Bombyx neuropeptide G protein–coupled receptor A7 is the third cognate receptor for short neuropeptide F from silkworm

The short neuropeptide F (sNPF) neuropeptides, closely related to vertebrate neuropeptide Y (NPY), have been suggested to exert pleiotropic effects on many physiological processes in insects. In the silkworm (Bombyx mori) two orphan G protein–coupled receptors, Bombyx neuropeptide G protein–coupled receptor (BNGR) A10 and A11, have been identified as cognate receptors for sNPFs, but other sNPF receptors and their signaling mechanisms in B. mori remain unknown. Here, we cloned the full-length cDNA of the orphan receptor BNGR-A7 from the brain of B. mori larvae and identified it as a receptor for Bombyx sNPFs. Further characterization of signaling and internalization indicated that BNGR-A7, -A10, and -A11 are activated by direct interaction with synthetic Bombyx sNPF-1 and -3 peptides. This activation inhibited forskolin or adipokinetic hormone-induced adenylyl cyclase activity and intracellular Ca2+ mobilization via a Gi/o-dependent pathway. Upon activation by sNPFs, BNGR-A7, -A10, and -A11 evoked ERK1/2 phosphorylation and underwent internalization. On the basis of these findings, we designated the receptors BNGR-A7, -A10, and -A11 as Bommo-sNPFR-1, -2, and -3, respectively. Moreover, the results obtained with quantitative RT-PCR analysis revealed that the three Bombyx sNPF receptor subtypes exhibit differential spatial and temporal expression patterns, suggesting possible roles of sNPF signaling in the regulation of a wide range of biological processes. Our findings provide the first in-depth information on sNPF signaling for further elucidation of the roles of the Bombyx sNPF/sNPFR system in the regulation of physiological activities.

