Identification, Characterization and Expression Analysis of TRP Channel Genes in the Vegetable Pest, Pieris rapae

Transient receptor potential (TRP) channels are critical for insects to detect environmental stimuli and regulate homeostasis. Moreover, this superfamily has become potential molecular targets for insecticides or repellents. Pieris rapae is one of the most common and widely spread pests of Brassicaceae plants. Therefore, it is necessary to study TRP channels (TRPs) in P. rapae. In this study, we identified 14 TRPs in P. rapae, including two Water witch (Wtrw) genes. By contrast, only one Wtrw gene exists in Drosophila and functions in hygrosensation. We also found splice isoforms of Pyrexia (Pyx), TRPgamma (TRPγ) and TRP-Melastatin (TRPM). These three genes are related to temperature and gravity sensation, fine motor control, homeostasis regulation of Mg2+ and Zn2+ in Drosophila, respectively. Evolutionary analysis showed that the TRPs of P. rapae were well clustered into their own subfamilies. Real-time quantitative PCR (qPCR) showed that PrTRPs were widely distributed in the external sensory organs, including antennae, mouthparts, legs, wings and in the internal physiological organs, including brains, fat bodies, guts, Malpighian tubules, ovaries, as well as testis. Our study established a solid foundation for functional studies of TRP channels in P. rapae, and would be benefit to developing new approaches to control P. rapae targeting these important ion channels.

The small cabbage white butterfly, Pieris rapae (Lepidoptera: Pieridae), known as imported cabbageworm or European cabbage butterfly, is one of the most common and widely spread pests of Brassicaceae plants [11]. P. rapae larvae eat vegetable leaves reducing cabbage plants to stems [12]. In order to reduce the loss of production, chemical control methods as the most convenient to spread in large area are still widely used in agricultural production. While as the long-term use of a few chemical pesticides, the P. rapae has produced resistances [13]. Therefore, it is necessary to develop alternative agents.
In recent years, TRP channels have gradually emerged as the molecular target of noxious chemicals. A previous study showed that Drosophila TRPA1 channel is required to avoid the insect repellent citronellal [14]. Painless mutant flies are defective in avoidance of isothiocyanate (ITC), the pungent ingredient of wasabi [15]. TRPA1 of Anopheles gambiae are directly activated by citronellal and benzaldehyde [14]. The commercial insecticides pymetrozine, pyrifluquinazon and afidopyropen directly act Nan-Iav complexes specifically expressed in chordotonal neurons, causing insects uncoordinated and eventually dead [16]. In general, TRP channels have become potential molecular targets for insecticides or repellents [17]. Thus, it is meaningful to identify the sequences and characteristics of TRP channels in important agricultural or hygienic insects.
In this study, we identified 14 TRP genes in P. rapae and investigated their spliced isoforms and tissue distributions. Supplementary identification of TRP genes was also done in three other lepidoptera insects (the monarch butterfly Danaus plexippus, the Asian swallowtail Papilio Xuthus and the tobacco hornworm Manduca sexta) and the brown planthopper Nilaparvata lugens (Stål) (Hemiptera: Delphacidae), one of the most important rice pests in temperate and tropical regions; the damp-wood termite Zootermopsis nevadensis (Isoptera: Archotermopsidae), one common pest in temperate forests; and the German cockroach Blattella germanica (Blattodea: Blattellidae), one notorious sanitary pest.

Insect
The colonies of P. rapae were collected from cabbage fields in Hangzhou, Zhejiang province, China and were all reared continuously on cabbage leaves in the temperature-controlled room (25 ± 1 • C, 70%-80% relative humidity (R.H.), and a photoperiod of 14:10 (L:D) h with the light intensity of 3500-4000 Lux). Adult P. rapae were fed with 10% sucrose water containing 0.15% riboflavin for several days.

Identification of TRP Channels
First of all, to search TRP genes in each species, we used TRP protein sequences of Drosophila to blast transcriptomic or proteomic data released in National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/). If any genes were not found, we would go to blast the genome database for supplementing or confirming our results. Then, candidate genes were further verified using BLASTP versus the Database of Drosophila Genes and Genomes (FlyBase) (http://flybase.org/). Insects 2020, 11, 192 3 of 14 One exception is TRPA5, because Drosophila does not have TRPA5. As long as the candidate genes were not Pyx orthologues, we considered that sequences as TRPA5.

