greatest sequence divergence between TcAce1 and TcAce2 using MEGAscript RNAi Kit (Ambion). Each 20-day larva was injected with 400 ng of dsTcAce1, dsTcAce2 or a total of 800 ng of the dsTcAce1 and dsTcAce2 mixture (i.e., 400 ng of dsTcAce1 and 400 ng of dsTcAce2). The mortality owing to injection damage was <10%. Our preliminary experiments showed very gene-specific phenotypes (i.e., insect mortality for TcAce1 vs. no mortality for TcAce2) after the expressions of these genes were individually suppressed by injecting their corresponding dsRNA in the larvae. Such gene-specific RNAi effects indicated that our observed phenotypes were not due to non-specific effect of exogenous nucleic acids (i.e., dsRNA). Thus, we used the buffer alone as a negative control in our injection experiments. After the injection, the insects were reared under standard conditions for visually monitoring of phenotypes and further analyses of the remaining transcript levels by qPCR. Three replications were carried out with at least 30 insects in each control or treatment.

To examine effects of dsTcAce2 injection on insect growth and development and gender-dependent effect of the RNAi, a large number of 20-day larvae were injected with dsTcAce2 or buffer as controls. After the injected larvae developed into pupae, the males and females were separated and paired as follows: 1) both a male and a female adults from the larvae injected with buffer; 2) both a male and a female adults from the larvae injected with dsTcAce2; 3) a male adult from the larva injected with dsTcAce2 but a female adult from larva injected with buffer; and 4) a male adult from the larva injected with buffer but a female adult from larva injected with dsTcAce2. Each treatment consisted of eight pairs of a male and a female and each treatment was repeated three times. Suppression of TcAce2 transcript was examined after four days following the injections of dsTcAce2 using total RNA prepared from pools of four individuals. Eggs were collected for five days after mated 13 days post-eclosion, whereas egg hatchability was examined five days after the eggs were collected.

Determinations of TcAce transcript levels after RNAi

Total RNA was isolated from a pool of four insects in each control or treatment on day 4 after 20-day larvae of T. castaneum were injected with buffer (control), dsTcAce1, dsTcAce2 or dsTcAce1+2 by using TRIzol reagent (Invitrogen). The RNA was treated with DNase I (Fermentas) and the first-strand cDNA was synthesized by using First Strand cDNA Synthesis Kit (Fermentas) with oligo (dT)18 as primer. qPCR was used to detect TcAce transcript levels after RNAi. A pair of TcRps3 primers was used as an internal control to monitor equal loading of cDNA for analysis of TcAce transcript levels.

AChE assays

Five larvae that were collected from the control and each dsTcAce-injected replication on day 4 were homogenized in 300 μl of ice-cold 0.1 M phosphate buffer (pH 7.5) containing 0.3% (vol./vol.) Triton X-100. The homogenates were centrifuged at 15,000 g for 15 min at 4°C and the supernatant was used as enzyme source. AChE activity was measured using model substrate acetylthiocholine (ATC, Sigma) according to the method of Ellman et al.43 with some modifications44. Specific activity was expressed as nmol of ATC hydrolyzed/min/mg protein. Remaining AChE activity was expressed as a percentage in relation to the activity of each control. Protein content of each enzyme preparation was determined by the BCA method45 using bovine serum albumin as protein standard. Each assay consisted of three biological replicates. Each reaction mixture including 50 μl AChE preparation, 0.25 mM ATC and 0.4 mM 5′5 dithio-bis (2-nitrobenzoic acid) (DTNB, Sigma) in 150 μl of 0.1 M phosphate buffer (pH 7.5). The enzyme activity expressed by Vmax mOD/min was determined using enzyme kinetic microplate reader (Molecular Devices) at 405 nm. AChE activities were expressed as nmol ATC hydrolyzed per min per mg protein using the extinction coefficient of 1.36×104 M−1 cm−1.

Analysis of AChE using non-denaturing electrophoresis

Non-denaturing polyacrylamide gel electrophoresis (non-denaturing PAGE) was performed on 4–20% Tris-glycine gels (Invitrogen). The running buffer contained 0.3% (vol./vol.) Triton X-100. The same volume of each AChE preparation containing same concentration of total protein was loaded onto each well. The gel was run at 150 V for 90 min in a cold chamber. The AChE band was visualized after the gel was stained for AChE activity using ATC as substrate46.

Insecticide bioassay

All four insecticides including dichlorvos (purity 99%), malathion (99.8%), carbaryl (99%) and carbofuran (98%) were obtained from Chem Service (West Chester, PA) and dissolved in acetone for bioassays on day 4 after 20-day larvae were injected with dsRNA. The concentration of each insecticide at LC50 (median lethal concentration) was used to treat 15 injected larvae in 50 μl of insecticide solution for 20 s. After the treated larvae were placed on a Whatman filter paper for drying in the air, they were transferred into an 8-ml glass vial and kept under the standard conditions as previously described. Control larvae were treated with acetone only. Each treatment was repeated three times. Mortality was recorded at 24, 48 and 72 h after insecticide treatment.

Statistical analysis

For the data obtained from qPCR, percent relative expression levels were calculated by dividing the relative expression value (REV) of each gene in the dsTcAce-injected larvae by the REV of the same gene in the buffer-injected larvae. The percent data of the relative TcAce expression were transformed using arcsine square root transformation and then the transformed data were subjected to ANOVA followed by Fisher's least significant difference (LSD) multiple comparisons to separate the means among the treatments by using ProStat software (Poly Software International). For the data obtained from AChE assay and insecticide bioassay, the percent data were first transformed using arcsine square root transformation and then subjected to ANOVA followed by Fisher's LSD multiple comparisons to separate the means among the treatments.