Hydrogen sulfide dilates rat mesenteric arteries by activating endothelial large-conductance Ca²⁺-activated K⁺ channels and smooth muscle Ca²⁺ sparks

Abstract

We have previously shown that hydrogen sulfide (H₂S) reduces myogenic tone and causes relaxation of phenylephrine (PE)-constricted mesenteric arteries. This effect of H₂S to cause vasodilation and vascular smooth muscle cell (VSMC) hyperpolarization was mediated by large-conductance Ca²⁺-activated potassium channels (BK_Ca). Ca²⁺ sparks are ryanodine receptor (RyR)-mediated Ca²⁺-release events that activate BK_Ca channels in VSMCs to cause membrane hyperpolarization and vasodilation. We hypothesized that H₂S activates Ca²⁺ sparks in small mesenteric arteries. Ca²⁺ sparks were measured using confocal microscopy in rat mesenteric arteries loaded with the Ca²⁺ indicator fluo-4. VSMC membrane potential (E_m) was measured in isolated arteries using sharp microelectrodes. In PE-constricted arteries, the H₂S donor NaHS caused vasodilation that was inhibited by ryanodine (RyR blocker), abluminal or luminal iberiotoxin (IbTx, BK_Ca blocker), endothelial cell (EC) disruption, and sulfaphenazole [cytochrome P-450 2C (Cyp2C) inhibitor]. The H₂S donor NaHS (10 μmol/l) increased Ca²⁺ sparks but only in the presence of intact EC and this was blocked by sulfaphenazole or luminal IbTx. Inhibiting cystathionine γ-lyase (CSE)-derived H₂S with β-cyano-l-alanine (BCA) also reduced VSMC Ca²⁺ spark frequency in mesenteric arteries, as did EC disruption. However, excess CSE substrate homocysteine did not affect spark activity. NaHS hyperpolarized VSMC E_m in PE-depolarized mesenteric arteries with intact EC and also hyperpolarized EC E_m in arteries cut open to expose the lumen. This hyperpolarization was prevented by ryanodine, sulfaphenazole, and abluminal or luminal IbTx. BCA reduced IbTx-sensitive K⁺ currents in freshly dispersed mesenteric ECs. These results suggest that H₂S increases Ca²⁺ spark activity in mesenteric artery VSMC through activation of endothelial BK_Ca channels and Cyp2C, a novel vasodilatory pathway for this emerging signaling molecule.

hydrogen sulfide (H₂S) is a recently described vasodilator produced in the vasculature by the enzymes cystathionine γ-lyase (CSE) and 3-mercaptopyruvate sulfurtransferase (3MST). H₂S has been proposed to cause vasodilation through a variety of mechanisms (1, 3, 26, 33) and genetic knockout of the CSE gene causes hypertension (32). We previously reported that inhibiting CSE or disrupting the endothelium enhances myogenic tone in small mesenteric arteries from rats and that exogenous H₂S hyperpolarizes vascular smooth muscle cell (VSMC) membrane potential (E_m) and dilates arteries through activation of large-conductance Ca²⁺-activated K⁺ (BK_Ca) channels (7).

Ca²⁺ sparks are spatially and temporally limited Ca²⁺-release events from ryanodine receptor Ca²⁺-release channels (RyR) in the sarcoplasmic reticulum (SR) of VSMCs and have been shown to activate VSMC BK_Ca channels to cause hyperpolarization, leading to a reduced open probability of L-type voltage-gated Ca²⁺ channels (VGCCs) and a decrease in cytosolic [Ca²⁺] (21). Ca²⁺ spark frequency can be increased by activation of transient receptor protein (TRP) channels (4) and has been hypothesized to act as an intrinsic negative-feedback mechanism to regulate stretch-induced VSMC depolarization and myogenic tone (10). Although RyR activity can be directly modulated in VSMCs (12), endothelial cell (EC)-derived factors can also increase VSMC Ca²⁺ spark activity (4, 8, 18), supporting a role for the endothelium in regulating VSMC RyR. Both nitric oxide (NO) and carbon monoxide (CO), two endothelium-derived dilators that function similarly to H₂S in many tissues, have been shown to modulate spark activity (20).

