4 Discussion
The main findings of this study are that BK-induced relaxations of resistance arteries from patients with resistant CVD, are not modified by c-PTIO, NAC or catalase and only slightly reduced by DETCA. They suggest that extracellular NO, HNO and H2O2 do not mediate communication from endothelium to smooth muscle during endothelium-dependent relaxation of resistance arteries from these patients and that these relaxations can proceed during excessive oxidative stress.
Endothelium-dependent vasodilatation of human resistance arteries are mediated by endothelium-derived prostacyclin, NOS and endothelium-dependent hyperpolarization in children, adults and patients with coronary artery disease, respectively(Beyer et al., 2017). Several laboratories including our own, reported that H2O2 can act as an endothelium-derived hyperpolarizing factor in resistance arteries from patients(Beyer et al., 2017; Ellinsworth, Sandow, Shukla, Liu, Jeremy & Gutterman, 2016; Leurgans et al., 2016; Schulz, Katunaric, Hockenberry, Gutterman & Freed, 2019; Shimokawa & Godo, 2020; Shimokawa & Morikawa, 2005). Relations between this factor, NOS and oxidative stress are controversial and were the focus of our study. The functional importance of endothelium-dependent hyperpolarization is largely unknown because it is usually investigated during pharmacological inhibition of NO-synthases. In large elastic conduit arteries, endothelial NOS produces NO that causes smooth muscle relaxation via production of cGMP. Superoxide anions produced by NADPH oxidases and mitochondria inhibit this pathway by uncoupling of eNOS, binding and inactivation of NO and damaging sGC(Daiber et al., 2017; Elbatreek et al., 2020; Evgenov, Pacher, Schmidt, Hasko, Schmidt & Stasch, 2006; Forstermann & Munzel, 2006; Monica, Bian & Murad, 2016; Zhang, Murugesan, Huang & Cai, 2020). In addition to NO, NOS can produce HNO and O2-(Forstermann & Munzel, 2006; Schmidt, Hofmann, Schindler, Shutenko, Cunningham & Feelisch, 1996). Nitroxyl is not inactivated by O2- and potently stimulates sGC(Fukuto, 2019; Irvine, Ritchie, Favaloro, Andrews, Widdop & Kemp-Harper, 2008; Leo, Joshi, Hart & Woodman, 2012; Toda & Okamura, 2003). Although L-NAME partly inhibits production of O2- by endothelial NOS(Kaesemeyer, Ogonowski, Jin, Caldwell & Caldwell, 2000), superoxide synthesized by NOS is transformed by SOD1 into H2O2 as an endothelium-derived hyperpolarizing factor in mouse mesenteric resistance arteries(Morikawa et al., 2003; Takaki et al., 2008). In coronary arterioles from patients with coronary artery disease, on the other hand, mitochondria rather than NOS are the source of vasodilator endothelium-derived H2O2(Beyer et al., 2017; Freed, Beyer, LoGiudice, Hockenberry & Gutterman, 2014; Schulz, Katunaric, Hockenberry, Gutterman & Freed, 2019).
With the aim to analyze clinically relevant relations between endothelium-dependent vasodilatation and oxidative mechanisms, we tested the hypotheses that BK-induced relaxation of resistance arteries from patients with resistant CVD is mediated by mechanisms that are either insensitive to oxidative stress or involve a reactive oxygen species. For this purpose, we studied resistance artery biopsies obtained during elective cardiothoracic surgery. The demographic, clinical and vascular pharmacological properties of the study group varied (Table 1 and Figure 1). Surgery was required despite normalization of classical risk factors, suggesting future benefit from novel pharmacotherapy inspired by proven pathogenic mechanisms. Resistance artery contractile and relaxing responses varied considerably between individual patients (Figure 1). The different size of the vessels, which is influenced by sampling bias, contributed little to this inter-assay variability. Because some of the results were unexpected, we had to confirm the endothelium-dependence of BK-induced relaxing responses in patient pericardial resistance arteries(Leurgans et al., 2016). Immunohistochemical staining of potential key players also displayed considerable inter-individual variability despite careful titration of the concentration of the primary antibodies. But, presence of all three isoforms of NOS, SOD1 and catalase could be demonstrated in the microvascular wall.