The vertebrate family of neuropeptide Y (NPY), 3 consisting of NPY, peptide YY (PYY), and pancreatic polypeptide (PP), is widely conserved in vertebrates and is responsible for the central regulation of pleiotropic physiological processes (1,2). The ortholog of NPY, neuropeptide F (NPF), with a characteristic C terminus ending with an amidated Phe residue instead of a Tyr residue, has been reported in most insects (3). By using an anti-NPF antibody, two shorter neuropeptide F (sNPF), sharing the C-terminal RF amide sequence, were first obtained from extracts of brains of the Colorado potato beetle, Leptinotarsa decemlineata (4). Subsequently, a cDNA encoding four sNPFs was identified in the fly, Drosophila melanogaster (5). Other insect sNPFs have been reported in the desert locust, Schistocerca gregaria, the African malaria mosquito, Anopheles gambiae, and the yellow fever mosquito, Aedes aegypti (6,7).
Considering that sNPFs are expressed in numerous small neurons in the brain (3,8,9), it is possible that sNPFs have a conserved role as a neurotransmitter or neuromodulator, likely functioning as regulators of feeding and reproduction. Overexpression of sNPF leads to increased food intake both in feeding larvae and adults, whereas knockdown of sNPF displays the opposite phenotype in D. melanogaster (10). In Locusta migratoria, the injection of sNPF stimulates ovarian development, suggesting a possible role in insect reproduction (6,11). Previous studies have suggested that sNPF signaling exhibits pleiotropic effects on osmotic stress, olfactory sensitivity, locomotor activity, and learning and memory (3,12,13).
Neuropeptides mediate their biological actions via interactions with specific receptors present on cell surface. The first insect sNPF receptor was identified in Drosophila. The Drosophila receptor termed NPFR76F (CG7395), a G proteincoupled receptor (GPCR), shows 62-66% similarity to and 32-34% identity with the vertebrate Y2 NPY receptors (14) and has been shown to be maximally activated by the four predicted sNPFs when expressed in both Xenopus oocytes and Chinese hamster ovary (CHO) cells in the presence of G␣ 16 (8,14,15). The direct interaction of this receptor with the four sNPFs has been confirmed by a radiolabeled sNPFbased binding assay (16). Subsequently, another two insect sNPFs were identified from the fire ant Solenopsis invicta and the mosquito A. gambiae (17)(18)(19). Several sNPF recep-tors have been identified in silico from the genomes of Apis mellifera, Tribolium castaneum, Nasonia vitripennis, and A. gambiae (17, 20 -22).
The cDNA encoding the sNPF precursor in B. mori was obtained, and the deduced precursor sequence contains three putative mature peptides (23). The presence of Bombyx sNPF-1, sNPF-2, and sNPF-3 in the brain and other larval ganglia was confirmed by direct MALDI-TOF mass spectrometric profiling and semiquantitative analyses (24). Three orphan receptor genes, Bombyx neuropeptide G protein-coupled receptor (BNGR)-A7, -A10 and -A11, have been identified in silico as putative sNPF receptors (23,25). Further investigation using Ca 2ϩ imaging analysis in human embryonic kidney cell line (HEK293) cells revealed that BNGR-A10 and -A11 are equally activated by Bombyx sNPF-1, -2, and -3 (23). However, BNGR-A7 remains to be deorphanized as a receptor for sNPFs by biochemical and pharmacological analysis. In addition, sNPF signaling is considered important in the regulation of a wide range of physiological functions. Bombyx sNPFs have been reported to stage-specifically inhibit juvenile hormone (JH) biosynthesis, suggesting its role in the regulation of molting and metamorphosis (23,26). Recent studies indicate that in B. mori larvae, sNPF-mediated signaling is involved in the regulation of locomotor activity associated with foraging behavior (24,27). The silkworm not only is an economically important insect, being a primary producer of silk, but also serves as a model of Lepidoptera for the investigation of the signaling and physiological functions of the sNPF/receptor system. Thus, identification of the cognate receptors for Bombyx sNPFs and further elucidation of the mechanism(s) involved in sNPF-mediated signaling will lead to a better understanding of its possible roles in the regulation of reproduction, osmotic stress, olfactory sensitivity, locomotor activity, and learning and memory in the silkworm (3,6,12,13).
In this study, we have reported on the cloning of the cDNA encoding a Bombyx neuropeptide GPCR A7 (BNGR-A7) sequence, which, as based on genomic data mining and phylogenetic analysis, is closely related to Drosophila receptor termed NPFR76F and the mammalian Y2 NPY receptor (23,25). Further functional expression and characterization in both mammalian and insect cells suggests that the orphan receptor BNGR-A7 is a specific receptor for Bombyx sNPFs. We have also investigated the spatial and temporal expression patterns of three Bombyx sNPF receptors. Our study provides the detailed information on the signaling and internalization of Bombyx sNPF receptors and perhaps aids in the interpretation of the signaling in the regulation of physiological processes.

Cloning and expression of BNGR-A7, -A10, and -A11 in HEK293 and sf21 cells
BNGR-A7, -A10, and -A11 have been identified in silico as sNPF-like receptors by using genomic data mining and phylogenetic analysis (23,25), and BNGR-A10 and -A11 have been confirmed as sNPF receptors by molecular and biochemical methods (23). In the present study, the full-length cDNA sequences encoding BNGR-A7 (GenBank TM accession no. AB330428), -A10 (GenBank TM accession no. AB330431), and -A11 (GenBank TM accession no. AB330432) were obtained by RT-PCR from brain tissue of B. mori larvae. As shown in Fig.  1A, an alignment was performed among BNGR-A7, -A10, and -A11 and the D. melanogaster sNPF receptor. The amino acid sequences of BNGR-A7, -A10, and -A11 are 35.4, 38.6, and 38.6%, respectively, identical with that of D. melanogaster sNPFR. To confirm the accurate expression and localization, BNGR-A7, -A10, and -A11 with an N-terminal FLAG tag or with EGFP fused to the C-terminal end were constructed and stably or transiently expressed in HEK293 and insect sf21 cells. Confocal microscopy revealed that BNGR-A10 and -A11 were expressed mainly and localized to the plasma membrane in the absence of the ligand in both HEK293 and sf21 cells (Fig. 1B). However, BNGR-A7 exhibited correct localization in the plasma membrane in sf21 cells but displayed serious intracellular accumulation in HEK293 cells (Fig. 1, B and C). Significant cell surface expression was further confirmed by ELISA (Fig. 1C). Moreover, upon stimulation by sNPF peptides, FLAG-BNGR-A11 and BNGR-A11-EGFP exhibited significant inhibition of forskolin-evoked CRE-luciferase activity with EC 50 values of 33.41 and 36.06 nM, respectively, comparable with the wild-type receptor (EC 50 57.09 nM) (Fig. 1D). These data suggest that N-terminally FLAGtagged or C-terminally EGFP-fused BNGR-A7, -A10 and -A11 exhibited normal expression and membrane translocation in HEK293 and sf21 cells.

