KCa channel activation normalizes endothelial function in Type 2 Diabetic resistance arteries by improving intracellular Ca²⁺ mobilization

Abstract

Background:

Endothelial dysfunction is an early pathogenic event in the progression of cardiovascular disease in patients with type 2 diabetes (T2D). Endothelial KCa2.3 and KCa3.1 K⁺ channels are important regulators of arterial diameter, and we thus hypothesized that SKA-31, a small molecule activator of KCa2.3 and KCa3.1, would positively influence agonist-evoked dilation in myogenically active resistance arteries in T2D.

Methodology:

Arterial pressure myography was utilized to investigate endothelium-dependent vasodilation in isolated cremaster skeletal muscle resistance arteries from 22–24 week old T2D Goto-Kakizaki rats, age-matched Wistar controls, and small human intra-thoracic resistance arteries from T2D subjects. Agonist stimulated changes in cytosolic free Ca²⁺ in acutely isolated, single endothelial cells from Wistar and T2D Goto-Kakizaki cremaster and cerebral arteries were examined using Fura-2 fluorescence imaging.

Main Findings:

Endothelium-dependent vasodilation in response to acetylcholine (ACh) or bradykinin (BK) was significantly impaired in isolated cremaster arteries from T2D Goto-Kakizaki rats compared with Wistar controls, and similar results were observed in human intra-thoracic arteries. In contrast, inhibition of myogenic tone by sodium nitroprusside, a direct smooth muscle relaxant, was unaltered in both rat and human T2D arteries. Treatment with a threshold concentration of SKA-31 (0.3 μM) significantly enhanced vasodilatory responses to ACh and BK in arteries from T2D Goto-Kakizaki rats and human subjects, whereas only modest effects were observed in non-diabetic arteries of both species. Mechanistically, SKA-31 enhancement of evoked dilation was independent of vascular NO synthase and COX activities. Remarkably, SKA-31 treatment improved agonist-stimulated Ca²⁺ elevation in acutely isolated endothelial cells from T2D Goto-Kakizaki cremaster and cerebral arteries, but not from Wistar control vessels. In contrast, SKA-31 treatment did not affect intracellular Ca²⁺ release by the sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) inhibitor cyclopiazonic acid.

Conclusions:

Collectively, our data demonstrate that KCa channel modulation can acutely restore endothelium-dependent vasodilatory responses in T2D resistance arteries from rats and humans, which appears to involve improved endothelial Ca²⁺ mobilization.

1. INTRODUCTION

Vascular endothelium regulates the contractile state and health of the arterial wall via the synthesis and release of nitric oxide (NO) [1–3], which is critical for the dynamic regulation of blood pressure, vascular bed perfusion and the prevention of atherogenesis, thrombosis and smooth muscle cell proliferation in the vessel wall [4–6]. These processes are typically disturbed in endothelial dysfunction, an early causative event in the development and progression of diabetes-related cardiovascular disease [7–10]. From a mechanistic perspective, endothelial small- and intermediate-conductance, Ca²⁺-activated K⁺ channels (i.e. KCa2.3 and KCa3.1, respectively) contribute to the maintenance of vascular health and function, and arterial blood pressure through the regulation of NO synthesis and vascular tone in large and small arteries [11–13]. Elevation of cytosolic free Ca²⁺ in the endothelium in response to vasodilatory hormones (e.g. acetylcholine and bradykinin) stimulates KCa2.3 and KCa3.1 activities, and the associated hyperpolarizing current can augment endothelial Ca²⁺ entry and NO synthesis [14–16]. This endothelium-dependent hyperpolarization (EDH) also transfers to the adjacent smooth muscle via myo-endothelial junctions to reduce L-type Ca²⁺ channel activity and vascular tone [17, 18].

Pharmacological inhibition or genetic inactivation of endothelial KCa2.3 and/or KCa3.1 channels impairs stimulated NO production in endothelial cells and agonist-evoked vasodilation of isolated, myogenically active arteries [12, 19, 20]. In contrast, small molecule activators of KCa2.3 and KCa3.1 channels (e.g. SKA-31 and NS309) produce concentration-dependent inhibition of contractile tone and can potentiate agonist-induced vasorelaxation [21–26]. Collectively, these observations support the idea that vascular KCa channel activities may be exploited therapeutically to improve vascular function [13].

In the present study, we therefore hypothesized that pharmacological manipulation of endothelial KCa channel activities would augment agonist-evoked vasodilation in resistance arteries exhibiting endothelial dysfunction. To test this hypothesis, we examined stimulated vasodilation in cannulated, myogenically active resistance arteries from Goto-Kakizaki rats, a well-characterized model of non-obese, Type 2 Diabetes (T2D), and from patients with established T2D, in the absence and presence of SKA-31, a prototypic, small molecule activator of small- and intermediate-conductance KCa channels [27, 28]. Our results suggest that enhancement of endothelial KCa channels by a threshold concentration of SKA-31 (i.e. one producing minimal direct vasodilation) can augment the endothelium-dependent relaxation of myogenically active, T2D resistance arteries in response to primary vasodilatory stimuli, and may represent a novel strategy to mitigate T2D-associated cardiovascular dysfunction.

2. MATERIALS AND METHODS

2.1. Isolated vessel preparation

Animal handling and euthanasia procedures described in the present study were approved by the University of Calgary Animal Care Committee, and conform to the guidelines for the care and use of laboratory animals established by the Canadian Council on Animal Care and the NIH. Intra-mammary arterial graft tissue was obtained from patients undergoing coronary artery bypass surgery; patients participating in this study provided informed consent prior to tissue collection (see Supplementary Table 1 for profile of patient characteristics). Collection and use of these materials were approved by the Calgary Conjoint Health Research Ethics Board (Protocol #REB15–1364). Goto-Kakizaki male rats (22–24 weeks of age), along with age-matched Wistar control rats were obtained from Charles River Laboratories and housed under pathogen-free, standard conditions (e.g. 12 hour day/light cycle) with continuous access to food and water. Rats were injected intraperioneally with sodium pentobarbital (50 mg/kg) to induce surgical anesthesia (i.e. stage 3, loss of blink reflex), and cremaster muscles were then surgically removed and pinned out in a cooled (4°C) dissection chamber containing Krebs’ buffer (115 NaCl, 5 mM KCl, 25 mM NaHCO₃, 1.2 mM MgCl₂, 2 mM CaCl₂, 1.2 mM KH₂PO₄ and 10 mM D-glucose); pH was adjusted to 7.4 with 1 N NaOH [29]. Euthanasia was carried out using an overdose of sodium pentobarbital. In the case of cerebral arteries, rats were decapitated following anesthesia and the brains were placed in the dissection chamber. Single rodent and human arteries (1–2 mm in length) were isolated and cleaned in the dissection chamber, following which, arteries were cannulated on glass pipettes fitted in the pressure myograph chamber (Living Systems, Burlington, VT). To help maintain endothelial function throughout the duration of the experiment (i.e. up to 5 h), the vessel lumen was filled with Krebs’ buffer containing 1% bovine serum albumin, as described by Duling and others [30, 31]; pH was adjusted to 7.4 with NaOH. The cannulated vessel/chamber apparatus was placed on the stage of an inverted microscope, and the vessel was superfused with Krebs’ buffer at a constant flow of ~7 ml/min using a peristaltic pump and vacuum suction line. Bath solution was maintained at 34°C for rat cremaster vessels and 36–37°C for human intra-mammary arteries and was bubbled with 95% air/5% CO₂ gas. The intra-luminal pressure of cannulated vessels was increased in a step-wise manner under no-flow conditions and then maintained at 70 mmHg; vessels typically developed myogenic tone within 20–30 minutes. Any cannulated vessels that developed leaks or could not maintain their passive diameter following an initial pressurization step to 100 mmHg were discarded. Continuous video measurement of the intraluminal vessel diameter was carried out using a diameter tracking system (IonOptix, Milton, MA). Drugs were prepared from stock solutions and added to the bath via a perfusion pump.

