Identification and functional characterization of D-fructose receptor in an egg parasitoid, Trichogramma chilonis

In insects, the gustatory system has a critical function not only in selecting food and feeding behaviours but also in growth and metabolism. Gustatory receptors play an irreplaceable role in insect gustatory signalling. Trichogramma chilonis is an effective biocontrol agent against agricultural insect pests. However, the molecular mechanism of gustation in T. chilonis remains elusive. In this study, we found that T. chilonis adults had a preference for D-fructose and that D-fructose contributed to prolong longevity and improve fecundity. Then, We also isolated the full-length cDNA encoding candidate gustatory receptor (TchiGR43a) based on the transcriptome data of T. chilonis, and observed that the candidate gustatory receptor gene was expressed from the larval to adult stages. The expression levels of TchiGR43a were similar between female and male. A Xenopus oocyte expression system and two-electrode voltage-clamp recording further verified the function analysis of TchiGR43a. Electrophysiological results showed that TchiGR43a was exclusively tuned to D-fructose. By the studies of behaviour, molecular biology and electrophysiology in T. chilonis, our results lay a basic fundation of further study on the molecular mechanisms of gustatory reception and provide theoretical basis for the nutritional requirement of T. chilonis in biocontrol.


Introduction
Animals evolved a gustatory system that possesses the ability to detect and distinguish different taste stimuli in their living environments. The taste sensory system, which identifies and evaluates potential foods by discriminating between nutrients that benefit feeding behaviour, growth, and metabolism and harmful or even toxic compounds that are adverse to survival, is essential for most animals, ranging from flies to humans [1,2]. effects of D-fructose on T. chilonis adults. Then, we cloned the gustatory receptor candidate gene, TchiGR43a, from the transcriptome of T. chilonis and confirmed the expression patterns of the candidate gene in different life stages and different genders of T. chilonis by qRT-PCR. Last, we identified the function and ligands of TchiGR43a by using a Xenopus laevis oocyte expression system. This study will provide a solid foundation for further research on gustatory reception of T. chilonis and a theoretical basis for the application of this egg parasitoid in biological control programs.

Insects
Adult T. chilonis and the host Corcyra cephalonica (Stainton) eggs were originally obtained from the Plant Protection Research Institute, Guangdong Academy of Agricultural Sciences, People's Republic of China. All C. cephalonica eggs were sterilized by ultraviolet radiation. The parasitized C. cephalonica eggs were reared at 25±1˚C with 75±5% relative humidity and 14 (h) L: 10 (h) D photoperiod. After ten generations, the T. chilonis that emerged from the C. cephalonica eggs were used in the experiments.

Behavioural, longevity and fecundity experiments
To understand the ability to detect fructose, adults of T. chilonis were deprived of food (only supplied with water) for 24 hours. Then, 50~60 parasitoids were randomly selected into a glass tube that contained fructose solution colored with sulforhodamine B (0.2mg/mL) (red dye) and distilled water coloured with brilliant blue FCF (0.125mg/mL) (blue dye), kept at 25±1˚C with 75±5% relative humidity for 2 hours. The results were verified by using a stereomicroscope, parasitoids that fed on fructose had their bellies dyed red, parasitoids that fed on distilled water had their bellies dyed blue and parasitoids that fed on both solutions had their bellies dyed purple. The preference index (PI) for D-fructose was calculated using the following formula: PI = (N Red + 0.5 N Mix ) / (N Red + N Blue + N Mix ) [53], where N Red , N Blue and N Mix represent the number of T. chilonis coloured red, blue and purple (Fig 1A), respectively. PI � 0.5 indicates no preference and PI of 0.5-1.0 indicates a preference. The attractiveness was induced by D-fructose with a set of concentrations (0.010 M, 0.025 M, 0.050 M, 0.100 M and 0.300 M). This experiment was tested in triplicate. For adult Trichogramma species, longevity and fecundity play crucial roles that affect oviposition and even the mass culture. To further evaluate the effects of D-fructose on T. chilonis, we investigated the longevity and fecundity of this parasitoid in conditions with D-fructose and without D-fructose. For the longevity experiments, 90 females were divided evenly into three groups, and these groups were fed a 0.050 M D-fructose solution, distilled water and nothing, respectively, D-fructose solution and distilled water were refreshed every 8 hours. The treatments feeding with distilled water and nothing were set as controls. In every group, T. chilonis female adults were individually introduced into a glass vial (diameter 2.2 cm, height 10 cm) that contained a piece of cardboard carrying approximately 1000 UV-sterilized C. cephalonica eggs and a droplet (10 μL) of D-fructose solution. Vials were closed with cotton gauze and kept in an environmental chamber (25±1˚C, 75±5% relative humidity, 14 (h) L:10 (h) D photoperiod). In every group, each female is an experiment replication. Dead parasitoids were checked daily using a stereomicroscope.
For the fecundity experiments, 180 females were divided evenly into three groups, and these groups were fed a 0.050 M D-fructose solution, distilled water and nothing, respectively. Every group has 60 females, each female T. chilonis adult was individually introduced into a glass vial (each female was in a separate glass vial). In every group, each female is an experiment replication. Feeding with distilled water and nothing were set as controls. The glass vial (diameter 2.2 cm, height 10 cm) contained a piece of cardboard carrying approximately 1000 UV-sterilized C. cephalonica eggs and a droplet (10 μL) of a D-fructose solution or distilled water. Vials were closed with cotton gauze and kept in an environmental chamber (25±1˚C, 75 ±5% relative humidity, 14 (h) L:10 (h) D photoperiod). C. cephalonica eggs were refreshed every day, and the foods (D-fructose solution, distilled water and nothing) were refreshed every 4 hours. The parasitized C. cephalonica eggs were cultured in the same environment, and the number of parasitized eggs was recorded when the host eggs turned dark.

