The nitric oxide receptor soluble guanylyl cyclase (sGC) exists in multimeric protein complexes, including heat shock protein (HSP) 90 and endothelial nitric oxide synthase. Inhibition of HSP90 by geldanamycin causes proteasomal degradation of sGC protein. In this study, we have investigated whether COOH terminus of heat shock protein 70-interacting protein (CHIP), a co-chaperone molecule that is involved in protein folding but is also a chaperone-dependent ubiquitin E3 ligase, could play a role in the process of degradation of sGC. Transient overexpression of CHIP in COS-7 cells degraded heterologous sGC in a concentration-related manner; this downregulation of sGC was abrogated by the proteasome inhibitor MG-132. Transfection of tetratricopeptide repeats and U-box domain CHIP mutants attenuated sGC degradation, suggesting that both domains are indispensable for CHIP function. Results from immunoprecipitation and indirect immunofluorescent microscopy experiments demonstrated that CHIP is associated with sGC, HSP90, and HSP70 in COS-7 cells. Furthermore, CHIP increased the association of HSP70 with sGC. In in vitro ubiquitination assays using purified proteins and ubiquitin enzymes, E3 ligase CHIP directly ubiquitinated sGC; this ubiquitination was potentiated by geldanamycin in COS-7 cells, followed by proteasomal degradation. In rat aortic smooth muscle cells, endogenous sGC was also degraded by adenovirus-infected wild-type CHIP but not by the chaperone interaction-deficient K30A CHIP, whereas CHIP, but not K30A, attenuated sGC expression in, and nitric oxide donor-induced relaxation of, rat aortic rings, suggesting that CHIP plays a regulatory role under physiological conditions. This study reveals a new mechanism for the regulation of sGC, an important mediator of cellular and vascular function.
- COOH terminus of heat shock protein 70-interacting protein
nitric oxide (NO) is an important modulator of vascular functions. Many of the effects of NO are mediated through the activation of cytosolic soluble guanylyl cyclase (sGC), an α- and β-subunit heterodimer. On NO binding to the sGC heme moiety, activity of sGC increases several hundredfold over basal levels and activated sGC converts GTP to cGMP (17) to initiate a cascade of multiple effector molecules, including protein kinases, phosphodiesterases, and ion channels (9, 21). Previous work (25) has shown the existence of endothelial NO synthase (eNOS)-heat shock protein (HSP) 90-sGC complexes in both endothelial and smooth muscle cells; these complexes function to stabilize sGC and reduce inactivation of NO by superoxide forming peroxynitrite. Inhibition of HSP90 by geldanamycin (GA) decreases both α- and β-sGC protein levels and activity in rat smooth muscle cells, and this effect was blocked by the proteasome inhibitor N-benzoyloxycarbonyl (Z)-Leu-Leu-leucinal (MG-132) (18), suggesting that the HSP90-based chaperone system is a regulator of sGC function.
HSP90 is one of the most abundant proteins in eukaryotic cells (1–2% total protein); together with other chaperones and co-chaperones, it helps maintain quality control of cellular proteins. Recently, COOH terminus of HSP70-interacting protein (CHIP) was identified as a protein quality-control, U-box-containing, E3 ubiquitin ligase, which facilitates processing of chaperone client proteins. Through its three tandem tetratricopeptide repeats (TPR) domain, CHIP binds to the TPR acceptor sites on HSP90 and HSP70 proteins (3, 7). CHIP has also been implicated in the ubiquitination and degradation of multiple chaperone client proteins, such as the glucocorticoid receptor (GR) (6), the cystic fibrosis transmembrane conductance regulator protein (CFTR) (15), and neuronal NO synthase (19). Additionally, CHIP interacts with eNOS by partitioning the soluble enzyme into the insoluble (inactive) cellular compartment, which eventually leads to the inactivation of eNOS (13).
In this study, we examined the relationship between CHIP and sGC and the role of CHIP in the ubiquitination and degradation of sGC. We show that CHIP forms a complex with sGC, ubiquitinates sGC in vitro and in intact cells, and promotes degradation of sGC. Our data suggest that CHIP plays an important role in regulating sGC function.