Drug-induced changes in vessel internal diameter were calculated as a percentage of the developed myogenic tone relative to the maximal passive intraluminal diameter, according to the following equation:

$$\%\text{Inhibition}\text{of}\text{Myogenic}\text{Tone}=\left[\left({\text{D}}_{\text{drug}}{\text{-D}}_{\text{basal}}\right)/\left({\text{D}}_{\text{pass}}{\text{-D}}_{\text{basal}}\right)\right]\times 100,$$

Where D_drug = intraluminal diameter in the presence of ACh, BK or SNP, D_basal = intraluminal diameter under basal myogenic tone at 70 mmHg and D_pass = maximal intraluminal diameter at 70 mmHg in bath solution containing 2 mM EGTA and no added CaCl₂.

2.2. Endothelial cell isolation and live cell Ca²⁺ imaging

Cremaster skeletal muscle and brain cerebral resistance arteries (i.e. middle and posterior cerebral arteries) were dissected from either a single Wistar or Goto Kakizaki rat and endothelial cells were isolated in parallel from both types of arteries by enzymatic digestion using the following procedure. Five to six isolated arteries from a given tissue were incubated for 30 min at 36°C in low Ca²⁺ physiological saline solution (135 NaCl, 5 mM KCl, 1.2 mM MgCl₂, 0.1 mM CaCl₂, 1.2 mM KH₂PO₄, 10 mM D-glucose and 10 mM HEPES, pH adjusted to 7.4 with 1 N NaOH) containing 1 mg/ml DL-dithiothreitol and 26 U/ml papain (Sigma-Aldrich P4762). The solution was then replaced by fresh PSS containing 2 U/ml collagenase (Sigma-Aldrich C8051), 75 U/ml elastase (Calbiochem cat# 324682) and 1 mg/ml soybean trypsin inhibitor (ThermoFisher Scientific cat# 17075–029). Following incubation, vessels were gently titurated with a fire-polished Pasteur glass pipette and dispersed endothelial cells were transferred to a clean tube. Immunocytochemistry revealed that >90% of dispersed cells stained positively for eNOS protein. Freshly dissociated endothelial cells were loaded with the ratiometric fluorescent Ca²⁺ sensitive dye Fura-2 AM (2 μM final, Molecular Probes) + 0.02% (w/v) pluronic acid (F-127) in PSS containing 2 mM CaCl₂ at 37°C in a CO₂ incubator for 30 min. An aliquot of Fura-2 loaded endothelial cells was placed into a bath chamber on the stage of a Nikon TE2000 inverted microscope and allowed to settle and adhere to the poly-lysine coated glass coverslip for ~10 min. The bath chamber (0.3 ml volume) was then perfused continuously at 35°C with PSS containing 2 mM CaCl₂ at 1–2 ml/min using a gravity-fed, multi-reservoir perfusion system controlled by switchable solenoid valves. Pharmacological agents were added to individual solution reservoirs and entered the bath via a multi-port connector positioned at the bath input. To prepare the nominally Ca²⁺ free physiological saline, CaCl₂ was omitted from the solution; no chelator was added. Enzymatically dissociated endothelial cells were kept at room temperature and were typically used within 6 h of isolation. The number of Wistar and Goto-Kakizaki rats used for Fura-2 imaging in cremaster skeletal muscle and cerebral arterial endothelial cells is indicated in the figure legend for each experimental protocol. To minimize animal usage, we carried out Fura-2 imaging in cremaster and cerebral artery endothelial cells isolated on the same day from a single animal using carbachol stimulation in the absence and presence of SKA-31 (i.e. 2 separate stimulation protocols per cell type per day). A similar experimental strategy was used for CPA stimulated endothelial cells. The finite stability of Fura-2 fluorescence in our acutely loaded endothelial cells (~1 hour) and the occasional movement of cells on the recording coverslip with bath solution changes limited the duration of our experiments, and as a result, we were unable to consistently record Ca²⁺ dynamics in a given endothelial cell in both the absence and presence of SKA-31. This situation may contribute to heterogeneity in the data set, and represents a limitation of our study. Note that the representative tracings displayed for a given stimulus in the absence and presence of SKA-31 (i.e. Figure 5A/B and Figure 6A/B and C/D, Supplementary Figure 4A/B and Supplementary Figure 5) were taken from recordings on the same experimental day for a given strain and type of artery. Efforts to isolate endothelial cells from human intra-thoracic arteries using a similar dissociation procedure did not yield a sufficient number of viable dissociated cells, likely due to the abundance of connective tissue in these arteries.

Figure 5 -.

Carbachol-stimulated Ca²⁺ mobilization is impaired in endothelial cells from T2D Goto-Kakizaki cremaster arteries and improved by acute SKA-31 treatment. Panel A displays representative Fura-2 fluorescence tracings recorded from single endothelial cells acutely isolated from cremaster arteries from a single Wistar rat in response to stimulation with 1 μM carbachol. Cells displaying robust (F/F₀ > 1.5, R1-type) vs. modest (F/F₀ < 1.5, R2-type) intracellular Ca²⁺ release in response to carbachol are represented by black and blue colored traces, respectively. The horizontal bars above the tracings denote the treatment protocol over the course of the recording. The histogram below the tracings quantifies changes in Fura-2 fluorescence at common time points, numbered 1–4, during the experimental protocol. The magnitude of the peak Fura-2 signal under a given test condition (i.e. time points 2, 3 and 4) has been normalized to the basal Fura-2 fluorescence observed at time point #1. Panel B shows representative Fura-2 recordings from single, isolated Wistar cremaster arterial endothelial cells that were acutely exposed to 300 nM SKA-31 prior to the addition of 1 μM carbachol. The R1- and R2-type Ca²⁺ mobilizing responses are denoted by the black and blue traces, respectively. The histogram beneath the tracings quantifies changes in Fura-2 fluorescence similar to that described for panel A. For the histograms displayed in Figures 5A and 5B, a total of 24 and 18 endothelial cells were analyzed, respectively. Four independent experiments were performed to collect both data sets, in which a separate animal was used for each experiment. Asterisks (*) displayed in histograms signify a statistical difference in the magnitude of the Fura-2 fluorescence ratio between the R1 and R2 groups under a given experimental condition, as determined by a Mann-Whitney U test for non-parametric data; P<0.05. The middle histogram depicts the proportion of Wistar endothelial cells exhibiting a R1 vs. R2 pattern of intracellular Ca²⁺ mobilization in response to carbachol (i.e. time point #3) in either the absence or presence of 300 nM SKA-31 treatment. The SKA-31 induced increase in the proportion of cells exhibiting a R1-type Ca²⁺ mobilization response to carbachol was not statistically significant, as determined by a two-tailed Chi-squared test (P = 0.22). Panels C and D display representative recordings of Fura-2 fluorescence from single cremaster arterial endothelial cells acutely isolated from a T2D Goto-Kakizaki rat in response to carbachol stimulation in either the absence (panel C) or presence (panel D) of 300 nM SKA-31. The grey and red tracings depict cells with robust (R1) vs. modest (R2) release, respectively, of intracellular Ca²⁺ in response to carbachol. The histogram beneath panel C quantifies changes in Fura-2 fluorescence at common time points during the experiment, which are numbered 1–4 in the tracings. The histogram in panel D displays fluorescence data for carbachol-stimulated cells in the presence of 300 nM SKA-31. The magnitude of peak Fura-2 signal under each specified condition (i.e. time points 2, 3 and 4) is normalized to the basal Fura-2 fluorescence observed at time point #1. For the histograms displayed in Figures 5C and 5D, 37 and 50 endothelial cells were analyzed, respectively. Six independent experiments were performed to collect both data sets, in which a separate animal was used for each experiment. Asterisks (*) displayed in histograms signify a statistical difference in the magnitude of the Fura-2 fluorescence ratio between the R1 and R2 groups under a given experimental condition, as determined by a Mann-Whitney U test for non-parametric data; P<0.05. The middle histogram depicts the proportion of endothelial cells exhibiting R1 vs. R2-type intracellular Ca²⁺ mobilization in response to carbachol (i.e. time point #3) in either the absence or presence of 300 nM SKA-31 treatment. The effect of SKA-31 treatment on the proportion of cells exhibiting R1-type Ca²⁺ elevation in response to carbachol was statistically significant, as determined by a two-tailed Chi-squared test (P = 0.009).

