GYY4137 and Sodium Hydrogen Sulfide Relaxations Are Inhibited by L-Cysteine and KV7 Channel Blockers in Rat Small Mesenteric Arteries

Donors of H2S may be beneficial in treating cardiovascular diseases where the plasma levels of H2S are decreased. Therefore, we investigated the mechanisms involved in relaxation of small arteries induced by GYY4137 [(4-methoxyphenyl)-morpholin-4-yl-sulfanylidene-sulfido-λ5-phosphane;morpholin-4-ium], which is considered a slow-releasing H2S donor. Sulfides were measured by use of 5,5′-dithiobis-(2-nitro benzoic acid), and small rat mesenteric arteries with internal diameters of 200–250 µm were mounted in microvascular myographs for isometric tension recordings. GYY4137 produced similar low levels of sulfides in the absence and the presence of arteries. In U46619-contracted small mesenteric arteries, GYY4137 (10−6–10–3 M) induced concentration-dependent relaxations, while a synthetic, sulfur-free, GYY4137 did not change the vascular tone. L-cysteine (10−6–10–3 M) induced only small relaxations reaching 24 ± 6% at 10–3 M. Premixing L-cysteine (10–3 M) with Na2S and GYY4137 decreased Na2S relaxation and abolished GYY4137 relaxation, an effect prevented by an nitric oxide (NO) synthase inhibitor, L-NAME (Nω-nitro-L-arginine methyl ester). In arteries without endothelium or in the presence of L-NAME, relaxation curves for GYY4137 were rightward shifted. High extracellular K+ concentrations decreased Na2S and abolished GYY4137 relaxation suggesting potassium channel-independent mechanisms are also involved Na2S relaxation while potassium channel activation is pivotal for GYY4137 relaxation in small arteries. Blockers of large-conductance calcium-activated (BKCa) and voltage-gated type 7 (KV7) potassium channels also inhibited GYY4137 relaxations. The present findings suggest that L-cysteine by reaction with Na2S and GYY4137 and formation of sulfides, inhibits relaxations by these compounds. The low rate of release of H2S species from GYY4137 is reflected by the different sensitivity of these relaxations towards high K+ concentration and potassium channel blockers compared with Na2S. The perspective is that the rate of release of sulfides plays an important for the effects of H2S salt vs. donors in small arteries, and hence for a beneficial effect of GYY4137 for treatment of cardiovascular disease.


INTRODUCTION
Hydrogen sulfide (H 2 S) is considered an essential signaling molecule in the cardiovascular and nervous systems (Szabó, 2007;Wallace et al., 2018) and a variety of pathophysiological changes including cancer, glycometabolic disorders, diabetes, sepsis, and human malignt hyperthermia are associated with altered endogenous levels of H 2 S (Szabó, 2007;Szabo and Papapetropoulos, 2017;Vellecco et al., 2020). In the cardiovascular system, endogenous H 2 S can lead to both vasodilatation and vasoconstriction (Li et al., 2015;Hedegaard et al., 2016;Gheibi et al., 2018).
GYY4137 exhibits vasorelaxant, hypotensive, antiinflammatory, and anti-cancer activity effects (Li et al., 2008;Martelli et al., 2013a;Lee et al., 2011;Liu et al., 2013), and it is considered as a slow-releasing H 2 S donor (Li et al., 2008;Feng et al., 2015). Different mechanisms have been reported to be involved in the release of sulfide species from GYY4137, including conversion by cystathionine γ-lyase (CSE) (Chitnis et al., 2013) and interaction with cysteine (Martelli et al., 2013b). In a recent study where changes in H 2 S gas were detected with microelectrodes, we observed that GYY4137 induced full relaxation of small mesenteric arteries without producing detectable changes in amperometric currents (Hedegaard et al., 2016). Hence, it is unclear whether H 2 S gas contributing to the pharmacodynamic effects of GYY4137 is below detection level or whether GYY4137 induces vasodilatation by mechanisms independent of H 2 S gas e.g., commercial GYY4137 has dichlormethane leading to formation of carbon monoxide (CO) (Alexander et al., 2015).