Activation of BNGR-A7, -A10, and -A11 via G i/o -dependent pathways by sNPFs
A previous study demonstrates that BNGR-A10 and -A11 are activated to trigger intracellular Ca 2ϩ mobilization in response to stimulation of sNPF (23). However, detailed information on

Identification of Bombyx sNPF receptors
sNPF-mediated signaling remains largely unavailable. To elucidate the G protein coupling in the activation of BNGR-A7, -A10, and -A11, a combination of functional assays with different inhibitors was performed. As shown in Fig. 5, sNPF-induced Ca 2ϩ mobilization (Fig. 5, A and B) and inhibition of forskolin-stimulated CRE-luciferase activity ( Fig. 5C) through receptors BNGR-A7, -A10, and -A11 were completely blocked by pretreatment with pertussis toxin (PTX), a specific inhibitor of G i/o proteins. The direct measurement of cAMP accumulation using a cAMP ELISA confirmed that receptors BNGR-A7, -A10, and -A11 exhibited inhibitory effects on AKH-induced intracellular cAMP production in sf21 cells in the response to sNPFs (Fig. 5D). Collectively, our data further strengthen the role of pertussis toxin-sensitive G i/o proteins in the Bombyx sNPF receptor-mediated signaling pathways.

sNPF-mediated activation of ERK1/2 and receptor internalization
It is well known that activated GPCRs signal to the mitogenactivated protein kinase (MAPK) cascades via G␣ i -, G␣ s -, and G␣ q -dependent signaling pathways (28). We next investigated whether sNPFs can induce phosphorylation of ERK1/2 through receptors BNGR-A7, -A10, and -A11. The sf21 cells transfected with BNGR-A7, -A10, or -A11, respectively, were seeded in 24-well plates and starved for 4 h in serum-free medium before stimulation. After stimulation with the indicated concentrations of agonist, the cell lysates were assayed using a phospho-specific antibody that binds only to the phosphorylated (Thr 202 and Tyr 204 of ERK1 and Thr 185 and Tyr 187 of ERK2) forms of these kinases (29). As shown in Fig. 6A, treatment with different concentrations of sNPF-1 induced a dose-dependent activation of ERK1/2 in sf21 cells. An analysis of the time course indicated that treatment of cells with sNPF-1 elicited transient phosphorylation of ERK1/2 with maximal phosphorylation evident at 5 min (BNGR-A7 and -A10) or 10 min (BNGR-A11), which returned to nearly basal levels by 60 min (Fig. 6B). Moreover, diverse specific inhibitors were used to elucidate the signaling pathways involved in sNPFR-mediated ERK1/2 phosphorylation in HEK293 cells. We found that the G i/o inhibitor PTX, PKC inhibitor Go6983, and MEK inhibitor U0126, respectively, exhibited inhibitory effects on the sNPF-induced activation of ERK1/2 (Fig. 6, C and D). In addition, to determine the biological activity of sNPF receptors under physiological conditions, we dissected Malpighian tubules from fifth instar larvae for the detection of ERK1/2 phosphorylation. As shown in Fig. 6E, the stimulation of Malpighian tubules with sNPF-1 induced phosphorylation of ERK1/2, which could be significantly inhibited by PTX treatment. These results demonstrate that sNPF receptors induce the activation of ERK1/2 via the G i/o /PKC/MEK pathway in response to sNPF.