Anti-sGC antibodies were purchased from Cayman Chemical (Ann Arbor, MI) or Sigma. Anti-hemagglutinin (HA), anti-Myc, anti-FLAG, and anti-β actin antibodies were purchased from Sigma. Anti-glutathione S-transferase (GST) was from Amersham Biosciences. Anti-ubiquitin was purchased from Covance. Anti-CHIP antibody was from ABR. Anti-HSP90 and anti-HSP70 antibodies were obtained from Stressgen. GA, cycloheximide (CHX), and MG-132 were purchased from Sigma. Bicinchoninic acid (BCA) protein assay was purchased from Pierce (Rockford, IL). The GST cloning vector pGEX-4T-l and glutathione-Sepharose 4B were purchased from Amersham Pharmacia Biotech. Protein A/G agarose came from Santa Cruz Biotechnology, and SuperSignal chemiluminescence substrate came from Pierce. Ubiquitin, ubiquitin-activating enzyme (E1), and UbcH5a were from Boston Biochem. Plasmids containing the cDNAs for the human sGC α1- and β1-subunits were constructed as previously described (18). The cDNA for HA-ubiquitin was a gift from Dr. Pietro DeCamilli (Yale University).
Cell culture and transient transfection.
COS-7 cells were obtained from the American Type Culture Collection and were cultured in DMEM supplemented with 10% fetal bovine serum. COS-7 cells were transiently transfected with various expression plasmids by using Lipofectamine 2000 (Invitrogen), according to manufacturer's instructions. In-house-harvested rat aortic smooth muscle cells (RASMC) were used during passages 2–4 as described previously (25). Cultures were maintained in DMEM plus F-12 that contained 10% fetal bovine serum, 500 IU/ml penicillin, and 500 μg/ml streptomycin. Transfection of RASMC was done by using engineered adenovirus containing wild-type or mutant CHIP gene (produced at the University of North Carolina Gene Therapy Vector Core Lab).
SDS-PAGE, Western blotting, and immunoprecipitation.
Cells grown on six-well plates were harvested 24–48 h after transfection, washed three times with ice-cold PBS, and lysed with RIPA buffer (50 mM Tris·HCl, pH 7.4, 1% Igepal CA-630, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EDTA) supplemented with protease inhibitor cocktail (Calbiochem). Cell lysates were clarified by centrifugation at 14,000 rpm for 15 min at 4°C, and protein concentration was determined with the BCA protein assay kit (Pierce). In some experiments, proteins in the pellet were recovered by boiling in SDS loading buffer. Proteins were resolved by SDS/PAGE, transferred to nitrocellulose membranes (Bio-Rad), and probed with appropriate antibodies. Immunoreactive bands were visualized with the chemiluminescence reagent Immobilon (Millipore). Immunoprecipitation was done after preclearing the cell lysates (300 μg) with mouse or rabbit IgG (Santa Cruz) and protein A/G agarose beads (20 μl) for 1 h at 4°C, then appropriate antibodies (1 μg) were added into the supernatant and samples were incubated overnight at 4°C. Fifty microliters of protein A/G agarose beads were added into the samples and were rotated for 1 h. Immunoprecipitated proteins bound to the agarose beads were transferred to spin filters (Microfilter) and were washed four times with RIPA buffer. Proteins were eluted from the beads by incubating with SDS sample buffer at room temperature and were analyzed by Western blotting. Optical densities of bands were estimated with ImageJ.
Pulse-chase analysis with CHX.
COS-7 cells were transiently transfected with various expression plasmids, as described in Cell culture and transient transfection. Twenty-four hours after transfection, CHX (50 μg/ml) was added and cells were incubated for various times. Cell lysates were then subjected to SDS-PAGE and immunoblotting analysis with antibodies to Myc or β-actin.
Expression and purification of GST-sGCα1 and GS-sGCβ1.
Rat sGC α1- and β1-subunit cDNA were subcloned into the GST fusion protein cloning vector pGEX-4T-1. Fusion GST-sGCα1 and GST-sGCβ1 proteins were expressed in Escherichia coli and were purified by affinity binding to glutathione-Sepharose beads as previously described (25).