To examine whether sulfides are involved in GYY4137 relaxations, measurements of sulfides were conducted and compared to a hydrolyzed version of GYY4137 (Alexander et al., 2015). Na 2 S was chosen for comparison as its vasodilating effects previously have been associated with increases in H 2 S gas (Hedegaard et al., 2016). To investigate whether CSE or L-cysteine contribute to the release of H 2 S from GYY4137, the effect of an inhibitor of CSE and L-cysteine were examined on GYY4137 relaxation. Release of H 2 S from Na 2 S and GYY4137 appear to have different kinetics and that may change the involvement of potassium channels, and therefore relaxations induced by the two compounds were investigated in the presence of blockers of potassium channels. Small mesenteric arteries contribute significantly to vascular resistance and blood pressure in intact animals (Fenger-Gron et al., 1995), and therefore the vasodilatation studies were performed in rat small mesenteric arteries.

Animals and Preparation of Samples
The investigation was carried out in accordance to the Guide for the Care and the Use of Laboratory Animals published by the United States National Institutes of Health (NIH Publication No. 85-23, revised 1996) and followed the ARRIVE guidelines (McGrath and Lilley, 2015). Adult male Wistar rats (12-14 weeks) were killed by decapitation and subsequent exsanguination. The protocol was approved by the Danish Animal Experiments Inspectorate (permission 2014-15-2934-0159).

Measurements of Release of Sulfide Species From GYY4137
Sulfide species (H 2 S and HS − ) released from GYY4137 were assessed in vitro as described previously (Li et al., 2008). In brief, 100 mM phosphate buffer pH 7.40 was incubated with 1.0 or 0.1 mM GYY4137 at 25 or 37°C. A phosphate buffer with pH of 3.01 was also tested as acidic conditions have been shown to promote H 2 S release from GYY4137 (Li et al., 2008;Hedegaard et al., 2016). At appropriate times, 20 µl aliquots were removed and added to 96-well microplates containing 50 µl 1 mM DTNB and 50 µl 1 M HEPES buffer pH 8.0, and absorbance was measured at 412 nm on a plate reader. The concentration of H 2 S formed from GYY4137 was calculated from a standard curve of NaSH (1-500 µM) for each of the respective time points.

Functional Studies in Small Mesenteric Arteries
Third branch mesenteric arteries were dissected from the mesenteric vascular bed and mounted on 40 µm steel wires in microvascular myographs (Danish Myotechnology, Aarhus, Denmark) for isometric tension recording as previously described (Mulvany and Halpern, 1976). The vessels were equilibrated in oxygenated (5% CO 2 , 20% O 2 , 75% N 2 ) PSS at 37°C and for 30 min, and by stretching normalized to a lumen diameter (d 100 ) equivalent to 100 mm Hg, after which tension was set to 90% x d 100 (Mulvany and Halpern, 1976). At this tension, the internal lumen diameters were 200-250 µm. After normalization, the arterial segments were stimulated with KPSS, washed in PSS, and stimulated with noradrenaline (10 µM). Arteries were only included if they developed an active force corresponding to a transmural pressure of 100 mmHg. The PowerLab data system and Chart 5.5 (ADInstruments, Oxfordshire, United Kingdom) was used to record the data. The mechanical responses of the vessel segments were measured as active wall tension (ΔT), which is the changes in force (ΔF) divided by twice the segment length (2L).

Experimental Protocol
To determine whether the effects of GYY4137 were due to H 2 S released from it, the parent compound was compared with an analog, which is normally produced by a two-step hydrolytic degradation of GYY4137 over weeks, but in this study synthesized as previously described (Alexander et al., 2015).
