Molecular cloning, expression, and functional analysis of the chitin synthase 1 gene and its two alternative splicing variants in the white-backed planthopper, Sogatella furcifera (Hemiptera: Delphacidae)

Chitin synthase is responsible for chitin synthesis in the cuticles and cuticular linings of other tissues in insects. We cloned two alternative splicing variants of the chitin synthase 1 gene (SfCHS1) from the white-backed planthopper, Sogatella furcifera. The full-length cDNA of the two variants (SfCHS1a and SfCHS1b) consists of 6408 bp, contains a 4719-bp open reading frame encoding 1572 amino acids, and has 5′ and 3′ non-coding regions of 283 and 1406 bp, respectively. The two splicing variants occur at the same position in the cDNA sequence between base pairs 4115 and 4291, and consist of 177 nucleotides that encode 59 amino acids but show 74.6% identity at the amino acid level. Analysis in different developmental stages showed that expression of SfCHS1 and SfCHS1a were highest just after molting, whereas SfCHS1b reached its highest expression level 2 days after molting. Further, SfCHS1 and SfCHS1a were mainly expressed in the integument, whereas SfCHS1b was predominately expressed in the gut and fat body. RNAi-based gene silencing inhibited transcript levels of the corresponding mRNAs in S. furcifera nymphs injected with double-stranded RNA of SfCHS1, SfCHS1a, and SfCHS1b, resulted in malformed phenotypes, and killed most of the treated nymphs. Our results indicate that SfCHS1 may be a potential target gene for RNAi-based S. furcifera control.


Results
Identification and characterization of SfCHS1. The full-length cDNA sequence of SfCHS1 was obtained by multiple PCR amplifications and RACE. The full-length nucleotide and deduced amino acid sequences of SfCHS1 are shown in Fig. 1. The complete cDNA sequence of SfCHS1 is 6,408 bp in size. The ORF of SfCHS1 is 4,719 bp long and encodes a protein of 1,572 amino acid residues with a predicted molecular weight of 180.6 kDa and a pI of 6.72. The SfCHS1 cDNA includes a 5′ non-coding region of 283 bp and a 3′ non-coding region of 1,406 bp.
On the basis of the deduced amino acid sequence, 16 transmembrane helices (TMHs) were predicted using the TMHMM Server v.2.0, suggesting that SfCHS1 is a membrane-associated protein. Similar to other known insect CHS proteins, SfCHS1 has an N-terminal domain (domain A) containing nine TMHs; a central domain (domain B) that contains two signature motifs, EDR (852-854) and QRRRW (889-893), and two other motifs that are highly conserved in insect chitin synthases, CATMWHET (579-586) and QMFEY (790-794) 41 ; and a C-terminal domain (domain C) that contains seven TMHs and another signature motif SWGTR (1071-1075) that may play a role in chitin translocation 2,42 . Using the 3DLigandSite Server 43 , a ligand-binding site was identified in the amino acid region 581-750, and a putative catalytic domain at position 579-900 was predicted using the SMART program. The Paircoil program identified a coiled-coil region following transmembrane helix five of the C domain. In addition, six possible N-glycosylation sites were predicted using the NetNGlyc 1.0 Server, suggesting that the SfCHS1 protein may be glycosylated. However, analysis of deduced amino acid sequences using the SignalP 4.1 Server did not identify a signal peptide.