Based on the previous description of VSMC BK_Ca channel activation by Ca²⁺ sparks downstream of CO and NO and our observations of EC-dependent dilation by H₂S, we hypothesized that H₂S activates Ca²⁺ sparks in an EC-dependent manner to contribute to VSMC hyperpolarization and vasodilation.

METHODS

Animals.

Male Sprague-Dawley rats (275–325 g) were used for all studies. On the day of the experiments, animals were euthanized with pentobarbital sodium (200 mg/kg ip) and mesenteric arteries were isolated for dilation, Ca²⁺ imaging, and E_m studies. All animal protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the University of New Mexico School of Medicine and conform to National Institutes of Health guidelines for animal use.

Isolated vessel preparation.

Intestinal arcades were removed and placed in a Silastic-coated Petri dish containing chilled physiological saline solution (PSS; in mmol/l: 129.8 NaCl, 5.4 KCl, 0.83 MgSO₄, 0.43 NaH₂PO₄, 19 NaHCO₃, 1.8 CaCl₂, and 5.5 glucose). Fourth- or fifth-order artery segments (inner diameter < 100 μm) were dissected from the mesenteric vascular arcade and transferred to a vessel chamber (Living Systems Instrumentation). Segments (1–2 mm in length) were cannulated onto glass micropipettes and secured with silk ligatures. The arteries were pressurized to 100 mmHg with PSS using a servocontrolled peristaltic pump (Living Systems Instrumentation) and superfused with warmed, oxygenated PSS at a rate of 5 ml/min.

Vasodilation studies.

Arterial inner diameter was recorded in cannulated arteries using edge-detection software (IonOptix). Arteries were equilibrated at 37°C in warmed, oxygenated PSS for 30 min prior to the start of the experiment. Following equilibration, arteries were preconstricted to 50% resting diameter using phenylephrine (PE) and vasodilation was measured during cumulative addition of the H₂S donor NaHS. Arteries were incubated with Ca²⁺-free PSS (in mmol/l: 129.8 NaCl, 5.4 KCl, 0.83 MgSO₄, 0.43 NaH₂PO₄, 19 NaHCO₃, 3.7 tetrasodium EGTA, and 5.5 glucose) to determine passive diameter at the end of the experiment.

Fluo-4 imaging.

Arteries used for VSMC Ca²⁺ spark imaging studies were incubated in a fluo-4 AM (10 μmol/l, Invitrogen) solution containing 0.25% pluronic acid in HEPES buffer (in mmol/l: 134 NaCl, 6 KCl, 1 MgCl₂, 2 CaCl₂, 10 HEPES, 0.026 EDTA, and 10 glucose) for 60 min at 28°C prior to cannulation. After loading with fluo-4, arteries were transferred to a vessel chamber and cannulated as described above. To measure endothelial Ca²⁺ events, arteries were first cannulated and then the endothelium was selectively loaded with fluo-4 AM through the lumen of the artery. After 10 min equilibration in oxygenated PSS at 32°C, arteries were excited at 488 nm by a solid-state laser, and emitted light > 500 nm was collected using an Olympus IX71 microscope with a 60X water-immersion lens and a spinning-disk confocal scanning unit (Andor). A 75 × 50-μm area was imaged at 50–60 Hz using laser power of 30%. All Ca²⁺ spark imaging studies were conducted in pressurized arteries in the absence of exogenous PE.

Spark analysis.

Collected images were analyzed using SparkAn software, developed by A. D. Bonev and M. T. Nelson (University of Vermont). Ten images without spark activity were averaged to determine background fluorescence levels (F₀). A minimum F/F₀ of 1.2 was required for an event to be considered a spark. Each spark site was analyzed as a region of interest (ROI; 25 pixels² or 3 μm²). Each image contained 15–25 cells, and spark frequency, amplitude, and duration were averaged for all ROIs in all visible cells to provide a single value for each of these parameters for each artery. These values were then averaged for all arteries in the study.

E_m recordings.

