Inhibition of Nitric Oxide–Stimulated Vasorelaxation by Carbon Monoxide-Releasing Molecules
Objective—Carbon monoxide (CO) is a weak soluble guanylyl cyclase stimulator, leading to transient increases in cGMP and vasodilation. The aim of the present work was to measure the effect of CO-releasing molecules (CORMs) on the cGMP/nitric oxide (NO) pathway and to evaluate how selected CORMs affect NO-induced vasorelaxation.
Methods and Results—Incubation of smooth muscle cells with some but not all of the CORMs caused a minor increase in cGMP levels. Concentration-response curves were bell-shaped, with higher CORMs concentrations producing lower increases in cGMP levels. Although exposure of cells to CORM-2 enhanced cGMP formation, we observed that the compound inhibited NO-stimulated cGMP accumulation in cells and NO-stimulated soluble guanylyl cyclase activity that could be reversed by superoxide anion scavengers. Reactive oxygen species generation from CORMs was confirmed using luminol-induced chemiluminescence and electron spin resonance. Furthermore, we observed that NO is scavenged by CORM-2. When used alone CORM-2 relaxed vessels through a cGMP-mediated pathway but attenuated NO donor-stimulated vasorelaxation.
Conclusion—We conclude that the CORMs examined have context-dependent effects on vessel tone, as they can directly dilate blood vessels, but also block NO-induced vasorelaxation.
Carbon monoxide (CO), which was previously considered only as a highly toxic and life-threatening pollutant, has been recognized as a signaling molecule having regulatory roles in many physiological and pathophysiological processes within the cardiovascular, immune, and nervous systems.1,2 It is generated in mammalian tissues via the endogenous degradation of heme by a family of constitutive (heme oxygenase-2 and heme oxygenase-3) and inducible (heme oxygenase-1) heme oxygenase enzymes.2 CO gas binds to heme-containing proteins and exerts antiinflammatory and antioxidant effects.2,3 In the cardiovascular system, CO exhibits vasodilatory properties, prevents endothelial cell apoptosis, controls vascular smooth muscle cell proliferation, and inhibits platelet aggregation.3–5
It has been suggested that CO can be used as a therapeutic agent; testing the beneficial actions of CO has been facilitated by the discovery of molecules that release CO under appropriate conditions.1,6 Initial CO-releasing molecules (CORMs) were transition metal carbonyls soluble only in organic solvents and required the use of physical or strong chemical stimuli to release CO.7,8 More recently, water-soluble CORMs have been reported that liberate CO in a controlled fashion in biological systems and have the potential to deliver it to tissues and organs.9,10 Several uses have been proposed for CORMs, ranging from inflammatory diseases and vascular dysfunction to tissue ischemia, organ rejection, and sepsis.1 In the cardiovascular system, CORMs recapitulate many of the effects of CO gas. CORMs induce vessel relaxation in isolated aortic tissue and prevent coronary vasoconstriction, as well as acute hypertension in vivo.11
Similar to nitric oxide (NO), CO can activate the hemoprotein soluble guanylyl cyclase (sGC), but only weakly; CO is reported to enhance sGC activity by 3- to 4-fold compared with the maximal 200-fold activation triggered by NO.12 On activation, sGC converts GTP to the second messenger molecule cyclic guanosine monophosphate (cGMP), which has an impact on many cellular functions.13 For example, sGC is important in lowering smooth muscle tone and preventing platelet aggregation and leukocyte adhesion to the vessel wall.13–15
In spite of the observations that CORMs reduce vessel tone, the potential interactions between CORMs and vasodilators that activate sGC (ie, NO donors) have not been studied. To this end, we have synthesized several CORMs and determined their ability to alter cGMP levels in cells and to modulate NO-stimulated cGMP formation and vasodilation.
Synthesis of ALFs and Detection of CO Release
The metal carbonyl CORMs (ALFXXX) were synthesized according to published or modified procedures for the synthesis of related complexes (see supplemental material, available online at http://atvb.ahajournals.org). All compounds were characterized by Fourier transform infrared spectroscopy, 1H nuclear magnetic resonance, and C,H,N analysis. The compounds were prepared, purified, and kept under N2 atmosphere in the dark. The solutions of CORMs were freshly prepared before the experiments by dissolving the compound in the appropriate solvent. The release of CO from the metal carbonyl complexes (ALFXXX) under normoxic conditions was assessed by a chromatographic method (see supplemental material).
