The Desaturase Gene Family is Crucially Required for Fatty Acid Metabolism and Survival of the Brown Planthopper, Nilaparvata lugens

Desaturases are essentially required for unsaturated fatty acid (UFA) biosynthesis. We identified 10 genes encoding putative desaturases in the transcriptome database of the brown planthopper (BPH), Nilaparvata lugens. These include eight First Desaturase family genes, one cytochrome b5 fused desaturase gene (Nlug-Cytb5r) and one Sphingolipid Desaturase gene (Nlug-ifc). Transcript level profiling revealed significant variation in the expression patterns of these genes across tissues and developmental stages, which occur in a gene-specific manner. Interestingly, their expression was also modulated by the insect food source: the mRNA levels of Nlug-desatC and Nlug-Cytb5r were down-regulated, but the expression level of Nlug-desatA1-b and Nlug-desatA1-c were elevated in the BPH fed on the resistant rice variety Babawee as compared to the non-resistant variety Taichun Native 1 (TN1). Silencing Nlug-desatA1-b, Nlug-desatA1-c, or Nlug-Ifc reduced fatty acid composition and abundance in female BPH 1-d-old-adults compared to controls. Whereas, single knockdown of all ten desaturase genes significantly increased mortality of BPH nymphs compared with controls. Of the ten desaturase genes, knockdown of Nlug-desatA1-b and Nlug-desatA2 caused the highest mortality in BPH (91% and 97%, respectively). Our findings offer a base for expression and functional characterization of newly identified desaturase genes in BPH, and may contribute to RNA interference-based pest management strategies.


Introduction
In eukaryotes, unsaturated fatty acids (UFA) perform diverse biological functions such as regulation of membrane fluidity [1,2], energy storage [3,4] and signaling [5,6]. In biosynthesis of UFAs, desaturases are the key enzymes. These catalyze the induction of unsaturated bonds into an acyl chain at specific positions. Desaturases are categorized into two independent phylogenetic groups namely soluble acyl-acyl carrier protein (ACP) desaturases and membrane-bound fatty acid desaturases [7]. ACP desaturases are mainly found in the plastids of higher plants. They specifically catalyze the conversion of saturated fatty acids into monounsaturated acids, e.g., oleic acid synthesis (conversion of C18:0 to C18:1) [8]. Membrane-bound fatty acid desaturases are identified and annotated the desaturase family genes in the BPH genome. By combining sequence alignments, phylogenetic construction, expression pattern analysis and RNA interference (RNAi), some functions of BPH desaturase genes have been elucidated. We report that desaturase genes play critical role in the survival and fatty acid metabolism of BPH. These findings improve our understanding of the molecular mechanisms underlying UFA metabolism and survival in BPH.

Developmental Expression Patterns of Nlug-desat Genes
The transcript level of each Nlug-desat gene varied with developmental stage, and each gene presented a specific developmental expression pattern. In general, Nlug-desatA1-a and Nlug-desatE showed low expression levels in egg and 1-d-old female, and their transcript levels peaked in 3-and 4-d-old males. Nlug-desatA1-b, Nlug-desatD, and Nlug-ifc had relatively constant expression levels at different developmental stages; meanwhile, the expression of Nlug-desatA1-b showed an increasing trend during the male adult stage. Nlug-desatA1-c had low expression levels at the egg stage but also showed high expression in the later adult stages with a considerable variation. Nlug-desatA2 exhibits significantly lower expression levels during all five nymph stages compared to that in the 3-d-old males. The expression levels of Nlug-desatB peaked in the 4-d-old female. Nlug-desatC showed a rising trend of expression before emerging but with insignificant difference among stages. Nlug-Cytb5r had lower expression levels in the egg stage, first-instar larvae, and 3-d-old female compared with that in the 3-d-old male. (Figure 4 and Table S2).  Letters indicate significant differences among different treatments (P < 0.05). Statistical information is provided in Table S2 and Table S3. Letters indicate significant differences among different treatments (p < 0.05). Statistical information is provided in Table S2 and Table S3.

Tissue-Specific Expression Patterns of Nlug-desat Genes
Tissue-specific expression pattern analysis revealed that Nlug-desatA1-a, Nlug-desatC, and Nlug-desatE had high but not significant expression in the integument. On the other hand, Nlug-desatA1-b showed a low expression level in the ovary. Nlug-desatA1-c was mainly expressed in the body fat and integument. Nlug-desatA2 had a maximum expression in the ovary and integument. On the other hand, Nlug-desatB, Nlug-desatD, and Nlug-ifc only showed a slight variation of expression levels among different tissues. Nlug-Cytb5r had a relatively high expression level in the head ( Figure 5 and Table S4).