Total RNA Isolation, Reverse Transcription and Splice Variants Detection
Total RNAs from fifth-instar P. rapae larvae were extracted using the TRIzol (Invitrogen, Carlsbad, CA, US) according to the manufacturer's instructions. Single-strand complementary DNA (cDNA), synthesized from 1000ng of total RNA using the TransScript First-Strand cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China), were as then used as templates for PCR reactions. The PCR primers for investigating the splice variants were designed with primer5 (Table S1) [18]. PCR products were separated by electrophoresis on 1% agarose gel. The gel where target fragments located were cut off and DNA were purified with Agarose Gel DNA Extraction Kit-250 prep (Esay-Do, Hangzhou, China). Then the purified PCR products were cloned into pEASY-Blunt Zero (Transgen, Beijing, China) and transformed into DH5α competent E. coli cells (Vazyme, Nanjing, China). Positive clones were sequenced by second-generation sequencing (Sunya, Hangzhou, China).

Bioinformatic Analysis
Sequence similarity for orthologous genes was performed using BLASTP program on the National Center for Biotechnology Information (NCBI). Protein domains were predicted by TMHMM 2.0 (http://www.cbs.dtu.dk/services/TMHMM-2.0/) and SMART (http://smart.embl-heidelberg.de/). Sequences were aligned by ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/). The amino acid sequences in phylogenetic tree were aligned through MAFFT software v7.123b with the default parameters [19], and then IQ-TREE was applied for the maximum likelihood analyses [20], under the LG + G4 nucleotide substitution model predicted by ModelFinder [21] and with 1000 bootstrap replicates for testing the credibility of evolutionary tree [22].

Real-Time Quantitative PCR and Statistical Analysis
Total RNAs from antennae, compound eyes (with optic lobes), mouthparts, legs, wings, brain (without optic lobes), malpighian tubules, gut and fat bodies as well as ovaries and testes of about 25 adults of P. rapae (3-5 days old) were extracted using the TRIzol (Invitrogen, Carlsbad, CA, US) according to the manufacturer's instructions. The concentrations and quantities of RNA were measured by using a Nanodrop 2000 spectrophotometer (Thermo Scientific Inc., Bremen, Germany). Then 1ug RNA was used to perform reverse transcription reaction in each sample. A total of 10×cDNA was used as a template for real-time quantitative PCR (qPCR). qPCR reactions were performed using CFX96™ Real-Time PCR Detection System (Bio-rad, Hercules, CA, USA) with ChamQ TM SYBR qPCR Master Mix (Without ROX) (Vazyme, Nanjing, China) according to the manufacturer's instructions. Three sample biological replicates and two mechanical repeats were carried out in each sample. Sequence-specific primers for qPCR were designed using AlleleID 6 (Table S2) [23]. cDNAs were normalized by amplification of 18s rRNA [24]. Quantification of transcript level was quantified using the 2 -∆∆CT method [25]. One-way analysis of variance (ANOVA) and Tukey's multiple comparison test were employed to analyze the data of relative expression levels in various tissues. For the comparison of expression levels between splice variants and gene duplications, unpaired t test was applied. The analysis software was Statistical Product and Service Solutions (SPASS).

