Heterodimeric Fly Glycoprotein Hormone-2 (GPA2) and Glycoprotein Hormone-?5 (GPB5) Activate Fly Leucine-Rich Repeat-Containing G Protein-Cou
http://www.100md.com
内分泌学杂志 2005年第8期
Division of Reproductive Biology, Department of Obstetrics and Gynecology, Stanford University School of Medicine, Stanford, California 94305-5317
Address all correspondence and requests for reprints to: Aaron J. W. Hsueh, Stanford University School of Medicine, Department of Obstetrics and Gynecology, 300 Pasteur Drive, Room A344, Stanford, California 94305-5317. E-mail: aaron.hsueh@stanford.edu.
Abstract
Glycoprotein hormones play important roles in thyroid and gonadal function in vertebrates. The glycoprotein hormone -subunit forms heterodimers with different ?-subunits to activate TSH or gonadotropin (LH and FSH) receptors. Recent genomic analyses allowed the identification of another -subunit, GPA2, and another ?-subunit, GPB5, in human, capable of forming heterodimers to activate TSH receptors. Based on comparative genomic searches, we isolated the fly orthologs for human GPA2 and GPB5, each consisting of 10 cysteine residues likely involved in cystine-knot formation. RT-PCR analyses in Drosophila melanogaster demonstrated the expression of GPA2 and GPB5 at different developmental stages. Immunoblot analyses further showed that fly GPA2 and GPB5 subunit proteins are of approximately 16 kDa, and coexpression of these subunits yielded heterodimers. Purified recombinant fly GPA2/GPB5 heterodimers were found to be glycoproteins with N-linked glycosylated -subunits and nonglycosylated ?-subunits, capable of stimulating cAMP production mediated by fly orphan receptor DLGR1 but not DLGR2. Although the fly GPA2/GPB5 heterodimers did not activate human TSH or gonadotropin receptors, chimeric fly GPA2/human GPB5 heterodimers stimulated human TSH receptors. These findings indicated that fly GPA2/GPB5 is a ligand for DLGR1, thus showing the ancient origin of this glycoprotein hormone-seven transmembrane receptor-G protein signaling system. The fly GPA2 also could form heterodimers with human GPB5 to activate human TSH receptors, indicating the evolutionary conservation of these genes and suggesting that the GPA2 subunit may serve as a scaffold for the ?-subunit to activate downstream G protein-mediated signaling.
Introduction
GLYCOPROTEIN HORMONES are heterodimers consisting of - (GPA) and ?-subunits (GPB) assembled by noncovalent bonds. The cystine-knot-containing glycoprotein hormones are essential for gonadal and thyroid function in all vertebrates by binding to seven-transmembrane G protein-coupled receptors leading to the activation of downstream signal transduction (1, 2, 3). Gonadotropins regulate the growth and differentiation of the ovary and testis, whereas TSH is essential for energy balance. In amphibians, it is known that TSH induces metamorphosis in tadpoles (4).
Recently a new glycoprotein hormone, thyrostimulin, was discovered (5). It consists of two newly identified subunits, GPA2 and GPB5, with homology to the common -subunit and the TSH ?-subunit, respectively (6). This heterodimeric glycoprotein hormone is capable of stimulating TSH receptors by inducing cAMP production and thymidine incorporation in cultured thyroid cells and increasing serum thyroxine levels in TSH-suppressed rats in vivo. In Drosophila melanogaster, a glycoprotein ?-subunit, fly GPB5, has been reported showing high homology to the mammalian GPB5 subunit with 10 cysteine residues as compared with the other ?-subunit carrying 12 cysteine residues (6).
There are at least two leucine-rich repeat-containing G protein-coupled receptor (LGR) genes in D. melanogaster (7), DLGR1 and DLGR2 (8, 9). Recent studies demonstrated that DLGR2 is the receptor for the insect neurohormone, bursicon, essential for hardening and darkening of the new cuticle (10, 30). Here we characterized a novel glycoprotein -subunit in Drosophila melanogaster and named it as fly GPA2. We also identified orthologs for GPA2 and GPB5 in mosquito and other species. After the isolation of both fly GPA2 and fly GPB5 cDNAs, we found that fly GPA2 is capable of forming heterodimers with fly GPB5. Based on the hypothesized coevolution of ligand/receptor pairs (11), we demonstrated the ability of the fly GPA2/GPB5 heterodimer to stimulate cAMP production mediated by the orphan fly receptor, DLGR1, showing high homology to the mammalian TSH receptor. Furthermore, we demonstrated that the chimeric heterodimeric molecule, fly GPA2/human GPB5, is capable of stimulating human TSH receptors.
Materials and Methods
Hormones and reagents
DMEM/Ham’s F-12 (DMEM/F12) and Dulbecco’s PBS were obtained from Life Technologies (Gaithersburg, MD). Anti-FLAG M1 monoclonal antibodies and anti-FLAG M1 affinity gels were obtained from Sigma (St. Louis, MO). Anti-fly GPB5 antibodies were generated using a synthetic peptide encoding Ac-DSSEISDWKFPYKRSFHPC-amide corresponding to amino acids 78–95 of fly GPB5 (Biosource, Camarillo, CA). The antiserum titer was checked by ELISA using the synthetic peptide as the antigen. This fly GPB5 antibody does not cross-react with human GPB5, fly GPA2, or human GPA2 (data not shown). Antihuman GPB5 antibodies were described previously (5) and do not cross-react with fly GPA2 or GPB5 subunits. Human recombinant FSH (Org32489 and human choriogonadotropin (hCG) (CR-129) were obtained from Organon (Oss, The Netherlands) and the National Hormone and Pituitary Program (Torrance, CA), respectively. Conditioned media containing recombinant bursicon were collected from human 293T cells expressing burs/pburs heterodimers (10).
Identification and cloning of fly GPA2
To identify fly GPA2, the BLAST server at the National Center for Biotechnology Information (National Institutes of Health, Bethesda, MD) was used to search the expressed sequence tag (EST) database based on the amino acid sequence of human GPA2 leading to the identification of an Anopheles gambiae EST sequence (GenBank accession no. XM_317164). Using the deduced amino acid sequence of Anopheles gambiae GPA2, a sequence for the putative GPA2 subunit was found in the high throughput genomic sequence of D. melanogaster (GenBank accession no. AC017657).
RT-PCR
D. melanogaster poly A+ RNA (CLONTECH Laboratories, Inc., Palo Alto, CA) was used for RT-PCR. RNAs obtained from flies at different developmental stages were transcribed into cDNAs using Omniscript reverse transcriptase (QIAGEN, Valencia, CA) and oligo (dT)12–18 (Invitrogen, Carlsbad, CA). PCR was performed using the FastStart high-fidelity PCR system (Roche Applied Science, Indianapolis, IN) with 100 ng cDNA. PCR amplification of cDNA was carried out under the following conditions: 94 C for 30 sec for denaturation, 59 C for 30 sec for annealing, and 72 C for 45 sec for extension. The specific primers are: fly GPA2 (426-bp product), 5'-atgccaaagccatggccaatctca-3' and 5'-ctaatcctttttgcagtgataacaactac-3'; fly GPB5 (510-bp product), 5'-atgctcagaataatttttttcaggac-3' and 5'-ctaatagtccaagtttttggtatc-3'; ribosomal protein 49 (391-bp product); and 5'-atgaccatccgcccagcatac-3' and 5'-gagaacgcaggcgaccgttgg-3'. The PCR products were analyzed using electrophoresis of 1.5% agarose gels stained with ethidium bromide.
Generation of glycoprotein hormone subunit expression plasmids
RT-PCR products were used to generate expression constructs for fly GPA2 and fly GPB5 as described previously (5). All cDNAs were subcloned into the bipromoter expression vector pBudCE4.1 (Invitrogen) according to different combinations of subunit cDNAs. To allow efficient detection and purification of the recombinant proteins, a FLAG-epitope (DYKDDDDK) was added to the N terminus of fly GPA2. For generating fly GPA2 or fly GPB5 subunits alone, each cDNA was subcloned into pcDNA3.1 Zeo (Invitrogen). Fidelity of the RT-PCR products was confirmed by sequencing of the final constructs on both strands before use in expression studies.
Expression of recombinant proteins in 293T cells
Human 293T cells were maintained in DMEM/F12 supplemented with 10% fetal bovine serum (FBS). Cells were transfected using 7 μg of plasmid suspended in FUGENE6 (Roche Applied Science). Clonal cell lines stably expressing fly and human GPA2/GPB5 were selected under 500 μg/ml of Zeocin (Invitrogen) and maintained in DMEM/F12 containing 10% FBS and 100 μg/ml Zeocin. When the cells reached 90–100% confluency, the medium was replaced with DMEM/F12 without FBS or Zeocin. After 4 d of serum-free culture, the condition media were harvested and cell debris was cleared. The supernatant was concentrated with the iCON concentrator 20K MWCO (Pierce, Rockford, IL). The concentrated sample was purified using the anti-FLAG M1 affinity gel according to the manufacturer’s instructions (Sigma). Protein concentration of fly GPA2/GPB5 was determined by Bradford protein assay (Bio-Rad Laboratories, Hercules, CA). Purified tagged fly GPA2/GPB5 protein was used as a standard to estimate the concentration of chimeric recombinant fly GPA2/human GPB5 in the conditioned media using immunoblotting against the FLAG-epitope.
Immunoblotting analyses of recombinant proteins
Immunoblotting analyses were performed using anti-FLAG M1 antibodies (Sigma) or rabbit polyclonal antibodies against fly GPB5 (1:1000 dilution) or human GPB5 (1:1000 dilution). Samples were mixed with loading buffer under reducing conditions (5% ?-mercaptoethanol) before immunoblotting. After electrophoresis using 4–20% gradient polyacrylamide gels (Bio-Rad Laboratories), proteins were transferred to polyvinylidene difluoride membranes (Hybond-P, Amersham, Buckinghamshire, UK). Signals were detected after immunofluorescent imaging using the ECL system (Amersham).
To remove N-linked carbohydrate side chains, conditioned media were diluted in the deglycosylation buffer [50 mM sodium phosphate (pH 7.5), 1% Nonidet P-40, 0.5% sodium dodecyl sulfate, and 1% ?-mercaptoethanol] and incubated with 7.7 mIU of peptide N-glycosidase F (PNGase F; New England BioLabs, Beverly, MA) at 37 C for 1 h. Samples were mixed with loading buffer under reducing conditions (5% ?-mercaptoethanol) before immunoblotting. Purified fly GPA2/GPB5 protein was cross-linked using 0.27 mM disuccinimidyl suberate (Pierce) for 30 min, and reaction was terminated by addition of 1 M Tris-HCl (pH 7.4). Cross-linked complexes also were monitored using SDS-PAGE (4–20%) under reducing conditions.
