NADPH Oxidase 5 Is Essential for Molting and Oviposition in a Rice Planthopper Nilaparvata lugens

Simple Summary The brown planthopper, Nilaparvata lugens, a plant phloem-sucking Hemipteran insect, has become the most destructive pest for rice—the major food source for half of the world’s population. Nilaparvata lugens possess robust fecundity, enabling population densities to increase quickly in a favorable environment. Nilaparvata lugens has also been used as a model system for ecological studies and for developing effective pest management. To better understand the regulation mechanisms of insect reproduction and to provide insights to improve pest control, we searched N. lugens genome and transcriptome databases and identified an NADPH oxidase 5 gene, which was specifically expressed in ovaries of female adults and freshly laid eggs in rice leaf sheaths. Although homologous NADPH oxidase 5 genes have been identified in many insect species, their reproductive functions remain unknown. In this work, our findings initially reveal a functional role of the NADPH oxidase 5 gene in molting and oviposition in N. lugens. This novel finding improves our understanding of the reproductive strategies in insects and provides a potential molecular target for effective pest control of the rice planthopper. Abstract The brown planthopper Nilaparvata lugens is a typical monophagous insect herbivore that feeds exclusively on rice sap. This insect pest causes serious damage to rice crops throughout East Asian countries. Chemical control remains the first choice for managing N. lugens populations; however, the use of insecticides has given rise to planthopper resurgence and additional environmental risks. Nilaparvata lugens is a model insect of Hemiptera because its whole genome sequence has been elucidated and is susceptible to RNA interference. In this study, our findings revealed that a superoxide-generating gene, NADPH oxidase 5 (Nox5), is essential for molting and oviposition in a Hemipteran insect Nilaparvata lugens. Knockdown of Nox5 transcript levels by RNA interference in 2nd–5th-instar nymphs results in significantly lethal deficits in the molting transitions from nymph–nymph and nymph–adult. Nox5 knockdown leads to a reduction of hydrogen peroxide in female ovaries and failure of oviposition from the insect ovipositor into the rice leaf sheath. Here, we provide in vivo evidence demonstrating that Nox5 is a key enzyme for regulating molting and oviposition in this insect species.


Introduction
The brown planthopper Nilaparvata lugens (Hemiptera: Delphacidae), a plant phloem-sucking Hemipteran insect, has emerged as a major pest affecting rice crops in tropical Asia and southern

Insects
The N. lugens population was originally collected from a rice field in the Huajiachi Campus of Zhejiang University, Hangzhou, China. The insects were reared at 26 • C ± 0.5 • C with 50% ± 5% humidity on rice seedlings (Oryza sativa strain Xiushui 134) under a 16:8 h light/dark photoperiod, as previously described [10].

Development-Stage and Tissue Sample Collection
For developmental stage-specific expression analysis of Nox5 gene, the samples were separately collected from the laid eggs taken from rice leaf sheaths at 24-h intervals, the 1st-4th-instar nymphs at 12-h intervals, the 5th-instar male nymphs and male adults at 24-h intervals, and the 5th-instar female nymphs and female adults at 12-h intervals. For tissue-specific expression analysis, the brain, salivary gland, fat body, gut, ovary, testis, cuticle, wing pad, macropterous wing and brachypterous wing were dissected from the 5th-instar nymphs-male and female adults on ice under a Leica S8AP0 stereomicroscope (Leica Microsystems GmbH, Wetzlar, Germany). To analyze Nox5 gene expression in each part of the ovaries, the ovipositor, bursa copulatrix, spermatheca and oviduct were individually collected from female adults. The samples were quickly washed in diethylpyrocarbonate-treated NaCl/Pi solution (pH 7.4), as previously described [9].

Developmental Stage-and Tissue-Expression Patterns of Nox5 by Quantitative Real-Time PCR Analysis
Total RNA was extracted from each developmental stage-and tissue-specific sample using RNAiso plus (TaKaRa, Dalian, China). The concentration and quality of the extracted RNA were measured with a NanoDrop 2000/2000 c spectrophotometer (Thermo Fisher Scientific, Bremen, Germany). First-strand cDNA was synthesized using a Hiscript ® II QRT SuperMix for qPCR (+gDNA wiper) Kit (Vazyme, Nanjing, China) to remove any contaminating genomic DNA. RNA with no-reverse-transcriptase was used as the no-template control. Quantitative real-time PCR was run on a CFX Connect TM Real-Time System (Bio-Rad, Hercules, CA, USA) using Hieff™ qPCR SYBR Green Master Mix (Yeasen, Shanghai, China) under the following reaction program: an initial denaturation step at 95 • C for 30 s, followed by 35 cycles at 95 • C for 5 s and 55 • C for 30 s. The pair of Nox5 gene-specific primers (sense 5 -AGCATGCCGTACTGATAGCC-3 and anti-sense 5 -TGCCTGCTCGATTTCCAGTT-3 ) were designed using the Primer Premier 6.0 program based on the N. lugens transcriptomic sequences. As the internal controls, expression of the N. lugens housekeeping gene for 18S ribosomal RNA, RPS11 and RPS15 (GenBank accession no. JN662398, XP_022188350 and ACN79501) was analyzed. The results were normalized to the expression level of the internal genes. The ∆∆C t method was used to evaluate the quantitative variation in the transcript levels, as described previously [20].