Indirect immunofluorescence was performed as described previously (1). Briefly, COS-7 cells grown on glass coverslips were first washed with PBS, fixed, and permeabilized in a methanol-acetone mixture (1:1) at −20°C for 20 min, then were washed in PBS, blocked with 1% BSA, and incubated with specific rabbit polyclonal anti-CHIP antibodies and mouse monoclonal anti-Myc antibodies for 1 h at 37°C. After a wash, cells were incubated with anti-rabbit secondary antibodies conjugated with Cy2, anti-mouse secondary antibodies conjugated with Cy3, and Hoechst 33258. Coverslips were mounted on slides with ProLong antifade mounting medium and were analyzed under a laser-scanning confocal microscope (LSM 510, Meta 3.2; Zeiss). Control, untransfected experiments did not exhibit any fluorescence. Sequential z-axis sections (0–2.9 μm) were examined at 0.36-μm intervals.
In vitro ubiquitination assays.
Bacterially expressed GST-sGCα1 and/or GST-sGCβ1 (0.25 μg) were incubated in reactions containing 0.1 μM human recombinant ubiquitin-activating enzyme E1, 8 μM UBCH5a, 4 μM His-tagged CHIP, 2.5 mg/ml ubiquitin, 1 mM DTT, and an 1× ATP-regenerating system for 2 h at 37°C. Reactions were stopped with 20 μl SDS loading buffer, and a 20-μl aliquot was subjected to SDS-PAGE and immunoblotted with appropriate antibodies.
Vascular relaxation studies.
The procedures were approved by the Institutional Committee on Animal Care in Research and Education and conform with guidelines for humane animal care of the American Physiological Society. Sprague-Dawley rats were euthanized by CO2 inhalation followed by thoracotomy, and the abdominal aorta was removed and sliced into 3-mm rings. Rings were incubated overnight at 4°C in either vehicle, Ad-CHIP or Ad-K30A (the TPR domain mutant CHIP) at 109 particles/ml. Next, rings were mounted on myographs (DMT, Copenhagen, Denmark) at an optimal tone of 2 g, allowed to equilibrate for 30 min, and constricted with 10−6 M U-46619 (thromboxane analog). Concentration-response relationships were then obtained to the NO donor spermine NONOate, which releases NO spontaneously at normal pH. Results are presented as percent of maximal relaxation for six to eight rings per group.
Data are expressed as means ± SE. All statistical tests were performed by using ANOVA. P < 0.05 was accepted as statistically significant.
CHIP decreases the stability of sGC in COS-7 cells.
Using CHX pulse-chase experiments, we investigated whether CHIP interferes with sGC stability. In the absence of CHIP, the half-life of overexpressed sGC α1- and sGC β1-subunits was ∼7 h, whereas in the presence of CHIP the turnover of both sGC α1- and sGC β1-subunits increased and its half-life was reduced to ∼3 h (Fig. 1). This data indicates that CHIP significantly reduces sGC protein levels.
CHIP promotes the proteasomal degradation of sGC subunits in COS-7 cells.
Because CHIP is a U-box ubiquitin E3 ligase responsible for the degradation of other HSP90 client proteins, we used transient transfection methods to investigate whether CHIP decreases sGC stability by enhancing its ubiquitination and proteasomal degradation. COS-7 cells were transiently transfected with Myc-tagged CHIP and sGC α1-subunit for 24 h, and the soluble fraction of the cell lysates was analyzed by Western blotting. As shown in Fig. 2A, transfection of COS-7 cells with increasing amounts of wild-type CHIP cDNA decreased the amount of expressed sGC α1-subunit, whereas HSP90, HSP70, and β-actin levels remained unaffected. When the selective proteasome inhibitor MG-132 (5 μM) was added to the cells for an additional 24 h after transfection, sGC α1-subunit degradation by CHIP was almost abolished (Fig. 2B); at the same time, increased total cell protein ubiquitination was observed. Results similar to those for sGC α1-subunit were obtained in COS-7 cells cotransfected with FLAG-tagged sGC β1-subunit and CHIP (data not shown). These results suggest that CHIP enhances sGC degradation through the ubiquitin-proteasome pathway. Because sGC functions as a heterodimer and single sGC subunits may be prone to degradation (16), we carried out experiments to confirm that CHIP acts similarly on the intact sGC heterodimer in COS-7 cells cotransfected with both sGC subunits and CHIP. As shown in Fig. 2C, both sGC α1- and sGC β1-subunit expression are decreased by wild-type CHIP, suggesting that intact sGC is similarly affected by CHIP.