Cloning of the candidate gustatory receptor of T. chilonis
To understand the molecular mechanism of the candidate gustatory receptor gene in T. chilonis, we first used the transcriptome data of T. chilonis from our previous work (SRA accession number: SRP137064) [52].
Total RNA was extracted from adult T. chilonis using TRIzol reagent (Invitrogen, USA) according to the manufacturer's instructions. The first-strand complementary DNA (cDNA) was synthesized using a PrimeScript RT reagent Kit with gDNA Eraser (Takara, Kyoto, Japan). Templates for 5' and 3' RACE were prepared using a SMART RACE cDNA Amplification Kit (Clontech, Mountain View, CA, USA) according to the manufacturer's instructions. Primers (S1 Table) were designed based on the nucleotide sequences. Nested polymerase chain reaction (PCR) was performed to obtain the 5'-end/3'-end sequence with primer pairs. PCR products were cloned and then sequenced by Invitrogen (Shanghai).

Phylogenetic analysis
The full-length protein sequence of the putative gustatory receptor gene in our study was phylogenetically analysed with the homologues from Diptera (D. melanogaster), Lepidoptera (B. mori) and Hymenoptera (T. pretiosum, N. vitripennis, Apis mellifera, C. floridanum, Cephus cinctus, Orussus abietinus, Pseudomyrmex gracilis and Athalia rosae). The phylogenetic tree was constructed using the neighbour-joining method with 1000 bootstrap replications in MEGA 6.06.

qRT-PCR analysis of TchiGR43a
Total RNA of T. chilonis was extracted using the TRIzol method (Taraka, Japan). To obtain the first-strand cDNAs, 1 μg of total RNA was used for reverse transcription in a reaction system with a total volume of 20 μL, according to the manufacturer's instructions (PrimeScript RT Reagent Kit, TaKaRa, Japan). qRT-PCR was performed using LightCycler480 SYBR-Green I Master (Roche Diagnostics, Basel, Switzerland) and run on the LightCycler480 Real-time PCR system (Roche Diagnostics Ltd). Each reaction was conducted in a reaction system with a total volume of 10 μL with 1 μL of cDNA (2 ng/μL), 5 μL of SYBR Green I Master (LightCycler480 SYBR Green I Master, Roche Diagnostics Ltd., Lewes, UK), 0.5 μL/primer, and 3 μL of ddH 2 O. The qRT-PCR was conducted using the following programme: denaturation at 95˚C for 5 min, followed by 40 cycles of 5 s at 95˚C, 20 s at 60˚C, and 20 s at 72˚C. gapdh was the internal reference gene. Each gene was tested in triplicate, and the experiments were conducted on three biological replicates. The relative expression levels of the genes normalized to the internal control gene, were calculated using the 2 -ΔΔCt method [54]. Analysis of relative gene expression data used a real time quantitative PCR and the 2 -ΔΔCt method.