Figure 6 -.

Endothelial cells from T2D Goto-Kakizaki cremaster arteries do not exhibit dampened intracellular Ca²⁺ release in response to the SERCA inhibitor cyclopiazonic acid (CPA). Panel A displays representative Fura-2 fluorescence recordings from single, cremaster arterial endothelial cells acutely isolated from a Wistar rat following bath application of 20 μM CPA. Cells displaying quantitatively distinct intracellular Ca²⁺ release in response to CPA treatment are represented by black (R1) and purple (R2) colored traces. The horizontal bars above the tracings denote the experimental conditions throughout the stimulation protocol. The histogram below the tracings quantifies changes in Fura-2 fluorescence at common time points, numbered 1–4, during the experimental protocol. The magnitude of the peak Fura-2 signal under a given test condition (i.e. time point 2, 3 and 4) has been normalized to the basal Fura-2 fluorescence observed at time point #1. Panel B shows representative Fura-2 recordings from single, isolated Wistar cremaster artery endothelial cells that were acutely exposed to 300 nM SKA-31 prior to the addition of 20 μM CPA. The R1- and R2-type Ca²⁺ mobilizing responses are denoted by the black and purple traces, respectively. The histogram beneath the tracings quantifies changes in Fura-2 fluorescence similar to that described for panel A. For the histograms displayed in Figures 6A and 6B, a total of 20 and 28 endothelial cells were analyzed, respectively. Four independent experiments were performed to collect both data sets, in which a separate animal was used for each experiment. Asterisks (*) displayed in histograms signify a statistical difference in the magnitude of the Fura-2 fluorescence ratio between the R1 and R2 groups under a given experimental condition, as determined by a Mann-Whitney U test for non-parametric data; P<0.05. The middle histogram depicts the proportion of Wistar endothelial cells exhibiting R1- vs. R2-type elevation of intracellular Ca²⁺ in response to CPA in either the absence or presence of 300 nM SKA-31 treatment. The SKA-31 induced increase in the proportion of cells exhibiting a R1-type Ca²⁺ mobilization response to CPA was not statistically significant, as determined by a two-tailed Chi-squared test (P = 0.37). Panels C and D display representative recordings of Fura-2 fluorescence from single cremaster arterial endothelial cells from a Goto-Kakizaki rat in response to CPA treatment in either the absence (panel C) or presence (panel D) of 300 nM SKA-31. The grey tracings depict cells with a R1 pattern of CPA-induced elevation of cytosolic Ca²⁺, whereas the green tracings depict cells with a R2 signature. The histogram beneath panel C quantifies changes in Fura-2 fluorescence corresponding to the common time points numbered 1–4 in the experimental tracings. The histogram in panel D presents fluorescence data in CPA-stimulated cells in the presence of 300 nM SKA-31. Peak Fura-2 fluorescence under a given test condition (i.e. time point 2, 3 and 4) is normalized to the basal Fura-2 fluorescence recorded at time point #1. For the histograms displayed in Figures 6C and 6D, 32 and 29 endothelial cells were analyzed, respectively. Six independent experiments were performed to collect both data sets, in which a separate animal was used for each experiment. Asterisks (*) displayed in histograms signify a statistical difference in the magnitude of the Fura-2 fluorescence ratio between the R1 and R2 groups under a given experimental condition, as determined by a Mann-Whitney U test for non-parametric data; P<0.05. The middle histogram depicts the proportion of endothelial cells exhibiting R1- vs. R2-type intracellular Ca²⁺ mobilization in response to CPA in either the absence or presence of 300 nM SKA-31 treatment. The SKA-31 induced increase in the proportion of cells exhibiting R1-type Ca²⁺ release activity was not statistically significant, as determined by a two-tailed Chi-squared test (P = 0.31).

Fura-2 loaded endothelial cells were placed in a temperature-controlled bath chamber on the stage of a Nikon TE2000S Eclipse inverted microscope equipped with a 20X Super Fluor objective (NA= 0.75). Cells were excited alternately for 200 msec using 340 and 380 nm wavelength light generated by a monochromator (Photon Technology Inc.) at 0.25 Hz. The switching time between the 340 and 380 nm excitation signals was <5 msec. Emitted fluorescence was collected by the objective and transmitted through a dichroic mirror with a center at 510 nm and bandwidth of 80 nm (Chroma Technology cat# 79001). Collected fluorescence was captured using a 4.2 megapixel CMOS camera and EasyRatio Pro acquisition software (Horiba Inc.). Cells were only selected for subsequent fluorescence analysis if they had a clear single cell appearance, were not obscured/associated with digested cellular debris, and did not change position over the course of the experiment (e.g. changes in bath flow often dislodged cells from the chamber bottom). Given these criteria, we were typically able to analyze < 10 cells in a given experimental trial. Fluorescence intensity recorded from individual cells was analyzed by manually outlining each selected cell within the field of view (i.e. regions of interest, 20x magnification) and plotting the ratio of the fluorescence intensity recorded in response to 340 and 380 nm excitation light over time. For quantification of stimulus-evoked Ca²⁺ mobilization in individual cells (i.e. F_test/F_basal), the level of stimulated Fura-2 fluorescence was normalized to basal Fura-2 fluorescence (i.e. F₀) recorded in normal physiological saline at the beginning of the experiment, prior to removal of external CaCl₂ and bath addition of the stimulus (refer to histograms shown in Figures 5 and 6).

Experimentally, we observed two distinct patterns of cytosolic Ca²⁺ elevation due to intracellular store release in stimulated, Fura-2 loaded endothelial cells. To address this phenomenon in a more quantitative manner, the magnitude of elevated cytosolic Ca²⁺ in the presence of either carbachol or cyclopiazonic acid was analyzed using a peak fluorescence-frequency distribution. The results revealed a bimodal-type of distribution for peak Fura-2 fluorescence, with a fluorescence minimum appearing around 1.5 (see Supplementary Figures 7 and 8). Based on these distributions, we grouped the stimulated Ca²⁺ responses into either a “R1” or “R2” category. Individual cells displaying a peak Fura-2 fluorescence ratio (i.e. F_test/F_basal) >1.5 were grouped into the R1 category, whereas cells with a peak Fura-2 ratio < 1.5 were labelled as R2 (see representative tracings in Figs. 5 and 6). In individual preparations of endothelial cells isolated from either cremaster or cerebral arteries, we did not observe Fura-2 loaded cell populations that were entirely either R1 or R2 in behavior in response to carbachol or CPA. In all cases, the R1:R2 proportion was >50% in Wistar endothelial cells and <50% in the T2D Goto-Kakizaki endothelial cells. In summary, this analytical approach revealed unexpected insights regarding stimulated intracellular Ca²⁺ elevation in control and T2D endothelial cells and the effects of acute SKA-31 treatment on this process.

