Introduction
Many insects have evolved various mechanisms for metabolizing toxic chemicals in their environment (Mazumdar-Leighton and Broadway, Reference Mazumdar-Leighton and Broadway2001; Gatehouse, Reference Gatehouse2002; Wittstock et al., Reference Wittstock, Agerbirk, Stauber, Olsen, Hippler, Mitchell-Olds, Gershenzon and Vogel2004). Similar to most other insects, social insects such as red imported fire ant (RIFA; Solenopsis invicta, Hymenoptera, Formicidae), and the honey bee, Apis mellifera, rely in part on a suite of detoxification enzymes to metabolize naturally occurring xenobiotic compounds and pesticides. Chief among these enzymes are the cytochrome P450 monooxygenases (P450) (Ffrench-Constant et al., Reference Ffrench-Constant, Daborn and Goff2004; Mao et al., Reference Mao, Rupasinghe, Johnson, Zangerl, Schuler and Berenbaum2009), especially the members of the CYP4 and CYP6 subfamily, which have been studied extensively in insect species (Liu et al., Reference Liu, Li, Gong, Liu and Li2015; Zhang et al., Reference Zhang, Su, Xie, Zhen, Hu, Jiang, Huang, Liu, Gao, Chen and Gao2020). Specifically, P450s play a role in the detoxification of phytochemicals and synthetic pesticides (Feyereisen, Reference Feyereisen2005). For example, compounds including tau-fluvalinate, a pyrethroid acaricide used to control Varroa mites (Johnson et al., Reference Johnson, Wen, Schuler and Berenbaum2006; Mao et al., Reference Mao, Schuler and Berenbaum2011), several pyrethroids including lambda-cyhalothrin (Pilling et al., Reference Pilling, Bromleychallenor, Walker and Jepson1995), and organophosphates such as coumaphos (Johnson et al., Reference Johnson, Pollock and Berenbaum2009; Mao et al., Reference Mao, Schuler and Berenbaum2011) are known to be metabolized by P450s.
The RIFA is recognized as the most invasive and destructive of alien species because of the complexity of its diet and its ferocious habits, rapid reproduction and strong competitive ability (Vinson, Reference Vinson1997). Many methods have been employed in attempts to control the fire ant. Chemical insecticides are still currently the most effective measures for control of the fire ant until a more effective control agent is found (Wang et al., Reference Wang, Lu, Xu and Zeng2013). Fipronil, a phenylpyrazole insecticide, exhibits neurotoxic activity by blocking the GABA-regulated chloride channels of neurons. It is useful for the control of many domestic and agricultural insect pests, especially since its mechanism of toxicity is different from traditional insecticides to which resistance has been developed. Fipronil that is formulated into granules or bait has been shown to be effective against of the fire ant (Greenberg et al., Reference Greenberg, Reierson and Rust2003; Marr et al., Reference Marr, O'Dowd and Green2003). In a study that evaluated broadcast application treatments with various contact insecticides, fipronil was the only insecticide to show a statistically significant reduction in the proportion of mounds that contained brood of the fire ant (Loftin et al., Reference Loftin, Hopkins, Gavin and Shanklin2003). Some success in killing the fire ant has been achieved in small areas, but eradication remains rare (Hoffmann et al., Reference Hoffmann, Luque, Bellard, Holmes and Donlan2016). Thus, new approaches to control the fire ant are needed.
RNA interference (RNAi) has been developed as an effective tool for both basic and applied applications in plants and animals (Fire et al., Reference Fire, Xu, Montgomery, Kostas, Driver and Mello1998; Tabara et al., Reference Tabara, Sarkissian, Kelly, Fleenor, Grishok, Timmons, Fire and Mello1999; Aravin et al., Reference Aravin, Naumova, Tulin, Vagin, Rozovsky and Gvozdev2001; Wesley et al., Reference Wesley, Helliwell, Smith, Wang, Rouse, Liu, Gooding, Singh, Abbott, Stoutjesdijk, Robinson, Gleave, Green, Waterhouse and Robinson2001). For example, gene silencing can inhibit virus replication (Tenllado and Dıaz-Ruız, Reference Tenllado and Dıaz-Ruız2001) and so plants engineered to produce artificial microRNAs targeting virus genes can resist viral infection (Niu et al., Reference Niu, Lin, Reyes, Chen, Wu, Yeh and Chua2006). Currently, many RNAi-based studies with various insects are in progress, and several reviews on their results have already been published. RNAi-based gene regulation has been reported in different insect orders including Lepidoptera, Hemiptera, Coleoptera, Diptera and Hymenoptera (Gordon and Waterhouse, Reference Gordon and Waterhouse2007; Huvenne and Smagghe, Reference Huvenne and Smagghe2010). In particular, Bellés (Reference Bellés2010) comprehensively reviewed RNAi-based studies on insects, covering some 30 species representing nine orders (Orthoptera, Dictyoptera, Isoptera, Hemiptera, Coleoptera, Neuroptera, Hymenoptera, Lepidoptera and Diptera). However, because RNAi approaches are still relatively new, this field will continue to expand rapidly as the number of insect genome projects increases and as RNAi methods are used ever more extensively in functional genomics.