To probe involvement of mediators in relaxing responses we tested effects of exogenously applied candidate mediators and of potential scavengers thereof. We confirmed that SNP, a NO-donor, and exogenous H2O2cause relaxation in patient pericardial resistance arteries(Leurgans et al., 2016; Leurgans, Bloksgaard, Irmukhamedov, Riber & De Mey, 2018). We found that CXL-1020, an HNO-donor compound(Arcaro, Lembo & Tocchetti, 2014), relaxes patient resistance arteries. Exogenous catalase prevented relaxing effects of exogenous H2O2, as expected. c-PTIO, ODQ and DETCA abolished relaxing effects of SNP confirming that they are due to NO stimulating sGC and sensitive to inactivation by O2-(Forstermann & Munzel, 2006; Ignarro et al., 1980; Omar, Cherry, Mortelliti, Burke-Wolin & Wolin, 1991; Vanhoutte, Zhao, Xu & Leung, 2016). NAC did not modify responses to SNP but reduced the relaxing potency of CXL-1020 which was not altered by the SOD-inhibitor DETCA. These properties are very similar to those of nitrergic nervous dilator mechanisms in various organs that are mediated by HNO generated by neuronal NOS, insensitive to scavenging and inactivation by endogenous O2- or c-PTIO but sensitive to inhibition by NAC(Arcaro, Lembo & Tocchetti, 2014; Fukuto, 2019; Irvine, Ritchie, Favaloro, Andrews, Widdop & Kemp-Harper, 2008; Schmidt, Hofmann, Schindler, Shutenko, Cunningham & Feelisch, 1996; Toda & Okamura, 2003). Apart from exogenous catalase, the actions of the pharmacological tools that we revalidated are not restricted to the extracellular space.
DETCA abolished relaxing responses to exogenous NO during K+-induced contraction and amitrole reduced these contractions. This suggests a large production of endogenous O2-leading via dismutation to H2O2 that is then decomposed by endogenous catalase. This endogenous antioxidant pathway seems efficacious because addition of exogenous catalase did not modify contractile responses. No signs of basal production of relaxing concentrations of HNO were observed since NAC, 7-NI and NPLA did not modify K+-induced contractions. In contrast to the selective inhibitors of nNOS, L-NAME increased contractile responses. This finding has frequently been interpreted as indicative of basal production of NO. However, c-PTIO and DETCA did not increase contractile responses and the effect of ODQ did not reach statistical significance (Figure 6A). Non-canonical effects of L-NAME that are not due to inhibition of NOS activity have been observed after chronic but not acute administration of the compound(Kopincova, Puzserova & Bernatova, 2012; Liu et al., 2019). Our findings therefore raise the possibility that, at least in resistance arteries from patients with CVD, NOS can generate a dilator compound that is distinct from NO, HNO and O2-.
We studied agonist-stimulated endothelium-dependent relaxation during i) contraction stimulated by K+-induced depolarization that mimics myogenic tone and inhibits hyperpolarizing influences and during ii) contraction stimulated with ET-1. This stimulator of NADPH oxidases is upregulated in several CVD (Barton & Yanagisawa, 2019; Davenport et al., 2016). In patient pericardial resistance arteries, its vasoconstrictor effect is mediated by ETA receptors and not modified by direct or endothelium-dependent ETBeffects(Leurgans et al., 2016). We confirmed that in these vessels, BK-induced relaxation of ET-1 stimulated contraction is larger and refractory to inhibition by L-NAME and ODQ compared to K+-induced contraction(Leurgans et al., 2016). Our earlier proposals on involvement of endothelium-derived NO during depolarization-induced contraction and of endogenous H2O2 as an endothelium-derived hyperpolarizing factor in the presence of ET-1, could however not be confirmed.
In depolarized arteries, the endothelium-dependent effects of BK were on average markedly reduced by L-NAME and ODQ. Unlike those of SNP, they were not modified by c-PTIO and only partly reduced by DETCA. These properties could suggest involvement of HNO instead of NO. Although the nitroxyl donor CXL-1020 caused relaxation with an efficacy comparable to that of BK, effects of the former but not the latter were, however, reduced by NAC. An involvement of extracellular H2O2 is equally unlikely. In theory, H2O2 can be generated by dismutation of O2- produced by NOS(Morikawa et al., 2003) and relaxes K+-induced contraction with low potency. However, BK-induced relaxation of depolarized arteries was only partly reduced by DETCA and not modified by even a high concentration of exogenous catalase. Rather, a NOS-derived dilator compound that stimulates sGC, is distinct from NO and HNO, and resists inactivation by O2- seems to be involved. This proposal for agonist-stimulated endothelium-dependent relaxation of depolarized patient pericardial arteries is comparable to the one described above for basal conditions. Observations with indomethacin suggest that such a mechanism is responsible for three quarters of the relaxation and a dilator prostaglandin for the remaining quarter.