2.3. Reagents

Acetylcholine chloride, bradykinin, carbamylcholine chloride, cyclopiazonic acid, DL-dithiothreitol, L-NAME (N^G nitro L-arginine methyl ester), indomethacin, sodium nitroprusside, DMSO (dimethyl sulfoxide), and all required chemicals to prepare physiological solutions were purchased from Sigma-Aldrich (Oakville, ON, Canada). Fura-2 acetoxymethylester was obtained from Molecular Probes/Invitrogen. Euthanyl (sodium pentobarbital, 250 mg/mL) was purchased from Bimeda-MTC Animal Health Inc, Cambridge, ON, Canada. SKA-31 (naphtho [1, 2-d] thiazole-2-ylamine) was synthesized as previously described [27]. SKA-31 and indomethacin were prepared as 10 mM stock solutions in DMSO and then diluted directly into the external bath solution. The final concentration of DMSO reaching the tissue was typically 0.05% (vol/vol) or less. In control experiments, bath application of 0.2% (v/v) DMSO had no effect on either developed myogenic tone or arterial responsiveness to acetylcholine and SKA-31 in Goto-Kakizaki cremaster arteries (n = 2) (data not shown).

2.4. Statistical analysis

Data are presented as mean ± SEM. In experiments involving arterial pressure myography (i.e. Figures 1–4), “N” denotes the number of animals or human subjects utilized for a given protocol (i.e. only a single artery was isolated from each preparation). For Fura-2 fluorescence imaging experiments, “n” denotes the number of individual cells analyzed from a total of “N” animals, as described in the figure legends. In essence, we followed the same convention utilized for analysis of single cell, patch clamp recordings, in which the total number of cells analyzed (i.e. “n”) is reported from “N” animals. Statistically significant differences between experimental conditions were evaluated using a one-way analysis of variance (ANOVA), followed by a Newman-Keuls post-hoc test. Note that a repeated measures ANOVA design was used to compare similar experimental conditions between rat strains or human subjects. A Mann-Whitney U test was utilized to analyze differences in normalized peak Fura-2 fluorescence between R1 and R2 designated cells (i.e. histograms in Figs 5 and 6, and Supplementary Figs 4 and 5). For other statistical comparisons, a two-tailed unpaired Student’s t-test or a Chi-squared test was utilized for two population comparisons and proportions, as noted. An observed difference was considered statistically significant at P < 0.05.

Figure 1 -.

Endothelium-dependent vasodilatory responses are impaired in cremaster skeletal muscle resistance arteries from T2D Goto-Kakizaki rats and restored in the presence of SKA-31. Panels A and B display representative tracings of acute vasoactive responses to endothelium-dependent and -independent agents in cannulated and pressurized cremaster arteries from Wistar (panel A) and age-matched, T2D Goto-Kakizaki rats (panel B). Following generation of myogenic tone at 70 mmHg, arteries were bath exposed to acetylcholine (ACh, 0.3 and 0.5 μM) and bradykinin (BK, 0.1 and 0.3 μM), and the endothelium-independent agent sodium nitroprusside (SNP, 10 μM) in the absence and presence of 0.3 μM SKA-31. Horizontal bars above and below the tracing indicate the treatment with each agent; W/O denotes washout of SKA-31. Maximal passive vessel diameter at the end of the protocol was determined in the presence of Krebs’ buffer containing 2 mM EGTA and no added CaCl₂. Panel C quantifies the percentage inhibition of myogenic tone in cremaster arteries by the indicated vasoactive agents ± SKA-31 treatment. Data are presented as mean ± SEM from 11 Wistar rats and 7 Goto-Kakizaki rats. The asterisk (*) indicates a statistical difference (P<0.05) vs. control response within a given strain; † indicates a difference between the Wistar and Goto-Kakizaki arteries for control responses under a defined experimental condition. Statistical analyses were performed using a repeated measures, one-way ANOVA, followed by a Newman-Keuls post-hoc test.

Figure 4 -.

The augmentation of endothelium-dependent vasodilation by SKA-31 treatment is not decreased by inhibitors of NO synthase (L-NAME, 100 μM) and COX (indomethacin, 10 μM) in intrathoracic resistance arteries from T2D patients. Experimental protocols were conducted as shown in Figure 3A and B, and the inhibition of myogenic tone by each vasoactive agent was quantified as described. Data are presented as mean ± SEM from four patients with T2D. An asterisk (*) indicates a statistically significant difference vs. the control response for each vasodilatory agent, as determined by a one-way ANOVA and a Newman-Keuls post-hoc test, P < 0.05. The § symbol denotes a significant difference compared with responses observed in the presence of L-NAME + indomethacin alone, P < 0.05.

3. RESULTS

3.1. Endothelium-dependent vasodilation is impaired in myogenically active resistance arteries from T2D cremaster skeletal muscle

Non-obese, T2D Goto-Kakizaki rats exhibit vascular endothelial dysfunction (i.e. impaired endothelium-dependent vasodilation) in both the macro- and microcirculation [23, 32–36]. To investigate this vascular impairment in greater depth, we utilized arterial pressure myography to examine both the endothelium-dependent and -independent vasodilatory responses evoked in resistance arteries (~150–175 μm maximal intraluminal diameter) isolated from cremaster skeletal muscle. Cannulated arteries from 24-week old T2D Goto-Kakizaki rats and euglycemic, age-matched, Wistar controls developed stable and comparable myogenic tone following elevation of intraluminal pressure to the physiological range (i.e. 70 mmHg). Functionally, the extent of pressure-induced myogenic vasoconstriction was similar in arteries from T2D Goto-Kakizaki (43.7 ± 8.8% of maximal passive diameter, n=13 animals) and Wistar rats (47.9 ± 13.4%, n=17 animals) (refer to Supplementary Table 2 for values of basal and maximal intraluminal diameters). Brief exposure of myogenically active cremaster arteries to the endothelium-dependent vasodilators acetylcholine (ACh, 0.3 and 0.5 μM) and bradykinin (BK, 0.1 and 0.3 μM) produced rapid and reversible increases in intraluminal diameter (i.e. inhibition of steady-state myogenic tone) in both Wistar (Fig. 1A) and T2D Goto-Kakizaki vessels (Fig. 1B). Critically, ACh and BK-evoked vasodilatory responses were significantly impaired in Goto-Kakizaki cremaster arteries compared with Wistar controls (Fig. 1C), consistent with the presence of endothelial dysfunction in T2D vessels. In contrast, arteries from both Goto-Kakizaki and Wistar rats exhibited similar relaxant responses to the smooth muscle-acting, nitrovasodilator sodium nitroprusside (SNP, 10 μM) (Fig. 1A–C) [29, 37, 38].