Comparative analysis of alternative splicing exons of SfCHS1. Analysis of the SfCHS1 cDNA sequence revealed two alternative splicing variants, named SfCHS1a and SfCHS1b (deposited in GenBank with accession numbers KY350143 and KY350144). The alternative exons are found in the same region (4115-4291) of the SfCHS1 cDNA (Fig. 1), and have 177 nucleotides that encode 59 amimo acid residues (Fig. 2). Alignment of the deduced amino acid sequences indicated that the identity between SfCHS1a and SfCHS1b is 74.6%. Each exon codes for a highly conserved transmembrane helix, and the flanking sequences consist of an intracellular and an extracellular domain, respectively 24,44 . Sequence alignment and phylogenetic analysis. Multiple sequence alignment of CHS1 proteins indicated a high degree of amino acid sequence homology among different insect species. For instance, the SfCHS1 protein shows 98% and 97% identity with that from the hemipteran L. striatellus (LsCHS1, AFC61179) and N. lugens (NlCHS1, AFC61181), respectively. It also shares identities of 81%, 73%, 71%, and 70% with the chitin  On the basis of the amino acid sequences of known insect CHSs, a phylogenetic tree was constructed using MEGA 6.06 based on the neighbor-joining method. The result indicated that the CHS1 and CHS2 genes originated from one ancestral gene and are closely related, but they clearly grouped into two different phylogenic branches (Fig. 3). The result is consistent with the findings of the previous studies 1,2,19,26 . Further, all hemipteran chitin synthases appeared to have a common ancestor in the lineage as indicated by the high bootstrap values (82~100), but they seemed to have lost the CHS2 gene during subsequent evolution. The chitin synthase from S. furcifera, SfCHS1, is clustered into the CHS1 family in the tree, and the identity of SfCHS1 to CHS1s was markedly  higher than identity to CHS2s from other insects (Fig. 3A). Moreover, the two splicing variants, SfCHS1a and SfCHS1b, grouped into two different phylogenetic classes (Fig. 3B).
Developmental-and tissue-specific expression of SfCHS1 and its two alternative splicing variants. qPCR was used to analyze the expression profiles of SfCHS1 and its two alternative splicing variants at different developmental stages (Fig. 4). The results revealed that SfCHS1 and its alternative variants were constitutively expressed in the 18 examined developmental stages. The relative expression levels of SfCHS1 were . Relative expression levels of SfCHS1 and its two alternative splicing variants in different developmental stages of S. furcifera. Expression levels at 18 different time points in eggs, nymphs (from firstinstar to fifth-instar nymphs), and adults were determined by qPCR. The S. furcifera 18S rRNA was used as an internal reference gene. The relative expression was calculated based on the value of the lowest expression which was arbitrarily set to 1. Data are means ± SE of three biological replications. The age in days of the insects is indicated, e.g., EG1, first day of eggs; lL1, first day of first-instar nymphs; AD1, first day of adults. higher just after each molting and reached a peak 1 day after eclosion. Specifically, the lowest expression levels for SfCHS1 were observed in third-day adults. For SfCHS1a, the expression patterns appeared to be similar to those of SfCHS1, but the relative transcript levels were lower in second-day adults. In contrast, SfCHS1b showed a different expression pattern to SfCHS1 and/or SfCHS1a, with the highest expression level being recorded 2 days after each molt. To investigate where SfCHS1 and its two alternative splicing variants are expressed, five different tissues from the integument, fat body, gut, ovary, and head were dissected for a tissue-specific expression experiment (Fig. 5). The results showed that SfCHS1 was mainly expressed in the integument, and that its expression was 75-, 11-, 42-, and 5-fold higher in the integument, fat body, ovary, and head than in the gut, respectively. SfCHS1a was also predominantly expressed in the integument, whereas SfCHS1b was primarily expressed in the gut and fat body.
RNAi response induced by injection of dsRNA. To verify whether RNAi is able to decrease target gene expression, sequence-specific dsRNAs for SfCHS1, SfCHS1a, and SfCHS1b were prepared in vitro and injected into first-day fifth-instar nymphs. Thereafter, qPCR was performed using total RNA isolated from dsRNA-injected insects as templates. The qPCR analysis indicated that the transcript levels of the target genes were markedly down-regulated at 72 h after dsRNA injection when compared with those of dsGFP-injected control insects (Fig. 6). More specifically, the expression of SfCHS1 was reduced by approximately 79% in the dsSfCHS1-injected nymphs. After RNAi of the SfCHS1a gene, there was no decrease in the level of SfCHS1b mRNA, even though Figure 5. Expression profiles of SfCHS1 and its two alternative splicing variants in different tissues of S. furcifera. The S. furcifera 18S rRNA was used as an internal reference gene. The relative expression was calculated based on the value of the lowest expression which was arbitrarily set to 1. Data are means ± SE of three biological replications. Different lower-case letters above the bars indicate significant differences (P < 0.05, Duncan's multiple range test in One-way ANOVA). SfCHS1a expression showed a 67% decrease. Similarly, after RNAi of SfCHS1b, the transcript level of SfCHS1b was reduced by approximately 64%, whereas SfCHS1a expression did not appear to be affected. Consequently, we assumed the dsRNA-mediated silencing to be gene specific.