Identification and Phylogenetic Analysis of TRP Channels in P. Rapae and Other Lepidoptera Insects
We identified 14 TRP channel genes in P. rapae, the monarch butterfly Danaus plexippus, the Asian swallowtail Papilio Xuthus, and 15 TRP channel genes in the tobacco hornworm Manduca sexta (Table 1). Phylogenetic analysis arranged these sequences into TRPC, TRPA, TRPN, TRPV, TRPM and TRML, but not into TRPP (Figure 1). Compared with other insects, Pkd2, Brv and Hymenoptera-specific TRPA Insects 2020, 11, 192 4 of 14 (HsTRPA) were not found in the above four Lepidoptera insects as well as the silkworm Bombyx mori (Table 2). However, TRPA5 were found in their genomes and two TRPA5 existed in M. sexta. In addition, all five lepidoptera insects had two Wtrw genes. They were well divided into two classes in evolutionary analysis ( Figure 1).   [18]. Under the LG + G4 nucleotide substitution model, IQ-TREE was applied for the maximum likelihood analyses with 1000 bootstrap replicates [19]. The bootstrap value is indicated next to each branch. The accession number of TRP channels identified in this study are listed in an Additional file (Table S3). Other accession numbers of identified TRP channels originated from previous studies [26,27].  Mus musculus (Mm) TRP channels were aligned through MAFFT software v7.123b with the default parameters [18]. Under the LG + G4 nucleotide substitution model, IQ-TREE was applied for the maximum likelihood analyses with 1000 bootstrap replicates [19]. The bootstrap value is indicated next to each branch. The accession number of TRP channels identified in this study are listed in an Additional file (Table S3). Other accession numbers of identified TRP channels originated from previous studies [26,27].

Sequence Analysis, Splice Variants and Gene Duplications of TRP Channelsin P. Rapae
We did a detailed sequence analysis of TRP channels in P. rapae. A blastp program revealed that most TRPs in P. rapae shared a high sequence identity (above 50%) with TRPs in D. melanogaster (Table 1). Only PrPain and PrWtrw, 2 of TRPA subfamily shared under 50% sequence identity with their homologous sequence in D. melanogaster, which may be due to the rapid evolution of TRPA subfamily in insects [28]. All other P. rapae TRPs had six transmembrane domains except that PrNan had only four transmembrane domains for its incomplete sequence. Consistent with other species [7], Insects 2020, 11,192 6 of 14 TRPN of P. rapae had the most N-terminal ankyrin repeats domain, followed by TRPA, TRPV and TRPC, TRPM and TRPML still have no ankyrin repeats (Table 1).
After we found two splice variants for PrPyx, we mapped the splice variants to the genomic sequence for checking the GA-AG rule about introns. Then we performed PCR on cDNA of P. rapae larvae using specific primers and sequenced the PCR products ( Figure 2A). Results of these two steps indicated that the two splice isoforms both consisted of 14 exons which included the first two alternative and mutually exclusive exons and the last 13 shared exons ( Figure 2B). Their open reading frames (ORFs) both encoded 928 amino acids, but there were still 9 different amino acids at the 5 'end of protein sequences ( Figure 2C). Even though, the two deduced protein sequences for PrPyx both had 9 N-terminal ankyrin repeats domains and 6 transmembrane domains. Then we verified two splice variants for PrTRPγ in the same way ( Figure 3A). The difference between two transcripts originated from the deletion of 9 nucleotides at the ends of 14th exon ( Figure 3B). Predictions of protein structure showed that the difference of variable splicing existed in TRP domain ( Figure 3C), which indicated that this splicing maybe played an important role in physiological function. Ankyrin repeats domains and transmembrane domains had no difference between the two transcripts ( Figure 3C). PrTRPM was also spliced into two isoforms ( Figure 4A). By investigating genomic and transcriptome data, we found the exons of their initial codons were mutually exclusive ( Figure 4B). Therefore, there were about 30 N-terminal amino-acid residues differences in their predicted protein sequences ( Figure 4C).  Pr Py x 1 : Pr Py x 2 : DmPy x :         As mentioned above, gene duplications of Wtrw were found in all five lepidoptera insects (Table 1). Two Wtrw genes of P. rapae existed in the same scaffold with the reverse direction of transcription ( Figure 5A). The results of functional domain prediction showed that PrWtrw-1/2 had the same transmembrane domains and different ankyrin repeats domains. In addition to the shared last eight ankyrin repeats, PrWtrw-1 had another unique ankyrin repeat in front of its N-terminal ankyrin repeats domain ( Figure 5B).
As mentioned above, gene duplications of Wtrw were found in all five lepidoptera insects (Table  1). Two Wtrw genes of P. rapae existed in the same scaffold with the reverse direction of transcription ( Figure 5A). The results of functional domain prediction showed that PrWtrw-1/2 had the same transmembrane domains and different ankyrin repeats domains. In addition to the shared last eight ankyrin repeats, PrWtrw-1 had another unique ankyrin repeat in front of its N-terminal ankyrin repeats domain ( Figure 5B). Arrows showed the direction of gene transcription. (B) Multiple sequence alignment deduced for PrWtrw-1, PrWtrw-2 and DmWtrw. Ankyrin repeats were indicated with dotted lines and marked as ANK1-2. Transmembrane regions were indicated with solid lines and marked as TM1-6. The protein sequences used for multiple sequence alignment were in the attached fasta file.