3.2. Acute treatment with SKA-31 restores endothelium-dependent vasodilation in T2D arteries

We and others have observed that endothelium-dependent vasodilation in arteries/tissues from T2D rats can be enhanced in the presence of a small molecule activator of small- and intermediate-conductance, Ca²⁺-activated K⁺ channels (KCa2.3 and KCa3.1, respectively) [23, 39]. However, the essential question of whether this effect occurs in myogenically active resistance arteries, as would be present under physiological conditions, has not been investigated. As shown in Figure 1, bath application of 0.3 μM SKA-31 (right-hand side of tracing in Fig. 1A) had little effect on basal myogenic tone in cannulated cremaster arteries from healthy Wistar rats (i.e. <5% inhibition of developed intraluminal constriction) (Supplementary Table 3). SKA-31 treatment also did not affect the magnitude of vasodilatory responses to ACh, with only minimal effects on BK responses (Fig. 1C). However, in T2D Goto-Kakizaki arteries, this same SKA-31 treatment significantly augmented vasodilatory responses to both ACh and BK, restoring the magnitude of these responses to the levels observed in Wistar arteries under control conditions (Figs. 1B and C). SKA-31 treatment produced <6% inhibition of basal myogenic tone in T2D arteries (see Supplementary Table 3). Supporting data indicate that the sensitivity of Wistar and T2D Goto-Kakizaki cremaster arteries to SKA-31 is similar (Supplementary Fig. 1) and that 3 μM SKA-31 has no vasoactive effect in rat cremaster and cerebral arteries following denudation of the vascular endothelium [29]. This latter observation demonstrates that SKA-31 does not directly act on contractile vascular smooth muscle. Functionally, SKA-31 treatment did not change vasorelaxant responses to the nitrovasodilator SNP in arteries from either Wistar or T2D Goto-Kakizaki rats. Following washout of SKA-31, we did not re-apply either ACh or BK as an additional control to verify that the observed enhancement of vasodilation by these agonists was strictly dependent upon the presence of SKA-31.

3.3. SKA-31 treatment improves endothelium-dependent vasodilatory responses in small intrathoracic arteries from T2D patients

Endothelial dysfunction is prominent in T2D subjects and contributes to the development and progression of cardiovascular complications/events leading to increased morbidity and mortality [8, 40, 41]. To examine the responsiveness of human arteries to SKA-31, we isolated small, intrathoracic resistance arteries (~200 μm maximal internal diameter) from consenting, non-diabetic and T2D patients undergoing coronary artery bypass graft surgery (see Supplementary Table 1 for patient characteristics). Overall, the clinical profiles of the two groups were similar. Functionally, cannulated human resistance arteries pressurized to 70 mmHg developed stable myogenic tone, the extent of which was not significantly different between the non-T2D arteries (38.1 ± 10.7% constriction of maximal passive diameter, n=10 subjects) and T2D arteries (42.9 ± 11.9% constriction, n=10 subjects) (refer to Supplementary Table 2 for average values of basal and maximal intraluminal diameters). In both non-diabetic and T2D human arteries, acute bath exposure to ACh (0.3 and 0.5 μM), BK (0.1 and 0.3 μM) and SNP (10 μM) evoked rapid and reversible vasodilatory responses (Fig. 2A and B). Functionally, T2D arteries exhibited blunted vasodilatory responses to both ACh and BK compared with vessels from non-diabetic subjects (Fig. 2C), consistent with the presence of endothelial dysfunction in T2D patients.

Figure 2 -.

Intrathoracic resistance arteries from Type 2 Diabetic (T2D) patients exhibit impaired endothelium-dependent vasodilation that is augmented in the presence of SKA-31. Panels A and B display representative recordings of acute vasoactive responses in pressurized, myogenically active intrathoracic resistance arteries isolated from non-diabetic (panel A) and T2D patients (panel B) that were treated with acetylcholine (ACh, 0.3 and 0.5 μM), bradykinin (BK, 0.1 and 0.3 μM), and sodium nitroprusside (SNP, 10 μM) in the absence and presence of 0.3 μM SKA-31. Horizontal bars above and below the tracing indicate the treatment with each agent; W/O denotes washout of SKA-31. Maximal passive arterial diameter at 70 mm Hg was determined at the end of the protocol in Krebs’ buffer containing 2 mM EGTA and no added CaCl₂. Panel C quantifies the percentage inhibition of myogenic tone in human intrathoracic arteries by the indicated vasoactive agents ± SKA-31 treatment. Data are presented as mean ± SEM from 7 non-diabetic patients and 6 T2D patients. The asterisk (*) indicates a statistical difference (P<0.05) vs. control response within a given patient group; † indicates that the control responses between the non-diabetic and T2D groups are statistically different for a given stimulus, P<0.05. Statistical analyses were performed using a repeated measures, one-way ANOVA and a Newman-Keuls post-hoc test.

Treatment of human arteries with a threshold concentration of SKA-31 (0.3 μM) (right-hand side of tracings in Fig. 2A and B) had little effect on basal myogenic tone, similar to observations in rat cremaster arteries (Fig. 1). In human T2D arteries, SKA-31 treatment robustly and significantly enhanced vasodilation in response to ACh and BK (Fig 2C), whereas this same treatment had a less pronounced effect in vessels from non-diabetic subjects. Functionally, SKA-31 treatment restored the magnitude of ACh and BK evoked vasodilation in human T2D arteries to the levels observed in non-diabetic vessels. Physiologically, these observations represent the first demonstration that myogenically active, human resistance arteries respond positively to an activator of endothelial KCa2.3 and KCa3.1 channels. Similar to cremaster arteries from Wistar and T2D Gto-Kakizaki rats, SKA-31 treatment did not augment vasodilatory responses to SNP in human arteries (Fig. 2C). Replotting the vasodilatory responses from T2D humans and Goto-Kakizaki rats (Supplementary Fig. 2) demonstrates that acute treatment with a threshold concentration of SKA-31 (i.e. one producing minimal direct vasodilation) augments agonist-evoked, endothelium-dependent vasodilation in small resistance arteries to a similar extent in both species.

3.4. Pharmacological inhibitors of eNOS and cyclo-oxygenase do not interfere with the SKA-31 induced augmentation of agonist-evoked vasodilation in rat and human T2D arteries

Key signaling mechanisms utilized by the endothelium to elicit relaxation of vascular smooth muscle in response to vasodilatory stimuli include the generation of nitric oxide (NO) and prostaglandins (e.g. prostacyclin), and the transfer of hyperpolarizing electrical current following activation of endothelial KCa channels [11, 12, 42]. To investigate whether increased NO and prostaglandin synthesis contribute to the cellular mechanism by which SKA-31 treatment augments agonist-evoked vasodilation in rat and human T2D resistance arteries, we examined SKA-31 mediated enhancement in the presence of pharmacological inhibitors of NO synthase (i.e. L-NAME) and cyclo-oxygenase (i.e. indomethacin).

In myogenically active Wistar cremaster arteries, treatment with L-NAME (100 μM) and indomethacin (10 μM) produced a modest vasoconstriction (i.e. decrease in intraluminal diameter, 8.9 ± 4.5 μm, n=6 animals), and significantly blunted the vasodilatory responses to ACh and BK, compared with control responses obtained in the absence of these inhibitors (Fig. 3A, C). These observations are consistent with the reported contribution of endothelium-derived NO and/or prostacyclin to the regulation of vascular tone under basal and stimulated conditions [4, 12, 43]. L-NAME + indomethacin treatment did not alter responses to the direct smooth muscle vasorelaxant SNP. Further addition of a threshold concentration of SKA-31 (0.3 μM) in the continued presence of L-NAME and indomethacin had little effect on baseline tone in Wistar arteries, and did not significantly alter vasodilatory responses to ACh, BK or SNP (Fig. 3A and 3C). These latter observations indicate that treatment of control vessels with a threshold concentration of SKA-31 was unable to improve endothelium-dependent, agonist-evoked vasodilation following pharmacological inhibition of NOS and cyclo-oxygenase (COX).

Figure 3 -.

