Cloning and characterization of the pheromone biosynthesis activating neuropeptide receptor from the silkmoth, Bombyx mori. Significance of the carboxyl terminus in receptor internalization.

In most Lepidoptera, pheromone biosynthesis is regulated by a neuropeptide termed pheromone biosynthesis activating neuropeptide (PBAN). Although much is known about the cellular targets of PBAN, identification and functional characterization of the PBAN receptor (PBANR) has proven to be elusive. Given the sequence similarity between the active C-terminal regions of PBAN and neuromedin U, it was hypothesized that their respective receptors might also be similar in structure (Park, Y., Kim, Y. J., and Adams, M. E. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 11423-11428). Consequently, utilizing primers constructed from the conserved regions of insect neuromedin U receptor homologues, a full-length 2780-nucleotide clone encoding a 46-kDa G protein-coupled receptor was amplified from a Bombyx mori pheromone gland cDNA library. Tissue distribution analyses revealed that the receptor transcript is specific to the pheromone gland where it undergoes significant up-regulation in the day preceding eclosion. When transiently expressed in Sf9 cells, the B. mori PBANR responds to PBAN by mobilizing extracellular calcium in a dose-dependent manner. Confocal microscopic studies demonstrated the specificity of enhanced green fluorescent protein-tagged B. mori PBANR for PBAN and showed that PBAN induces internalization of the PBANR.PBAN complex. The rapid onset of internalization is mediated by a 67-amino acid C-terminal extension absent in the cloned Helicoverpa zea PBANR, which suggests that receptor internalization in that species likely utilizes a different mechanism. From these results, we have concluded that the cloned receptor gene encodes the B. mori PBANR and that it is both structurally and functionally distinct from the H. zea PBANR.

For many lepidopteran insects (i.e. moths and butterflies), the successful propagation of individual species is dependent on the female's ability to attract conspecific males; an event mediated by species-specific blends of volatile compounds known as sex pheromones (1). These sex pheromones are produced and released by a specialized gland, appropriately known as the pheromone gland (PG), 1 located between the eighth and ninth abdominal segments of the female. A major class of pheromone components produced by female moths is the C 10 -C 18 unsaturated, acyclic, aliphatic compounds that contain an oxygenated functional group such as aldehyde, alcohol, or acetate ester. They are synthesized de novo from acetyl-CoA through fatty acid synthesis, desaturation, and limited chainshortening reactions followed by reductive modification of the acyl group (2,3).
In most Lepidoptera, pheromone biosynthesis is regulated by pheromone biosynthesis activating neuropeptide (PBAN), a 33-amino acid neuropeptide amidated at its C terminus that originates from the subesophageal ganglion (4). Initially isolated in 1989 from Helicoverpa zea (5) and Bombyx mori (6), it has since been identified in a variety of species (7). Structure-function studies have determined that the C-terminal FSPRLamide is the minimal sequence necessary for pheromonotropic activity (7,8). Although the PBAN signal transduction cascade is initiated by an influx of extracellular calcium (7), both the site of PBAN regulation and the downstream effector molecules utilized appear to be species-dependent. In H. zea, an increase in intracellular cAMP levels is necessary (9) for PBAN regulation of the one or more enzymatic functions that occur at a step in or prior to fatty acid synthesis (10). In B. mori, however, a similar rise in the levels of the second messenger is not required (11) for PBANinduced reduction of the fatty-acyl group (12), i.e. the final step in the B. mori pheromone biosynthetic pathway. These findings could be an indication that the PBAN receptors (PBANRs) of these two insect species are functionally different.
PBAN actions had long been proposed to be mediated by the stimulation of a specific G protein-coupled receptor (GPCR) located on the surface of pheromone-producing cells (7). Although molecular elucidation of the PBANR lagged significantly behind identification of other proteins involved in the pheromone biosynthetic pathways, advances in insect genome projects, most notably the sequencing of the Drosophila and Anopheles genomes (13,14), made it a more tenable option. A 346-amino acid GPCR recently identified and characterized as the H. zea PBANR was homology-cloned based on sequence similarity with the receptor for the mammalian neuropeptide, neuromedin U (NmU) (15). Based on the sequence similarities between the C-terminal pentapeptide motifs necessary for biological activity in NmU, FRPRNamide (16), and PBAN, FSPRLamide, it had been postulated that the respective receptors may likewise exhibit regions of similarity. In support of this hypothesis, Hewes and Taghert (17) had earlier reported the presence of multiple GPCRs in the Drosophila genome homologous with the NmU/arginine vasopressin family of receptors specific for peptides containing the C-terminal PRXamide motif. Park et al. (18) further elaborated on those findings by reporting that Drosophila GPCR CG8795 (GenBank TM accession no. AF522190), putatively identified as a PRXamide receptor, interacts with a number of C-terminal FXPRLamide peptides, including H. zea PBAN.
