NO and cGMP have antigrowth and anti-inflammatory effects on the vessel wall in response to injury. It is well established that after vascular injury proinflammatory cytokines are involved in vascular wall remodeling. The purpose of this study was to ascertain the signaling mechanisms involved in cGMP-dependent protein kinase (PKG) suppression by inflammatory cytokines in primary bovine aortic vascular smooth muscle cells (VSMC). Interleukin (IL)-Iβ, tumor necrosis factor (TNF)-α, and LPS decreased the mRNA and protein levels of PKG in VSMC. IL-Iβ, TNF-α, and LPS increased inducible nitric oxide synthase (iNOS) expression and cGMP production. Treatment of cells with selective inhibitors of iNOS or soluble guanylate cyclase (sGC) reversed the downregulation of PKG expression induced by cytokines and LPS. The NO donor (Z)-1-[2-(2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-diolate (DETA NONOate) and 3-(5-hydroxymethyl-2-furyl)-1-benzylindazole (YC-1), a NO-independent sGC activator, decreased PKG mRNA and protein expression in bovine aortic VSMC. Cyclic nucleotide analogs [8-(4-chlorophenylthio)guanosine 3′,5′-cyclic monophosphate (CPT-cGMP) and 8-(4-chlorophenylthio)adenosine 3,5′-cyclic monophosphate (CPT-cAMP)] also suppressed PKG mRNA and protein expression. However, CPT-cAMP was more effective than CPT-cGMP in decreasing PKG mRNA levels. Selective inhibition of PKA with the Rp isomer of 8-(4-chlorophenylthio)adenosine 3′,5′-cyclic monophosphorothioate (Rp-8p-CPT cAMPS) prevented the downregulation of PKG by LPS. In contrast, the Rp isomer of 8-(4-chlorophenylthio)guanosine 3,5′-cyclic monophosphorothioate (Rp-8p-CPT cGMPS; inhibitor of PKG) had no effect on LPS-induced inhibition of PKG mRNA and protein expression. These studies suggest that cross-activation of PKA in response to iNOS expression by inflammatory mediators downregulates PKG expression in bovine aortic VSMC.
- vascular injury
- nitric oxide
cyclic gmp (cGMP) and cGMP-dependent protein kinase (PKG) have numerous effects on the vessel wall, including vasodilation, vascular permeability, inhibition of platelet aggregation, and inhibition of smooth muscle cell proliferation (11, 16, 17, 24). PKG plays a key role in mediating NO-dependent signaling in vascular tissue. PKG is a homodimeric serine-threonine protein kinase that belongs to the very large protein kinase family. Two types of PKG are found in mammalian cells, type I and type II, and they are the products of two distinct genes (12, 19, 34). Type I is expressed as two isoforms, PKG-Iα and PKG-Iβ, generated by alternate mRNA splicing of the type I gene. Although PKG-I is widely distributed in mammalian cells, it is most abundant in smooth muscle cells (SMC), platelets, and Purkinje cells (15, 30, 32). Activation of PKG-I mediates smooth muscle relaxation through lowering of intracellular Ca2+ levels in vascular smooth muscle (2, 20, 21) or inhibition of myosin light chain (MLC) phosphorylation (28, 29). Although other downstream mechanisms have been proposed for PKG in the vasculature, the end result of this activation of the pathway is relaxation of SMC and the regulation of vascular tone.
In several pathophysiological situations, including atherosclerosis and restenosis after organ transplantation or balloon angioplasty, the intimal layer of the vessel thickens. A key component of this process is excessive vascular SMC (VSMC) proliferation and migration from the media to the intima. In the intima, VSMC secrete abundant extracellular matrix proteins, thus further expanding the intimal area. In culture, VSMC not only become highly proliferative but also change phenotype, such that contractile protein expression decreases and extracellular matrix protein increases (6, 18, 23). In some VSMC, the loss of PKG expression parallels the decline in contractile protein expression and the increase in extracellular matrix protein synthesis. Our laboratory (4, 10) has demonstrated that transfection of PKG cDNA (holoenzyme or active catalytic domain) into PKG-deficient, synthetic, and secretory VSMC results in the partial restoration of the contractile phenotype. These data suggest that PKG may regulate VSMC phenotype and may have a possible role in vascular pathophysiology. The mechanisms underlying the decrease in PKG expression and phenotypic modulation in VSMC in vitro and in vivo are not well understood.