Transfection of cells and analysis of signal transduction
Human 293T cells were maintained in DMEM/F12 supplemented with 10% FBS, 100 μg/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine. Before transfection, 2 x 105 cells were seeded in 12-well tissue culture plates (Corning, Corning, NY). When cells were 80–90% confluent, transient transfection was performed using 1 μg of plasmids encoding DLGR1 (9) or human TSH receptor cDNA with Lipofectamine 2000 (Invitrogen). After 24 h incubation with Lipofectamine, cells (2 x 105/ml) were placed on 24-well tissue culture plates (Corning, Corning NY) and preincubated at 37 C for 30 min in the presence of 0.25 mM 3-isobutyl-1-methyl xanthine (Sigma) before treatment with or without hormones for 16 h. At the end of incubation, cells and medium were frozen. After thawing, samples were heated to 95 C for 3 min to inactivate phosphodiesterase activity and total cAMP levels were measured in triplicate by a specific RIA as described previously (12). All experiments were repeated at least three times using cells from independent transfections. To monitor transfection efficiency, 0.5 μg of the ?-galactosidase plasmid (13) was routinely included in the transfection mixture, and ?-galactosidase activity in the cell lysate was measured as described (14).
Structural modeling
Comparative protein modeling was performed with the SWISS-MODEL server (http://swissmodel.expasy.org) using the experimentally determined structure for hCG as the template (Protein Data Bank, accession code 1HRP). DeepView-Swiss-PdbViewer and Protein Explorer (http://molvis.sdsc.edu/protexpl) were used to visualize the three-dimensional structure as well as to conduct structural comparisons between hCG and fly GPA2/GPB5.
Results
Isolation of fly GPA2 and fly GPB5 cDNAs and comparison with their invertebrate and vertebrate homologs
Based on the amino acid sequence of human GPA2, a homologous EST sequence of the Anopheles gambiae GPA2 was found in GenBank (accession no. XM_317164) using the BLAST search program. The open reading frame of Anopheles gambiae consisting of 121 amino acids was used for another BLAST search to identify a partial EST sequence of D. subobscura GPA2 (GenBank accession no. U73802). This stretch of amino acid sequence was used further to identify a genomic sequence of D. melanogaster using Mega-BLAST. Figure 1A shows nucleotide and predicted protein sequences of D. melanogaster GPA2 (GenBank accession no. AY940435).
FIG. 1. Isolation of GPA2 and GPB5 cDNAs from D. melanogaster and comparison with orthologs from invertebrates and vertebrates. A, Nucleotide and deduced protein sequences of the fly GPA2 subunit. Underlined sequences and the asterisk indicate the predicted signal peptide and a putative N-linked glycosylation site, respectively. B, Sequence alignment of GPA2 from fly, mosquito, nematode, human, mouse, rat, fugu, and zebrafish. Residues conserved in all eight species are shown in upper-case letters. Lower-case letters represent residues conserved in five of eight species. C, Nucleotide and deduced protein sequences of the fly GPB5 subunit. Underlined sequence represents the predicted signal peptide. D, Sequence alignment of GPB5 from fly, mosquito, nematode, human, mouse, rat, and tetraodon. Residues conserved in all seven species are shown in upper-case letters. Lower-case letters represent conserved residues in four of seven species. E, Comparison of cysteine signatures among fly GPA2, fly GPB5, human GPB5, and the human TSH?. Different pairs of cysteine residues forming disulfide bridges are connected by lines. Bold lines: disulfide bonds for the cystine knot structure; thin lines: additional disulfide bridges; dashed line: the latch disulfide bridge found in the seat belt structure. F, RT-PCR amplification of fly GPA2 and fly GPB5 mRNAs. Transcripts for GPA2 and GPB5 genes were amplified using PCR from cDNAs derived from D. melanogaster embryo, larva, and adults as described in Materials and Methods. Ribosomal protein 49 (RP49) expression served as an internal control. (Figure continues on next page.)
FIG. 1A. Continued
To confirm the predicted sequence of fly GPA2 cDNA, RT-PCR was carried out using gene-specific primers from templates derived from adult D. melanogaster RNA. Comparison of this cDNA with the corresponding genomic sequence revealed that the fly GPA2 gene has two exons divided by a 58-bp intron. Deduced fly GPA2 consists of 141 amino acid residues with a 31-amino acid-long signal peptide (Fig. 1A). Similar to its human counterpart, the mature fly GPA2 has 10 cysteine residues capable of forming a cystine-knot structure (15). There is also a potential N-linked glycosylation site at 124N. Sequence alignment of GPA2 from fly, mosquito, nematode, fish, and other species (Fig. 1B) indicated a complete conservation of cysteine residues in the predicted mature proteins. Although the fly GPB5 cDNA has already been predicted (6), its expression has not been verified. To confirm the expression of fly GPB5, RT-PCR was performed using gene-specific primers to amplify transcripts from adult D. melanogaster RNA. The deduced amino acid sequence of fly GPB5 cDNA consists of a signal peptide of 25-amino acid residues and a mature region with 144 amino acids but without a putative N-linked glycosylation site (Fig. 1C). Sequence alignment of GPB5 molecules from fly, mosquito, nematode, and other species (Fig. 1D) indicated the conservation of key cysteine residues for cystine-knot formation; however, the fly and mosquito GPB5 molecules have an extended C tail.
Predicted fly GPA2 and fly GPB5 subunits show 22 and 19% homology in amino acid sequences in the mature protein as compared with human GPA2 and human GPB5, respectively. To characterize the putative structure of fly GPA2 and fly GPB5, tertiary structural models were generated by comparative protein modeling using the crystal structure of - and ?-subunits of hCG as templates (FUGUE Profile Lib Search for fly GPA2 and 3D-JIGSAW for fly GPB5) (16). As shown in Fig. 1E, the predicted cystine-knot structure of fly GPA2 is comprised of three disulfide bridges (44C-104C, 68C-133C, 72C-135C) and two additional disulfide bonds (58C-118C and 103C-138C) outside the cystine knot. Likewise, fly GPB5 has a cystine-knot structure consisting of three disulfide bridges (45C-97C, 73C-131C, and 77C-133C) and two additional disulfide bonds (64C-112C and 136C-143C). Different from human GPB5 (6), fly GPB5 has an extended C-terminal tail corresponding to the seat belt structure of the mammalian ?-subunits. However, the corresponding C-tail region of TSH?, truncated in human GPB5, has one additional cysteine residue (125C) forming an additional disulfide bridge with the 39C residue of TSH? to form the seat belt structure presumably holding the -subunit tightly in the hole (17, 18).
Expression of fly GPA2 and fly GPB5 transcripts during different developmental stages of D. melanogaster
To elucidate the expression of fly GPA2 and fly GPB5 transcripts during development, RT-PCR was performed using cDNAs from different developmental stages of D. melanogaster. As shown in Fig. 1F, both GPA2 and GPB5 were amplified as DNA fragments of predicted sizes from RNA obtained from flies at embryonic, larval, and adult stages.
Immunoblotting analyses of purified recombinant fly GPA2/GPB5 heterodimers
We hypothesized that fly GPA2 and fly GPB5, similar to their human counterparts, could form heterodimers. 293T cells were transfected with a bipromoter expression vector encoding both FLAG-tagged fly GPA2 and nontagged fly GPB5. Conditioned media expressing fly GPA2/GPB5 was affinity purified using anti-FLAG M1 affinity gels. In Fig. 2A, Coomassie blue staining analyses indicates an abundance of proteins of 66 kDa in the conditioned media, likely representing human albumin secreted by 293T cells. In contrast, a single band of approximately 16 kDa was found after affinity purification of the tagged GPA2/GPB5 heterodimer. To further characterize the purified recombinant proteins, immunoblotting analyses were performed. Under reducing conditions, purified fly GPA2/GPB5 proteins were detected as bands of approximately 16 kDa using the anti-FLAG M1 antibody against the FLAG-tagged fly GPA2 subunit or the anti-fly GPB5 antibody against GPB5 (Fig. 2B). After treatment with N-glycosidase, a 12-kDa band was detected using the anti-FLAG M1 antibody corresponding to the predicted size of the deglycosylated form of fly GPA2. This finding is consistent with the presence of a putative N-glycosylation site at 124N of fly GPA2. In contrast, the migration of the fly GPB5 subunit was not affected by N-glycosidase treatment, indicating this subunit is not N-glycosylated. To further demonstrate the heterodimeric nature of purified GPA2/GPB5 proteins, we performed cross-linking analyses using disuccinimidyl suberate. After cross-linking of the purified fly GPA2/GPB5, immunoblotting using either the anti-FLAG or the anti-GPB5 antibodies revealed the presence of high-molecular-mass complexes of approximately 32 kDa, corresponding to the size of the predicted heterodimer (Fig. 2C).
FIG. 2. Characterization of recombinant fly GPA2/GPB5 heterodimers. A, Purification of fly GPA2/GPB5 using M1 affinity chromatography. Coomassie blue staining of conditioned media (CM; left lane) and M1 affinity-purified fly GPA2/GPB5 heterodimers (right lane). B, Immunoblotting analyses of purified fly GPA2/GPB5 appended at the N terminus of GPA2 with the FLAG epitope. Left, Detection using anti-FLAG M1 antibodies; right, detection using antibodies against fly GPB5. Some samples were pretreated with N-glycosidase F to remove N-linked carbohydrate side chains. C, Cross-linking of purified fly GPA2/GPB5 using disuccinimidyl suberate. D, Tertiary structure comparison between the fly GPA2/GPB5 heterodimer and hCG. Structural model for the fly GPA2/GPB5 heterodimer was generated based on the hCG crystal structure template. To optimize the modeling of the tertiary structure of fly GPA2/GPB5, the histidine residue in position 137 of fly GPA2 was replaced by the corresponding threonine residue in position 110 of the human common -subunit. Views from two different angles (view A with the -subunit in the foreground; view B with the ?-subunit in the foreground) are presented. The -subunits are indicated by the white color, whereas the ?-subunits are in green. Cysteine residues important for the formation of structure-determining cystine bonds in the - and ?-subunits are denoted in brown and purple, respectively. The seat belt structure formed by the C terminus of the hCG? is circled for comparison with the corresponding region in the fly molecule (dashed circle).
Based on the crystal structure of hCG, we generated a structural model for the fly GPA2/GPB5 heterodimer. As shown in Fig. 2D, the fly GPA2/GPB5 likely forms a heterodimer similar to hCG. Although the fly GPB5 lacks the last cysteine residue in its C terminus for forming the S-S bond found in the seat belt region of CG-? (Fig. 2D, circled), a similar structure could still be modeled in the fly GPA2/GPB5 heterodimer (Fig. 2D, dashed circle).