After successful silencing of SfCHS1 and the two alternative splicing variants, mortality rates and lethal phenotypes of injected insects were recorded. It was clearly apparent that nymphs injected with 100 ng/head SfCHS1 dsRNA could not shed their old cuticle, and were trapped within the exuviae, leading to 100% mortality (Fig. 7). Following SfCHS1a dsRNA injection, 42% of individuals died before reaching the adult stage. Nevertheless, 49% of individuals died after eclosion, among which 36% of nymphs were able to molt to become adults but exhibited a notably abnormal phenotype. Moreover, 13% failed to shed their appendages and eventually died (Fig. 7). Following SfCHS1b dsRNA injection, only 15% of nymphs died before eclosion, whereas 85% of individuals successfully underwent molting to become adults. In contrast, 92% of individuals in the dsGFP-injected control group survived and had a normal phenotype (Fig. 7).
The fifth-instar nymphs of S. furcifera subjected to RNAi for the SfCHS1 gene displayed several distinct phenotypes. When injected with dsRNA of SfCHS1, three abnormal phenotypes were observed, and the insects eventually died: shrunken abdomen that was smaller than that of normal nymphs (I); the old cuticle only slightly splitted open on the head and thorax (II); and the old cuticle cracked to certain level but the whole insect body was still encased (III) (Fig. 8). After injection with SfCHS1a dsRNA, three typical lethal phenotypes were present, which included: nymths partially shed their old cuticle but the old cuticle could not be completely detached from the body, particularly from the tail (IV); nymphs were able to molt and become adults, but the adults were unable to extricate their appendages (V); and nymphs molted successfully but the new cuticle was crimpled and the wings were malformed (VI) (Fig. 8). However, we found no obvious differences in visible phenotypes between individuals in the dsSfCHS1b-and dsGFP-injected groups (Fig. 8).  The survival rate of insects following the injection of dsRNAs on the first-day of fifth-instar nymphs. 100 ng dsRNA was injected into each nymph. The age in days of the insects is indicated, e.g., 5L1, first day of fifth-instar nymphs; 5L2 and 5L2′ represent the two 12 hours in 1 day; AD, adults. Data are mean ± SE from three biological replications with fifty insects in each group.

Discussion
Chitin synthases play important roles in chitin biosynthesis during insect growth and development. It is known that most insects usually possess both CHS1 and CHS2. CHS1 is primarily expressed in the exoskeleton structures and is crucial for the synthesis of chitin required for the cuticle and tracheae, whereas CHS2 is expressed in midgut epithelial cells for production of chitin in the PM 25 . In this study, we obtained the full-length cDNA encoding chitin synthase from the hemipteran S. furcifera. Alignment and phylogenetic analysis indicated that CHS from S. furcifera belongs to the CHS1 group. By searching of the genomes and transcriptomes of the hemipteran insects, it was demonstrated that these species seem to lost one of the two CHS genes during evolution, and only one CHS gene exists [18][19][20] . This result is probably associated with the fact that Hemiptera insects lack the PM 26 . Our result also indicated that the SfCHS1 cDNA sequence is 6,408 bp in length and encodes a protein with a predicted pI of 6.72. The slightly acidic pI is conducive to its function in the cuticle. Similar to the CHS1 protein of other insects, SfCHS1 was predicted to be a 180.6-kDa membrane protein that contains 16 TMHs. The distribution and conserved number of these transmembrane segments in SfCHS1 allow the central catalytic domain (domain B) to face the cytoplasm, where the UDP-N-acetylglucosamine (UDP-GlcNAc) substrate is accessible. Its catalytic domain contains the highly conserved chitin synthase signature motifs CATMWHET, QMFEY, EDR, and QRRRW, which have been implicated to be essential for the catalytic mechanism 1,41,45 . Among the 16 TMHs, five are located immediately adjacent to the catalytic domain, forming a topological feature named the five-transmembrane span (5-TMS) region. This topology is found in all insect chitin synthases 18,19,23,24,46 . Consistent with other insect CHS1 proteins, SfCHS1 was predicted to include a conserved coiled-coil region immediately following the 5-TMS region, which is orientated toward the extracellular space and is a potential region for protein-protein