The biosynthetic enzymes for H 2 S production (CBS and CSE) were reported to be involved in the relaxant effects of GYY4137 in bovine ciliary arteries (Chitnis et al., 2013). Therefore, small mesenteric arteries were incubated with D, L-propargylglycine (PPG, 10 mM), which is an irreversible inhibitor of CSE, and concentration-response curves were constructed for Na 2 S and GYY4137. L-cysteine is considered substrate for formation of endogenous H 2 S (Wang, 2012), and was also reported as a scavenger of HS − giving rise to formation of sulfides . Concentration-response curves for L-cysteine were constructed in U46619 (0.3 µM)-contracted preparations. To investigate an eventual scavenger effect, we conducted two set of experiments. In a first set of experiments, solutions of cysteine and, respectively, Na 2 S and GYY4137 were pre-mixed in airtight containers and then added to the arteries contracted with U46619 at 0 and 10 h after the mixing. At 10 h the amount of sulfides was also measured using the DTNB assay as described above. In a second series of experiments, small mesenteric arteries were incubated with L-cysteine (10 -3 M) or the thiol reducing agent, 1,4-dithiothreitol (10 −3 M), and concentration-response curves were constructed for GYY4137 and Na 2 S. The control and examination of drugs were run in parallel, and only one concentration-response curve was constructed for each vasodilator per animal.
To investigate the role of the endothelium in relaxations induced by Na 2 S and GYY4137, arterial segments with and without endothelium were mounted. The endothelial cells were removed by introducing into to the lumen a human scalp hair and rubbing back and forth several times (Hedegaard et al., 2016). The effectiveness of the procedure was assessed by absence of relaxation to acetylcholine in noradrenaline-contracted arteries, while vessel with endothelium were accepted only if acetylcholine-induced (10 -5 M) relaxation on noradrenalin-induced (5 × 10 -6 M) contraction was larger than 50%, and exclusion following these criteria explains unequal group numbers are reported. The preparations were contracted with U46619 (0.3 µM) giving a contraction level corresponding to 50-60% of the contractions induced by 125 mM KPSS, and when the contraction was stable cumulative concentration-response curves were constructed for Na 2 S (10 −6 -10 -3 M) or GYY4137 (10 −6 -10 -3 M). Preparations were incubated with the NO synthase inhibitor, L-NAME (300 µM), the preparations were contracted with U46619 (0.1 µM) to obtain contraction similar to controls levels, concentration-response curves were constructed for increasing concentrations of Na 2 S, GYY4137, and acetylcholine.
High concentrations of the NO donor, sodium nitroprusside (SNP) with NaSH yield formation of a nitrosothiol and inhibits NO-induced rat aorta relaxation (Ali et al., 2006;Whiteman et al., 2006). To investigate the interaction with NO, concentrationresponse curves for SNP (10 −9 -10 -4 M) were obtained in the absence and presence of GYY4137 or Na 2 S.
The involvement of K channels in Na 2 S and GYY4137 induced relaxation were examined by comparing relaxations in U46619 FIGURE 2 | Effect of L-cysteine on Na 2 S and GYY4137 relaxations in rat small mesenteric arteries. Mixture of Na 2 S (300 µM) or GYY4137 (1 mM) with and without L-cysteine (1 mM) was added to air-tight bottles and relaxations measured to time 0 and 10 h for (A) Na 2 S and (B) GYY4137, and the amount of sulfides measured at 10 h for (C) Na 2 S and GYY4137 in the absence, and for (D) Na 2 S and GYY4137 in the presence of L-cysteine, where the concentration of sulfides generated by L-cysteine were subtracted. Please note the difference in scale comparing data in (C and D). Data are means ± SEM (n 6). *p < 0.05, Students t-test.

Data and Statistical Analysis
All data were presented as mean ± S.E.M. with a significance level of p < 0.05, and n representing the number of individual animals (n > 5 for each protocol). Statistical comparisons between H2S release at time 0 and 10 h by GYY4137 and Na2S were performed by Student´s t-test. The two-way analysis of variance (ANOVA) was used to compare the different conditions affecting release of sulfide species from GYY4137 and concentration-response curves obtained in functional studies of isolated mesenteric arteries. The assumptions of the ANOVA approach were verified by inspection of Q-Q plots. The graphs and statistical analyses were performed using GraphPad Prism 7.0 (GraphPad Software, La Jolla, CA).