Delivery of double-stranded RNA (dsRNA) through feeding has also been shown to effectively induce RNAi of target genes. Examples of this include dsRNA-diet incorporated feeding to Diabrotica punctate and D. undecimpunctata howardi, and Leptinotarsa decemlineata to silence genes involved in growth and survival (Baum et al., Reference Baum, Bogaert, Clinton, Heck, Feldmann, Ilagan, Johnson, Plaetinck, Munyikwa, Pleau, Vaughn and Vaughn2007); plants expressing short hairpin dsRNA to feeding Helicoverpa armigera (Mao et al., Reference Mao, Cai, Wang, Hong, Tao, Wang, Huang and Chen2007) to silence a P450 gene (CYP6AE14); droplet feeding to Epiphyas postvittana (Turner et al., Reference Turner, Davy, MacDiarmid, Plummer, Birch and Newcomb2006) to silence a gut carboxylesterase and a pheromone binding protein gene; feeding dsRNA to Plutella xylostella (Bautista et al., Reference Bautista, Tanaka and Miyata2007) to silence a P450 gene (CYP6BG1) involved in permethrin tolerance; feeding dsRNA to H. armigera (Asokan et al., Reference Asokan, Chandra, Manamohan, Kumar and Sita2014) to silence five target genes related to the digestion of proteins and the detoxification host allelochemicals; and feeding dsRNA to Sitobion avenae (Zhang et al., Reference Zhang, Su, Xie, Zhen, Hu, Jiang, Huang, Liu, Gao, Chen and Gao2020) to silence an aphid carboxylesterase gene related to tolerance to phoxim insecticides.
Our previous study indicated that CYP6AS160 (12H3) could be involved in detoxification of fipronil because the expression of CYP6AS160 in workers was induced strongly by fipronil (Zhang et al., Reference Zhang, Kong, Wang, Gao, Zeng and Shi2016). Here, we further explore the function of CYP6AS160 in workers, through RNAi-mediated gene silencing of CYP6AS160 transcripts via voluntary feeding bioassays to determine the effects of knockdown of CYP6AS160 on susceptibility to fipronil and total P450 activity in workers. Furthermore, we also determine the effect of silencing a non-overexpressed P450 gene (CYP4AB73), whose expression cannot be induced significantly by fipronil, to obtain further insight into the potential role of the overexpressed CYP6AS160 (table 1).
Table 1. Fipronil toxicity to workers fed with dsRNA (dsCYP6AS160 and dsCYP4AB73) vs. dsGFP
Notes: Different letters within the same time point indicate significant differences (P < 0.05). The data below the activity are the relative ratio of treatments/controls.
Materials and methods
Insect cultures
RIFAs were collected in Wushan (Coordinates, N23.16 E113.23) and Zengcheng (Coordinates, N23.3 E113.8) in Guangdong Province in China. The methods of collection and transport were based on Kuriachan and Vinson (Reference Kuriachan and Vinson2000). Rearing conditions were 27 ± 1°C, with 70–90% relative humidity and a 16:8h light:dark photoperiod.
The samples of different castes and tissues of S. invicta were collected as follows: the two castes were minor workers (n = 5) and dealate queens (N = 3) from polygyne nests. The different tissues were taken from minor workers (N = 10), each sample was repeated three times. All samples were snapped frozen in a liquid nitrogen flash freezer and then stored at −80°C for RNA extraction.