Functional characterization of TchiGR43a
To identify the function of the candidate D-fructose receptor of T. chilonis, we examined electrophysiological responses of Xenopus oocytes expressing TchiGR43a to 11 sugars at the concentration of 0.100 M. The full-length open reading frame sequence of TchiGR43a cDNA was amplified by RT-PCR, and first cloned into pMD 19-T vectors (Taraka, Japan). Then, the sequence was subcloned into a pCS2+ vector. The cRNA of TchiGR43a was synthesized from linearized modified pCS2+ vectors with a mMESSAGE mMACHINE SP6 Transcription Kit Foster City, CA, USA). Dose-response data were analysed with GraphPad Prism 6, and EC 50 (concentration for 50% of maximal effect) was calculated to show the sensitivity of the receptor to the ligand.

Data analyses
All results are expressed as the mean±SEM, and the data were analysed using one-way analysis of variance (ANOVA), followed by Duncan's multiple range test for multiple comparisons. To compare the sensitivity of GR to D-fructose and myo-inosotiol, a t-test was used. Statistical significance was determined at the P<0.05 level. Statistical analyses were performed using the SPSS 19.0 statistical software package.

Behaviour al preference of D-fructose
From the tests, T. chilonis adults showed a significant preference for D-fructose compared with water. To verify the sensitivity of T. chilonis adults to D-fructose, we calculated the attractive preference index. The PI of D-fructose rised with the increase in concentration of D-fructose ( Fig 1B). These data indicated that T. chilonis has an ability of sensing different concentrations of D-fructose.

Effect of D-fructose on the longevity and fecundity of T. chilonis adults
D-fructose significantly increased the longevity of T. chilonis adults when compared with the controls (Fig 2A). Between the two controls, adults fed water lived longer than those not fed. The fecundity of T. chilonis varied significantly between the D-fructose treatment and controls. The 0.050 M D-fructose solutions significantly improved the fecundity (Fig 2B). The fecundity of the two controls was similar. From the experiments, fecundity and longevity of T. chilonis highly benefited from feeding on D-fructose.

The candidate gustatory receptors identified from T. chilonis
To understand the molecular mechanism of taste detection in T. chilonis, we identified the candidate gustatory receptor gene that might be sensitive to some special tastants based on previous analysis of information from the transcriptome database of T. chilonis [53]. With RACE PCR, we obtained the full-length cDNA that encoded putative gustatory receptor in T. chilonis.
The full-length cDNA sequence of the candidate TchiGR gene was 1930 bp, with the GC content of 54.72% and an open reading frames (ORFs) of 1650 bp that encoded protein sequences of 549 amino acid residues. The candidate TchiGR had an identity higher than 60% with GRs from Trichogramma pretiosum, Nasonia vitripennis and Copidosoma floridanum. To assign putative functions to the candidate TchiGR gene, phylogenetic analysis of the TchiGR and GRs from Diptera (D. melanogaster), Lepidoptera (B. mori) and Hymenoptera (T. pretiosum, N. vitripennis, Apis mellifera, C. floridanum, Cephus cinctus, Orussus abietinus, Pseudomyrmex gracilis and Athalia rosae) insects was performed. According to this GR phylogenetic tree, the candidate TchiGR clustered phylogenetically with DmelGR43a, BmorGR9 and the members of a GR43a subclade from Hymenoptera insects (Fig 3). Most of the splits in the phylogenetic tree were strongly supported by high bootstrap values. According to the results of phylogenetic tree analysis and conventions of GR nomenclature, we named the putative gustatory receptor gene TchiGR43a (GenBank accession numbers: MH816967).