SKA-31 mediated augmentation of endothelium-dependent vasodilatory responses in T2D Goto-Kakizaki cremaster arteries is not altered by inhibition of vascular NO synthase and COX activities. Myogenically active arteries from Wistar (panel A) and Goto-Kakizaki rats (panel B) were stimulated with ACh (0.3 μM), BK (0.1 μM) and SNP (10 μM) under control conditions, in the presence of L-NAME (100 μM) + indomethacin (Indo, 10 μM) and then following treatment with L-NAME + indomethacin + 0.3 μM SKA-31. Horizontal bars above and below each tracing indicate the treatment with each stimulus and agent; W/O denotes washout of L-NAME + indomethacin + SKA-31. Maximal passive intraluminal diameter at 70 mm Hg was determined in Krebs’ buffer containing 2 mM EGTA and no added CaCl₂. The histograms in panels C and D quantify vasoactive responses to ACh, BK and SNP in cremaster arteries from Wistar and Goto-Kakizaki rats, respectively, in the absence (i.e. control conditions) and the presence of 100 μM L-NAME + 10 μM indomethacin and L-NAME + indomethacin + 0.3 μM SKA-31. Data are presented as mean ± SEM from 6–7 animals in each group. An asterisk (*) indicates a statistically significant difference vs. the control response for a given vasodilatory agent, as determined by a one-way ANOVA, followed by a Newman-Keuls post-hoc test, P < 0.05. The † symbol denotes a significant difference compared with responses observed in the presence of L-NAME + indomethacin, P < 0.05.

In contrast to results from healthy animals, treatment of cremaster arteries from T2D Goto-Kakizaki rats with L-NAME and indomethacin did not constrict intraluminal diameter or decrease the magnitude of vasodilatory responses to either ACh or BK (Fig. 3B and D). These findings are thus consistent with reports of reduced NO bioavailability and endothelial dysfunction in T2D arteries [23, 42]. Remarkably, treatment of T2D arteries with SKA-31 led to a significant improvement of the ACh and BK-evoked dilatory responses compared with those observed in the presence of L-NAME + indomethacin alone. SKA-31 treatment, however, did not alter vasodilatory responses to SNP in the continued presence of L-NAME + indomethacin. Our observations that SKA-31 treatment produced similar enhancement of endothelium-dependent dilation in the absence and presence of acute eNOS and COX inhibition strongly argue that de novo synthesis of NO and/or prostacyclin do not contribute to the rescue of agonist-evoked dilation in T2D vessels (see Supplementary Fig. 3 for a summary of L-NAME, indomethacin and SKA-31 effects in control and T2D arteries). Importantly, parallel experiments carried out using intra-thoracic resistance arteries from T2D patients revealed similar results (Fig. 4) as those observed in T2D Goto-Kakizaki cremaster arteries, indicating a lack of involvement of NO and prostacyclin in SKA-31 mediated augmentation of agonist-evoked vasodilation in human T2D arteries. Collectively, these data suggest that SKA-31 treatment augments endothelium-dependent, agonist-evoked dilation in a similar manner in myogenically active resistance arteries from T2D patients and Goto-Kakizaki rats, likely via a conserved cellular mechanism.

3.5. Acute SKA-31 treatment improves agonist-evoked Ca²⁺ mobilization in isolated endothelial cells from T2D rats

Endothelial cell Ca²⁺ mobilization and signaling play critical roles in the activation of KCa2.3 and KCa3.1 channels in response to vasodilatory agonists [13, 44, 45]. To examine if the enhancement of agonist-evoked vasodilation by SKA-31 in T2D resistance arteries involved altered intracellular Ca²⁺ release and/or external Ca²⁺ entry, acutely dissociated cremaster artery endothelial cells were loaded with the fluorescent Ca²⁺ indicator Fura-2 and stimulated with either the muscarinic receptor agonist carbachol or the sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) inhibitor cyclopiazonic acid (CPA). Carbachol was utilized in place of ACh as it was more chemically stable under our experimental conditions and produced robust inhibition of myogenic tone in Wistar rat cremaster arteries, although the IC₅₀ was right-shifted approximately 7-fold compared with ACh (Supplementary Fig. 4). Stimulation of dissociated endothelial cells with carbachol (1 μM) in the absence of external Ca²⁺ elicited a rapid and transient increase in cytosolic free Ca²⁺ due to internal store release, followed by secondary rise upon re-introduction of Ca²⁺ to the bath solution (Fig. 5A and C). Experimentally, this well-established protocol allowed us to differentiate precisely stimulus-evoked elevation of cytosolic Ca²⁺ due to intracellular release vs. external Ca²⁺ entry. Unexpectedly, two types of Ca²⁺ mobilizing responses were observed in both control and T2D endothelial cells exposed to carbachol; in one group (R1), the stimulated intracellular Ca²⁺ release was large (e.g. peak F/F₀ > 1.5), and in the other group (R2), this release event was noticeably smaller (i.e. peak F/F₀ < 1.5). In both Wistar and T2D Goto-Kakizaki endothelial cells, all other measurements of cytosolic Ca²⁺, including baseline Ca²⁺ levels and external Ca²⁺ entry, were of similar magnitude over the course of the experimental protocol (see representative tracings and histograms in Figs. 5A and C). In Wistar endothelial cells, the proportion of cells displaying robust (R1) vs. weak (R2) intracellular Ca²⁺ release was 67%, whereas a robust R1 response was observed in only 19% of endothelial cells from T2D Goto-Kakizaki cremaster arteries. A similar pattern of carbachol-stimulated intracellular Ca²⁺ release was observed in endothelial cells acutely isolated from Wistar and T2D Goto-Kakizaki cerebral arteries (70% vs. 17%, respectively; Supplementary Fig. 5), indicating that this phenomenon was not unique to skeletal muscle arteries. These observations demonstrate that receptor-evoked intracellular Ca²⁺ release was functionally blunted in the majority of endothelial cells obtained from T2D Goto-Kakizaki cremaster and cerebral arteries. Remarkably, brief (2–3 min) treatment of T2D Goto-Kakizaki endothelial cells with 0.3 μM SKA-31 (i.e. a threshold concentration producing minimal vasodilation) prior to carbachol stimulation significantly increased the proportion of cremaster endothelial cells displaying robust intracellular Ca²⁺ release from 19% to 49% (P = 0.009, Fig. 5C and 5D). In cerebral artery endothelial cells, SKA-31 treatment increased this proportion from 17% to 68% (P < 0.001, Supplementary Fig. 5C and 5D). In contrast, SKA-31 pre-treatment produced a non-significant increase in the proportion of cells exhibiting robust intracellular Ca²⁺ release in response to carbachol in Wistar cremaster endothelial cells (i.e. 67% to 85%, p = 0.22) (Fig. 5A and 5B). A similar pattern was observed in carbachol stimulated endothelial cells from Wistar cerebral arteries in the absence and presence of SKA-31 pre-treatment (i.e. 70% vs 74%, respectively; P = 0.73) (Supplementary Fig. 5A and 5B). Quantitatively, treatment of cremaster endothelial cells with 0.3 μM SKA-31 had no effect on the magnitude of carbachol-evoked intracellular Ca²⁺ release (i.e. mean F/F₀ values) compared with control responses in cells from either Wistar or T2D Goto-Kakizaki arteries (Fig. 5A vs. 5B; Fig. 5C vs. 5D). Similar results were observed in cerebral artery endothelial cells from both groups (Supplementary Fig. 5A vs. 5B and 5C vs. 5 D). SKA-31 treatment also did not modify the basal level of cytosolic free Ca²⁺ in unstimulated cremaster or cerebral artery endothelial cells from Wistar and T2D Goto-Kakizaki rats. These data thus demonstrate that a large proportion of endothelial cells from T2D cremaster and cerebral arteries exhibit blunted intracellular Ca²⁺ mobilization in response to carbachol and that brief exposure to the KCa channel activator SKA-31 can restore normal intracellular Ca²⁺ release in individual T2D Goto-Kakizaki endothelial cells.