Fly DLGR1 is activated by recombinant fly GPA2/GPB5 heterodimers
Based on the hypothesis that ligand-receptor pairs coevolved during evolution, the fly GPA2/GPB5 heterodimer is likely the ligand for fly DLGR1 with closest sequence homology to type A LGRs including mammalian TSH, LH, and FSH receptors (19). We tested the ability of this recombinant protein to activate DLGR1. 293T cells were transiently transfected with the DLGR1 expression plasmid and ligand signaling was estimated after treatment with purified fly GPA2/GPB5. Figure 3A shows in cells expressing DLGR1, treatment with fly GPA2/GPB5 led to dose-dependent increases in total cAMP production with an EC50 value of 3.2 nM. In contrast, conditioned media expressing fly GPA2 or fly GPB5 alone did not stimulate cAMP production. Furthermore, neither fly GPA2/GPB5 heterodimers nor the individual subunits alone were capable of activating DLGR2. In contrast, DLGR2 responded to treatment of 50 nM bursicon with a major increase in cAMP production (Fig. 3B) (10).
FIG. 3. Activation of fly DLGR1 by recombinant fly GPA2/GPB5 heterodimers. A, Dose-dependent effects of purified fly GPA2/GPB5 to stimulate cAMP production by 293T cells expressing DLGR1. 293T cells were treated with purified fly GPA2/GPB5 or conditioned media containing fly GPA2 or fly GPB5. B, Lack of activation of fly DLGR2 by fly GPA2/GPB5 heterodimers. 293T cells transiently expressing DLGR2 were treated with purified fly GPA2/GPB5, individual subunits, or recombinant bursicon, all at 50 nM. Mean ± SD (n = 3).
Chimeric fly GPA2/human GPB5 heterodimers are capable of activating human TSH receptors but not fly DLGR1
We further tested the ability of fly GPA2/GPB5 heterodimers to activate human glycoprotein hormone receptors. As shown in Fig. 4, A–C, neither fly GPA2/GPB5 heterodimers nor the individual fly subunit alone stimulated cAMP production in cells expressing human TSH, LH, or FSH receptors. Because GPA2 genes showed high sequence conservation between fly and human orthologs, we hypothesized that chimeric heterodimers consisting of fly and human subunits could be formed. We constructed expression vectors encoding fly GPA2 and human GPB5 and expressed the recombinant heterodimers. After affinity purification based on the FLAG tag appended to fly GPA2, the heterodimers showed immunoreactive bands at 16 kDa corresponding to each subunit detected by anti-FLAG M1 antibodies and human GPB5 antibodies, respectively (Fig. 5A). After treatment with glycosidase F, both subunits showed a decrease in their size to 12 kDa as detected using anti-FLAG and antihuman GPB5 antibodies, respectively. These results are consistent with the presence of the potential N-linked glycosylation sites in fly GPA2 and at 63N in the human GPB5 subunit (5).
FIG. 4. Fly GPA2/GPB5 heterodimers are not capable of stimulating human TSH or gonadotropin receptors. Cells expressing human TSH receptors (A), human LH receptors (B), and human FSH receptors (C) were incubated with 10 nM of fly GPA2/GPB5, fly GPA2, or fly GPB5. For positive controls, cells were treated with 10 nM human GPA2/GPB5 (thyrostimulin), 10 ng/ml of hCG, or 100 mIU/ml of human FSH. Mean ± SD (n = 3).
FIG. 5. Chimeric fly GPA2/human GPB5 heterodimers are capable of activating human TSH receptors. A, Characterization of chimeric fly GPA2/human GPB5 heterodimers using immunoblots. Conditioned media containing fly GPA2 (appended at N terminus with the FLAG-epitope) and human GPB5 were subjected to M1 affinity chromatography before immunoblotting analyses. Left, Detection using anti-FLAG M1 antibodies; right, detection using antibodies against human GPB5. Some samples were pretreated with N-glycosidase F to remove N-linked carbohydrate side chains. B, Stimulation of human TSH receptors by fly GPA2/human GPB5 heterodimers and purified human GPA2/GPB5 (thyrostimulin). The concentration of fly GPA2/human GPB5 in the conditioned media was estimated by immunoblotting analysis as described in Materials and Methods. Mean ± SD (n = 3). C, Lack of stimulation of cAMP production by DLGR1-expressing cells treated with chimeric fly GPA2/human GPB5 and human GPA2/GPB5 (thyrostimulin). Mean ± SD (n = 3).
We further tested the ability of the chimeric heterodimer molecule, fly GPA2/human GPB5, to stimulate cAMP production in cells expressing human TSH receptors. As shown in Fig. 5B, conditioned media containing chimeric fly GPA2/human GPB5 heterodimers stimulated cAMP production in cells expressing human TSH receptors. However, the potency of the chimeric molecules is much lower than human GPA2/GPB5 (thyrostimulin) with EC50 values of 9.8 nM and 3.0 pM, respectively. In contrast, treatment with 100 nM of fly GPA2 or human GPB5 alone did not alter cAMP production (data not shown). These results indicated that, despite a separation of about 1 billion years during evolution, the fly GPA2 subunit retains the ability to heterodimerize with human GPB5 for activation of the human TSH receptors. In contrast, neither the chimeric fly GPA2/human GPB5 heterodimer nor human GPA2/GPB5 was capable of stimulating cAMP production by cells expressing DLGR1 (Fig. 5C).
Discussion
Recent genomic analyses allowed the identification of the new -subunit GPA2 and ?-subunit GPB5 in human capable of forming heterodimers to activate TSH receptors (5, 6). Taking advantage of comparative genomic searches using human sequences for GPA2 and GPB5, we characterized the fly orthologs for human GPA2 and GPB5 consisting of 10 cysteine residues capable of forming a cystine-knot structure. RT-PCR results demonstrated the existence of the transcripts for these genes in different developmental stages in fly. Furthermore, two fly glycoprotein hormone subunits formed a heterodimeric glycoprotein hormone capable of activating the DLGR1 receptor. In addition, we demonstrated that fly GPA2 also could form heterodimers with human GPB5 to activate the human TSH receptor, albeit with a lower potency as compared with human GPA2/GPB5 (thyrostimulin). These findings indicate the evolutionary conservation of these genes and suggest that the GPA2 subunit could serve as a scaffold to present the ?-subunit to the DLGR1 receptor, thus activating G protein-mediated downstream signaling.
Similar to human GPA2/GPB5, the fly GPA2/GPB5 heterodimer is less stable than known glycoprotein hormones after SDS-PAGE (5). To allow efficient detection and purification of the recombinant protein, a FLAG-epitope was appended to the N terminus of fly GPA2. Consistent with the presence of a putative N-linked glycosylation site (124N), the fly GPA2 subunit migrated as a 16-kDa band under reducing conditions but was detected as a 12-kDa band after treatment with glycosidase F. In contrast, the 16-kDa band corresponding to the fly GPB5 subunit retained its migration pattern after treatment with glycosidase F. The existence of GPA2/GPB5 heterodimers was demonstrated by three approaches. First, cells overexpressing both GPA2 and GPB5, but not the individual subunit alone, secrete bioactive molecules capable of stimulating DLGR1. Second, immunoreactive fly GPB5 could be detected after FLAG M1 affinity chromatography against the FLAG-tagged GPA2. Third, cross-linking tests using purified GPA2/GPB5 after M1 affinity chromatography showed a single band of heterodimers.
In mammals, glycoprotein hormone subunits share a high degree of sequence similarity, 85% between LH? and CG? and approximately 30% among TSH?, FSH?, and LH?. Previous analysis of all ?-subunits of glycoproteins from chondrostean to human have demonstrated that ?-subunits in vertebrates can be separated into three groups composed of orthologs of the FSH/GTH1, LH/GTH2, and TSH clusters (20). It has been presumed that all four ?-subunits in human are derived from gene duplication during vertebrate evolution. Because only the newly discovered human GPA2 and human GPB5 have orthologs in invertebrates, it is likely that mammalian subunit genes are derived from ancestral GPA2 and GPB5 genes. We propose the following diagram of the molecular evolution of glycoprotein hormone ligands (Fig. 6). Fly and human are believed to branch apart almost 1 billion years ago when an ancestral GP -subunit already existed. In invertebrates, this gene evolved into GPA2 represented by fly GPA2 whereas the vertebrate GPA2 duplicated into the common -subunit and GPA2 found in human. Likewise, an ancestral GP ?-subunit existed in the common ancestor and was retained in fly as fly GPB5. In contrast, the vertebrate GPB5 duplicated several times to derive the five ?-subunit genes found in human.
FIG. 6. Diagrammatic drawing of the evolution of - and ?-subunit genes in the glycoprotein hormone family together with type A LGRs. Ancestral glycoprotein - (GP) and ?-subunit genes likely existed in the common ancestor of invertebrates and vertebrates. Unlike the presence of only one -subunit gene in invertebrates (represented by fly GPA2), gene duplications in vertebrates led to the derivation of the common -subunit and GPA2 as well as GPB5 and several ?-subunit genes. Although only one type A LGR is present in the fly genome, gene duplications in vertebrates led to the derivation of three glycoprotein hormone receptors.
Based on sequence homology and phylogenetic relatedness, receptors for the human glycoprotein hormones belong to the larger family of LGRs. Analysis of the completely sequenced human genome indicates the existence of five LGRs in addition to TSH, LH, and FSH receptors (6, 19, 21). In D. melanogaster, two LGRs (DLGR1 and DLGR2) have been reported (6, 8, 9), whereas only one LGR each has been reported in Caenorhabditis elegans (22), sea anemone, and snail (23, 24). Phylogenetic and functional analyses have led to the hypothesis that mammalian LGRs are classified into three subgroups (21). The type A group consists of gonadotropin and TSH receptors, whereas the type B group has the orphan receptors, LGR4, LGR5, and LGR6. The type C group consists of LGR7 and LGR8, the receptors for relaxin and INSL3, respectively (2, 25, 26).
In D. melanogaster, DLGR1 and DLGR2 belong to the type A and B groups, respectively. In addition to showing 50% amino acid sequence homology to mammalian type A receptors (7, 9), the fly DLGR1 exhibited constitutive activity when overexpressed in mammalian cells as exemplified by high basal levels of cAMP production (9). Of interest, the orthologous human TSH receptor also exhibited constitutive activity when overexpressed (27). We hypothesized that an ancestral type A LGR evolved before the divergence of invertebrates and vertebrates (Fig. 6). Although the vertebrates evolved three type A LGRs represented by human TSH, LH, and FSH receptors, there is only one type A LGR in fly. Because fly DLGR1 is likely the ortholog for the human TSH receptor that binds to the GPA2/GPB5 heterodimer and TSH, we further hypothesized that the candidate ligand for the fly receptor is the fly GPA2/GPB5 heterodimer. Indeed, the noncovalently linked heterodimeric fly GPA2/GPB5 is capable of activating DLGR1, similar to the ability of human GPA2/GPB5 (thyrostimulin) to activate human TSH receptors (5). Of interest, our recent study indicated that another heterodimeric cystine-knot hormone, bursicon, is the ligand for DLGR2 essential for tanning and eclosion in insects (10).