In addition to promoting the coupling, and eventual activation, of the heterotrimeric G proteins necessary for promulgating the appropriate signal transduction cascade, ligand-receptor interactions also initiate the molecular events that lead to receptor desensitization. The current paradigm of receptor desensitization is a multistep, multicomponent process that includes: 1) uncoupling of the heterotrimeric G proteins in response to the phosphorylation of specific residues in the GPCR C terminus and/or intracellular loops, 2) endocytosis of the GPCR, and 3) recycling of the GPCR back to the cell surface following dephosphorylation, or targeting of the GPCR⅐ligand complex for degradation followed by down-regulation of the total cellular complement of the GPCR as result of decreased protein synthesis (19). An intriguing aspect of the desensitization process is the myriad of functions attributed to the GPCR C terminus, from harboring putative phosphorylation sites (19), to containing both positive and negative regulators of the endocytotic machinery (19), to serving as an anchor point upon which numerous protein-protein interactions can be built (20,21). Consequently, the functionality of the GPCR C-terminal tail has been greatly expanded in recent years to include a role in receptor trafficking (22)(23)(24), intracellular signaling (25,26), and receptor dimerization (27).
Given the relatively recent identification of the PBANR, nothing is known concerning the residues that comprise the PBAN binding pocket, the structural motifs utilized in heterotrimeric G protein coupling, or the functionality, if any, of the PBANR C terminus. In this report, we describe the cloning of a GPCR gene from a B. mori PG cDNA library that we have concluded, based on tissue and stage specificity, in conjunction with expression studies that demonstrate dose-dependent calcium mobilization, specific PBAN binding, and PBAN-induced cellular dynamics, encodes the B. mori PBANR. Although the cloned B. mori PBANR exhibits high homology with the H. zea PBANR throughout the transmembrane (TM) regions, the presence of a 67-amino acid C-terminal extension in the B. mori PBANR indicates that there is a significant structural difference between the two PBANRs. Furthermore, we provide the first evidence that PBAN induces rapid internalization of the B. mori PBANR and that the C-terminal extension is essential for that internalization.

EXPERIMENTAL PROCEDURES
Insects-B. mori eggs (Shuko ϫ Ryuhaku) were purchased from Katakura Kogyo (Matsumoto, Japan) and larvae were raised on an artificial diet as described previously (28).
Cloning of Receptor cDNA-mRNA was isolated from PGs of newly eclosed B. mori females with a Micro Fast Track Kit (Invitrogen) and used as a template for first strand cDNA synthesis employing random hexamers and Superscript II reverse transcriptase (Invitrogen). To amplify receptor DNA fragments, a set of degenerate primers were designed from the conserved sequences encoding the TMs of Drosophila GPCRs CG8795 and CG8784 (GenBank TM accession no. AF522189) and two Anopheles sequences (GenBank TM accession nos. BK0001383 and BK0001384). The primer set included the sense primers 5Ј-ATGCAY-ACIGCIACNAAY-3Ј (corresponding to the second TM sequence -MHT-ATN-) and 5Ј-GCIAAYGCNACIGTNCTIAC-3Ј (corresponding to the third TM sequence -ANATVLT-) and the antisense primers 5Ј-YTGIG-CRTGRAAIGGIGCCCA-3Ј (corresponding to the sixth TM sequence -WAPFHAQ-) and 5Ј-ICGRAAYTTRTGIGACAT-3Ј (corresponding to the seventh TM sequence -MSHKFR-). Touchdown PCR amplification was performed using primers to the second and seventh TMs, the conditions consisted of 95°C for 1 min then 22 cycles of 94°C at 45 s, 52°C for 45 s (lowered by 0.5°C/cycle), 72°C for 90 s followed by 20 cycles consisting of 94°C for 30 s, 50°C for 40 s, and 72°C for 1 min, and ended at 72°C for 5 min. These products then served as templates for amplification by means of nested PCR using primers to the third and sixth TMs, the conditions of which were 95°C for 90 s followed by 35 cycles at 94°C for 30 s, 53°C for 45 s, 72°C for 90 s, and ended at 72°C for 10 min. Both amplifications were performed using LA Taq (Takara Bio, Otsu, Japan). Products corresponding to the expected sizes were gel-excised, sub-cloned using the pGEM-T Easy vector system (Promega), and sequenced. Specific primers, 5Ј-TACTGACTATTACGGCAT-T-3Ј (sense) and 5Ј-ATGAAAAACGACAGCG-3Ј (antisense), were then designed and utilized in conjunction with T3 and T7 primers to amplify the 5Ј-and 3Ј-ends of the cDNA fragment from a p50 PG cDNA library (29). The resulting products were sub-cloned into the pGEM-T Easy vector and sequenced. Primers corresponding to the 5Ј and 3Ј termini of the cDNA were used to amplify the full-length cDNA from the library, and the resulting product was sub-cloned, sequenced as before, and submitted to GenBank TM .
