Vascular Smooth Muscle Cell Heme Oxygenases Generate Guanylyl Cyclase–Stimulatory Carbon Monoxide
Background Carbon monoxide (CO), like nitric oxide (NO), stimulates soluble guanylyl cyclase and thereby raises intracellular levels of cGMP. We examined the endogenous capacity of vascular smooth muscle cells (SMCs) to produce CO from heme through the activity of heme oxygenases.
Methods and Results Cultured SMCs from rat aorta (RASMCs) expressed immunoreactive inducible heme oxygenase-1 (HO-1) and constitutive HO-2. Treatment of RASMCs with hemin and sodium arsenite, which are inducers of HO-1, stimulated RASMC cGMP without stimulating nitrite release or inducible NO synthase expression, and the induced elevations of cGMP were not inhibited by the NO synthase inhibitor NG-methyl-l-arginine. Induced CO from RASMCs likewise caused elevation of cGMP levels in platelets coincubated with the vascular cells. Zinc protoporphyrin IX, an inhibitor of HO, reversed the inducible increases in platelet cGMP.
Conclusions These results indicate that vascular SMCs have both constitutive and inducible HO activity, and they respond to specific stimuli to generate guanylyl cyclase–stimulatory CO in the same SMCs and in coincubated platelets.
The role of nitric oxide (NO) as a physiological messenger molecule is clearly established. NO is produced from l-arginine by the action of NO synthases (NOSs). Also known as endothelium-derived relaxing factor, NO regulates blood vessel tone and blood flow by inhibiting smooth muscle contraction and platelet aggregation. Both of these NO-induced effects are mediated by the stimulation of soluble guanylyl cyclase, resulting in an increase in the intracellular levels of cGMP.1
Carbon monoxide (CO) is a second simple diatomic gas molecule that shares some of the physicochemical properties of NO. Like NO, CO has the ability to bind to the iron atom of the heme moiety associated with soluble guanylyl cyclase, thereby activating the enzyme and increasing intracellular cGMP production. Exogenously added CO inhibits platelet aggregation and relaxes isolated blood vessel preparations by elevating intracellular levels of cGMP.2
CO is produced endogenously by various cell types as a byproduct of heme catabolism, in which heme oxygenase (HO) catalyzes the degradation of heme to biliverdin and CO. Two forms of HO have been identified. HO-1 is induced by diverse factors, including its substrate heme, as well as metal ions such as cobalt and arsenite.3 HO-2 is a constitutive enzyme that cannot be induced. Both forms of HO are inhibited by metal protoporphyrins, the most potent being zinc protoporphyrin IX (ZnPP-IX). There is great variability in the tissue distribution of the two isoforms of HO: HO-1 has high expression in the spleen and the liver, whereas the brain and testes have the highest levels of HO-2 expression.3
Verma et al4 recently reported the potential physiological role of endogenous CO as a neural messenger. HO-2 was localized to discrete brain neuronal populations in which NOS activity was absent and cGMP levels could be depleted by exposure to ZnPP-IX. Since cGMP controls vascular tone and platelet function, the aim of the present study was to investigate the role of endogenously produced CO as a potentially important vascular modulator of hemostasis. Results of these experiments suggest that the capacity of vascular SMCs to regulate vascular tone and platelet thrombus formation may be due, in part, to heme oxygenase–mediated CO production.
Culture of Rat Aortic SMCs
Rat aortic SMCs (RASMCs) were obtained from the laboratory of Dr Tim Scott-Burden (University of Texas, Houston). These cells were isolated from rat thoracic aorta by enzymatic digestion using standard methods and characterized morphologically and immunocytochemically as described previously.5 RASMCs were cultured serially in minimum essential medium (MEM) containing 10% (vol/vol) heat-inactivated FCS, glutamine (4 mmol/L), and the antibiotics penicillin (0.1 g/L) and streptomycin (0.2 g/L). RASMCs were passaged twice weekly by harvesting with trypsin-EDTA and were either plated as stocks at a 1:4 ratio in T75 culture flasks or seeded into six-well multiwell plates for incubation with the different reagents. Confluent monolayers of RASMCs were used between passages 10 and 20.