RNA Interference (RNAi)
The double-stranded RNA of Nox5 was synthesized by in vitro transcription with the PCR-generated DNA template, which was amplified from the Nox5 sequence using the MEGAscript T7 Transcription Kit (Vazyme, Nanjing, China). A pair of primers containing the T7 promoter were designed as follows: 5 -TAATACGACTCACTATAGGGAGATGCAAGTACCGCTGATGTGT-3 (sense primer) and 5 -TAATACGACTCACTATAGGGAGACACTCCAATGCCTGTGGCTA-3 (anti-sense primer). The amplified sequence, around 500 bp long, was cloned into the pMD19-T vector (TaKaRa). Aequorea victoria green fluorescent protein (GFP) was used as a negative control. Following transcription, the DNA templates were removed with TURBO DNase (Ambion). The sizes of double-stranded RNA (dsRNA) products were confirmed by electrophoresis on a 1% agarose gel with TAE buffer. The 2nd-, 3rd-, 4th-and 5th-instar nymphs were anesthetized with carbon dioxide for 15-20 s and each instar nymph was microinjected with approximately 250 ng of dsRNA using the FemtoJet Microinjection System (Eppendorf-Netheler-Hinz, Hamburg, Germany). The treated insects were reared on fresh rice Insects 2020, 11, 642 4 of 13 seedlings at 26 • C ± 0.5 • C with 50% ± 5% humidity under a 16:8 h light/dark photoperiod. RNAi effects were determined three days after dsRNA-injection, using quantitative real-time PCR. The transcript level changes of the Nox5 gene were determined by quantitative real-time PCR analysis.

Observation of Oviposition and Egg Hatching
Oviposition and egg hatching experiments were carried out according to Xu et al. [21] and Shen et al. [22]. Briefly, the 5th-instar nymphs were microinjected with dsNox5 and dsGFP, respectively, and were reared on fresh rice seedlings until emergence. To ensure successful mating, the mating was performed between the newly emerged female and male adults as follows: a single dsNox5-treated female adult was mated with two dsNox5-treated male adults; similarly, a single dsNox5-treated female adult was mated with two dsGFP-treated male adults; and a single dsGFP-treated female adult was mated with two dsNox5-treated or two dsGFP-treated male adults. The mating between female and male adults was performed in a long glass tube containing three-leaf stage fresh rice seedlings (6.5 ± 0.5 cm long) at 26 • C ± 0.5 • C with 50% ± 5% relative humidity under a 16:8 h light/dark photoperiod for 3 days, then the female and male adults were moved. The hatched nymphs in the rice seedlings were observed and counted at 24-h intervals, then were moved out of the rice seedlings 10 days later. Unhatched eggs in rice leaf sheaths were dissected and counted. Biological replicates were carried out for each mating (n = 15-20 ♀ × Figure 5. RNAi of Nox5 effects on egg-laying and hatching. (A) Calculation of egg-laying rates. The 5th-instar nymphs were injected with dsNox5 and the newly emerged female adults were mated with male adults. The number of eggs laid in rice leaf sheaths from each mating pair was calculated. The x-axis indicates each mating pair, namely, dsGFP-treated females and males; dsGFP-treated females and dsNox5-treated males; dsNox5-treated females and dsGFP-treated males; dsNox5-treated females and males. The y-axis shows the number of eggs laid from each mating pair. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (B) Determination of hatching rates. Hatching rates were calculated from the eggs laid in rice leaf sheaths from each mating pair as described above. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (C) Observation of the eggs laid in the rice leaf sheath. The eggs were dissected from the rice leaf sheath at the 3rd day after oviposition of dsNox5-and dsGFP-injected females that mated with male adults. The arrows indicate the egg-laying sites in the rice leaf sheath.
). The egg-laying sites in rice leaf sheaths were observed under a microscope and oviposition behavior was recorded by photography.

Measurement of H 2 O 2 Production
The production of H 2 O 2 was measured spectrophotometrically with a Varioskan ® Flash microplate reader (Thermo Fisher Scientific, Vantaa, Finland) using the Hydrogen Peroxide Assay Kit (Beyotime Biotech, Shanghai, China). In this assay, H 2 O 2 converts Fe 2+ to Fe 3+ , which then complexes with xylenol orange dye to yield a purple product having an absorbance maximum at 560 nm, which could be detected by a spectrometer [23]. Ten micrograms of ovaries of female adults were homogenized in 200 µL lysis buffer using a hand-drive tissue homogenizer on ice and centrifuged at 12,000× g at 4 • C for 5 min. The protein concentrations of the supernatant were quantified using a Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA). Aliquots of 50 µL of supernatants and 100 µL of test solutions from the Hydrogen Peroxide Assay Kit were added into each well of a 96-well plate and gently mixed to incubate at room temperature for 30 min and measured immediately using a microplate reader at a wavelength of 560 nm (OD 560 ). The concentration of H 2 O 2 was calculated according to the H 2 O 2 standard curve. Three independent biological replicates were performed.

Analysis of Evolutionary Relationships
The predicted amino acid sequence of N. lugens Nox5 was aligned with the orthologs of other insect species from the NCBI database using the CLUSTAL W program. The phylogenic tree was constructed using the full-length amino acid sequences of Nox5 proteins by the Maximum Likelihood program Mega X (http://www.megasoftware.net/). Phylogenetic relationships were determined using bootstrap analysis with 1000 replications.

