There is evidence to suggest that the pharmacological activity of endogenous NO (i.e. NO formed by NOS) can be attributed to the combination of different redox forms of nitrogen monoxide. For example, we recently reported that both NO• and NO− may contribute to the activity of EDRF (Ellis et al., 2000). Therefore the interconversion between the redox forms of nitrogen monoxide needs to be studied in greater detail. In the present study we investigated interactions between the nitroxyl anion derived from Angeli's salt and agents that inactivate the free radical form, NO•, using both pharmacodynamic experiments and an electrochemical system with a NO electrode sensor that selectively detects the free radical.
Oxidation of nitroxyl by Cu[II]
When copper (as Cu[II] in CuSO4, 100 μM) was present in the PSS, the amount of NO• detected from 0.3 μM Angeli's salt was increased from an undetectable level to 142.7 nM, indicating that NO− was being oxidized to NO• coupled to the reduction of copper from Cu[II] to Cu[I], an effect reported recently by Nelli et al. (2000). However, the conversion ratio, that is the proportion of Angeli's salt added to the bath detected by the sensor as NO•, was only 47.6%, yet when a saturated solution of NO gas was used, almost 100% of the added NO• was detected by the electrode. It appears, therefore, that the oxidation of NO− from Angeli's salt to NO• is incomplete.
The NO• signal produced by Angeli's salt in the presence of copper was almost abolished by carboxy-PTIO (to about 8% of control), markedly attenuated by hydroxocobalamin (to about 18% of control), and reduced to a lesser extent by pyrogallol (to about 48% of control), indicating that NO• was rapidly inactivated. Likewise, the presence of these agents abolished or reduced the NO• signal when NO• from a saturated solution of NO gas was added in the presence or absence of copper. However, pyrogallol reduced the NO• signal to a greater extent in the absence of copper than in its presence. A likely explanation for this is that the copper participates in a dismutation reaction with superoxide with the net reaction products being hydrogen peroxide and molecular oxygen (Halliwell & Gutteridge, 1999), as in the reaction catalysed by Cu/Zn superoxide dismutase.
Difference between aorta and anococcygeus
As with previous findings (Li et al., 1999), relaxations of anococcygeus muscles induced by Angeli's salt were not affected by carboxy-PTIO, hydroxocobalamin or pyrogallol, agents which are widely reported to reduce NO•-induced relaxations (Gillespie & Sheng, 1990; La & Rand, 1998; Li et al., 1999). These agents, however, reduced relaxations to Angeli's salt in endothelium-denuded rat aortic rings.
Many agents that reduce relaxations in vascular tissue to endogenously generated NO (EDRF) or exogenously applied NO• (as an aqueous solution of the gas) are ineffective in blocking nitrergic transmission in nitrergically-innervated tissues (Rajanayagam et al., 1993; Rand & Li, 1993; 1995; La et al., 1996; Jiang et al., 1997; La & Rand, 1998). A possible reason for the difference in susceptibility between exogenous NO• and the nitrergic transmitter is that the latter may be the nitroxyl redox form of nitrogen monoxide, as recently suggested (Li et al., 1999). However, this does not explain the difference between anococcygeus muscles and aortic rings in the susceptibility to blockade of relaxant responses to Angeli's salt. A possible explanation for this is that the oxidative environment may differ between the two tissues.
Such a difference between the aorta and a non-vascular tissue (the gastric fundus of the rat) was invoked by Guilmard et al. (1998). They suggested that differences in reactivities to NO-mediated relaxants and of inhibition of relaxations by guanylate cyclase inhibitors were attributable to greater inactivation by superoxide anions of NO• released into the extracellular space in vascular tissue than in the gastric fundus. It is unlikely that in the present study the different sensitivities to the agents could be accounted for by the different contractile agents used to raise tone in the two issues. When the aorta was incubated with 20 μM guanethidine to replicate conditions used in the anococcygeus muscle, relaxations to Angeli's salt were still reduced by carboxy-PTIO, hydroxocobalamin and pyrogallol (data not shown).