RASMCs were incubated with or without various HO inducers in MEM containing 5% FCS for 24 hours. The inducers used in this study were (in μmol/L): hemin 20, cobalt chloride 100, iodoacetamide 10, and sodium arsenite 10. In indicated experiments, the inhibitor of NO synthase NG-methyl-l-arginine (L-NMA; 0.1 mmol/L) or the inhibitor of HO ZnPP-IX (10 μmol/L) was added at the beginning of the 24-hour incubation period.
NO in the supernatants of the RASMC monolayers was determined by measuring the levels of nitrite (the predominant oxidation product of NO) by diazotization and absorbance reading at 540 nm, as previously described.6 Nitrite concentrations were determined by comparison with a standard curve using sodium nitrite as a standard and water as a blank. The detection limit of this assay has been reported at 200 pmol/L.7
After the 24-hour treatment period, the culture media were removed and the monolayers of cells rinsed with PBS. The RASMCs were treated with the phosphodiesterase inhibitor isobutyl methylxanthine (IBMX; 0.5 mmol/L) for 45 minutes, and then 1 mL of trichloroacetic acid (TCA; 6% wt/vol) was added. The cells were scraped off the plate, collected into Eppendorf tubes, sonicated for 20 seconds, and assayed for cGMP with a commercially available radioimmunoassay kit (NEN-DuPont).
Platelet Detector System
Platelets were isolated from blood obtained from healthy volunteers, and washed platelet suspensions in Ca2+-free Tyrode’s buffer supplemented with HEPES (2 mmol/L) were prepared as previously described.8 The washed platelet suspensions were then treated with IBMX (0.1 mmol/L) and added onto monolayers of RASMCs. Before the addition of the platelet suspensions to the individual wells, the media containing the various treatments were aspirated, and the RASMC monolayers were rinsed with PBS. This ensured that platelets were not exposed to the inducers and inhibitors used in the pretreatments of the RASMCs. After 45 minutes of coincubation with RASMCs, the platelet suspensions were aspirated from the wells and placed in TCA (6% wt/vol). The samples were then processed for cGMP radioimmunoassay as described above.
Western Blot Analysis
RASMCs were exposed to HO inducers for 24 hours. The cells were then lysed in gel electrophoresis sample buffer (12 mmol/L Tris-HCl, pH 6.8, 5% glycerol, 0.4% SDS), boiled for 5 minutes, sonicated for 10 seconds, and microcentrifuged at 10 000g for 10 minutes. The resulting supernatants were collected and subsequently quantified for protein. Cell extracts (50 μg) were then subjected to SDS-PAGE on a 17.5% gel. The separated proteins were transferred to a nitrocellulose membrane, and the blot was then blocked with a 5% solution of Carnation dried milk and 3% BSA in PBS for 1 hour. The blot was washed with PBS and incubated for 1 hour at 24°C with an HO-1–specific antibody (OSA-100) diluted 400-fold. The blot was washed again with PBS and incubated for an additional 1 hour with goat alkaline phosphatase–conjugated F(ab′)2 fragment against rabbit IgG, diluted 2000-fold. After the final washes, the immune-complexed bands were visualized with nitro blue tetrazolium and BCIP. This procedure allowed visual comparison of the relative staining intensities of individual bands as an indication of induction of HO-1. Alternatively, an HO-2–specific antibody (OSA-200) was used as the primary antibody to detect the presence of the constitutive HO-2 protein in these cells. In a parallel experiment, samples from control and hemin-treated (20.0 μmol/L), sodium arsenite–treated (10.0 μmol/L), or interleukin 1β–treated (IL-1β, 3.0 ng/mL) RASMCs were exposed to Western blot analysis using a mouse monoclonal antibody specific for inducible NOS (iNOS).