Expression patterns of TchiGRs of T. chilonis
The relative expression levels of the gustatory receptor gene in different developmental stages and in male and female adult T. chilonis were quantified by qRT-PCR with specific primers (Fig 4). T. chilonis is a holometabolous insect with several developmental stages, including the 26-hour egg stage, 36-hour larval stage, 48-hour prepupal stage, 84-hour pupal stage and 1-3-day adult stage [55]. The results of qRT-PCR showed that TchiGR43a expressed from the larval stage (d2) to prepupal stage (d3-d4) to pupal stage (d5-d8). The expression of the gustatory receptor gene showed a trend in which the expression level of TchiGR43a declined from the larval stage to the prepupal stage, then increased from the prepupal stage to the pupal stage and declined again at the late stage of the pupa (Fig 4A). The highest expression level of Tchi-GR43a was detected in the prometaphase of the pupal stage (d6), while the lowest expression level appeared in the prometaphase of the prepupal stage (d3). In adult male and female T. chilonis, the expression levels of TchiGR43a showed a similar pattern (Fig 4B).

Functional assay of TchiGR43a using two-electrode voltage-clamp recording
We found that the oocytes expressing TchiGR43a showed responses to D-fructose and myoinositol. There were significant differences between the response to D-fructose and myo-inositol, with a stronger response observed for D-fructose (Fig 5A and 5D). The D-fructose-induced current increased with fructose concentration from 0.005 to 0.300 M (Fig 5E). Based on the dose-response curve, the D-fructose evaluated EC 50 value was 0.023 M (n = 6) for TchiGR43a (Fig 5F). In the tests, one kind of the control cells were without injected cRNA of TchiGR43a and the other kind of the control cells were only injected with ddH 2 O; all of these control cells showed no response to the 11 sugars (Fig 5B and 5C).

