Evolutionarily conserved role of calcineurin in phosphodegron-dependent degradation of phosphodiesterase 4D

Abstract

Calcineurin is a widely expressed and highly conserved Ser/Thr phosphatase. Calcineurin is inhibited by the immunosuppressant drug cyclosporine A (CsA) or tacrolimus (FK506). The critical role of CsA/FK506 as an immunosuppressant following transplantation surgery provides a strong incentive to understand the phosphatase calcineurin. Here we uncover a novel regulatory pathway for cyclic AMP (cAMP) signaling by the phosphatase calcineurin which is also evolutionarily conserved in Caenorhabditis elegans. We found that calcineurin binds directly to and inhibits the proteosomal degradation of cAMP-hydrolyzing phosphodiesterase 4D (PDE4D). We show that ubiquitin conjugation and proteosomal degradation of PDE4D are controlled by a cullin 1-containing E(3) ubiquitin ligase complex upon dual phosphorylation by casein kinase 1 (CK1) and glycogen synthase kinase 3beta (GSK3beta) in a phosphodegron motif. Our findings identify a novel signaling process governing G-protein-coupled cAMP signal transduction-opposing actions of the phosphatase calcineurin and the CK1/GSK3beta protein kinases on the phosphodegron-dependent degradation of PDE4D. This novel signaling system also provides unique functional insights into the complications elicited by CsA in transplant patients.

FIG. 1.

Posttranscriptional regulation of PDE4D expression by calcineurin. (A) Extracts prepared from COS7 cells transiently transfected with PDE4D3, PDE4D5, or βArr-1 were examined by immunoblot analysis. The effect of coexpression of CnAβ is also shown. Expression of endogenous IGF-1 receptor (IGF-1) and tubulin is shown as a control. (B) Expression of endogenous PDE4D in MEFs was detected by immunoprecipitation (IP) using rabbit (Rb) polyclonal PDE4D antibody (sc25814) and subsequent immunoblot analysis (IB) using goat (Gt) polyclonal PDE4D antibody (9). The effect of the calcium ionophore ionomycin and mitogenic stimuli (20% fetal calf serum) is also shown. (C) Extracts prepared from COS7 cells transiently transfected with PDE4D3 were treated with the calcium ionophore Ion (2 μM) and/or the calcineurin inhibitor CsA (5 μM). Expression of PDE4D3 was determined by immunoblot analysis. (D) COS7 cells were transiently transfected with PDE4D3. The half-life of PDE4D3 was determined in the presence CHX (0.5 mM). The effect of CsA (5 μM) is also shown. Expression of tubulin was used as a control. (E) COS7 cells were transiently transfected with PDE4D3. The half-life of PDE4D3 was determined in the presence of CsA (5 μM). The effect of the transcription inhibitor ActD (0.5 μg/ml) or the translational inhibitor CHX (0.5 mM) was examined. *, P < 0.05; #, P < 0.005.

FIG. 2.

Ubiquitin-mediated degradation of PDE4D via a phosphodegron motif. (A) Schematic illustration of the location of the DSGSQVEED phosphodegron in PDE4D3. Sequences of known phosphodegrons are also shown. Conserved amino acid residues and phosphorylated Ser are highlighted and underlined, respectively. A phosphodegron-like motif (DSGSQVEED) and potential phosphorylation sites (Ser⁶¹⁶ and Ser⁶¹⁸) in PDE4D3 are also shown. (B) COS7 cells were transiently transfected with PDE4D3. Ubiquitin conjugation of PDE4D3 was determined by coimmunoprecipitation assays. The presence of ubiquitin in PDE4D3 immunoprecipitates (IP) was analyzed by immunoblot (IB) analysis. The effect of the proteosome inhibitor MG132 (10 μM) is also shown. (C) COS7 cells were transiently transfected with PDE4D3. The expression levels of PDE4D3 in the presence and absence of dominant negative Cul1 (ΔCul1) were determined by immunoblot analysis. (D) COS7 cells were transiently transfected with wild-type or phosphodegron-mutated (Ala⁶¹⁶, Ala⁶¹⁸, Ala^616,618, or Ala^621,622,623) PDE4D3. The expression levels of wild-type and phosphodegron-mutated PDE4D3 were determined in the presence and absence of ΔCul1. (E) COS7 cells were transiently transfected with wild-type or phosphodegron-mutated (Ala⁶¹⁶, Ala⁶¹⁸, Ala^616,618, or Ala^621,622,623) PDE4D3. The expression levels of wild-type and phosphodegron-mutated PDE4D3 were determined in the presence and absence of the proteosome inhibitor MG132 (10 μM). (F) COS7 cells were transiently transfected with wild-type or phosphodegron-mutated (Ala⁶¹⁶, Ala⁶¹⁸, Ala^616,618, or Ala^621,622,623) PDE4D3. Ubiquitin conjugation of wild-type and phosphorylation-defective PDE4D3 was determined by coimmunoprecipitation assays. The presence of ubiquitin in PDE4D3 immunoprecipitates was analyzed by immunoblot analysis. The effect of the proteosome inhibitor MG132 (10 μM) is also shown. (G) COS7 cells were transiently transfected with wild-type or phosphodegron-mutated (Ala^616,618) PDE4D3. The half-lives of wild-type and phosphorylation-defective PDE4D3 were determined in the presence of CHX (0.5 mM) and CsA (5 μM). *, P < 0.05; #, P < 0.005.