Small mesenteric arteries were dissected and cannulated as described above. VSMCs were impaled with glass microelectrodes (tip resistance 40–120 MΩ) filled with 1 mol/l KCl. Endothelial E_m was measured by impaling ECs in arteries cut open to expose the luminal surface. A Neuroprobe amplifier (A-M Systems) was used for recording E_m. Analog output from the amplifier was low pass filtered at 1 kHz and recorded and analyzed using Axoscope software (Axon Instruments). Criteria for acceptance of E_m recordings were 1) an abrupt negative deflection of potential as the microelectrode was advanced into a cell; 2) stable membrane potential for at least 1 min; and 3) an abrupt change in potential to ∼0 mV after the electrode was retracted from the cell.

Whole cell patch-clamp studies in isolated ECs.

Mesenteric arteries were cut into 2-mm segments and exposed to mild digestion solution containing 0.2 mg/ml dithiothreitol and 0.2 mg/ml papain in HEPES buffer for 45 min at 37°C. Arteries were removed from the digestion solution and placed in 1 ml of HEPES buffer containing 2 mg/ml BSA. Single ECs were released by gentle trituration with a small-bore Pasteur pipette and were stored at 4°C until use (for up to 5 h). One to two drops of the resulting cell suspension were seeded on a glass cover slip mounted on an inverted fluorescence microscope (Olympus IX71) for 30 min prior to superfusion. Single ECs were identified by the selective uptake of the fluorescently labeled acylated low-density lipoprotein Ac-LDL-Dil using a rhodamine filter (29) prior to each electrophysiological experiment. ECs were superfused under constant flow (2 ml/min) at room temperature (22–23°C) in an extracellular solution (in mmol/l: 141 NaCl, 4.0 KCl, 1 MgCl₂, 1 CaCl₂, 10 HEPES, 10 glucose, buffered to pH 7.4 with NaOH). Whole cell current data were generated using an Axopatch 200B amplifier (Axon Instruments). Biophysical criteria (seal resistance > 1 GΩ, series resistance < 25 MΩ) were checked following membrane rupture and monitored throughout the course of the experiment. Cells were held at −60 mV and were dialyzed for 5 min with an intracellular solution (in mmol/l: 140 KCl, 0.5 MgCl₂, 5 Mg₂ATP, 10 HEPES, 1 EGTA adjusted to pH 7.2 with KOH). CaCl₂ was added to yield a free-Ca²⁺ concentration of 1 μmol/l, as calculated using WinMAXC chelator software.

Immunohistochemistry.

Intestinal arcades were removed and placed in ice-cold HEPES buffer. Vessels were dilated with 100 μM papaverine for 10 min at room temperature and then frozen in Tissue-Tek O.C.T. Compound (VWR International). Frozen tissue was cut in 10-μm cross-sections and thaw mounted on slide. Slides were air dried for 20 min and then fixed with 2% formalin for 15 min at room temperature. Following a 5 min wash in phosphate-buffered saline (PBS; Mediatech) slides were blocked and permeabilized with PBS + 0.1% Triton X-100 (J. T. Baker) + 2% donkey serum (Sigma) for 25 min at room temperature. Slides were stained overnight at 4°C with primary antibodies against cystathionase (1:100; AbCam) and BK_Ca (1:50; Alamone Labs) in PBS + 2% donkey serum + 0.1% Tween-20 (Bio-Rad). DyLight 549 donkey anti-mouse and DyLight 649 donkey anti-rabbit secondary antibodies (1:200; Jackson Immunoresearch) used to detect cystathionase and BK_Ca, respectively, were applied to slides for 1.5 h at room temperature. Nuclei were stained with Sytox Green (1:5,000; Invitrogen) for 15 min at room temperature.

Statistical analysis.

Data are presented as means ± SE and were analyzed using one-way or two-way ANOVA as appropriate with Student-Newman-Keuls post hoc analysis for differences between groups, concentrations, and interactions for studies with multiple groups (SigmaStat). Paired data were analyzed using a paired Student's t-test. P < 0.05 was considered statistically significant for all analyses.

RESULTS

H₂S vasodilation.

The H₂S donor NaHS produced a robust vasodilation in PE-constricted arteries, and this dilation was significantly reduced by EC disruption (Fig. 1A). Due to recent evidence that H₂S can activate EC BK_Ca channels (34), we measured vasodilation in arteries following BK_Ca channel blockade with IbTx administered in the superfusion (abluminal) or in the lumen (Fig. 1B). Both routes of IbTx administration abolished dilation to NaHS. To assess the EC-specific effect of luminal IbTx application, VSMC E_m was measured in isolated arteries with luminal or abluminal IbTx in the absence of PE. VSMC E_m was depolarized by abluminal IbTx whereas luminal IbTx had no significant effect (Fig. 1C), suggesting that luminal IbTx does not cross the EC layer.