Rat aortic smooth muscle cells (RASMCs) were isolated from 12- to 14-week-old male Wistar rats, 5 rats per isolation, as previously described.16 More than 95% of cells isolated stained positive for smooth muscle α-actin. Cells between passages 2 and 5 were used for all experiments. RASMCs were routinely cultured in Dulbecco's modified Eagle medium containing 4.5 g/L glucose and supplemented with 10% fetal bovine serum and antibiotics.
Determination of cGMP Levels in RASMCs
RASMCs were plated in 24-well plates. Cells were washed twice with Hanks' balanced salt solution and incubated in Hanks' balanced salt solution in the presence of 3-isobutyl-1-methylxanthine (1 mmol/L) for 5 minutes. Cells were then treated with vehicle, tempol (1 mmol/L), sodium nitroprusside (SNP) (10 μM), diethylamine NONOate (DEA/NO) (1 μmol/L), a nonsaturated CO solution (100 μmol/L), or CORMs at various concentrations (1 to 300 μmol/L) for 15 minutes. Media were then aspirated, and 200 μL of 0.1 N HCl was added into each well to extract cGMP. After 30 minutes, HCl extracts were collected and centrifuged at 600g for 10 minutes to remove debris. The supernatants were directly analyzed for cGMP by enzyme immunoassay.
Determination of sGC Activity
Guanylyl cyclase activity was determined by formation of [α-32P]cGMP from [α-32P]GTP. In short, reactions were performed for 10 minutes at 30°C in a final volume of 100 μL, in a 50 mmol/L HEPES, pH 8.0, reaction buffer containing 500 μmol/L GTP, 1 mmol/L dithiothreitol, and 5 mmol/L MgCl2. Basal and DEA/NO (1 μmol/L)–stimulated guanylyl cyclase activity was measured in the presence of CORMs (10 and 100 μmol/L). To inhibit reactive oxygen species (ROS) production superoxide dismutase (SOD) (100 IU/mL) was added in the mixture. Typically, 5 μL of purified sGC (10 ng/μL) was used in each assay reaction.
Assessment of ROS Production by CORMs
Chemiluminescence was used to measure ROS production by freshly prepared CORMs. L-012 (8-amino-5-chloro-7-phenylpyrido[3,4-day]pyridazine-1,4-(2H,3H)dione sodium salt) is a very sensitive chemiluminescence probe used to detect reactive oxygen and nitrogen species formation. Various CORMs (1 to 300 μmol/L) were added to 40 μL of a PBS buffer containing 400 μmol/L L-012. L-012-enhanced chemiluminescence was immediately measured with a chemiluminometer (Luminoskan, Labsystems, Melbourne, Victoria, Australia) and reported as relative light units. In some cases, 4-hydroxy-2,2,6,6-tetramethylpiperidinyloxy (tempol) (1 mmol/L) was added to the mixture.
NO Quantification in Solution in the Presence of CORM-2
The release of NO was measured using an amperometric NO-specific probe (ISO-NOP) connected to the single channel free radical analyzer (TBR 1025, World Precision Instruments). A 4-channel recording system (Laboratory-Trax-4/16, World Precision Instruments) was used to digitize the analog signal from the probe. The ISO-NOP sensor was calibrated by mixing NaNO2 solution of desired concentration with H2SO4 (0.1 mol/L) and NaI (0.1 mol/L). The release of NO by DEA/NO in phosphate buffer, pH 7.4, was then monitored continuously. Cumulative concentration responses curves to DEA/NO were performed in the presence of vehicle or CORM-2 (10 and 100 μmol/L); the solution was constantly bubbled with N2 to remove atmospheric oxygen.