To investigate whether the blunted intracellular Ca²⁺ mobilization observed in carbachol-stimulated T2D Goto-Kakizaki endothelial cells may be due to reduced Ca²⁺ content in the ER store, we utilized cyclopiazonic acid (CPA, 20 μM) to elicit direct release of stored Ca²⁺ from the ER. We followed the same experimental protocol as used for carbachol stimulation in order to avoid procedural differences that may confound comparison and interpretation of the data. As shown in Figure 6A, 80% of endothelial cells from Wistar cremaster arteries exhibited robust, R1-type intracellular Ca²⁺ release (i.e. peak F/F₀ > 1.5) in response to acute CPA application, whereas 72% of endothelial cells from T2D Goto-Kakizaki arteries exhibited robust intracellular Ca²⁺ release following CPA treatment (Fig. 6C) (difference not significant, P > 0.4). SKA-31 treatment modestly increased the proportion of CPA-stimulated cells exhibiting robust intracellular Ca²⁺ release from 80% to 89% in Wistar endothelial cells, and from 72% to 83% in T2D Goto-Kakizaki arteries (Fig. 6B and 6D); however, this observed increase within each group was not significant (P > 0.3). In cerebral artery endothelial cells from Wistar and T2D Goto-Kakizaki rats stimulated with CPA, a similar pattern of results was observed in the absence and presence of 0.3 μM SKA-31 (Supplementary Fig. 6). Collectively, these observations indicate that the release of intracellular stored Ca²⁺ in response to CPA is comparable in isolated endothelial cells from Wistar and T2D Goto-Kakizaki resistance arteries, and is not modified in the presence of a threshold concentration of SKA-31.

4. DISCUSSION

Endothelial dysfunction is a prominent vascular pathology contributing to the development and progression of cardiovascular complications in T2D (i.e. hypertension, atherosclerosis, coronary artery disease) [8, 9, 42]. Clinically, mitigation of T2D-associated endothelial dysfunction is expected to improve cardiovascular performance and reduce morbidity and mortality [41, 46]. In the present study, we provide novel and direct evidence that pharmacological targeting of endothelial KCa2.3 and KCa3.1 activities can significantly augment agonist-evoked, endothelium-dependent vasodilation in myogenically active, small resistance arteries from T2D rats and human subjects, in part by improving stimulated Ca²⁺ mobilization in individual endothelial cells.

Physiologically, endothelial KCa channel activation and signaling represents a powerful vasodilatory mechanism in the vasculature, as direct pharmacological stimulation of these channels with small molecule activators (e.g. SKA-31, NS309) can evoke rapid and robust vasodilation [24, 27, 29, 47] and acutely lower systemic blood pressure in vivo [24, 27, 48–50]. We thus reasoned that treatment of T2D resistance arteries with a threshold concentration of SKA-31 (i.e. one producing minimal direct vasodilation) may enhance endothelial function by pharmacologically sensitizing or “priming” endothelial KCa channels (i.e. KCa2.3 and KCa3.1 channels), thereby facilitating their activation in response to Ca²⁺-mobilizing agonists (e.g. ACh and BK) that evoke endothelium-dependent hyperpolarization (EDH) [11, 13, 51]. We further predicted that this strategy would normalize endothelium-dependent vasodilation in T2D vessels.

As shown in Figures 1 and 2, acute treatment of myogenically active resistance arteries from T2D Goto-Kakizaki rats and T2D human subjects with 0.3 μM SKA-31 (i.e. a concentration near the reported EC₅₀ for KCa3.1 channel activation and far from saturating) had little effect on baseline intraluminal diameter and myogenic tone (Supplementary Table 3), but significantly enhanced the magnitude of agonist-evoked, endothelium-dependent vasodilation. This demonstration represents the first report of this kind in myogenically constricted, T2D resistance arteries, although similar effects of a KCa channel activator have been described previously in pharmacologically pre-constricted rodent arteries [21, 22, 39]. SKA-31 treatment had minimal functional effect on arteries from non-diabetic Wistar rats lacking endothelial dysfunction, although in non-diabetic human arteries, SKA-31 did enhance vasodilatory responses to low concentrations of ACh and BK (Fig. 2C). As these latter vessels were obtained from patients with a mean age of 71 years (Supplementary Table 1) and aging is associated with endothelial dysfunction [52, 53], it is perhaps not surprising that SKA-31 treatment had a modest effect in these arteries. In contrast, SKA-31 did not alter vasodilatory responses to the smooth muscle acting, nitrovasodilator sodium nitroprusside (SNP), which is consistent with the lack of expression of SKA-31-sensitive KCa channels in contractile vascular smooth muscle [54, 55] and the insensitivity of endothelium-denuded, rat resistance arteries to SKA-31 [29]. Critically, these observations demonstrate that SKA-31 is able to normalize agonist-evoked vasodilation in both human and rodent T2D resistance arteries, possibly via a conserved cellular mechanism(s). Remarkably, this enhancement remained robust in human resistance arteries, despite years of T2D-associated comorbidities in these subjects (average duration of clinically established T2D = 17 ± 6 years, Supplementary Table 1). It is further noteworthy that the vasodilatory effects of SKA-31 and related KCa activators have been reported in multiple arterial beds and species, despite expected differences in endothelial cell phenotype and function in vessels from different regions of the vasculature [22–24, 26, 29, 39, 49, 50]. Although not apparent in our study, increases or decreases in endothelial KCa channel expression/activity have been reported in other rodent models of T2D [56–58], which may relate to differences in animal phenotype (e.g. lean vs. obese), age, genetic alterations, vascular bed, etc. Factors affecting endothelial KCa channel expression may influence the ability of SKA-31 and other KCa channel activators to modify evoked vasodilation in T2D resistance arteries.

Functionally, the enhancement of endothelium-dependent vasodilation by SKA-31 in cremaster arteries from both T2D Goto-Kakizaki rats and intra-thoracic resistance arteries from T2D subjects was unaltered by pharmacological inhibition of vascular eNOS and COX activities (Figs. 3 and 4), indicating that NO bioavailability and/or prostacyclin signaling are not mechanistically required for this functional improvement. In contrast, inhibition of eNOS and COX activities significantly reduced agonist-stimulated vasodilation in control cremaster arteries (Fig. 3C). These observations are thus in line with the reported vasoactive effects of NO and prostaglandins in healthy resistance arteries and reduction of NO bioavailability in T2D arteries [4, 42]. Importantly, these findings strongly suggest that endothelial KCa channel activity alone may be sufficient to support the restoration of endothelial function by SKA-31 in T2D arteries, which is consistent with the recognized functional importance of EDH as a vasodilatory mechanism in small resistance arteries [51, 59, 60]. In contrast, Hamilton et al. [61] have reported that NO synthase inhibition does decrease carbachol-evoked relaxation in larger intrathoracic arteries from humans, in line with a more prominent vasoactive role for NO in peripheral conduit arteries [42, 59].

