Brown planthoppers were reared on a susceptible rice (Oryza sativa) variety, Taichung Native 1 (TN1), cultivated at 27°C ± 0.5°C and 75% ± 5% relative humidity under a 14L: 10D (h) photoperiod, according to a previously described method (Li et al., 2018).

Total RNA was extracted with the TRIzol Total RNA Isolation Kit (Invitrogen, Carlsbad, CA, United States). The concentration and purity were measured with the NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, Rockford, IL, United States) and determined by agarose gel electrophoresis. cDNA was synthesized by using the ReverTra Ace qPCR RT Kit (Toyobo Co., Ltd., Osaka, Japan), following the manufacturer’s manual.

Based on the published N. lugens genomic and transcriptomic data (Xue et al., 2014; Wan et al., 2015), NlHR3 and NlFTZ-F1 homologies were identified and their sequences were confirmed by the reverse transcription polymerase chain reaction (RT-PCR) using primers, as shown in Supplementary Table S1. The PCR product was gel purified, ligated into the vector TOPO2.1 (Invitrogen, Carlsbad, CA), and transformed into Escherichia coli DH5α competent cells (Novagen, Darmstadt, Germany). A total of 10 recombinant plasmids from several independent subclones were fully sequenced on the Applied Biosystems 3730 automated sequencer (Foster City, CA) from both directions. The newly described transcript variants of NlHR3 and NlFTZ-F1 were submitted to GenBank.

ClustalW2 was used to perform a homologous sequence alignment of HR3 and FTZ-F1 proteins from Nilaparvata lugens, Drosophila melanogaster, Tribolium castaneum, Bombyx mori, Apis mellifera, Aedes aegypti, Pediculus humanus corporis, Blattella germanica, and Acyrthosiphon pisum (Larkin et al., 2007). The conserved domains were predicted by using the Simple Modular Architectural Research Tool (SMART; http://smart.embl-heidelberg.de/) and InterPro: protein sequence analysis and classification (http://www.ebi.ac.uk/interpro/).

For temporal expression analysis of NlHR3 and NlFTZ-F1, total RNA samples were prepared from eggs; first (N1), second (N2), and third (N3) instar nymphs; each day of the fourth and fifth instar nymphs (N4D1, N4D2, N4D3, N5D1, N5D2, N5D3, and N5D4); and newly emerged (New-A) and 2-day-old female adults (AD2). To analyze the tissue-specific expression patterns, the different tissues were obtained by dissecting the fourth and fifth instar nymphs. The first group comprised the head (He), thorax (Th), and abdomen (Ab). The second group comprised the integument (In), wingbud (Wi), midgut (Mg), leg (Lg), and fat body (Fb). Quantitative real-time PCR (qRT-PCR) was conducted to estimate expression levels of NlHR3 and NlFTZ-F1 of various samples using internal control genes (RSP15 and RSP11), according to the published methods (Wang et al., 2018). All the qRT-PCR primers are shown in Supplementary Table S1. There were three independent replications of each biological sample with three technical replicates. Data were analyzed by the 2^−ΔΔCT method (Livak and Schmittgen, 2001) and in accordance with the MIQE guidelines (Bustin et al., 2009).

The response of NlHR3 and NlFTZ-F1 to 20E application was determined by referring to previous methods (Gao et al., 2022). 20E was purchased from Sigma (Sigma-Aldrich, Shanghai, China). The fourth instar nymphs (0–12 h) were collected and treated with 0.2 μL 20E (0.1 μg/μL in acetone). After treatment with 20E, samples were taken at 6, 12, and 24 h after application to extract total RNA, and the expression of NlHR3 and NlFTZ-F1 was examined. Three replicates (10*3 nymphs, a total of 30 nymphs) were used to extract total RNA, and acetone was used as the control.