Fig. 1.

Small mesenteric artery dilation to increasing concentrations of the H₂S donor NaHS in endothelial cell (EC)-intact and EC-disrupted arteries (A) and the presence of vehicle or iberiotoxin (IbTx) (100 nmol/l) administered abluminally or luminally (B). Arteries were pressurized to 100 mmHg and preconstricted to ∼50% with PE. C: vascular smooth muscle cell (VSMC) membrane potential (E_m) measured by sharp microelectrode impalement in the absence of PE, with and without luminal or abluminal IbTx application. n = 5–6 per group. *P < 0.05 vs. baseline, #P < 0.05 vs. luminal IbTx.

Endogenous and exogenous H₂S regulation of Ca²⁺ sparks.

Importantly, endogenous H₂S produced by CSE opposes myogenic tone in small mesenteric arteries (7). To determine whether endogenous H₂S activates Ca²⁺ sparks, the CSE inhibitor BCA was applied to arteries and Ca²⁺ spark frequency was assessed. Example traces of Ca²⁺ spark activity under basal and NaHS-stimulated conditions are shown in Fig. 2, A and B, respectively. BCA reduced Ca²⁺ spark frequency, as did EC disruption, and these effects were not additive (Fig. 2C). Conversely, adding exogenous NaHS (10 μmol/l) to mesenteric arteries significantly increased spark frequency (Fig. 2C). To determine whether the decreased spark frequency observed following BCA treatment could be an effect of homocysteine accumulation due to CSE inhibition, Ca²⁺ sparks were measured in the presence of exogenous homocysteine. Addition of homocysteine did not decrease spark frequency in a manner parallel to CSE inhibition with BCA (Fig. 2D). Although 3-MST is expressed in these arteries (data not shown), BCA completely mimics H₂S scavenging, suggesting that vasoactive H₂S is synthesized primarily by CSE in these small mesenteric arteries.

Fig. 2.

Example traces of Ca²⁺ spark activity in two regions of interest (ROI) in VSMC of mesenteric arteries loaded with fluo-4 and pressurized to 100 mmHg in the presence of vehicle (A) or β-cyano-l-alanine (BCA) (B). C: summary data showing Ca²⁺ spark frequency in arteries under EC-intact vehicle conditions or with BCA (100 μmol/l; EC intact), disrupted EC, both BCA and EC disruption, or NaHS (10 μmol/l; EC intact). ROIs = 1.7 μm per side. D: Ca²⁺ spark frequency before and after addition of exogenous homocysteine. E: the change in endothelial cell Ca²⁺ as measured by fura-2 AM in response to NaHS (10 μmol/l) or acetylcholine (10 μmol/l). F: the frequency of endothelial Ca²⁺ events before and after addition of NaHS (10 μmol/l) measured with the Ca²⁺ indicator fluo-4. n = 5–6 per group. *P < 0.05 vs. vehicle, #P < 0.05 vs. basal, %P < 0.05 vs. NaHS.

EC Ca²⁺, H₂S vasodilation, and Ca²⁺ sparks.

It has been proposed that EC K⁺ channel activation increases EC [Ca²⁺], which can activate several EC enzymes to produce vasodilators (27). The effect of NaHS on EC [Ca²⁺] was assessed in small mesenteric arteries. NaHS did not elicit a detectable change in global EC [Ca²⁺] as measured with the Ca²⁺ indicator fura-2 AM (Fig. 2E). Conversely, acetylcholine caused a robust change in EC [Ca²⁺]. Localized Ca²⁺ events were also measured in EC of mesenteric arteries in response to NaHS. There were few EC Ca²⁺ events under basal conditions, but frequency of these events was increased upon addition of NaHS (Fig. 2F).

