ATP-sensitive K⁺ Channel Inhibition in Rats Decreases Kidney and Skeletal Muscle Blood Flow Without Increasing Sympathetic Nerve Discharge

Abstract

ATP-sensitive K⁺ (K_ATP) channels contribute to exercise-induced hyperemia in skeletal muscle either locally by vascular hyperpolarization or by sympathoinhibition and decreased sympathetic vasoconstriction. However, mean arterial pressure (MAP) regulation via baroreceptors and subsequent efferent activity may confound assessment of vascular versus neural K_ATP channel function. We hypothesized that systemic K_ATP channel inhibition via glibenclamide (GLI) would increase MAP without increasing sympathetic nerve discharge (SND). Lumbar and renal nerve SND were measured in anesthetized male rats with intact baroreceptors (n=12) and sinoaortic denervated (SAD; n=4) counterparts and blood flow (BF) and vascular conductance (VC) assessed in conscious rats (n=6). GLI increased MAP (p<0.05) and transiently decreased HR in intact (p<0.05), but not SAD rats. Renal (−30%) and lumbar (−40%) ΔSND decreased in intact but increased in SAD rats (~40% and 20%; p<0.05). BF and VC decreased in kidneys and total hindlimb skeletal muscle (p<0.05). Thus, because K_ATP inhibition decreases SND, GLI-induced reductions in blood flow cannot result from enhanced sympathetic activity.

Introduction

At rest and during exercise the regulation of systemic mean arterial pressure (MAP = Cardiac Output x SVR) is a competition between the sympathetic nervous system maintaining MAP (↑cardiac output and renal and splanchnic vasoconstriction) and contracting skeletal muscle decreasing systemic vascular resistance (↓SVR) to ensure adequate blood flow and O₂ delivery. Peripheral mechanisms for increasing O₂ delivery (i.e. originating within skeletal muscle) include shear stress-mediated vasodilation (e.g. ↑nitric oxide and prostaglandin production), myogenic tone, blunted sympathetic-mediated adrenergic vasoconstriction (i.e. sympatholysis), and regulation of vascular smooth muscle membrane potential (rev. Laughlin et al., 2012; Joyner and Casey, 2015). Importantly, inadequate O₂ delivery leads to elevated production of fatigue-related metabolites (i.e. ADP, Pi, H⁺, etc.), decreased muscle contractile performance and exhaustion (Hogan et al, 1992; Richardson et al., 1998; Wilson et al., 1977).

ATP-sensitive potassium (K_ATP) channels may constitute a key mechanism in sensing inadequate O₂ supply relative to demand within skeletal muscle (i.e., ↓ATP:ADP ratio) leading to vasodilation. K_ATP channels have been shown to contribute significantly to reactive and functional hyperemia in human and animal models (Banitt et al., 1996; Bjilstra et al., 1996; Holdsworth et al., 2015; Saito et al., 1996) due, in part, to blunting sympathetic vasoconstriction (Keller et al., 2004; Thomas et al., 1997). This increased blood flow via activation of K_ATP channels following the onset of contractions improves O₂ delivery-utilization matching at the microvascular-myocyte interface (Holdsworth et al., 2016), thereby improving exercise tolerance. Local K_ATP channel inhibition via glibenclamide (GLI) superfusion demonstrates that proper K_ATP channel function within skeletal muscle (peripheral regulation of vascular tone) is obligatory for providing adequate O₂ delivery (Holdsworth et al., 2017). Importantly, oral sulphonylureas which aim to inhibit pancreatic K_ATP channels are the most widely prescribed second-line anti-diabetic drug (~50% of new prescriptions; Montvida et al., 2018) and their systemic effects may therefore contribute to the exercise intolerance symptomatic of patient populations.

However, K_ATP channels are also active in neural tissue (i.e. rostral ventrolateral medulla and carotid sinus baroreceptors) and produce sympathoinhibition when directly activated via microinjection or isolated perfusion techniques (Guo et al., 2011; Guo et al., 2016; Xiao et al., 2006). It can thus be argued that systemic K_ATP channel inhibition during exercise may increase vascular tone via increased sympathetic nerve discharge (SND; centrally-mediated regulation of vascular tone) resulting in decreased blood flow and increased MAP. As such, GLI augments baroreflex-mediated vasoconstriction at rest and during muscle contractions (Guo et al., 2016; Keller et al., 2004; Xiao et al., 2006). Nevertheless, during systemic delivery of GLI (Keller et al., 2004) muscle sympathetic activity has not been recorded. Therefore, to determine the mechanistic bases for the effects of GLI on blood pressure regulation and muscle hyperemia, it is imperative to separate the central from peripheral contributions to vascular tone in the kidneys (renal SND) and hindlimb muscles (lumbar SND) following systemic K_ATP channel inhibition.