The dsRNA synthesis and RNAi bioassay method were applied, as previously described (Li et al., 2018). The dsDNA fragments were amplified by RT-PCR using specific primers (Supplementary Table S1) and used as templates to synthesize dsRNA using the MEGAscript T7 High-Yield Transcription Kit (Ambion, Austin, TX, United States). The quality and concentration of the dsRNA were detected and kept at −80°C until further use. The dsRNA targeting the gene-encoding green fluorescence protein (dsGFP) served as a negative control.

RNA interference (RNAi) was performed by the FemtoJet microinjector (Eppendorf), as previously reported (Wang et al., 2018). Approximately 200 ng and 400 ng dsRNA were microinjected into each individual of the third and fifth instar nymphs, respectively (Xu et al., 2015). A total of 250 nymphs (10 replicates) were used for each treatment, including three survival evaluation replicates, three phenotypic evaluation replicates, three qRT-PCR verification replicates, and one backup replicate. QRT-PCR verification was conducted 3 days after injection.

Data were analyzed using Student’s t-test (the difference between two samples), Tukey’s test (the difference among three or more samples), and analysis of variance (ANOVA) by data processing system software (Tang and Zhang, 2013).

The cDNA of putative HR3 and FTZ-F1 transcripts in N. lugens was cloned and submitted to GenBank (OP937347, OP937348, KU928171.1, and KU928171.1). We identified two transcript variants of HR3 (NlHR3a and NlHR3b) that encoded 520 and 462 amino acid residues and two transcript variants of FTZ-F1 (NlFTZ-F1a and NlFTZ-F1b) that encoded 609 and 633 amino acid residues. The HR3 and FTZ-F1 family proteins contain two conserved nuclear receptor functional domains, among which the DNA-binding domain (DBD) at the amino terminal and the ligand-binding domain (LBD) at the carboxy terminal are highly conserved (Supplementary Figures S1A, S2A). The phylogenetic tree analysis showed that NlHR3 has a close relationship with that of A. pisum, and NlFTZ-F1 has a close relationship with that of A. pisum and P. humanus (Supplementary Figures S1C, S2C).

Developmental expression analysis showed that the transcript levels of NlHR3 were the highest in the first instar period (Figure 1A), while NlFTZ-F1 had two expression peaks at the later stages of the fourth and fifth instar periods (Figure1C). The expression profiles of NlHR3 and NlFTZ-F1 in different tissues are similar and relatively higher in the thorax and appendage of the nymphs (Figures 1B,D). The spatiotemporal data are compatible with the common idea that NlHR3 and NlFTZ-F1 play a vital role in the 20E signaling pathway.

In order to test whether NlHR3 and NlFTZ-F1 are induced by 20E, the fourth instar nymphs were treated with 20E or acetone (control) for 6, 12, and 24 h. The expression of NlHR3 significantly increased by 7.4, 11.4, and 8.3 fold (p-value < 0.01) after 6, 12, and 24 h of 20E application, respectively, compared to the acetone control (Figure 2A), while the expression of NlFTZ-F1 significantly increased 10.7-, 8.9-, and 12.1-fold (p-value < 0.01) (Figure 2B).

To explore the regulatory network between HR3, FTZ-F1, and the hormone signaling pathway, dsRNA of NlHR3, NlFTZ-F1, NlCYP314A, and NlKr-h1 (Krüppel homolog 1) were injected into the third instar nymphs. Then, the transcript levels of NlHR3, NlFTZ-F1, NlCYP314A, and NlKr-h1 were measured 3 days later. In the nymphs injected with dsNlHR3, the transcript levels of NlCYP314A were not significantly altered, but NlFTZ-F1 was significantly decreased by 54.8% (p-value < 0.01) and NlKr-h1 was significantly increased by 739.7% (p-value <0.01) compared to the dsGFP control (Figure 3A).