Northern Blot Hybridization-The Micro Fast Track kit was used to isolate mRNA from various tissues of adult (day 0) and fifth instar (day 3) B. mori larvae. After heat denaturation, 2 g of mRNA from each tissue was electrophoresed on a 1% agarose gel and then transferred to a Hybond-XL nylon membrane (Amersham Biosciences). Pre-hybridization was performed for 1 h at 68°C in Perfect Hybridization Plus buffer (Sigma). The [␣-32 P]dCTP-labeled probe was prepared from the receptor cDNA ORF (1239 bp) using a Random Primer DNA Labeling Kit Version 2 (Takara Bio) and purified with Probe Quant G50 Micro Columns (Amersham Biosciences). Probe hybridization was performed at 68°C for 18 h at which point the blot was washed in an initial solution of 0.1% SDS/1 ϫ SSC for 5 min at 68°C and then transferred to 0.1% SDS/0.2 ϫ SSC for two 5-min washes at 68°C. The blot was exposed for 18 h at Ϫ80°C to BioMax film (Kodak). For estimation of the relative abundance of B. mori PBANR transcripts, ratios of the relative blot intensities corresponding to B. mori PBANR mRNA and actin-3 mRNA were quantified with ImageGauge 4.0 software (Fuji PhotoFilm Co., Ltd.). The ratios were expressed in terms of arbitrary units and presented in histogram form.
Construction and Transfection of Expression Plasmids-To obtain cells transiently expressing B. mori PBANR, the expression plasmid, pIB-PBANR, containing the 413-amino acid ORF behind the Opie-2 promoter was generated using the pIB/V5-His-TOPO TA Expression kit (Invitrogen). Gene-specific primers 5Ј-GAGCTCATATTCGAAATGAT-G-3Ј (from Ϫ9, including an SacI site, referred to as PBANRSacF) and 5Ј-GAATTCCTATGGAGAGATCGC-3Ј (plus stop codon and an EcoRI site) allowed PCR amplification of the receptor ORF using LA Taq. A second expression plasmid, pIB-⌬C-PBANR, encoding a copy of B. mori PBANR lacking the C-terminal extension, was generated using PBAN-RSacF and the antisense gene-specific primer 5Ј-CTAAGCCGCTAAT-GTCATCTT-3Ј (including stop codon). The PCR products were cloned into the vector pIB/V5-His-TOPO and sequenced to confirm the presence and orientation of the inserts. Transfections were performed according to the Cellfectin (Invitrogen) transfection reagent protocol using adherent monolayer Sf9 cells and ϳ1.5 g of DNA/transfection. Twelve hours after transfection, the medium was replaced with fresh media containing streptomycin/kanamycin, and cells were incubated for an additional 24 -36 h at 27°C.
