The shy Angeli and his elusive creature: the HNO route to vasodilation

what we now call Angeli's salt (AS) was discovered more than a century ago by the Italian chemist Angelo Angeli (Fig. 1) . He found that “salts of nitroxylaminic acid are readily resolved into the corresponding nitrites and the unsaturated residue nitroxyl” (HNO), proposing their use as suppliers of HNO (3). However, by analogy with the genial but very reclusive personality of the discoverer (17), the physiological properties of HNO didn't draw too much attention until early 1990s. In their seminal work, Fukuto and colleagues (19) reported that HNO elicits vasorelaxation in both the rabbit aorta and bovine intrapulmonary artery by a soluble guanylate cyclase (sGC)- dependent pathway. Later, it was evident that the thiol-donating agent l-cysteine can discriminate the vasodilatative profile of HNO from that of nitric oxide (NO·) or nitrosothiols: AS/HNO action can be blocked by this agent, whereas the NO· effect is potentiated (36). In combination with the suggestion that NO· is the only species able to directly activate sGC (9), these data led to the conclusion that HNO-induced vasorelaxation is due to its intracellular conversion to NO· with subsequent activation of sGC. For several years, this idea has relegated HNO to a mere precursor of the presence and biological activity of NO·, de facto negating to HNO its own right to exist.

Fig. 1.

Angelo Angeli (1864–1931) in his laboratory. [Courtesy of Dr. Marco Fontana and Prof. Maria Grazia Costa (Dipartimento di Chimica Organica, University of Florence, Florence, Italy)].

Recent thorough reevaluation of the chemical properties and reactivity of HNO has definitely proved that this molecule is distinct from NO· (28, 33) and other reactive nitrogen species (RNS) (24, 29). Likewise, the use of HNO donors is delineating biological effects that are unique and different from those of NO·-donating compounds, NO· oxidized products, and other ROS/RNS (ONOO⁻, NO₂, or N₂O₃). This is particularly true in the cardiovascular system, where HNO action is often “orthogonal” to that of NO·/organic nitrates (27, 42) despite some shared features. As a matter of fact, the in vivo infusion of AS/HNO leads to increased inotropy in normal and failing in vivo hearts (34, 35), whereas NO·/organic nitrates are negative or neutral. In vitro vasodilation from AS/HNO doesn't lead to tolerance (23), in stark contrast to the organic nitrate glyceryl trinitrate. Similarly to exogenous NO·, however, when given before ischemia HNO protects the heart against reperfusion damage (32). Furthermore, HNO, like NO·, exerts antiaggregating effects on platelets (5, 21). All this evidence concurs in praising HNO donors as a possible new attractive avenue to treat decompensated hearts and vascular dysfunction. Nevertheless, by and large HNO per se remains an enigma, as “elusive” and “shy” as the scientist who first synthesized the white and soft powder that donates it (26). First and foremost, it is still undetermined whether this chemical entity is set to remain a bench “oddity” (or a pharmacological tool) or if it will finally reveal itself as an endogenous, physiological mediator. Over the last decade, the possible generation of HNO in vivo has been suggested to derive from 1) the direct oxidation of N^G-hydroxy-l-arginine, which is an intermediate in the enzymatic conversion of l-arginine to NO· (19); 2) direct reduction of NO· via ubiquinol, cytochrome c, MnSOD, and xanthine oxidase (20, 37); 3) decomposition of nitrosothiols (44); 4) NO· synthase (NOS) in the absence of tetrahydrobiopterin (1); and 5) heme protein-mediated peroxidation of hydroxylamine (11). No definitive proof has been attained yet, and tracking down HNO footprints in vitro and in vivo remains a major current challenge for physicists and biochemists. In theory, HNO should elude spin trapping because, unlike NO·, it is not a radical. However, deoxymyoglobin and metmyoglobin can trap free HNO to form a stable adduct, HNO-myoglobin, or myoglobin-NO (12, 14), and, more recently, selective detection of HNO in solution has been achieved by reductive nitrosylation in xerogel films (10). Still, in physiological buffers, HNO rapidly disappears via dimerization to hyponitrous acid followed by dehydration and the formation of N₂O, which, however, doesn't relax vascular beds (19). Alternatively, HNO may evanesce after its fast reaction with thiols (20, 33), and likely this may serve as a HNO signature. As a matter of fact, in vitro HNO reaction with GSH generates sulfinamide and GSSG (12). Whether this is an amenable tool to detect HNO in vivo is currently unclear.

