Edited by Ruma Banerjee, University of Michigan, Ann Arbor, MI, and accepted by the Editorial Board June 26, 2012 (received for review April 6, 2012)
Heme insertion is key during maturation of soluble guanylyl cyclase (sGC) because it enables sGC to recognize NO and transduce its multiple biological effects. Although sGC is often associated with the 90-kDa heat shock protein (hsp90) in cells, the implications are unclear. The present study reveals that hsp90 is required to drive heme insertion into sGC and complete its maturation. We used a mammalian cell culture approach and followed heme insertion into transiently and endogenously expressed heme-free sGC. We used pharmacological hsp90 inhibitors, an ATP-ase inactive hsp90 mutant, and heme-dependent or heme-independent sGC activators as tools to decipher the role of hsp90. Our findings suggest that hsp90 complexes with apo-sGC, drives heme insertion through its inherent ATPase activity, and then dissociates from the mature, heme-replete sGC. Together, this improves our understanding of sGC maturation and reveals a unique means to control sGC activity in cells, and it has important implications for hsp90 inhibitor-based cancer therapy.
The molecular chaperone 90-kDa heat-shock protein (hsp90) helps orchestrate fundamental processes including gene expression, signal transduction, innate immunity, oncogenesis, and influencing evolutionary phenotypes (1⇓⇓⇓–5). Hsp90 functions through its subdomain molecular motions and an inherent ATPase activity to help control client protein maturation, trafficking, and lifetime in cells (2, 6, 7). The molecular-level impacts of hsp90 on various client proteins are just beginning to be elucidated (8⇓⇓⇓⇓–13).
Soluble guanylyl cyclase (sGC) is the primary intracellular receptor for the signal molecule NO and is often found associated with hsp90 in cells (14⇓⇓–17). The NO-active sGC is a heterodimer of α1 and β1 subunits and contains a heme prosthetic group axially ligated to His105 in the β1 subunit (18). NO activates sGC by binding to its heme group, which enables sGC to catalyze conversion of GTP to cGMP and mediate many of the biological actions of NO (18⇓⇓–21). Accordingly, the pathogenesis of several diseases appears linked to insufficient sGC activity (22), which may be linked to oxidation of the sGC heme and/or buildup of heme-free sGC (23, 24). However, our understanding of sGC maturation in cells, processes regulating the levels of active sGC, and the potential roles of hsp90 is incomplete. Thus far, hsp90 has been reported to improve sGC NO response (25), and pharmacologically inhibiting hsp90 over a period of several days was shown to lower sGC activity by increasing its proteasomal degradation (26⇓⇓–29).
Recently, we reported that maturation of the hemeprotein-inducible NO synthase (iNOS) is hsp90-dependent (30). Here, the hsp90 associates with the heme-free (i.e., apo) iNOS in cells and then helps to drive heme insertion into iNOS through an inherent ATPase activity. Heme insertion into the related neuronal NOS also requires hsp90 (31). Given the pervasive nature of the sGC–hsp90 association, we wondered if hsp90 might play a similar role in enabling heme insertion into sGC during its maturation. To address this, we adopted a mammalian cell culture approach (30, 32) that allowed us to follow heme insertion into transiently transfected and endogenously expressed apo-sGC. We used pharmacological hsp90 inhibitors, an ATPase-inactive hsp90 mutant, and heme-dependent or heme-independent sGC activators as tools to decipher the role of hsp90. Our findings reveal that hsp90 plays a key role in promoting heme insertion into sGC. This hsp90-dependent process, which is essential for sGC maturation, may be a previously unknown but important feature to control sGC activity in cells, and has important implications regarding hsp90 inhibitor-based cancer therapy.