For confocal microscopic examination, we utilized overlap extension PCR to construct the chimeric receptor expression plasmid, pIB-PBANR-EGFP, by fusing the enhanced green fluorescent protein (EGFP) coding sequence to the C terminus of B. mori PBANR. The first PCR entailed amplification of the complete coding region of the B. mori PBANR minus the stop codon from the plasmid, pIB-PBANR, by using the specific sense PBANR primer PBANRSacF and the chimeric antisense primer 5Ј-GCCCTTGCTCACCATTGGAGAGATCGCGAT-3Ј and KOD-Plus-(Toyobo, Osaka, Japan). The second PCR amplification involved amplification of the EGFP coding sequence from pEGFP-1 (Clontech) using the antisense gene-specific primer 5Ј-CTTGTACAGC-TCGTCC-3Ј and the chimeric sense primer 5Ј-ATCGCGATCTCTCCA-ATGGTGAGCAAAGGGC-3Ј. Similarly, an EGFP-tagged version of the truncated receptor, pIB-⌬C-PBANR-EGFP, was constructed using PB-ANRSacF and the chimeric antisense primer 5Ј-GCCCTTGCTCACCA-TAGCCGCTAATGTCAT-3Ј. The second reaction using pEGFP-1 as a template was performed with the antisense gene-specific primer and the chimeric sense primer 5Ј-ATGACATTAGCGGCTATGGTGAGCAA-GGGC-3Ј. The final reactions for each construct involved the products from the two previous reactions as templates and utilized the PBANR gene-specific sense primer and the EGFP-specific antisense primer. The resulting products were sub-cloned into the expression vector as described earlier and sequenced. A control expression plasmid, pIB/EGFP, was also constructed using the gene-specific primers 5Ј-ATGGTGAGC-AAGGGCG-3Ј (sense) and 5Ј-CTTGTACAGCTCGTCC-3Ј (antisense) and pEGFP-1 as the PCR template. After ligation, the expression vector was likewise sequenced.
Measurement of Calcium Influx-On the day of the experiment, Sf9 cells transiently expressing B. mori PBANR were harvested, washed in cold phosphate-buffered saline, and suspended in a modified Ringer's buffer (15) at ϳ9 ϫ 10 5 cells/ml. From this suspension, 150-l aliquots were centrifuged (700 ϫ g for 1 min), and the cells were re-suspended in 150 l of buffer plus Fluo-4AM (Molecular Probes) at a final concentration of 2 M. Cells were incubated at room temp for 40 min in the dark on a rotating shaker. After incubation, the cells were washed twice with buffer and placed in the dark for another 20 min to allow hydrolysis of the Fluo-4AM ester bond. Fluorescence was measured on a Wallac ARVO SX plate reader (Amersham Biosciences) using a 485-nm CWlamp filter and a 535-nm emission filter. Fluorescence was monitored for 1 min prior to addition of samples and then every 10 s for 2 min following addition of samples (10 l). Collected data were analyzed using GraphPad Prism 3.0 (GraphPad Software Inc.).
Preparation of a Fluorescent PBAN Analog-To prepare a fluorescent analog of synthetic B. mori PBAN (LSEDMPATPADQEMYQPD-PEEMESRTRYFSPRLamide), we modified the N-terminal amino group with Rhodamine Red succinimidyl ester (Molecular Probes). In a 50-l reaction, 2.5 nmol of peptide, and 10 nmol of Rhodamine Red succinimidyl ester were added to 0.15 M sodium bicarbonate (pH 8.2) and stirred overnight at room temperature. The conjugated peptide, designated as Rhodamine Red-labeled PBAN (RR-PBAN), was purified by reversed-phase high-performance liquid chromatography using a dual solvent system consisting of 0.1% trifluoroacetic acid (solvent A) and 80% CH 3 CN/0.1% trifluoroacetic acid (solvent B) on a Pegasil 300-C 4 column (4.6 ϫ 250 mm; Senshu Scientific Co., Ltd., Tokyo, Japan) with Nanospace SI-2 pumps (Shiseido Co., Ltd.) and the software package Chromquest. RR-PBAN was eluted at a flow rate of 0.5 ml/min using a step-gradient that went from 1-10% B in 10 min, then to 40% B over the next 30 min, and concluded with a 10 min increase to 99% B. Absorbance was monitored at 220, 280, and 570 nm with a SpectraSystem UV6000LP detector (ThermoFinnigan). Pheromonotropic activity of RR-PBAN was confirmed using a B. mori in vivo assay (28), and RR-PBAN was stored at 4°C until needed.
Confocal Microscopic Examination of PBANR-EGFP Expression-To examine the cellular localizations of EGFP-tagged PBANR (PBANR-EGFP) and RR-PBAN, confocal microscopy imaging analyses of both live and fixed cells were performed. Sf9 cells attached to 27-mm glassbottom dishes (Matsunami, Tokyo, Japan) were transfected with ϳ1.5 g of plasmid DNA. Twelve hours after transfection, the medium was replaced with fresh media containing streptomycin/kanamycin, and cells were incubated for an additional 24 -36 h at 27°C. For live cell imaging, cells were washed with fresh media and EGFP fluorescence imaged with a Leica TCS NT confocal system using a 488-nm laser line.