Both TPR and U-box domains are essential for CHIP-mediated sGC degradation.
The E3 ligase CHIP has an NH2-terminal TPR domain, which binds to HSP90/HSP70, and a COOH-terminal U-box domain, which is responsible the ubiquitin ligase activities (27). To test whether both domains were required for sGC subunit degradation, we used two mutant CHIP constructs: CHIP (K30A), a TPR domain mutant, which does not bind HSP90 or HSP70, and CHIP (H260Q), a U-box domain mutant, which lacks protein ubiquitination abilities (11). Both the TPR and U-box domain CHIP mutants decreased sGC subunit expression to a much lesser extent than wild-type CHIP (Fig. 2C). These data suggest that both chaperone binding and ubiquitin ligase activities are necessary for CHIP-mediated sGC degradation. Although both mutants exhibited less sGC degradation than the wild-type CHIP, we observed a lesser amount of sGC protein with the H260Q mutant compared with the K30A mutant in the supernatant fraction of cell lysates. When we probed the pellet fraction, it became evident that, in the presence of the H260Q mutant CHIP, sGC subunit expression remained unchanged compared with noninfected cells; however, both sGC subunits were relocated to the detergent-insoluble pellet fraction, along with the H260Q CHIP mutant (Fig. 2C). Thus the U-box mutant CHIP H260Q functions like a dominant-negative E3 ligase, which lacks ubiquitin ligase activity. Instead of stimulating the ubiquitin-proteasome pathway, it causes aggregation of misfolded proteins in the detergent-insoluble pellet fraction. This appears similar to what has been described for nucleophosmin-anaplastic lymphoma kinase (5).
CHIP associates with sGC.
To examine whether CHIP associates with sGC in intact cells, COS-7 cells were cotransfected with Myc-tagged sGC α1-subunit and wild-type CHIP; nontransfected cells served as controls. Twenty-four hours later, cells were fixed, stained with anti-CHIP primary and secondary antibodies (Cy2, green), anti-Myc primary and secondary antibodies (Cy3, red), and Hoechst 33258 dye to visualize the nuclei (blue) and were examined under a confocal microscope. As shown in Fig. 3A, there was strong association of sGC α1-subunit and CHIP in the cytoplasmic region (1.09-μm section in a 0- to 2.9-μm z-axis) of COS-7 cells (yellow). Cells that were not infected with either sGC α1-subunit (white arrows) or both sGC α1-subunit and CHIP (yellow arrows) did not exhibit yellow color in the merge mode. This observation was further verified with immunoprecipitation experiments. sGC α1-subunit was immunoprecipitated from COS-7 cells cotransfected with sGC α1-subunit with or without wild-type CHIP; immunoprecipitates were then immunoblotted with anti-CHIP antibody. As shown in Fig. 3B, strong expression of CHIP was observed in immunocomplexes pulled down by anti-sGC α1-subunit antibody. In agreement with previous reports, there were also associations between sGC, HSP90, and HSP70 (2, 25). To increase the stability of sGC protein and immunoprecipitation efficiency, in additional experiments we included the proteasome inhibitor MG-132 (5 μM) for 6 h before cell harvesting. Coexpression of sGC α1-subunit and CHIP greatly increased the association of HSP70 with sGC α1-subunit (lane 2), suggesting that HSP70 is an HSP90 co-chaperone associated with client protein degradation. As with other HSP90 client proteins, inhibition of HSP90 by ansamycin drugs, such as GA, caused significant degradation of sGC α1-subunit, even in the presence of MG-132 (lane 3). The specificity of the sGC α1-subunit association with CHIP was demonstrated in experiments where lysates were immunoprecipitated with anti-IgG antibodies; neither sGC α1-subunit nor CHIP was detected (Fig. 3C).