Role of H 2 S in Na 2 S and GYY4137 Relaxation
The release of free H 2 S from GYY4137 was examined spectrophotometrically by the use of DTNB. Following previous studies (Li et al., 2008), incubation of 0.1 mM GYY4137 at pH 7.4 25°C resulted in a slow-release, which reached an end value of 8.33 µM after 90 min of incubation (Supplementary Figure S1A). This release was augmented by an increased starting concentration of GYY4137 (1 mM) (Supplementary Figure S1A). Sulfide release was also significantly increased under acidic conditions and by increased temperatures (Supplementary Figures S1B,C  respectively).
L-cysteine by itself increased the spectrophotometrically measured absorbance, but there was a further increase in sulfide release by combining cysteine and GYY4137 (Supplementary Figure S1D). The presence of rat mesenteric artery did not affect sulfide release from GYY4137 ( Figure 1A). The results suggest that the release of sulfides from GYY4137 is independent of the presence of mesenteric artery.
Concentration-response curves were obtained in U46619 (0.3 µM)-contracted preparations for GYY4137 and the hydrolyzed product of GYY4137, GYY4137H to investigate whether the relaxant effects of GYY4137 were due to H 2 S release. We found that GYY4137 induced concentrationdependent relaxations while there was no change in vascular tone by adding the GYY4137H (Figures1B,C), suggesting that release of sulfides is pivotal for GYY4137 relaxation.
In U46619-contracted arteries with endothelium, the H 2 S substrate, and thiol-containing amino acid, L-cysteine (10 −6 -10 -2 M) induced small relaxations which were 24 ± 6% at 10 -3 M, and these relaxations were not inhibited in the presence of PPG ( Figure 1D). Pre-mixing L-cysteine (10 -3 M) with either Na 2 S or GYY4137 and adding it immediately after the mixing or 10 h later showed, that the presence of L-cysteine inhibited Na 2 S and GYY4137 relaxations (Figures 2A,B). The relaxations induced by Na 2 S and GYY4137 were comparable 10 h after the mixing of the solutions (Figures 2A,B), and this was also the case for the amount of sulfides measured in the solutions ( Figures 2C,D). After preincubation with L-cysteine (10 −3 M), instead of relaxations,3 × 10 -6 to 10 -3 M Na 2 S and 10 −4 -10 -3 M GYY4137 induced contractions, which at the highest FIGURE 4 | Effect of the cystathionine γ-lyase (CSE) inhibitor, PPG, and DTT on Na 2 S and GYY4137 relaxations in rat mesenteric arteries (A) Effect of PPG (1 mM) on Na 2 S induced vascular relaxation (B) effect of PPG on GYY4137 induced vascular relaxation (C) effect of DTT (1 mM) on Na 2 S induced vascular relaxation (D) effect of DTT on GYY4137 induced vascular relaxation. Where no error bars are indicated, error lies within dimensions of the symbol. *p < 0.05, two-way ANOVA. All data is represented as mean ± SEM (n 5-6).
Frontiers in Pharmacology | www.frontiersin.org March 2021 | Volume 12 | Article 613989 concentrations were contractions followed by relaxations ( Figures 3A-D). In the presence of both L-NAME and L-cysteine, Na 2 S-induced contractions were reduced ( Figures  3C,G), while the inhibitory effect of L-cysteine on GYY4137 relaxation was prevented ( Figures 3F,H). To examine whether enzymatic conversion by CSE or interaction with thiol-groups play a role for Na 2 S and GYY4137 relaxations, the preparations were pre-incubated with an inhibitor of CSE, PPG (10 -3 M), or the thiol reducing agent DTT (10 -3 M), but these treatments did not change concentration-response curves for Na 2 S and GYY4137 (Figure 4). Taken together our results show that L-cysteine converts Na 2 S and GYY4137 relaxations to contraction associated with markedly higher sulfide concentrations. This effect of L-cysteine on vascular tone was partly reversed in the presence of the endothelial NO synthase inhibitor, L-NAME.