To determine the degree of co-localization between PBANR-EGFP and RR-PBAN, transfected cells were washed twice with cold phosphatebuffered saline and then incubated in the presence of 50 nM RR-PBAN with or without 5 M unlabeled B. mori PBAN or 5 M B. mori adipokinetic hormone (Bachem, Budendorf, Switzerland) for 1 h at 4°C. The low temperature was utilized to minimize loss of bound RR-PBAN from the cell surface by way of receptor internalization. Cells were washed twice with cold phosphate-buffered saline to remove unbound label and fixed for 30 min with 4% paraformaldehyde at 4°C. To examine internalization of the receptor/ligand complex, cells were treated as before and then incubated at 27°C for 30 min prior to fixation. EGFP fluorescence was again imaged using a 488-nm laser line while Rhodamine Red was excited with a 568-nm laser line. For these experiments, only one laser line was used for single channel recordings to minimize cross-talk. Images were processed and merged using Photoshop 6.0 (Adobe Systems Inc., San Jose, CA).
To further evaluate receptor internalization, the fluorescence attributable to internalized RR-PBAN (i.e. 568-nm fluorescence) was measured following 30  plasma membrane. The fluorescence intensity of each region was calculated, adjusted for pixel number, and expressed as a ratio of the intracellular fluorescence (not including plasma membrane) to the total fluorescence of a cell (including the plasma membrane). Data were analyzed in GraphPad Prism 3.0 using a Student's t test.

RESULTS
Isolation of B. mori PBANR cDNA Clone-Given the potential sequence similarities with neuromedin U receptors (NmURs) (17,18), we chose a homology-based molecular approach to identify the B. mori PBANR. A search of sequences in the Drosophila and Anopheles genomes exhibiting significant homology with the Drosophila receptor CG8795 identified three additional dipteran homologues, the Drosophila receptor CG8784, and two sequences from the Anopheles genome, BK001383 and BK001384, which have subsequently been putatively identified as pyrokinin-2 receptors (31). Based on alignment of these sequences, degenerate primers to the second, third, sixth, and seventh TMs were designed. A touchdown PCR strategy using the outermost primers was initially utilized to amplify receptor candidates from the PGs of recently eclosed B. mori females. The ensuing products were re-amplified by means of nested PCR using the inner primers. From this amplification, we identified a 494-nt fragment that exhibited 50% identity with the Drosophila CG8795 receptor. Specific primers were then designed and used in conjunction with the pBluescript primers, T7 and T3, to obtain a full-length 2780-nt clone from a B. mori p50 PG cDNA library (29). The clone is comprised of a 1239-nt ORF encoding a 413-amino acid protein flanked by a 261-nt 5Ј-untranslated region and a 1280-nt 3Јuntranslated region (Fig. 1). Hydropathic analysis of the ORF generated a plot consistent with the heptahelical class of receptors (i.e. 7 TM spanning regions with an extracellular N terminus and an intracellular C terminus). The protein, which we have designated as B. mori PBANR, contains several features characteristic of the rhodopsin-like family of GPCRs (32) such as a number of highly conserved residues implicated in structural stabilization of the receptor and/or in promoting G protein-coupling (Arg 137 ), the presence of N-linked glycosylation sites (Asn 18 and Asn 21 ) in the extracellular N terminus, a potential disulfide bridge (Cys 112 -Cys 193 ) connecting the first two extracellular loops, as well as numerous potential phosphorylation sites throughout the intracellular loops and the C terminus. An ionic interaction involving the highly conserved Arg in the D/ERY motif located at the C-terminal end of TM3 (Arg 137 in B. mori PBANR) and a Glu at the C terminus of TM6 has been proposed to stabilize the inactive state of some GPCRs (33,34). In the B. mori PBANR the crucial Glu residue has been replaced by a non-charged Gln (Gln 262 ). It is possible, however, that stabilization of the inactivate receptor may still occur, albeit via a weaker interaction utilizing hydrogen bonds (33).