Statistical Analyses
Data were analyzed by Student's t-test or one-way ANOVA. All statistical analysis and data plots were performed using Prism software. In the figures, statistically significant differences are presented as ** and indicate p < 0.01.

Availability of Supporting Data
The Nox5 sequence was searched against the N. lugens genome database deposited in GenBank under accession number AOSB00000000 (BioProject PRJNA177647) and the N. lugens transcriptome Insects 2020, 11, 642 5 of 13 in the Sequence Read Archive (SRA) database (http://www.ncbi.nlm.nih.gov/sra, SRX023419) using the tBLASTX algorithm with a cut-off E-value of 10 −10 . The conserved domains were determined using the Simple Modular Architecture Research Tool (SMART) (http://smart.embl.de/) and the Conserved Domains database of the National Center for Biotechnology Information (NCBI) website (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi). The transcript sequence of N. lugens Nox5 gene has been submitted to NCBI (http://www.ncbi.nlm.nih.gov/). The accession number is MT629918.

Identification of an NADPH Oxidase 5 Gene in N. lugens
By searching the N. lugens genomic and transcriptomic databases, we obtained the complete cDNA sequence of an NADPH oxidase 5 (Nox5). The full-length Nox5 sequence is 4401 base pair (bp) long and contains an open reading frame of 3633 bp, a 445-bp-long 5 untranslated region and a 323-bp-long 3 untranslated region (GenBank accession no. MT629918). The deduced protein consists of 1210 amino acid residues ( Figure 1A). Amino acid sequence alignment showed that Nox5 proteins of the different insects contain the characteristic calcium-binding EF-hand motifs at the N-terminus, two transmembrane ferric reductase domains and the binding domains for flavin adenine dinucleotide (FAD) and nicotinamide adenine dinucleotide (NAD) at the C-terminus ( Figure 1A). The Nox5 proteins of Hemipteran N. lugens, Hymenopteran Apis mellifera and Coleopteran Tribolium castaneum have five classical EF-hand motifs at their N-terminus, whereas Hemipteran Acyrthosiphon pisum, Dipteran D. melanogaster and Lepidopteran B. mori Nox5 proteins only have four EF-hand motifs ( Figure 1A). The phylogenetic analysis revealed that N. lugens Nox5 cluster with the orthologs of Hemipteran insect species, which are closely related to the Coleoptera cluster, but are distant from the Hymenopteran, Lepidopteran and Dipteran clusters ( Figure 1B).
Insects 2020, 11, x FOR PEER REVIEW 5 of 13 The Nox5 sequence was searched against the N. lugens genome database deposited in GenBank under accession number AOSB00000000 (BioProject PRJNA177647) and the N. lugens transcriptome in the Sequence Read Archive (SRA) database (http://www.ncbi.nlm.nih.gov/sra, SRX023419) using the tBLASTX algorithm with a cut-off E-value of 10 −10 . The conserved domains were determined using the Simple Modular Architecture Research Tool (SMART) (http://smart.embl.de/) and the Conserved Domains database of the National Center for Biotechnology Information (NCBI) website (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi). The transcript sequence of N. lugens Nox5 gene has been submitted to NCBI (http://www.ncbi.nlm.nih.gov/). The accession number is MT629918.

Identification of an NADPH Oxidase 5 Gene in N. Lugens
By searching the N. lugens genomic and transcriptomic databases, we obtained the complete cDNA sequence of an NADPH oxidase 5 (Nox5). The full-length Nox5 sequence is 4401 base pair (bp) long and contains an open reading frame of 3633 bp, a 445-bp-long 5′ untranslated region and a 323bp-long 3′ untranslated region (GenBank accession no. MT629918). The deduced protein consists of 1210 amino acid residues ( Figure 1A). Amino acid sequence alignment showed that Nox5 proteins of the different insects contain the characteristic calcium-binding EF-hand motifs at the N-terminus, two transmembrane ferric reductase domains and the binding domains for flavin adenine dinucleotide (FAD) and nicotinamide adenine dinucleotide (NAD) at the C-terminus ( Figure 1A). The Nox5 proteins of Hemipteran N. lugens, Hymenopteran Apis mellifera and Coleopteran Tribolium castaneum have five classical EF-hand motifs at their N-terminus, whereas Hemipteran Acyrthosiphon pisum, Dipteran D. melanogaster and Lepidopteran B. mori Nox5 proteins only have four EF-hand motifs ( Figure 1A). The phylogenetic analysis revealed that N. lugens Nox5 cluster with the orthologs of Hemipteran insect species, which are closely related to the Coleoptera cluster, but are distant from the Hymenopteran, Lepidopteran and Dipteran clusters ( Figure 1B).