Chemicals
Fipronil (95% technical powder) was obtained from Wuxi Ruize Pesticide Co., Ltd. Chlorpyrifos (96% technical oil) was obtained from the Tianjin Jingjin Pesticide Factory. Nicotinamide adenine dinucleotide phosphate (NADPH), diethylpyrocarbonate and phenylmethanesulphonyl fluoride (PMSF) were obtained from Sigma-Aldrich (USA); dithiothreitol (DTT) and Tris base were purchased from Promega (USA). Ethylenediaminetetraacetate acid (EDTA) and bovine serum albumin (BSA) were purchased from Beijing Tongzheng Biological Company; p-nitroanisole (P-NA) and p-nitrophenol were purchased from Beijing Chemical Reagents; TRIzol reagent was purchased from Invitrogen (USA). Taq DNA polymerase and DNA Marker DL 2000 were purchased from Sangon Company (Shanghai, China). Agarose, DNase I and SYBR Green I were purchased from TaKaRa (Dalian, China). The MEGAscript® RNAi kit was purchased from Ambion (USA). All other chemicals used were of reagent grade.
Enzyme preparations
Workers (dsRNA-fed) samples were homogenized in 0.1 M phosphate buffer (pH 7.5), containing 1 mM EDTA, 1 mM DTT, 1 mM PMSF and 10% glycerol in an ice-cold mortar. The homogenate was centrifuged at 10,800×g, at 4°C, for 20 min. The supernatant was collected, filtered through cotton and either used immediately for the P450 assays or stored at −80°C.
The o-demethylase activity of total P450
The o-demethylase activity of total P450 was measured using methods described by Rose et al. (Reference Rose, Barbhaiya, Roe, Rock and Hodgson1995). Reactions were carried out in 1.5 ml centrifuge tubes in a water bath for 30 min at 30°C. About 345 μl of enzyme preparation was added to initiate the reactions; the initial reaction solutions contained 30 μl of 9.6 mmol l−1 NADPH, 375 μl of 2 μmol l−1 P-NA (dissolved in 0.1 mol l−1 PBS, pH 7.8). The final volume of the reaction system was 750 μl. The control samples used 1 mol l−1 PBS (pH 7.8) instead of the enzyme preparation. For analysis, we added 200 μl of reaction solution (three replicates per reaction) to 96-well microtitre plates, then used a vmax microplate reader and at a wavelength of 405 nm to record spectrophotometric values. Protein content was determined by the method of Bradford (Reference Bradford1976), using BSA as a standard. The o-demethylase activity of total P450 was expressed as μmol mg−1 protein/30 min (table 2).
Table 2. o-Demethylase activity of total P450 in workers treated with dsRNA (dsCYP6AS160 and dsCYP4AB73) vs. controls (dsGFP)
Notes: Different letters within the same time point indicate significant differences (P < 0.05). The o-demethylase activity of total cytochrome P450 is here reported as μmol mg−1 protein/30 min; the data below the activity are the ratio of treatments/controls.
Primer design
Quantitative real-time polymerase chain reaction (qRT-PCR) analysis primers were designed using primer3 (http://primer3.ut.ee/). The primers were designed on the basis of the following sequences published in NCBI: CYP6AS160 (accession no. MT482289), CYP6AS161 (accession no. MT482313), CYP4AB73 (accession no. MT482314) and CYP4G232 (accession no. MT482315) (personal communication by Professor Nelson, D.R., University of Tennessee, Memphis, TN, USA), CYP4AB2 (accession no. AY345971), RPL18 (accession no. EH413666) and EF1-beta (accession no. EH413796). RPL18 and EF1-beta were chosen as an internal control according to Cheng et al. (Reference Cheng, Zhang, He and Liang2013). The primers used in this study are detailed in table 3.