The physiological roles of fly GPA2/GPB5 and their receptor, DLGR1, are still unknown. Earlier Northern blot analyses indicated that expression of DLGR1 increased starting 8–16 h after oviposition and remained high until after pupation (7). Although DLGR1 expression decreased in adult female flies, high levels of the transcript were maintained in adult males. Coupled with the present observation of GPA2 and GPB5 expression in embryos, larvae, and adult flies, the findings suggest potential sex-related differences in these ligand-receptor pairs. Further studies on the tissue expression patterns and mutant phenotypes of fly GPA2, GPB5, and DLGR1 genes would reveal their functions in insects.
Due to the orthologous relationship between fly and human GPA2 and GPB5 genes, we generated a chimeric heterodimer formed by fly GPA2 and human GPB5. Of interest, this chimeric molecule was found to stimulate cAMP production by cells expressing human TSH receptors, albeit with a lower potency than human GPA2/GPB5. These results suggest that the fly GPA2 could serve as a scaffold to present not only fly GPB5 but also human GPB5 to specific receptors, thus underlying the extreme conservation of ligand and receptor genes in this family during evolution. These findings also are consistent with earlier observations showing that the human TSH receptors show promiscuous activation by the hCG-FSH chimeras (28). In contrast, our attempts to generate recombinant human GPA2 and fly GPB5 did not lead to bioactive molecules capable of activating the human TSH receptor or fly DLGR1 (data not shown).
Although fly GPA2/human GPB5 heterodimers activated the human TSH receptor, it could not activate fly DLRG1, suggesting the importance of the ?-subunit in ligand-receptor interactions. These findings are consistent with the well-established ability of the common -subunit to form heterodimers with different ?-subunits in the activation of different receptors. However, a recent study on the crystal structure of the complexes formed by the human FSH and the FSH receptor ectodomain indicated that four regions in the FSH molecule interact with its receptor (29). In addition to the seat belt region of the ?-subunit (?89–105 of the mature peptide) and the heel of the ?-subunit (?40–45 of the mature peptide), the heel and C-terminal segments of the -subunit also are important. Further mutagenesis studies using the present chimeric heterodimers may reveal the exact role of each subunit in receptor activation.
The known glycoprotein hormone heterodimers are believed to be stabilized by a segment of the ?-subunit, which wraps around the -subunit like a seat belt (17). Different from the TSH ?-subunit, both human and fly GPB5 subunits lack the C-tail region with a cysteine residue required for forming the seat belt structure (Fig. 1E). The lack of the unique C-tail region likely contributed to the instability of the GPA2 and GPB5 heterodimers. Although we could detect biological activities in the conditioned media of cells expressing the chimeric fly GPA2 and human GPB5, this activity was lost on elution of the chimeric molecules from the M1 affinity column, suggesting the chimeric molecule is more unstable than fly or human GPA2/GPB5 heterodimers.
In conclusion, we demonstrated that fly GPA2/GPB5 is a ligand for DLGR1. We also demonstrated the ability of the fly GPA2 subunit to form heterodimers with fly and human GPB5 subunits to activate fly DLGR1 and human TSH receptor, respectively. These studies provide a basis for future investigation of the interactions between glycoprotein hormones and their receptors and will allow further elucidation of the physiological roles of fly GPA2/GPB5 and DLGR1 genes in insects.
Acknowledgments
We thank Caren Spencer for editorial assistance.
References
Ji TH, Grossmann M, Ji I 1998 G protein-coupled receptors. I. Diversity of receptor-ligand interactions. J Biol Chem 273:17299–17302
Grossmann M, Weintraub BD, Szkudlinski MW 1997 Novel insights into the molecular mechanisms of human thyrotropin action: structural, physiological, and therapeutic implications for the glycoprotein hormone family. Endocr Rev 18:476–501
Isaacs NW 1995 Cystine knots. Curr Opin Struct Biol 5:391–395
Buckbinder L, Brown DD 1992 Thyroid hormone-induced gene expression changes in the developing frog limb. J Biol Chem 267:25786–25791
Nakabayashi K, Matsumi H, Bhalla A, Bae J, Mosselman S, Hsu SY, Hsueh AJ 2002 Thyrostimulin, a heterodimer of two new human glycoprotein hormone subunits, activates the thyroid-stimulating hormone receptor. J Clin Invest 109:1445–1452
Hsu SY, Nakabayashi K, Bhalla A 2002 Evolution of glycoprotein hormone subunit genes in bilateral metazoa: identification of two novel human glycoprotein hormone subunit family genes, GPA2 and GPB5. Mol Endocrinol 16:1538–1551
Hauser F, Nothacker HP, Grimmelikhuijzen CJ 1997 Molecular cloning, genomic organization, and developmental regulation of a novel receptor from Drosophila melanogaster structurally related to members of the thyroid-stimulating hormone, follicle-stimulating hormone, luteinizing hormone/choriogonadotropin receptor family from mammals. J Biol Chem 272:1002–1010
Eriksen KK, Hauser F, Schiott M, Pedersen KM, Sondergaard L, Grimmelikhuijzen CJ 2000 Molecular cloning, genomic organization, developmental regulation, and a knock-out mutant of a novel leu-rich repeats-containing G protein-coupled receptor (DLGR-2) from Drosophila melanogaster. Genome Res 10:924–938
Nishi S, Hsu SY, Zell K, Hsueh AJ 2000 Characterization of two fly LGR (leucine-rich repeat-containing, G protein-coupled receptor) proteins homologous to vertebrate glycoprotein hormone receptors: constitutive activation of wild-type fly LGR1 but not LGR2 in transfected mammalian cells. Endocrinology 141:4081–4090
Luo CW, Dewey EM, Sudo S, Ewer J, Hsu SY, Honegger HW, Hsueh AJ 2005 Bursicon, the insect cuticle-hardening hormone, is a heterodimeric cystine knot protein that activates G protein-coupled receptor LGR2. Proc Natl Acad Sci USA 102:2820–2825
Ben-Shlomo I, Yu Hsu S, Rauch R, Kowalski HW, Hsueh AJ 2003 Signaling receptome: a genomic and evolutionary perspective of plasma membrane receptors involved in signal transduction. Sci STKE RE9
Davoren JB, Hsueh AJ 1985 Vasoactive intestinal peptide: a novel stimulator of steroidogenesis by cultured rat granulosa cells. Biol Reprod 33:37–52
Hall CV, Jacob PE, Ringold GM, Lee F 1983 Expression and regulation of Escherichia coli lacZ gene fusions in mammalian cells. J Mol Appl Genet 2:101–109
Su JG, Hsueh AJ 1992 Characterization of mouse inhibin gene and its promoter. Biochem Biophys Res Commun 186:293–300
Vitt UA, Hsu SY, Hsueh AJ 2001 Evolution and classification of cystine knot-containing hormones and related extracellular signaling molecules. Mol Endocrinol 15:681–694
Erbel PJ, Karimi-Nejad Y, De Beer T, Boelens R, Kamerling JP, Vliegenthart JF 1999 Solution structure of the -subunit of human chorionic gonadotropin. Eur J Biochem 260:490–498
Grossmann M, Szkudlinski MW, Wong R, Dias JA, Ji TH, Weintraub BD 1997 Substitution of the seat-belt region of the thyroid-stimulating hormone (TSH) ?-subunit with the corresponding regions of choriogonadotropin or follitropin confers luteotropic but not follitropic activity to chimeric TSH. J Biol Chem 272:15532–15540
Lapthorn AJ, Harris DC, Littlejohn A, Lustbader JW, Canfield RE, Machin KJ, Morgan FJ, Isaacs NW 1994 Crystal structure of human chorionic gonadotropin. Nature 369:455–461
Hsu SY, Kudo M, Chen T, Nakabayashi K, Bhalla A, van der Spek PJ, van Duin M, Hsueh AJ 2000 The three subfamilies of leucine-rich repeat-containing G protein-coupled receptors (LGR): identification of LGR6 and LGR7 and the signaling mechanism for LGR7. Mol Endocrinol 14:1257–1271
Querat B, Sellouk A, Salmon C 2000 Phylogenetic analysis of the vertebrate glycoprotein hormone family including new sequences of sturgeon (Acipenser baeri) ? subunits of the two gonadotropins and the thyroid-stimulating hormone. Biol Reprod 63:222–228
Hsu SY, Nakabayashi K, Nishi S, Kumagai J, Kudo M, Sherwood OD, Hsueh AJ 2002 Activation of orphan receptors by the hormone relaxin. Science 295:671–674
Kudo M, Chen T, Nakabayashi K, Hsu SY, Hsueh AJ 2000 The nematode leucine-rich repeat-containing, G protein-coupled receptor (LGR) protein homologous to vertebrate gonadotropin and thyrotropin receptors is constitutively active in mammalian cells. Mol Endocrinol 14:272–284
Nothacker HP, Grimmelikhuijzen CJ 1993 Molecular cloning of a novel, putative G protein-coupled receptor from sea anemones structurally related to members of the FSH, TSH, LH/CG receptor family from mammals. Biochem Biophys Res Commun 197:1062–1069
Tensen CP, Van Kesteren ER, Planta RJ, Cox KJ, Burke JF, van Heerikhuizen H, Vreugdenhil E 1994 A G protein-coupled receptor with low density lipoprotein-binding motifs suggests a role for lipoproteins in G-linked signal transduction. Proc Natl Acad Sci USA 91:4816–4820
Kumagai J, Hsu SY, Matsumi H, Roh JS, Fu P, Wade JD, Bathgate RA, Hsueh AJ 2002 INSL3/Leydig insulin-like peptide activates the LGR8 receptor important in testis descent. J Biol Chem 277:31283–31286
Sudo S, Kumagai J, Nishi S, Layfield S, Ferraro T, Bathgate RA, Hsueh AJ 2003 H3 relaxin is a specific ligand for LGR7 and activates the receptor by interacting with both the ectodomain and the exoloop 2. J Biol Chem 278:7855–7862
Vassart G, Dumont JE 1992 The thyrotropin receptor and the regulation of thyrocyte function and growth. Endocr Rev 13:596–611
Campbell RK, Bergert ER, Wang Y, Morris JC, Moyle WR 1997 Chimeric proteins can exceed the sum of their parts: implications for evolution and protein design. Nat Biotechnol 15:439–443
Fan QR, Hendrickson WA 2005 Structure of human follicle-stimulating hormone in complex with its receptor. Nature 433:269–277
Mendive FM, Van Loy T, Claeysen S, Poels J, Williamson M, Hauser F, Grimmelikhuijzen CJ, Vassart G, Vanden Broeck J 2005 Drosophila molting neurohormone bursicon is a heterodimer and the natural agonist of the orphan receptor DLGR2. FEBS Lett 579:2171–2176(Satoko Sudo, Yoshimitsu K)
Address all correspondence and requests for reprints to: Aaron J. W. Hsueh, Stanford University School of Medicine, Department of Obstetrics and Gynecology, 300 Pasteur Drive, Room A344, Stanford, California 94305-5317. E-mail: aaron.hsueh@stanford.edu.