Spatial and temporal expression profiles of BNGR-A7, -A10, and -A11
To better understand the physiological role of Bombyx sNPF signaling, we examined the mRNA expression patterns of BNGR-A7, -A10, and -A11. We dissected different tissues from the fifth instar larvae, pupae, and moths, respectively, and determined the sNPF receptor expression using real-time RT-PCR. Reference genes including Bombyx actin-A3, GAPDH, Rp49, and RpL3 were used for normalization of the qRT-PCR results. As shown in Fig. 8, transcripts of BNGR-A7, -A10, and -A11 were detectable in the brain, Malpighian tubules, midgut, testis, and ovary, but a higher level expression of BNGR-A10 was detected throughout development compared with BNGR-A7 and -A11. We found relatively high-level expression of the three receptors in the moth brain (Fig. 8A), in agreement with previous observations (14,19,(31)(32)(33)(34). We also observed obvious expression of BNGR-A7 and -A10 in moth brains and Malpighian tubules of the larvae and pupae, whereas BNGR-A11 mRNA could be detected at a relatively high level in the testis, Malpighian tubules, midgut, brain, and ovary at different developmental stages. Collectively, our data suggest that Bombyx sNPF signaling is likely to function as a pleiotropic regulator.
So far, mammalian derived-cell lines HEK293 and CHO-K1 have been widely selected as model cellular systems for the molecular and functional characterization of insect GPCRs in many studies. Interestingly, receptors BNGR-A10 and -A11