In the nymphs injected with dsNlFTZ-F1, the transcript levels of NlCYP314A and NlHR3 were significantly decreased by 74.2% and 80.3% (p-value <0.01), respectively, and NlKr-h1 was significantly increased by 1771.3% (p-value <0.01) (Figure 3B). In the nymphs injected with dsNlCYP314A (an ecdysone synthesis key gene), the transcript levels of NlHR3 and NlFTZ-F1 were significantly decreased by 56.6% and 64.1% (p-value <0.01), respectively (Figure 3C), while in the nymphs injected with dsNlKr-h1 (a juvenile hormone response key gene), the transcript levels of NlHR3 and NlFTZ-F1 were not significantly different from that of the control (Figure 3D). Ecdysone receptor (EcR) is the starting point of an ecdysone cascade reaction and the upstream of HR3 and FTZ-F1. We analyzed the expression level of NlEcR, but there was no significant difference after the knockdown of NlHR3 and NlFTZ-F1 (Supplementary Figure S3).

In summary, the transcript levels of 20E biosynthetic and cascade regulated genes were decreased significantly by silencing NlFTZ-F1, while only the ecdysone cascade regulated genes were decreased significantly by silencing NlHR3. The transcript level of the JH response key gene was increased significantly by silencing NlHR3 and NlFTZ-F1. The transcript level of NlFTZ-F1 was decreased significantly by silencing NlHR3.

In order to investigate the role of NlHR3 and NlFTZ-F1 in nymph–nymph molting, dsNlHR3 and dsNlFTZ-F1 from the common region (marked with black lines in Supplementary Figure S1) were synthesized in vitro and injected into the third instar nymphs. Compared to the control, the transcript levels of NlHR3 and NlFTZ-F1 were significantly decreased by 69.6% and 57.8% (p-value <0.01), respectively, 48 h after injecting with dsNlHR3 (Figure 4D).

Upon injecting with dsRNA into the early third instar nymphs, the knockdown of NlHR3 resulted in a significantly lower survival rate (81.7%; p-value < 0.05), which started on the second day after dsRNA injection (Figure 4D). The development of these nymphs was inhibited on the third instar stage, and all individuals (n = 60 out of 60) were unable to molt normally into the fourth instar stage and eventually died on the sixth day after dsRNA injection (Figures 4A,D). The knockdown of NlFTZ-F1 resulted in a significantly lower survival rate (78.3%; p-value < 0.05) which started on the first day after dsRNA injection (Figure 4D). The insects injected with dsRNA, targeting NlFTZ-F1, exhibited lethal phenotypes at approximately 3 days after injection, and the development was inhibited on the third nymph–nymph molting transition; as a result, the insect bodies became slender and extended. The insects then failed to shed their old cuticle and died (Figure 4A). In contrast, the dsGFP-treated nymphs successfully completed all nymph–nymph molting transition, and the survival rate on the seventh day after injection was 93.3% (Figures 4A,D). Furthermore, NlFTZ-F1 silencing caused a lower survival rate than NlHR3 from the second day after injection. The survival rate decreased to 0% on the fifth day after downregulating NlHR3, while all individuals died on the sixth day for downregulating NlFTZ-F1 (Figure 4D).

In order to investigate the role of NlHR3 and NlFTZ-F1 in nymph–adult molting, dsRNAs were injected into the fifth instar nymphs. The transcript levels of NlHR3 and NlFTZ-F1 were significantly decreased from the controls. The knockdown of NlHR3 and NlFTZ-F1 resulted in a significantly higher mortality rate (42.7% and 53.3%, p-value < 0.05) on the fourth day after dsRNA injection, and these individuals were unable to molt normally into adults; the old cuticle of the notum was split open, and the insects then failed to shed their old cuticle and died (Figure 4A). The individuals that molt normally into female adults were dissected to observe the development of ovaries. The ovaries of female individuals injected with dsNlHR3 and dsNlFTZ-F1 were deficient of oocytes and abnormally formed. The oocytes of individuals injected with dsNlHR3 were smaller than those of the dsGFP, and the egg shell of oocytes in individuals injected with dsNlFTZ-F1 had no luster and elasticity (Figure 4B), while the ovaries of those injected with dsGFP were full of oocytes (banana-shaped), and the egg shell of oocytes had luster and elasticity (Figure 4B). No offspring was produced by female individuals injected with dsNlHR3 and dsNlFTZ-F1.