Tissue-Specific Expression Patterns of Nlug-desat Genes
Tissue-specific expression pattern analysis revealed that Nlug-desatA1-a, Nlug-desatC, and Nlug-desatE had high but not significant expression in the integument. On the other hand, Nlug-desatA1-b showed a low expression level in the ovary. Nlug-desatA1-c was mainly expressed in the body fat and integument. Nlug-desatA2 had a maximum expression in the ovary and integument. On the other hand, Nlug-desatB, Nlug-desatD, and Nlug-ifc only showed a slight variation of expression levels among different tissues. Nlug-Cytb5r had a relatively high expression level in the head ( Figure 5 and Table S4).

Figure 5.
Mean transcript levels (+SE, n = 3) of 10 desaturase genes in different tissues of BPH. Hd, head; Sg, salivary gland; In, integument; Mg, midgut; Fb, body fat; Ov, ovary. All tissues were dissected from 4-d-old-female adults. The results (threshold cycle values) of the qRT-PCR assays were normalized to the expression level of RPS11 (ribosomal protein S11, GenBank accession number: ACN79505.1). Letters indicate significant differences among different treatments (p < 0.05). Statistical information is provided in Table S4 and Table S5.

Differences in Nlug-desat Genes Expression between TN1 and Babawee Populations
To test for variation in expression of Nlug-desat genes with host variety, we compared the transcript levels of 10 Nlug-desat genes in 4-d-old female adults from populations of BPH reared on TN1 rice vs. Babawee rice using qRT-PCR methods. The mRNA levels of Nlug-desatA1-b and Nlug-desatA1-c were significantly higher in the whole body of Babawee-BPH compared with TN1-BPH. In contrast, Nlug-desatC and Nlug-Cytb5r had significantly higher expression levels in the whole body of TN1-BPH compared with Babawee-BPH ( Figure 6). The transcript levels of the other 6 Nlug-desat genes were similar between two different populations (Supplementary Figure S3). ACN79505.1). Letters indicate significant differences among different treatments (p < 0.05). Statistical information is provided in Table S4 and Table S5.

. Differences in Nlug-desat Genes Expression between TN1 and Babawee Populations
To test for variation in expression of Nlug-desat genes with host variety, we compared th nscript levels of 10 Nlug-desat genes in 4-d-old female adults from populations of BPH reared o 1 rice vs. Babawee rice using qRT-PCR methods. The mRNA levels of Nlug-desatA1-b and Nlu atA1-c were significantly higher in the whole body of Babawee-BPH compared with TN1-BPH. I trast, Nlug-desatC and Nlug-Cytb5r had significantly higher expression levels in the whole bod TN1-BPH compared with Babawee-BPH ( Figure 6). The transcript levels of the other 6 Nlug-des es were similar between two different populations (Supplementary Figure S3).

. Influence of Nlug-desats Suppression on Fatty Acids of BPH Female Adults
To investigate the possible involvement of Nlug-desats in the BPH fatty acid metabolism, Nlu at genes used in this study were knocked down individually, and the fatty acids from each RNA ated BPH sample were analyzed. The dsRNA for each Nlug-desat gene was injected into the third tar nymphs and the transcript levels of each target gene were efficiently suppressed by 77.7% 0% at 3 days post injection (dpi) (Supplementary Figure S4). Comparing the presence of saturate well as unsaturated fatty acids, we noticed four saturated fatty acids: lauric acid (C12:0), myrist d (C14:0), palmitic acid (C16:0) and stearic acid (C18:0). Three unsaturated fatty acids; palmitole d (C16:1), oleic acid (C18:1) and linoleic acid (C18:2), were identified in the female BPH with C16 C18:1 as the main constituent composition. Compared to the fatty acid profiles in dsGFP-BPH ockdown of Nlug-desatA1-b significantly reduced the levels of C14:0, C16:0, C18:1, and C18: reover, C12:0 and C16:1 were not found in Nlug-desatA1-b-silenced BPH, suggesting that thes The results (threshold cycle values) of the qRT-PCR assays were normalized to the expression level of RPS15 (ribosomal protein S15e, GenBank accession number: ACN79501.1). Asterisks indicate significant difference between treatments (* P < 0.05, ** P < 0.01 and *** P < 0.001, Student's t test).