Elucidation of the HNO mechanism of action and its associated signaling in the cardiovascular system are other pending issues. In the heart, HNO positive inotropy/lusitropy stems from its ability to interact with thiols, likely cysteines, located in key structures of both the sarcoplasmic reticulum (SR) (6, 18, 38) and myofilaments (8). In this context, HNO enhances Ca²⁺ cycling and the sensitivity of contractile/regulatory proteins to Ca²⁺, fully independently from cGMP/PKG and cAMP/PKA (38). Conversely, in the vasculature, the HNO route to vasodilation appears to be more complex, raising several major questions. Are the mechanistic underpins of HNO-induced dilation always similar to those of NO·/organic nitrates? Does HNO impact the same messengers and channels in different vascular beds? Finally, if the HNO vasorelaxant action differs substantially from that of NO·, is HNO vicarious in diseased situations where NO·-dependent endothelial function is impaired? If so, how this happens? In in vitro large conduit vessels, HNO-induced vasorelaxation is accompanied by cGMP accumulation (13, 19, 23, 40). Yet, the direct intravenous infusion of AS (35) or the other HNO donor isopropylamine NONOate (27) into large mammals does not appreciably rise systemic levels of cGMP. There are many possibility to explain this apparent discrepancy, including the presence of adventitious oxidants in in vitro preparations (possibly converting HNO to NO·), the possible in situ compartmentalization of cGMP levels, the dose of HNO donors employed in vivo versus in vitro studies, and, finally, the sensitivity of the cGMP detection methods employed in vivo. This observation, however, combined with in vivo evidence that HNO donors may affect more markedly venous capacitance than arterial resistance (35), calls for different mechanisms and/or sensitivity to HNO in these systemic vascular compartments. It is tempting to speculate that HNO may affect the compartment side through cGMP signaling, whereas it may lower peripheral arterial resistance by releasing the potent vasorelaxant CGRP. Indeed, CGRP per se does not affect venous capacitance because preload is not reduced by CGRP in vivo (25). Thus, sGC/cGMP signaling may not be the only route followed by HNO to induce vasorelaxation in large conduit arterial compartments. This hypothesis is supported by data obtained in coronary arteries showing that the use of the CGRP antagonist CGRP_(8–37) significantly blunted HNO-evoked coronary relaxation (15). Considering that 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) abolished the response to AS/HNO, yet CGRP receptor and ATP-sensitive K⁺ (K_ATP) channel blockade only suppressed the maximal effect, it seems plausible that CGRP release is dependent on sGC activation. Likewise, the fact that glibenclamide, in part, inhibited the AS/HNO vasodilatory response makes possible that K_ATP channel activation occurs subsequent to CGRP release. Overall, these mechanistic intricacies unravel a complex interplay of possible mediators of the HNO vascular response that requires further investigation. The role of HNO in the modulation of the tone of resistance vessels is even more complex. In contrast to NO·, which is known to activate Ca²⁺-activated K⁺ (K_Ca) channels (30), Irvine and colleagues (22) have shown that HNO instead targets voltage-gated K⁺ (K_v) channels, leading to vascular smooth muscle cell hyperpolarization that, at least in part, is sGC/cGMP dependent. When coupled with the possibility of NOS-independent generation of HNO (4, 20, 44), this finding has raised the interesting possibility that HNO may act as an endothelium-derived hyperpolarization factor (EDHF).