Abstract
Glycoprotein hormones play important roles in thyroid and gonadal function in vertebrates. The glycoprotein hormone -subunit forms heterodimers with different ?-subunits to activate TSH or gonadotropin (LH and FSH) receptors. Recent genomic analyses allowed the identification of another -subunit, GPA2, and another ?-subunit, GPB5, in human, capable of forming heterodimers to activate TSH receptors. Based on comparative genomic searches, we isolated the fly orthologs for human GPA2 and GPB5, each consisting of 10 cysteine residues likely involved in cystine-knot formation. RT-PCR analyses in Drosophila melanogaster demonstrated the expression of GPA2 and GPB5 at different developmental stages. Immunoblot analyses further showed that fly GPA2 and GPB5 subunit proteins are of approximately 16 kDa, and coexpression of these subunits yielded heterodimers. Purified recombinant fly GPA2/GPB5 heterodimers were found to be glycoproteins with N-linked glycosylated -subunits and nonglycosylated ?-subunits, capable of stimulating cAMP production mediated by fly orphan receptor DLGR1 but not DLGR2. Although the fly GPA2/GPB5 heterodimers did not activate human TSH or gonadotropin receptors, chimeric fly GPA2/human GPB5 heterodimers stimulated human TSH receptors. These findings indicated that fly GPA2/GPB5 is a ligand for DLGR1, thus showing the ancient origin of this glycoprotein hormone-seven transmembrane receptor-G protein signaling system. The fly GPA2 also could form heterodimers with human GPB5 to activate human TSH receptors, indicating the evolutionary conservation of these genes and suggesting that the GPA2 subunit may serve as a scaffold for the ?-subunit to activate downstream G protein-mediated signaling.
Introduction
GLYCOPROTEIN HORMONES are heterodimers consisting of - (GPA) and ?-subunits (GPB) assembled by noncovalent bonds. The cystine-knot-containing glycoprotein hormones are essential for gonadal and thyroid function in all vertebrates by binding to seven-transmembrane G protein-coupled receptors leading to the activation of downstream signal transduction (1, 2, 3). Gonadotropins regulate the growth and differentiation of the ovary and testis, whereas TSH is essential for energy balance. In amphibians, it is known that TSH induces metamorphosis in tadpoles (4).
Recently a new glycoprotein hormone, thyrostimulin, was discovered (5). It consists of two newly identified subunits, GPA2 and GPB5, with homology to the common -subunit and the TSH ?-subunit, respectively (6). This heterodimeric glycoprotein hormone is capable of stimulating TSH receptors by inducing cAMP production and thymidine incorporation in cultured thyroid cells and increasing serum thyroxine levels in TSH-suppressed rats in vivo. In Drosophila melanogaster, a glycoprotein ?-subunit, fly GPB5, has been reported showing high homology to the mammalian GPB5 subunit with 10 cysteine residues as compared with the other ?-subunit carrying 12 cysteine residues (6).
There are at least two leucine-rich repeat-containing G protein-coupled receptor (LGR) genes in D. melanogaster (7), DLGR1 and DLGR2 (8, 9). Recent studies demonstrated that DLGR2 is the receptor for the insect neurohormone, bursicon, essential for hardening and darkening of the new cuticle (10, 30). Here we characterized a novel glycoprotein -subunit in Drosophila melanogaster and named it as fly GPA2. We also identified orthologs for GPA2 and GPB5 in mosquito and other species. After the isolation of both fly GPA2 and fly GPB5 cDNAs, we found that fly GPA2 is capable of forming heterodimers with fly GPB5. Based on the hypothesized coevolution of ligand/receptor pairs (11), we demonstrated the ability of the fly GPA2/GPB5 heterodimer to stimulate cAMP production mediated by the orphan fly receptor, DLGR1, showing high homology to the mammalian TSH receptor. Furthermore, we demonstrated that the chimeric heterodimeric molecule, fly GPA2/human GPB5, is capable of stimulating human TSH receptors.
Materials and Methods
Hormones and reagents
DMEM/Ham’s F-12 (DMEM/F12) and Dulbecco’s PBS were obtained from Life Technologies (Gaithersburg, MD). Anti-FLAG M1 monoclonal antibodies and anti-FLAG M1 affinity gels were obtained from Sigma (St. Louis, MO). Anti-fly GPB5 antibodies were generated using a synthetic peptide encoding Ac-DSSEISDWKFPYKRSFHPC-amide corresponding to amino acids 78–95 of fly GPB5 (Biosource, Camarillo, CA). The antiserum titer was checked by ELISA using the synthetic peptide as the antigen. This fly GPB5 antibody does not cross-react with human GPB5, fly GPA2, or human GPA2 (data not shown). Antihuman GPB5 antibodies were described previously (5) and do not cross-react with fly GPA2 or GPB5 subunits. Human recombinant FSH (Org32489 and human choriogonadotropin (hCG) (CR-129) were obtained from Organon (Oss, The Netherlands) and the National Hormone and Pituitary Program (Torrance, CA), respectively. Conditioned media containing recombinant bursicon were collected from human 293T cells expressing burs/pburs heterodimers (10).
Identification and cloning of fly GPA2
To identify fly GPA2, the BLAST server at the National Center for Biotechnology Information (National Institutes of Health, Bethesda, MD) was used to search the expressed sequence tag (EST) database based on the amino acid sequence of human GPA2 leading to the identification of an Anopheles gambiae EST sequence (GenBank accession no. XM_317164). Using the deduced amino acid sequence of Anopheles gambiae GPA2, a sequence for the putative GPA2 subunit was found in the high throughput genomic sequence of D. melanogaster (GenBank accession no. AC017657).
RT-PCR
D. melanogaster poly A+ RNA (CLONTECH Laboratories, Inc., Palo Alto, CA) was used for RT-PCR. RNAs obtained from flies at different developmental stages were transcribed into cDNAs using Omniscript reverse transcriptase (QIAGEN, Valencia, CA) and oligo (dT)12–18 (Invitrogen, Carlsbad, CA). PCR was performed using the FastStart high-fidelity PCR system (Roche Applied Science, Indianapolis, IN) with 100 ng cDNA. PCR amplification of cDNA was carried out under the following conditions: 94 C for 30 sec for denaturation, 59 C for 30 sec for annealing, and 72 C for 45 sec for extension. The specific primers are: fly GPA2 (426-bp product), 5'-atgccaaagccatggccaatctca-3' and 5'-ctaatcctttttgcagtgataacaactac-3'; fly GPB5 (510-bp product), 5'-atgctcagaataatttttttcaggac-3' and 5'-ctaatagtccaagtttttggtatc-3'; ribosomal protein 49 (391-bp product); and 5'-atgaccatccgcccagcatac-3' and 5'-gagaacgcaggcgaccgttgg-3'. The PCR products were analyzed using electrophoresis of 1.5% agarose gels stained with ethidium bromide.
Generation of glycoprotein hormone subunit expression plasmids
RT-PCR products were used to generate expression constructs for fly GPA2 and fly GPB5 as described previously (5). All cDNAs were subcloned into the bipromoter expression vector pBudCE4.1 (Invitrogen) according to different combinations of subunit cDNAs. To allow efficient detection and purification of the recombinant proteins, a FLAG-epitope (DYKDDDDK) was added to the N terminus of fly GPA2. For generating fly GPA2 or fly GPB5 subunits alone, each cDNA was subcloned into pcDNA3.1 Zeo (Invitrogen). Fidelity of the RT-PCR products was confirmed by sequencing of the final constructs on both strands before use in expression studies.
Expression of recombinant proteins in 293T cells
Human 293T cells were maintained in DMEM/F12 supplemented with 10% fetal bovine serum (FBS). Cells were transfected using 7 μg of plasmid suspended in FUGENE6 (Roche Applied Science). Clonal cell lines stably expressing fly and human GPA2/GPB5 were selected under 500 μg/ml of Zeocin (Invitrogen) and maintained in DMEM/F12 containing 10% FBS and 100 μg/ml Zeocin. When the cells reached 90–100% confluency, the medium was replaced with DMEM/F12 without FBS or Zeocin. After 4 d of serum-free culture, the condition media were harvested and cell debris was cleared. The supernatant was concentrated with the iCON concentrator 20K MWCO (Pierce, Rockford, IL). The concentrated sample was purified using the anti-FLAG M1 affinity gel according to the manufacturer’s instructions (Sigma). Protein concentration of fly GPA2/GPB5 was determined by Bradford protein assay (Bio-Rad Laboratories, Hercules, CA). Purified tagged fly GPA2/GPB5 protein was used as a standard to estimate the concentration of chimeric recombinant fly GPA2/human GPB5 in the conditioned media using immunoblotting against the FLAG-epitope.
Immunoblotting analyses of recombinant proteins
Immunoblotting analyses were performed using anti-FLAG M1 antibodies (Sigma) or rabbit polyclonal antibodies against fly GPB5 (1:1000 dilution) or human GPB5 (1:1000 dilution). Samples were mixed with loading buffer under reducing conditions (5% ?-mercaptoethanol) before immunoblotting. After electrophoresis using 4–20% gradient polyacrylamide gels (Bio-Rad Laboratories), proteins were transferred to polyvinylidene difluoride membranes (Hybond-P, Amersham, Buckinghamshire, UK). Signals were detected after immunofluorescent imaging using the ECL system (Amersham).
To remove N-linked carbohydrate side chains, conditioned media were diluted in the deglycosylation buffer [50 mM sodium phosphate (pH 7.5), 1% Nonidet P-40, 0.5% sodium dodecyl sulfate, and 1% ?-mercaptoethanol] and incubated with 7.7 mIU of peptide N-glycosidase F (PNGase F; New England BioLabs, Beverly, MA) at 37 C for 1 h. Samples were mixed with loading buffer under reducing conditions (5% ?-mercaptoethanol) before immunoblotting. Purified fly GPA2/GPB5 protein was cross-linked using 0.27 mM disuccinimidyl suberate (Pierce) for 30 min, and reaction was terminated by addition of 1 M Tris-HCl (pH 7.4). Cross-linked complexes also were monitored using SDS-PAGE (4–20%) under reducing conditions.