The response to injury by the vessel wall is a complex process involving the influx of several cell types (leukocytes, lymphocytes, and smooth muscle) and culminating in an inflammatory response that increases the fibroproliferative activity of VSMC. The inflammatory response generates a cascade of events, including upregulation of inducible nitric oxide synthase (iNOS) in immune cells, endothelial cells, and VSMC by proinflammatory cytokines such as interleukin (IL)-1β and tumor necrosis factor (TNF)-α (3, 14, 31). Cytokine-induced increases in iNOS expression cause the subsequent production of pathophysiologically high levels of NO and cGMP in the vascular wall. Studies have reported that high concentrations of NO donor drugs and cyclic nucleotide analogs reduce PKG protein and mRNA expression in low-passaged bovine aortic VSMC (27) and pulmonary veins of newborn lambs (13). Other studies have shown that PKG expression is suppressed in arterial VSMC after balloon catheter injury in vivo (1, 26). These findings suggest that downregulation of PKG expression in response to injury and inflammation could contribute to the vascular fibroproliferative response. However, the mechanism of PKG downregulation is still unknown. Likewise, there are no studies addressing the role of inflammatory cytokines or their signaling pathways on PKG expression in VSMC. The goal of the present study was to elucidate the mechanism of PKG suppression under conditions that mimic the inflammatory response that occurs after vascular injury.
MATERIALS AND METHODS
Human recombinant IL-1β and TNF-α were purchased from R&D Systems (Minneapolis, MN). LPS, amphotericin B, soybean trypsin inhibitor (SBTI), deoxyribonuclease I (DNase I), and monoclonal anti-β-actin antibody were obtained from Sigma (St. Louis, MO). 8-(4-chlorophenylthio)adenosine 3′,5′-cyclic monophosphate (CPT-cAMP), 8-(4-chlorophenylthio)guanosine 3′,5′-cyclic monophosphate (CPT-cGMP), the Rp isomer of 8-(4-chlorophenylthio)adenosine 3′,5′-cyclic monophosphorothioate (Rp-8p-CPT-cAMPS), and the Rp isomer of 8-(4-chlorophenylthio)guanosine 3′,5′-cyclic monophosphorothioate (Rp-8p-CPT cGMPS) were purchased from Biolog (San Diego, CA). Fetal bovine serum (FBS), Dulbecco's modified Eagle's medium (DMEM), penicillin-streptomycin, gentamicin, and salmon sperm DNA were purchased from GIBCO (Grand Island, NY). Collagenase III and elastase were acquired from Worthington Biochemicals (Freehold, NJ). (Z)-1-[2-(2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-diolate (DETA NONOate), 3-(5-hydroxymethyl-2-furyl)-1-benzylindazole (YC-1) and 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) were obtained from Alexis Biochemicals (San Diego, CA). Nitrocellulose and iQ SYBR green supermix were obtained from Bio-Rad (Hercules, CA). The Random Prime DNA labeling kit and QuickHyb were purchased from Stratagene (La Jolla, CA). The polyclonal carboxy terminus anti-PKG antibody was purchased from Stressgen (Carlsbad, CA). All other reagents were purchased from Fisher Biochemicals (Pittsburgh, PA) or VWR Scientific (West Chester, PA).
Primary culture and subculturing of bovine aortic VSMC.
Aortas from freshly killed cattle were generously provided by Richardson Meat (Tuscaloosa, AL) and Kimbrell Slaughterhouse (Saraland, AL). Bovine aortic VSMC were prepared from the anterior region below the aortic arch. Loose fat and connective tissue were carefully stripped, the aorta was opened, and the endothelium was removed by gentle scraping. The tunica adventitia was then removed, and medial smooth muscle strips were excised. The strips were rinsed three times in a wash medium of DMEM containing 25 mM HEPES, pH 7.4, 100 μg/ml gentamicin, 2.5 μg/ml amphotericin B, and 1 mg/ml bovine serum albumin (BSA). Medial strips were minced with scissors and incubated in digestion medium (DMEM containing 200 U/ml collagenase III, 1 mg/ml elastase, 0.1 mg/ml SBTI, and 50 μg/ml DNase I). The minced smooth muscle was digested for 6–8 h until a single-cell suspension was obtained. Suspended cells were plated (7,500 cells/cm2) in DMEM containing 10% FBS, 50 μg/ml gentamicin, 2.5 μg/ml amphotericin B, and 200 μg/ml penicillin-streptomycin. Cells were grown in a 10% CO2 incubator at 37°C. For most of the experiments described, only freshly plated, non-subcultured cells were used. For subculturing of cells, monolayers were rinsed once with phosphate-buffered saline (PBS) and treated briefly with trypsin (1.25% in PBS). The cell suspension was added to 4 ml of DMEM containing FBS and split into separate culture dishes (3,500 cells/cm2).