Temporospatial Expression Patterns of Nox5 Gene in N. lugens
To understand the functional roles of the Nox5 gene in N. lugens, we investigated its expression patterns through developmental stages and tissue distributions using quantitative real-time PCR. Nox5 transcripts were detected at a notably high level in freshly laid eggs at 0 h in rice leaf sheaths and sharply decreased to a low level in the laid eggs at 24 h, then slowly increased to attach a high level at 168 h, just before egg hatching ( Figure 2A). Nox5 displayed similar expression patterns in the 1st-5th-instar nymphal stages. Its transcripts were detected at very low levels in newly hatched nymphs at 0 h and gradually increased to the peak levels at a timepoint toward the end of each nymphal stage for nymph-nymph and nymph-adult transitions, namely at 36 h for the 1st-2nd-instar nymphs, 48 h for the 3rd-instar nymphs, 60 h for the 4th-instar nymphs and 72 h for the 5th-instar female and male nymphs. In adults, Nox5 transcripts were at very low levels in males at 0-72 h after emergence, whereas the transcript levels gradually increased from 0 to 72 h in females. The development expression pattern suggested that Nox5 has important functions in the molting transition between nymph-nymph, nymph-adult and oviposition process.   (n = 50); the ovary, testis, macropterous and brachypterous wings (n = 50) were extracted from the female and male adults, respectively. Br, brain; Sg, salivary gland, Gu, gut; Cu, cuticle; Fb, fat body, Wp, wing pad; Te, testis; Ov, ovary, Mw, macropterous wing; Bw, brachypterous wing. (C) Ovary-specific expression of Nox5 gene. Total RNA was extracted from the ovipositor, bursa copulatrix, spermatheca and oviduct of female adults (n = 100). M-ovipositor, the ovipositor of female adults mated with males; V-ovipositor, the ovipositor of virgin females. Quantitative real-time PCR was conducted to analyze the Nox5 gene expression change. The relative transcript levels of Nox5 gene in each development stage, each tissue and each part of the ovary were normalized using N. lugens RPS11, RPS15 and 18S rRNA Ct values as previously described [24]. The reactions were performed with specific primers for amplifying the Nox5 gene. Three biological replications (mean ± SD) were carried out based on the independent RNA sample and the ∆∆Ct method was used to measure the relative transcript levels in each developmental stage and each tissue. The results of triplicate experiments are shown with standard deviations. p values are indicated (one-way ANOVA by least significant difference-test). Asterisks (**) indicate significant difference (p < 0.01) in the laid eggs at 0 h from other developmental stages; significant difference (p < 0.01) in the ovaries of the female adults from other tissues of the 5th instar nymph, male and female adults; significant difference (p < 0.01) in M-ovipositor from V-ovipositor, bursa copulatrix, spermatheca and oviduct.
Tissue-specific analysis displayed that Nox5 transcripts were at a high level in the wing pad, but at low levels in the brain, salivary gland, gut, fat body and cuticle of the 5th-instar nymphs ( Figure 2B). In female adults, the Nox5 transcripts had a notably high level in the ovary and a low level in the fat body but were almost undetectable in the brain, salivary gland, gut, cuticle, macropterous and brachypterous wings. In male adults, Nox5 transcripts were almost undetectable or at extremely low levels in all tested tissues ( Figure 2B). Tissue-specificity implicated that Nox5 is of functional importance for female reproduction. A further investigation aiming at the female ovaries showed that Nox5 had the highest transcript level in the ovipositor of female adults that mated with males. High transcript levels were also detected in the bursa copulatrix followed by the spermatheca and oviduct of the females. Interestingly, a very low transcript level was observed in the ovipositor of virgin females that did not mate with male adults ( Figure 2C).

Effect of Nox5 Knockdown on Development of Nymphs
To understand whether Nox5 has functional roles during the development process in N. lugens, we conducted phenotypic analysis through the RNAi approach at the in vivo level. Knockdown of Nox5 transcript levels generated lethal defects throughout the nymph and adult stages. The dsNox5-treated nymphs showed a deficiency in the molting 3rd-4th-instar, 4th-5th-instar and 5th-instar-adult transitions ( Figure 3). By contrast, dsGFP-injected nymphs successfully completed the transitions from nymph-nymph and nymph-adult.
RNAi led to a significant reduction in the survival rates of the insects. The survival rates of dsNox5-treated 2nd-instar nymphs were 84% at 3 d.p.i. (days post injection) and gradually decreased to 49% at 15 d.p.i ( Figure 4A). The survival rates of dsNox5-treated 3rd-and 4th-instar nymphs were 88%-92% at 3 d.p.i. and decreased to 46%-49% at 15 d.p.i. The survival rates of dsNox5-treated 5th-instar nymphs were 84% at 3 d.p.i. and quickly declined to 33% at 15 d.p.i. By contrast, the survival rates of non-and dsGFP-treated 2nd-, 3rd-or 4th-instar nymphs were more than 85% and the 5th-instar nymphs were more than 75% at 15 d.p.i. Quantitative real-time PCR confirmed that the transcript levels of Nox5 in each developmental stage (2nd-5th nymphs) were notably reduced in RNAi-treated nymphs compared with non-and dsGFP-treated controls ( Figure 4B). RNAi led to a significant reduction in the survival rates of the insects. The survival rates of dsNox5-treated 2nd-instar nymphs were 84% at 3 d.p.i. (days post injection) and gradually decreased to 49% at 15 d.p.i ( Figure 4A). The survival rates of dsNox5-treated 3rd-and 4th-instar nymphs were 88%-92% at 3 d.p.i. and decreased to 46%-49% at 15 d.p.i. The survival rates of dsNox5-treated 5thinstar nymphs were 84% at 3 d.p.i. and quickly declined to 33% at 15 d.p.i. By contrast, the survival rates of non-and dsGFP-treated 2nd-, 3rd-or 4th-instar nymphs were more than 85% and the 5thinstar nymphs were more than 75% at 15 d.p.i. Quantitative real-time PCR confirmed that the transcript levels of Nox5 in each developmental stage (2nd-5th nymphs) were notably reduced in RNAi-treated nymphs compared with non-and dsGFP-treated controls ( Figure 4B).   RNAi led to a significant reduction in the survival rates of the insects. The survival rates of dsNox5-treated 2nd-instar nymphs were 84% at 3 d.p.i. (days post injection) and gradually decreased to 49% at 15 d.p.i ( Figure 4A). The survival rates of dsNox5-treated 3rd-and 4th-instar nymphs were 88%-92% at 3 d.p.i. and decreased to 46%-49% at 15 d.p.i. The survival rates of dsNox5-treated 5thinstar nymphs were 84% at 3 d.p.i. and quickly declined to 33% at 15 d.p.i. By contrast, the survival rates of non-and dsGFP-treated 2nd-, 3rd-or 4th-instar nymphs were more than 85% and the 5thinstar nymphs were more than 75% at 15 d.p.i. Quantitative real-time PCR confirmed that the transcript levels of Nox5 in each developmental stage (2nd-5th nymphs) were notably reduced in RNAi-treated nymphs compared with non-and dsGFP-treated controls ( Figure 4B).  The newly hatched 2nd-, 3rd-, 4th-or 5th-instar nymphs were individually injected with dsNox5. Non-and dsGFP-treated nymphs were used as controls, respectively. The survival rates were calculated from three biological replicates (mean ± SD). For each treatment, n = 100 nymphs. p < 0.01 was considered statistically significant (**), different from non-and dsGFP treatments. (B) The transcript level variations (in fold change) of Nox5 in 2nd-, 3rd-, 4th-and 5th-instar nymphs were determined by quantitative real-time PCR, as described in Figure 2. Naive and dsGFP refer to non-and dsGFP-treated controls, respectively. The results of triplicate experiments are shown with standard deviations. p values are indicated (Student's t-test). Asterisks (**) indicate significant difference (p < 0.01) from controls.