Hemin, sodium m-arsenite, cobalt chloride, iodoacetamide, IBMX, creatine phosphate, creatine phosphokinase, fatty acid–free BSA, sodium nitrite, N-(1-naphthyl) ethylenediamine dihydrochloride, sulfanilamide, DMSO, L-NMA, TCA, and alkaline phosphatase–conjugated F(ab′)2 goat fragment against rabbit IgG were purchased from Sigma Chemical Co; 3-morpholinosydnonimine hydrochloride (SIN-1) was from Biomolecular Research Laboratories; antibiotics, MEM, FCS, l-glutamine, and trypsin-EDTA were from Gibco BRL, Life Technologies; antibodies against HO-1 (OSA-100) and HO-2 (OSA-200) were purchased from StressGen; ZnPP-IX was from Porphyrin Products; nitro blue tetrazolium, BCIP, and goat anti-mouse alkaline phosphatase–conjugated IgG (H+L) were from Upstate Biotechnology, Inc; and mouse anti-iNOS antibody was from Transduction Laboratories.
Results are expressed as mean±SEM. Statistical analyses were performed with Student’s two-tailed paired t test and ANOVA when more than two treatments were compared. Values of P<.05 were considered to be statistically significant.
The two HO-specific antibodies used in the Western blot analysis, OSA-100 and OSA-200, produced bands demonstrating that RASMCs constitutively express HO-1 and HO-2, respectively (Fig 1⇓). Western blotting of untreated RASMCs with antibody OSA-100 produced an immunoreactive band at approximately 30 kD (Fig 1A⇓, lane 2). OSA-100 exhibited specificity for HO-1 in that it did not recognize noninducible HO-2; HO-2 was identified specifically by OSA-200 at approximately 36 kD (Fig 1B⇓, lane 2). Treatment of RASMCs with inducers of HO-1 resulted in a significant increase in the amount of HO-1 protein (30 kD) expressed in these cells. Fig 1A⇓ (lanes 3 through 6) shows that hemin (20 μmol/L), iodoacetamide (10 μmol/L), sodium arsenite (10 μmol/L), and cobalt chloride (100 μmol/L) induced a significant increase in HO-1 expression compared with the untreated cells (lane 2). A comparison of the relative intensities of these bands revealed that sodium arsenite was the most potent of these inducers of HO-1. Using Western blot analysis with an anti-iNOS monoclonal antibody, we ruled out induction of iNOS by the heme oxygenase inducers. Samples from IL-1β–treated RASMCs produced the iNOS band at 130 kD (Fig 1C⇓, lane 1). The same antibody did not react with samples from control, sodium arsenite–treated, or hemin-treated RASMCs (lanes 2 through 4).
Since exogenous CO is known to increase intracellular cGMP concentrations, we measured RASMC cGMP to monitor the induction of HO activity. As shown in Fig 2A⇓, 24-hour treatment of RASMCs with either 20 μmol/L hemin or 10 μmol/L sodium arsenite resulted in 36% and 290% increases of intracellular cGMP levels, respectively. Direct activation of guanylyl cyclase by the heme oxygenase inducers was ruled out, since treatment of RASMCs with hemin or sodium arsenite for 15 minutes (an incubation time too brief for HO-1 induction) did not result in cGMP elevations (data not shown). In a number of tissues, including vascular SMCs, NO generated by NOSs likewise regulates intracellular cGMP concentrations.1 To exclude the possibility that the cGMP increases seen in Fig 2A⇓ were due to the induction of inducible NOSs, two approaches were taken. First, as shown in Fig 2B⇓, the NOS inhibitor L-NMA (0.1 mmol/L) failed to alter the cGMP elevations induced by hemin and sodium arsenite. Second, levels of NO in the media, as determined by nitrite measurements, were not significantly affected by the above treatments (Fig 2C⇓ and 2D⇓).