Assessment of Hydroxyl Production by ALF186, ALF409, and ALF436
The formation of oxygen-based radicals and ROS in solution from the reaction between low-oxidation-state complexes and O2 was tested in electron spin resonance experiments using 5-tert-butoxycarbonyl 5-methyl-1-pyrroline N-oxide (BMPO) as a spin trap.17 A stock solution of each CORM was prepared by dissolving the compound in 1 mL of deoxygenated distilled water (MeOH was used in the case of ALF436) under N2. Another stock solution of BMPO was prepared with a concentration of 250 mmol/L. An aliquot was taken from the CORM solution and 100 μL from the BMPO solution to a closed vial under N2, and deoxygenated distilled water was added to make 1 mL total volume. The final concentration of CORM was 200 μmol/L, whereas the final concentration of BMPO was 12.5 mmol/L for the ALF186 experiment and 25 mmol/L for the others, ALF409 and ALF436. A sample was taken from this solution, and a control spectrum was acquired before bubbling O2. Oxygen was then bubbled in the solution for 3 to 4 minutes, and another sample was taken and analyzed. The experiment was performed in a quartz, planar, electron spin resonance cell, at room temperature, microwave power of 2 mW, modulation frequency 100 kHz, modulation amplitude 0.1 mT.
Aortic Ring Preparation
Male Wistar rats (250 to 300 g; Harlan, Bresso, Italy) of 8 to 10 weeks of age were euthanized, and the thoracic aorta was rapidly dissected and cleaned from fat and connective tissue. Rings were denuded of the endothelium and placed in organ baths filled with oxygenated (95% O2 to 5% CO2) Krebs solution at 37°C, mounted to isometric force transducers, and connected to a Graphtec recorder. Rings were initially stretched until a resting tension of 0.5 g was reached and allowed to equilibrate for at least 30 minutes, during which tension was adjusted. In each experiment aortic rings were first challenged with phenylephrine (PE) (1 μmol/L) until the responses were reproducible. Aortic rings were contracted with PE (1 μmol/L). Once the plateau was reached, a cumulative concentration-response curve to DEA/NO or S-nitroso-N-acetyl-l,l-penicillamine (SNAP) (1 nmol/L to 10 μmol/L) and CORM-2 (1 μmol/L to 200 μmol/L) in the presence or absence of 1H-(1,2,4)oxadiazole(4,3-a)quinoxalin-1-one (ODQ) (5 μmol/L) were performed. In another set of experiments, rings were pretreated with 200 μmol/L CORM-2 for 15 minutes at resting tension, and then PE was added to constrict the preparation. The increase in tension induced by PE after CORM-2 preincubation was not significantly different from control. DEA/NO or SNAP cumulative concentration-response curves were then performed in the presence or absence of SOD (100 IU/mL) or tempol (3 mmol/L).
Data are expressed as means±SE. Statistical comparisons between groups were performed using 1- or 2-way ANOVA (data from cell culture experiments were analyzed with 1-way ANOVA, whereas data from vasorelaxation studies were analyzed by 2-way ANOVA) followed by a post hoc test or Student t test as appropriate. Differences were considered significant at P<0.05. GraphPad Prism software (version 4.02, GraphPad Software, San Diego, CA) was used for all the statistical analysis.
Detection of CO Liberated From CORMs
CO release form CORMs in cell-culture medium (Roswell Park Memorial Institute medium/fetal bovine serum) is summarized in the supplemental Table. Some CORMs (ALF186, ALF409, ALF411, ALF432, and ALF436) liberate >0.9 molar equivalents of CO after 2 hours. Interestingly, when dissolved in PBS (pH 7.4), CO liberation was enhanced in all cases, with the exception of ALF411. This indicates that molecules present in the Roswell Park Memorial Institute medium/fetal bovine serum medium protect CORMs complexes from decomposition or quench the CO formed. ALF411 behaves differently because it dissolves in Roswell Park Memorial Institute medium/fetal bovine serum, but is completely insoluble in PBS. As expected, once in solution its stability decreases and CO release occurs. The role of O2 as the trigger for the liberation of CO from these complexes is quite clear because under anaerobic conditions (N2), CO release is blocked in all cases except that of ALF409, where a small amount of CO is still liberated. In contrast to other CORMs and in accordance to previous reports,1 the well-known Ru(II) CORMs, CORM-2 [Ru(CO)3Cl2]2 and CORM-3 [Ru(CO)3Cl(H2NCH2COO)] do not release any CO in the headspace. On the contrary, they slowly release CO2 over 24 hours in solution.18
Effect of CORMs on cGMP Levels in Cultured Cells
Exposure to a nonsaturated CO solution (100 μmol/L) caused a small, ≈50%, increase in cGMP levels in rat aortic smooth muscle cells (Figure 1A). For comparison, under the same conditions, stimulation of cells with the NO donor at 10 μmol/L leads to a 30-fold increase in cGMP accumulation (Figure 1B). Most CORMs when used at 1 and 10 μmol/L failed to increase cGMP levels in smooth muscle cells (Supplemental Table I); among the compounds tested, ALF436 increased cGMP by 3-fold at the lowest concentration used. For the more widely studied CORM-2 and CORM-3, higher concentrations were also used. Cells treated with CORM-2 showed a bell-shaped increase in cGMP accumulation, whereas exposure to CORM-3 caused an inhibition at the highest concentration used (300 μmol/L) (Figure 1C and 1D). The ability of CORM-2 to stimulate cGMP levels was blocked by the sGC inhibitor ODQ or by incubation of cells with hemoglobin (Supplemental Figure IA and IB). Moreover, the inactive form of CORM-2 (lacking the CO groups) did not alter cGMP accumulation (Supplemental Figure IC). The above data taken together suggest that CORM-2 increased cGMP accumulation through CO-dependent activation of sGC.