In the current investigation changes in renal and lumbar SND in anesthetized rats, and those in kidney and skeletal muscle blood flow (BF) and vascular conductance (VC), within resting conscious rats, were assessed following GLI-induced K_ATP channel inhibition. Based on GLI decreasing O₂ delivery following local (peripheral) application (Holdsworth et al., 2017), we hypothesized that increased MAP following GLI-induced systemic K_ATP channel inhibition would occur simultaneously with decreased BF and VC to kidneys and hindlimb skeletal muscle but that GLI would not incur increased lumbar or renal SND. SND recordings were also performed in rats following sinoaortic denervation (SAD) because baroreflex-mediated regulation of MAP may limit the effects of K_ATP channel inhibition on SND (i.e. removal of sympathoinhibition and subsequent ↑SND). SAD rats were thus hypothesized to have greater MAP elevation compared to intact rats due to more robust increases in lumbar and renal SND, highlighting the functional balance between K_ATP channel-mediated sympathoinhibition and baroreceptor function at rest.

Methods

Sixteen young adult male Sprague-Dawley rats with (intact; n=12, 402 ± 24 g) and without (SAD; n=4, 405 ± 19 g) intact sinoaortic baroreceptors were utilized during the assessment of sympathetic activity (Charles River Laboratories, Wilmington, MA). For the assessment of GLI’s impact on resting vascular control within kidneys and skeletal muscle, all rats examined at rest in Holdsworth et al. (2015; n=6, 365 ± 7 g) were statistically re-analyzed separate from exercise. Rats were housed in cages in accredited facilities on a 12:12 h light:dark cycle with standard rat chow and water provided ad libitum. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Kansas State University and conducted according to National Institutes of Health guidelines.

Drugs

The pharmacological sulphonylurea derivative GLI (494 g mol⁻¹; 5-chloro-N-(4-[N-(cyclohexylcarbamoyl)-sulfamoyl]phenethyl)-2-methoxybenzamide; Sigma-Aldrich, St Louis, MO, USA) was utilized to inhibit K_ATP channels. To obtain a 2.5 mg ml⁻¹ stock solution, 25 mg of GLI was dissolved in 0.8 ml of NaOH (0.1 M) and 0.2 ml dimethyl sulfoxide (DMSO). The desired 5 mg kg⁻¹ dose of GLI was drawn from the stock solution and diluted to ~1 ml with heparinized saline and administered as a bolus infusion via the femoral artery catheter (during SND recordings) or the carotid catheter (tissue BF and VC; Holdsworth et al. 2015). GLI has been reported previously to be a selective K_ATP channel blocker at concentrations below 5 μmol L⁻¹ (Beech et al., 1993 a,b; Sadraei and Beech, 1995). With the 5 mg kg⁻¹ dose used herein in rats averaging ~366-405 g, the blood concentration of GLI equates to 153 μmol L⁻¹. Since 98-99% of GLI binds to plasma protein, the effective concentration of GLI in the blood is ~2-3 μmol L⁻¹ (George et al., 1990) and in the range for selective K_ATP channel inhibition without inhibiting Ca²⁺ channel currents (Sadraei and Beech, 1995).

Study 1: Effect of GLI on Sympathetic Nerve Discharge

Surgical Instrumentation

Anesthesia was introduced with an isoflurane-O₂ mixture (3-5%) and maintained using isoflurane-O₂ (1.25-1.75%), α-chloralose (80 mg kg⁻¹ i.p.), and urethane (800 mg kg⁻¹ i.p.) during all surgical procedures. Catheters (PE-50) were placed in the femoral vein and femoral artery. Throughout experiments maintenance doses of α-chloralose (35-45 mg g⁻¹ hr⁻¹) were administered intravenously, whereas urethane (200 mg kg⁻¹ every 4 h) was administered intraperitoneally. Adequacy of anesthesia was assessed by the absence of increased SND or MAP following mechanical stimulation of the hindlimb or tail. The trachea was cannulated for artificial ventilation with the frequency and depth of breathing set to regulate end-tidal CO₂ partial pressure between 35 and 40 mmHg. Femoral arterial pressure was monitored via a pressure transducer connected to a blood pressure analyzer. Pulsatile arterial pressure was also measured by the blood pressure analyzer and subsequently used to calculate HR. A temperature-controlled table was used to maintain core temperature at 37-38°C, measured by a rectal thermistor probe.