Tissue Expression Profiles of TRP Channel in Adult P. Rapae
Insects have to constantly sense the environmental stimuli and adjust homeostasis to ensure their normal life activities. TRP channel functions not only in sensory physiology, but also in homeostasis [2]. Thus, in order to investigate the possible roles of TRPs in P. rapae, we evaluated the relative transcription levels of TRPs in the external sensory organs, including antennae, mouthparts, legs, wings, and in the internal physiological organs, including brains, fat bodies, guts, Malpighian tubules, ovaries, testis. Results showed that transcription levels of PrTRP were different in various tissues, and the expression level in legs was the highest, while that in wings, fat bodies, Malpighian tubes and ovaries were low ( Figure 6A). PrTRPL was mainly transcribed in antenna and brains ( Figure 6B). Abundant transcripts of PrTRPγ were detected not only in brains but also in guts ( Figure   Pr      Transmembrane regions were indicated with solid lines and marked as TM1-6. The protein sequences used for multiple sequence alignment were in the attached fasta file.

Tissue Expression Profiles of TRP Channel in Adult P. Rapae
Insects have to constantly sense the environmental stimuli and adjust homeostasis to ensure their normal life activities. TRP channel functions not only in sensory physiology, but also in homeostasis [2]. Thus, in order to investigate the possible roles of TRPs in P. rapae, we evaluated the relative transcription levels of TRPs in the external sensory organs, including antennae, mouthparts, legs, wings, and in the internal physiological organs, including brains, fat bodies, guts, Malpighian tubules, ovaries, testis. Results showed that transcription levels of PrTRP were different in various tissues, and the expression level in legs was the highest, while that in wings, fat bodies, Malpighian tubes and ovaries were low ( Figure 6A). PrTRPL was mainly transcribed in antenna and brains ( Figure 6B). Abundant transcripts of PrTRPγ were detected not only in brains but also in guts ( Figure 6C). PrTRPA1 was highly expressed in the brains, but it is rarely detected in the Malpighian tubes ( Figure 6D). Compared with the high transcription level of PrTRPA5 in guts, the transcription levels in other tissues were very low and had no significant difference ( Figure 6E). Like PrTRPA5, the highest messenger RNA (mRNA) levels of PrPyx were observed in guts ( Figure 6F). mRNA levels of PrPyx(A) in antenna and testis were significantly higher than that of other tissues ( Figure 6F). In mouthparts and brains, mRNA levels of PrPyx (primer-amplified both PrPyx(A) and PrPyx(B) were significantly higher than that of PrPyx(A), indicating that PrPyx(B) may be the main form in these two tissues of adult P. rapae ( Figure 6F). The highest expression levels of both PrWtrw-1 and PrWtrw-2 were detected in antenna ( Figure 6G). Higher transcription levels of PrWtrw-1 than PrWtrw-2 were found in legs, wings, brains, guts and Malpighian tubules ( Figure 6G). Transcription levels of PrPain were higher in brains, mouthparts and legs than in other tissues ( Figure 6H). PrNompC, PrIav and PrNan were all detected the highest mRNA levels in antennae ( Figure 6I-K). In addition, PrIav was highly transcribed in testis, and PrNan was highly transcribed in legs, brains and guts ( Figure 6J-K). TRPM(A) had higher mRNA levels in antenna, guts, Malpighian tubules and ovaries, while TRPM(B) only had higher mRNA levels in guts and Malpighian tubules ( Figure 6L). TRPM(B) mRNA was significantly more than TRPM(A) mRNA in guts and Malpighian tubules, and was significantly fewer that TRPM(B) mRNA in ovaries ( Figure 6L). TRPML had the highest mRNA level in Malpighian tubules, followed by brains and guts ( Figure 6M). 6C). PrTRPA1 was highly expressed in the brains, but it is rarely detected in the Malpighian tubes ( Figure 6D). Compared with the high transcription level of PrTRPA5 in guts, the transcription levels in other tissues were very low and had no significant difference ( Figure 6E). Like PrTRPA5, the highest messenger RNA (mRNA) levels of PrPyx were observed in guts ( Figure 6F). mRNA levels of PrPyx(A) in antenna and testis were significantly higher than that of other tissues ( Figure 6F). In mouthparts and brains, mRNA levels of PrPyx (primer-amplified both PrPyx(A) and PrPyx(B) were significantly higher than that of PrPyx(A), indicating that PrPyx(B) may be the main form in these two tissues of adult P. rapae ( Figure 6F). The highest expression levels of both PrWtrw-1 and PrWtrw-2 were detected in antenna ( Figure 6G). Higher transcription levels of PrWtrw-1 than PrWtrw-2 were found in legs, wings, brains, guts and Malpighian tubules ( Figure 6G). Transcription levels of PrPain were higher in brains, mouthparts and legs than in other tissues ( Figure 6H). PrNompC, PrIav and PrNan were all detected the highest mRNA levels in antennae ( Figure 6I-K). In addition, PrIav was highly transcribed in testis, and PrNan was highly transcribed in legs, brains and guts ( Figure 6J-K). TRPM(A) had higher mRNA levels in antenna, guts, Malpighian tubules and ovaries, while TRPM(B) only had higher mRNA levels in guts and Malpighian tubules ( Figure 6L). TRPM(B) mRNA was significantly more than TRPM(A) mRNA in guts and Malpighian tubules, and was significantly fewer that TRPM(B) mRNA in ovaries ( Figure 6L). TRPML had the highest mRNA level in Malpighian tubules, followed by brains and guts ( Figure 6M). PrWtrw-1 ns ns *** * ns * *** ** ns ns Relative expression levels