Transfection of cells and analysis of signal transduction
Human 293T cells were maintained in DMEM/F12 supplemented with 10% FBS, 100 μg/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine. Before transfection, 2 x 105 cells were seeded in 12-well tissue culture plates (Corning, Corning, NY). When cells were 80–90% confluent, transient transfection was performed using 1 μg of plasmids encoding DLGR1 (9) or human TSH receptor cDNA with Lipofectamine 2000 (Invitrogen). After 24 h incubation with Lipofectamine, cells (2 x 105/ml) were placed on 24-well tissue culture plates (Corning, Corning NY) and preincubated at 37 C for 30 min in the presence of 0.25 mM 3-isobutyl-1-methyl xanthine (Sigma) before treatment with or without hormones for 16 h. At the end of incubation, cells and medium were frozen. After thawing, samples were heated to 95 C for 3 min to inactivate phosphodiesterase activity and total cAMP levels were measured in triplicate by a specific RIA as described previously (12). All experiments were repeated at least three times using cells from independent transfections. To monitor transfection efficiency, 0.5 μg of the ?-galactosidase plasmid (13) was routinely included in the transfection mixture, and ?-galactosidase activity in the cell lysate was measured as described (14).
Structural modeling
Comparative protein modeling was performed with the SWISS-MODEL server (http://swissmodel.expasy.org) using the experimentally determined structure for hCG as the template (Protein Data Bank, accession code 1HRP). DeepView-Swiss-PdbViewer and Protein Explorer (http://molvis.sdsc.edu/protexpl) were used to visualize the three-dimensional structure as well as to conduct structural comparisons between hCG and fly GPA2/GPB5.
Results
Isolation of fly GPA2 and fly GPB5 cDNAs and comparison with their invertebrate and vertebrate homologs
Based on the amino acid sequence of human GPA2, a homologous EST sequence of the Anopheles gambiae GPA2 was found in GenBank (accession no. XM_317164) using the BLAST search program. The open reading frame of Anopheles gambiae consisting of 121 amino acids was used for another BLAST search to identify a partial EST sequence of D. subobscura GPA2 (GenBank accession no. U73802). This stretch of amino acid sequence was used further to identify a genomic sequence of D. melanogaster using Mega-BLAST. Figure 1A shows nucleotide and predicted protein sequences of D. melanogaster GPA2 (GenBank accession no. AY940435).
FIG. 1. Isolation of GPA2 and GPB5 cDNAs from D. melanogaster and comparison with orthologs from invertebrates and vertebrates. A, Nucleotide and deduced protein sequences of the fly GPA2 subunit. Underlined sequences and the asterisk indicate the predicted signal peptide and a putative N-linked glycosylation site, respectively. B, Sequence alignment of GPA2 from fly, mosquito, nematode, human, mouse, rat, fugu, and zebrafish. Residues conserved in all eight species are shown in upper-case letters. Lower-case letters represent residues conserved in five of eight species. C, Nucleotide and deduced protein sequences of the fly GPB5 subunit. Underlined sequence represents the predicted signal peptide. D, Sequence alignment of GPB5 from fly, mosquito, nematode, human, mouse, rat, and tetraodon. Residues conserved in all seven species are shown in upper-case letters. Lower-case letters represent conserved residues in four of seven species. E, Comparison of cysteine signatures among fly GPA2, fly GPB5, human GPB5, and the human TSH?. Different pairs of cysteine residues forming disulfide bridges are connected by lines. Bold lines: disulfide bonds for the cystine knot structure; thin lines: additional disulfide bridges; dashed line: the latch disulfide bridge found in the seat belt structure. F, RT-PCR amplification of fly GPA2 and fly GPB5 mRNAs. Transcripts for GPA2 and GPB5 genes were amplified using PCR from cDNAs derived from D. melanogaster embryo, larva, and adults as described in Materials and Methods. Ribosomal protein 49 (RP49) expression served as an internal control. (Figure continues on next page.)
FIG. 1A. Continued
To confirm the predicted sequence of fly GPA2 cDNA, RT-PCR was carried out using gene-specific primers from templates derived from adult D. melanogaster RNA. Comparison of this cDNA with the corresponding genomic sequence revealed that the fly GPA2 gene has two exons divided by a 58-bp intron. Deduced fly GPA2 consists of 141 amino acid residues with a 31-amino acid-long signal peptide (Fig. 1A). Similar to its human counterpart, the mature fly GPA2 has 10 cysteine residues capable of forming a cystine-knot structure (15). There is also a potential N-linked glycosylation site at 124N. Sequence alignment of GPA2 from fly, mosquito, nematode, fish, and other species (Fig. 1B) indicated a complete conservation of cysteine residues in the predicted mature proteins. Although the fly GPB5 cDNA has already been predicted (6), its expression has not been verified. To confirm the expression of fly GPB5, RT-PCR was performed using gene-specific primers to amplify transcripts from adult D. melanogaster RNA. The deduced amino acid sequence of fly GPB5 cDNA consists of a signal peptide of 25-amino acid residues and a mature region with 144 amino acids but without a putative N-linked glycosylation site (Fig. 1C). Sequence alignment of GPB5 molecules from fly, mosquito, nematode, and other species (Fig. 1D) indicated the conservation of key cysteine residues for cystine-knot formation; however, the fly and mosquito GPB5 molecules have an extended C tail.
Predicted fly GPA2 and fly GPB5 subunits show 22 and 19% homology in amino acid sequences in the mature protein as compared with human GPA2 and human GPB5, respectively. To characterize the putative structure of fly GPA2 and fly GPB5, tertiary structural models were generated by comparative protein modeling using the crystal structure of - and ?-subunits of hCG as templates (FUGUE Profile Lib Search for fly GPA2 and 3D-JIGSAW for fly GPB5) (16). As shown in Fig. 1E, the predicted cystine-knot structure of fly GPA2 is comprised of three disulfide bridges (44C-104C, 68C-133C, 72C-135C) and two additional disulfide bonds (58C-118C and 103C-138C) outside the cystine knot. Likewise, fly GPB5 has a cystine-knot structure consisting of three disulfide bridges (45C-97C, 73C-131C, and 77C-133C) and two additional disulfide bonds (64C-112C and 136C-143C). Different from human GPB5 (6), fly GPB5 has an extended C-terminal tail corresponding to the seat belt structure of the mammalian ?-subunits. However, the corresponding C-tail region of TSH?, truncated in human GPB5, has one additional cysteine residue (125C) forming an additional disulfide bridge with the 39C residue of TSH? to form the seat belt structure presumably holding the -subunit tightly in the hole (17, 18).
Expression of fly GPA2 and fly GPB5 transcripts during different developmental stages of D. melanogaster
To elucidate the expression of fly GPA2 and fly GPB5 transcripts during development, RT-PCR was performed using cDNAs from different developmental stages of D. melanogaster. As shown in Fig. 1F, both GPA2 and GPB5 were amplified as DNA fragments of predicted sizes from RNA obtained from flies at embryonic, larval, and adult stages.
Immunoblotting analyses of purified recombinant fly GPA2/GPB5 heterodimers
We hypothesized that fly GPA2 and fly GPB5, similar to their human counterparts, could form heterodimers. 293T cells were transfected with a bipromoter expression vector encoding both FLAG-tagged fly GPA2 and nontagged fly GPB5. Conditioned media expressing fly GPA2/GPB5 was affinity purified using anti-FLAG M1 affinity gels. In Fig. 2A, Coomassie blue staining analyses indicates an abundance of proteins of 66 kDa in the conditioned media, likely representing human albumin secreted by 293T cells. In contrast, a single band of approximately 16 kDa was found after affinity purification of the tagged GPA2/GPB5 heterodimer. To further characterize the purified recombinant proteins, immunoblotting analyses were performed. Under reducing conditions, purified fly GPA2/GPB5 proteins were detected as bands of approximately 16 kDa using the anti-FLAG M1 antibody against the FLAG-tagged fly GPA2 subunit or the anti-fly GPB5 antibody against GPB5 (Fig. 2B). After treatment with N-glycosidase, a 12-kDa band was detected using the anti-FLAG M1 antibody corresponding to the predicted size of the deglycosylated form of fly GPA2. This finding is consistent with the presence of a putative N-glycosylation site at 124N of fly GPA2. In contrast, the migration of the fly GPB5 subunit was not affected by N-glycosidase treatment, indicating this subunit is not N-glycosylated. To further demonstrate the heterodimeric nature of purified GPA2/GPB5 proteins, we performed cross-linking analyses using disuccinimidyl suberate. After cross-linking of the purified fly GPA2/GPB5, immunoblotting using either the anti-FLAG or the anti-GPB5 antibodies revealed the presence of high-molecular-mass complexes of approximately 32 kDa, corresponding to the size of the predicted heterodimer (Fig. 2C).
FIG. 2. Characterization of recombinant fly GPA2/GPB5 heterodimers. A, Purification of fly GPA2/GPB5 using M1 affinity chromatography. Coomassie blue staining of conditioned media (CM; left lane) and M1 affinity-purified fly GPA2/GPB5 heterodimers (right lane). B, Immunoblotting analyses of purified fly GPA2/GPB5 appended at the N terminus of GPA2 with the FLAG epitope. Left, Detection using anti-FLAG M1 antibodies; right, detection using antibodies against fly GPB5. Some samples were pretreated with N-glycosidase F to remove N-linked carbohydrate side chains. C, Cross-linking of purified fly GPA2/GPB5 using disuccinimidyl suberate. D, Tertiary structure comparison between the fly GPA2/GPB5 heterodimer and hCG. Structural model for the fly GPA2/GPB5 heterodimer was generated based on the hCG crystal structure template. To optimize the modeling of the tertiary structure of fly GPA2/GPB5, the histidine residue in position 137 of fly GPA2 was replaced by the corresponding threonine residue in position 110 of the human common -subunit. Views from two different angles (view A with the -subunit in the foreground; view B with the ?-subunit in the foreground) are presented. The -subunits are indicated by the white color, whereas the ?-subunits are in green. Cysteine residues important for the formation of structure-determining cystine bonds in the - and ?-subunits are denoted in brown and purple, respectively. The seat belt structure formed by the C terminus of the hCG? is circled for comparison with the corresponding region in the fly molecule (dashed circle).
Based on the crystal structure of hCG, we generated a structural model for the fly GPA2/GPB5 heterodimer. As shown in Fig. 2D, the fly GPA2/GPB5 likely forms a heterodimer similar to hCG. Although the fly GPB5 lacks the last cysteine residue in its C terminus for forming the S-S bond found in the seat belt region of CG-? (Fig. 2D, circled), a similar structure could still be modeled in the fly GPA2/GPB5 heterodimer (Fig. 2D, dashed circle).