To confirm that the hemin- and sodium arsenite–induced cGMP increases were due to HO-mediated CO production, the potent HO inhibitor ZnPP-IX was used. ZnPP-IX lowered cGMP concentrations of the RASMCs to levels 31.1% lower than those of basal controls (data not shown). This effect was not due to ZnPP-IX activation of RASMC phosphodiesterase, since all experiments were performed in the presence of IBMX, a phosphodiesterase inhibitor. We considered these results to be due to either ZnPP-IX–mediated trapping of basal NO or to ZnPP-IX–mediated inhibition of heme oxygenase and/or guanylyl cyclase.
We tested the ability of ZnPP-IX (10.0 μmol/L) to trap NO in a cell-free system using SIN-1, an NO-generating agent. The nitrite levels of SIN-1 solutions exposed to ZnPP-IX for 24 hours were not significantly different from those of untreated controls (data not shown). Therefore, the reductions of cGMP in RASMCs in the presence of ZnPP-IX are not due to trapping of basal NO by ZnPP-IX. To examine whether ZnPP-IX inhibits guanylyl cyclase itself, SMCs were exposed to sodium nitroprusside (SNP, 0.1 mmol/L), a guanylyl cyclase stimulator, in the presence or absence of ZnPP-IX. The SNP-induced increases in RASMC cGMP were reduced significantly, by 63%, in the presence of ZnPP-IX, indicating that this protoporphyrin inhibits guanylyl cyclase as well as HO.
Treatment of RASMCs with hemin or sodium arsenite also resulted in an increase of intracellular cGMP levels of coincubated platelets (Table⇓). The increases paralleled those observed in RASMCs under the same conditions. Platelets exposed to the hemin-treated SMCs, but not to hemin itself, demonstrated a 23.6% increase in their intracellular cGMP levels. Similarly, platelets coincubated with RASMCs exposed to the more potent HO inducer sodium arsenite demonstrated a 111% increase in their intracellular cGMP levels. The addition of ZnPP-IX reversed the hemin- or sodium arsenite–induced rise in platelet cGMP concentrations, even though the platelets were not themselves exposed to the above treatments. When RASMCs were treated with ZnPP-IX alone, the basal cGMP content of the coincubated platelets was reduced by approximately 25%. To confirm that this observed decrease was not due to trapping of basal NO by ZnPP-IX, we repeated the experiment in the presence of L-NMA. L-NMA failed to reverse the ZnPP-IX–mediated reduction of basal platelet cGMP (data not shown).
We demonstrate the expression of both the constitutive and inducible isoforms of HO in RASMCs. Treatment of RASMCs with hemin or sodium arsenite resulted in the induction of HO-1 and in a significant increase in the cGMP content of the cultured SMCs. These heme oxygenase inducers had no effect on NO production or iNOS expression by RASMCs; furthermore, the NOS inhibitor L-NMA did not alter the hemin- and sodium arsenite–induced cGMP elevations in the SMCs. Therefore, NO was excluded as the guanylyl cyclase activator in this system. Coincubation of platelets with hemin- or sodium arsenite–treated RASMCs also increased the intracellular cGMP content of the platelets. These cGMP increases were blocked by ZnPP-IX, an HO inhibitor. Furthermore, ZnPP-IX treatment of the SMCs also reduced the basal platelet cGMP levels, indicating constitutive activity of SMC HO-2 under basal conditions. Vascular SMCs thus have the capacity to generate endogenous carbon monoxide (CO).
CO shares some of the chemical and biological properties of NO.2 Exogenously added CO, like NO, stimulates SMC guanylyl cyclase and causes vasorelaxation9 10 11 12 ; it also stimulates platelet guanylyl cyclase and inhibits platelet activation.13 14 CO has recently been identified as an endogenous biological messenger in the brain.4 An NO-independent pathway of guanylyl cyclase activation has recently been proposed in cytokine-stimulated vascular smooth muscle.15 Our results suggest that CO is the diffusible endogenous messenger molecule that stimulates the cGMP elevations produced by the hemin- and arsenite-treated RASMCs. To demonstrate that these effects were due to HO-derived CO, we used ZnPP-IX, an inhibitor of HO. However, control experiments revealed that ZnPP-IX also inhibited cGMP accumulation in SMCs in response to the guanylyl cyclase inducer SNP, indicating that ZnPP-IX inhibits guanylyl cyclase as well as HO. To circumvent the effects of ZnPP-IX on guanylyl cyclase, we developed a platelet detector system. Under these conditions, platelets coincubated with the SMCs were not directly exposed to the various inducers and/or inhibitors of SMC HO; therefore, any effects observed on the platelet cGMP content were due solely to guanylyl cyclase regulation by a diffusible SMC-derived messenger molecule. The results of these experiments demonstrated that RASMCs release CO by both constitutive and inducible heme oxygenase mechanisms to regulate platelet cGMP levels.