Protein extraction and immunoblotting.
Cell monolayers were rinsed and scraped with ice-cold PBS and centrifuged at 14,000 g for 10 min. Cell pellets were homogenized in (in mM) 20 potassium phosphate, pH 6.8, 150 NaCl, 0.1 phenylmethylsulfonyl fluoride, and 10 benzamidine with 10 μM leupeptin, 10 μM pepstatin A, and 1% Triton X-100. The homogenate was sonicated at 20 duty cycles, 50% power for 20 s, and centrifuged at 10,000 g for 5 min at 4°C. The supernatants were removed, and protein was determined with BSA as the standard in the Bradford assay. The supernatants (10 μg) were mixed with an equal volume of 2× sample loading buffer (125 mM Tris·HCl, pH 6.9, 4% SDS, 50% glycerol, 0.02% bromophenyl blue, and 1.4% β-mercaptoethanol). Samples were boiled for 5 min and applied to 10% SDS-polyacrylamide gels. The proteins were resolved by electrophoresis and transferred to nitrocellulose. The blots were blocked in 5% nonfat milk and probed with anti-PKG and anti-β-actin antibodies as previously described (10). Densities of the bands were quantified by computer-assisted analysis with the Image Gauge software program and expressed as area units. All values were normalized to values for β-actin, which served as the housekeeping gene for quantitative analysis.
Measurement of cyclic nucleotide levels.
Primary bovine aortic VSMC were treated with cytokines. After treatment, the medium was aspirated and 0.1 N HCl-50% methanol was added to the monolayer. After 5 min, the acidified extracts were collected and evaporated in the SpeedVac. The dried extracts were suspended in 500 μl of water, and cGMP was quantified by radioimmunoassay as previously described (5).
Total RNA extraction and cDNA synthesis.
Total cellular RNA was isolated from cell monolayers with a Qiagen RNeasy Mini Kit (Valencia, CA) as instructed by the manufacturer's protocol. The integrity of RNA was confirmed by spectrophotometric analysis [optical density (OD) 260/280] and agarose gel electrophoresis. A total of 1 μg of RNA (OD 260/280 > 1.8) was reverse transcribed with reagents from Applied Biosystems (Foster City, CA). Each 25-μl reaction contained 1× reaction buffer, 2.5 μM random hexamers, 500 μM dNTPs, 5.5 mM MgCl2, 0.4 U/μl RNase inhibitor, and 1.25 U/μl Multiscribe reverse transcriptase. The reactions were incubated at 25°C for 10 min, 45°C for 30 min, and 95°C for 5 min to inactivate the enzyme and stop the reaction.
Northern blot analysis.
RNA (15 μg) was resolved on 1% formaldehyde agarose gels, transferred to Nytran Plus membranes, and UV cross-linked. After prehybridization in QuickHyb, the membranes were incubated with 32P-labeled cDNA probes (∼1 × 106 cpm/ml hybridization solution) in 100 μg/ml salmon sperm DNA. Hybridization was conducted at 42°C for at least 3–4 h. Membranes were then washed with 2× SSC and 0.1% SDS at room temperature, 2× SSC and 0.1% SDS at 42°C, and 0.1× SSC and 0.1% SDS at 65°C, consecutively. Blots were developed, and densities of the bands were quantified by computer-assisted analysis with the Image Gauge software program and expressed as area units. All values were normalized to values for GAPDH, which served as the housekeeping gene for quantitative analysis. None of the treatments affected the levels of GAPDH mRNA. The probes for bovine iNOS and GAPDH were generated as purified insert DNAs and labeled with the Random Prime DNA labeling kit.
Absolute quantitative real-time PCR.
