GFAT and PFK genes show contrasting regulation of chitin metabolism in Nilaparvata lugens

Glutamine:fructose-6-phosphate aminotransferase (GFAT) and phosphofructokinase (PFK) are enzymes related to chitin metabolism. RNA interference (RNAi) technology was used to explore the role of these two enzyme genes in chitin metabolism. In this study, we found that GFAT and PFK were highly expressed in the wing bud of Nilaparvata lugens and were increased significantly during molting. RNAi of GFAT and PFK both caused severe malformation rates and mortality rates in N. lugens. GFAT inhibition also downregulated GFAT, GNPNA, PGM1, PGM2, UAP, CHS1, CHS1a, CHS1b, Cht1-10, and ENGase. PFK inhibition significantly downregulated GFAT; upregulated GNPNA, PGM2, UAP, Cht2-4, Cht6-7 at 48 h and then downregulated them at 72 h; upregulated Cht5, Cht8, Cht10, and ENGase; downregulated Cht9 at 48 h and then upregulated it at 72 h; and upregulated CHS1, CHS1a, and CHS1b. In conclusion, GFAT and PFK regulated chitin degradation and remodeling by regulating the expression of genes related to the chitin metabolism and exert opposite effects on these genes. These results may be beneficial to develop new chitin synthesis inhibitors for pest control.

The relative expression levels of target genes and phenotype observation after dsRNA injection. The results of qRT-PCR showed that the relative expression levels of NlGFAT (Fig. 3A) and NlPFK ( Fig. 3B) were significantly decreased at 48 h and 72 h after dsGFAT and dsPFK injection, respectively, which suggests that dsRNA successfully inhibited the expression of target genes. In addition, after dsGFAT and dsPFK separate injection into the nymphs, three kinds of abnormal phenotypes including only molting difficulties, only wing deformities, and both, were observed (Fig. 3C). As seen above, silencing of both NlGFAT and NlPFK impacted on the development of N. lugens. www.nature.com/scientificreports/ Malformation rates and mortality rates of N. lugens after dsRNA injection. As shown in the figures, after 48 h or 72 h of GFAT inhibition, the malformation rates were 19.19% and 23.77% (Fig. 4A), respectively, and the mortality rates were 25.63% and 32.51% (Fig. 4B), respectively. Similarly, after dsPFK injection, the malformation rates were 13.76% at 48 h and 16.14% at 72 h (Fig. 4A), and the mortality rates were 18.06% at 48 h and 25.59% at 72 h (Fig. 4B). Thus, the mortality and malformation rates of N. lugens showed significant increases (p < 0.01) after dsGFAT or dsPFK injection alone; both GFAT and PFK thus have a great influence on the growth and development of N. lugens individuals.

Relative expression levels of chitin biosynthesis-related genes after dsRNA injection. One
GFAT, GNPNA, UAP and two PGM genes were identified in N. lugens 6 . When the PFK of N. lugens was inhibited, the relative expression levels of GFAT were decreased extremely significantly (p < 0.01) at 48 h and 72 h (Fig. 5A), but the expression of GNPNA, PGM2, and UAP showed an extremely significant increase (p < 0.01) at 48 h and an extremely significant decrease (p < 0.01) at 72 h (Fig. 5B,D,E). In contrast, the expression of PGM1 showed little change at 48 h whereas it increased extremely significantly (p < 0.01) at 72 h (Fig. 5C). The qRT-PCR results showed that the relative expression levels of GFAT, GNPNA, PGM1, and UAP were significantly decreased (p < 0.01) at 48 h and 72 h after dsGFAT injection (Fig. 5B,C,E) whereas that of PGM2 decreased extremely significantly (p < 0.01) at 48 h but showed little change at 72 h (Fig. 5D).   GFAT, glutamine:fructose-6-phosphate aminotransferase; PFK, phosphofructokinase. The head, leg, ovary, cuticle, fat body, midgut, and wing bud of N. lugens used to detect the tissue expression level of genes, were evenly collected from individuals of nymphs in different instars and from adults. Gene expression levels were measured by quantitative real-time PCR with 18S RNA as the internal control. Values are means ± SE from three independent measurements. Different letters indicate significant differences according to Duncan's test (p < 0.05).  Figure 3. Relative expression levels of GFAT (A) and PFK (B) after dsGFAT and dsPFK injection, and phenotype changes (C) after dsRNA injection in Nilaparvata lugens. GFAT, glutamine:fructose-6-phosphate aminotransferase; PFK, phosphofructokinase. The N. lugens used for the microinjection of RNAi were at the 1st day of 5th instar nymph stage. dsGFAT or dsPFK injection were used as the test groups whereas dsGFP injection was used as the control group. Gene expression levels were measured by quantitative real-time PCR with 18S RNA as the internal control. Values are means ± SE from three independent measurements. Different letters indicate significant differences according to Duncan's test (p < 0.05).  