oligomerization, or functions as a signal for vesicular trafficking 19,23,[47][48][49] . Alternative splicing plays a vital role in regulating gene function by expanding the diversity of expressed mRNA transcripts 46 . Many previous studies have demonstrated that alternative splicing appears to occur in the CHS1 gene 1,46,50 . In the present study, we also detected the presence of two alternative splicing exons of 177 bp in SfCHS1. However, it is surprising that no alternative exons have been identified in the genome of the hemipteran insect A. glycines 18 . A similar absence of alternative exons has also been reported in the hemipteran Toxoptera citricida 20 and thus it appears that alternative exons of the CHS1 gene are present in S. furcifera but are absent in aphids. The relationship between the production and evolution of alternative splicing thus requires further investigation.
In the present study, we performed qPCR expression analysis of SfCHS1 and its two alternative exons at different developmental stages in S. furcifera. Our results indicated that the expression of SfCHS1 was periodically repeated at each molting cycle. The transcript level of SfCHS1 peaked after molting, declined during each inter-molting phase and then increased again before the next molt, which may be associated with the requirement of chitin. Similar phenomena have also been observed for the transcript patterns of CHS1 in N. lugens 19 , Manduca sexta 41 , T. castaneum 46 and Ostrinia furnacalis 14 . Indeed, previous studies have shown that CHS1 is essential for eggshell formation and egg hatching in T. castaneum 30 , and that CHS1a mRNA expression plays a vital role in chitin synthesis of the serosal cuticle in Aedes aegypti 44 . In the current study, we also observed a relatively high expression of SfCHS1 in S. furcifera eggs. These results indicate that constitutive expression of SfCHS1 might be necessary in S. furcifera. Furthermore, the developmental expression patterns of SfCHS1a were similar to those of SfCHS1, but differed from those of SfCHS1b. Similar results were obtained by Wang et al. 19 in N. lugens and Yang et al. 23 in B. dorsalis. These results accordingly indicate that SfCHS1a and SfCHS1b probably play different roles in the biosynthesis of chitin during insect growth and development.
Further, the expression profiles of SfCHS1 and its two alternative exons were also investigated in various tissues. The results showed that SfCHS1 was predominately expressed in the integument, and ovary, with the highest levels of expression being observed in the integument. This is consistent with the fact that CHS1 is responsible for chitin biosynthesis in the epidermis. However, SfCHS1 was expressed at very low levels in the gut. Although the hemipteran insects lack PM, chitin was also detected in the lining of the gut of Myzus persicae 51 . The trace amounts of SfCHS1 transcripts in the gut might be responsible for the chitin-containing structures. Additionally, the observed low expression of SfCHS1 mRNA in the gut might be alternatively explained by the fact that the tracheae are tightly integrated into gut tissues and thus it is very difficult to completely remove these from the gut due to small size of the body 52 . The weaker expression of CHS1 in the gut was also detected in L. migratoria 16 , N. lugens 19 and Plutella xylostella 53 and these were believed to be due to contamination from the tracheal tissues. Also, we had detected a relatively high level of expression in the ovary. Similar results have been observed in Mythimna separata 54 , where MsCHS1 was highly expressed in the ovary. A previous study using the fluorescently labeled lectin technique had also documented that chitin was present in A. aegypti ovaries as well as in the eggs and egg shells 55 , suggesting the importance of CHS1 gene in insect reproduction. A low expression of SfCHS1 in S. furcifera head was also observed. Similar results have also been observed in P. xylostella 53 and Bombyx mori 15 , where the CHS1 gene was expressed in their head. Expression of CHS1 is known to be integument-specific. Therefore, we speculated that expression in the head was probably due to the CHS1 gene in the epidermis of the head. Moreover, we noted that the expression patterns of SfCHS1a were similar to those of SfCHS1, with the highest levels in the integument, whereas an exceedingly high expression of SfCHS1b was detected in the gut and fat body. However, a previous study on Anopheles gambiae has