EC-derived vasodilators such as 11,12-epoxyeicosatrienoic acid (11,12-EET) produced by cytochrome P-450 epoxygenase can activate VSMC Ca²⁺ sparks (4). Therefore, NaHS vasodilation was measured in the presence of the Cyp2C inhibitor sulfaphenazole (10 μmol/l); inhibition of Cyp2C reduced vasodilation (Fig. 3A) using a concentration that we have previously demonstrated decreases synthesis of 11,12-EET without affecting synthesis of other EET products (5). Additionally, adding ryanodine at a concentration that eliminates Ca²⁺ sparks (10 μmol/l; Fig. 3B) diminished vasodilation to the H₂S donor NaHS (Fig. 3C).

Fig. 3.

A: NaHS dilation curves in the presence of vehicle and the cytochrome P-450 2C (Cyp2C) inhibitor sulfaphenazole (10 μmol/l) in arteries preconstricted to ∼50% with PE. B: effect of 10 μmol/l ryanodine on Ca²⁺ spark frequency in cannulated small EC-intact mesenteric arteries loaded with fluo-4. C: NaHS dilation curves in the presence of vehicle and the RyR inhibitor ryanodine (10 μmol/l) in arteries preconstricted to ∼50% with PE. n = 5–6 per group. *P < 0.05 vs. vehicle.

H₂S-induced Ca²⁺ spark activity.

The mechanism of NaHS-induced VSMC Ca²⁺ spark activity was assessed. The increase in spark frequency elicited by NaHS was blocked by EC disruption, sulfaphenazole, or luminal IbTx (Fig. 4A). NaHS actually reduced Ca²⁺ spark frequency in EC-disrupted arteries. NaHS did not affect Ca²⁺ spark amplitude (Fig. 4B) or duration (Fig. 4C), except in the presence of luminal IbTx, in this case reducing amplitude and extending duration. EC disruption alone reduced spark amplitude, and amplitude was not affected by the presence of NaHS in the EC-disrupted group (Fig. 4B).

Fig. 4.

A: summary data of Ca²⁺ spark frequency at 100 mmHg in the presence of vehicle or NaHS (10 μmol/l) under control conditions (EC intact), EC disruption, sulfaphenazole (Sulf; 10 μmol/l; EC intact), or luminal IbTx (100 nmol/l; EC intact). Sparks were recorded before and after addition of NaHS in the same artery. Spark amplitude (B) and duration (C) at half-maximal amplitude under the same conditions as in A. n = 5–7 per group. *P < 0.05 vs. vehicle, #P < 0.05 vs. control within treatment.

H₂S regulation of E_m and EC K⁺ current.

Ca²⁺ sparks mediate vasodilation through activation of VSMC BK_Ca channels and subsequent E_m hyperpolarization of VSMC (21). We therefore assessed the effect of NaHS on VSMC E_m in small mesenteric arteries depolarized with 1 μmol/l PE. Representative recordings are shown in Fig. 5A. NaHS caused a hyperpolarization of mesenteric artery VSMC E_m measured by sharp electrodes (Fig. 5B). This E_m hyperpolarizing effect of NaHS was prevented with ryanodine, sulfaphenazole, abluminal IbTx, or luminal IbTx (Fig. 5B). The effect of luminal IbTx suggested an effect of H₂S on EC BK_Ca channels; therefore, whole cell outward K⁺ currents were measured in freshly dispersed mesenteric artery EC to directly assess this possibility. These cells demonstrate an IbTx-sensitive current (Fig. 5C), demonstrating a contribution of BK_Ca channels to the K⁺ current. Inhibition of either BK_Ca or CSE reduced K⁺ currents. The combination of IbTx and BCA also decreased currents but currents in the presence of both inhibitors were not different from either BCA or IbTx alone. The effect of NaHS on EC E_m was also assessed in the absence and presence of IbTx and is shown in Fig. 6. IbTx depolarized EC E_m under basal conditions. NaHS hyperpolarized EC E_m, but this was prevented in the presence of IbTx.

Fig. 5.