GFAT, glutamine:fructose-6-phosphate aminotransferase; PFK, phosphofructokinase. The N. lugens used for the microinjection of RNAi were at the 1st day of the 5th instar nymph stage and dsGFAT or dsPFK injection were used as the test groups whereas dsGFP injection was used as the control group. This experiment had three biological repeats. *, significant differences (P < 0.05); **, extremely significant differences (p < 0.01). The deformity rate of the dsGFP treatment group was zero, with "•" indicating the data.  (Fig. 6D,K). After injecting dsPFK into N. lugens on the 1st day of the 5th instar nymph, the relative expression levels of Cht2 to Cht8 (Fig. 6B-H), Cht10 (Fig. 6J) and ENGase (Fig. 6L) showed extremely significant increases (p < 0.01) at 48 h. But their subsequent trends were not the same. The expression levels of Cht2, Cht3, Cht4, Cht6, and Cht7 decreased significantly (p < 0.05) (Fig. 6B,C) and extremely significantly (p < 0.01) (Fig. 6D,F,G), whereas the expression levels of Cht8, Cht10, ENGase showed extremely significant declines at 72 h (p < 0.01) (Fig. 6H,J,L); the expression of Cht5 was restored to the same levels as the control group (Fig. 6E). Further, silencing PFK gene did not affect the expression of Cht1 and IDGF (Fig. 6A,K), but affected Cht9 as its expression decreased significantly (p < 0.05) at 48 h and showed an extremely significant increase (p < 0.01) at 72 h (Fig. 6I). GFAT, glutamine:fructose-6-phosphate aminotransferase; PFK, phosphofructokinase; Cht, chitinase; IDEF, imaginal disc growth factor; ENGase, endo-β-N-acetylglucosaminidase. Nilaparvata lugens larvae at the 1st day of the 5th instar stage were divided into three groups and injected with dsGFP, dsGFAT, and dsPFK, respectively. Insects were collected and used to determine the relative expression levels of Cht1 to Cht10 (A to J), IDGF (K), and ENGase (L) at 48 h and 72 h after dsRNA injection. Three replicates were performed per group. *, significant differences (P < 0.05); **, extremely significant differences (p < 0.01).

Discussion
GFAT catalyzes the rate-limiting step of the UDP-GlcNAc synthesis pathway. Because of its role in the development of insulin resistance in type 2 diabetes 39 , studies on GFAT are mostly focused on mammals, whereas there have been rather few studies on the GFAT gene in insects for a long time and it has only been reported in Drosophila melanogaster, Aedes aegypti, Haemaphysalis longicornis at present 21,40,41 . Northern blot analysis of Drosophila melanogaster and Aedes aegypti showed two bands for GFAT1, the ratios of which varied in different developmental stages; GFAT1 was localized by whole mount in situ hybridization to chitin synthesis-related tissues, suggesting that DmGFAT and AeGFAT are involved in chitin synthesis 21,40,42 . Our studies have shown that NlGFAT was expressed in all stages after the 4th instar, and that there was a significant difference in its expression during the molt period from the 5th instar to the adult stage (Fig. 1A). In addition, NlGFAT was highly expressed at the wing bud and cuticle ( Fig. 2A), which contains significant amounts chitin 43,44 . Therefore, GFAT and chitin metabolism are closely related. As rate limiting enzyme in the glycolytic pathway, PFK is closely related to diabetic cardiomyopathy 45 , and there have been few studies on PFK gene in insects. In Spodoptera litura, transcriptional expression of PFK occurs at a stable and low level during the period from larval stage to pupa, but its enzyme activity decreased dramatically in the pre-pupae and was recovered in pupae during metamorphosis 46 . In N. lugens, the expression levels of PFK were decreased dramatically during the period of the 4th instar to 5th instar, but were increased extremely significantly during the 5th instar nymph to adult stage (Fig. 1B). Glycolysis in the cytosol could produce ATP, the chemical energy in cells, which is used to run the reactions that maintain viability, growth, and proper function of individuals 47 . Flight muscles are tissues that require a large supply of energy 48 . In our experiments, NlPFK showed the highest expression in wing buds (Fig. 2B), which is consistent with the requirement of energy. Overall, the expression levels of GFAT and PFK change significantly during molting, consistent with the pace of chitin metabolism, and high expression in chitinous tissues, suggesting a link between them and chitin metabolism.