Our studies suggest that CO is not simply an irrelevant byproduct of heme catabolism but rather may assume physiologically important regulatory functions in the vasculature. We propose that CO is a SMC-derived messenger molecule with autocrine and paracrine actions. Thus, vessel wall–derived CO potentially serves as an endogenous regulator of vascular tone and platelet reactivity. The heme–heme oxygenase–CO and the l-arginine–NO synthase–NO systems have certain parallels. Both gas products cause vasorelaxation and platelet inhibition by elevating the cGMP content of vascular SMCs and platelets, respectively. Furthermore, both constitutive and inducible forms of heme oxygenase and NO synthase are active in the vessel wall. However, there are also some important differences between the properties of CO and NO.2 It is unlikely that CO and NO represent redundant messenger molecules. Further studies will be required to elucidate the specific inducers and regulators of the enzymes that lead to the production of these biologically active simple gases.
Note added in proof. Since submission of this article, Morita et al have reported that hypoxia induces vascular SMC CO production.16
This study was supported by VA Merit Review grants (Dr Schafer and Dr Kroll), National Institutes of Health grants HL-36045 (Dr Schafer) and HL-02311 (Dr Kroll); and grants from the American Heart Association, Texas Affiliate (Dr Durante), and an Established Investigator award (Dr Kroll).
- Received February 14, 1995.
- Accepted March 5, 1995.
- Copyright © 1995 by American Heart Association
Stamler JS, Singel DJ, Loscalzo J. Biochemistry of nitric oxide and its redox-activated forms. Science. 1992;258:1898-1902.
Marks GS, Brien JF, Nakatsu K, McLauglin BE. Does carbon monoxide have a physiological function? Trends Pharmacol Sci. 1991;121:185-188.
Maines MD. Heme oxygenase: function, multiplicity, regulatory mechanisms, and clinical applications. FASEB J. 1988;2:2557-2568.
Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH. Carbon monoxide: a putative neural messenger. Science. 1993;259:381-384.
Scott-Burden T, Resink TJ, Baur U, Burgin M, Buhler FR. Epidermal growth factor responsiveness in smooth muscle cells from hypertensive and normotensive rats. Hypertension. 1989;13:295-304.
Durante W, Schini VB, Catovsky S, Kroll MH, Vanhoutte PM, Schafer AI. Plasmin potentiates the induction of nitric oxide synthesis by interleukin-1β in vascular smooth cells. Am J Physiol. 1993;264:H617-H624.
Schmidt HHH, Warner TD, Nakane M, Forstermann U, Murad F. Regulation and subcellular location of nitrogen oxide synthases in RAW264.7 macrophages. Mol Pharmacol. 1992;41:615-624.
Brune B, Ullrich V. Inhibition of platelet aggregation by carbon monoxide is mediated by activation of guanylyl cyclase. Mol Pharmacol. 1987;32:497-504.
Beasley D, McGuiggin M. Interleukin 1 activates soluble guanylyl cyclase in human vascular smooth muscle cells through a novel nitric oxide-independent pathway. J Exp Med. 1994;179:71-80.
Morita T, Perrella MA, Lee ME, Kourembanas S. Smooth muscle cell-derived carbon monoxide is a regulator of vascular cGMP. Proc Natl Acad Sci U S A. 1995;92:1475-1479.