Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Both human endothelial and vascular smooth muscle cells have the ability to convert organic nitrates, e.g., glyceryl trinitrate (GTN) to nitric oxide (NO) (Benjamin et al., 1991; Feelisch et al., 1995). NO binds to the ferroheme prosthetic group of soluble guanylate cyclase, with increased production of guanosine 3',5'-cGMP, resulting in vascular smooth muscle relaxation and inhibition of platelet aggregation (Ignarro et al., 1981; Hibbs et al., 1990).

It remains unclear how organic nitrates are converted to NO. It has been proposed that organic nitrates interact with reduced sulfhydryl groups (SH) to form NO and/or S-nitrosothiols (Ignarro et al., 1981; Loscalzo, 1985). Supplementation with SH, especially cysteine, increases the vasodilatory and antiaggregant effects of organic nitrates in vitro (Loscalzo, 1985; Boesgaard et al., 1993; Salvemini et al., 1993) and in vivo (Horowitz et al., 1983). NO is rapidly inactivated intracellularly or in the plasma; therefore, SH may react readily with NO to yield biologically active and stable S-nitrosothiols, which act as carrier molecules for NO and preserve its activity (Ignarro et al., 1981; Stamler et al., 1992).

The evidence for an enzymatic process in the bioconversion of organic nitrates to NO arises from the demonstration that when vascular smooth muscle and endothelial cells are denatured by boiling, the conversion of GTN to NO is decreased by 80 to 90% (Feelisch and Kelm, 1991; Salvemini et al., 1992). In addition, human umbilical vein endothelial cells (HUVEC), when pretreated with the cross-linking agent glutaraldehyde, lost the capacity to generate NO from organic nitrates, indicating that NO generation is unlikely to result from a nonenzymatic reaction between nitrate ester and membrane-bound SH (Feelisch et al., 1995). Various enzymatic pathways have been proposed to take part in the bioconversion of GTN to NO. Although glutathione S-transferase has been shown to metabolize GTN in the liver and a cytochrome P-450 has been shown to metabolize GTN in porcine kidney cells (Servent et al., 1989; Lau and Benet, 1990; Schroder and Schror, 1992), this is not the case in the vascular smooth muscle and endothelial cell (Chung et al., 1992; Kurz et al., 1993; Salvemini et al., 1993).

Bacterial nitrate and nitrite reductases are a group of complex enzymes that possess a series of redox centers including Fe/S. These redox centers are used for shuttling electrons in the anaerobic growth of bacteria when inorganic nitrate is reduced to nitrite and NO (Adams and Mortenson, 1982; Chen et al., 1996). Xanthine oxidoreductase (XOR) is an enzyme that possesses a similar series of redox centers in mammals (Fig. 1). This enzyme is a metalloflavoprotein composed of two identical and catalytically independent subunits. Each subunit contains one molybdenum center, two Fe/S centers, and an FAD (Fig. 1). The two Fe/S centers are located near the NH₂-terminal portion of the enzyme. This is the only region of the deduced amino acid sequence that has sufficient numbers of conserved cysteine residues to serve as ligands for the Fe/S center (Wootton et al., 1991; Wright et al., 1993).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Schematic of the domain structure of XOR. A, a generalized scheme of electron (e) transport with the three redox centers of the enzyme: molybdenum (Mo), FAD, and Fe/S. B, the xanthine-to-urate reaction with the generation of O2 or reduced NADH. C, the oxidation of NADH to NAD+ at the FAD center with the generation of O2 (Sanders et al., 1997). D, the proposed reduction of GTN to NO.

XOR exists in vivo in the constitutive form of xanthine dehydrogenase (XDH), which oxidizes the purines xanthine and hypoxanthine to uric acid. NAD⁺ is reduced to NADH during this reaction. The dehydrogenase form of the enzyme can be proteolytically converted to the oxidase form usually through a stimulus such as hypoxia. Oxidation of purines still occurs with xanthine oxidase (XO), however, coupled to the reduction of molecular oxygen rather than NAD⁺, with the formation of O₂. More recently, Sanders et al. (1997) have described a less well recognized function of XOR, the oxidation of NADH using molecular oxygen with the formation of O₂. This appears to be more efficient with the dehydrogenase form of the enzyme. Whether the conversion of XDH to XO is the main source of O₂ in ischemia-reperfusion injury, or whether an NADH oxidase function of XDH is the source, is the subject of current debate.