Two microliters of reverse transcription product were used for subsequent amplification. The primer sets were designed on the bovine sequences of PKG-I isoforms and β-actin (Table 1). Throughout this article, PKG refers specifically to the PKG-I isoform. Bovine PKG cDNA was ligated into the Bluescript vector as previously described and used as a standard in the reaction (4). The β-actin standard was generously provided by Dr. Steven P. Suchyta (Michigan State University, East Lansing, MI). Briefly, the bovine sequence of β-actin was ligated into pCMV SPORT6, and serial dilutions were conducted and served as a housekeeping gene for quantitative analysis. PKG and β-actin primers were generated by MWG Biotech (High Point, NC). Each reaction contained Bio-Rad's 2× SYBR master mix, and primers for PKG or β-actin were added to a final concentration of 500 or 100 nM, respectively. Cycling conditions were 95°C for 3 min followed by 40 cycles for 95°C for 30 s and 60°C for 30 s. PKG and β-actin mRNA were quantified by real-time PCR with the Bio-Rad iCycler Real Time PCR, and the data generated were analyzed with the accompanying software. A threshold value and baseline cycles were assigned by the analysis software. This threshold setting was adjusted to obtain the best correlation coefficient for the standard curve. The number of cycles required for each sample to cross the threshold was compared with the number of cycles required for the standards to cross the threshold. This threshold setting was based on finding the best correlation coefficient (≥0.990) of the standards. The ratio between the quantity of PKG and β-actin was calculated and expressed as percentage of control.
All data are expressed as means ± SE. Each experiment was performed in duplicate with at least n = 3 independent experiments. Statistical differences between values were evaluated by Student's t-test. Comparisons among multiple groups were made by one-way ANOVA for multiple comparisons. Statistical significance was then determined by Tukey's posttest with GraphPad Prism software, with significant probabilities at P < 0.05.
PKG expression is suppressed in VSMC treated with cytokines and LPS.
Freshly isolated bovine aortic VSMC retain normal tissue levels of expression of PKG, soluble guanylate cyclase (sGC), and smooth muscle-specific genes such as smooth muscle myosin heavy chain in primary culture (data not shown). For these reasons, we used freshly isolated, unpassaged cells to examine the role of inflammatory cytokines and LPS on PKG expression. As shown in Fig. 1, treatment of bovine aortic VSMC with IL-1β (10 ng/ml), TNF-α (10 ng/ml), or LPS (100 μg/ml) decreased PKG mRNA and protein levels as determined by real-time PCR and Western blotting. Using real-time PCR, we were able to determine the absolute number of mRNA transcripts encoding PKG-I in bovine aortic VSMC, which was 4.4 × 10−5 copies/μl (normalized to β-actin) for control. The suppression of PKG mRNA was detected at 4 h of treatment with the cytokines (20–25%) and was maximal after 24 h (80–85%). At the protein level, cytokines or LPS caused a 70–80% reduction of PKG expression within 48 h of treatment (Fig. 1B). None of the cytokine treatments affected the expression of β-actin mRNA and protein levels in the cells. Prolonged treatment (72–96 h) maintained suppressed PKG expression in bovine aortic VSMC. Concentrations of IL-1β and TNF-α as low as 1 ng/ml and of LPS as low as 10 μg/ml decreased PKG expression (data not shown).
It is well documented that proinflammatory cytokines such as IL-1β and TNF-α and LPS increase iNOS expression in various types of cells, including VSMC. To determine whether these biologically active compounds increase iNOS expression in bovine aortic VSMC, we performed Northern blot analysis in cells treated with various combinations of cytokines. As shown in Fig. 2, LPS, TNF-α, or IL-Iβ increased iNOS mRNA expression in VSMC. All three iNOS transcripts were increased, as described previously (3). To assess whether decreased levels of PKG expression were caused by increased iNOS expression, bovine aortic VSMC were incubated with iNOS inhibitors and PKG mRNA and protein levels were quantified. The addition of nitro-l-arginine methyl ester (l-NAME; 1 mM), a nonselective iNOS inhibitor, or 1400W (10 μM), a selective iNOS inhibitor, blocked the capacity of TNF-α or LPS to downregulate PKG mRNA at 24 h (Fig. 3A). Similarly, l-NAME and 1400W inhibited the ability of TNF-α or LPS to suppress PKG protein expression in bovine aortic VSMC at 48 h (Fig. 3B).