Table 3. Sequences of primers used for relative expression of the genes tested from the fire ant
qRT-PCR experiments
Total RNA was isolated from worker samples using TRIzol reagent. Sample RNA concentrations were measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA) at 260 nm. After total RNA (1.0 μg sample−1) was treated with DNase I (Fermentas, USA) to remove possible genomic DNA contamination, first-strand cDNA was synthesized in a 20 μl reaction system using a First Strand cDNA Synthesis Kit (Fermentas, USA) with oligo (dT)18 as the primer. The relative expression levels of the six target P450 genes (CYP6AS160, CYP6AS161, CYP4AB1, CYP4AB2, CYP4AB73 and CYP4G232) in the fire ant were examined using qRT-PCR. A total of 1 μg RNA was used as the template, three biological replicates were conducted; each sample was analyzed with two technical replicates and then immediately frozen in liquid nitrogen and stored at −80°C for total RNA isolation. qRT-PCR reactions were performed in a 20 μl mixture containing 1 μl of cDNA, 10 μl of SYBR Green qRT-PCR SuperMix-UDG, 0.15 μl of each primer and 8.7 μl of H2O. The amplification efficiency of the target genes and housekeeping genes (RPL18 and EF1-beta) was estimated using E = 10−1/slope − 1, where the slope was derived from the plot of the cycle threshold (C t) value vs. the log of the serially diluted template concentration. The optimized qRT-PCR program consisted of an initial step at 50°C for 2 min, 94°C or 2 min, followed by 50 cycles of 94°C for 15 s and 60°C for 30 s. After the cycling protocol, melting curves were obtained by increasing the temperature from 60 to 95°C (0.2°C s−1) to denature the dsDNA. The qRT-PCR amplifications were carried out in 96-well plates. The assays were performed in an ABI 7500 system using SDS v.1.4 application software (Applied Biosystems). Quantification of the transcript levels of six P450 genes was performed using the comparative 2–ΔΔCT method (Pfaffl, Reference Pfaffl2001).
CYP6AS160 and CYP4AB73 gene silencing
Synthesis of dsRNAs
dsRNAs were synthesized using an Ambion®MEGAscript® RNAi kit (USA) according to the manufacturer's instructions. The primers used to produce cDNA with T7 promoter sequences were designed based on the nucleotide sequences flanking the 391–726 positions of CYP6AS160 and the nucleotide sequences flanking the 581–959 positions of CYP4AB73 (table 4). As there is a high risk of cross-suppression or co-suppression in RNAi among closely related P450s, the primers targeted the sequences outside the signature motifs of P450s. The PCR products were examined on agarose gels prior to in vitro transcription to verify that the products showed a single band of the expected sizes. The two amplification products possessing a single T7 promoter sequence on opposite ends were combined (1 mg:2 mg). All dsRNA preparations were quantified spectrophotometrically and stored at −20°C until use.
Table 4. Primer sequences for production of dsRNA transcription templates
Administration of dsRNA by feeding
RNAi-mediated gene silencing was accomplished using voluntary feeding according to the method of Zhou et al. (Reference Zhou, Wheeler, Oi and Scharf2008). Workers (minors) of the fire ant were prestarved for 12 h before use, and the dealate queens were used as samples of different castes to compared expression levels of P450s between workers (minors) and queens. Feeding dsRNA assays were performed by placing groups of 80 workers into disposable plastic cups containing a paper disc moist with honey water (5%) containing 400 ng μl−1 of dsRNA (dsCYP6AS160 (336 bp) or dsCYP4AB73 (379 bp)) or the control which was 400 ng μl−1 of dsGFP (green fluorescent protein gene), and paper disc moisture with honey water (5%) containing 400 ng μl−1 of dsRNA was monitored each day. Mortality owing to the dsGFP-feeding was less than 10%. Three replicates were used for four survival timeframes: 24, 48, 72 and 96 h. Survivors at the end of each timeframe were subjected to qRT-PCR to quantify the expression levels of the P450.
To assess the sensitivity of workers to fipronil after silencing the target genes by RNAi, at 12, 24, 36, 48, and 60 h of dsRNA feeding, the workers were transferred to fipronil at the sublethal dose (0.05 μg ml−1) according to the method of the residual film in a disposable plastic cup (Hu et al., Reference Hu, Song and Scherer2005). One milliliter of the liquid was pipetted into a disposable plastic cup (caliber, bottom diameter and height were 6.6, 4.4 and 6.8 cm, respectively) and the cup was shaken until the acetone completely evaporated to evenly coat the cup with fipronil. The caliber was then lined with talcum powder to prevent the workers climbing out. At least 30 workers were then placed into the cup and allowed to feed on honey water (5%) soaked into a small cotton ball. Workers with dsGFP feeding were used as the control. The mortality of workers was assessed at 24 h after insecticide exposure. Each treatment was replicated three times.