A: example recordings of sharp microelectrode impalements of VSMC in arteries pressurized to 100 mmHg and depolarized with 1 μmol/l PE in the presence of vehicle or NaHS under control conditions. Negative deflection of trace indicates cell impalement; positive deflection to 0 mV indicates retraction of electrode. B: summary data of E_m recordings in EC-intact arteries in the presence of vehicle or NaHS under control conditions or in the presence of ryanodine (10 μmol/l), sulfaphenazole (10 μmol/l), or IbTx (100 nmol/l) in superfusate (abluminal) or in the artery lumen. n = 4–5 per group. C: effect of BCA (100 μmol/l) and IbTx (100 nmol/l) or IbTx + BCA on whole cell K⁺ currents in freshly dispersed mesenteric EC. n = 5–7 per group. *P < 0.05 vs. PE (B); #P < 0.05 vs. control or vehicle (B and C).

Fig. 6.

EC E_m measured in the absence or presence of NaHS (10 μmol/l) with and without IbTx (100 nmol/l) pretreatment in mesenteric arteries cut open to expose the endothelium. n = 6–7 per group. *P < 0.05 vs. vehicle, #P < 0.05 vs. baseline.

Vascular expression of CSE and BK_Ca.

CSE and BK_Ca expression were assessed in the mesenteric vasculature by immunohistochemistry. CSE expression appeared to be more localized to the endothelial and adventitial layers (Fig. 7A), whereas BK_Ca staining was apparent throughout the vascular wall (Fig. 7B). Overlay of images from the same vessel revealed that both CSE and BK_Ca are expressed in the endothelium (Fig. 7, C and D).

Fig. 7.

Representative images of cystathionine γ-lyase (CSE) (A), large-conductance Ca²⁺-activated potassium channels (BK_Ca) (B), overlay of the images in A and B (C), and mgnification of C (D).

DISCUSSION

Our findings show that H₂S dilates small mesenteric arteries through activation of EC BK_Ca channels and Cyp2C with downstream activation of VSMC Ca²⁺ sparks. In addition, a Cyp2C product is necessary for H₂S-induced dilation. H₂S is a recently described vasodilator and novel modes of action are being described at a rapid pace, but this is the first demonstration of an H₂S-initiated endothelium-dependent mechanism mediated in part by activation of VSMC Ca²⁺ sparks. These results are consistent with our previous observation in the same vascular bed that dilation and VSMC E_m hyperpolarization evoked by the H₂S donor NaHS require BK_Ca channels (7). It is also consistent with a recent study in vas deferens smooth muscle showing that H₂S can directly activate BK_Ca channels through a redox mechanism (14), and a study in cerebral arteries demonstrating activation of VSMC Ca²⁺ sparks by H₂S (15).

The effect of H₂S on the activity of BK_Ca channels is a contentious issue. As noted above, some studies suggest H₂S activates the channel (7, 15, 17, 28, 34), whereas others find H₂S inhibits BK_Ca activity (13, 30, 31). Because the cell types in these two groups of studies do not overlap, it is possible that tissue-specific expression of subunits or other associated proteins modify the effect of H₂S on BK_Ca channels. The demonstration that H₂S activates EC BK_Ca channels (34) led us to test whether EC BK_Ca channels were the IbTx-sensitive target of H₂S in mesenteric arteries. We used luminal application of IbTx to specifically target EC BK_Ca channels as previously described (6). Importantly, luminal application of IbTx did not significantly affect VSMC E_m although abluminal application of the same concentration depolarized VSMC (Fig. 1C). In contrast, luminal application of IbTx blocked NaHS-induced dilations similarly to abluminal application of the BK_Ca blocker (Fig. 1B). Furthermore, EC disruption also blocked NaHS dilation (Fig. 1A). We observed expression of both BK_Ca and the H₂S-producing enzyme CSE in the endothelium of the mesenteric vasculature (Fig. 7). Thus multiple functional tests suggest EC BK_Ca channels are critically important in the vasodilatory response to H₂S in small mesenteric arteries and that EC contain the primary target for this compound.