Effect of Nox5 Knockdown on Oviposition of Female Adults
As the Nox5 gene was highly expressed in the ovaries of female adults and the freshly laid eggs at 0 h in rice leaf sheaths, we investigated the function of Nox5 in oviposition of female adults using RNAi. No eggs were laid in rice leaf sheaths from dsNox5-injected females that mated with dsNox5-injected males or dsGFP-treated male controls ( Figure 5A). By contrast, about 60-80 eggs were laid in rice leaf sheaths from each dsGFP-treated female that mated with dsNox5-injected males or dsGFP-treated male controls and 80%-90% of the laid eggs were successfully hatched into the nymphs ( Figure 5B), suggesting that Nox5 plays a vital role for oviposition of female adults but is not necessary for male reproduction. Subsequently, we investigated the oviposition behavior of female adults. The dsNox5-injected females had oviposition behavior by inserting the ovipositor into rice leaf sheaths and leaving the egg-laying sites but no eggs in rice leaf sheaths ( Figure 5C). SD). For each treatment, n = 100 nymphs. p < 0.01 was considered statistically significant (**), different from non-and dsGFP treatments. (B) The transcript level variations (in fold change) of Nox5 in 2nd-, 3rd-, 4th-and 5th-instar nymphs were determined by quantitative real-time PCR, as described in Figure 2. Naive and dsGFP refer to non-and dsGFP-treated controls, respectively. The results of triplicate experiments are shown with standard deviations. p values are indicated (Student's t-test). Asterisks (**) indicate significant difference (p < 0.01) from controls.

Effect of Nox5 Knockdown on Oviposition of Female Adults
As the Nox5 gene was highly expressed in the ovaries of female adults and the freshly laid eggs at 0 h in rice leaf sheaths, we investigated the function of Nox5 in oviposition of female adults using RNAi. No eggs were laid in rice leaf sheaths from dsNox5-injected females that mated with dsNox5injected males or dsGFP-treated male controls ( Figure 5A). By contrast, about 60-80 eggs were laid in rice leaf sheaths from each dsGFP-treated female that mated with dsNox5-injected males or dsGFPtreated male controls and 80%-90% of the laid eggs were successfully hatched into the nymphs ( Figure 5B), suggesting that Nox5 plays a vital role for oviposition of female adults but is not necessary for male reproduction. Subsequently, we investigated the oviposition behavior of female adults. The dsNox5-injected females had oviposition behavior by inserting the ovipositor into rice leaf sheaths and leaving the egg-laying sites but no eggs in rice leaf sheaths ( Figure 5C). The 5th-instar nymphs were injected with dsNox5 and the newly emerged female adults were mated with male adults. The number of eggs laid in rice leaf sheaths from each mating pair was calculated. The x-axis indicates each mating pair, namely, dsGFP-treated females and males; dsGFP-treated females and dsNox5-treated males; dsNox5-treated females and dsGFP-treated males; dsNox5-treated females and males. The y-axis shows the number of eggs laid from each mating pair. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (B) Determination of hatching rates. Hatching rates were calculated from the eggs laid in rice leaf sheaths from each mating pair as described above. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (C) Observation of the eggs laid in the rice leaf sheath. The eggs were dissected from the rice leaf sheath at the 3rd day after oviposition of dsNox5-and dsGFP-injected females that mated with male adults. The arrows indicate the egg-laying sites in the rice leaf sheath. The 5th-instar nymphs were injected with dsNox5 and the newly emerged female adults were mated with male adults. The number of eggs laid in rice leaf sheaths from each mating pair was calculated. The x-axis indicates each mating pair, namely, dsGFP-treated females and males; dsGFP-treated females and dsNox5-treated males; dsNox5-treated females and dsGFP-treated males; dsNox5-treated females and males. The y-axis shows the number of eggs laid from each mating pair. Three biological replicates (mean ± SD) were carried out (n = 15-20 ♀ × SD). For each treatment, n = 100 nymphs. p < 0.01 was considered statistically significant (**), different from non-and dsGFP treatments. (B) The transcript level variations (in fold change) of Nox5 in 2nd-, 3rd-, 4th-and 5th-instar nymphs were determined by quantitative real-time PCR, as described in Figure 2. Naive and dsGFP refer to non-and dsGFP-treated controls, respectively. The results of triplicate experiments are shown with standard deviations. p values are indicated (Student's t-test). Asterisks (**) indicate significant difference (p < 0.01) from controls.