Comparison of the B. mori PBANR with sequences from other organisms (Fig. 2) reveals high identity (82%) with the recently cloned H. zea PBANR (GenBank TM accession no. AY319852) (15). Similarities are also seen with the Drosophila pyrokinin-2 receptors, CG8795 and CG8784 (64% similarity to each), the putative Anopheles pyrokinin-2 receptors, BK001383 and BK001384 (70 and 54% similarity, respectively), and to a lesser extent with a number of mammalian sequences coding for the NmU/growth hormone secretagogue/ thyrotropin-releasing hormone family of GPCRs. Interestingly, the homology between the B. mori and H. zea PBANR sequences deviates at both the N terminus (32% identity) and the C terminus with the B. mori PBANR containing a 67amino acid extension.
Tissue and Stage Specificity-To examine the tissue distribution of the B. mori PBANR transcript, Northern blot analyses were carried out using various adult and larval tissues. We found that expression of the ϳ2.8-kb transcript is specific to the PG as evidenced by the presence of a single resolvable band with no detectable signal observed in any of the other tissues examined (Fig. 3A). Furthermore, stage-specific expression of the B. mori PBANR within the PG indicates that it undergoes significant up-regulation 1 day prior to adult emergence (Fig.  3B). The expression profile is reminiscent of other PG-specific proteins involved in B. mori pheromonogenesis, i.e. fatty-acyl reductase, acyl-CoA desaturase, and two acyl-CoA-binding proteins (35)(36)(37), and, given that B. mori sex pheromone is initiated on the day of adult emergence, correlates well with the temporal aspect of pheromonogenesis. Accordingly, the present results suggest the functional involvement of the cloned GPCR gene in pheromonogenesis in this species as well. In addition, a Southern blot analysis was performed using genomic DNAs digested with multiple restriction enzymes. The presence of a single detectable band in each digest demonstrated that the B. mori PBANR is a single-copy gene (data not shown).
Functional Expression in Sf9 Cells-Initiation of the B. mori PBAN signal transduction cascade involves receptor-activated opening of calcium channels (7). Consequently, Sf9 cells transiently expressing B. mori PBANR were examined for their ability to mobilize extracellular calcium in response to exogenous B. mori PBAN. When challenged with varying concentrations of synthetic B. mori PBAN, the transfected cells exhibited a dose-dependent increase in fluorescence derived from the calcium indicator Fluo-4/AM (Fig. 4). No change in fluorescence To further characterize the cellular functions of B. mori PBANR, we transfected Sf9 cells with an expression plasmid containing a chimera gene encoding B. mori PBANR tagged at its C terminus with EGFP, referred to as PBANR-EGFP. In addition, for confocal microscopic examination of the specificity of B. mori PBANR binding we prepared a red fluorescent analogue of B. mori PBAN labeled at the N terminus with Rhodamine Red (RR-PBAN), which was shown using in vivo bioassays to retain the same pheromonotropic activity as unlabeled B. mori PBAN (data not shown). In Sf9 cells transfected with PBANR-EGFP, green fluorescence was localized exclusively to the plasma membrane (Fig. 5A), whereas control cells transfected with EGFP alone exhibited diffuse fluorescence through-out the cytosol (Fig. 5B), indicating that the EGFP-tagged B. mori PBANR trafficks correctly and verifies that the B. mori PBANR sequence encodes a cell surface protein. When the PBANR-EGFP-transfected cells were incubated with RR-PBAN, the ligand bound to the cell surface and co-localized with PBANR-EGFP (Fig. 5, I-K). Neither PBANR-EGFP cells incubated with Rhodamine Red alone (Fig. 5, C-E) nor EGFPtransfected cells incubated with RR-PBAN (Fig. 5, F-H) exhibited co-localization of the green and red fluorescence signals, indicating that RR-PBAN binding is not the result of nonspecific interactions. Furthermore, RR-PBAN binding and co-localization with PBANR-EGFP was greatly diminished when incubated with 100-fold excess unlabeled B. mori PBAN (Fig. 5, L-N), an effect that was not observed in cells incubated with the unrelated insect neuropeptide adipokinetic hormone (Fig. 5, O-Q). These results confirm that RR-PBAN binding is the result of a specific, reversible interaction with B. mori PBANR.