Discussion
In this study, we identified 14 TRPs in P. rapae, D. plexippus, P. Xuthus, and 15 TRPs in M. sexta that be divided into TRPC, TRPA, TRPN, TRPV, TRPM and TRML, but not into TRPP based on their structure and phylogenetic analyses. As the most ancient TRP subfamily, TRPP channels are gated by Ca 2+ flux [29]. Drosophila TRPP involves in not only larval feeding behavior, but also in directional sperm movement [30,31]. However, TRPP were absent in the five lepidoptera species we investigated. It indicated the functions of TRPP may be compensated by other TRPs or alternative pathways in lepidoptera insects. In addition, we identified 3 TRPPs in Z. nevadensis, 2 TRPPs in B. germanica and 1 TRPPs in N. lugens ( Table 2), suggesting that the numbers and functions of TRPPs varied among different insects. Compared with only 1 TRPA channel in mammals, the member of this subfamily was highly diverse in arthropod species, especially insects ( Table 2) [26]. As shown above, although TRPA5 was absent in Drosophila and N. lugens, it was present in all five lepidoptera insects, Z. nevadensis and B. germanica (Table 2). However, the physiological roles and gating mechanism of TRPA5 channel had yet to be determined in any organism. Previous researches have shown the lack of TRPA1 probably can be compensated by HsTRPA [28,32]. TRPA1 was present in all eight insects we investigated, which may explain why HsTRPA was absent in their genomes. In our study, the duplication of Wtrw was only found in Lepidoptera insects, which may be due to their specific habitats and life histories. So far, studies on the functions of Wtrw have been limited in Drosophila, duplications of Wtrw in Lepidoptera will provide more perspectives for its physiological roles. We identified more than one Pain channels in Z. nevadensis and B. germanica. While only ZnPain-1 and BgPain-1 were clustered together with Pain channels of other ten insects, and their branch evolved later (Figure 1). It implied various physiological functions which required multiple Pain channels in ancient insects could be undertaken by only one Pain channel through evolution. The functions of disappeared Pyx in N. lugens may be compensated by other TRPs or alternative pathways.
Studies in Drosophila have shown that TRPC subfamily expressed in photoreceptors and the brain participate in phototransduction [9,33]. In this study, we failed to get clean samples of compound eyes for quantitative expression analysis, while we found high expression of TRPCs in the brains of P. rapae. Many bristles on leg, mouthparts and antennae of insects sense various stimuli all the time, so the highly transcribed PrTRP and PrTRPL in these three tissues may play a role similar to the mammalian TRPC sensing ever-changing environmental conditions in the primary cilia [34]. In mammal, TRPC channels function in intestinal fibrogenesis [35], which may explain the relatively high transcription levels of PrTRP and PrTRPγ, the homologs of TRPC, in P. rapae guts. Drosophila TRPA1 channel is expressed in the AC (anterior cell) neurons that are located in the brain for temperature sensation [36], and robust transcripts of PrTRPA1 in P. rapae brains were also observed. In addition, TRPA1 is involved in other sensory processes, such as the avoidance of