Data analysis
Relative transcript levels of P450 in workers and queens, and time-dependent suppression of P450 transcript expression in workers were analyzed with an unpaired t-test using GraphPad InStat 3.0 software (GraphPad Software, San Diego, CA, USA). Mortality of workers exposed to fipronil after feeding dsRNA, the o-demethylase activity of total P450 and relative transcript levels of CYP6AS160 and CYP4AB73 in different tissues of minor workers were analyzed by an analysis of variance followed by Tukey's multiple comparison test (P < 0.05) using software InStat v.3.0 (GraphPad Software, San Diego, CA, USA).
Results
Castes (workers and queens) and tissue-specific (minor workers) expression patterns of CYP6AS160 and CYP4AB73
The expression levels of CYP6AS160 and CYP4AB73 in workers were significantly higher than those in queens (19.9- and 3.2-fold, respectively) (fig. 1). The expression amount of CYP6AS160 was relatively more in the abdomen compared to that of the head (4.2-fold), whereas CYP4AB73 was relatively more abundant in the head compared to the abdomen (1.6-fold) (fig. 2).
Figure 1. Transcript levels of CYP6AS160 and CYP4AB73 in tissues of multiple body parts combined from fire ant workers and queens. Different letters above the standard error bars indicate significant differences (P < 0.01).
Figure 2. Transcript levels of CYP6AS160 and CYP4AB73 in tissues of multiple body parts from fire ant minor workers as determined by qRT-PCR. Different letters above the standard error bars indicate significant differences (P < 0.05).
CYP6AS160 silencing and specificity
The expression of CYP6AS160 was significantly reduced in dsCYP6AS160-fed workers by 61.8-, 64.6-, 60.3- and 61.8% in comparison with controls (dsGFP-fed workers) at 24, 48, 72 and 96 h, respectively (fig. 3). The efficiency of knockdown of CYP6AS160 was relatively stable from 24 to 96 h.
Figure 3. Time-dependent suppression of CYP6AS160 transcript expression in workers of the fire ant fed with dsCYP6AS160 and dsGFP as determined by qRT-PCR. *Indicates significant differences (P < 0.05).
To examine whether the knockdown was specific to CYP6AS160 and did not affect the expression of other P450s belonging to the same CYP6, and CYP4 families, qRT-PCR analysis was also performed on similar samples with primer pairs specific for each of the P450s. The analysis revealed that the CYP6AS160 transcript level was greatly reduced by dsRNA feeding, whereas the levels for CYP4AB73 and CYP4G232 remain unchanged (0.97- and 0.98-fold difference between dsCYP6AS160-fed and dsGFP-fed workers) (fig. 4). Slight increases were observed for the expression of CYP6AS161, CYP4AB1 and CYP4AB2 (1.3-, 1.2- and 1.2-fold, respectively) in dsCYP6AS160-fed workers compared to controls (dsGFP-fed workers).
Figure 4. Transcript levels of P450s in tissues of multiple body parts of fire ant minor workers after exposure to dsCYP6AS160 and dsGFP as determined by qRT-PCR analysis. Different letters above the standard error bars indicate significant differences (P < 0.05).
Silencing CYP4AB73 and specificity
The expression of CYP4AB73 in dsCYP4AB73-fed workers was significantly reduced, by 40.2-, 77.1-, 51.6- and 62.4% in comparison with controls (dsGFP-fed workers) at 24, 48, 72 and 96 h, respectively (fig. 5). The knockdown efficiency of CYP6AS160 was relatively stable from 24 to 96 h. The knockdown efficiency of CYP4AB73 was not as effective as CYP6AS160, which suggests that knockdown efficiency may vary depending on the P450 gene target. In terms of knockdown efficiency of CYP4AB73, qRT-PCR analysis showed that the expression levels for CYP6AS160, CYP4AB1 and CYP4G232 remain unchanged (1.1-, 1.1- and 1.2-fold differences) between dsCYP6AS160-fed workers and controls (dsGFP-fed workers). A slight increase (1.3- and 1.9-fold) in the expression of CYP6AS161 and CYP4AB1 was observed in dsCYP4AB73-fed workers compared to controls (dsGFP-fed workers) (fig. 6).
Figure 5. Time-dependent suppression of CYP4AB73 transcript in fire ant workers fed with dsCYP4AB73 and dsGFP as determined by qRT-PCR. *Below the standard error bars indicates significant differences (P < 0.05).