Identification of Bombyx sNPF receptors
localized correctly in the plasma membrane, whereas BNGR-A7 displayed serious intracellular retention, although only a small portion of the expressed receptors could reach to the cell surface because signals in the intracellular cAMP formation and Ca 2ϩ mobilization were detectable when expressed in HEK293 cells. The human CXCR1 chemokine receptor (39) and the murine CB2 cannabinoid receptor (40) expressed in mammalian cells and the human D2S dopamine receptor (41) and human bradykinin B2 receptor (42) expressed in insect cells were found to accumulate intracellularly for the most part, probably because of the high level of overexpression and saturation of trafficking machinery of the host cell (43). Post-translational glycosylation was demonstrated to play a crucial role in surface trafficking for the human angiotensin II type 1a receptor (43), the human bradykinin receptor (44), and the FSH receptor (45). However, the mechanism underlying the intracellular retention of the BNGR-A7 receptor expressed in HEK293 cells remains to be investigated.
Insect sNPF receptors belong to the rhodopsin family of GPCRs and display a strong sequence resemblance to the vertebrate type 2 neuropeptide Y receptors (14). Coexpression of Drosophila NPFR76F with the promiscuous G␣ 16 in Xenopus oocytes revealed that the receptor is maximally activated by Drosophilas NPFs to produce inward currents due to the activation of an endogenous oocyte calcium-dependent chloride current (14). This activation could be blocked by pretreatment with pertussis toxin, indicating that Drosophila NPFR76F signals via the G i/o -dependent pathway in the Xenopus oocytes system (15). Drosophila sNPF-1 and sNPF-2 elicited a calcium response when the NPFR76F receptor was expressed in a CHO cell line (8). In mammalian cells expressing the sNPF receptor of the mosquito, A. gambiae, and the desert locust, S. gregaria, cognate sNPFs were found to potently inhibit forskolin-stimulated cAMP production, suggesting the coupling of G i/o to the activated sNPF receptor (19,32). These results derived from heterologous systems suggest that insect sNPF receptor couples preferentially to pertussis toxin-sensitive G i/o proteins, leading to the inhibition of adenylyl cyclase activity and the reduction of cAMP levels on activation, which resembles the structurally related vertebrate NPY Y2 subtype receptor. In addition, a recent study using a genetically encoded cAMP sensor demonstrates that sNPF induces a decrease in intracellular cAMP in the Drosophila third instar larval motor neuron (46). However, more recent studies have demonstrated, controversially, that sNPFs stimulate intracellular cAMP production via a G s -dependent pathway in both an in vivo model and in vitro BG2-c6 cells expressing Drosophila sNPF receptor (38,47). In the current study, we used both a mammalian cell system and an insect cell system to characterize three Bombyx sNPF receptors to clarify this controversial issue. Our data derived from a CRE-Luc assay and an ELISA-based cAMP determination showed that activation of sNPF receptors resulted in a significant reduction of forskolin-induced intracellular cAMP formation; this inhibitory effect was completely blocked by pretreatment with the G i inhibitor PTX. In addition, sNPF-induced Ca 2ϩ mobilization and ERK1/2 phosphorylation were also significantly suppressed by PTX. These data suggest that Bombyx sNPF receptors preferentially couple to pertussis toxin-sensi-tive G i/o proteins, leading to the inhibition of adenylate cyclase activity. However, the reasons for the discrepancy between our results and those of Hong et al. (47) and Chen et al. (38) are not clear but might be attributable to the different cell systems used in the original study.
Insect sNPFs were initially identified by cross-reactivity with an anti-NPF antibody and are found in every insect genome discovered thus far (3,48). However, information on their physiological function is still limited. sNPF neuropeptides have been proposed to act as neuromodulator(s) and neurohormone(s) in the regulation of feeding behavior, locomotion, reproduction, sleep, and learning and memory in insects (3,6,10,37,49). In B. mori, sNPF signaling has been suggested to be involved in the regulation of feeding behavior and JH biosynthesis and secretion (23,26,27). In the present study, in identifying potential target tissues of sNPFs, we used quantitative RT-PCR to examine the tissue and developmental expression profile of three Bommo-sNPF receptors. Overall, Bommo-sNPFR-2 showed the highest expression throughout development compared with Bommo-sNPFR-1 and -3. Additionally, three subtypes, Bommo-sNPFR-1, -2, and -3, exhibited differential spatial and temporal expression patterns. The expression of three receptors in the brain was detected at a relatively lower level in larvae and pupae but at the maximal levels in moths, consistent with previous observations that sNPF receptor mRNA mainly appears to be limited to the nervous system in other insects (14,19,(31)(32)(33)(34). This suggests the possible roles of sNPF signaling in the regulation of feeding and growth, locomotor activity, and learning and memory. Interestingly, it is noteworthy that high-level expression of Bommo-sNPFR-1, -2, and -3 was detected in Malpighian tubules as compared with other peripheral tissues such as the gut, fat body, and ovaries, suggesting a regulatory role of sNPF signaling in the control of osmotic homeostasis in diuresis. This finding is in agreement with studies in S. invicta, Glossina morsitans, and A. gambiae (18,19,48). However, receptor transcript presence may not exactly reflect protein level. In addition, because of the high level of expression of the sNPF receptors in the nervous system, the RT-PCR products potentially could be amplified from the neuronal contamination (34). Therefore, further efforts are required to examine the spatial and temporal distribution of sNPF receptor proteins for better understanding the physiological roles of sNPF signaling.
In conclusion, in this study, we have paired the Bombyx orphan receptor BNGR-A7 as a cognate receptor with the Bombyx neuropeptide sNPFs. Our results on the characterization of Bommo-sNPF-mediated signaling have demonstrated that BNGR-A7 and previously deorphanized BNGR-A10 and -A11 are activated by direct interaction with Bommo-sNPFs via a G i/o -dependent pathway, leading to the inhibition of forskolin-or AKH-mediated cAMP production, intracellular Ca 2ϩ mobilization, and ERK1/2 activation in a PTX-sensitive manner. Furthermore, quantitative RT-PCR analysis has indicated that three subtypes of Bommo-sNPF receptors show developmental stage-specific and tissue-specific expression patterns, suggesting that these three receptors may play different physiological roles in different tissues and stages. Our Bommo-sNPF receptor-based func-

Identification of Bombyx sNPF receptors
tional assay systems and molecular data provide a foundation for further characterization and validation of this signaling system in vitro and in vivo.