aversive odorants, non-volatile irritants and mechanical stimuli [2,14]. Most sensory organs are located in insect antennae, legs, mouthparts and wings [2], so we successfully detected mRNAs of PrTRPA1 in these tissues. In the case that the gating mechanism and function of TRPA5 are not clear, we found mRNA level of PrTRPA5 was significantly higher than that of other tissues, which may provide some basis for the function studies of TRPA5 channel. Pyx [26], the closest relative of TRPA5 (Figure 1), was also observed to have high transcription levels in P. rapae guts, implying that they may regulate the function of guts. Drosophila Pyx channel that expressed in the Johnston's organ of antennae could regulate anti-gravitaxis behavior [37]. Accordingly, high mRNA level of PrPyx(A) was detected in antennae of P. rapae. Abundant transcripts of PrPyx in testis may affect reproductive behaviors. Consistent with previous reports, Wtrw and NompC were mainly transcribed in antennae, therefore they may be also related to hygrosensation and hearing respectively in P. rapae [38,39]. Pain was widely distributed in the external sensory organs and internal physiological organs. High mRNA levels of PrPain in the external sensory organs, such as legs, wings and mouthparts, suggested that PrPain was probably required for both thermal, chemical and mechanical nociception [40]. Abundant mRNA of PrPain in the internal physiological organs, such as brains, guts and ovaries, indicated that PrPain may had other extended roles in regulating various physiological processes. The TRPV channels Iav and Nan, another TRP channel expressed in insect antennae for hearing and gravity sensation [37,41], were also detected to have robust transcription in P. rapae antennae. Abundant mRNA of PrIav in testis may explain that mRNAs of Iav in male insects are higher than that in female [42]. This result also suggested that Iav may be involved in the reproductive physiology of male insects. Moderate expression levels of PrNan were detected in legs for the existence of peripheral neurons and brains for antennal neuron projections [39,43]. Moreover, the relatively abundant mRNA of PrNan in guts may be due to its roles similar to TRPV1 in mammalian gastrointestinal tract for secreting substances and protecting digestive tracts [44]. TRPM protein is required for Mg 2+ and Zn 2+ homeostasis and expressed in the Malpighian tubules to remove electrolytes and toxic components from the hemolymph [45,46]. Indeed, TRPM was transcribed at a high level in P. rapae Malpighian tubules as well as in guts, which indicated PrTRPM may not only regulate Mg 2+ /Zn 2+ homeostasis but also maintain normal physiological functions of guts. Similar to the high transcription levels of TRPML3 in the intestine and kidney [47,48], abundant transcripts of PrTRPML were also detected in guts and Malpighian tubules of P. rapae, indicating its roles in digestion and excretion. Neurodegeneration and accumulation of late-apoptotic cells always occurred in the brain of TRPML mutant Drosophila, which may be the reason why abundant transcripts of PrTRPML were transcribed in the P. rapae brain.  Table S3: Accession number of TRP channels identified in this study, Fasta file: The protein sequences used in multiple sequence alignments.