Figure 6. Transcript levels of P450s in tissues of multiple body parts of fire ant minor workers after exposure to dsCYP4AB73 and dsGFP as determined by qRT-PCR analysis. Different letters above the standard error bars indicate significant differences (P < 0.05).
Susceptibility to fipronil and total P450 activity in workers fed with dsRNA
Mortality was significantly higher in dsCYP6AS160-fed workers (35.6, 77.8, 84.5, 95.6 and 97.8%, respectively) compared to controls (dsGFP-fed workers) (11.1, 20.0, 51.1, 77.8 and 84.5%, respectively at 24, 36, 48, 60 and 72 h) whereas no significant differences was observed in mortality between dsCYP4AB73-fed and controls (dsGFP-fed workers) (table 2).
The o-demethylase activity of total cytochrome P450 significantly decreased in dsCYP6AS160-fed workers (0.85, 0.56, 0.33 and 0.45-fold) compared to controls (dsGFP-fed workers) at 24, 48, 72 and 96 h respectively. Minimal o-demethylase activity of total cytochrome P450 were observed at 72 h (0.33–0.56 fold less than controls), whereas the o-demethylase activity of total P450 in workers treated with dsRNA (dsCYP4AB73) were slightly increased (1.22, 1.50, 1.48 and 1.32-fold) compared to controls (dsGFP-fed workers) at 24, 48, 72 and 96 h, respectively. Maximum of the o-demethylase activity of total cytochrome P450 was observed at 48 h (1.5-fold higher activity than the controls) (table 3).
Discussion
The midgut and fat body tissues in insects are generally considered to be the primary detoxification sites where most insect P450s related to detoxification are expressed (Liu and Scott, Reference Liu and Scott1998). Furthermore, other tissues (the brain and nervous system) can also be important for P450 gene expression and response to insecticide resistance (Scott et al., Reference Scott, Liu and Wen1998). CYP6AS160 was found to be more abundantly expressed in the abdomen compared to the head of workers, whereas CYP4AB73 was not relatively more expressed in the abdomen compared to the head. The observed higher expression of CYP6AS160 in the abdomen might reflect a role for CYP6AS160 in the metabolism of insecticides. On the contrary, the observed higher expression of CYP4AB73 in the head might not reflect a role for CYP4AB73 in the metabolism of insecticides.
In social insects, queens produce pheromones to regulate the development, reproduction and sex ratios of their colonies (Passera et al., Reference Passera, Aron, Vargo and Keller2001; Pennisi, Reference Pennisi2001; Bloch et al., Reference Bloch, Wheeler and Robinson2009). Worker individuals perform the functions of defending and maintaining the colony (Robinson, Reference Robinson2002). Because previous studies have shown that P450 genes may be doubly important in caste differentiation and detoxification in social insects (Liu and Zhang, Reference Liu and Zhang2004; Cornette et al., Reference Cornette, Koshikawa, Hojo, Matsumoto and Miura2006; Mao et al., Reference Mao, Rupasinghe, Johnson, Zangerl, Schuler and Berenbaum2009; Tarver et al., Reference Tarver, Coy and Scharf2012), we decided to investigate the role of P450 genes in these processes in the fire ant. Given that workers forage outside of mounds and can easily encounter insecticides, whereas queens would remain unexposed within the nest, it is reasonable to infer that the P450 genes from workers may ultimately contribute more to detoxification compared to queens. Thus, we measured the expression level of CYP6AS160 and CYP4AB73 in workers and queens. Indeed the expression levels of CYP6AS160 and CYP4AB73 in workers were significantly higher than those in queens. Furthermore, our previous study indicated that CYP6AS160 was induced significantly following exposure to fipronil, whereas CYP4AB73 was not induced significantly following exposure to fipronil (Zhang et al., Reference Zhang, Kong, Wang, Gao, Zeng and Shi2016). These findings strengthened the possibility that CYP6AS160 plays a significant role in detoxification of insecticide in the fire ant, whereas CYP4AB73 may not be involved in increased metabolism of insecticides.