Molecular cloning and plasmid construction
Total RNA was isolated from the brain of B. mori larvae using the RNAiso Plus reagent (Takara, Tokyo, Japan) following the manufacturer's instructions. cDNA was synthesized using a PrimeScript first strand cDNA synthesis kit (Takara) according to the manufacturer's instructions. The entire coding region of the BNGR-A7, -A10, and -A11 genes were cloned and sequenced. The primers used for the PCR cloning of BNGR-A7, -A10, and -A11 see supplemental Table 1. The PCR products obtained were directly cloned into the pcDNA-3.1 vector, the pCMV-FLAG vector, and the pEGFP-N1 vector for expression in mammalian cells. For expression in insect cells, the immediate-early gene promoter (IE1) and homologous region 3 (Hr3) of B. mori nucleopolyhedrovirus (BmNPV) and the promoter of BmActin A3 were used to replace the corresponding sites of pCMV-FLAG and pEGFP-N1 as described previously (50). pBmCRE-Luc was also reconstructed using the promoter of BmHsp20.4 (B. mori heat shock protein 20.4, GenBank TM accession no. EU350577) and BmFibL (B. mori fibroin light chain, Gen-Bank TM accession no. NM_001044023) polyadenylation signal to replace pVIP (promoter of vasointestinal peptide) and the SV40polyadenylationsignalofpCRE-Luc,respectively.Allconstructs were verified by sequencing.

Cell culture and transfection
The insect Spodoptera frugiperda ovarian cell line sf21 cells were maintained in insect cell culture medium TC100 from Applichem (Darmstadt, Germany) supplemented with 10% FBS (HyClone) at 28°C and seeded onto a 6-well tissue culture plate 2 h prior to transfection. The BNGR-A7, -A10, and -A11 cDNA plasmid constructs were transfected into sf21 cells using an X-tremeGENE HP protocol (Roche) according to the manufacturer's instruction. HEK293 cells were maintained in DMEM (Hyclone) supplemented with 10% heat-inactivated FBS (Hyclone) and were incubated at 37°C in a humidified atmosphere with 5% CO 2 , 95% air. The BNGR-A7, -A10, and -A11 cDNA plasmid constructs were transfected into HEK293 cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Selection for cell-stable expression was initiated by the addition of G418 (800 mg/ml, Invitrogen) 1-2 days after transfection.

Luciferase activity assay
A reporter gene assay was performed to investigate changes in the intracellular levels of cAMP. In the present study, the pCRE-Luc or pBmCRE-Luc reporter gene systems, consisting of the firefly luciferase coding region under the control of a minimal promoter containing CREs, were used to measure the intracellular cAMP levels. HEK293 and sf21 cells co-expressing BNGR-A7, -A10, or -A11, respectively, with the reporter protein were seeded into 96-well plates and incubated overnight. Cells were then stimulated with different concentrations of sNPF in serum-free medium and incubated for 4 h. Luciferase activity was detected using a firefly luciferase assay kit (Kenreal, Shanghai, China). When required, cells were treated with PTX (50 ng ml Ϫ1 ) for 16 h prior to the start of the experiment. The cAMP concentration was assessed using a commercially available cAMP detection kit (R&D Systems, Minneapolis, MN).

ERK1/2 activation assay
sf21 cells expressing BNGR-A7, -A10, or -A11, respectively, were seeded in 24-well plates and starved for 4 h in serum-free medium to reduce background ERK1/2 activation and eliminate the effects of the change of medium. If required, the cells were pretreated with PTX for 10 -16 h before activation. After stimulation with the indicated agonist, the cells were lysed in 80 l of lysis buffer (20 mM HEPES (pH 7.5), 10 mM EDTA, 150 mM NaCl, and 1% Triton X-100) with protease inhibitors (Roche) at 4°C for 30 min. Equal amounts of total cell lysate were sizefractionated by SDS-PAGE (10%) and transferred to a PVDF membrane (Millipore, catalog no. SCHVU01RE). The membranes were blocked in TBS containing 0.05% Tween 20 and 5% nonfat dry milk for 1 h at room temperature and then incubated with rabbit monoclonal anti-pERK1/2 antibody (Cell Signaling