Since baroreceptor-mediated afferent feedback can alter central SND responses, baroreceptor (sino-aortic)-denervated (SAD) rats were utilized in separate experiments. SADs were completed 3-4 hours prior to the initiation of experimental protocols (see Experimental Protocol section). Bilateral denervation of the aortic arch was completed by cutting the superior laryngeal nerve near its junction with the vagus nerve and removal of the superior cervical ganglion. Bilateral carotid sinus denervation was completed by removal of the adventitia from the area of the carotid sinus bifurcation. Efficacy of the denervation procedure has been demonstrated previously (Kenney, 1994) by using the coherence function to relate arterial pressure to SND. Coherence analysis provides a measure of the strength of linear correlation between two signals as a function of frequency. A lack of coherence between arterial pressure and SND at the frequency of HR demonstrated a complete SAD (Kenney, 1994).

For SND measurements, the left renal nerve was isolated retroperitoneally and the left lumbar nerve was isolated from a midline approach. Platinum bipolar electrodes were attached to the central end of cut or distally crushed lumbar or renal nerves and SND was measured and recorded biphasically after capacity-coupled preamplification (30-300 Hz). Nerve-electrode preparations were covered with silicone gel to prevent exposure to room air. Filtered neurograms were routed to an oscilloscope and a nerve traffic analyzer where sympathetic nerve potentials were full-wave rectified and integrated (10-ms time constant). The total power in lumbar and renal SND was quantified as microvolts x seconds (μVs) and SND recordings were corrected for background noise after administration of the ganglionic blocker chlorisondamine (5 mg kg⁻¹ i.v.) or distal nerve crush.

Experimental Protocol 1

Prior to the baseline measurements, anesthetized rats were allowed 60 min to stabilize. MAP, HR, and lumbar and renal SND were measured and recorded for ~10-15 min with the final 60 s averaged for baseline values (time point “zero” in Figures 1–4). For controls, saline (0.5 ml) was infused into the femoral vein and MAP, HR, lumbar SND, and renal SND were measured and recorded continuously for 10 min. Following another ~15 min stabilization period, baseline measurements were recorded as described above. GLI (K_ATP channel inhibitor; 5 mg kg⁻¹) was then infused and MAP, HR, lumbar SND, and renal SND were measured and recorded for 10 min. For each rat the entire experiment (surgery, stabilization, and experimental protocol) lasted ~6-8 h. Euthanasia was then administered via methohexital sodium overdose (150 mg kg⁻¹ i.v.).

Figure 1.

The effect of K_ATP channel inhibition (GLI; glibenclamide) on changes in mean arterial pressure (MAP). MAP increases across time in sinoaortic baroreceptor denervated rats (SAD (n=4); open bars) and baroreceptor intact rats (intact (n=12); solid bars). Data are presented as mean ± SEM. * p < 0.05 vs. baseline; # p < 0.05 vs. intact

Figure 4.

Changes in Lumbar Sympathetic Activity following K_ATP channel inhibition. The change in lumbar sympathetic nerve discharge (SND) is directionally different between sinoaortic baroreceptor intact rats (intact (n=8); solid bars) and baroreceptor denervated rats (SAD (n=3); open bars). Data are presented as mean ± SEM. * p < 0.05 vs. baseline; # p < 0.05 vs. intact

Study 2: Effect of GLI on Resting Blood Flow and Vascular Conductance

Surgical Instrumentation

In a separate group of rats, the role of K_ATP channels for supporting resting BF and VC was measured in conscious rats. Briefly, rats were initially anesthetized on 5% isoflurane-O₂ mixture and maintained on 3% isoflurane-O₂ mixture for the instrumentation of carotid (microsphere infusions and MAP and HR recordings) and caudal (GLI infusion and reference sample for blood flow) artery catheters using PE-10 connected to PE-50 (Intra-Medic polyethylene tubing, BD, Franklin Lakes, NJ). Catheters were subsequently tunneled subcutaneously to the dorsal aspect of the cervical region and exteriorized through a puncture wound in the skin. Incisions were then closed and rats removed from anesthesia for a recovery period of at least 2 h.