K⁺ channel activation has been demonstrated to increase EC [Ca²⁺] (11, 27). This finding, however, is a controversial issue and the opposite relationship, that of EC [Ca²⁺] activating K⁺ channels, has also been demonstrated (29). Recent studies demonstrated that NaHS increases microvascular and aortic EC [Ca²⁺] (19, 22); our findings are consistent with these studies (Fig. 2F). In conjunction with the current finding that NaHS-induced vasodilation requires intact endothelium and, specifically, EC BK_Ca channels, one of these two mechanisms, or perhaps both, appear to contribute to H₂S mediated dilation. In either case, increased EC [Ca²⁺] should elevate the production of endothelial dilators, several of which have been shown to affect Ca²⁺ spark signaling. For example, NO increases spark frequency and EC denudation decreases basal spark activity in cerebral artery VSMC (18). The EC product CO also increases spark frequency and enhances coupling between sparks and “spontaneous transient outward currents” (STOCs) in cerebral artery VSMCs (8). Finally, the EC-derived Cyp2C product 11,12-EET increases spark frequency downstream of VSMC TRP vanilloid type 4 (TRPV4) channel activation in cerebral arteries (4), and cytochrome P-450 enzymes contribute to NaHS-mediated relaxation in the perfused mesenteric bed (3). Consistent with these last two observations, our data demonstrate that inhibiting Cyp2C with sulfaphenazole blocks NaHS-induced dilation (Fig. 3A) and hyperpolarization of VSMC (Fig. 5B). However, mesenteric arteries display extensive myoendothelial electrical coupling through gap junctions (25). Thus it is possible that the endothelium-dependent vasodilation seen in these experiments occurs without a diffusible factor released from the EC. Furthermore, the Cyp2C product 11,12-EET could act on EC TRPV4 channels to mediate its effect (29), suggesting the need for future experiments to address these questions.

A major downstream target for Cyp2C products is VSMC BK_Ca channel activation by Ca²⁺ sparks, and we further investigated this pathway. Rather than increasing global intracellular [Ca²⁺], Ca²⁺ sparks increase [Ca²⁺] locally in the subsarcolemmal space to increase the open probability of BK_Ca channels, measured electrophysiologically as STOCs (21). Therefore, the net effect of Ca²⁺ sparks is E_m hyperpolarization, reduced open probability of VGCCs, and a reduction in global intracellular [Ca²⁺] (9, 21).

We observed that concentrations of NaHS that cause dilation also increase VSMC Ca²⁺ spark activity (Fig. 2C) in pressurized arteries but this activation is absent in arteries with abraded EC (Fig. 4A). This is consistent with H₂S releasing an endothelial factor that subsequently activates Ca²⁺ spark activity rather than acting directly on the VSMC as recently described in cerebral arteries (15). Sulfaphenazole administration and luminal IbTx administration abolished NaHS stimulation of increased spark frequency (Fig. 4A). Thus NaHS dilation in small mesenteric arteries appears to be at least partly mediated through enhanced VSMC Ca²⁺ spark activity dependent on EC BK_Ca channels and Cyp2C activation. In EC-disrupted arteries, NaHS in fact reduced Ca²⁺ spark frequency, suggesting H₂S may inhibit sparks in mesenteric VSMC through a mechanism separate from effects on EC BK_Ca channels and Cyp2C. This is in contrast to observations in cerebral arteries where a portion of the H₂S-induced dilation appears to be caused by direct effects on VSMC SR to increase spark activity (15). However, similar to the studies in cerebral arteries, H₂S minimally affected spark kinetics (Fig. 4, B and C). Importantly, both ryanodine and IbTx largely blocked NaHS-induced vasodilation (Figs. 1 and 3), as expected if Ca²⁺ spark activation of VSMC BK_Ca channels contributes to the H₂S effect. However, dilation induced by concentrations of NaHS above 10 μmol/l is not inhibited by these treatments, similar to our previous observations (7) suggesting these concentrations of NaHS exert nonspecific effects.

We also observed a pressure effect on NaHS-induced dilation. In the current study arteries were held at 100 mmHg luminal pressure and preconstricted to ∼50% of resting diameter with PE. Under these conditions, NaHS induced a robust dilation (∼70%) at relatively low concentrations (1 μmol/l), whereas in arteries pressurized to 60 mmHg luminal pressure, NaHS elicited a much smaller dilation (7). This enhanced H₂S-induced vasodilation with increasing luminal pressure is consistent with H₂S acting as a hyperpolarizing agent and an endogenous inhibitor of myogenic tone (7, 15). That is, since increasing luminal pressure stimulates VSMC depolarization and myogenic tone, H₂S-induced vasodilation is amplified at higher pressures, perhaps due to enhanced spark-hyperpolarization coupling at these pressures.