XOR is present in the human endothelium and heart (Bulkley, 1993; Hellsten-Westing, 1993; Moriwaki et al., 1993), and its enzymatic activity is increased during hypoxia (Poss et al., 1996). This distribution of enzymatic activity in a relatively hypoxic environment, in addition to the requirement of a three-electron reduction, would favor the NADH oxidase activity of XOR as a possible bioactivating enzyme for organic nitrates.

Using platelet aggregation and the antiaggregant effect of NO as a bioassay, we evaluated the ability of XOR to bioconvert GTN to NO.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Preparation of Platelet-Rich Plasma (PRP). Peripheral venous blood was drawn from drug-free healthy human volunteers into trisodium citrate (0.38% final concentration) and acetylsalicylic acid (1 mM final concentration). PRP was separated after centrifugation of plasma at 250g for 10 min. Platelet-poor plasma (PPP) was separated after further centrifugation at 1500g for 10 min at room temperature.

Platelet Aggregation Studies. Platelet aggregation was studied with a Payton aggregometer model 300B using the Born method (Born and Cross, 1963). The output from the aggregometer was processed by a Maclab system connected to an Apple Macintosh computer that allowed automated analysis. PRP (0.5 ml) was added to each cuvette and made up to a final volume of 0.65 ml with PBS. For each set of experiments, the addition to the PRP of the different components was as follows: 220 µM GTN and 28 to 340 U/l XOR (n = 10 healthy volunteers); 220 µM GTN, 340 U/l XOR, and 5 to 1000 µM allopurinol (n = 4); 220 µM GTN, 28 to 340 U/l XOR, and 100 U/ml superoxide dismutase (SOD) (n = 5); and 220 µM GTN, 85 to 340 U/l XOR, and 1 to 10 µM oxyhemoglobin (n = 4). After addition of the various components, PRP was incubated at 37°C for 1 min, before stimulation of platelet aggregation with a submaximal dose of U46619 (11,9-epoxymethano-prostaglandin H₂) 2 µM. All measurements of platelet aggregation were performed in duplicate. Aggregation was recorded as percent change in light transmission 2 min after stimulation of platelet aggregation, with 100% aggregation taken as light transmission of PPP. Baseline calibrations were performed for each recording to compensate for changes in light transmission due to the opacity of the preparation. All aggregation studies were performed within 2 h of blood sampling.

Materials. Bovine XOR was obtained from Biozyme Laboratories (Blaenavon, Gwent, UK). Hemoglobin from bovine erythrocytes was obtained from Calbiochem/Novabiochem (Nottingham, UK). Oxyhemoglobin was freshly prepared as follows: hemoglobin was dissolved in PBS, bubbled with oxygen for 3 min, reduced with 10 M excess sodium dithionite, and subsequently bubbled with oxygen for an additional 15 min. The solution was then desalted and purified by passing it through a Sephadex G-25 column (Pharmacia Biotech, Uppsala, Sweden). GTN was obtained from David Bull Laboratories (Warwick, UK). All other materials were obtained from Sigma Chemical Co. (Poole, UK).

Analysis. Data acquired were analyzed with one-way ANOVA. Percent inhibition of platelet aggregation was log transformed before statistical analysis. Results are expressed as means ± S.E., except for percent inhibition of platelet aggregation in the experiments with SOD. Data in these experiments were fitted to a sigmoid E_max model with nonlinear least-squares regression. Both parallel and nonparallel sigmoid E_max shift models were considered, and the choice of model was determined by Akaike's Information Criterion (Jackson et al., 1987).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

U46619 (2 µM) caused platelet aggregation in the presence of 1 mM acetylsalicylic acid. When added individually, 220 µM GTN and 340 U/l XOR did not inhibit the aggregant effect of U46619, producing 73.1 ± 3.8 and 72.5 ± 3.1% aggregation, respectively. This decreased to 4.2 ± 1.2% aggregation when XOR and GTN were added in combination (Fig. 2).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Representative traces of platelet aggregation as portrayed by an increase in light transmission in the presence of U46619, GTN, or XOR alone and the antiaggregant effect of the GTN plus XOR combination. GTN (220 µM) and XOR (340 U/l), separately or in combination, were added to PRP and incubated at 37°C for 1 min before stimulation of platelet aggregation with 2 µM U46619. Aggregation measurements were recorded for 2 min.