Knockdown of Nlug-desats Decrease the Survival Rate of BPH Nymphs
We investigated whether knockdown of Nlug-desats could influence the survival rate of BPH. Compared with dsGFP-BPH (BPH injected with the dsRNA of GFP) and C-BPH (non-injected BPH), BPH injected with dsRNAs of seven Nlug-desats (Nlug-desatA1-b, Nlug-desatA1-c, Nlug-desatA2, Nlug-desatD, Nlug-desatE, Nlug-Cytb5r, and Nlug-Ifc) separately exhibited a significantly lower survival rate 7 dpi. The survival rates of dsdesatA1-b-, dsdesatA1-c-, dsdesatA2-, and dsdesatE-BPH were lower than 40% at 10 dpi (9%, 37%, 3%, and 34%, respectively). Interestingly, compared with dsGFP-BPH and C-BPH, knockdown of Nlug-desatC did not affect the survival rate of BPH during 7 dpi, but the survival rate of dsdesatC-BPH rapidly declined from 8 dpi and dropped to 40% at 10 dpi. Of the ten desaturase genes, silencing of Nlug-desatA2 resulted in the lowest survival rate of BPH (Figure 8). ND, no detection. Differences in specific fatty acid levels between the control group (dsGFP-BPH) and each treatment group were determined by Brown-Forsythe and Welch ANOVA followed by Dunnett's T3 multiple comparisons test. Asterisks indicate significant difference between dsGFP injection and each dsdesat injection treatments (p < 0.05). Statistical information is provided in Table S6.

Discussion
In this study, we identified and cloned 10 putative desaturase genes from the N. lugens transcriptome database. Phylogenetic analysis revealed that eight members of First Desaturase family genes in BPH could be clustered into five subfamilies (Figure 3). The Nlug-Cytb5r and Nlug-ifc belong to Cyt-b5-r [13] (Supplementary Figure S1) and infertile crescent (Ifc) protein, a Sphingolipid Desaturase with Δ4 activity [14] (Supplementary Figure S2), respectively.
Nlug-Cytb5r ortholog, commonly known as Cyt-b5-r, has reportedly been down-regulated in the activation of immune defense in honey bee larvae [15] and cold acclimatization of Drosophila virilis [16]. A low copy-number in the Cyt-b5-r subfamily of insects suggests their fundamental role in lipid metabolism [13]. The single-copy gene Nlug-Cytb5r, which had a minor variation of expression levels