Experimental Protocol 2

The carotid catheter was attached to a pressure transducer (P23ID, Gould Statham Instruments, Hato Rey, Puerto Rico) for central hemodynamic recordings of MAP and HR. The caudal artery catheter was connected to a 1-ml syringe attached to a Harvard Pump (model 907, Harvard, Holliston, MA). After MAP and HR were recorded, blood withdrawal was initiated from the caudal catheter (0.25 ml min⁻¹) and the carotid catheter was disconnected from the transducer for rapid infusion of ~0.5-0.6 10⁶ 15-um-diameter microspheres (⁵⁷Co or ⁸⁵Sr in random order, Perkin Elmer Life and Analytical Sciences, Waltham, MA) into the aortic arch for the determination of tissue BF. A second pressure reading was recorded immediately following reconnection of the carotid catheter to the pressure transducer. Following 30 min, GLI (5 mg kg⁻¹) was infused via the caudal artery catheter and pressure transducer was monitored until the persistent rise in MAP occurred (~5-6 min). Administration of microspheres and pressure recordings were repeated as described above. Rats were then euthanized via pentobarbital overdose (>50 mg kg⁻¹ body mass).

Determination of BF and VC

The carotid catheter was checked for correct placement following anatomic dissection. Hindlimb muscles and muscle portions were removed, weighed, and placed in counting vials for the determination of radioactivity. Each tissue and reference sample was measured for radioactivity using a γ-scintillation counter (model 5230, Packard Auto Gamma Spectrometer, Downers Grove, IL). Radioactivity for the separate microsphere injections (⁵⁷Co or ⁸⁵Sr) was enabled after taking into account the cross-talk fraction between isotopes. Using these radioactivity counts, BF was determined for each tissue under each condition (control and GLI) by comparison to the reference sample of known flow rate and measured activity (Ishise et al. 1980; Musch and Terrell, 1992). BFs were expressed as milliliters per minute per 100 g of tissue, and also normalized to MAP to express VC (ml min⁻¹ 100g tissue⁻¹ mmHg⁻¹). Adequate mixing of microspheres for BF determination within control and GLI conditions were verified by a <15% difference in BF between the right and left kidneys or right and left hindlimbs.

Statistical Analysis

For each minute following infusion, the changes in MAP, HR, and renal and lumbar SND resulting from K_ATP channel inhibition (ΔGLI) were compared to saline infusion (ΔCON) via two-way repeated measures analysis of variance (ANOVA) with Tukey’s post hoc analysis. Changes in HR within the SAD group were then compared using a two-tailed z-test. Differences in the effect of K_ATP channel inhibition (ΔGLI-ΔCON) between groups, with and without baroreceptor regulation of MAP (Intact vs. SAD, respectively), were compared using a two-tailed t-test. When normality was not achieved utilizing the Shapiro-Wilk test statistic, Mann-Whitney Rank Sum Tests were performed. During the assessment of BF and VC in conscious rats, all data were compared using two-tailed paired t-tests. Data are presented as mean ± SE. Significance was accepted at p<0.05.

Results

Study 1: Effect of GLI on Sympathetic Nerve Discharge

Mean Arterial Pressure

Pre-infusion MAP was slightly higher prior to GLI compared to CON in the SAD group (99 ± 11 vs 105 ± 13, p<0.05); but was not different between CON and GLI infusions in the intact group (93 ± 4 vs 94 ± 4 mmHg, p>0.05). HR was not different prior to CON and GLI infusions in either group (intact: 398 ± 9 vs 412 ± 9; SAD: 433 ± 23 vs 449 ± 19 bpm; p>0.05).

In both baroceptor intact (solid bars) and denervated (SAD; open bars) rats, systemic GLI administration increased MAP significantly compared to CON (Figure 1, 3-10 min and 2-10 min, intact and SAD respectively; p<0.05). GLI-induced increases in MAP were significantly higher in SAD compared with baroreceptor intact rats (2-7 min; p<0.05).

Heart Rate

Administration of GLI led to a transient reduction in HR in rats with intact baroreceptors (Figure 2; p<0.05). However, the HR response of SAD rats was significantly increased in GLI compared to CON infusions (6-10 min; p<0.05) resulting in significantly different HR responses to K_ATP channel inhibition between groups (p<0.05).

Figure 2.

The effect of K_ATP channel inhibition (GLI; glibenclamide) on changes in heart rate. Heart rate transiently decreases in sinoaortic baroreceptor intact rats (intact (n=12); solid bars) but increases in baroreceptor denervated rats (SAD (n=4); open bars). Data are presented as mean ± SEM. * p < 0.05 vs. baseline; # p < 0.05 vs. intact

Sympathetic Nerve Discharge (SND)

Renal and lumbar SND decreased following the systemic administration of GLI in baroreceptor intact rats whereas GLI produced an increase in renal and lumbar SND in SAD rats (Figures 3 and 4; p<0.05 for all). Therefore GLI resulted in divergent responses in lumbar (i.e., hindlimb muscle) and renal SND depending on the presence or absence of intact baroreceptors (p<0.05).