RNAi is a biological process that may be mediated by exogenous dsRNA, which is sliced into small RNAs, causing endogenous complementary mRNA silencing 49 . RNAi is considered as an important tool for gene function research 50 . The results of qRT-PCR showed that the relative expression levels of NlGFAT (Fig. 3A) and NlPFK (Fig. 3B) were significantly decreased after dsRNA injection, respectively, which suggesting that dsRNA successfully inhibited the expression of target genes. In the present study, we obtained many interesting experimental results after GFAT-knockdown or PFK-knockdown using RNAi. Most insects possess two CHS genes (CHS1 and CHS2), but N. lugens possesses only CHS1 with two transcript variants (CHS1a and CHS1b) 38 . RNAi against Nilaparvata lugens larvae on the 1st day of 5th instar stage were divided into three groups and injected with dsGFP, dsGFAT, and dsPFK, respectively. The dsGFP-treatment group was used as the control group. Three replicates were performed per group. *, significant differences (p < 0.05); **, extremely significant differences (p < 0.01). www.nature.com/scientificreports/ NlCHS1 and NlCHS1a causes high mortality rates and severe morphological malformations 38,51 . Knockdown of NlTRE1 could downregulate CHS1 and CHS1a and cause abnormal phenotypes 44 , and knockdown of TPS1 could downregulate the expression of CHS1, CHS1a, and CHS1b, resulting in extremely high malformation and mortality rates in N. lugens 52 , as well as HK-knockdown also could result in the downregulation of CHS1, CHS1a, CHS1b 53 . In our study, the mRNA levels of CHS1, CHS1a, and CHS1b were acutely decreased at 48 h and 72 h after dsGFAT injection (Fig. 7). In addition, molting difficulties and wing deformities were observed with dsGFAT injection (Fig. 3); the malformation rates and mortality rates of N. lugens were also increased extremely significantly after dsGFAT injection, compared to the dsGFP injection group (Fig. 4A,B). These results are consistent with previous studies. In locusts, reduced expression of miR-71 and miR-263 increased CHS1 and CHS10 mRNA expression, thus resulting in molting defects 54 . Similarly, in these experiments, reduced expression of NlPFK increased CHS1, CHS1a, and CHS1b (Fig. 7), along with high malformation rates and mortality rates in N. lugens (Fig. 4).
To further investigate the effects of GFAT and PFK genes on chitin metabolism in N. lugens, we detected the expression of chitin synthesis pathway genes after silencing GFAT and PFK. When TRE1-1, TRE1-2, and TRE2 in N. lugens were co-inhibited using RNAi, the relative expression levels of GFAT, GNPNA, PGM1, PGM2 and UAP were decreased significantly, but the relative expression of PGM2 was increased significantly at 72 h 55 . In addition, the same effects were achieved by injecting validamycin, a kind of trehalase inhibitor 6 . This suggested that TRE could regulate chitin synthesis by regulating the transcriptional levels of other enzymes involved in chitin synthesis, and that PGM1 and PGM2 might be functionally complementary 6,55 . HK-knockdown could also result in downregulation of GFAT, GNPNA, and UAP 53 . In our experiment, when dsGFAT was injected into the 5th instar nymph of N. lugens, other than PGM2 being expressed normally at 72 h compared with the control group (Fig. 5D), the relative expression levels of GFAT, GNPNA, PGM1, PGM2, and UAP were dramatically decreased (Fig. 5), similar to the result of N. lugens TRE and HK gene inhibition 6,53,55 . Therefore, silencing GFAT expression directly leads to impaired chitin synthesis by inhibiting the chitin pathway genes.