This HNO-EDHF hypothesis is in part the subject of the contribution by Favaloro and Kemp-Harper in the current issue of the American Journal of Physiology-Heart and Circulatory Physiology (16) that complements other in press observations made by this group on the same topic (2). Here, the authors compared the vasorelaxation mechanisms of HNO and NO· in rat resistance vessels. They evaluated the ability of AS/HNO (0.1 nM–10 μM) and an aqueous solution of authentic NO· gas (0.1 nM–1 μM) to hyperpolarize rat mesenteric arteries and to repolarize them after contraction and depolarization with methoxamine. They found that the endothelium-independent vasorelaxation and repolarization by AS/HNO were attenuated by l-cysteine, 4-aminopyridine (4-AP), and ODQ. Conversely, NO·-mediated repolarization was resistant to both ODQ and 4-AP, suggesting different pathways for HNO signaling in this vascular bed. These findings confirm the redox-sensitive nature of HNO vasorelaxation (22, 36) and the involvement of K_v channels in HNO-induced repolarization, which is mostly sGC/cGMP dependent. The possibility of NO-independent K_v channel activation is supported by the observation that the direct sGC stimulator YC-1 also activated these channels (although more potently than HNO) in a 4-AP-sensitive manner. In contrast, glibenclamide and large-conductance K_Ca/intermediate-conductance K_Ca channel inhibitor charybdotoxin were unable to abolish HNO-induced repolarization. Thus, one major conclusion of the present report is that in resistance arteries, in the same concentration range, HNO action is mechanistically and quantitatively distinct from that of pure NO·. HNO donated by AS is more potent than NO in eliciting vascular repolarization (48% vs. 14%), and nitrite has no part in the AS action. Moreover, these data hint at the concrete possibility that in this vascular bed HNO activates sGC directly with an as-yet-undefined mechanism. Indeed, sGC inhibition by ODQ more markedly blunts HNO than NO· action. The authors excluded the extracellular conversion of HNO to NO· using the metal chelator EDTA and the NO· scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide, although the possibility of intracellular oxidation still persists. In another study (2), the authors made another important observation, identifying a putative contribution of HNO to ACh-mediated vasorelaxation in mouse mesenteric arteries. They found that in these vessels the response to ACh was in part inhibited by l-cysteine or 4-AP (up to a 25-fold reduction in potency). ACh-induced hyperpolarization in these vessels was reduced by ∼45% with 4-AP, and this response was further attenuated by combining 4-AP with the NOS inhibitor N-nitro-l-arginine methyl ester (l-NAME). In this setting, vasorelaxation induced by the NO· donor diethylamine NONOate was unchanged by 4-AP. Therefore, the authors concluded that endogenously generated HNO may contribute to endothelium-dependent hyperpolarization evoked by ACh in mouse mesenteric arteries. However, they also highlighted the fact that NO· may take some part in mouse mesenteric artery vascular smooth muscle hyperpolarization as l-NAME combined with 4-AP virtually abolished ACh effects. Thus, HNO and NO· vasorelaxation may complement each other in resistance vessels, but the lack of physical and/or biochemical evidence for in situ HNO production warrants further studies.

Several major considerations and practical implications for the cardiovascular system follow from these studies. Given its peculiar profile, exogenous (or endogenous) HNO may not only complement NO· signaling but also be vicarious when the latter is impaired. This hypothesis is based on the following observations. ROS may trap and neutralize NO·. Vascular smooth muscle typifies this situation as it has multiple sources of ROS generation and many ways for redox-based mechanisms to negatively or positively affect vascular function (43). However, HNO has very different chemical properties than ROS and NO·, which involve radical chemistry (33). HNO would rather act as an electrophil than participate in the propagation of this chemistry (28, 33, 42). This view is supported by the fact that HNO vasorelaxing properties appear preserved in in vivo heart failure (34), where increased oxidative stress is certainly present. Furthermore, loss of nitrovasodilator responsiveness (i.e., vascular tolerance) has been associated with increased vascular superoxide production (31). This is not the case with AS/HNO, both in vivo (35) and in vitro (23). Since some of these outcomes may also be dose dependent, further studies are needed to fully compare HNO responses in normal versus redox-altered vascular tissue and to dissect out the complex network of reactions likely going on between HNO and other ROS/RNS in the vasculature. Finally, HNO from AS has some similarities with nitrite/nitrate (35), which have long been used as unloading tools for the failing heart. In certain vessels, such as the resistance ones, HNO appears to be even more potent than nitrite (22). In the end, Angeli created for us a compound that generates two substances: nitrite and HNO. Both of them may provide alternative to traditional nitrosovasodilators in diseased conditions. Under hypoxia, nitrite can be converted to NO·, providing a redox buffer for the consequences of lack of O₂ (7). Conversely, under oxidative stress conditions, NO· is consumed through radical interactions with ROS, leading to decreased bioavailability. In stark contrast, HNO does not readily interact with these radicals. Thus, its chemistry, cardiotropic action, and vasodilatory profile make this molecule (and its donors) almost an ideal pharmacological tool to treat cardiovascular conditions in which pump failure, endothelial dysfunction, and oxidative stress concur.

Angeli was nominated nine times for the Nobel Prize by chemists all over Europe (17). Yet, his reluctance to attend congresses, give lectures, and publish his studies (still lying on his chest of drawers for the most part) very likely precluded the full recognition of his creativity and innovative contributions. He expanded our knowledge on pyrrole and indolic compounds, discovering the analogies of behaviour of the para- and ortho-derivatives of benzene. His work on azoxy compounds led him to demonstrate their asymmetric structure and the possibility of obtaining isomers. Angeli also pioneered studies on “constitution and olfactory function,” correlating this sensing ability to oxidative reactions. He died in Florence in an anonymous hotel room where he lived in for many years (17). While HNO has been found already in space as a gas (39), we hope that one day this molecule will also be discovered in our bodies. On that day, Angeli's creature, HNO, will finally leave its niche in the kingdom of chemistry and pharmacology to enter the realm of physiology and perhaps pathobiology (37, 41).

GRANTS

The present work was supported by National Heart, Lung, and Blood Institute Grant RO1 HL-075265 (to N. Paolocci).