Effect of Nox5 Knockdown on Oviposition of Female Adults
As the Nox5 gene was highly expressed in the ovaries of female adults and the freshly laid eggs at 0 h in rice leaf sheaths, we investigated the function of Nox5 in oviposition of female adults using RNAi. No eggs were laid in rice leaf sheaths from dsNox5-injected females that mated with dsNox5injected males or dsGFP-treated male controls ( Figure 5A). By contrast, about 60-80 eggs were laid in rice leaf sheaths from each dsGFP-treated female that mated with dsNox5-injected males or dsGFPtreated male controls and 80%-90% of the laid eggs were successfully hatched into the nymphs ( Figure 5B), suggesting that Nox5 plays a vital role for oviposition of female adults but is not necessary for male reproduction. Subsequently, we investigated the oviposition behavior of female adults. The dsNox5-injected females had oviposition behavior by inserting the ovipositor into rice leaf sheaths and leaving the egg-laying sites but no eggs in rice leaf sheaths ( Figure 5C).

Figure 5.
RNAi of Nox5 effects on egg-laying and hatching. (A) Calculation of egg-laying rates. The 5th-instar nymphs were injected with dsNox5 and the newly emerged female adults were mated with male adults. The number of eggs laid in rice leaf sheaths from each mating pair was calculated. The x-axis indicates each mating pair, namely, dsGFP-treated females and males; dsGFP-treated females and dsNox5-treated males; dsNox5-treated females and dsGFP-treated males; dsNox5-treated females and males. The y-axis shows the number of eggs laid from each mating pair. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (B) Determination of hatching rates. Hatching rates were calculated from the eggs laid in rice leaf sheaths from each mating pair as described above. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (C) Observation of the eggs laid in the rice leaf sheath. The eggs were dissected from the rice leaf sheath at the 3rd day after oviposition of dsNox5-and dsGFP-injected females that mated with male adults. The arrows indicate the egg-laying sites in the rice leaf sheath.
). (B) Determination of hatching rates. Hatching rates were calculated from the eggs laid in rice leaf sheaths from each mating pair as described above. Three biological replicates (mean ± SD) were carried out (n = 15-20 ♀ × instar nymphs were individually injected with dsNox5. Non-and dsGFP-treated nymphs were used as controls, respectively. The survival rates were calculated from three biological replicates (mean ± SD). For each treatment, n = 100 nymphs. p < 0.01 was considered statistically significant (**), different from non-and dsGFP treatments. (B) The transcript level variations (in fold change) of Nox5 in 2nd-, 3rd-, 4th-and 5th-instar nymphs were determined by quantitative real-time PCR, as described in Figure 2. Naive and dsGFP refer to non-and dsGFP-treated controls, respectively. The results of triplicate experiments are shown with standard deviations. p values are indicated (Student's t-test). Asterisks (**) indicate significant difference (p < 0.01) from controls.

Effect of Nox5 Knockdown on Oviposition of Female Adults
As the Nox5 gene was highly expressed in the ovaries of female adults and the freshly laid eggs at 0 h in rice leaf sheaths, we investigated the function of Nox5 in oviposition of female adults using RNAi. No eggs were laid in rice leaf sheaths from dsNox5-injected females that mated with dsNox5injected males or dsGFP-treated male controls ( Figure 5A). By contrast, about 60-80 eggs were laid in rice leaf sheaths from each dsGFP-treated female that mated with dsNox5-injected males or dsGFPtreated male controls and 80%-90% of the laid eggs were successfully hatched into the nymphs ( Figure 5B), suggesting that Nox5 plays a vital role for oviposition of female adults but is not necessary for male reproduction. Subsequently, we investigated the oviposition behavior of female adults. The dsNox5-injected females had oviposition behavior by inserting the ovipositor into rice leaf sheaths and leaving the egg-laying sites but no eggs in rice leaf sheaths ( Figure 5C).