XOR produced a dose-dependent inhibition of platelet aggregation when incubated with GTN for 1 min (Fig. 3). Allopurinol, a specific inhibitor of XOR, produced a dose-dependent inhibition of the antiaggregant effect of XOR plus GTN, with an IC₅₀ of ~100 µM (Fig. 4).

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Inhibition of platelet aggregation on incubation of XOR with 220 µM GTN for 1 min at 37°C. A dose-dependent increase in inhibition of platelet aggregation occurred. n = 10; mean ± S.E. ***P < .001 versus 340 U/l.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   The effect of increasing doses of allopurinol on the antiaggregant effect of XOR (340 U/l) and GTN (220 µM). Allopurinol produced a dose-dependent reversal of the antiaggregant effect of XOR plus GTN with an IC50 of 100 µM. n = 4; mean ± S.E. ***P < .001 versus baseline.

The addition of SOD (100 U/ml) to the incubation of PRP, GTN, and XOR shifted the antiaggregant dose-response curve for XOR plus GTN to the left. The IC₅₀ for inhibition of platelet aggregation of 200 U/l for XOR plus GTN without SOD was reduced to an IC₅₀ of 80 U/l in the presence of SOD (Fig. 5). The addition of 1, 5, and 10 µM oxyhemoglobin, an extracellular NO scavenger, produced a dose-dependent decrease in the antiaggregant effect of XOR plus GTN. Percent inhibition of platelet aggregation decreased from 98.5 ± 0.8% without oxyhemoglobin to 35.3 ± 4.8% with 10 µM oxyhemoglobin (Fig. 6).

View larger version (13K): [in this window] [in a new window]   Fig. 5.   The effect of SOD on the antiaggregant effect of XOR and GTN (220 µM). SOD produced a shift to the left in the XOR plus GTN dose-response curve. The IC50 of 200 U/l for XOR plus GTN without SOD decreased to 80 U/l for XOR plus GTN with SOD. n = 5.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   The effect of increasing doses of oxyhemoglobin (HbO2) on the dose-response curve for XOR with 220 µM GTN. There was a noncompetitive inhibition of the antiaggregant effect of XOR plus GTN with the addition of increasing doses of oxyhemoglobin to the incubation. n = 4; mean ± S.E. ***P < .001 versus XOR plus GTN alone.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

GTN is an effective antianginal agent for use in ischemic heart disease because of its combined vasodilatory and platelet antiaggregant activity. These pharmacodynamic effects are achieved by the conversion of GTN to the active component NO. The exact mechanism of bioactivation of GTN to NO has not been elucidated. However, there is sufficient evidence to suggest that bioactivation occurs through an enzymatic and thiol-dependent pathway (Loscalzo, 1985; Feelisch and Kelm, 1991; Salvemini et al., 1992; Boesgaard et al., 1993; Salvemini et al., 1993; Feelisch et al., 1995). This enzymatic bioactivation of GTN will require a chemical reduction.

Using the thromboxane mimetic U46619 as a platelet aggregant in a suspension of PRP, we have shown that the combination of GTN plus XOR produces an antiaggregant effect that does not occur when GTN or XOR is incubated separately. Although GTN alone is a weak inhibitor of platelet aggregation (Salvemini et al., 1993; Kampf and Ritter, 1994), this depends on the duration and temperature of incubation of GTN with the platelet preparation. In this study, the experimental design included an incubation of GTN for 1 min at 37°C before the addition of U46619, and under these conditions, GTN alone did not produce significant inhibition of platelet aggregation. PRP was kept at room temperature before aliquots were incubated at 37°C. PRP stored at room temperature, as opposed to PRP stored at 37°C, does not affect the antiaggregant effects of GTN (Kampf and Ritter, 1994).

Millar et al. (1998) recently showed that XOR, with NADH acting as reducing substrate, can reduce GTN to NO under hypoxic conditions. The rate of production of NO from GTN plus XOR in these experiments equates with that reported for GTN when incubated with endothelial cells (Feelisch et al., 1995; Millar et al., 1998). In addition, the production of NO from XOR plus GTN was inhibited by inhibitors of the molybdenum center of XOR [oxypurinol and ()-BOF-4272]. This is in keeping with our finding that the antiaggregant effect of GTN plus XOR in PRP was inhibited by allopurinol in a dose-dependent manner.