COS-7 cells were used to examine a possible role for hsp90 in heme insertion into apo-sGC. Normal or heme-deficient cells were transfected to express sGC-α1 (Myc-tagged) and sGC-β1 (V5-tagged) subunits, and then we compared cellular sGC expression levels and sGC activities in response to S-nitroso-N-acetyl-DL-penicillamine (SNAP), which releases NO and can activate only the heme-replete form of sGC. The normal and heme-deficient cell groups expressed similar amounts of sGC proteins (Fig. S1), indicating that apo-sGC-β1 was stable and able to accumulate in the heme-deficient cells. The cells transfected under normal conditions displayed good sGC activation by SNAP, whereas the transfected heme-deficient cells showed much less SNAP activation (Fig. 1). However, subsequent addition of hemin for 3 h converted the apo-sGC-β1 in the cells to the heme-replete form, as judged by their recovering a normal sGC activation by SNAP (Fig. 1). When hsp90 inhibitors radicicol or novobiocin were included during the hemin treatment, they prevented recovery of the SNAP response (Fig. 1). This suggests the hsp90 inhibitors may have prevented the cells from inserting heme into the apo-sGC-β1.
Effect of hsp90 inhibition during heme reconstitution of NO-dependent sGC activity. COS-7 cells were pretreated with or without SA for 48 h and then cotransfected with sGC α1 and β1 constructs for 42 h. This was followed by addition of vehicle or hemin (5 μM) to cells and hsp90 inhibitors radicicol or novobiocin (added for 30 min before hemin). After 3 h, cells were treated with SNAP (50 μM) for 5 min to activate sGC and then harvested, and sGC protein expression and cGMP product were measured. Bar graph shows GMP concentration in supernatants determined by ELISA. Values are mean ± SD of three independent experiments, each containing five replicates (***P < 0.001 by one-way ANOVA).
To further explore this possibility, we used sGC activators BAY 41-2272 and BAY 60-2770, which do not release NO but still activate the heme-containing sGC or the apo-sGC, respectively (33, 34). Fig. 2 shows that both compounds stimulated comparable sGC activation in sGC-transfected cells cultured under normal conditions, but, in cells made heme-deficient, the degree of sGC activation by BAY 60-2770 increased whereas that by BAY 41-2272 decreased by fourfold. This is consistent with apo-sGC being predominant in the heme-deficient cells, and BAY 60-2770 activating the apo-sGC (33, 34). Hemin reconstitution of the heme-deficient cells allowed BAY 41-2272 to markedly recover its sGC activation and diminished somewhat the ability of BAY 60-2770 to activate sGC (Fig. 2). Together, these data suggest a majority of the sGC in the heme-deficient cells was in the apo form, and confirm that functional heme insertion into the apo-sGC population is achieved upon exogenous hemin treatment.
Effect of inhibiting hsp90 during heme reconstitution on heme-dependent and heme-independent sGC activity. COS-7 cells were pretreated with or without SA for 48 h and then cotransfected with sGC α1 and β1 constructs, followed by hemin treatment with/without hsp90 inhibitor radicicol (added for 30 min before hemin). After 3 h, the cells were treated with a heme-independent (BAY 60-2770, 10 μM) or heme-dependent (BAY 41-2272, 10 µM) sGC activator for 30 min and then harvested. Bar graph shows cGMP concentration in supernatants determined by ELISA. Values are mean ± SD of three independent experiments, each containing five replicates (*P < 0.05, **P < 0.01, and ***P < 0.001 by one-way ANOVA).
In this context, we added the hsp90 inhibitor radicicol to the heme-deficient cells during their hemin treatment to see if it would alter the subsequent sGC activation in response to the two drugs. Radicicol still allowed good sGC activation by the heme-independent activator BAY 60-2770, but it inhibited the recovery of sGC activation by the heme-dependent activator BAY 41-2272 (Fig. 2). Western analysis showed that none of the activity differences could be ascribed to unequal sGC protein expression (Fig. S2). Together, our findings demonstrate that active hsp90 is needed for cells to insert heme into apo-sGC. Inhibiting hsp90 prevented heme insertion but otherwise did not hinder the apo-sGC from being activated by the heme-independent pathway.