Effect of CORMs on Recombinant sGC Activity
We next evaluated whether CORMs exert a direct effect on sGC activity by using recombinant rat sGC. In contrast to the increase in cGMP accumulation observed with cultured cells, no increase in sGC activity was observed after incubation with CORM-3 and CORM-2 at the concentrations used (Figure 2A). On the other hand, the NO-stimulated sGC activity was markedly reduced in a concentration-dependent manner by these CORMs (Figure 2B).
Generation of ROS and Scavenging of NO by CORMs
To study the mechanism through which CORMs reduce NO-stimulated cGMP formation, we initially measured the ability of CORMs to generate ROS. ROS production from CORM-2 was concentration dependent, plateauing at a 100 μmol/L concentration of this CORM, and could be abolished by addition of tempol, an O2− scavenger (Figure 3A and 3B). Two other molybdenum-based CORMs, ALF186 and ALF436, also increased ROS production (Supplemental Figure IIA and IIB). The ALF186- and ALF436-induced ROS released was inhibited by tempol (Supplemental Figures IIIB and IVB). In a separate series of experiments, we observed that the electron spin resonance spectra obtained for the solutions of ALF186 (or CORM-3, ALF409, ALF436; data not shown) after bubbling O2 revealed a very strong and clean signal assignable to the hydroxyl adduct of BMPO (Figure 3C). Thus, although CO from CORMs activates sGC in an NO-independent and, thus, superoxide-insensitive manner, the ROS generated by CORMs would be expected to inhibit NO-stimulated guanylyl cyclase activity.
To assess whether CORM-2 can react with NO, we measured NO using an NO electrode (Figure 3D). In these experiments, we observed that CORM-2 attenuated the amount of NO in solution released by DEA/NO in a concentration-dependent manner. When NO levels were measured in a 10 μmol/L DEA/NO solution, they were found to be decreased by 80% when a 10-fold excess of CORM-2 was present. Similar observations were made using SNP as a donor (data not shown). Using a different experimental setup, we confirmed that CORM-2 quenched NO; scavenging of NO by CORM-2 was favored when CORM-2 is present in excess (>10-fold) relative to NO (Supplemental Figure V).
Interaction of CORMs With NO Donors in Cells
To study whether CORM-2 reduces NO donor-stimulated cGMP accumulation in cells, RASMCs were exposed to SNP or DEA/NO, and cGMP was measured. Indeed, CORM-2 attenuated cGMP formation in response to both NO donors, confirming and extending in a cellular milieu the in vitro biochemical findings that CORMs release ROS and attenuate cGMP formation. Interestingly, responses to SNP (Figure 4A) were affected more than those to DEA/NO (Figure 4B). In cells treated with tempol, the ability of CORM-2 to reduce DEA/NO-induced cGMP accumulation was attenuated (Supplemental Figure VI). The 2 molybdenum-based CORMs, ALF186 and ALF436, also reduced NO-stimulated cGMP accumulation in cultured cells (Supplemental Figure IIC and IID) in a tempol-reversible manner (Supplemental Figures IIIA and IVA).