Figure 5.
RNAi of Nox5 effects on egg-laying and hatching. (A) Calculation of egg-laying rates. The 5th-instar nymphs were injected with dsNox5 and the newly emerged female adults were mated with male adults. The number of eggs laid in rice leaf sheaths from each mating pair was calculated. The x-axis indicates each mating pair, namely, dsGFP-treated females and males; dsGFP-treated females and dsNox5-treated males; dsNox5-treated females and dsGFP-treated males; dsNox5-treated females and males. The y-axis shows the number of eggs laid from each mating pair. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (B) Determination of hatching rates. Hatching rates were calculated from the eggs laid in rice leaf sheaths from each mating pair as described above. Three biological replicates (mean ± SD) were carried out (n=15-20 ♀×♂). (C) Observation of the eggs laid in the rice leaf sheath. The eggs were dissected from the rice leaf sheath at the 3rd day after oviposition of dsNox5-and dsGFP-injected females that mated with male adults. The arrows indicate the egg-laying sites in the rice leaf sheath.
). (C) Observation of the eggs laid in the rice leaf sheath. The eggs were dissected from the rice leaf sheath at the 3rd day after oviposition of dsNox5-and dsGFP-injected females that mated with male adults. The arrows indicate the egg-laying sites in the rice leaf sheath.
Furthermore, we observed that dsNox5-treated female adults displayed an obviously swollen abdomen and the stretched intersegmental membranes in the tergum at the 12th day after mating with males, whereas the female controls completed oviposition and showed flat abdomens at the 12th day after mating with males ( Figure 6A). We dissected the ovaries of dsNox5-injected female adults that mated with dsGFP-treated males to calculate the number of mature oocytes and to observe the morphological changes of oocytes. Reduction in the number and malformation of oocytes was not observed in the ovaries of dsNox5-injected females from the 4th to the 12th days after mating with males. Generally, N. lugens female oviposition begins at the 3rd-4th days after mating and egg-laying is completed around the 8th-12th days, at which time a few mature oocytes remain in the ovaries ( Figure 6B). However, in dsNox5-injected females, the ovaries were full of banana-shaped mature oocytes at the 4th-12th days after mating. Interestingly, we found a mature oocyte located at the lateral oviduct in a dsNox5-treated female individual, but this oocyte could not be laid in the rice leaf sheaths ( Figure 6B). Video indicated that the dsNox5-treated females tried their best to lay eggs from the ovipositor onto rice leaf sheaths but failed, while dsGFP-treated females successfully laid eggs on the rice leaf sheaths (Supplementary Video 1). To understand whether dsNox5 treatment influences ROS production, we measured H 2 O 2 content in the ovaries of females at the 8th day after mating. Nox5 knockdown led to a significant reduction of H 2 O 2 content by 87% and 89% in the ovaries of dsNox5-injected females compared with those of dsGFP-treated and non-treated female controls, respectively ( Figure 6C), suggesting that Nox5 is important for H 2 O 2 production.
Furthermore, we observed that dsNox5-treated female adults displayed an obviously swollen abdomen and the stretched intersegmental membranes in the tergum at the 12th day after mating with males, whereas the female controls completed oviposition and showed flat abdomens at the 12th day after mating with males ( Figure 6A). We dissected the ovaries of dsNox5-injected female adults that mated with dsGFP-treated males to calculate the number of mature oocytes and to observe the morphological changes of oocytes. Reduction in the number and malformation of oocytes was not observed in the ovaries of dsNox5-injected females from the 4th to the 12th days after mating with males. Generally, N. lugens female oviposition begins at the 3rd-4th days after mating and egg-laying is completed around the 8th-12th days, at which time a few mature oocytes remain in the ovaries ( Figure 6B). However, in dsNox5-injected females, the ovaries were full of banana-shaped mature oocytes at the 4th-12th days after mating. Interestingly, we found a mature oocyte located at the lateral oviduct in a dsNox5-treated female individual, but this oocyte could not be laid in the rice leaf sheaths ( Figure 6B). Video indicated that the dsNox5-treated females tried their best to lay eggs from the ovipositor onto rice leaf sheaths but failed, while dsGFP-treated females successfully laid eggs on the rice leaf sheaths (Supplementary Video 1). To understand whether dsNox5 treatment influences ROS production, we measured H2O2 content in the ovaries of females at the 8th day after mating. Nox5 knockdown led to a significant reduction of H2O2 content by 87% and 89% in the ovaries of dsNox5-injected females compared with those of dsGFP-treated and non-treated female controls, respectively ( Figure 6C), suggesting that Nox5 is important for H2O2 production. The 5th-instar nymphs were injected with dsNox5 and the female adults that mated with males at the 10th day were used to observe the body morphology. The left and right images show the back and lateral views of dsGFP-and dsNox5-injected females, respectively. (B) The ovaries were dissected from the dsNox5-treated female adults at the 2nd, 4th, 6th, 8th, 10th and 12th days after mating and were observed under a Leica S8AP0 stereomicroscope. The enlarged circles show the mature oocytes in the lateral oviduct of the female adults. Sp, spermatheca, Ooc, oocyte, Ov, oviduct, Bc, bursa copulatrix. (C) The H2O2 content in the ovaries of dsNox5-treated female adults at the 8th day after mating was determined by means of a Hydrogen Peroxide Assay Kit (Beyotime Biotech, Shanghai, China). Naive and dsGFP-treated insects were used as controls.

Discussion
In this study, we investigated the in vivo function of the NADPH oxidase gene, Nox5, in a Hemipteran model insect, N. lugens. The deduced amino acid sequence of the Nox5 gene contains the characteristic domains, including calcium-binding EF-hand motifs, transmembrane components and Figure 6. RNAi of Nox5 effects on oocyte formation in female ovaries. (A) Observation of the body change in female adults after RNAi. The 5th-instar nymphs were injected with dsNox5 and the female adults that mated with males at the 10th day were used to observe the body morphology. The left and right images show the back and lateral views of dsGFP-and dsNox5-injected females, respectively. (B) The ovaries were dissected from the dsNox5-treated female adults at the 2nd, 4th, 6th, 8th, 10th and 12th days after mating and were observed under a Leica S8AP0 stereomicroscope. The enlarged circles show the mature oocytes in the lateral oviduct of the female adults. Sp, spermatheca, Ooc, oocyte, Ov, oviduct, Bc, bursa copulatrix. (C) The H 2 O 2 content in the ovaries of dsNox5-treated female adults at the 8th day after mating was determined by means of a Hydrogen Peroxide Assay Kit (Beyotime Biotech, Shanghai, China). Naive and dsGFP-treated insects were used as controls.