Effects of NO donor, ODQ, and YC-1 on PKG expression.
Because the iNOS inhibitors were effective in blocking the suppression of PKG expression induced by IL-1β, TNF-α, or LPS, it was of interest to examine the effect of a NO donor drug on PKG mRNA and protein expression. Bovine aortic VSMC were stimulated with DETA NONOate, which has a half-life of ∼17 h and allows a chronic, sustained delivery of NO. Treatment with DETA NONOate resulted in decreased PKG mRNA (Fig. 4A) and protein (Fig. 4B) levels in a concentration-dependent manner. Examination of cell viability showed that the NO donor did not have any cytotoxic effect (data not shown). The major receptor protein for NO in VSMC is sGC, which catalyzes the conversion of GTP to cGMP. To determine the role of sGC in PKG suppression, bovine aortic VSMC were preincubated with the sGC inhibitor ODQ for 30 min followed by LPS stimulation. ODQ (10 μM) prevented the ability of LPS to downregulate PKG mRNA expression (Fig. 5A). The ability of ODQ to block the downregulation of PKG expression was also observed at the protein level (Fig. 5B). In addition, the non-NO-dependent sGC activator YC-1 decreased PKG protein expression by itself (Fig. 6). Together, these results suggest that sustained cGMP generated by persistent activation of sGC appears to inhibit PKG expression at both mRNA and protein levels in primary bovine aortic VSMC.
Cyclic nucleotide analogs suppress PKG expression in VSMC.
It has been proposed that cytokines such as IL-1β and TNF-α as well as LPS increase cGMP levels in VSMC (3). As shown in Table 2, basal levels of cGMP were increased >20-fold by IL-1β, TNF-α, and LPS even after 24 h. This increase in cGMP production was inhibited by iNOS inhibitors (data not shown).
To determine the effects of cyclic nucleotides on PKG expression, bovine aortic VSMC were treated with cell-permeant and phosphodiesterase (PDE)-resistant cyclic nucleotide analogs. As shown in Fig. 7, both cyclic nucleotide analogs suppressed PKG mRNA synthesis. However, there was a clear difference in the degree of inhibition produced by these compounds. The addition of CPT-cGMP (10 μM) only slightly reduced PKG mRNA expression (20–25%), whereas CPT-cAMP (10 μM) dramatically decreased PKG mRNA expression (60–65%). This suggests that CPT-cAMP may be more potent than CPT-cGMP in downregulating PKG mRNA expression. Thus we examined whether the suppression of PKG mRNA expression might have involved the activation of PKA in response to both cyclic nucleotide analogs. Previously, we (9) showed that cytokines that induce iNOS expression also increase the activity of PKA by measuring the in situ activation of the kinase with the activity ratio assay. Thus this possibility was further tested with selective cyclic nucleotide kinase antagonist compounds, the cyclic nucleotide phosphorothioate derivatives Rp-8p-CPT cGMPS and Rp-8p-CPT cAMPS. The addition of the PKG inhibitor Rp-8p-CPT cGMPS had no effect on LPS-induced suppression of PKG expression (Fig. 8A). In contrast, bovine aortic VSMC treatment with Rp-8p-CPT cAMPS, a PKA inhibitor, blocked the capacity of LPS to downregulate PKG mRNA (90–95%). Examination of PKG protein levels showed similar results, with the PKA antagonist reversing the downregulation of PKG by LPS (Fig. 8B). Together, these results suggest that it is the activation of PKA by inflammatory mediators that downregulates PKG expression in primary bovine aortic VSMC.
Studies from several laboratories have shown that the NO/cGMP/PKG pathway has an inhibitory effect on VSMC proliferation and ECM protein production, two hallmarks of vascular diseases such as atherosclerosis and restenosis after injury (4, 6–10). In culture, VSMC often express lower levels of either sGC or PKG (4, 7, 26). We previously reported (4) that rat aortic VSMC that are deficient in PKG expression and have undergone phenotypic transition can be restored to a more contractile state simply by restoration of PKG expression. These findings suggest that a functional NO/cGMP/PKG pathway is important for the maintenance of a normal contractile phenotype of adult VSMC and that the suppression of PKG expression in VSMC may play an important pathophysiological role in phenotypic modulation of VSMC to a fibroproliferative state.