Relaxing responses to BK in arteries contracted with ET-1 were on average and in contrast to depolarized segments, not modified by L-NAME or ODQ in the absence and presence of indomethacin. They were shifted to endothelium-dependent hyperpolarizing responses that could not be attenuated by scavengers of NO or HNO. This shift is selectively induced by ET-1 because in our previous study of CVD patient pericardial resistance arteries, BK-induced relaxation of contractile responses to the TXA2-analogue U46618, but not ET-1, were significantly attenuated by L-NAME and indomethacin(Leurgans et al., 2016). H2O2 did not play a major role because inhibition of SOD, which is an important albeit not exclusive source of the peroxide, had only a small effect and especially because exogenous catalase did not reduce the potency or efficacy of BK in the presence of ET-1. This seemingly contrasts with previous reports from several laboratories including our own(Beyer et al., 2017; Ellinsworth, Sandow, Shukla, Liu, Jeremy & Gutterman, 2016; Freed, Beyer, LoGiudice, Hockenberry & Gutterman, 2014; Leurgans et al., 2016; Matoba et al., 2000; Munoz et al., 2018; Schulz, Katunaric, Hockenberry, Gutterman & Freed, 2019; Shimokawa & Morikawa, 2005). We previously evaluated the effect of exogenous catalase during simultaneous inhibition of NOS, sGC, COX and endothelial Ca2+-activated K+-channels(Leurgans et al., 2016). Almost by definition, continuous presence of both L-NAME and indomethacin has been part of most studies of endothelium-dependent hyperpolarization and evaluations of involvement of endothelium-derived H2O2 using catalase. Indomethacin was absent in our present experiments. It may be of interest to analyze in future studies how cyclooxygenase products modulate the role of H2O2 as an endothelium-derived hyperpolarizing factor in the microvascular wall. Earlier conclusions were strengthened by us and others with imaging studies using dyes that were designed to demonstrate reactive oxygen species within cells and not in the extracellular space. In line with this, PEGylated-catalase is used in recent investigations of H2O2 in flow-induced endothelium-dependent dilatation of coronary and subcutaneous arterioles from patients with and without coronary artery disease(Beyer et al., 2017; Ellinsworth, Sandow, Shukla, Liu, Jeremy & Gutterman, 2016; Schulz, Katunaric, Hockenberry, Gutterman & Freed, 2019). Several other mechanisms have been proposed to mediate endothelium-dependent vasodilatation in addition to endothelium-derived prostaglandins, NO, HNO and H2O2(Ellinsworth, Sandow, Shukla, Liu, Jeremy & Gutterman, 2016; Shimokawa & Godo, 2020; Vanhoutte, Shimokawa, Feletou & Tang, 2017). These include alternative chemical mediators and electrical communication via heterocellular gap junctions. Although we did not investigate the latter, it can not explain that endothelium-dependent relaxation of depolarization-induced contraction could be inhibited by L-NAME and ODQ but not by a scavenger of NO or by an inhibitor of SOD.
To better define the molecular mechanisms underlying the observed profiles of endothelium-dependent vasodilatation, indirect pharmacological analyses of small number of vessels from the type of patients that we investigated will not suffice. The patients and their vessels varied considerably in several aspects. For instance, not only the amplitudes of the contractions and of the endothelium-dependent relaxations (Figure 1) but also the extent to which BK-induced relaxation could be inhibited by L-NAME (Figure 10), varied considerably between individual patients. As a result, a large number of experiments was needed to detect for instance i) a significant difference of the maximal relaxation between patients with and without coronary artery disease and ii) that the relaxing response in the presence of ET-1 was not significantly modified by L-NAME. Experimental animal research is here not an immediate or straightforward solution. To the best of our knowledge, no type of vessel from an animal model has been described with comparable vascular pharmacological properties. Future studies can focus on human resistance arteries that are more widely available and that can be harvested from groups of individuals with or without cardiovascular and non-cardiovascular disease. To this end, human subcutaneous and omental resistance arteries can be considered(Beyer et al., 2017; Munoz et al., 2018; Shimokawa & Morikawa, 2005; Zinkevich, Fancher, Gutterman & Phillips, 2017). Pharmacological experiments using a standardized protocol in vessels from many diverse individuals, can be of value provided that these individuals are characterized in detail in terms of not only their demographics and standard risk factors. Large numbers can form the basis of association studies. Selection of novel circulating factors can be hypothesis-driven and lead to confirmatory studies in dedicated experimental animals. These approaches were recently applied with respect to the role of a subtype of NADPH oxidases (NOX5) in the development of regional endothelial dysfunction and systolic hypertension with ageing(Elbatreek et al., 2020).
In resistance arteries from patients with CVD requiring surgery, we confirmed that endothelium-dependent relaxation can be inhibited by L-NAME and ODQ during depolarization-induced contraction but not in the presence of ET-1. A role for NO, HNO or H2O2 in these endothelium-dependent relaxations could not be confirmed. Alternative mechanisms seem to be involved. They do not depend on reactive oxygen species and are resistant to elevated levels of superoxide anions.