In order to further clarify the function of CYP6AS160 involved in detoxification of fipronil, we deemed workers as a more amenable caste for use in RNAi experiments, as previous studies into dsRNA acquisition by workers have suggested that they can rapidly acquire dsRNA through a combination of feeding and trophallaxis. Through the process of trophallaxis, worker termites readily share food and other resources with nestmates such as pheromones, semiochemicals and nest building materials (Hamilton, Reference Hamilton1972; Zhou et al., Reference Zhou, Wheeler, Oi and Scharf2008). Trophallaxis can lead to the rapid transfer of materials throughout entire colonies in short periods of time (Cabrera and Rust, Reference Cabrera and Rust1999; Buczkowski et al., Reference Buczkowski, Wang and Bennett2007). Our study shows that oral delivery is capable of inducing RNAi in social insect workers, as was also observed in Reticulitermes flavipes (Zhou et al., Reference Zhou, Wheeler, Oi and Scharf2008). The high specificity of dsRNA for CYP6AS160 knockdown was confirmed by the unchanged levels of other P450 transcripts (i.e. CYP4AB73 and CYP4AB1) between dsCYP6AS160-fed workers and controls (dsGFP-fed workers) and the slightly increased expression level of CYP6AS161 in workers 24 h after dsCYP6AS160 feeding. CYP4AB73 knockdown also showed high specificity of dsRNA for CYP4AB73; this was confirmed by unchanged levels of other P450 transcripts (i.e. CYP6AS160, CYP4AB1 and CYP4G232) between dsRNA (CYP6AS160)-fed and controls (dsGFP-fed workers) and the slightly increased expression level of CYP6AS161 and CYP4AB2 in workers 24 h after dsRNA (CYP4AB73) feeding. Although a small subset of P450s was analyzed, it is likely that the RNAi employed here was highly sequence-specific. Analysis of the knockdown effects of CYP6AS160 on worker susceptibility to fipronil indicates that CYP6AS160 is responsible for detoxification of fipronil.
In addition, the o-demethylase activity of total cytochrome P450 was significantly lower in dsCYP6AS160-fed workers than that in controls, whereas the o-demethylase activity of total cytochrome P450 was a little higher in dsCYP4AB73-fed workers than that in controls. Although there were slightly increased levels of CYP6AS161 expression in workers, we observed significantly higher mortalities of the knockdown workers. Analysis of the knockdown effects of CYP4AB73 on worker susceptibility to fipronil indicates that the gene CYP4AB73 is not responsible for detoxification, although there was a significantly increased level of CYP6AS161 expression in workers, lower mortalities of knockdown workers were observed. We think that CYP6AS161 but not CYP4AB2, possibly be involved in insecticide detoxification as described above. These CYP450s that were upregulated may compensate for the silencing of CYP6AS160 or CYP4AB73.
The use of RNAi is a promising tool for insect management, but currently remains problematic, especially for field applications. There have been some successful RNAi studies such as those performed on Schistocerca gregaria that targeted different genes in a wide variety of tissues and developmental stages (Badisco et al., Reference Badisco, Marchal, Van Wielendaele, Verlinden, Vleugels and Broeck2011; Ott et al., Reference Ott, Verlinden, Rogers, Brighton, Quah, Vleugels, Verdonck and Broeck2012; Van Wielendaele et al., Reference Van Wielendaele, Dillen, Marchal, Badisco and Broeck2012). However, there have also been many unsuccessful RNAi attempts, such as for feeding dsRNA to Lepidopteran pests (Shukla et al., Reference Shukla, Kalsi, Sethi, Narva, Fishilevich, Singh, Mogilicherla and Palli2016). Our results demonstrate that RNAi is feasible by feeding dsRNA to worker ants, which is a base requirement for using RNAi as a pest management tool. Future studies should focus on demonstrating CYP6AS160 metabolism of fipronil to determine a more precise definition of its role in detoxification. Knockdown of CYP6AS160 in another larval caste of the fire ant would also be insightful.
Acknowledgements
This study was supported by the Key Science and Technology Program of Henan (Agriculture) (212102110441), the Key Scientific Projects of Institutions of Henan (21A210008) and the Project of Plant Protection Key Discipline of Henan Province (1070202190011005).
Author contributions
X.G. conceived and designed the experiments, B.Z., G.H. and L.L. performed the experiments and B.Z. and G.H. analyzed the data and wrote the manuscript. All authors reviewed the manuscript.
Conflict of interest
The authors declare no financial and non-financial interests.