Because PKG activity leads to the expression of genes associated with a contractile-like phenotype and the suppression of genes associated with a synthetic state of VSMC, we predict that cytokine-induced suppression of PKG leads to a fibroproliferative wound healing phenotype for VSMC. Anderson et al. (1) and Sinnaeve et al. (26) showed in two animal models that PKG expression is decreased after balloon coronary artery or carotid artery injury. In the case of pig coronary artery, loss of PKG expression was correlated with an increase in iNOS expression (1). Furthermore, in vivo gene transfer of the constitutively active PKG catalytic domain reduced neointima formation in response to balloon catheter injury in rat carotid arteries and reduced in-stent restenosis in swine as well (26). These findings demonstrate the functional importance of PKG expression in the response to injury. However, the mechanism underlying PKG suppression is not well understood.
Hence, this study was initiated to determine the mechanisms by which inflammatory cytokines lead to PKG suppression in VSMC. Two early inflammatory mediators, IL-Iβ and TNF-α, as well as LPS were found to be potent inhibitors of PKG mRNA and protein expression in primary, freshly isolated bovine aortic VSMC. Notably, IL-Iβ and TNF-α are induced and released by leukocytes and VSMC at the sites of vascular lesions in response to injury and inflammation in vivo. LPS, on the other hand, has potent inflammatory properties by itself and can induce the expression of other inflammatory mediators. LPS, like IL-Iβ and TNF-α, suppressed PKG mRNA and protein expression in bovine aortic VSMC.
It is well established that inflammatory mediators alter the expression of a number of genes in many target cells. One of the more prominently expressed genes induced by IL-Iβ and TNF-α is iNOS. It has been proposed by several laboratories that high NO production in response to iNOS expression mediates many of the inflammatory effects of IL-Iβ and TNF-α (3, 31). For instance, persistent, high-output NO, by itself or in combination with reactive oxygen species, may alter cell function. The findings reported in this study suggest another important role for iNOS and NO production in the inflammatory response, namely the suppression of PKG expression, which would facilitate the fibroproliferative response of VSMC to injury and inflammation. The critical role for iNOS-mediated NO production in suppressing PKG expression is supported by the finding in the present study that both the nonselective NOS inhibitor l-NAME and the selective iNOS inhibitor 1400W reversed the suppression of PKG expression induced by inflammatory cytokines.
The mechanism by which iNOS expression leads to suppression of PKG expression appears to be dependent on NO production. Gao et al. (13) showed that chronic delivery of NO significantly reduces PKG activity and protein and mRNA levels in vessels. NO has both cGMP-dependent and cGMP-independent effects on cells, especially when present at high concentrations such as during the inflammatory response. Our results indicate that NO-dependent activation of sGC is responsible, in part at least, for the suppression of PKG expression, because the sGC inhibitor ODQ effectively prevented the capacity of cytokines and LPS to suppress PKG expression. As further support of this notion, the NO-independent sGC activator YC-1 suppressed PKG expression as well. Similarly, we also observed that adenovirus-induced overexpression of sGC α- and β-subunits in passaged bovine aortic VSMC, which are sGC deficient, suppress PKG mRNA and protein expression after chronic NO stimulation (unpublished data).
Previous studies from several laboratories including our own have shown that “cross talk” occurs between the NO/cGMP pathway and the cAMP pathway (9, 10). Hence, robust elevations in cGMP have been shown to lead to the activation of PKA in cells, and the growth inhibitory effects of high concentrations of NO have been linked, at least in part, to cGMP elevations and the cross-activation of PKA in rat aortic VSMC (9). The PDE-resistant cyclic nucleotide analogs CPT-cGMP and CPT-cAMP inhibited PKG expression. However, the cAMP analog appeared to be more effective than the cGMP analog in suppressing PKG expression in VSMC. These findings suggest that cyclic nucleotide-dependent activation of PKA, but not PKG, mediates the PKG downregulation in bovine aortic VSMC. This possibility was tested with selective phosphorothioate protein kinase inhibitors. Micromolar concentrations of the PKG antagonist Rp-8p-CPT cGMPS did not reverse the effects of LPS to suppress PKG expression, whereas micromolar concentrations of the PKA antagonist Rp-8p-CPT cAMPS blocked the capacity of LPS to suppress PKG mRNA and protein expression in bovine aortic VSMC. The results described here indicate that the persistent, high levels of NO produced in response to iNOS expression in bovine aortic VSMC activate sGC to generate high concentrations of cGMP, and this leads to activation of PKA. Persistent PKA activation, in turn, leads to the suppression of PKG mRNA expression. Sellak et al. (25) showed that NO inhibits PKG gene expression in VSMC by interfering with Sp1-dependent gene transcription, in part through PKA activation. The exact mechanism by which PKA-dependent protein phosphorylation interferes with Sp1 is not known, although it has been reported that PKA activation inhibits Sp1-mediated gene transcription in other cell types (22).