The role of PFK in regulating energy metabolism during insect development has been studied in Spodoptera litura, but it is unclear whether it affects the chitin synthesis pathway 46 . Radiometric glycolysis assays have demonstrated that low rates of glycolysis did not affect the overall level of incorporation of glucose-derived carbon into HP, but low PFK activity promotes channeling of F-6-P into HP 56 . In our study, when PFK was inhibited, the mRNA levels of GFAT were sharply declined at 48 h and 72 h (Fig. 5A). However, contrary to the interference results of GFAT, the expression levels of GNPNA, PGM2 and UAP were increased sharply at 48 h after PFK inhibition, but decreased significantly after 72 h (Fig. 5B,D,E), and the expression of PGM1 was still in contrast to PGM2 (Fig. 5C). We speculated that when PFK is inhibited, more fructose-6-phosphate flows into HP, as shown by radioactive glycolysis experiments 56 . Therefore, inhibition of PFK expression might have promoted CHS transcription by upregulating chitin synthesis pathway genes, which is contrary to GFAT inhibition. However, inhibition of NlPFK resulted in reduced transcription levels of NlGFAT, which may indicate the existence of other regulatory pathways.
In summary, silencing of GFAT or PFK affects the synthesis and degradation of chitin by interfering with the transcription levels of crucial chitin metabolizing enzymes, thus resulting in extremely high malformation rates and mortality rates. Moreover, GFAT and PFK have opposite effects on chitin synthesis in N. lugens. All the above results provide theoretical support for the discovery of new targets for pest control. However, all measurements were based on transcriptional levels, but data on protein levels are lacking, so measurements of enzyme activity and related metabolites at the tissue level, rather than at the individual level, will be considered.

Methods
Insect sourcing and culture conditions. The N. lugens used in this study were provided by the China National Rice Research Institute (Hangzhou, China), and the variety of all rice (Oryza sativa L.) cultivars was Taichung Native 1 (TN1) planted in cement tanks from April to October and in a greenhouse or growth chamber during winter. Insects were reared on fresh TN1 rice seedlings in an artificial climate chamber at 26 ± 1 ºC, 70% relative humidity, and 16 L:8 D (light:dark) photoperiod 55 . All experiments were performed under the same conditions. Developmental stages were synchronized by collecting new eggs laid by N. lugens, and the instar was judged based on the hind foot and antennae of the nymph.

Collection and dissection of N. lugens in different developmental stage. N. lugens individuals
used in gene expression stage analyses were obtained from the 4th instar nymphs on their first day, and after every 24 h until they reached the adult stage; 10 individuals were taken from each stage. Besides, female adults and male adults were also collected separately. The N. lugens used to detect the tissue expression level of genes were collected from 50 individuals of adults, and with a 1:1 ratio of male to female. The head, leg, ovary, cuticle, fat body, midgut, and wing bud of N. lugens were dissected in a saline solution (0.75% NaCl) under an EZ4 NlGFAT and NlPFK expression in several tissues and developmental stages using quantitative real-time polymerase chain reaction (qRT-PCR). cDNA synthesis and qRT-PCR were performed to analyze the distribution of NlGFAT and NlPFK using gene-specific primers (Table 1). Using 1 µg of total RNA as template, and a specifically designed Nl18S primer pair ( The expression of NlGFAT and NlPFK in several tissues and developmental stages was estimated by qRT-PCR with a SYBR Green master mix (SYBR Green Premix Ex Taq, Takara, Japan) in a Bio-Rad CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories Inc., Hercules, CA, USA). Each PCR was performed in a 20 µL volume, containing 1 µL cDNA, 1 µL (10 µM) of each primer, 7 µL ultrapure water, and 10 µL SYBR buffer 44,55 . The reactions were performed under the following conditions: preincubation at 95 ºC for 2 min; 39 cycles of 95 ºC for 5 s and annealing at 59 °C for 30 s; and a melting curve at 65-95 °C. Amplification of 18S RNA was used as an internal control 44,55 . Double-stranded RNA (dsRNA) synthesis and injections. The N. lugens cDNA template and specific primers ( Table 2) were used to amplify the NlGFAT and NlPFK genes with reverse transcription polymerase chain reaction (RT-PCR). The reaction procedure is set as follows: preincubation at 95 ºC for 3 min, 35 cycles at 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 1 min; and a final extension at 72 °C for 10 min. Purified GFAT and PFK amplicons were transcribed to synthesize dsRNA using the T7 RiboMax Express RNAi System (Promega Corporation, Madison, WI) 44 . A green fluorescence protein (GFP) gene amplicon was also used to synthesize dsRNA for being control group. Sense and anti-sense strands were separately produced using PCR and were then mixed for annealing. Reactions were incubated for 10 min at 70 °C and then placed on ice for 20 min. Finally, dsRNAs were purified with 95% ethanol and 4.4 M sodium acetate (pH 5.2), then washed with 70% ethanol, air Table 1. Gene-specific primers used for quantitative real-time polymerase chain reaction (qRT-PCR).