We also observed PBAN-induced internalization of PBANR-EGFP and RR-PBAN. In the absence of exogenous PBAN, PBANR-EGFP fluorescence is predominantly localized to the plasma membrane (Fig. 5A). Exposure to RR-PBAN for 30 min at 27°C, however, results in internalization of PBANR and the bound RR-PBAN, an event that can be visually discerned by the presence of green and red fluorescent vesicles scattered throughout the perinuclear region (Fig. 5, R and S). Furthermore, the co-localization of the green and red vesicles within the interior of the cell could clearly be seen following a 5-min incubation at 27°C, demonstrating the rapidity with which receptor desensitization is invoked (data not shown). Once internalized, the PBANR⅐RR-PBAN complex appears to remain intact as evidenced by the co-localization of the red and green fluorescent signals (Fig. 5, panel T1). To examine the functionality that may reside in the B. mori C-terminal extension, we constructed expression plasmids harboring genes encoding a C-terminal deleted B. mori PBANR (⌬C-PBANR), in which the new terminal residue, Ala 346 , corresponds to the terminal residue of the H. zea PBANR. An expression plasmid harboring an EGFP-tagged ⌬C-PBANR (⌬C-PBANR-EGFP) was likewise constructed. When transiently expressed in Sf9 cells, ⌬C-PBANR-EGFP localized to the cell surface and bound RR-PBAN (Fig. 5, U-W) but failed to internalize (Fig. 5, X-Z and  Z1). Similarly, the non-EGFP-tagged construct ⌬C-PBANR also failed to undergo PBAN-induced internalization (Fig. 6A) indicating that the internalization defect is not the result of EGFP interference. Furthermore, the absence of observable internalized ligand in the ⌬C-PBANR construct cannot be attributed to differences in the microscopic focal plane, because no vesicles could be seen when cells were sectionally imaged along their z-axis (Fig. 6A). In contrast, when cells expressing the full-length PBANR are similarly examined, numerous vesicles corresponding to internalized RR-PBAN can be seen dispersed throughout the cell interior (Fig. 6B). DISCUSSION Despite recent successes in elucidating many of the components of the pheromone biosynthetic pathways, the characterization of PBANRs has lagged significantly behind. In the present experiments, we built on the conclusion of Hewes and Taghert (17) that the B. mori PBANR would be homologous to the mammalian NmUR. Consequently, we designed degenerate primers based on the alignment of deduced dipteran homologues. Using a 456-bp fragment initially amplified from PGs of a hybrid B. mori strain, we were able to obtain a full-length clone encoding the B. mori PBANR from a PG cDNA library of the inbred B. mori p50 strain (Fig. 1). Expression of the B. mori PBANR transcript is specific to the PG (Fig. 3A), and it, like other PG-specific proteins involved in B. mori pheromonogenesis (35)(36)(37), undergoes significant up-regulation prior to adult eclosion (Fig. 3B). Calcium mobilization assays demonstrated that the B. mori PBANR is fully functional and capable of initiating the first phase of the PBAN signal transduction cascade, i.e. extracellular calcium influx (Fig. 4). Our functional studies using PBANR-EGFP demonstrated that the B. mori PBANR gene encodes a cell surface receptor (Fig. 5A) that specifically binds PBAN (Fig. 5, I-Q) and that the PBANR⅐RR-PBAN complex undergoes rapid agonist-induced internalization (Fig. 5, R-T). Taken together, these results allow us to conclude that the cloned GPCR gene encodes the B. mori PBANR, the GPCR utilized by PBAN in this species to regulate pheromone biosynthesis.
Although B. mori PBANR exhibits considerable homology with the recently cloned H. zea PBANR throughout the TM regions, the B. mori PBANR sequence encodes a structurally distinct GPCR as the aforementioned homology diminishes within the N terminus (32% identity) and the C terminus where the B. mori PBANR has a 67-amino acid extension (Fig.  2). It is possible, however, given the presence of potential protease cleavage sites in the C-terminal extension (Arg 350 / Arg 351 , Arg 383 /Arg 384 , and Arg 386 /Arg 387 ) that the B. mori PBANR may undergo post-translational processing. Indeed, processing of the first potential cleavage site, Arg 350 /Arg 351 , would result in a C terminus that more closely resembles that of the H. zea PBANR. To examine this possibility, as well as any functionality that may reside in the B. mori C terminus, we constructed the truncation mutants, ⌬C-PBANR and ⌬C-PBANR-EGFP. When expressed in Sf9 cells, ⌬C-PBANR-EGFP localized to the plasma membrane and bound RR-PBAN (Fig. 5, U-W). However, exposure to ligand at 27°C for 30 min failed to promote internalization as both ⌬C-PBANR-EGFP and RR-PBAN remained at the cell surface (Figs. 5, X-Z, and 6A). To further assess the internalization event, the fluorescence attributable to internalized RR-PBAN in both ⌬C-PBANR-and PBANR-expressing cells was measured and quantified as a ratio of the intracellular fluorescence (not including plasma membrane) to the total fluorescence of a cell. The results shown in Fig. 6C demonstrate a clear difference in internalization efficiencies between PBANR and ⌬C-PBANR and indicate that the B. mori PBANR C-terminal extension is functionally indispensable and that post-translational processing of the potential cleavage sites likely does not occur. Indeed, recent studies indicate that basic residues in the C-terminal tails of GPCRs may be necessary for receptor function (38).