Fly DLGR1 is activated by recombinant fly GPA2/GPB5 heterodimers
Based on the hypothesis that ligand-receptor pairs coevolved during evolution, the fly GPA2/GPB5 heterodimer is likely the ligand for fly DLGR1 with closest sequence homology to type A LGRs including mammalian TSH, LH, and FSH receptors (19). We tested the ability of this recombinant protein to activate DLGR1. 293T cells were transiently transfected with the DLGR1 expression plasmid and ligand signaling was estimated after treatment with purified fly GPA2/GPB5. Figure 3A shows in cells expressing DLGR1, treatment with fly GPA2/GPB5 led to dose-dependent increases in total cAMP production with an EC50 value of 3.2 nM. In contrast, conditioned media expressing fly GPA2 or fly GPB5 alone did not stimulate cAMP production. Furthermore, neither fly GPA2/GPB5 heterodimers nor the individual subunits alone were capable of activating DLGR2. In contrast, DLGR2 responded to treatment of 50 nM bursicon with a major increase in cAMP production (Fig. 3B) (10).
FIG. 3. Activation of fly DLGR1 by recombinant fly GPA2/GPB5 heterodimers. A, Dose-dependent effects of purified fly GPA2/GPB5 to stimulate cAMP production by 293T cells expressing DLGR1. 293T cells were treated with purified fly GPA2/GPB5 or conditioned media containing fly GPA2 or fly GPB5. B, Lack of activation of fly DLGR2 by fly GPA2/GPB5 heterodimers. 293T cells transiently expressing DLGR2 were treated with purified fly GPA2/GPB5, individual subunits, or recombinant bursicon, all at 50 nM. Mean ± SD (n = 3).
Chimeric fly GPA2/human GPB5 heterodimers are capable of activating human TSH receptors but not fly DLGR1
We further tested the ability of fly GPA2/GPB5 heterodimers to activate human glycoprotein hormone receptors. As shown in Fig. 4, A–C, neither fly GPA2/GPB5 heterodimers nor the individual fly subunit alone stimulated cAMP production in cells expressing human TSH, LH, or FSH receptors. Because GPA2 genes showed high sequence conservation between fly and human orthologs, we hypothesized that chimeric heterodimers consisting of fly and human subunits could be formed. We constructed expression vectors encoding fly GPA2 and human GPB5 and expressed the recombinant heterodimers. After affinity purification based on the FLAG tag appended to fly GPA2, the heterodimers showed immunoreactive bands at 16 kDa corresponding to each subunit detected by anti-FLAG M1 antibodies and human GPB5 antibodies, respectively (Fig. 5A). After treatment with glycosidase F, both subunits showed a decrease in their size to 12 kDa as detected using anti-FLAG and antihuman GPB5 antibodies, respectively. These results are consistent with the presence of the potential N-linked glycosylation sites in fly GPA2 and at 63N in the human GPB5 subunit (5).
FIG. 4. Fly GPA2/GPB5 heterodimers are not capable of stimulating human TSH or gonadotropin receptors. Cells expressing human TSH receptors (A), human LH receptors (B), and human FSH receptors (C) were incubated with 10 nM of fly GPA2/GPB5, fly GPA2, or fly GPB5. For positive controls, cells were treated with 10 nM human GPA2/GPB5 (thyrostimulin), 10 ng/ml of hCG, or 100 mIU/ml of human FSH. Mean ± SD (n = 3).
FIG. 5. Chimeric fly GPA2/human GPB5 heterodimers are capable of activating human TSH receptors. A, Characterization of chimeric fly GPA2/human GPB5 heterodimers using immunoblots. Conditioned media containing fly GPA2 (appended at N terminus with the FLAG-epitope) and human GPB5 were subjected to M1 affinity chromatography before immunoblotting analyses. Left, Detection using anti-FLAG M1 antibodies; right, detection using antibodies against human GPB5. Some samples were pretreated with N-glycosidase F to remove N-linked carbohydrate side chains. B, Stimulation of human TSH receptors by fly GPA2/human GPB5 heterodimers and purified human GPA2/GPB5 (thyrostimulin). The concentration of fly GPA2/human GPB5 in the conditioned media was estimated by immunoblotting analysis as described in Materials and Methods. Mean ± SD (n = 3). C, Lack of stimulation of cAMP production by DLGR1-expressing cells treated with chimeric fly GPA2/human GPB5 and human GPA2/GPB5 (thyrostimulin). Mean ± SD (n = 3).
We further tested the ability of the chimeric heterodimer molecule, fly GPA2/human GPB5, to stimulate cAMP production in cells expressing human TSH receptors. As shown in Fig. 5B, conditioned media containing chimeric fly GPA2/human GPB5 heterodimers stimulated cAMP production in cells expressing human TSH receptors. However, the potency of the chimeric molecules is much lower than human GPA2/GPB5 (thyrostimulin) with EC50 values of 9.8 nM and 3.0 pM, respectively. In contrast, treatment with 100 nM of fly GPA2 or human GPB5 alone did not alter cAMP production (data not shown). These results indicated that, despite a separation of about 1 billion years during evolution, the fly GPA2 subunit retains the ability to heterodimerize with human GPB5 for activation of the human TSH receptors. In contrast, neither the chimeric fly GPA2/human GPB5 heterodimer nor human GPA2/GPB5 was capable of stimulating cAMP production by cells expressing DLGR1 (Fig. 5C).
Discussion
Recent genomic analyses allowed the identification of the new -subunit GPA2 and ?-subunit GPB5 in human capable of forming heterodimers to activate TSH receptors (5, 6). Taking advantage of comparative genomic searches using human sequences for GPA2 and GPB5, we characterized the fly orthologs for human GPA2 and GPB5 consisting of 10 cysteine residues capable of forming a cystine-knot structure. RT-PCR results demonstrated the existence of the transcripts for these genes in different developmental stages in fly. Furthermore, two fly glycoprotein hormone subunits formed a heterodimeric glycoprotein hormone capable of activating the DLGR1 receptor. In addition, we demonstrated that fly GPA2 also could form heterodimers with human GPB5 to activate the human TSH receptor, albeit with a lower potency as compared with human GPA2/GPB5 (thyrostimulin). These findings indicate the evolutionary conservation of these genes and suggest that the GPA2 subunit could serve as a scaffold to present the ?-subunit to the DLGR1 receptor, thus activating G protein-mediated downstream signaling.
Similar to human GPA2/GPB5, the fly GPA2/GPB5 heterodimer is less stable than known glycoprotein hormones after SDS-PAGE (5). To allow efficient detection and purification of the recombinant protein, a FLAG-epitope was appended to the N terminus of fly GPA2. Consistent with the presence of a putative N-linked glycosylation site (124N), the fly GPA2 subunit migrated as a 16-kDa band under reducing conditions but was detected as a 12-kDa band after treatment with glycosidase F. In contrast, the 16-kDa band corresponding to the fly GPB5 subunit retained its migration pattern after treatment with glycosidase F. The existence of GPA2/GPB5 heterodimers was demonstrated by three approaches. First, cells overexpressing both GPA2 and GPB5, but not the individual subunit alone, secrete bioactive molecules capable of stimulating DLGR1. Second, immunoreactive fly GPB5 could be detected after FLAG M1 affinity chromatography against the FLAG-tagged GPA2. Third, cross-linking tests using purified GPA2/GPB5 after M1 affinity chromatography showed a single band of heterodimers.
In mammals, glycoprotein hormone subunits share a high degree of sequence similarity, 85% between LH? and CG? and approximately 30% among TSH?, FSH?, and LH?. Previous analysis of all ?-subunits of glycoproteins from chondrostean to human have demonstrated that ?-subunits in vertebrates can be separated into three groups composed of orthologs of the FSH/GTH1, LH/GTH2, and TSH clusters (20). It has been presumed that all four ?-subunits in human are derived from gene duplication during vertebrate evolution. Because only the newly discovered human GPA2 and human GPB5 have orthologs in invertebrates, it is likely that mammalian subunit genes are derived from ancestral GPA2 and GPB5 genes. We propose the following diagram of the molecular evolution of glycoprotein hormone ligands (Fig. 6). Fly and human are believed to branch apart almost 1 billion years ago when an ancestral GP -subunit already existed. In invertebrates, this gene evolved into GPA2 represented by fly GPA2 whereas the vertebrate GPA2 duplicated into the common -subunit and GPA2 found in human. Likewise, an ancestral GP ?-subunit existed in the common ancestor and was retained in fly as fly GPB5. In contrast, the vertebrate GPB5 duplicated several times to derive the five ?-subunit genes found in human.
FIG. 6. Diagrammatic drawing of the evolution of - and ?-subunit genes in the glycoprotein hormone family together with type A LGRs. Ancestral glycoprotein - (GP) and ?-subunit genes likely existed in the common ancestor of invertebrates and vertebrates. Unlike the presence of only one -subunit gene in invertebrates (represented by fly GPA2), gene duplications in vertebrates led to the derivation of the common -subunit and GPA2 as well as GPB5 and several ?-subunit genes. Although only one type A LGR is present in the fly genome, gene duplications in vertebrates led to the derivation of three glycoprotein hormone receptors.
Based on sequence homology and phylogenetic relatedness, receptors for the human glycoprotein hormones belong to the larger family of LGRs. Analysis of the completely sequenced human genome indicates the existence of five LGRs in addition to TSH, LH, and FSH receptors (6, 19, 21). In D. melanogaster, two LGRs (DLGR1 and DLGR2) have been reported (6, 8, 9), whereas only one LGR each has been reported in Caenorhabditis elegans (22), sea anemone, and snail (23, 24). Phylogenetic and functional analyses have led to the hypothesis that mammalian LGRs are classified into three subgroups (21). The type A group consists of gonadotropin and TSH receptors, whereas the type B group has the orphan receptors, LGR4, LGR5, and LGR6. The type C group consists of LGR7 and LGR8, the receptors for relaxin and INSL3, respectively (2, 25, 26).
In D. melanogaster, DLGR1 and DLGR2 belong to the type A and B groups, respectively. In addition to showing 50% amino acid sequence homology to mammalian type A receptors (7, 9), the fly DLGR1 exhibited constitutive activity when overexpressed in mammalian cells as exemplified by high basal levels of cAMP production (9). Of interest, the orthologous human TSH receptor also exhibited constitutive activity when overexpressed (27). We hypothesized that an ancestral type A LGR evolved before the divergence of invertebrates and vertebrates (Fig. 6). Although the vertebrates evolved three type A LGRs represented by human TSH, LH, and FSH receptors, there is only one type A LGR in fly. Because fly DLGR1 is likely the ortholog for the human TSH receptor that binds to the GPA2/GPB5 heterodimer and TSH, we further hypothesized that the candidate ligand for the fly receptor is the fly GPA2/GPB5 heterodimer. Indeed, the noncovalently linked heterodimeric fly GPA2/GPB5 is capable of activating DLGR1, similar to the ability of human GPA2/GPB5 (thyrostimulin) to activate human TSH receptors (5). Of interest, our recent study indicated that another heterodimeric cystine-knot hormone, bursicon, is the ligand for DLGR2 essential for tanning and eclosion in insects (10).