Additional evidence for the role of ROS in CORMs-mediated inhibition of NO-induced cGMP formation was obtained in sGC activity assays. High concentrations of CO (700 μmol/L) decreased NO-stimulated sGC activity by ≈20% (data not shown); cGMP formation was even more reduced after exposure to CORM-2; when SOD was added to the sGC activity assay mixture to remove any O2− formed, the effect of CORM-2 was partially reversed (Figure 4C).
Effect of CORM-2 on Vascular Tone
CORM-2 administration caused a concentration-dependent relaxation that reached approximately 60% relaxation at the highest concentration used (Figure 5A). CORM-2-induced relaxation was abolished by ODQ, suggesting that the relaxation is mediated by sGC. Exposure to the NO donor DEA/NO led to a full relaxation that was observed at a much lower concentration than the one needed for maximal CORM-2 relaxation (Figure 5B). To assess whether CORM-2 could interfere with the vasodilation triggered by exogenously administered NO, relaxation curves to DEA/NO were repeated in the presence of a maximally dilating CORM-2 concentration. It should be noted that the contractile response to PE was not significantly affected by the preincubation with CORM-2. This might be due to the fact that the CORM-2-induced vasodilation is short-lived and is no longer observed after the preincubation, or due to the fact that the vasorelaxing action of CORM-2 is only displayed when this agent is applied to a tissue after, but not before, it develops an elevated tone. In line with what we observed in the experiments with purified sGC and cultured cells, NO-induced relaxation was attenuated by CORM-2. The inhibitory effect of CORM-2 in NO-induced vasodilation was reduced by SOD (Figure 5C) and tempol (Figure 5D). Similar results were obtained using a different NO donor SNAP. In a manner analogous to that observed with SNP responses in cultured cells, the vasorelaxing action of SNP was inhibited to a greater extent by CORM-2 (Supplemental Figure VII).
The derivatives of the fragment fac-[RuII(CO)3)], CORM-2, and CORM-3 are the most commonly reported examples of metal carbonyl complexes used as vehicles to deliver CO to cells and tissues.1 The CO releasing capacity of these compounds has been measured by their ability to rapidly (<1 minute) transfer 1 CO to deoxy-Mb, forming carbonmonoxy-myoglobin. However, we and others1,19 have observed that these compounds do not release CO as a free gas. Thus, CO from CORM-2 and CORM-3 is assumed to be first transferred to a biological molecule forming an intermediate that delivers CO in living systems.1 In the present study, we introduced a different group of CORMs based on zero-valent molybdenum carbonyls. In contrast to the previous Ru(II) complexes, these Mo(0) compounds are able to release CO as a free gas to their biological environment once they are activated by molecular oxygen.
All but one of the molybdenum-based CORMs, in spite of releasing measurable amounts of CO in cell culture media, failed to stimulate cGMP production in cultured vascular smooth muscle cells. The one that did (ALF436) increased cGMP only marginally, and this effect was lost when higher concentrations were used. From our in vitro studies, it became apparent that cGMP accumulation in cultured smooth muscle cells is not a good bioassay to test for biological activity of molybdenum-based CORMs. Among the ruthenium-based CORMs (CORM-2 and CORM-3), neither of which release CO to the headspace but rather transfer CO to heme-containing biological molecules, only CORM-2 increased cGMP accumulation. However, CORM-3 has been reported to promote vasodilation in an ODQ-inhibitable manner and to increase cGMP levels in vascular tissue.9 The cGMP-elevating effect of CORM-3 was shown to be diminished by endothelium removal or endothelial nitric oxide synthase inhibition, suggesting that CORM-3 might be acting, at least in part, by enhancing endogenous NO release rather than through a direct effect on sGC.9 Of note, cGMP-independent mechanisms of CORM-induced smooth muscle relaxation have also been shown to exist. These are mediated through big conductance calcium-activated potassium channels: the opening of these channels leads to membrane hyperpolarization, which in turn closes voltage-dependent calcium channels, reduces calcium levels, and relaxes smooth muscle.20
To examine whether CORMs directly activate sGC, we tested the 2 most studied CORMs to date (CORM-2 and CORM-3) for their ability to stimulate recombinant sGC. In line with previous observations, CORM-3 did not enhance sGC activity in vitro9; neither did CORM-2. CORM-3 has been shown to increase sGC activity in the presence of the sGC sensitizer YC-1.9 The fact that CORM-2 increased cGMP formation in cells but not in the sGC activity assay could be explained by the existence of an endogenous YC-1-like sGC sensitizer21 in smooth muscle cells. Alternatively, the formation of an intermediate/metabolite might be required for the transfer of CO from these CORMs to sGC and this only occurs in cells. Surprisingly, we found that both CORMs markedly and dose-dependently inhibited NO stimulated sGC activity. This inhibitory action of CORM-2 and CORM-3 on NO-stimulated sGC has not been reported before and could be explained by (1) a partial agonist effect of CO on sGC activation, (2) ROS generation by the CORMs that inactivates NO, or (3) quenching of NO by the transition metal complexes. Experiments were therefore designed to test the above-mentioned hypotheses.