We next sought to document the hsp90–sGC interaction in cells and determine whether it depends on the heme content of sGC-β1. Because sGC is a heterodimer of α1/β1 subunits, we also examined the importance of sGC α1 by transfecting cells with sGC α1/β1 together or with sGC-β1 alone. The cells were or were not made heme-deficient and then were transfected to express sGC α1/β1 together or only sGC-β1 under various conditions, and then we examined the hsp90-sGC protein interaction by immunoprecipitation by using anti-V5 (sGC-β1) antibody. Fig. 3A and Fig. S3 compare the amount of hsp90–sGC complex isolated from normal vs. heme-depleted cells, and also show the effect of incubation with hemin in the presence or absence of two different hsp90 inhibitors. There was fourfold greater hsp90-sGC association in heme-depleted cells compared with normal cells that were overexpressing sGC α1/β1 (Fig. 3A and Fig. S3B), despite their expressing equal levels of the sGC proteins. The hsp90–sGC interaction was greatly decreased after incubating the heme-deficient cells with hemin for 3 h to reconstitute their apo-sGC-β1 [i.e., plus succinyl acetone (SA) plus hemin]. The hsp90–sGC interaction was partly retained if the hsp90 inhibitor radicicol was present during the hemin incubation, but this was not the case with novobiocin. We observed similar results for the hsp90–sGC interactions in cells that overexpressed sGC-β1 alone (Fig. S3A), indicating that overexpression of the α-subunit is not required. Together, our data suggest that hsp90 associates predominantly with apo-sGC-β1, and this association weakens after heme insertion occurs.
Hsp90-sGC interactions before and after heme reconstitution of sGC-β1 and effect of hsp90 inhibition. (A) COS-7 cells were or were not made heme-deficient (deficiency indicated by +SA) and then were transfected 42 h to express sGC α1 and β1. Cells were then incubated 3 h with or without hemin (5 μM) and hsp90 inhibitors radicicol or novobiocin, and the supernatants were prepared, subjected to immunoprecipitation with anti-V5 (sGC-β1) antibody, and analyzed by SDS/PAGE and Western blotting. (A) Immunoprecipitated hsp90 with associated sGC-β1 and α1 subunits (input 10%) froms GCα1 plus β1 transfected cells. (B) COS-7 cells were transfected with sGC-β1 or sGC-β1H105F for 42 h and then harvested. Right: Immunoprecipitation shows hsp90 associated with sGC-β1 or sGC-β1H105F (input 10%). Bar graph shows densitometric quantification of the associated hsp90 bands. Data are mean of three independent experiments. (C and D) Heme-deficient COS-7 cells were transfected to express sGC-β1 for 42 h, followed by addition of hemin (5 μM) at various time points between 0 and 180 min, and then treated with or without SNAP (50 μM) for 5 min before harvesting. (C) Bar graph compares the level of hsp90 associated with immunoprecipitated sGC-β1 (input 10%) in each sample (data are mean of three independent experiments). (D) Supernatant cGMP concentration as determined by ELISA. Values are mean ± SD of three independent experiments, with each containing three replicates (*P < 0.05 and **P < 0.01, by one-way ANOVA; ns, not statistically significant).
To independently test if the hsp90–sGC interaction depends on sGC heme content, we separately expressed V5-tagged versions of the WT sGC-β1 and the heme-free mutant sGC-β1H105F in COS-7 cells and then immunoprecipitated by using anti-V5 antibody to compare their levels of hsp90 binding. As shown in Fig. 3B, the extent of hsp90 binding to the H105F mutant was more than threefold greater compared with WT sGC-β1. This result matches our seeing a greater hsp90–sGC association in heme-depleted cells, and argues that greater association is directly related to the level of heme deficiency of sGC-β1, rather than to any indirect effects of cellular heme depletion.