shown that AgCHS1a and AgCHS1b share the same transcript patterns and are expressed at considerable levels in the carcass (ie the insect body after its digestive canal is removed) 56 . Future work will be needed to address how CHS1a and CHS1b are involved in the physiological function of the various tissues in different insect species. Gene silencing through dsRNA feeding and dsRNA injection has been successfully used for studying the functions of essential genes in hemipteran insects 10,19,20,39,[57][58][59][60] . In the present study, to ascertain the functional difference among SfCHS1 and its two transcript variants, specific dsRNAs targeting SfCHS1, SfCHS1a, and SfCHS1b were synthesized and injected into fifth-instar nymphs. When fifth-instar nymphs on day 1 were injected with SfCHS1 dsRNA, qPCR result showed that RNAi of SfCHS1 strongly suppressed the expression of SfCHS1, thus new cuticle could not form normally due to the reduction of chitin. This result was supported by a similar study from T. castaneum 29 . In this species, TcCHS1-specific RNAi reduced the chitin content of whole larvae. Indeed, the morphological observation indicated that all treated planthoppers were unable to shed their old cuticle and died before reaching the adult stage. Such altered phenotypes are similar to those of B. dorsalis 23 , Leptinotarsa decemlineata 61 and L. migratoria 62 whose CHS1 and/or UDP-N-acetylglucosamine pyrophosphorylases (UAP), two important components in chitin biosynthesis pathway, were silenced by RNAi. Further, in L. migratoria, knockdown of LmUAP1 or LmCHS1 led to synthesize the very thin new cuticle during their molting 62 . These results suggest once again that UAPs and CHSs play crucial role during insect ecdysis and metamorphosis.
When CHS1a and CHS1b dsRNA of the two alternative variants was injected into fifth-instar nymphs, respectively, qPCR showed no cross-silencing between SfCHS1a and SfCHS1b. SfCHS1a dsRNA-mediated silencing affected the growth and development of treated insects, leading to lethal phenotypes. In contrast, dsRNA-mediated silencing of SfCHS1b caused no obviously phenotypic defects, although the mortality was slightly increased compared with the dsGFP-injected control group. Our result suggested that SfCHS1a was essential for insect molting and metamorphosis. Similar results have been observed in N. lugens 19 and B. dorsalis 23 , in which silencing of CHS1a expression by in vivo RNAi caused phenotypic defects in molting and resulted in mortality of the injected insects, whereas nymphs also injected with CHS1b dsRNA exhibited a normal phenotype. However, in L. migratoria, nymphs injected with CHS1b dsRNA exhibited a crimpled cuticle phenotype, resulting in over 50% mortality 16 . These results indicate that there is considerable variation in the efficiency of RNAi-mediated silencing of CHS1b in various insect orders.
S. furcifera is an important insect pests on rice in some Asia-Pacific countries. In recent years, destructive outbreaks of S. furcifera have been increasing in China, causing severe losses in rice yield. At present, control of planthoppers still relies upon spraying chemical insecticides. However, considering the adverse impact of insecticides on the ecological environment and on human health, new pest management strategies urgently need to be developed. A previous study demonstrated that feeding with the trehalose phosphate synthase (TPS) dsRNA in N. lugens led to reduction levels of TPS mRNA and disturbed the development of nymphs, suggesting that administering dsRNA corresponding to important genes by oral delivery may be a means for the control of phloem-sucking insects 39 . In another study, when N. lugens nymphs were fed on the transgenic rice plants expressing dsRNAs of the hexose transporter gene, the carboxypeptedase gene and the trypsin-like serine protease gene, levels of expression of the target genes in the midgut were suppressed; nevertheless, lethal phenotypic effects after dsRNA feeding were not observed 40 , either because the amount of dsRNA-uptake by the insects was insufficient or because RNAi target genes were not sensitive in this species. Therefore, there is an urgent need to elucidate the physiological functions of vital candidate genes from different insect species. Overall, our results indicated that injecting dsRNA of CHS1 into S. furcifera nymphs could lead to a significant mortality, suggesting that SfCHS1 may be a candidate gene for use in S. furcifera control.