CO binds to the sGC heme with considerably lower affinity and is a much weaker sGC stimulator compared with NO.22,23 tion of NO-stimulated sGC activity, suggesting that CO, can act as partial agonist if present in great excess. Because of the different affinity for sGC exhibited by NO and CO, the antagonistic action of CO would only be observed if pharmacological amounts of CO are delivered and would be of no importance under physiological conditions. Incubation of recombinant sGC with CORM-2 caused an even greater decrease in NO-stimulated sGC activity than that caused by the CO-saturated solution. The decrease in sGC activity was partially restored by SOD, suggesting that the inhibition of sGC activity by CORM-2 involves at least 2 mechanisms (partial agonism and ROS generation). To provide experimental evidence that CORM-2 generates ROS, we measured the production of ROS from CORM-2 using a chemiluminescent dye. In these experiments, we could demonstrate a concentration-dependent increase in chemiluminescence that was blocked by incubation with tempol, indicating that superoxide anions are generated from CORM-2. In addition, scavenging of superoxide anions limited the ability of CORM-2 to inhibit DEA/NO-stimulated cGMP formation in cells. ROS generation was also noted with the molybdenum complexes (ALF186 and ALF436). In addition, the inhibition of NO-induced cGMP formation by ALF186 and ALF436 could be rescued by tempol. It thus seems that Mo-based CORMs are not devoid of the drawbacks seen with the Ru-based compounds.
CORM-3, ALF186, ALF409, and ALF436 released hydroxyl radicals after exposure to O2. For the Mo(0) CORMs, it is quite plausible that hydroxyl radicals originate from the metal catalyzed decomposition of O2− initially formed, but we have not investigated this issue further. It is worth noting that in spite of the known antioxidant properties of CO, CORM-3 has also been shown to increase O2− production and constrict renal arteries in a dose-dependent manner.24 The effect of CORM-3 in renal arteries was, however, proposed to involve the activity of multiple oxidases including nitric oxide synthase, NADPH oxidase, xanthine oxidase, and complex IV of the mitochondrial electron chain. Thus, CORM-3 might generate ROS both directly and indirectly in biological systems. As sGC activity is known to depend on the existence of reduced thiol groups,13,25 it is tempting to speculate that cGMP accumulation in cells is biphasic, showing reduction at higher CORM concentrations (see Figure 1) due to the generation of ROS from CORMs and oxidation of critical sGC cysteine residues. If ROS production from CORM is excessive, it would be possible to observe an inhibition of sGC activity that masks any activation of sGC by the CO delivered.
It has been previously reported that ruthenium complexes have a high affinity for NO and effectively scavenge NO in biological systems.26,27 In fact, efforts to design drugs that scavenge NO and prevent some of its deleterious effects include the synthesis of structures based on Ru complexes.28 To test whether CORM-2 quenches NO, we used 2 different approaches. Both in the NO electrode experiments and in the measurements of NO remaining in solution after incubation with CORM-2, we observed that CORM-2 scavenged NO and removed it from the solution in a time and concentration-dependent manner. Thus, when ruthenium-based CORMs are used as pharmacological tools, they would be expected to lower NO concentrations in tissues.