The C-terminal extension contains a number of sites that may be utilized to regulate the internalization process. Initiation of the internalization process involves phosphorylation of specific residues within the C-terminal tail by either second messenger-dependent protein kinases such as cAMP-dependent protein kinases A and C or G protein-coupled receptor kinases (19). The B. mori PBANR C-terminal extension contains two consensus second messenger-dependent protein kinase sites, a protein kinase C site (SQR; residues Ser 366 -Arg 368 ) and a protein kinase A site (RRLS; residues Arg 386 -Ser 389 ). However, any one, or combination of, the five Thr and six Ser residues may potentially serve as the phosphorylation site(s) utilized in promoting PBANR internalization. The B. mori Cterminal extension also contains a consensus tyrosine-based sorting signal (YXXL) between residues Tyr 360 and Leu 363 that has been reported to regulate the intracellular trafficking of protease-activated receptor-1 (39). Although not situated within the 67-amino acid C-terminal extension, the B. mori PBANR also contains the consensus sequence NPXXY spanning residues 325-329 (NPFLY) in the seventh TM. This motif is necessary for regulating the internalization mechanism of a number of receptors, including the human N-formyl peptide receptor (40), the ␤ 2 -adrenergic receptor (41,42), and the platelet-activating factor receptor (43). The complexity of the molecular mechanisms underlying receptor internalization, however, is highlighted by the fact that the NPXXY motif is not utilized as a general internalization sequence in either the gastrin-releasing hormone receptor (44) or the angiotensin II receptor (45,46).
The absence of the C-terminal extension in the H. zea PBANR is intriguing because it suggests that the receptor either does not undergo agonist-induced internalization, or at the very least, indicates that internalization proceeds by a very different mechanism. The observed structural differences between the two PBANRs could be an indication of species-specific PBANR types, a long form (B. mori PBANR) and a short form (H. zea PBANR), reminiscent of the type I gonadotropinreleasing hormone receptors (GnRHRs). The non-mammalian type I GnRHRs possess a C-terminal tail and exhibit rapid agonist-induced receptor internalization, whereas the mammalian type I GnRHRs, in contrast, lack a C-terminal tail and exhibit significantly different internalization kinetics (47). Interestingly, the signal transduction cascades utilized by the mammalian and non-mammalian type I GnRHRs are also different with the mammalian type I GnRHRs preferentially signaling through the G q/11 -linked protein kinase C pathway, whereas the non-mammalian type I GnRHRs utilize the G slinked protein kinase A pathway (48). Despite the striking similarities between the PBANRs and the type I GnRHRs, there are two concerns that must be addressed before the comparison can be taken any further. First, the structural differences observed in the type I GnRHRs occur between species from different taxonomic classes (mammalian versus nonmammalian), whereas the structural difference between the PBANRs occurs at a much lower taxonomic classification (Bombycidae versus Noctuidae). Second, and most importantly, because no transcript analyses regarding tissue specificity have been performed, concerns remain regarding the biological relevancy of the H. zea PBANR.
In summary, we have provided the first evidence demonstrating that PBAN promotes internalization of the B. mori PBANR and that the internalization event is mediated by the actions of the 67-amino acid C-terminal extension. These data are a strong indication that the B. mori PBANR is structurally and functionally distinct from H. zea PBANR. Only with further characterization of the two PBANRs will it be possible to account for the different signaling pathways utilized in the two species and take us one step closer to unraveling the complexity of the molecular mechanisms underlying lepidopteran sex pheromone production.