Conclusion
In conclusion, we successfully cloned and characterized two alternative splicing variants of the chitin synthase 1 gene (SfCHS1) from S. furcifera. Phylogenetic analysis demonstrated that these genes belong to the CHS1 gene family. The genes were expressed at all developmental stages. Further, SfCHS1 and SfCHS1a were mainly expressed in the integument, whereas SfCHS1b was predominately expressed in the gut and fat body. Our RNAi-based gene silencing inhibited the transcript levels of the corresponding variants, resulted in malformed phenotypes, and killed most of the treated nymphs. These results indicate that SfCHS1 may be a potential target gene for RNAi-based S. furcifera control.

Materials and Methods
Insect rearing. The planthoppers used in the present study were originally collected from a rice paddy field On the basis of the transcriptome sequencing data (SRR116252) of S. furcifera 63 , four short cDNA sequences encoding SfCHS1 were identified. To obtain a larger cDNA fragment, six pairs of gene-specific primers (Table 1) were designed using Primer Premier 6.0 (Palo Alto, CA, USA). The ends were amplified by 3′-and 5′-RACE using a SMARTer RACE Kit following the manufacturer's instructions (Clontech, Mountain View, CA, USA).
PCR amplifications were carried out using LA Taq ® polymerase (TaKaRa, Dalian, China) in 25-μL reaction mixtures containing 2 μL dNTP (2.5 mM), 2.5 μL 10 × LA PCR Buffer (Mg 2+ plus), 1 μL each primer (10 mM), and l μL cDNA templates. The thermal cycling conditions were as follows: one cycle of pre-denaturation at 94 °C for 3 min, followed by 30 cycles of denaturation at 94 °C for 30 s, annealing at 50-55 °C (according to primer annealing temperature) for 30 s, and extension at 72 °C for 1-2 min (according to amplified fragment size), with a final extension at 72 °C for 10 min. The amplified products were examined by 1% agarose gel electrophoresis, and the target band of products was purified using an EasyPure ® Quick Gel Extraction Kit (Transgen Biotech, Beijing, China). Purified DNA was cloned into a pMD18-T vector (TaKaRa, Dalian, China) and sequenced by Sangon Biotech (Shanghai, China).

Identification of alternative splicing exons of SfCHS1.
It is known that the insect CHS1 gene exists as two alternative splicing variants. To identify the alternatively spliced exons of SfCHS1, one pair of gene-specific primers (ASV-F: 5′-TGACGATAACAGTGATACCA-3′ and ASV-R: 5′-GAATCGGCGTCATAGTCC-3′) were designed based on the full-length sequence of SfCHS1. cDNA was synthesized as described above. PCR was carried out via one cycle of pre-denaturation at 94 °C for 3 min, followed by 30 cycles of denaturation at 94 °C for 30 s, annealing at 51 °C for 30 s, and extension at 72 °C for 1 min, with a final extension at 72 °C for 10 min. A 648-bp amplified product was cloned into a pMD18-T vector and sequenced. cDNA and amino acid sequence analysis. The sequenced fragments were assembled using SeqMan software to obtain the full-length sequence of SfCHS1 cDNA. The nucleotide sequence was edited using DNAMAN 7.0 (Lynnon Biosoft, CA, USA). Homology searches were performed using the NCBI BLAST program (https:// blast.ncbi.nlm.nih.gov/Blast.cgi). The open reading frame (ORF) of SfCHS1 cDNA was identified using ORF finder (https://www.ncbi.nlm.nih.gov/orffinder/). The ProtParam tool at ExPASy (https://www.expasy.org/) was used to compute the molecular weight and theoretical isoelectric point (pI) of the deduced protein sequence 64 . N-glycosylation sites were analyzed using the NetNGlyc 1.0 Server (http://www.cbs.dtu.dk/services/NetNGlyc/), and the signal peptide was predicted using the SignalP 4.1 Server (http://www.cbs.dtu.dk/services/SignalP/). The TMHMM v.2.0 program (http://www.cbs.dtu.dk/services/TMHMM/) was used to analyze the transmembrane helices 65 . The putative coiled-coil regions were predicted using the