Identification of Bombyx sNPF receptors
Technology) and anti-rabbit horseradish peroxidase-conjugated secondary antibody (Beyotime) according to the manufacturers' protocols. The blots were stripped and relabeled using an anti-␤-actin (1:2000) monoclonal antibody as a control for protein loading. The levels of ERK1/2 phosphorylation were normalized to ␤-actin. All immunoblots were quantified using a Bio-Rad Quantity One imaging system.
Malpighian tubules from fifth instar larvae were dissected and placed in cold HBM buffer. Following dissection, the HBM buffer was replaced with fresh TC100 medium (with or without the G i inhibitor PTX) for a 6-h preincubation with or without the G i inhibitor PTX. Then the Malpighian tubules were rapidly transferred to fresh medium containing 100 nM sNPF-1 and incubated for 1 h. All treatment samples (n Ͼ 5) were collected and homogenized in 300 ml of radioimmune precipitation assay lysis buffer and centrifuged. Supernatant of the lysate was diluted and boiled in an equal volume of SDS buffer for 10 min. Lysates were loaded onto 12% SDS gels according to the total protein content determined by a BCA protein assay kit, and then phosphor-ERK and ␤-actin levels were detected.

Internalization assay
To detect the expression of receptors on the cell membrane, BNGR-A7, -A10, or -A11/EGFP cells were seeded in glass-bottom 6-well plates. Cells were washed with PBS and fixed with 3% paraformaldehyde in PBS for 10 min at room temperature. The cells were then incubated with DAPI (Beyotime) for 10 min to stain the cell nuclei. For the internalization assay, the cells were treated with sNPF1 at 28°C for 60 min. After fixation with 3% paraformaldehyde, the cells were mounted in 50% glycerol and visualized by fluorescence microscopy using a Zeiss LSM510 laser-scanning confocal microscope attached to a Zeiss Axiovert 200 microscope with a Zeiss Plan-Apo 63ϫ 1.40 NA oil immersion lens. Images were collected using an LSM 5 Image Examiner and processed with Adobe Photoshop.

Binding assay
The binding assay for HEK293 cells that had been transiently transfected with BNGR-A7, -A10, or -A11 was performed as described previously (51). In brief, the cells were detached using non-enzyme cell dissociation buffer and washed twice with 1 ml of ice-cold Hanks' balanced salt solution. Binding experiments were performed in a final volume of 200 l of binding buffer containing 5 ϫ 10 5 cells and incubated with 1 M FITC-labeled sNPF3 in the absence or presence of different concentrations of unlabeled sNPF3 at room temperature for 60 min. Cells were then washed three times with 500 l of ice-cold standard solution with 0.1% BSA, and fluorescence intensity was determined using a flow cytometer (Beckman Coulter). The binding displacement curves were analyzed with GraphPad Prism.

Quantitative real-time PCR
Quantitative RT-PCR was performed as described previously with slight modifications (48). Tissues from the fifth instar larvae, pupae, and moths were dissected under a binocular microscope, collected in tubes containing MagNaLyser Green Beads (Roche), and homogenized with the MagNA Lyser (Roche). Total RNA was then extracted using the RNeasy lipid tissue mini kit (Qiagen, Duesseldorf, Germany) in combination with a DNase digestion (RNase-free DNase set, Qiagen). The total RNA (0.5 g) was converted to cDNA using a PrimeScript first strand cDNA synthesis kit (TaKaRa). The cDNA from the samples were quantified on a real-time PCR machine (CFX-Touch, Bio-Rad) using SYBR Premix ExTaq (TaKaRa). The possibility of genomic DNA contamination was excluded by DNase treatment. All samples were measured in three technical replicates and contained a no-template control for all primer pairs to check for the presence of unwanted genomic DNA. Reference genes, including Bombyx actin-A3, GAPDH, Rp49, and RpL3, were determined with geNorm software (52) and used for normalization of the qRT-PCR results (see supplemental Table 1 for primer sequences). Results were expressed using the comparative cycle threshold (⌬⌬Ct) method as described previously (53). Briefly, data were normalized by subtracting the Ct value of the geometric average of the reference genes from that of the target gene. ⌬⌬Ct was calculated as the difference in the normalized Ct value (⌬Ct) of the different tissue samples. The comparative expression level of target genes is equal to 2 Ϫ⌬⌬Ct .