Finally, we examined the dynamics of the hsp90–sGC interaction in relation to the change in sGC heme content by monitoring the kinetics of both facets in heme-deficient cells that were transfected to express apo-sGC-β1 and then had hemin added. At each time point, we added SNAP for 5 min before cell harvest and then measured cell cGMP levels as an indicator of heme-dependent sGC activity (Fig. S4), which served to indirectly determine the heme content of sGC. [The time point of SNAP treatment was done for 5 min as it correlated with maximum cGMP accumulation (Fig. S4).] As shown in Fig. 3C and Fig. S5, there was a strong hsp90–sGC-β1 association at the initial time point followed by a sharp decrease within the 15-min period following hemin addition. The decrease in sGC–hsp90 interaction was somewhat greater for cells undergoing the SNAP treatment. The corresponding cGMP measures (Fig. 3D) showed that the apo-sGC became enzymatically active over the same time course after hemin was added, consistent with it incorporating heme within the first 15 min. The time courses show that the hsp90–apo-sGC association is relatively strong before heme insertion and then decreases rapidly with the onset of heme insertion into the sGC-β1 subunit.
To further examine the importance of hsp90 ATPase activity in enabling heme insertion into apo-sGC-β1, we used the D88N hsp90 mutant, which has no ATPase activity but is known to form in-cell complexes with client proteins such as endothelial NOS and iNOS (30, 35). We coexpressed the D88N mutant or WT hsp90β (both HA-tagged) with sGC-β1 in COS-7 cells for 42 h, and then activated the sGC by SNAP before cell harvest to measure heme-dependent sGC activation. As shown in Fig. 4A, cells expressing D88N-hsp90 had a 70% reduction in cGMP accumulation after SNAP treatment compared with cells expressing WT hsp90, without impacting the cell sGC-β1 protein level (Fig. S6A). This shows D88N-hsp90 acts as a dominant-negative inhibitor of sGC activation during its expression in cells, confirming the original observation by Miao et al. (35). We next studied if D88N-hsp90 expression would impact heme insertion into apo-sGC-β1. In this case, heme-deficient COS-7 cells were cotransfected with sGC-β1 and the WT hsp90 or D88N mutant and cultured for an additional 48 h, followed by hemin addition for 1 h, SNAP treatment for 5 min, and then cell harvest. Heme incorporation into apo-sGC was assessed by cGMP accumulation in response to SNAP. Coexpression of D88N-hsp90 completely inhibited heme incorporation into apo-sGC relative to the control without impacting the cell sGC protein level (Fig. 4B and Fig. S6B). These results confirm that the ATP-ase activity of hsp90 is critical for heme insertion into apo-sGC-β1.
ATPase-defective D88N-hsp90 down-regulates NO activation of sGC by antagonizing heme insertion. (A) COS-7 cells were cotransfected with sGC-β1 and hsp90β or D88N-hsp90β for 42 h, given SNAP (50 μM) for 5 min, and then harvested. (A) cGMP concentration in the supernatants as assayed by ELISA. (B) COS-7 cells were made heme deficient with SA before 42 h cotransfection with sGC-β1 plus hsp90β or D88N-hsp90, followed by 1 h hemin addition (5 μM) and then 5 min SNAP treatment. (B) cGMP concentrations in supernatants as assayed by ELISA. Values are mean ± SD of three independent experiments, with each containing three replicates (*P < 0.05 and **P < 0.01, one-way ANOVA; ns, not statistically significant).
To test if hsp90 is required for heme insertion in a more natural setting, we used a rat lung fibroblast cell line (RLF-6) that expresses high levels of endogenous sGC (36). RLF-6 cells were cultured for 3 d with SA to make them heme-deficient. This did not lower sGC protein expression (Fig. S7) or its activation by heme-independent activator BAY 60-2770, but greatly lowered sGC activation by SNAP (Fig. 5A), indicating apo-sGC had built up in the heme-deficient RLF-6 cells. Further culture for 1 h with hemin restored sGC activation by SNAP (Fig. 5A), consistent with heme insertion taking place during this time period, and radicicol inhibited this recovery. In related experiments, we transfected RLF-6 cells with WT or D88N hsp90β and then compared their protein expression, association with native sGC, and sGC activation by SNAP. Both hsp90 proteins were expressed well in RLF-6 cells (Fig. 5B). Pull-down studies using an anti-sGC antibody showed that the HA-tagged WT and D88N hsp90 proteins both associated with sGC-β1 in the RLF-6 cells (Fig. 5C). There was a decrease in hsp90–sGC-β1 interaction during sGC activation by SNAP that was not evident for the D88N hsp90–sGC-β1 complex, which may indicate a more stable complex formed between D88N hsp90 and sGC-β1. D88N-hsp90 expression diminished SNAP activation of sGC (Fig. 5D). Our results show that hsp90 and its ATP-ase activity are also required for heme insertion into the sGC-β1 that is endogenously expressed in lung fibroblast cells.