Paircoil program 66 . phylogenetic analysis of insect chitin synthases. Phylogenetic trees were constructed using MEGA 6.06 based on the neighbor-joining (NJ) method 67   Developmental-and tissue-specific expression of SfCHS1 and its two alternative splicing variants. S. furcifera at stages ranging from eggs to adults were sampled to determine the developmental stage expression profiles by quantitative real-time PCR (qPCR). Five different tissue samples from the integument, fat body, gut, ovary, and head were dissected from first-day fifth-instar nymphs and third-day adults to examine tissue-specific expression. Three biological replications were performed for each sample. Total RNA was isolated from the whole body of nymphs and adults at each stage or from the different tissues using an HP Total RNA Kit (with gDNA removal columns; Omega bio-tek, Norcross, GA, USA). An AMV RT reagent Kit (Sangon Biotech) with an oligodT primer was used to synthesize first-strand cDNA. The most unique nucleotide regions of SfCHS1, SfCHS1a, and SfCHS1b were selected for expression analysis (the selected regions are shown in Figs 1 and 2), and the primers used for qPCR are listed in Functional analysis of SfCHS1 and its two alternative splicing variants using RNAi. To further investigate the biological functions of SfCHS1 and its two alternative splicing variants, SfCHS1a and SfCHS1b, RNAi was carried out by injecting S. furcifera nymphs with sequence-specific dsRNA. The most unique nucleotide regions of SfCHS1, SfCHS1a and SfCHS1b were selected for dsRNA synthesis (the synthesized regions are shown in Figs 1 and 2), and the primers added a T7 RNA polymerase promoter (Table 2) were used to synthesize dsRNA. Templates for in vitro transcription reactions were synthesized by PCR from the plasmid DNA of SfCHS1, SfCHS1a, and SfCHS1b using primers. The PCR products of SfCHS1, SfCHS1a, and SfCHS1b were subcloned and sequenced to determine the specificity. The expected fragments were then purified using an EasyPure ® Quick Gel Extraction Kit (Transgen Biotech). The concentration of the purified products was determined using a Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific) and these products were then used for in vitro transcription reactions.  dsRNAs were synthesized using a MEGAscript ® RNAi Kit (Ambion, Carlsbad, CA, USA) according to the user manual provided by the manufacturer. In vivo RNAi in S. furcifera nymphs was carried out as previously described 19,70 . First-day fifth-instar nymphs were anesthetized with carbon dioxide for approximately 30 s and subsequently used for microinjection. Each group included 50 nymphs and treatments were performed in triplicate. One hundred nanograms of dsRNA was injected into nymphs between the prothorax and mesothorax using a Nanoliter 2010 Injector (injection speed, 25 nL/s) (World Precision Instruments, FL, USA). Equivalent volumes of dsGFP were used for control injections. Injected nymphs were maintained on fresh rice under the conditions described above until eclosion, and thereafter phenotype and mortality were observed daily. Photographs were taken using a Keyence VH-Z20R stereoscopic microscope (Keyence, Osaka, Japan). Subsequent to injection, 10 nymphs were selected randomly from each replication for mRNA-level detection.
statistical analysis. Statistical analysis of all data was performed using SPSS 13.0 software (IBM Inc., Chicago, IL, USA). Data values are represented as the mean ± SE of three replications. A one-way ANOVA and Duncan's multiple range test (P < 0.05) were used to calculate the relative expression of each sample. For RNAi experiments, significant differences in mRNA levels between each of the dsRNA-injected groups and the dsGFP group were analyzed using t-tests.

Data Availability
The data were deposited in GenBank with accession numbers KY350143 (SfCHS1a) and KY350144 (SfCHS1b).