QNl18S
CGC TAC TAC CGA TTGAA  GGA AAC CTT GTT ACG ACT T   QNlGFAT  CTG GAC TTT GAC AGC GTT AC  GTG GTC GTT GTC GGAGC   QNlPFK  TGA CGT GAC AGG GTG GGT  ATG GCT TGG ATT TGG AAC T   QNlGNPNA  TGA GCT GCT GAA GAC ACT  CCT GAA TAA CGG TGA TGT A   QNlPGM1  AAC GAG ACG GTG GGA GAC  TCC TGG TAA GTG TTG AGC C   QNlPGM2  AGA GGA AGG TTG GGA GTG  CAT AAT TCG CGG AGA TAA  www.nature.com/scientificreports/ dried, and redissolved with DEPC. The integrity and quantity of dsRNAs were determined by spectrophotometer with Nanodrop 2000 (Thermo Fisher Scientific) and agarose gel electrophoresis 44 .
Using an IM-31 microinjector (NARISHIGE, Tokyo, Japan), dsGFAT and dsPFK (3000 ng of each) were injected into the abdomen of N. lugens on the 1st day of the 5th instar nymphs. Control groups were injected with dsGFP.
Sample statistics, collection and phenotype observations after injection. After dsRNA was injected into fifth-instar larvae of N. lugens, the malformation rates and mortality rates of N. lugens were counted at 48 h and 72 h, respectively. In addition, insects were randomly collected (excluding abnormal individuals) at 48 h and 72 h after injection to detect the relative expression of chitin metabolism-related genes. Collected samples were stored at -80℃. Photographs of abnormal insects were taken in different dsRNA injection treatments.
Quantification of chitin metabolism-related gene expression levels. N. lugens treated with dsRNA were used to extract the total RNA using the TRIzol reagent (Invitrogen, Carlsbad, California, USA), then firststrand cDNA synthesis was performed using the PrimeScript RT reagent kit with gDNA Eraser (Takara, Kyoto, Japan). Relative expression levels of chitin metabolism-related genes were estimated by qRT-PCR using genespecific primers (Table 1) with a SYBR Green master mix (SYBR Green Premix Ex Taq, Takara, Japan) in a Bio-Rad CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories Inc., Hercules, CA, USA). The specific steps have been mentioned previously. The 2 −△△CT method was used for analyzing relative gene expression 59 .
Statistical analyses. In this study, all data were analyzed using one-way analysis of variance (ANOVA) and are shown as the mean ± standard error (SE) of three biological replicates. Data on developmental and tissues expression patterns were analyzed using Duncan's test. In Duncan's test, different letter indicates a significant difference (p < 0.05). Other data was analyzed using the Tukey's test. In Tukey's test, a double asterisk indicates an extremely significant difference in mRNA levels (p < 0.01), and an asterisk indicates a significant difference (p < 0.05).

Data availability
The datasets generated or analysed during the current study are not publicly available but are available from the corresponding author on reasonable request. Table 2. Gene-specific primers used for double-stranded RNA synthesis. T7 sequence: 5′-GGA TCC TAA TAC GAC TCA CTA TAG G-3′.