Report

Activation of Orphan Receptors by the Hormone Relaxin

See allHide authors and affiliations

Science  25 Jan 2002:
Vol. 295, Issue 5555, pp. 671-674
DOI: 10.1126/science.1065654

Abstract

Relaxin is a hormone important for the growth and remodeling of reproductive and other tissues during pregnancy. Although binding sites for relaxin are widely distributed, the nature of its receptor has been elusive. Here, we demonstrate that two orphan heterotrimeric guanine nucleotide binding protein (G protein)–coupled receptors, LGR7 and LGR8, are capable of mediating the action of relaxin through an adenosine 3′,5′-monophosphate (cAMP)–dependent pathway distinct from that of the structurally related insulin and insulin-like growth factor family ligand. Treatment of antepartum mice with the soluble ligand-binding region of LGR7 caused parturition delay. The wide and divergent distribution of the two relaxin receptors implicates their roles in reproductive, brain, renal, cardiovascular, and other functions.

Relaxin has diverse actions in the reproductive tract and other tissues during pregnancy (1). These actions include promotion of growth and dilation of the cervix, growth and quiescence of the uterus, growth and development of the mammary gland and nipple, and regulation of cardiovascular function. Although binding sites for relaxin have been found in reproductive tissues (2), brain (3), and heart (4), the nature of the relaxin receptor has not been determined. Prorelaxin, the precursor form of relaxin, has a domain arrangement similar to that of insulin and insulin-like growth factor (IGF) precursors, and several relaxin and insulin-related genes have been identified, including those encoding INSL3 (or Leydig cell relaxin), INSL4, INSL5, INSL6, and relaxin 3 (5–7). The abnormal testis descent phenotype of INSL3-null mice (5, 8) is similar to that of mice with a disruption of a G protein–coupled receptor (GPCR) encoded by the GREAT gene (9); this finding suggests that relaxin-related proteins may be ligands for GPCRs. Indeed, relaxin stimulates cAMP production in endometrial, anterior pituitary, and other cells (1), an event mediated by GPCRs.

The orphan leucine-rich repeat–containing GPCRs (LGRs) designated as LGR4 through LGR7 are structurally similar to the LGRs for gonadotropins and thyrotropin (10, 11). LGR7 can be distinguished from the other three orphan LGRs on the basis of structural motifs and phylogenetic analysis. Moreover, LGR7 likely couples to Gs proteins, because constitutively active LGR7 mutants show ligand-independent cAMP production (11). We screened the human genome with LGR7 in search of novel paralogs, and isolated LGR8 (12, 13). LGR8 proved to be the human ortholog of the mouse GREAT GPCR and shares about 60% sequence identity with LGR7 (13). Phylogenetic analysis also showed that LGR8 is closely related to an orthologous Drosophila receptor, DmLGR3, and to the snail LGR (13). Similar to the gain-of-function mutants identified for LGR7 (Asp637 → Tyr) (11) and the luteinizing hormone (LH) receptor (Asp578 → Tyr), an LGR8 mutant with the same amino acid substitution in transmembrane helix VI conferred a ligand-independent increase in basal cAMP production in transfected human fetal kidney 293T cells (13). Thus, LGR7 and LGR8 likely signal through the adenylate cyclase pathway.

Treatment of transfected cells expressing either LGR7 or LGR8 with porcine relaxin resulted in a dose-dependent increase in cAMP production, with median effective concentrations of 1.5 and 5.0 nM, respectively (Fig. 1). In contrast, treatment with structural homologs (insulin, IGF-I, or IGF-II) or with an unrelated peptide (glucagon) was ineffective. These findings indicate that relaxin is a cognate ligand for these two orphan GPCRs and that it activates adenylate cyclases through Gs proteins.

Figure 1

LGR7 and LGR8 are relaxin receptors. Porcine relaxin stimulated dose-dependent cAMP production in transfected 293T cells (105 cells per culture) expressing LGR7 (A) or LGR8 (B) using the pcDNA3.1-Zeo expression vector (11). In contrast, treatment with insulin, IGF-1, IGF-II, or glucagon (a known Gs-coupled receptor activator) had no effect on cAMP production. Treatment with relaxin did not increase cAMP production by nontransfected cells or 293T cells overexpressing related receptors, LH receptor, LGR4, LGR5, or a short splicing variant of LGR7 (11). Total cAMP production was measured in triplicate by a specific radioimmunoassay (10,28).

To determine whether the expression of LGR7 and LGR8 is consistent with known relaxin binding sites, we determined their expression patterns (13). Reverse transcription polymerase chain reaction (RT-PCR) analysis of 22 different human cDNAs indicated that LGR7 transcript is expressed in the brain, kidney, testis, placenta, uterus, ovary, adrenal, prostate, skin, and heart. However, LGR8 transcript is mainly present in brain, kidney, muscle, testis, thyroid, uterus, peripheral blood cells, and bone marrow. Specific antibodies were generated against the ectodomain of LGR7 (14). Immunohistochemical analysis showed that the expression of LGR7 is cell type–specific in different rodent tissues (Fig. 2). In the uterus, LGR7 was expressed mainly in the myometrium and in the epithelial layer of the endometrium. In the vagina and cervix, LGR7 was found in muscularis layers and the myometrium, respectively.