CHIP enhances sGC ubiquitination in vitro.
For a protein to be recognized and then degraded by the proteasome system, it first needs to be ubiquitinated. To investigate whether CHIP directly ubiquitinates sGC, we performed an in vitro ubiquitination assay by using purified proteins: purified GST-sGCα1, His-tagged wild-type or U-box mutant CHIP, human recombinant E1, E2 (UbcH5a), and ubiquitin. As seen in Fig. 4, addition of wild-type CHIP to the assay mixture caused the ubiquitination of sGC α1-subunit (reflected in a high-molecular-weight smear); reprobing of the same blot with anti-ubiquitin antibody confirmed the enhanced ubiquitination by the wild-type CHIP. These data suggest that CHIP can serve as an E3 ubiquitin ligase for sGC. Conversely, the U-box mutant CHIP lacked E3 ligase activity and did not ubiquitinate sGC α1-subunit. Our results indicated that CHIP directly ubiquitinates sGC in vitro and promotes its degradation by the proteasome system. Similar results were obtained by using purified GST-sGCβ1 protein (data not shown).
CHIP induces sGC ubiquitination in COS-7 cells.
To investigate whether CHIP induces sGC ubiquitination in intact cells, we tested whether cotransfection of CHIP with both sGC subunits affects the ubiquitination of sGC in COS-7 cells. Cell lysates were immunoprecipitated by using either anti-sGC α1- or anti-sGC β1-subunit antibody and were immunoblotted for sGC α1, sGC β1, and HA-tagged ubiquitin. As shown in Fig. 5, coexpression of CHIP, sGC β1- and sGC α1-subunit in COS-7 cells increased the ubiquitination levels of sGC (comparing lanes 3 and 2, quantified in bar graph at right), indicating that the degradation of sGC caused by CHIP was linked to the ubiquitination of sGC. Furthermore, treatment of cells with the HSP90 inhibitor GA significantly increased the level of ubiquitination of both sGC β1- and sGC α1-subunit (comparing lanes 5 and 2, lanes 6 and 3, and in bar graph at right) and was accompanied by degradation of sGC α1- and sGC β1-subunit (comparing lanes 5 and 6 to 2 and 3 and in bar graph at right). Also, coexpression of CHIP enhanced the sGC degradation induced by GA (comparing lanes 6 and 5). These data support the concept that CHIP ubiquitinates sGC α1- and sGC β1-subunits in intact cells. This effect of CHIP was further augmented by MG-132. As shown in Fig. 6, the proteasome inhibitor increased ubiquitination and expression of both sGC subunits.
Overexpression of CHIP enhances GA-induced sGC α1-subunit degradation.
The HSP90 inhibitor GA binds to the ATP/ADP pocket of HSP90 to induce an open conformation of the HSP90 dimer, recruitment of specific co-chaperones, and the destabilization and degradation of many client proteins. We examined the possible synergy between CHIP and GA on heterologous sGC α1-subunit in COS-7 cells. As seen in Fig. 7, top, GA produced a time-dependent sGC α1-subunit degradation as early as 30 min after exposure; coexpression of CHIP enhanced the effect of GA (Fig. 7, bottom) in an additive way. Pretreatment of cells with the specific proteasome inhibitor MG-132 for 3 h attenuated the CHIP- and GA-induced sGC α1-subunit degradation.
Adenovirus-CHIP promotes endogenous sGC degradation in vascular smooth muscle cells.
To investigate whether CHIP reduces the stability of endogenous sGC, we used early-passage (2–4) RASMC, harvested in our lab, because they exhibit high endogenous expression of both sGC subunits. Expression of wild-type or mutant CHIP was achieved by adenovirus (Ad)-CHIP infection into RASMC for 24 h. Cell lysates were collected and analyzed by Western blotting. Overexpression of wild-type CHIP degraded endogenous sGC α1- and sGC β1-subunits (Fig. 8A). Compared with noninfected cells, low concentration of wild-type Ad-CHIP caused ∼40% decrease in sGC α1- and sGC β1-subunit expression, whereas the higher concentration of wild-type CHIP slightly further decreased sGC α1- and sGC β1-subunit expression (Fig. 8B). Conversely, the chaperone interaction-deficient TPR domain mutant K30A CHIP failed to reduce sGC expression, indicating that CHIP-induced sGC degradation is chaperone dependent. Furthermore, in agreement with previous reports (8), wild-type CHIP, but not the TPR mutant CHIP, dose-dependently increased HSP70 expression. Together, these results suggest that increased expression of HSP70 plays a role in the degradation of endogenous sGC by CHIP in RASMC.