Fura-2 based Ca²⁺ imaging in acutely isolated, single endothelial cells revealed that muscarinic receptor-evoked release of intracellular Ca²⁺ stores was significantly reduced in the majority of endothelial cells from T2D Goto-Kakizaki cremaster and cerebral arteries compared with cells from Wistar control vessels (Fig. 5 and Supplementary Fig. 5). Collectively, these observations highlight a major, previously unrecognized alteration in GPCR-mediated vasodilatory signaling in endothelial cells from T2D resistance arteries. In contrast, release of stored ER Ca²⁺ by the SERCA pump inhibitor cyclopiazonic acid (CPA, 20 μM) (Fig. 6 and Supplementary Fig. 6) exhibited a similar pattern in cremaster and cerebral artery endothelial cells from T2D and Wistar rats. These latter observations suggest that the total pool of releasable ER stored Ca²⁺ was comparable in cells from each group. Remarkably, acute treatment of T2D cremaster and cerebral endothelial cells with SKA-31 dramatically improved carbachol-evoked cytosolic Ca²⁺ elevation by significantly increasing the proportion of cells displaying robust ER Ca²⁺ release (i.e. R1 response), whereas only a minimal change was observed in endothelial cells from healthy arteries. In contrast, SKA-31 treatment had no effect on the responsiveness of endothelial cells from either Wistar or T2D rats to CPA, suggesting that SKA-31 treatment does not directly affect the total pool of releasable ER Ca²⁺. Physiologically, restoration of robust, agonist-evoked Ca²⁺ mobilization in the endothelium would be predicted to increase EDH-type signaling and vasodilation, in line with our functional data (Figs 1 and 2). This restoration of “normal” Ca²⁺ mobilization may also be important for conducted vasodilation within the vascular network via KCa channel-mediated hyperpolarization and the spread of endothelial Ca²⁺ via gap junctions through the endothelial layer [62–64]. Collectively, these observations reveal a previously unappreciated molecular action of endothelial KCa channel activity that likely contributes to the observed functional enhancement of agonist-evoked vasodilation in T2D arteries (see Figures 1 and 2).

Exactly how acute SKA-31 treatment increases the proportion of T2D endothelial cells exhibiting a robust Ca²⁺-mobilizing response remains unclear. However, the literature provides several clues. Stored Ca²⁺ within the ER may exist in stimulus-sensitive compartments or microdomains that can be preferentially released via IP3- or ryanodine-sensitive Ca²⁺ release channels [65–67], suggesting that bulk ER Ca²⁺ may not be uniformly distributed or available for release within the ER organelle. Although data have not been reported for endothelial cells, it remains conceivable that the size of the agonist-sensitive, IP3-releasable ER Ca²⁺ pool may be smaller in T2D endothelial cells compared with controls and could explain why the majority of T2D endothelial cells exhibit weak, R2-type ER Ca²⁺ release in response to carbachol. If SKA-31 treatment were able to promote a redistribution of bulk ER Ca²⁺ to increase the size of the agonist releasable pool, then a greater proportion of T2D cells could exhibit a robust R1-type ER Ca²⁺ release in response to carbachol. Whether KCa channel activators are able to promote a redistribution of stored ER Ca²⁺ to a particular compartment is currently unknown. FRET-based structure-function studies in cell expression systems that have further demonstrated that membrane hyperpolarization can increase agonist-binding affinity for M3 and M5 muscarinic receptors [68], the predominant Gq-linked receptor sub-types coupled to IP3 generation and intracellular Ca²⁺ release in peripheral [69] and cerebrovascular endothelium [70], respectively. These, and related findings [71], raise the speculative, but intriguing possibility that enhancement of agonist-evoked endothelial cell hyperpolarization by a KCa channel activator (e.g. SKA-31) could promote a more robust Ca²⁺ mobilizing response by increasing the efficacy of muscarinic receptor activation in individual T2D endothelial cells, or possibly the recruitment of silent receptors (see Supplementary Figure 9). Hyperpolarization may also augment muscarinic receptor-activated, TRPV4-mediated Ca²⁺ influx in endothelium, which contributes to KCa channel activation and evoked vasodilation [72, 73]. However, our Fura-2 measurements of global cytosolic Ca²⁺ were unable to detect such localized activity in single endothelial cells following re-introduction of external Ca²⁺, given the small magnitude of TRPV4-mediated Ca²⁺ influx events (i.e. Ca²⁺ sparklets). Increasing the spatial and temporal resolution of endothelial Ca²⁺ imaging in our cells would help address this question. At the cellular level, these speculative actions of SKA-31 at the plasma membrane would not be expected to influence ER Ca²⁺ release by the SERCA inhibitor CPA, in agreement with our observations (Fig. 6 and Suppl. Fig. 6). Finally, McCarron and colleagues have demonstrated that endothelial cells within an intact monolayer exhibit non-uniform sensitivity to Ca²⁺-mobilizing agonists, such as carbachol and ATP, that can range over a 1000-fold of agonist concentration [74–76]. The higher proportion of T2D endothelial cells exhibiting a modest, R2-type Ca²⁺-mobilizing response may thus reflect a difference in muscarinic GPCR activity and/or the size of the IP3-releasable ER Ca²⁺ pool. We speculate that these processes may be acutely affected by a KCa channel activator. Resolving such issues experimentally will require simultaneous measurements of endothelial membrane potential, cytosolic Ca²⁺ and intraluminal diameter in myogenically contracted resistance arteries; the absence of such data is a limitation of the current study.

From a pre-clinical perspective, it is unclear from our results whether the extent and/or severity of T2D-associated conditions (i.e. hyperglycemia, obesity, metabolic syndrome, insulin resistance, hypertension, etc.) may alter the beneficial effects of a KCa channel activator on endothelial dysfunction and impaired vasodilation. However, our results demonstrating that treatment with a KCa channel activator effectively normalized endothelium-dependent vasodilation in myogenically active resistance arteries from human subjects with chronic T2D (i.e. 17 ± 6 years) strongly suggest that the cellular mechanism(s) underlying this functional effect remains intact in the face of prolonged vascular pathology (e.g. hyperglycemia, insulin resistance, reduced NO bioavailability, increased oxidative stress, etc.). Taken together, our data directly support the targeting of endothelial KCa channel activity and signaling as a novel therapeutic strategy in the treatment of T2D-associated vascular dysfunction and cardiovascular disease [13, 25, 42, 51].

5. Conclusions

In summary, our results demonstrate that the KCa channel activator SKA-31 is able to normalize agonist-evoked, endothelium-dependent vasodilation in myogenically active resistance arteries from T2D Goto-Kakizaki rats and T2D subjects independent of vascular NO synthase and COX enzyme activities. Mechanistically, we suggest that the observed enhancement of agonist-evoked, intracellular Ca²⁺ mobilization in T2D endothelial cells by SKA-31 contributes to improved endothelial function and vasodilatory signaling in T2D arteries (Supplementary Figure 9). This action may be due, in part, to SKA-31 mediated “priming” or sensitization of endothelial KCa channel activity and the subsequent enhancement of EDH-mediated dilation [15, 24]. Taken together, our study outlines a novel cellular mechanism linking KCa channel activity to augmented endothelium-dependent vasodilation in myogenically active, T2D resistance arteries.

Article Highlights:

• Endothelial dysfunction contributes to cardiovascular disease in Type 2 Diabetes

• Pharmacological “priming” of KCa channels restores vasodilation in T2D arteries

• Functional restoration does not require nitric oxide or prostaglandin synthesis

• KCa channel targeting improves agonist-evoked Ca²⁺ release in T2D endothelial cells

• Endothelial Ca²⁺ dynamics may contribute to restored vasodilation in T2D arteries

Abbreviations:

ACh

acetylcholine

BK

bradykinin

CABG

coronary artery bypass graft

COX

cyclooxygenase

CPA

cyclopiazonic acid

KCa channel

calcium-activated K⁺ channel

NO

nitric oxide

PSS

physiological saline solution

SERCA

sarco/endoplasmic reticulum calcium ATPase

SNP

sodium nitroprusside

T2D

type 2 diabetes