Discussion
In this study, we identified and cloned 10 putative desaturase genes from the N. lugens transcriptome database. Phylogenetic analysis revealed that eight members of First Desaturase family genes in BPH could be clustered into five subfamilies (Figure 3). The Nlug-Cytb5r and Nlug-ifc belong to Cyt-b5-r [13] (Supplementary Figure S1) and infertile crescent (Ifc) protein, a Sphingolipid Desaturase with ∆4 activity [14] (Supplementary Figure S2), respectively.
Nlug-Cytb5r ortholog, commonly known as Cyt-b5-r, has reportedly been down-regulated in the activation of immune defense in honey bee larvae [15] and cold acclimatization of Drosophila virilis [16]. A low copy-number in the Cyt-b5-r subfamily of insects suggests their fundamental role in lipid metabolism [13]. The single-copy gene Nlug-Cytb5r, which had a minor variation of expression levels at developmental stages ( Figure 4 and Table S2) or in different tissues ( Figure 5 and Table S4), might serve as a housekeeping gene involved in lipid metabolism. However, knockdown of Nlug-Cytb5r did not significantly alter the fatty acid composition of BPH (Figure 7 and Supplementary Figure S5). These results suggest that Nlug-Cytb5r might not be essentially needed in the regulation of fatty acid levels during BPH development. However, the Nlug-Cytb5r gene might also be necessary for other essential metabolic functions in BPH. The expression of Nlug-Cytb5r might be stabilized within compensatory mechanisms past day 3 of RNAi, which might be sufficient to rescue the potential lethal phenotype of dsCytb5r-BPH. Previously, some beetle strains, such as Tribolium castaneum [42], Rhyzopertha dominica (F.) [43] and Sitophilus oryzae (L.) [44], lacking in functional copies of Cyt-b5-r, did not show adverse effects on development or reproduction under laboratory conditions [45]. It is also possible that other functional homologs of Cyt-b5-r gene, e.g., Cytochrome b5 (cyb5), can maintain the primary function of Cyt-b5-r as well as the survival of nymphs in the absence of Nlug-Cytb5r activity.
It has already been reported that Ifc protein is associated with meiosis [46] and thereby involved in the reproduction of Drosophila [18]. However, very little is on record regarding the involvement of Ifc protein in lipid metabolism. Similar to the Cyt-b5-r gene Nlug-Cytb5r, the Ifc gene Nlug-Ifc has a single-copy as well (Supplementary Figure S2). The constant mRNA level of Nlug-ifc during different developmental stages ( Figure 4) and in different tissues ( Figure 5) implies its crucial function in the lipid metabolism pathway. Additionally, knockdown of Nlug-ifc significantly down-regulated the composition of unsaturated and saturated fatty acids (Figure 7 and Supplementary Figure S5). In a recent report, disruption of fatty acid composition in Drosophila melanogaster larvae fed with SCD1 (also commonly known as ∆9 desaturases) inhibitor CAY10566 resulted in a lethal phenotype [22]. Here, we speculate that fatty acid deficiency caused by Nlug-ifc-knockdown might contribute to decreasing the survival rate of dsIfc-BPH (Figure 8).
In the present study, the size of the First Desaturase gene family in N. lugens (8 genes) is approximately similar to that in A. pisum (10 genes), D. melanogaster (7 genes) and Apis mellifera (9 genes). However, this is much smaller than the gene family size in T. castaneum (15 genes) or Bombyx mori (20 genes) [13]. Hahn et al. [47] attributed such differences in gene family size between genomes to adaptation. The replacement of conserved His residues in Nlug-desatE suggests the potential loss of desaturation activity (Figure 2). This is based upon the essential catalytic role of conserved histidine residues in the membrane-bound fatty acid desaturases [9,12]; however, additional research would be needed to verify its function. Interestingly, Nlug-desatE did not significantly affect fatty acid abundance or composition (Figure 7 and Supplementary Figure S5), but the survival rate of dsdesatE-BPH nymphs decreased significantly (Figure 8). These results combined with the high expression level of Nlug-desatE in the integument suggest the essential requirement of Nlug-desatE in basic metabolism excluding fatty acid desaturation during development.
The three most abundant fatty acids identified in BPH female adults are palmitic acid (C16:0), oleic acid (C18:1) and linoleic acid (C18:2) (Figure 7). This is consistent with previous research about BPH [41]. An important function of DESAT1 in D. melanogaster is C16:1 and C18:1 production by introducing a double bond in C16:0 and C18:0 [48]. The content of C18:0 in dsdesatA1-a-BPH is 1.59-fold higher than that in dsGFP-BPH (Figure 7), which suggests that Nlug-desatA1-a may be involved in the induction of an unsaturated bond in C18:0. Further, due to the tissue expression pattern of Nlug-desatA1-a ( Figure 5), Nlug-desatA1-a might play a role in cuticular hydrocarbon production. However, the total contents of saturated fatty acids and unsaturated fatty acids were not altered in the dsdesatA1-a-BPH (Supplementary Figure S5). Moreover, Nlug-desatA1-a silencing had only a small effect on the survival rate of BPH (Figure 8). We are of the opinion that other Nlug-desats in BPH may maintain the content of fatty acids in the absence of Nlug-desatA1-a activity.
Metabolic deficit is one of the important reasons for the lethal phenotype of DESAT1-deficient or -silenced D. melanogaster larvae. However, artificial diet supplemented with unsaturated fatty acids can rescue the survival of larvae [49]. Knockdown of Nlug-desatA1-b, Nlug-desatA1-c, and Nlug-Ifc not only significantly down-regulated unsaturated and saturated fatty acid composition but also significantly reduced the survival rate of BPH (Figures 7 and 8). So we suggest that these two desaturases play an essential role in lipid metabolism and thereby maintain the survival of BPH. The lethal phenotype in Nlug-desatA1-b-, Nlug-desatA1-c-, and Nlug-Ifc-knockdown BPH may be due to deficiency of unsaturated fatty acids (Figure 7 and Supplementary Figure S5). Furthermore, this lethal phenotype may have appeared partly due to effects on feeding behavior in the previously mentioned desaturase gene-knockdown BPH. Previous results have confirmed that the feeding behavior of D. melanogaster larvae can be blocked by using desaturase inhibitor [22]. Contrarily, knockdown of Nlug-desatB did not affect the unsaturated and saturated fatty acids (Figure 7), and only slightly decreased the survival rate of Nlug-desatB-knockdown BPH (Figure 8). However, the relative expression level of Nlug-desatB was higher in 4-d-old-female adults compared to second-instar nymphs (Figure 4) and showed a rising trend from 1-to 4-d-old females, suggesting that the Nlug-desatB may still play a role in the lipid metabolism of the adult female. Similar to Nlug-desatB, Nlug-desatE and Nlug-Cytb5r, single knockdown of Nlug-desatA2, Nlug-desatC, and Nlug-desatD does not affect the level of each single fatty acid composition (Figure 7) or the total content of saturated and unsaturated fatty acids (Supplementary Figure S5). However, silencing any of these genes individually increased the mortality of BPH nymphs (Figure 8). These genes may perform other crucial functions besides FA metabolism, or their roles in maintaining the dynamic balance of lipid metabolism are firmly related to the developmental stage and tissue of BPH. Furthermore, fatty acid analysis across tissues and lifecycles will be necessary to determine the role of these desaturase genes during the development of BPH.
Previously, high expression of desaturases has been reported in the generalist species Spodoptera littoralis, as well as the less polyphagous species Spodoptera frugiperda (rice strain), but not in S. frugiperda (corn strain) [50]. Desaturases may have a potential role in the insect-plant interaction and be involved in the adaptation of herbivores. Interestingly, the expression of some desaturase genes in BPH was also modulated by food source. Higher expression levels of Nlug-desatA1-b and Nlug-desatA1-c in Babawee-BPH suggest their probable involvement in the adaptation of BPH to the resistant rice variety Babawee. Conversely, significantly higher transcript levels of Nlug-desatC and Nlug-Cytb5r in TN1-BPH compared with Babawee-BPH suggest the acceleration of lipid metabolism in TN1-BPH ( Figure 6).