One possible reason for the difference between the aorta and the anococcygeus muscle is that the NO− released from Angeli's salt can be oxidized to NO• by a cellular component of the aorta which is not present in the anococcygeus muscle. A recent study indicates that cytochrome P450 may play a role in the oxidation of NO− into NO• in vascular tissue (Nelli et al., 2001), therefore this could represent a potential mechanism to explain our present findings. An attempt to demonstrate this possibility by adding Angeli's salt to PSS containing a crude homogenate of aorta failed to show an increase in the amount of NO• detected by the sensor electrode (data not shown). However, the fact that no change was observed in the NO• signal does not necessarily indicate the absence of a potential oxidant. In the intact tissue, it would be expected that the oxidation of NO− would occur in close proximity to the endogenous ‘detector’ of NO, guanylate cyclase, and may even be coupled to the ‘detector’. Experiments with the electrode sensor and an aortic homogenate dispersed throughout the 8 ml bath may not faithfully emulate the conditions that operate in the intact organized tissue.
Another possible explanation for the difference between the aorta and the anococcygeus muscle is that the anococcygeus muscle may contain an antioxidant that may prevent the oxidation of NO− to NO•, the form which is more susceptible to inactivation. Lilley & Gibson (1997) reported that the antioxidants, ascorbate and urate were released from the mouse anococcygeus muscle, and suggested that their presence explained the resistance of the nitrergic transmitter to inactivation by superoxide generators. We found that homogenates of anococcygeus muscles slightly decreased the NO• signal from Angeli's salt in the presence of copper. However, this effect was also produced by aortic homogenates, indicating that this decrease was not due specifically to an antioxidant factor found only in the anococcygeus muscle.
It has been stated that only the free radical form of nitrogen monoxide can activate guanylate cyclase (Dierks & Burstyn, 1996). If this is so, then the NO− from Angeli's salt must be oxidized to NO• before it can produce a relaxation. However, our findings with the anococcygeus muscle in the presence of NO•-inactivating agents indicate that NO− is not oxidized to NO•. It was also shown that Angeli's salt can stimulate human neutrophil migration and elevate cyclic GMP levels in either aerobic or anaerobic conditions, and this was taken as evidence that NO− does not need to be oxidized extracellularly to mediate a guanylate cyclase-dependent effect (Vanuffelen et al., 1998). It should be noted that Dierks & Burstyn (1996) assayed the activity of soluble guanylate cyclase in the presence of a high concentration of the thiol, dithiothreitol (10 mM). Since, thiols are known to trap NO− (Pino & Feelisch, 1994; Hughes, 1999) it is likely that this may have interfered with the activity of the NO− derived from Angeli's salt, therefore the possibility that the NO− anion can directly activate guanylate cyclase cannot be completely ruled out yet. However, since NO− is charged and presumably not lipid-soluble, the mechanism by which it traverses the cell membrane is not clear and requires investigation.
In the aorta on the other hand, our present study indicates that NO− may be oxidized to NO• at the smooth muscle cell surface, where the NO• would be exposed to the inactivating agents. Furthermore, the uncharged, lipid-soluble NO• could diffuse readily across the cell membrane.
Nitroxyl anions generated by Angeli's salt appear to mediate relaxations differently in the aorta than in non-vascular smooth muscle since NO-inactivating agents inhibited these responses in the aorta but not in the anococcygeus muscle. Such differences can be explained by the possibility that the aorta may have a more oxidative environment, thus allowing the oxidation of NO− to NO•, while in the anococcygeus muscle, little or no conversion occurs. Because of mounting evidence for the participation of the NO− anion in the activity of endogenously-generated NO, the findings from this study could provide an insight into how oxidant agents, whether occurring endogenously or used in an experimental setting, alter its reactivity.