Phosphorylation-Dependent Regulation of Guanylyl Cyclase (GC)-A and Other Membrane GC Receptors

Abstract

Receptor guanylyl cyclases (GCs) are single membrane spanning, multidomain enzymes, that synthesize cGMP in response to natriuretic peptides or other ligands. They are evolutionarily conserved from sea urchins to humans and regulate diverse physiologies. Most family members are phosphorylated on 4 to 7 conserved serines or threonines at the beginning of their kinase homology domains. This review describes studies that demonstrate that phosphorylation and dephosphorylation are required for activation and inactivation of these enzymes, respectively. Phosphorylation sites in GC-A, GC-B, GC-E, and sea urchin receptors are discussed, as are mutant receptors that mimic the dephosphorylated inactive or phosphorylated active forms of GC-A and GC-B, respectively. A salt bridge model is described that explains why phosphorylation is required for enzyme activation. Potential kinases, phosphatases, and ATP regulation of GC receptors are also discussed. Critically, knock-in mice with glutamate substitutions for receptor phosphorylation sites are described. The inability of opposing signaling pathways to inhibit cGMP synthesis in mice where GC-A or GC-B cannot be dephosphorylated demonstrates the necessity of receptor dephosphorylation in vivo. Cardiac hypertrophy, oocyte meiosis, long-bone growth/achondroplasia, and bone density are regulated by GC phosphorylation, but additional processes are likely to be identified in the future.

Keywords: Npr1; Npr2; achondroplasia; cardiac hypertrophy; osteoporosis.

Graphical Abstract

Figure 1.

A cartoon model shows how highly phosphorylated versions of GC-A and GC-B are activated by NPs and how prolonged exposure to NPs or brief exposure to antagonizing factors result in receptor dephosphorylation and inactivation. Under basal conditions, both dimeric receptors are highly phosphorylated (yellow circles are phosphates) on their kinase homology domains and have low GC activity. Once bound by natriuretic peptide (NP), the affinity of the enzyme for GTP increases, which elevates cGMP synthesis. Over time, the NP-dependent bound receptor is dephosphorylated, resulting in receptor inactivation by a process known as homologous desensitization. More importantly, brief exposure to antagonizing factors like fibroblast growth factor (FGF) and luteinizing hormone (LH) for GC-B and possibly angiotensin II (AngII) for GC-A, stimulate receptor dephosphorylation and inactivation by a process known as heterologous desensitization.

Figure 2.

Phosphorylation of GC-A is positively correlated with ANP-dependent GC activity. (A) ANP exposure to 293 cells overexpressing GC-A causes a time-dependent decrease in the ³²PO₄ content and ANP-dependent GC activity of GC-A. Duplicate plates of ³²PO₄-labeled or unlabeled 293-GC-A cells were incubated with 1 μM ANP for the periods of time indicated, followed by preparation of either whole cell detergent extracts (from labeled cells) for purification of GC-A or particulate fractions (from unlabeled cells) for ANP-dependent guanylyl cyclase assays. (Inset) A 2 hours ANP exposure increased the electrophoretic mobility of ³⁵S-methionine-labeled GC-A, but did not decrease GC-A protein content, which is consistent with receptor dephosphorylation, but not receptor degradation. These data were originally reported by Potter and Garbers (86). (B) 293 cells transfected with alanine mutants or wild type GC-A constructs were metabolically labeled with ³²PO₄ overnight. GC-A was purified by immunoprecipitation, SDS-PAGE, and blotting to a PVDF membrane. The ³²PO₄ content was visualized by exposure to film. (C) The same membrane was subsequently probed with an antibody to GC-A to identify receptor protein content by Western blot. (D) Crude membranes from unlabeled cells transfected with alanine substitutions for known phosphorylation sites, displayed reduced ANP-dependent GC activity. (E) In contrast, conversion of Ser-487 to Glu to mimic the negative charge of phosphate reduced the activity of the enzyme, without affecting the protein content as depicted in the inset. Panels B, C and D are from Potter and Hunter (87) and panel E is from Schroter et al (88). (F) Conserved phosphorylation sites identified for the indicated transmembrane guanylyl cyclase receptors are shown in red. Numbers above the sequences are the amino acids for phosphorylated residues in rat GC-A. Phosphorylation site identifications are from the following references (87-92). Abbreviations: A. punGC, sea urchin; h, human; m, mouse; r, rat.

Figure 3.

Salt bridges between negatively charged phosphates and positively charged amino acids on the backside of the KHDs of GC-A and GC-B explain why phosphorylation is required for enzyme activation. (A) The back side of the KHD of GC-A contains a high concentration of positive charge as indicated by increased red color due to an elevated number of conserved arginines and lysines. (B) A back-to-back, head-to-tail, homodimer of the KHDs from GC-A and GC-B places the phosphates from the phosphorylated amino acids in one molecule in close proximity to the conserved arginines and lysines in the second molecule of the dimer. (C) Potential salt bridges between negatively charged phosphorylated amino acids and positively charged conserved arginine and lysine residues in GC-A and GC-B are shown.

Figure 4.

Reduced ERK activity decreases cardiac hypertrophy in mice where GC-A cannot be inactivated by dephosphorylation. (A) GC-A-8E, a glutamate-substituted mutant, was knocked into mice. (B) GC activity in kidney tissue from male (squares) or female (circles) in GC-A^8E/8E mice (red lines) is greater than the activity measured in male or female GC-A^WT/WT mice (black lines). (C) Heart weight to body weight ratios are lower in GC-A^8E/8E male mice. (D) Cardiomyocyte cross sectional areas are lower in GC-A^8E/8E male mice. (E) Ventricular ERK1 and ERK2 activities from male GC-A^8E/8E mice are lower than activities from GC-A^WT/WT male mice. All data are from Wagner and colleagues (38).

Figure 5.

GC-B^7E/7E mice are unresponsive to luteinizing hormone or fibroblast growth factor and have increased bone mineral density. (A) GC-B-7E that is not inactivated by dephosphorylation was knocked into mice (120). (B) Luteinizing hormone (LH) decreases GC-B activity in ovarian follicle membranes from GC-B^WT/WT but not GC-B^7E/7E mice (120). (C) FGF18 decreases CNP-dependent cGMP elevation in GC-B^WT/WT but not GC-B^7E/7E murine tibias (145). (D) Crossing FGFR3G380R/G380R mice with GC-B7E/7E mice rescues achondroplasia (146). (E) Tibias from male GC-B^7E/7E mice (red squares) have greater trabecular bone volume fraction (BV/TV) than wild type littermates (black circles) (147). (F) Tibias from male GC-B^7E/7E mice (red squares) have greater ultimate load than wild type littermates (black circles) (147). (G) Injection of the CNP analog, BMN-111, dramatically increases sclerostin (Sost) mRNA in tibias from adult male mice (47).