FIG. 3.

Calcineurin regulation of PDE4D degradation is phosphodegron dependent. Wild-type or phosphodegron-mutated (Ala⁶¹⁶, Ala⁶¹⁸, Ala^616,618, or Ala^621,622,623) PDE4D3 was coexpressed with CnAβ in COS7 cells. The expression levels of PDE4D3 and CnAβ were determined by immunoblot analysis. The expression level of tubulin was used as a control.

FIG. 4.

Mapping of functional PXIXIT-like calcineurin docking motifs in PDE4D. (A) Schematic illustration of the locations of the two functional PXIXIT-like motifs (PLNLYR⁴⁹⁶ and PEACVI⁶⁶⁶) in the catalytic core of PDE4D3. The location of the DSGSQVEED phosphodegron in PDE4D3 is also indicated. A sequence comparison of known calcineurin docking PXIXIT motifs found in human NFAT and yeast proteins is shown. Potential PXIXIT-like motifs in PDE4D3 and an optimized PXIXIT motif (PVIVIT) are also shown. Functional PXIXIT-like motifs in PDE4D3 are highlighted. (B) Wild-type and PXIXIT-mutated PDE4D3 were transiently transfected into COS7 cells. The expression levels of wild-type and PXIXIT-mutated PDE4D3 were determined in the presence and absence of Ion (2 μM) and/or CsA (5 μM). The expression of tubulin was used as a control. Replacements of critical amino acid residues with Ala in potential PXIXIT-like motifs of PDE4D3 are underlined. Functional PXIXIT-like motifs in PDE4D3 are highlighted.

FIG. 5.

Direct interaction between calcineurin and PDE4D. (A) Schematic illustration of PDE4D3 and PDE4D3 GST fusion protein. Functional calcineurin docking motifs (PLNLTR⁴⁹⁶ and PEACVI⁶⁶⁶) are indicated. Calcineurin binding-defective PDE4D3 (ALALAA⁴⁹⁶ and AEACAA⁶⁶⁶) is also shown. (B) Binding of endogenous PDE4D and calcineurin was determined by coimmunoprecipitation assays. Endogenous PDE4D in COS7 cells was immunoprecipitated (IP) using goat polyclonal PDE4D antibody (sc25097), mouse monoclonal calcineurin A antibody (sc17808), goat polyclonal calcineurin Aβ antibody (sc6124), or goat polyclonal calcineurin B antibody (sc6119). The presence of endogenous PDE4D in calcineurin immunoprecipitates was analyzed by immunoblot (IB) analysis using rabbit polyclonal PDE4D antibody (sc25814). Mouse M2 monoclonal antibody was used as a control. (C) PDE4D3 was coexpressed with CnAβ in COS7 cells. Binding of PDE4D3 and CnAβ was determined by coimmunoprecipitation assays. The presence of PDE4D3 in CnAβ immunoprecipitates and in cell extracts was analyzed by immunoblot analysis. The effect of the PXIXIT-containing calcineurin docking inhibitor dnNFAT is also shown. (D) Wild-type and calcineurin binding-defective PDE4D3 were coexpressed in the presence and absence of CnAβ in COS7 cells. The presence of PDE4D3 in CnAβ immunoprecipitates and in cell extracts was analyzed by immunoblotting.

FIG. 6.