It has been previously reported that CORM-2 relaxes precontracted vessels.11 In agreement with these observations, we observed a vasorelaxing effect of this CORM that was first evident at 100 μmol/L and reached an ≈60% decrease in vessel tone at 200 μmol/L. In addition, inhibition of sGC abolished the effects of CORM-2, indicating that sGC activation and cGMP elevation mediate the vasodilatory effect of CORM-2 in this preparation. However, in a manner analogous to that seen with recombinant sGC and cells, we also observed that CORM-2 attenuated NO donor–induced vasorelaxation. The magnitude of the inhibitory responses of CORM-2 toward NO-driven dilation varied with the donor used; responses to SNAP were affected more than those to DEA/NO. In agreement with what we observed with purified sGC and cultured cells we saw that scavenging O2− reduced the defect in relaxation caused by CORM-2. However, the effect of both SOD and tempol in preventing the effects of CORM-2 on NO-stimulated dilation was weaker compared with that observed with the purified enzyme or cells. Given that tissues have a higher antioxidant capacity, the inhibitory action of CORM-2 in aortic ring relaxations might be mainly due to scavenging of the NO released by CORM-2; the CORM-2:DEA/NO ratio was, for most of the concentration-response curves in the relaxation experiments, much higher than the CORM-2:NO donor ratio in the activity experiments; thus, even if ROS are removed by antioxidants, NO would still be scavenged in the organ bath experiments by CORM-2.
In summary, we have shown that CORMs do not activate purified sGC, but are capable of elevating cGMP levels in cells and causing sGC-mediated vasorelaxation. More importantly, we have shown that currently available CORMs based on transition metal carbonyl scaffolds inhibit NO-induced signaling and vasorelaxation through multiple mechanisms that include ROS generation and direct scavenging of NO. The final biological effect of any CORM would depend on the amount of CO released by the CORM, the levels of endogenously produced NO, the concentration of CORM present that might react with NO, the amount and species of ROS produced as a byproduct of CORM breakdown, and the levels of antioxidants enzymes and chemicals present in the tissues. Design of improved, “second-generation” CORMs devoid of unwanted properties (ROS generation, NO scavenging) is needed for this class of compounds to be used in establishing the therapeutic utility of CO in preclinical and clinical models.
Sources of Funding
Dr Seixas thanks Fundação para a Ciência e Tecnologia, Portugal for Doctoral Grant SFRH/BDE/15501/2004. The nuclear magnetic resonance spectrometers are part of the National NMR Network and were purchased in the framework of the National Program for Scientific Re-equipment, contract REDE/1517/RMN/2005, with funds from POCI 2010 (FEDER) and Fundação para a Ciência e a Tecnologia. This work was supported by grants from the Greek Ministry of Education, the Secretariat of Research and Technology, and a University of Patras grant to fund the PHARMANET network (to A.P.), by Grant 8661 from the Shriners Burns Hospitals (to C.S.), and by the COST Action BM1005 (European Network on Gasotransmitters).
We thank Conceição Almeida, Instituto de Tecnologia Química e Biológica, for elemental analysis.
- Received December 8, 2010.
- Accepted August 2, 2011.
- © 2011 American Heart Association, Inc.
- Wu L,
- Wang R
- Ryter SW,
- Alam J,
- Choi AMK
- Brouard S,
- Otterbein LE,
- Anrather J,
- Tobiasch E,
- Bach FH,
- Choi AM,
- Soares MP
- Szabo S
- Wang R
- Motterlini R,
- Foresti R,
- Green CJ
- Motterlini R,
- Sawle P,
- Hammad J,
- Bains S,
- Alberto R,
- Foresti R,
- Green CJ
- Motterlini R,
- Clark JE,
- Foresti R,
- Sarathchandra P,
- Mann BE,
- Green CJ
- Kharitonov VG,
- Sharma VS,
- Pilz RB,
- Magde D,
- Koesling D
- Lucas KA,
- Kazerounian S,
- Ruiz-Stewart I,
- Park J,
- Schulz S,
- Chepenik KP,
- Waldman SA
- Ahluwalia A,
- Scotland RS,
- McLean PG,
- Mathur A,
- Perretti M,
- Moncada S,
- Hobbs AJ
- Papapetropoulos A,
- Marczin N,
- Mora G,
- Milici A,
- Murad F,
- Catravas JD
- Wang R,
- Wu L
- Lamon BD,
- Zhang FF,
- Puri N,
- Brodsky SV,
- Goligorsky MS,
- Nasjletti A
- Sayed N,
- Baskaran P,
- Ma X,
- van den Akker F,
- Beuve A
- Chatterjee D,
- Mitra A,
- van Eldik R