NaHS hyperpolarized VSMC E_m by ∼7 mV (Fig. 5B). Consistent with this effect being mediated by Ca²⁺ spark activation, superfusing with ryanodine to block Ca²⁺ spark generation or with IbTx to block VSMC BK_Ca channels abolished the hyperpolarization (Fig. 5B). Luminal IbTx, sulfaphenazole and EC inactivation also prevented hyperpolarization, suggesting both EC and VSMC BK_Ca channels participate in H₂S-induced VSMC hyperpolarization in intact arteries. The observation that outward K⁺ currents in EC were greatly diminished by BCA (Fig. 5C) provides additional support that endogenous H₂S activates EC BK_Ca channels in an autocrine manner. Indeed, endogenous H₂S activation of EC BK_Ca channels appears to account for about half of the resting BK_Ca current in mesenteric artery EC. This is consistent with recent reports that H₂S can activate BK_Ca channels by a redox method (14) and that H₂S increases the open probability of BK_Ca in cultured mesenteric artery endothelial cells (34).

Based on these studies, we propose that in small mesenteric arteries, H₂S activates EC BK_Ca channels, increasing EC [Ca²⁺] to activate Cyp2C and TRPV4 to increase Ca²⁺ spark activity in VSMC (4). VSMC sparks then activate VSMC BK_Ca channels to promote vasodilation. This hypothesized pathway explains several apparently disparate observations of H₂S dilation in small arteries. Specifically, it has been observed that H₂S 1) activates BK_Ca channels in microvascular EC (34); 2) increases microvascular EC [Ca²⁺] (22); 3) dilates perfused mesenteric beds in a cytochrome P-450, K_Ca channel, and EC-dependent manner (2, 3, 3); and 4) hyperpolarizes small mesenteric artery VSMC E_m in a BK_Ca-dependent manner (7). In this model, EC BK_Ca channels are the primary target of H₂S in small mesenteric arteries.

It is not clear why abluminal NaHS activates EC BK_Ca channels but does not directly activate VSMC BK_Ca channels. We have described differences in Ca²⁺ sensitivity of BK_Ca channels in EC and VSMC (23, 24), directly demonstrating potential differences in regulation of BK_Ca channels in these two cell types. Specifically, EC BK_Ca channels display a greater Ca²⁺ sensitivity despite the apparent lack of functional BK-β1 subunits, and H₂S may affect the moiety responsible for this difference. Future studies investigating the mechanism of H₂S activation of EC BK_Ca channels and the molecular differences between VSMC and EC BK_Ca channels are needed to resolve these questions.

In conclusion, the present work demonstrates for the first time that exogenous and endogenous H₂S act on the endothelium to activate VSMC Ca²⁺ sparks to elicit vasodilation, adding to and complementing the growing list of vascular effects of this molecule.

GRANTS

O. Jackson-Weaver was supported by National Heart, Lung, and Blood Institute (NHLBI) Grant HL-07736 and American Heart Association (AHA) Grant 10-PRE-4050028; J. M. Osmond by NHLBI Grant HL-07736 and AHA Grant 12-POST-8690008; M. A. Riddle by NHLBI Grant HL-07736; L. V. Gonzalez Bosc by NHLBI Grant HL-088151; B. R. Walker by NHLBI Grant HL-95640; and N. L. Kanagy by NHLBI Grant HL-82799.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: O.J.-W., J.O., L.V.G.B., B.R.W., and N.L.K. conception and design of research; O.J.-W., J.O., M.A.R., and J.S.N. performed experiments; O.J.-W., J.O., M.A.R., J.S.N., and N.L.K. analyzed data; O.J.-W., J.O., M.A.R., J.S.N., L.V.G.B., B.R.W., and N.L.K. interpreted results of experiments; O.J.-W., J.O., M.A.R., J.S.N., and N.L.K. prepared figures; O.J.-W. drafted manuscript; O.J.-W., J.O., J.S.N., L.V.G.B., B.R.W., and N.L.K. edited and revised manuscript; O.J.-W., J.O., M.A.R., J.S.N., L.V.G.B., B.R.W., and N.L.K. approved final version of manuscript.