Hsp90 drives heme insertion into endogenous sGC expressed in RLF-6 cells. (A) RLF-6 cells were pretreated with or without SA for 72 h, incubated 3 h with hemin (5 μM) with or without hsp90 inhibitors radicicol or novobiocin, and then treated with SNAP (50 µM) for 5 min or with heme-dependent (BAY 41-2272, 10 μM) or heme-independent (BAY 60-2770, 10 µM) sGC activators for 30 min before being harvested. (A) Cell supernatant cGMP concentrations determined by ELISA. (B–D) RLF-6 cells were transfected with WT or D88N-hsp90 for 42 h, treated with SNAP for 5 min, and harvested. (B) Representative Western blot shows protein expression levels (sGC-β1, hsp90β or D88N-hsp90, and total hsp90) in supernatants. (C) Immunoprecipitations depict hsp90β or D88N-hsp90β associated with sGC-β1 (input 10%) with or without SNAP treatment. (D) cGMP concentration in supernatants as determined by ELISA. Data are mean ± SD of three independent experiments, with each containing three replicates (*P < 0.05 and **P < 0.01, one-way ANOVA).
Heme insertion into apo-sGC-β1 is essential for maturation of the NO-responsive enzyme. We found that heme insertion into apo-sGC-β1 depends on hsp90. An hsp90 requirement was apparent whether we followed incorporation of exogenously added heme or endogenous cell-derived heme into apo-sGC-β1, and it held for sGC that was transiently expressed and for sGC that was expressed naturally in a lung fibroblast cell line. sGC is the third protein whose heme insertion has now been found to be hsp90-dependent (the others are iNOS and neuronal NOS) (30, 31). Given that sGC and NOS enzymes have markedly different protein structures and heme environments (37, 38), our findings suggest that hsp90 may have a more general role in heme protein maturation than was previously realized.
In the present study, hsp90 associated most strongly with apo-sGC-β1 in cells and the association weakened or fell apart when heme had become inserted. Likewise, there was greater hsp90 association with heme-free mutant sGC-β1H105F compared with WT. Although their association did not require hsp90 to have an intact ATPase activity, as judged from our results with the use of radicicol, novobiocin, and an ATPase-defective hsp90 mutant, an intact ATPase activity was essential to drive heme insertion into apo-sGC. The fact that radicicol or novobiocin had no “downstream” impact on the NO-dependent activation of mature sGC or on activation of apo-sGC by the heme-independent stimulator BAY 60-2770, implies that hsp90 restricts its participation to the heme insertion step during the protein maturation phase and is not involved in activating the catalysis of heme-replete or heme-free sGC. Together, our data suggest a model for hsp90 function that is illustrated in Fig. 6. Interestingly, this model for hsp90 mimics its proposed role in driving heme insertion into apo-iNOS (30). The similarity implies hsp90 may operate through a common mechanism to target and stabilize heme-free forms of client heme proteins, and then enable their maturation by driving heme-insertion in an ATP-dependent process.
Model for hsp90 function during maturation of active sGC. Hsp90 binds to heme-free sGC in cells, and this complex likely interacts with a heme carrier protein (HCP). Hsp90 then uses its ATPase activity to help drive heme insertion into apo-sGC-β1. This process is blocked by hsp90 inhibitors radicicol or novobiocin and antagonized by the ATPase-defective D88N-hsp90. Hsp90 then dissociates from the heme-replete sGC, whose catalysis can now be activated by NO. In contrast, sGC activation by heme-independent sGC activators does not require hsp90. Further details are provided in the text.