The decrease in PKG expression with inflammation and the release of growth factors and other mediators of cell proliferation presumably allow VSMC to modulate to a wound-healing phenotype. These findings suggest that iNOS expression is critical in the response to injury process and is involved in neointima formation after vascular injury. At first glance, this notion seems counterintuitive, given the accepted inhibitory role of NO on VSMC proliferation. However, it is the persistent, high levels of NO produced by iNOS, as opposed to low-level, transient NO production by the calcium-dependent forms of NOS, that appears to be specifically involved in the VSMC fibroproliferative response. Further support for this concept has been reported in iNOS knockout mice. Genetically deleted iNOS mice have decreased medial and intimal VSMC proliferation, reduced VCAM-1 expression, and decreased neointimal development in response to injury or in the progression of atherosclerosis (8). Other studies have found that the lack of iNOS expression delays the wound-healing response in iNOS knockout mice (36). These findings suggest that iNOS expression is critical to the fibroproliferative response to injury process. In contrast to studies that suggest that the NO/cGMP/PKG pathway inhibits VSMC growth and the progression of atherosclerosis, a recent study by Wolfsgruber et al. (35) argues that this pathway is proatherogenic in vivo. In that study, a VSMC-specific reduction of PKG-I gene expression decreased atherosclerotic lesions in ApoE-deficient mice after 16 wk. No other vascular abnormalities were associated with the disruption of NO signaling, such as hypertension, and the PKG-deficient animals were not phenotypically different from control littermates at 8 wk. These findings were interpreted to suggest that PKG has a proatherogenic role, at least at 16 wk after targeted gene disruption. These studies are difficult to resolve with the literature suggesting an opposite role for the cGMP signaling pathway in atherosclerotic lesions. Perhaps reduction in PKG expression triggers an antiatherosclerotic compensation pathway in these animals at 16 wk.
It is well established that PKG expression in mammalian cells is highly variable and nonuniform. Several studies have shown that PKG is not as widely expressed or abundant as its closest molecular relative, PKA, in cells. Some cells such as erythrocytes and skeletal muscle myofibers do not even express PKG, whereas in other cells such as leukocytes, endothelial cells, and cardiac myocytes the levels of PKG expression are 10-fold or more lower than those in smooth muscle cells. Furthermore, in several types of vascular smooth muscle, the levels of PKG expression vary greatly under different growth conditions or in response to injury, as described above. Thus it is tempting to speculate that the rapid reduction in expression of PKG in cells such as VSMC may be an important prerequisite for the response to injury. In other mesenchymal cells, such as renal mesangial cells, PKG appears to play a similar antifibroproliferative role. Hence, the high-glucose-induced fibrotic response of mesangial cells is inhibited in cells expressing the constitutively active PKG catalytic domain (33). Perhaps inflammation and fibrosis in vascular tissue and in a wider variety of cell types depend on suppression of PKG expression.
In summary, we have demonstrated that the expression of PKG is suppressed by proinflammatory cytokines such as IL-Iβ or TNF-α and LPS in bovine aortic VSMC. The decrease in PKG mRNA and protein expression is due to the activation of the NO-cGMP pathway in response to iNOS induction. The pathophysiological production of cGMP appears to lead to activation of PKA and a subsequent decrease in PKG expression. We suggest that the ability of VSMC to decrease PKG allows the VSMC to respond to injury by undergoing a phenotypic modulation to a fibroproliferative state.
This work was supported by National Heart, Lung, and Blood Institute Grants HL-53426 and HL-66164.
We are grateful to Dr. Steven P. Suchyta for generously providing the β-actin standard and to Dr. Solomon Fiifi Ofori-Acquah for skillful assistance with real-time PCR.
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 © 2004 the American Physiological Society