CK1 and GSK3β phosphorylate PDE4D phosphodegron. (A) Schematic illustration of PDE4D3 and PDE4D3 GST fusion protein. The location of the DSGSQVEED phosphodegron in PDE4D3 are also shown. Replacements of critical amino acid residues in the DSGSQVEED phosphodegron of PDE4D3 with Ala are highlighted. (B) GST recombinant proteins encompassing either the NH₂-terminal or the COOH-terminal portion of the catalytic core of PDE4D3 were subjected to in vitro protein kinase assays in the presence of recombinant GSK3β or CK1 and [γ-³²P]ATP. Autoradiography and Coomassie blue staining of recombinant PDE4D3 are shown. (C) COS7 cells transiently transfected with wild-type or Ala^616,618 mutant PDE4D3. The effect of GSK3β was also examined. Phosphorylation of Ser⁶¹⁶ of PDE4D3 was assessed by phospho-PDE4D antibody (Phos-PDE4D) in immunoblot analysis. Expression of PDE4D3 and tubulin is also shown.

FIG. 7.

CK1 and GSK3β promote PDE4D degradation. (A) PDE4D3 or βArr-1 was coexpressed with GSK3β or CK1 in COS7 cells. The expression levels of PDE4D3, βArr-1, GSK3β, and CK1 were determined by immunoblot analysis. The expression level of tubulin was used as a control. (B) PDE4D3 was transiently expressed in COS7 cells. The expression levels of PDE4D3 in the presence and absence of protein kinase inhibitors for 4 h were determined by immunoblot analysis. The effect of Ion (2 μM) is also shown. The protein kinase inhibitors administered included the GSK3 inhibitor LiCl (20 mM), the CK1 inhibitor D4476 (1 μM), and the CK2 inhibitor DMAT (400 nM). (C and D) Wild-type or phosphodegron-mutated (Ala⁶¹⁶, Ala⁶¹⁸, Ala^616,618, or Ala^621,622,623) PDE4D3 was coexpressed with GSK3β (C) or CK1 (D) in COS7 cells. The expression levels of PDE4D3, GSK3β, and CK1 were determined by immunoblot analysis. The expression level of tubulin was used as a control.

FIG. 8.

Evolutionarily conserved regulation of PDE4D degradation via calcineurin, CK1, and GSK3β. (A) Schematic illustration of C. elegans PDE4 (cePDE4). The locations of the potential phosphodegron and calcineurin docking motif are also shown. Sequence alignment and percent identity with human PDE4D (hPDE4D) are indicated. Bullets indicated conserved amino acid residues. Colons indicated similar amino acid residues. Asterisks indicated potential phosphorylation sites. (B) Extracts prepared from CnAβ^−/− and CnAβ^+/+ mouse skeletal muscle were subjected to immunoprecipitation and immunoblot analysis to determine the endogenous expression of PDE4D. The expression level of tubulin was used as a control. (C) Extracts prepared from N2 wild-type control C. elegans and loss-of-function calcineurin mutant worms [tax-6(p675lf)] were subjected to immunoblot analysis to determine the endogenous expression of cePDE4. Extracts prepared from PDE4 null worms [pde-4(ok1290)] were used to show the specificity of the cePDE4 antibody. The expression level of C. elegans tubulin was used as a control. (D) cePDE4 was transiently cotransfected with either CK1 or GSK3β into COS7 cells. Phosphorylation of cePDE4 was determined by immunoprecipitation (IP) using M2 monoclonal antibody against FLAG-tagged cePDE4 and subsequent immunoblotting (IB) analysis using phospho-Thr polyclonal antibodies. The expression levels of cePDE4, CK1, GSK3β, and tubulin are also shown.

FIG. 9.

Model of the phosphatase calcineurin in the β-adrenergic signaling pathway via regulation of PDE4D degradation. PDE4D encompasses a DSGSQVEED phosphodegron. Phosphorylation at the phosphodegron by the protein kinases GSK3β and CK1 promotes PDE4D degradation, which is mediated by the SCF E3 ubiquitin (Ub) ligase complex. Ubiquitinated PDE4D was degraded by the proteosome complex. Conversely, dephosphorylation mediated by the phosphatase calcineurin opposes PDE4D degradation. Thus, phosphorylation-dependent degradation controls the expression level of PDE4D, which in turn regulates the intracellular concentration of the second messenger cAMP upon activation of the β-adrenergic receptor signaling pathway—a key component of the control of diverse physiological responses. The calcineurin/ubiquitin/PDE4D signaling axis is also evolutionarily conserved in C. elegans, indicating its important role in critical cellular processes to regulate cAMP duration. We propose that potentiation of the G_s-coupled signaling pathway due to sustained elevation of cAMP, which is a consequence of reduced PDE4D expression in the absence of calcineurin function, accounts for the metabolic complications (e.g., hyperlipidemia and hemodynamic dysregulations) found in transplant patients treated with CsA. AC, adenylyl cyclase.