Some apo-sGC appeared to be present in our cell cultures under normal conditions, as judged by BAY 60-2770 causing significant additional stimulation of sGC activity relative to SNAP alone (Fig. 2). Indeed, our study suggests that observing any hsp90-sGC association in cells may itself indicate that heme-free sGC is present. Interestingly, Ignarro et al. purified sGC from bovine lung in a heme-deficient form (39). Overall, these concepts support the claim by Roy et al. (40) that heme-free sGC normally exists in cells and that BAY 60-2770 can activate the heme-free sGC (34).
Our cell culture systems were purposely designed to test how short-term hsp90 inhibition (0–3 h) impacts a critical posttranslational step in sGC maturation, namely heme insertion into apo-sGC-β1. However, in regard to hsp90 inhibitor therapy, one must also consider longer-term effects. Previous studies that investigated how hsp90 inhibitors impact sGC activity all incubated cells with the inhibitors overnight to 1-d time periods (25, 26, 28). Inhibiting hsp90 over these longer periods diminished sGC activity mainly by shunting sGC toward the proteasomal degradation pathway, thus lowering its overall expression level in cells. Increased degradation may also explain why long-term hsp90 inhibition diminishes the expression levels of several hsp90 client proteins (41). In contrast, our study reveals a more immediate mechanism by which hsp90 inhibition can diminish sGC activity, namely by preventing heme insertion into apo-sGC-β1. Thus, over long periods of hsp90 inhibition, we surmise that any newly translated sGC-β1 protein would fail to mature and would remain in its heme-free form. Whether apo-sGC-β1 is inherently shorter-lived and degraded faster over a period of days in cells, or whether hsp90 inhibition works in a less direct manner to alter the sGC-β1 level, are important questions that can now be addressed.
There is a tremendous interest in developing hsp90-targeted cancer therapeutic agents (42). Whether newly developed hsp90 inhibitors would block heme insertion into apo-sGC-β1 or the NOS enzymes is still untested, but the fact that most drugs are designed to target the hsp90 ATPase activity (shown here to be essential for heme insertion) suggests the possibility is real and may have important implications for the development of hsp90 inhibitors for cancer therapy. We predict that inhibiting hsp90-based heme insertion into both the sGC and NOS enzymes would cause a synergistic lowering of cGMP levels because it would diminish production of the signal molecule (i.e., NO) and diminish the responsiveness of the amplifier (i.e., sGC) toward NO. Our study shows that one could circumvent the potential impact of hsp90 inhibitors by administering heme-independent sGC activators like BAY 58-2667, which can still activate heme-free sGC and thus allow it to function in the signal cascades, leading to vasorelaxation or other cGMP-dependent processes. Our work provides a platform to explore these concepts.
BAY 60-2770 and BAY 41-2272 were provided by Bayer, SNAP was purchased from Cayman Chemicals, and lipofectamine was purchased from Invitrogen. All other chemicals were purchased from Sigma or Fisher. Stock solutions of novobiocin were prepared in water whereas 3-isobutyl-1-methylxanthine (IBMX) and radicicol were dissolved in DMSO. Stock solutions of hemin were freshly prepared in 0.1 N NaOH (32). cDNAs for human hsp90β and D88N-hsp90β mutant were gifts from Bill Sessa (Yale University, New Haven, CT). cDNAs for sGC α1, β1, and sGC-H105F β1 mutant were gifts from Andreas Papapetropoulos (University of Patras, Patras, Greece). Monkey COS-7 and rat RLF-6 cells were purchased from American Type Culture Collection. Rabbit polyclonal hsp90 antibody was purchased from Cell Signaling Technology. Anti-V5 antibody was purchased from Invitrogen, and anti-Myc and anti-HA tag antibodies were purchased from Sigma. A cGMP ELISA kit was obtained from Cell Signaling Technology.