The physiological roles of fly GPA2/GPB5 and their receptor, DLGR1, are still unknown. Earlier Northern blot analyses indicated that expression of DLGR1 increased starting 8–16 h after oviposition and remained high until after pupation (7). Although DLGR1 expression decreased in adult female flies, high levels of the transcript were maintained in adult males. Coupled with the present observation of GPA2 and GPB5 expression in embryos, larvae, and adult flies, the findings suggest potential sex-related differences in these ligand-receptor pairs. Further studies on the tissue expression patterns and mutant phenotypes of fly GPA2, GPB5, and DLGR1 genes would reveal their functions in insects.
Due to the orthologous relationship between fly and human GPA2 and GPB5 genes, we generated a chimeric heterodimer formed by fly GPA2 and human GPB5. Of interest, this chimeric molecule was found to stimulate cAMP production by cells expressing human TSH receptors, albeit with a lower potency than human GPA2/GPB5. These results suggest that the fly GPA2 could serve as a scaffold to present not only fly GPB5 but also human GPB5 to specific receptors, thus underlying the extreme conservation of ligand and receptor genes in this family during evolution. These findings also are consistent with earlier observations showing that the human TSH receptors show promiscuous activation by the hCG-FSH chimeras (28). In contrast, our attempts to generate recombinant human GPA2 and fly GPB5 did not lead to bioactive molecules capable of activating the human TSH receptor or fly DLGR1 (data not shown).
Although fly GPA2/human GPB5 heterodimers activated the human TSH receptor, it could not activate fly DLRG1, suggesting the importance of the ?-subunit in ligand-receptor interactions. These findings are consistent with the well-established ability of the common -subunit to form heterodimers with different ?-subunits in the activation of different receptors. However, a recent study on the crystal structure of the complexes formed by the human FSH and the FSH receptor ectodomain indicated that four regions in the FSH molecule interact with its receptor (29). In addition to the seat belt region of the ?-subunit (?89–105 of the mature peptide) and the heel of the ?-subunit (?40–45 of the mature peptide), the heel and C-terminal segments of the -subunit also are important. Further mutagenesis studies using the present chimeric heterodimers may reveal the exact role of each subunit in receptor activation.
The known glycoprotein hormone heterodimers are believed to be stabilized by a segment of the ?-subunit, which wraps around the -subunit like a seat belt (17). Different from the TSH ?-subunit, both human and fly GPB5 subunits lack the C-tail region with a cysteine residue required for forming the seat belt structure (Fig. 1E). The lack of the unique C-tail region likely contributed to the instability of the GPA2 and GPB5 heterodimers. Although we could detect biological activities in the conditioned media of cells expressing the chimeric fly GPA2 and human GPB5, this activity was lost on elution of the chimeric molecules from the M1 affinity column, suggesting the chimeric molecule is more unstable than fly or human GPA2/GPB5 heterodimers.
In conclusion, we demonstrated that fly GPA2/GPB5 is a ligand for DLGR1. We also demonstrated the ability of the fly GPA2 subunit to form heterodimers with fly and human GPB5 subunits to activate fly DLGR1 and human TSH receptor, respectively. These studies provide a basis for future investigation of the interactions between glycoprotein hormones and their receptors and will allow further elucidation of the physiological roles of fly GPA2/GPB5 and DLGR1 genes in insects.
Acknowledgments
We thank Caren Spencer for editorial assistance.
References
Ji TH, Grossmann M, Ji I 1998 G protein-coupled receptors. I. Diversity of receptor-ligand interactions. J Biol Chem 273:17299–17302
Grossmann M, Weintraub BD, Szkudlinski MW 1997 Novel insights into the molecular mechanisms of human thyrotropin action: structural, physiological, and therapeutic implications for the glycoprotein hormone family. Endocr Rev 18:476–501
Isaacs NW 1995 Cystine knots. Curr Opin Struct Biol 5:391–395
Buckbinder L, Brown DD 1992 Thyroid hormone-induced gene expression changes in the developing frog limb. J Biol Chem 267:25786–25791
Nakabayashi K, Matsumi H, Bhalla A, Bae J, Mosselman S, Hsu SY, Hsueh AJ 2002 Thyrostimulin, a heterodimer of two new human glycoprotein hormone subunits, activates the thyroid-stimulating hormone receptor. J Clin Invest 109:1445–1452
Hsu SY, Nakabayashi K, Bhalla A 2002 Evolution of glycoprotein hormone subunit genes in bilateral metazoa: identification of two novel human glycoprotein hormone subunit family genes, GPA2 and GPB5. Mol Endocrinol 16:1538–1551
Hauser F, Nothacker HP, Grimmelikhuijzen CJ 1997 Molecular cloning, genomic organization, and developmental regulation of a novel receptor from Drosophila melanogaster structurally related to members of the thyroid-stimulating hormone, follicle-stimulating hormone, luteinizing hormone/choriogonadotropin receptor family from mammals. J Biol Chem 272:1002–1010
Eriksen KK, Hauser F, Schiott M, Pedersen KM, Sondergaard L, Grimmelikhuijzen CJ 2000 Molecular cloning, genomic organization, developmental regulation, and a knock-out mutant of a novel leu-rich repeats-containing G protein-coupled receptor (DLGR-2) from Drosophila melanogaster. Genome Res 10:924–938
Nishi S, Hsu SY, Zell K, Hsueh AJ 2000 Characterization of two fly LGR (leucine-rich repeat-containing, G protein-coupled receptor) proteins homologous to vertebrate glycoprotein hormone receptors: constitutive activation of wild-type fly LGR1 but not LGR2 in transfected mammalian cells. Endocrinology 141:4081–4090
Luo CW, Dewey EM, Sudo S, Ewer J, Hsu SY, Honegger HW, Hsueh AJ 2005 Bursicon, the insect cuticle-hardening hormone, is a heterodimeric cystine knot protein that activates G protein-coupled receptor LGR2. Proc Natl Acad Sci USA 102:2820–2825
Ben-Shlomo I, Yu Hsu S, Rauch R, Kowalski HW, Hsueh AJ 2003 Signaling receptome: a genomic and evolutionary perspective of plasma membrane receptors involved in signal transduction. Sci STKE RE9
Davoren JB, Hsueh AJ 1985 Vasoactive intestinal peptide: a novel stimulator of steroidogenesis by cultured rat granulosa cells. Biol Reprod 33:37–52
Hall CV, Jacob PE, Ringold GM, Lee F 1983 Expression and regulation of Escherichia coli lacZ gene fusions in mammalian cells. J Mol Appl Genet 2:101–109
Su JG, Hsueh AJ 1992 Characterization of mouse inhibin gene and its promoter. Biochem Biophys Res Commun 186:293–300
Vitt UA, Hsu SY, Hsueh AJ 2001 Evolution and classification of cystine knot-containing hormones and related extracellular signaling molecules. Mol Endocrinol 15:681–694
Erbel PJ, Karimi-Nejad Y, De Beer T, Boelens R, Kamerling JP, Vliegenthart JF 1999 Solution structure of the -subunit of human chorionic gonadotropin. Eur J Biochem 260:490–498
Grossmann M, Szkudlinski MW, Wong R, Dias JA, Ji TH, Weintraub BD 1997 Substitution of the seat-belt region of the thyroid-stimulating hormone (TSH) ?-subunit with the corresponding regions of choriogonadotropin or follitropin confers luteotropic but not follitropic activity to chimeric TSH. J Biol Chem 272:15532–15540
Lapthorn AJ, Harris DC, Littlejohn A, Lustbader JW, Canfield RE, Machin KJ, Morgan FJ, Isaacs NW 1994 Crystal structure of human chorionic gonadotropin. Nature 369:455–461
Hsu SY, Kudo M, Chen T, Nakabayashi K, Bhalla A, van der Spek PJ, van Duin M, Hsueh AJ 2000 The three subfamilies of leucine-rich repeat-containing G protein-coupled receptors (LGR): identification of LGR6 and LGR7 and the signaling mechanism for LGR7. Mol Endocrinol 14:1257–1271
Querat B, Sellouk A, Salmon C 2000 Phylogenetic analysis of the vertebrate glycoprotein hormone family including new sequences of sturgeon (Acipenser baeri) ? subunits of the two gonadotropins and the thyroid-stimulating hormone. Biol Reprod 63:222–228
Hsu SY, Nakabayashi K, Nishi S, Kumagai J, Kudo M, Sherwood OD, Hsueh AJ 2002 Activation of orphan receptors by the hormone relaxin. Science 295:671–674
Kudo M, Chen T, Nakabayashi K, Hsu SY, Hsueh AJ 2000 The nematode leucine-rich repeat-containing, G protein-coupled receptor (LGR) protein homologous to vertebrate gonadotropin and thyrotropin receptors is constitutively active in mammalian cells. Mol Endocrinol 14:272–284
Nothacker HP, Grimmelikhuijzen CJ 1993 Molecular cloning of a novel, putative G protein-coupled receptor from sea anemones structurally related to members of the FSH, TSH, LH/CG receptor family from mammals. Biochem Biophys Res Commun 197:1062–1069
Tensen CP, Van Kesteren ER, Planta RJ, Cox KJ, Burke JF, van Heerikhuizen H, Vreugdenhil E 1994 A G protein-coupled receptor with low density lipoprotein-binding motifs suggests a role for lipoproteins in G-linked signal transduction. Proc Natl Acad Sci USA 91:4816–4820
Kumagai J, Hsu SY, Matsumi H, Roh JS, Fu P, Wade JD, Bathgate RA, Hsueh AJ 2002 INSL3/Leydig insulin-like peptide activates the LGR8 receptor important in testis descent. J Biol Chem 277:31283–31286
Sudo S, Kumagai J, Nishi S, Layfield S, Ferraro T, Bathgate RA, Hsueh AJ 2003 H3 relaxin is a specific ligand for LGR7 and activates the receptor by interacting with both the ectodomain and the exoloop 2. J Biol Chem 278:7855–7862
Vassart G, Dumont JE 1992 The thyrotropin receptor and the regulation of thyrocyte function and growth. Endocr Rev 13:596–611
Campbell RK, Bergert ER, Wang Y, Morris JC, Moyle WR 1997 Chimeric proteins can exceed the sum of their parts: implications for evolution and protein design. Nat Biotechnol 15:439–443
Fan QR, Hendrickson WA 2005 Structure of human follicle-stimulating hormone in complex with its receptor. Nature 433:269–277
Mendive FM, Van Loy T, Claeysen S, Poels J, Williamson M, Hauser F, Grimmelikhuijzen CJ, Vassart G, Vanden Broeck J 2005 Drosophila molting neurohormone bursicon is a heterodimer and the natural agonist of the orphan receptor DLGR2. FEBS Lett 579:2171–2176(Satoko Sudo, Yoshimitsu K)