CHIP attenuates the vasorelaxant effects of NO.
NO relaxes blood vessels by stimulating sGC to produce cGMP. We tested the hypothesis that because CHIP promotes sGC degradation, it would also attenuate NO-mediated vascular relaxation. Rat aortic rings exposed to vehicle exhibited a robust relaxation to the NO donor spermine NONOate (Fig. 9A). Conversely, vessels exposed to Ad-CHIP exhibited a significant rightward shift in the concentration-response curve and a significant reduction in the magnitude of maximal relaxation to spermine NONOate. Vessels exposed to the CHIP mutant K30A, which lacks the ability to bind to HSP90/HSP70, responded similarly to vehicle-exposed, control rings. Western blot analysis of vessel ring lysates confirmed that Ad-CHIP, but not K30A, reduced expression of both sGC α1- and sGC β1-subunits.
Posttranslational protein-protein interaction is an important regulatory mechanism for sGC functions. We have previously demonstrated (25) that eNOS and sGC are associated with HSP90 in multimeric complexes. The existence of a HSP90-eNOS-sGC complex in endothelial cells would increase the efficacy of NO, because the proximity of eNOS to sGC would attenuate inactivation of NO by superoxide anions and therefore improve the responsiveness of sGC to NO (25). Later, Balashova et al. (2) demonstrated that HSP70 is also a sGC-associated protein, which could increase the cGMP-forming ability of the enzyme. Inhibition of HSP90 by GA leads to proteasomal degradation of sGC subunits (18), further verifying that sGC is an HSP90 client protein. In the present study, we report on another sGC-interacting molecule, CHIP, which downregulates exogenous and endogenous sGC through the ubiquitin proteasome pathway (UPP). Removal of misfolded or damaged proteins via the UPP is an important mechanism in terms of protein quality control. For a protein to be recognized by the 26S proteasome, it must be ubiquitinated by a multiple-enzyme cascade involving at least three distinct types of enzymes: a ubiquitin-activating enzyme (E1), a ubiquitin-conjugating enzyme (E2), and a ubiquitin ligase (E3). E3 enzymes catalyze the final step for substrate recognition in UPP and add ubiquitin chains to target proteins. Hundreds of E3 ligases have been described, reflecting the diversity and specificity of proteins that they interact with (26). As a U-box E3 ligase, CHIP was originally identified as a TPR-containing protein that interacts with heat shock cognate/HSP70 (3). CHIP binds directly to the molecular chaperones HSP90 or HSP70 via their TPR domains, and, through its COOH-terminal E3 ligase activity, it causes degradation of multiple target proteins, such as GR (6), CFTR (15), and erythroblastic leukemia viral oncogene homolog 2 (Erb-B2) (27). Our current data demonstrate that sGC is a new CHIP target protein. Thus the HSP90/HSP70 chaperones not only enhance sGC activity, but on some occasions they also help decrease sGC protein levels through the UPP. It appears that these opposing actions of chaperones are under elegant regulation, but conditions that favor enhancing or downregulating sGC activities need further examination. As an HSP90/HSP70-interacting protein, CHIP might play a role in determining the fate of sGC. Under physiological conditions, intracellular concentrations of CHIP are low relative to those of folding co-chaperones, such as HSP40 and HSP90, so that conditions that favor folding rather than degradation prevail. However, two- to fourfold elevation of CHIP is sufficient to drive CFTR and GR to the UPP for degradation (6, 15).