Effect of Endothelial Cell Removal and NO in Na 2 S and GYY4137 Relaxation
In contrast to L-cysteine (10 -3 M), incubation with L-NAME significantly rightward shifted concentration-response curves for acetylcholine relaxation in small mesenteric arteries ( Figure 5A). In U46619-contracted arteries, Na 2 S induced concentrationdependent relaxations, which were of similar magnitude in vessel segments with and without endothelium ( Figure 5B), while concentration-response curves for GYY4137 were significantly rightward shifted in vessels without endothelium ( Figure 5C). In the presence of an inhibitor of NO synthase, L-NAME (10 -4 M), the concentration-response curves for Na 2 S were unaltered ( Figure 5B), while L-NAME rightward shifted concentration-response curves for GYY4137 ( Figure 5C). These results suggest that in rat small mesenteric arteries, endotheliumderived NO is of importance for some of the effects of GYY4137 on vascular tone, while there were no significant differences for Na 2 S.
To investigate the effect of Na 2 S and GYY4137 on NO donorinduced relaxations, the small mesenteric arteries were incubated with vehicle, Na 2 S (3 × 10 -4 M), or GYY4137 (10 -3 M), then contracted to the same level with U46619 (0.3 µM), and increasing concentrations of SNP was added. We found that in the presence of Na 2 S, concentration-response curves for SNP were leftward shifted, while the presence of GYY4137 did not change the relaxation responses for SNP in small mesenteric arteries ( Figure 6).

Involvement of K Channels in GYY4137 and Na 2 S-Induced Vascular Relaxation
In preparations contracted with high extracellular potassium (60 mM KPSS), relaxations induced by GYY4137 were abolished, while relaxations induced by Na 2 S were significantly decreased compared with responses obtained in U46619 (0.3 µM)-contracted preparations ( Figures 7A-D). In contrast to GYY4137, Na 2 S still induced 60% maximum relaxation in 60 mM KPSS-contracted preparations ( Figures 7E,F), hence FIGURE 5 | Role of the endothelium and NO for Na 2 S and GYY4137 relaxation (A) Average relaxations induced by acetylcholine (ACh) in the absence (n 7) and the presence of either L-NAME (100 µM) (n 6) or L-cysteine (1 mM) (n 7) in rat mesenteric arteries (B) Na 2 S induced vascular relaxation in vessels with (n 11) and without endothelium (n 5), or with endothelium in the presence of an NO synthase inhibitor, L-NAME (n 6) (C) GYY4137 induced vascular relaxation in vessels with (n 12) and without endothelium (n 6) or with endothelium in the presence L-NAME (n 6). Where no error bars are indicated, error lies within dimensions of symbol. All data are means ± SEM. *p < 0.05, two-way ANOVA. suggesting that K channels are pivotal for GYY4137-induced relaxations, while K channels and also other mechanisms contribute to Na 2 S relaxation. To investigate the K channel subtypes involved in the relaxations, the preparations were incubated with blockers of ATP-sensitive K channels (glibenclamide), BK Ca (iberiotoxin), and of K V 7 channels (XE991). Glibenclamide decreased Na 2 S relaxation, while GYY4137 relaxation was unaltered in U46619-contracted arteries ( Figures 8A,B). Iberiotoxin and XE991 significantly decreased relaxations induced by both Na 2 S and GYY4137 ( Figures 8C-F).

DISCUSSION
The main findings in the present study are that GYY4137 spontaneously releases low amounts of sulfides leading to relaxation, and that L-cysteine by direct chemical interaction inhibits Na 2 S and GYY4137 relaxations. The observation of sulfide release is supported by the tissue-independent effect on sulfide release measured from GYY4137 and that the hydrolyzed control, GYY4137H, in contrast to the parent molecule, fails to relax small mesenteric arteries. Moreover, Na 2 S induced comparable relaxations after dissolving at 0 h compared to 10 h storage, while GYY4137 relaxation was markedly increased by storage for 10 h in airtight containers and yielded sulfide accumulation similar to Na 2 S, and suggesting GYY4137 is associated with slow release of H 2 S. These findings suggest rate of H 2 S release plays an essential role for the effect on vascular tone, where high levels of H 2 S from Na 2 S interacts leftward shift concentration-response curves for the NO donor SNP, while low levels of H 2 S from GYY4137 interact with endogenous endothelium-derived NO leading to relaxation. Moreover, blockers of K ATP , BK Ca , and K V 7 channels affected Na 2 S and GYY4137 concentration-response curves differently. It has previously been shown that GYY4137 releases H 2 S (Li et al., 2008;Whiteman et al., 2010;Martelli et al., 2013b), and in agreement with these studies, we found that GYY4137 concentration-dependently releases small amounts of sulfides. This release is markedly enhanced by increasing the temperature and acidifying the solutions. In contrast, at physiological conditions (pH 7.4, 37°C), we found by simultaneous measurements of H 2 S gas and relaxation that GYY4137 caused relaxation of small rat arteries without releasing detectable amounts of H 2 S gas (Hedegaard et al., 2016). However, the lack of relaxant effect of the hydrolyzed GYY4137 control compound, GYY4137H (Alexander et al., 2015), suggests that GYY4137-induced vessel relaxation requires H 2 S.