Discussion
Detecting sugars in the living environment is critical for adult Hymenoptera. As a food source, sugars consumed are determined not only at the sensory but also at the physiological level. T. chilonis showed a behavioural preference response to D-fructose. Our results are consistent with those of other hymenopteran parasitoids such as Trichogramma japonicum [41] and Diadegma semiclausum [56], indicating that D-fructose could induce feeding behaviour, which might help with the exploitation of fructose as a main food source in nature. As a main sugar in various floral nectars and honeydew, fructose induces a feeding stimulation effect on insects, and this is caused by the palatability and nutritional value of fructose [57].
Feeding with D-fructose significantly prolonged T. chilonis longevity, as observed in other studies with parasitoids [41,56,58]. This finding indicates that D-fructose could be highly attractive to parasitoids and have an important role in their biological activity. Increasing the reproductive potential is crucial for improving the efficiency and ability of parasitoids to control pests. In general, fecundity increase with an increase in life span [59,42]. In our study, Dfructose increased the longevity of T. chilonis and also contributed to parasitoid fecundity. In contrast to controls, the fecundity of T. chilonis feeding on D-fructose almost doubled. The results indicated that diet, to a large extent, affected egg load during the oviposition period, which is consistent with the results from research on the parasitoids Microplitis mediator and Gonatocerus ashmeadi [42,60]. The levels of sugars were low in the newly emerged parasitoids, which is supported by previous studies [61 , 62]. After emerging, the wasps consumed sugars when D-fructose diets were provided, which revealed that the wasps might effectively utilize D-fructose [43]. Another study also showed that the D-fructose is correlated with reproduction, indicating that an increase in progeny is related with sugar intake in T. chilonis [41].
Fructose is suitable as a food source that prolongs the longevity of T. chilonis adults, increases their oviposition and prolongs the oviposition period. This study provided valuable insights for further understanding of T. chilonis in the field nutrition physiology research, and provided possibilities for the optimal utilization of sugar feedstuffs, for example, planting flowering plants could enhance the activity and efficiency of Trichogramma spp. [63][64][65][66], and the application of sugar sprays with certain concentration may also increase Trichogramma spp. efficiency [67].
The candidate gustatory receptor from T. chilonis, DmelGR43a, BmorGR9 and the members of a GR43a family from Hymenoptera formed a monophyletic subclade distinct from others (Fig 3). DmelGR43a and BmorGR9 were identified as having the function of sensing fructose, from the phylogenetic tree analysis we speculated that TchiGR43a may also have the same function. These members of the GR43a family from Hymenopteran species, including TchiGR43a, might be evolutionarily homologous with similar mechanisms or modalities for sensing sugars.
Gustatory receptors are in the membrane of gustatory neurons, which are housed in sensilla [4][5][6][7]. In this study, the results showed that TchiGR43a expressed from the larval stage to the pupal stage, indicating that the gustatory neurons related to TchiGR43a might exist from larval to pupal stages; thus, we suggest that some gustatory neurons are persistent larval neurons that form a new system in the pupa and adult [68]. When in the prepupal stage, the expression levels of TchiGR43a gene slightly declined, which might be due to the decrease in taste sensilla. Some larval sensilla are lost during metamorphosis and are replaced by new sensilla that originate from imaginal discs [69][70][71], indicating that gustatory neurons are reorganized in metamorphosis [72]. However, in the early phase of the pupal stage, the expression levels of TchiGR43a gene increased, which might be the result of an increase in gustatory neurons. For example, in Phormia regina, several hours after pupation some groups of gustatory neurons largely developed, and then, many new gustatory neurons appeared in all tarsal segments and the tibia [73]. Thus, we speculated that during metamorphosis, the increase in gustatory neurons and the production of new gustatory neurons might contribute to the high expression of the gustatory receptor gene-TchiGR43a. However, in the later period of the pupal stage, the expression levels of TchiGR43a declined, which could be caused by a decrease in gustatory neurons. In previous studies, the apoptosis of sensory neurons is observed at a late pupal stage [68,72,73].
In our study, the expression levels of TchiGR43a were similar between male and female adults, and we speculated that TchiGR43a not only acted as a sensor of fructose in chemosensory sensilla but also detected internal nutrients in other organs, which is supported by recent reports. In Drosophila adults, GR43a is also found in the brain and gut and is sufficient to evaluate nutritious carbohydrates and regulate feeding behaviour [24,74,75]. In the gut of Helicoverpa armigera, HaGR9 acts as a nutrient sensor to guide digestive processes and to protect from harmful substances [76]. The expression levels of TchiGR43a among different developmental stages and different genders suggested the existence of stage-specific and sex-specific gustatory tasks.
In the Xenopus-based functional studies, TchiGR43a cells showed response to D-fructose and myo-inositol and no response to the other nine sweet tastants. The response to D-fructose was significantly greater than that to myo-inositol. The results indicated that TchiGR43a was the receptor protein for D-fructose, which is also demonstrated in other in vitro studies. For example, DmGr43a of D. melanogaster and BmGr9 of B. mori shows a response only to Dfructose but not to other sugar tastants [3]. Similar results were also obtained from Helicovepa armigera, in which HarmGR4 were tuned to D-fructose [36]. In the present study, TchiGR43a, the orthologous gene of DmGR43a and BmGR9, showed responses to D-fructose and myo-inositol. Three possibilities might explain these differences. First, because TchiGR43a showed 27% and 28% identity to DmGR43a and BmGR9, respectively, at the amino acid level, the ligand-binding ability might differ. Second, in vivo, all the tastants must pass through the pores on the sensillum and diffuse into the lymph, with lymph then conveying the diffused tastants to a dendrite at which the final tastants and concentration may be different compared with the experimental cells that are bathed in tastant solution [8]. Third, the density of receptor proteins may be different in in vivo than in a heterologous expression system. However, in previous studies, based on experimental gene knockout and transgene rescue, DmGR43a responded to fructose and sucrose [24]. In H. armigera, HaGR9 responds to fructose, galactose and maltose [76]. Those studies could support our findings, thus, we speculate that D-fructose receptors share certain similarities and differences among insects, which may be due to differences among species or different ecological conditions and natural habitats. The EC 50 values of fructose are 0.055 mM for BmGr9 in B. mori [3], 0.045 M for HarmGr4 in H. armigera [36] and 0.069 M for AmGr3 in Apis mellifera [77], whereas in T. chilonis, the EC 50 of fructose was 0.023 M, indicating that the receptor of T. chilonis are more sensitive than those of lepidopterans.

Conclusions
We identified, for the first time, the D-fructose receptor in T. chilonis and verified its function. Behavioural and electrophysiological evidence was provided that T. chilonis responded to and had a preference for D-fructose. Moreover, the expression of TchiGR43a in different developmental stages and genders might also indicate involvement in stage-specific and sex-specific gustatory tasks. Studies on the relationship among D-fructose, D-fructose receptor and physiological behaviours may not only help us understand the underlying molecular mechanism of insect feeding behaviour but also shed light on developing a new strategy in massive production of T. chilonis for use in biocontrol.