Figure 2

Tissue distribution of LGR7, as assessed by immunohistochemical analysis in rodent tissues. Uterine tissues were obtained from postpartum rats; vagina and cervix were from rats at 19 days of pregnancy. Specific staining with antibody to LGR7 (14) in uterus (a), vagina (c), and cervix (e) is indicated by arrows. Nonimmune serum showed negligible staining in uterus (b), vagina (d), and cervix (f). Abbreviations: EM, endometrium; MM, myometrium; ML, muscularis layer; MU, mucosal layer; EC, endocervix. Immunohistochemical analysis was performed as described (6). Magnifications, ×100.

Following an anchored receptor approach previously used to generate soluble ectodomains of the gonadotropin and thyrotropin receptors for use as functional antagonists (15), we established stable cell lines expressing a fusion protein that comprised the ectodomain of LGR7 and the transmembrane region of a T cell surface antigen CD8. Upon treatment of cells with thrombin, a thrombin cleavage site located at the LGR7-CD8 junction allowed the release of the soluble ectodomain of LGR7, designated as 7BP. Western blot analysis indicated a single 7BP band of ∼60 kD (Fig. 3A). Cross-linking analysis demonstrated the formation of high molecular weight complexes between relaxin and 7BP (Fig. 3B). In contrast, the soluble ectodomain from rat LGR4 (4BP) showed negligible interaction with relaxin.

Figure 3

Relaxin binding to the ligand-binding domain of LGR7. (A) A soluble ectodomain of LGR7 (7BP) was detected after Western blot (WB) analysis with anti-FLAG and Coomassie blue staining (CS) of total isolated protein (10 μg) (arrowhead). A homologous domain of LGR4 (4BP) was also generated by a similar approach and detected by Western blot with anti-FLAG (arrowhead). (B) Specific interaction of relaxin and 7BP. Upper panel: Purified porcine relaxin (1 μg) (29) was preincubated with increasing concentrations (1, 10, 100, and 1000 ng) of purified 7BP for 1 hour and cross-linked with disuccinimidyl suberate at room temperature for 15 min before boiling under denaturing conditions and resolved by 7.5% SDS–polyacrylamide gel electrophoresis. The relaxin-7BP complexes (arrowhead) were detected by antibodies to porcine relaxin (lanes 1 to 4). In contrast, cross-linking of purified 4BP with relaxin showed negligible interaction (lane 5). Lower panel: Purity of the 5.9-kD relaxin is shown by Western blot analysis with antibodies to porcine relaxin (WB) or Coomassie blue staining (CS, 5 μg) under nondenaturing conditions.

To determine whether soluble 7BP could serve as a functional antagonist by sequestering relaxin, we simultaneously treated transfected 293T cells expressing LGR7 with relaxin and 7BP (Fig. 4A). Both 7BP and antibodies to porcine relaxin effectively blocked the effects of relaxin. However, cotreatment of relaxin with boiled 7BP or purified 4BP had minimal effect. In addition, the stimulatory effect of relaxin on rat myometrial cells (16) was also antagonized by cotreatment with 7BP (Fig. 4B).

Figure 4

Blockage of relaxin action by recombinant 7BP. (A) Purified 7BP blocked the stimulatory effects of relaxin on LGR7. Transfected 293T cells (105 cells per culture) stably expressing LGR7 from the pcDNA3.1-Zeo expression vector were isolated (14) and treated (105 cells per culture) with 0.1 nM porcine relaxin in combination with different doses of 7BP, boiled 7BP (B7BP), or 4BP for 24 hours under serum-free conditions. In some cultures, antibodies to porcine relaxin (RLXAb) was also included with relaxin. (B) Treatment of cultured rat myometrial cells with 7BP blocked the relaxin stimulation of cAMP production. Uterine tissues were obtained from 25-day-old female rats implanted with diethylstilbestrol for 3 days. Myometrial cells were prepared by digestions with trypsin and collagenase as described (16). Cells (105 per culture) were treated with 1 nM porcine relaxin with or without 7BP for 24 hours.

The antagonistic action of 7BP was further tested in vivo. Subcutaneous administration of 7BP (500 μg/day) for 4 days (days 17 to 20 after conception) in pregnant mice led to parturition delay by 27 hours (17). Most living pups had minimal milk in their stomachs. In antepartum mice, treatment with 7BP also led to the underdevelopment of nipples, as demonstrated by a 29% decrease in nipple size at about 12 hours after parturition (17). This finding is consistent with the deficiency of nipple development found in relaxin-null mice (18). An earlier study (19) also found a disruption of normal delivery in pregnant rats treated with antibodies to relaxin.

The signaling of relaxin through GPCRs is different from that of insulin and IGFs, which involves tyrosine kinase receptors (20). The existence of divergent receptor types for relaxin and insulin is consistent with the crystal structures of these ligands (21). However, the participation of downstream tyrosine kinases in relaxin signaling could not be ruled out (22). The common phenotypes of INSL3-null and GREAT-null mice indicate that INSL3 may be another ligand for LGR8, and possibly for LGR7.

Preterm labor and delivery remain major obstetrical problems. Studies on relaxin receptors could allow the design of agonistic or antagonistic relaxin analogs for the treatment of disorders of labor onset. Relaxin also has a role in regulating pituitary hormone release (1, 23), renal vasculature (24), and lung and skin remodeling (25) as well as in heart failure, angiogenesis, and tumor formation (26, 27). The identification of two relaxin receptors with overlapping tissue expression patterns could facilitate our understanding of relaxin actions in diverse physiological and pathological conditions.

  • * To whom correspondence should be addressed. E-mail: aaron.hsueh{at}stanford.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article