Because both HSP90 and HSP70 appear in sGC/CHIP immunoprecipitates, it is not yet clear what their relative role is in sGC ubiquitination and degradation. In sGC α1-subunit immunoprecipitates from naïve cells (not transfected with CHIP), it is unclear whether HSP90 and HSP70 expression are different (Fig. 3B), but when CHIP was cotransfected or cells were treated with CHIP and GA, HSP70 was preferentially overexpressed and was the dominant coprecipitated chaperone, suggesting that CHIP and GA increase the amount of HSP70 associated with sGC. This is similar to what was reported for Erb-B2 (27). However, in contrast to Erb-B2, we did not observe a reduction in the amount of coprecipitated HSP90 in either condition; instead, GA and overexpression of CHIP increased HSP90 association with sGC. Therefore, a better understanding of the role of each chaperone is still needed.
The HSP90 inhibitor GA replaces the nucleotide in the HSP90-binding “pocket” and forces chaperone complexes into an ADP-like bound conformational state, which favors client protein degradation (22). Interestingly, we found that CHIP and GA exert similar negative effects on sGC stability (Fig. 7), suggesting a cooperative action between the two molecules, as previously reported for Erb-B2 and nucleophosmin-anaplastic lymphoma kinase (5, 27). In addition, when COS-7 cells were exposed to GA, drug-induced degradation of exogenous sGC was enhanced in the presence of wild-type CHIP.
Our intact cell data revealed that both TPR and U-box domains are necessary for the CHIP-mediated degradation of sGC, suggesting that this process is chaperone dependent. However, molecular chaperones such as HSP90 or HSP70 were not required for CHIP-mediated in vitro sGC ubiquitination (HSP90 or HSP70 protein levels were not detectable in purified GST-sGC and His-CHIP preparations; data not shown). Because the in vitro reactions contained both E1 and E2, it may be that assembly of activating and conjugating enzymes is a primary role of these chaperones. Indeed, the following evidence supports the hypothesis that chaperone function is necessary for CHIP activity in sGC ubiquitination and degradation: 1) the TPR mutant CHIP degrades sGC substantially less than the wild-type CHIP, and 2) in in vitro conditions, both sGC and CHIP are in excess concentrations, hence precluding the need for HSP90/HSP70 to recruit CHIP for sGC ubiquitination. Hatakeyama (10) and Timsit (24) described similar observations in the case of Tau and Epsin proteins. However, it is still possible that CHIP not only associates with sGC indirectly via chaperones but that it also binds to sGC directly, as in the case of the androgen receptors (12).
Our ex vivo, vascular findings support the cell data. CHIP-exposed but not K30A-exposed vessels exhibited a severely attenuated response to NO and reduced sGC expression, suggesting that CHIP-induced degradation of sGC has important physiological consequences and that it requires association of CHIP with HSP90. These findings also agree with our previous work showing that HSP90 inhibitors reduce vascular relaxation in response to spermine NONOate and other NO donors (28).
In summary, our results demonstrate for the first time that CHIP is a novel sGC-interacting protein that decreases the stability of both heterologous sGC subunits, ubiquitinates sGC α1- and sGC β1-subunits in intact cells and in vitro, and prepares the sGC heterodimer for proteasome degradation. We also show that the HSP90 inhibitor GA degrades sGC α1- and sGC β1-subunits through the UPP, a process that is enhanced by CHIP. Various cardiovascular diseases are associated with the downregulation of the activities and expression of the heme-containing NO receptor sGC (4, 14, 20). It is noteworthy that in certain cardiovascular diseases, increased reactive oxygen species production increases sGC oxidation and heme loss, which may be linked to the downregulation of sGC function. Recently, Stasch (23) showed that heme oxidation induced by the sGC inhibitor 1H-[1,2,4]oxadiazalo[4,3-a]quinoxalin-1-one (ODQ) promotes sGC ubiquitination, followed by proteasomal degradation; although the contribution of CHIP to this phenomenon was not addressed, it is possible that ODQ-induced proteasomal degradation of sGC is CHIP dependent. Studying alterations in CHIP expression and function in cardiovascular disease may provide new insights into their pathogenesis and management.
This work was supported by National Heart, Lung, and Blood Institute Grants HL-70214 and HL-66993 to J. D. Catravas.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by the American Physiological Society