Several H 2 S releasing compounds with slow releasing rates, including organic polysulphides of garlic, e.g., diallyl disulfide and arylthiamides require the presence of reduced glutathione or thiols to release H 2 S (Benavides et al., 2007;Martelli et al., 2013b). However, in the presence of arterial tissue, the amount of sulfides measured from GYY4137 was not increased suggesting the H 2 S release is tissue-independent. Moreover, Na 2 S induced comparable relaxations immediately after dissolving the salt compared to solutions stored in airtight containers for 10 h, while GYY4137 stored for 10 h yielded sulfide accumulations similar to Na 2 S and markedly increased relaxation. These findings suggest rate of H 2 S release plays an essential role for the effect on vascular tone of GYY4137.
Plasma L-cysteine concentrations are in the range of 3.5-11 μmol/L (Chawla et al., 1984) and L-cysteine is considered one of the primary substrates leading to formation of H 2 S. It has at high concentrations been found to increase formation of H 2 S in mammalian tissues such as kidney (Jackson-Weaver et al., 2013), and to cause relaxations in small cerebral (Streeter et al., 2012) and retinal arteries (Takır et al., 2015), although 1-300 μM L-cysteine had no effect (Takır et al., 2015). In agreement with the latter study, we only observed small relaxations by adding increasing concentrations of L-cysteine to U46619-contracted arteries and no effect on acetylcholine relaxation. In rat mesenteric arteries, the expression of CSE is high in perivascular and adventitial tissue and associated with formation of H 2 S (Jackson-Weaver et al., 2011;. In the present study, we carefully removed adhering tissue and cannot exclude L-cysteine will contribute to endogenous H 2 S formation to a larger degree in mesenteric arteries with adhering perivascular tissue. L-cysteine is considered a scavenger of nitroxyl (HNO) (Andrews et al., 2009), and studies in cell cultures reported that L-cysteine by direct interaction may scavenge HS − and lead to formation of inactive sulfides (Miyamoto et al., 2017). By mixing L-cysteine with Na 2 S or GYY4137, we observed increased accumulation of sulfides ( Figure 2). Considering FIGURE 6 | Effect of Na 2 S and GYY4137 on relaxations induced by the NO donor, sodium nitroprusside (SNP) in small mesenteric arteries. The arteries were incubated with and inhibitor of NO synthase, L-NAME (10 µM), and Na 2 S (300 µM) (n 6), or GYY4137 (1 mM) (n 5), contracted with U46619 (0.3 µM) and concentration-response curves constructed for SNP. *p < 0.05, two-way ANOVA. All data are means ± SEM.
Frontiers in Pharmacology | www.frontiersin.org March 2021 | Volume 12 | Article 613989 8 that L-cysteine by itself only had small effect on vascular tone and did not change acetylcholine relaxation, the effect of L-cysteine on Na 2 S and GYY4137 may be ascribed to a direct chemical reaction, and thereby inactivating Na 2 S and GYY4137 relaxation.