Discussion
In this study, we investigated the in vivo function of the NADPH oxidase gene, Nox5, in a Hemipteran model insect, N. lugens. The deduced amino acid sequence of the Nox5 gene contains the characteristic domains, including calcium-binding EF-hand motifs, transmembrane components and FAD-and NAD-binding domains, which are highly conserved in the orthologs of many insect species and mammals, implying its primary functions in the production of superoxide and the regulation of intracellular calcium flux. Here, our findings revealed that Nox5 played essential roles in the development and oviposition of N. lugens. Knockdown of Nox5 transcript levels in each nymphal stage led to significantly lethal deficits. The nymphs treated with dsNox5 hardly completed the molting transition of 3rd-4th-instar, 4th-5th-instar and 5th-instar-adult, suggesting that Nox5 has important functions in molting and development mechanisms in this insect species.
Tissue specificity indicated that the Nox5 gene had the highest transcript level in the ovaries among all tested tissues of the 5th-instar nymphs and adults. A further investigation on the ovaries revealed that the Nox5 gene had a much higher transcript level in the ovipositor than in other parts, including the bursa copulatrix, spermatheca and oviduct, in the mated females. Moreover, the Nox5 gene showed a notably higher transcript level in the ovipositor of the mated females than the virgin females without mating, implying the functional importance of the Nox5 gene for the oviposition of the fertilizable oocytes in ovipositor. Knockdown of Nox5 by RNAi did not reduce the number of mature oocytes in female ovarioles, nor did it cause morphologic changes in mature oocytes, implicating that this enzyme may not be required for oocyte formation and maturation in the ovaries. In N. lugens, the female adults generally lay the mature oocytes at the 3rd-4th days after mating, and a few mature oocytes remain in the ovaries around the 8th-12th days after mating with male adults. However, dsNox5-treated females retained many mature oocytes in the ovaries at the 12th day after mating with either dsNox5-injected or dsGFP-treated male adults. Furthermore, we observed that the female adults showed abnormally expanded abdomens with accumulated oocytes in the ovaries. These females displayed apparent oviposition behavior by inserting the ovipositor into rice leaf sheaths. We found that knockdown of Nox5 did not affect the release of the mature oocytes from ovarioles to the lateral oviduct, but inhibited the laying of the oocytes from the ovipositor into rice leaf sheaths, which led to the failure of oviposition. By contrast, dsNox5-treated males that mated with dsGFP-treated females successfully laid the eggs in rice leaf sheaths, and around 80% of the laid eggs hatched into nymphs. These observations indicated that Nox5 plays a determinant role in female oviposition but is not essential for male reproduction.
The primary function of NADPH oxidase is to produce ROS by transferring an electron from NADPH to oxygen. ROS are a collection of highly reactive molecules containing molecular oxygen (O 2 ), superoxide anion (O 2 •− ), hydrogen peroxide (H 2 O 2 ) and hydroxyl radicals ( • OH) [25]. In insects, ROS are involved in immune defense reactions against a broad range of pathogens. Nevertheless, little is known about the role of ROS in insect oviposition. For oviparous insects, the fertilizable oocytes are released from the ovipositor into the external environment. During this process, ROS production may be necessary for successful oviposition and survival of the newly laid eggs. In the present study, to understand the reasons for oviposition failure, we investigated ROS generation in female ovaries by RNAi. Because of the diversity of ROS and their intrinsic instability, we mainly measured H 2 O 2 levels in ovaries. Knockdown of Nox5 significantly reduced H 2 O 2 content in female ovaries, suggesting that Nox5 is involved in the production of H 2 O 2 . The Nox5 gene had a notably high transcript level in the laid eggs at a particular time point (0 h) and rapidly decreased levels at 24 h after egg-laying in rice leaf sheaths, implicating a possibility that Nox5-generated ROS are required for oviposition of mature oocytes from the ovipositor to rice tissue and for the survival of freshly laid eggs to overcome new environmental challenges, i.e., defenses from plant hosts and pathogenic infection. Ritsick et al. reported that Drosophila Nox modulates intracellular calcium flux via H 2 O 2 [13]. As N. lugens Nox5 contains five calcium-binding EF-hand motifs, we hypothesize another possibility-that Nox5 regulates calcium flux in the ovaries of female adults. The detailed regulation mechanisms of the Nox5 gene will need to be further elucidated. The knowledge of in vivo function of the Nox gene in molting and metamorphosis mechanisms is very limited in insects. In this study, the observation of the lethal phenotypes in dsNox5-treated nymphs, combined with the expression patterns of Nox5 in nymphs with the lowest levels at 0 h and the highest levels at the end of each instar, imply that Nox-generated ROS are essential for molting transition and the survival of newly hatched nymphs or emerged adults when encountering the external environment. Further studies investigating the actions of Nox-associated ROS and calcium flux will be required to understand development regulation mechanisms in insects.