Insect Rearing and Plant Growth
BPHs were originally provided by the Chinese National Rice Research Institute, Hangzhou, Zhejiang, China, and maintained on fresh Taichun Native 1 (TN1) rice seedlings under controlled conditions (27 ± 1 • C, 70 ± 10% relative humidity and 14/10 h light/dark photoperiod). The Babawee-BPH for gene-expression analysis of two different BPH populations were maintained on Babawee, a rice variety containing the resistance gene Bph4, for more than 40 generations under the same controlled conditions of the TN1-BPH.

Identification and Amplification of Nlug-desat Genes
Ten desaturase genes characterized by a fatty acid desaturase type I domain (InterPro ID: IPR005804) in D. melanogaster [51] were used to find potential Nlug-desat genes in BPH based on its fat body transcriptome database [41] by using BLAST search. The candidate Nlug-desat genes were confirmed by blastx search against the non-redundant protein sequences database (nr). Ten Nlug-desat genes were obtained by RT-PCR from total RNA isolated from 4-d-old brachypterous BPH females. The primers (Table S7) were designed on the basis of transcriptome data of BPH fat bodies [41]. The PCR products were cloned into the pMD19-T vector (TaKaRa) and sequenced. Sequences were deposited in GenBank with the accession numbers (MH271225-MH271234).

Sequence Analysis and Phylogenetic Construction
The open reading frame (ORF) of Nlug-desats was predicted by using the ORF Finder (https:// www.ncbi.nlm.nih.gov/orffinder/). Domains of Nlug-desats were searched using the Batch CD-search (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) and InterProScan (http://www.ebi. ac.uk/interpro/search/sequence-search). Amino acid sequences of insect desaturase sequences downloaded from NCBI (National Center for Biotechnology Information, http://www.ncbi.nlm. nih.gov) were aligned using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/). The evolutionary history was inferred by using the Maximum Likelihood method based on the Le Gascuel (2008) model [52]. This method was chosen after carrying a best-fit substitution model analysis in MEGA7 [53]. The tree with the highest log likelihood (−58066.33) is shown. The proportion of trees in which the associated taxa clustered together in the bootstrap test (100 replicates) is shown next to the branches. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value. A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories (+G, parameter = 0.8376)). The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The bar represents 0.1 amino acid substitutions per sequence position. There were a total of 311 positions (from 177 putatively functional genes) in the final dataset. Evolutionary analyses were conducted in MEGA7 [53]. Graphical representation of the phylogenetic tree were performed with Interactive Tree of Life (version 3.5.1) [54,55]. Phylogenetic analysis of Cyt-b5-r and Ifc were constructed as outlined above for Cyt-b5-r (425 amino acid positions derived from 26 putatively functional genes) and Ifc (321 amino acid positions from 18 putatively functional genes).