In previous studies, NaSH and Na 2 S induced contraction followed by relaxation is observed in the perfused mesenteric vascular bed and in isolated arteries (Ali et al., 2006;Di Villa Bianca et al., 2011;Hedegaard et al., 2016). In the presence of L-cysteine, low concentrations of Na 2 S and GYY4137 induced marked contractions of the isolated rat mesenteric arteries, and when L-cysteine was mixed with Na 2 S or GYY4137, we observed an increased sulfide accumulation. Polysulfides (H 2 S 2 and H 2 S 3 ) have been suggested to play a role in the effects of H 2 S or to produce many of the effects previously attributed to H 2 S (Kimura, 2019), but the effects of these unstable compounds were reported to be inhibited in the presence of L-cysteine (Miyamoto et al., 2017), and in previous studies we observed that polysulfides (K 2 S n ) induce relaxations in rat mesenteric arteries (Hedegaard et al., 2016). Therefore, it seems unlikely that L-cysteine by interaction with Na 2 S and GYY4137 just leads to formation of inactive sulfides. Instead the conversion of the Na 2 S and GYY4137 relaxations to contractions by L-cysteine treatment may result from inhibition of polysulfides. However, inhibition of formation of polysulfides with DTT (Miyamoto et al., 2017) did not affect Na 2 S and GYY4137 relaxation ( Figures 4C,D). Therefore, a more speculative mechanism is that L-cysteine together with Na 2 S or GYY4137 result in the formation of a product that may interfere with endothelial NO synthase leading to contraction, e.g., formation of cIMP instead of cGMP by NO as described in large coronary arteries (Chen et al., 2014) ( Figure 9A). Indeed in the presence of L-cysteine, L-NAME by inhibition of NO synthase restored the GYY4137 relaxations. In agreement with these findings we in preliminary studies observed in 1 H-, 31 P-NMR spectra that L-cysteine incubation with GYY4137 formed a product (results not shown) which support the formation of a product that may interfere with vascular tone in small mesenteric arteries, although other experimental approaches will be required to characterize the product formed by L-cysteine and GYY4137.
While 500 and 1,000 nmol/kg Na 2 S failed to change blood pressure in normotensive rats (Tomasova et al., 2015), intravenous injection of 39 μmol/kg Na 2 S decreased mean arterial blood pressure with 45 mmHg in anesthetized mice (Eberhardt et al., 2014). Treatment with 56 μmol/kg/day of NaSH administered by intraperitoneal injection lowered also blood pressure in spontaneously hypertensive rats (Ni et al.,  2018). These findings suggest that high doses of H 2 S salts lowers blood pressure by vasodilatation. As mentioned in the introduction, several mechanisms have been suggested to mediate NaSH and Na 2 S vasodilatation, depending on the vascular preparations that have been studied. In previous studies in resistance arteries, H 2 S vasodilatation was found to involve K ATP channels (Tang et al., 2005), BK Ca channels (Jackson-Weaver et al., 2011;Jackson-Weaver et al., 2013), and K V 7 channels (Schleifenbaum et al., 2010;Hedegaard et al., 2016).
In patch-clamp studies H 2 S gas 30 μM to 1 mM caused activation of K ATP channels in vascular smooth muscle from mesenteric arteries (Tang et al., 2005), and 10 µM NaSH hyperpolarized mesenteric arteries by iberiotoxin-sensitive mechanism also suggesting the involvement BK Ca channels (Jackson-Weaver et al., 2011, and NaHS (1 mM) hyperpolarized rat aorta and directly activated K V 7 channels in CHO cells (Martelli et al., 2013a). Recent studies have also shown that direct activation of K V 7 channels by H 2 S donors protects against neuropathic pain (Di Cesare Mannelli et al., 2017), and that direct persulfidation of K V 7 channels by H 2 S plays an important role in skeletal muscle hypercontractility in human malignant hyperthermia syndrome (Vellecco et al., 2020). In agreement with studies in resistance arteries, in small mesenteric arteries contracted with the U46619, Na 2 S in the present study induced relaxations sensitive to high extracellular potassium and blockers of both ATP-sensitive, K V 7, and BK Ca channels suggesting involvement of these channels in Na 2 S relaxation, although electrophysiological measurements, e.g., membrane potential measurements or patch-clamp will be required to confirm the activation of K V 7 channels by Na 2 S in this preparation.