All cell lines were grown in 100 mm tissue culture dishes. COS-7 cells were cultured in DMEM containing 10% (vol/vol) FBS and 5,000 U of penicillin-streptomycin. RLF-6 cells were cultured in Ham F-12 K media containing l-glutamine and pyruvate, 5,000 U/L of penicillin/streptomycin and 20% (vol/vol) FBS. Heme depletion was achieved by culturing cells with 400 μM SA for 48 h before transfection or activation of sGC (30, 32). RFL-6 cells expressing endogenous sGC were activated by NO donor SNAP (50 μM) or with heme-dependent (BAY 41-2272; 10 μM) or heme-independent activators (BAY 60-2770; 10 μM) from 0 to 30 min as indicated. At the point of cell harvest, the monolayers were washed twice with 4 mL cold PBS solution plus 1 mg/mL glucose, collected by scraping in 500 μL lysis buffer [40 mM 3-(4-[2-hydroxyethyl]-1-piperazinyl)propanesulfonic acid buffer, pH 7.6, 10% glycerol, 3 mM DTT, 150 mM NaCl, and .5% Nonidet P-40). Cells were lysed by three cycles of freeze/thawing, the lysates were centrifuged for 30 min at 4 °C, and the resulting supernatants were assayed for total protein content by using a Bio-Rad protein assay kit.
Cultures (50–60% confluent) of COS-7 or RLF-6 cells or SA-pretreated COS-7 cells were transfected for 42 h with expression constructs of sGC α1, β1, sGC-H105F β1, hsp90β, or D88N-hsp90β in various combinations using Lipofectamine [5 μg DNA was used for single transfections and 10 μg DNA (i.e., 5µg plus 5 μg) was used for cotransfections]. Cells were then treated with 0.5 mM IBMX for 10 min followed by SNAP or with Bayer compounds (i.e., BAY 60-2770 or BAY 41-2272) for various times (0–30 min) before being harvested. In some cases, cells were pretreated with SA (400 μM) before transfection with sGC expression constructs. After 42 h, cells were given radicicol (20 μM), novobiocin (500 μM), or vehicle for 30 min and then treated with hemin for 3 h. Cell sGC was then activated by adding SNAP or Bayer compounds at the indicated time points before harvesting. In all cases, the cells were treated with cycloheximide (10 mg/mL) for 30 min before hemin addition. Transfections were carried out in duplicate plates, and the experiments were repeated at least three times.
For Western blotting, 30 μg of cell supernatants were loaded on 8% SDS/PAGE gels. Proteins were electroblotted to PVDF membrane and probed with the respective antibodies. For immunoprecipitation, 500 μg of the total cell supernatants was precleared with 20 µL of protein G–sepharose beads (Amersham) for 1 h at 4 °C, beads were pelleted, and the supernatants were incubated overnight at 4 °C with 3 μg of anti-HA or anti-iNOS antibody. Protein G–sepharose beads (20 μL) were then added and incubated for 1 h at 4 °C. The beads were microcentrifuged (3,220 × g), washed three times with wash buffer (50 mM Hepes, pH 7.6, 100 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40), and then boiled with loading buffer for SDS-PAGE gels and subsequent Western blotting.
sGC activity in cell supernatants was measured as a function of cGMP concentration by using a cGMP assay kit (Cell Signaling Technology).
We thank Drs. W. Sessa and A. Papapetropoulos for providing hsp90 and sGC constructs and Dr. Hans-Peter Stasch (Bayer, Leverkusen, Germany) for providing sGC activators BAY 60-2770 and BAY 41-2272. This work was supported by National Institutes of Health Grants GM51491 and HL076491 (to D.J.S.).
Author contributions: A.G. and D.J.S. designed research; A.G. performed research; A.G. and D.J.S. analyzed data; and A.G. and D.J.S. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission. R.B. is a guest editor invited by the Editorial Board.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1205854109/-/DCSupplemental.