RNAi Experiment
A unique region of each Nlug-desat gene was amplified by RT-PCR with primers including a T7 promoter sequence (Table S7). The purified PCR products were used to synthesize dsRNAs by using MEGAscript T7 High Yield Transcription Kit (Ambion, Austin, TX, USA). Third or fifth-instar nymphs were injected by using the FemtoJet (Eppendorf, Hamburg, Germany) microinjection device [57]. Each nymph was injected with about 0.25 µg of dsRNA of individual Nlug-desat or GFP (dsGFP, control) or not injected (C-BPH, control). The levels of Nlug-desat transcripts in the whole body of injected and control samples (third-instar nymphs were injected) were investigated 3 dpi. Individual nymphs were pooled (n = 20) for RNA extraction. The results (threshold cycle values) of the qRT-PCR assays were normalized to the expression level of RPS15. Three replicates were analyzed.
Due to the massive consumption of BPH in bioassay, the survival rates of BPH were investigated in four independent experiments on a different date. The Nlug-desatA1-a-, Nlug-desatD-, and Nlug-Cytb5r-silenced BPH were used in experiment 1; the Nlug-desatCand Nlug-desatE-silenced BPH were used in experiment 2; the Nlug-desatA1-c-, Nlug-desatB-, and Nlug-ifc-silenced BPH were used in experiment 3; and the Nlug-desatA1-band Nlug-desatA2-silenced BPH were used in experiment 4. Each experiment contained a dsGFP-BPH group and a C-BPH group as controls. Twenty insects were used for each treatment, and each treatment was performed five times in parallel. The number of surviving BPH nymphs was recorded every day.

Fatty Acid Analysis
Total lipids were extracted from groups of BPH female adults at 3 dpi of Nlug-desats or GFP (fifth-instar nymphs were injected). The BPH samples (9 female adults) were homogenized in 600 µL petroleum ether using a Precellys 24 homogenizer (Bertin) and centrifuged at 10,000 × g for 5 min at 4 • C. The supernatant was transferred into pre-weighed centrifuge tubes, evaporated by a rotavapor till dried and re-weighed.
Fatty acid concentrations were measured by forming fatty acid methyl esters (FAMEs) as described by Li et al. [58]. Ten microliters of methyl heptadecanoate (4 µg/µL) was added into 100 µL supernatant as internal standard. FAMEs were analyzed by using Agilent GC6890 gas chromatography equipped with a flame ionization detector (280 • C) through Agilent DB-23 column (60-m × 0.25-mm id 0.15-µm). Chromatography conditions: 250 • C inlet temperature, 242.3 kPa, splitless; 1 µL injection volume; N 2 carrier gas (30 mL/min); H 2 flow 40 mL/min, air flow 445 mL/min; the oven temperature was held at 50 • C for 1 min, increased to 175 • C at 25 • C /min, then at 3.5 • C /min to 230 • C and finally held at 230 • C for 5 min. FAMEs were identified by comparison with the Supelco ® 37 Component FAME Mix. The fatty acid contents were calculated based on three experiments.

Data Analysis
The statistical analyses were conducted by using the PASW ® Statistics 18 and Prism 8.0.2 software. Developmental and tissue-specific gene expression patterns were analyzed by using Brown-Forsythe and Welch ANOVA followed by Tamhane's T2 multiple comparisons test. Mean transcript levels of Nlug-desatB and Nlug-desatD in different tissues of BPH were analyzed by using ordinary one-way ANOVA followed by Tukey's test. For fatty acid analysis, differences in specific fatty acid levels between a control group (dsGFP-BPH) and each treatment group were determined by Brown-Forsythe and Welch ANOVA followed by Dunnett's T3 multiple comparisons test. Due to the undetectable levels of C12:0 and C16:1 in dsdesatA1-b-BPH, they were excluded from the multiple comparisons, and differences in the ratio of C16:1/C16:0 between dsdesatA1-b-BPH and dsGFP-BPH was analyzed by using t-test with Welch's correction. Differences in BPH survival rate between treatments and silencing efficiency of BPH desaturase genes were determined by one-way ANOVA followed by Duncan's test. Student's t-test for comparing differences in Nlug-desat genes expression between TN1 and Babawee populations.

Conclusions
We have identified and annotated the acyl-CoA desaturase gene family in BPH, which contains eight First Desaturases, one sphingolipid delta-4 desaturase, and one cytochrome-b5-related desaturase. Results have confirmed the role of Nlug-desats in lipid metabolism and survival of BPH upon different rice varieties. These findings are of paramount significance with respect to BPH desaturases and offer aid in developing novel strategies for pest control.