In contrast to findings under hypoxic conditions, where GTN is reduced to NO by XOR and NADH, Sanders et al. (1997) showed that XOR catalyzes the oxidation of NADH with oxygen, generating O₂ in an oxygenated environment. This redox reaction takes place at the FAD center of the enzyme. As outlined by Millar et al. (1998), the production of NO from the reduction of GTN by XOR decreases as the oxygen tension increases. It would seem there is competition for the reduction of GTN from the FAD center of the enzyme, at which any residual oxygen in the incubation medium will compete for the reducing electron and form O₂. Our experiments were conducted under atmospheric oxygen (21% O₂) conditions, and because platelets are exquisitely sensitive to NO, sufficient amounts of NO appear to have been produced to exert an antiaggregant effect that was inhibited by oxyhemoglobin, an extracellular scavenger of NO.

It is conceivable that O₂ is generated in addition to NO in the incubation of XOR plus GTN under atmospheric oxygen conditions. O₂ would be expected to decrease any platelet antiaggregant effect mediated through NO, and this theory is supported by the increase in the antiaggregant effect of XOR plus GTN observed with the addition of SOD. However, there are other mechanisms in addition to the dismutation of O₂ whereby SOD could facilitate the antiaggregant effect of XOR plus GTN. It is well recognized that the affinity of O₂ for NO far exceeds that for SOD, leading to the production of peroxynitrite (ONOO; Radi et al., 1991; Thomson et al., 1995). Interestingly, ONOO has been shown to inhibit platelet aggregation induced by U46619, and this inhibition is thought to be mediated through the nitration of protein, e.g., tyrosine residues, by ONOO (Yin et al., 1995). It has also been shown that nitration of tyrosine by ONOO is catalyzed by SOD (Ischiropoulos et al., 1992), and this may be a possible explanation for the increase in the antiaggregant effect of XOR plus GTN observed with the addition of SOD. Furthermore, H₂O₂ has been shown to strongly enhance the inhibitory effect of NO donors on platelet aggregation (Naseem and Bruckdorfer, 1995). We, therefore, cannot outrule the role of H₂O₂ in facilitating the antiaggregatory effect produced from the combination of XOR plus GTN on the addition of SOD. The inhibition of platelet aggregation by XOR plus GTN is unlikely to have occurred as a result of oxidation of U46619, which is a stable endoperoxide analog and an effective thromboxane mimetic in the presence of ONOO (Chabot et al., 1997) and conditions of oxidative stress (Kromer and Tippins, 1999).

The ability of an NADH oxidase to produce both NO and O₂ in vivo, which needs further study, will depend on the balance between the availability of molecular oxygen and an organic nitrate, e.g., GTN. In the relatively hypoxic environment of the venous circulation, the conversion of GTN to NO by an NADH oxidase activity of XOR may account for the venoselective hemodynamic responses seen with GTN in vivo. In support of this, XOR has been located on the plasma membrane of capillary and postcapillary endothelium (Stevens et al., 1991). Furthermore, nanomolar concentrations of NO have been shown to reversibly inhibit XOR activity, and the duration of inhibition of XOR has been shown to be dose dependent (Fukahori et al., 1994). The FAD center was proposed as the most likely site of NO-induced alteration of enzyme function. This interaction provides an interesting concept whereby NO may be acting as a defense mechanism against XOR production of O₂. However, in the presence of organic nitrates, the inhibitory effect of NO on XOR may conceivably be inhibiting NO production from the reduction of organic nitrates and may therefore contribute to the development of nitrate tolerance.

In summary, we have given evidence to support the hypothesis that GTN may be reduced to NO in vitro by the enzyme XOR in sufficient amounts to inhibit platelet aggregation. The antiaggregant effect of GTN plus XOR was inhibited by oxyhemoglobin, an extracellular scavenger of NO. Allopurinol, a specific inhibitor of the molybdenum center of XOR, inhibited the antiaggregant effects of GTN plus XOR, and this antiaggregant effect was enhanced by SOD. These results support the hypothesis that XOR, through a less well recognized NADH oxidase activity of the enzyme, may be important in the bioactivation of organic nitrates.