Interaction of H 2 S with the NO pathway is thought to be important for the vascular effects of Na 2 S. Thus, it was proposed that NO and H 2 S may act co-operatively to generate nitroxyl (HNO), and that this activates transient receptor potential ankyrin 1 (TRPA1) channels on sensory nerves with subsequent calcitonin gene-related peptide release and relaxation in meningeal and mouse mesenteric arteries (Eberhardt et al., 2014). In constrast, based on studies in mice with downregulation of CSE, it was suggested that physiological concentrations of H 2 S scavenge endothelium-derived NO, and in the absence of NO leads to activation of smooth muscle K ATP and K V channels (Szijártó et al., 2018). In the present study, endothelial cell removal or inhibition of NO synthase failed to change relaxations induced by Na 2 S in mesenteric arteries. These findings agree with our previous studies showing that NaSH relaxation in rat mesenteric arteries is NO and endothelium-independent. However, incubating the preparations with Na 2 S leftward shifted concentration-response curves for an exogenous NO donor, SNP suggesting that high concentrations of Na 2 S and NO synergistically cause vasodilatation in rat mesenteric arteries. Moreover, our results support that Na 2 S causes relaxation by activation of K channels in the smooth muscle layer. High concentrations of Na 2 S also relax contractions induced by high extracellular potassium suggest that K channel independent mechanisms are involved ( Figure 5A), and may similar to NaSH involve inhibition of mitochondrial complex I and III (Hedegaard et al., 2016).
The mechanism of GYY4137 induced vascular relaxation has previously been observed to be endothelium-dependent in rat aorta rings (Li et al., 2008). In small mesenteric arteries, GYY4137 relaxations were reduced in preparations without endothelium, and by an inhibitor of NO synthase suggesting endotheliumderived NO is involved in relaxations induced by GYY4137. Interestingly, incubation with GYY4137 failed to change relaxations induced by SNP suggesting that high H 2 S FIGURE 9 | Schematic figure proposing the mechanisms involved in (A) contraction in small mesenteric arteries induced by Na 2 S and GYY4137 in the presence of L-cysteine. This involve a formation of a sulfide product (XSn) that may modulate guanylate cyclase leading to contraction of the vascular smooth muscle cells (VSMC). The effect is partly reversed by the NO synthase inhibitor, L-NAME (B) Relaxation induced by Na 2 S and GYY4137 leads to, respectively, fast and slow release of H 2 S and followed by activation of subsets of different potassium channels: ATP-sensitive K (K ATP ) channels by Na 2 S, and of large-conductance calcium activated (BKCa) and voltage-gated type 7 (K V 7) channels by GYY4137.
Frontiers in Pharmacology | www.frontiersin.org March 2021 | Volume 12 | Article 613989 concentrations are required to act synergistically with an NO donor, but also implying that the interaction of GYY4137 with endothelium-derived NO is likely at smooth muscle level. In aorta segments and ciliary arteries K ATP channels were found involved in GYY4137 relaxation (Li et al., 2008;Chitnis et al., 2013). Here, we provide evidence that potassium channels may play a pivotal role in the vascular relaxations induced by GYY4137, as high extracellular potassium completely inhibited GYY4137 relaxation. Also, blockers of smooth muscle K V 7 and BK Ca channels, XE991 and iberiotoxin markedly inhibited relaxation, suggesting these channels are involved in relaxations induced by GYY4137 in rat mesenteric arteries ( Figure 9B). Therefore, the mechanisms involved in GYY4137 relaxation regarding both the endothelium and involvement of K channels seems different from the mechanisms involved in Na 2 S relaxation, reflecting different rate and levels of H 2 S reaching the vascular preparations when Na 2 S salt and GYY4137 are added to an organ bath in similar conditions.

CONCLUSION AND PERSPECTIVES
The present findings suggest that L-cysteine by reaction with Na 2 S and GYY4137 and formation of sulfides, inhibits relaxations by these compounds. The low rate of release of H 2 S species from GYY4137 is reflected by the different sensitivity of these relaxations towards high K + concentration and K channel blockers compared with Na 2 S. The perspective is that the rate of release of sulfides plays an important for the effects of H 2 S salt vs. donors in small arteries.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by Danish Animal Experiments Inspectorate permission 2014-15-2934-01059.