Figure 3.

Changes in Renal Sympathetic Activity following K_ATP channel inhibition. The change in renal sympathetic nerve discharge (SND) is directionally different between sinoaortic baroreceptor intact rats (intact (n=8); solid bars) and baroreceptor denervated rats (SAD (n=3); open bars). Data are presented as mean ± SEM. * p < 0.05 vs. baseline; # p < 0.05 vs. intact

Study 2: Effect of GLI on Blood Flow and Vascular Conductance in Conscious Intact Rats

Mean Arterial Pressure and Heart Rate

GLI increased mean arterial pressure (130 ± 5 vs 152 ± 5 mmHg), but decreased heart rate (440 ± 12 vs 410 ± 12 bpm, CON vs GLI respectively; p < 0.05 for both).

Blood Flow and Vascular Conductance

GLI reduced resting BF to both kidneys (left: 502 ± 59 vs 375 ± 28; right: 474 ± 42 vs 384 ± 27) and total hindlimb skeletal muscle (left: 26 ± 7 vs 17 ± 5; right: 27 ± 8 vs 21 ± 6 ml min⁻¹ 100g tissue⁻¹; p<0.05 CON vs GLI, respectively). VC was also reduced in kidneys (left: 3.93 ± 0.57 vs 2.56 ± 0.25; right: 3.72 ± 0.44 vs 2.50 ± 0.25) and total hindlimb skeletal muscle (left: 0.21 ± 0.6 vs 0.14 ± 0.04; right: 0.21 ± 0.06 vs 0.11 ± 0.03 ml min⁻¹ 100g tissue⁻¹ mmHg⁻¹; p<0.05 CON vs GLI, respectively). There were no differences between left and right renal or hindlimb BF or VC (p>0.05), thus mean data are presented in Figures 5 and 6.

Figure 5.

Decreases in total hindlimb blood flow (A) and vascular conductance (B) following systemic K_ATP channel inhibition (n=6 rats with 28 individual muscles or muscle portions). * p < 0.05 vs control for total hindlimb blood flow and vascular conductance.

Figure 6.

Decreases in kidney blood flow (A) and vascular conductance (B) following systemic K_ATP channel inhibition (n=6). * p < 0.05 vs control for kidney blood flow and vascular conductance.

Discussion

The principal original finding of the present investigation is that systemic K_ATP channel inhibition via GLI (5 mg kg⁻¹) increased MAP simultaneously with decreased blood flow and vascular conductance to kidneys and hindlimb skeletal muscle, but, in baroreceptor-intact rats, did not increase sympathetic activity (SND) to these regions. Rather, decreases in SND occurred following K_ATP inhibition with increased SND only seen in the absence of sinoaortic baroreceptors (SAD rats). Thus GLI-induced increases in MAP in baroreceptor-intact rats activate the arterial baroreceptors which in turn act to reduce central sympathetic nerve outflow as measured by reductions in efferent SND directed to the kidneys and hindlimb vasculature.

Consistent with our first hypothesis in baroreceptor-intact rats, these findings suggest that the increased MAP with K_ATP channel inhibition (~20 mmHg in both preparations) was due, at least in part, to peripheral vasoconstriction originating within kidney and muscle tissue themselves (Figure 5) and not a consequence of a centrally-mediated increase in sympathetic nerve discharge leading to a vasoconstrictor response (Figures 3 and 4). These results highlight the need to preserve vascular K_ATP channel function in health and especially in disease conditions where K_ATP channels may be relied upon heavily for O₂ delivery (Holdsworth et al., 2017). This is especially the case for heart failure patients in whom endogenous nitric oxide bioavailability is compromised and muscle vascular O₂ pressures fall to levels that impair blood-myocyte O₂ flux (revs. by Hirai et al., 2015; Poole et al., 2018). Moreover, as sulphonylurea medications, such as GLI, that inhibit pancreatic K_ATP channels and increase insulin release, are the most commonly prescribed second-line diabetes medication these K_ATP channel inhibitors may also increase the risk of adverse cardiovascular events in diabetic patients with heart failure (Abdelmoneim et al., 2016; Qaseem et al., 2017). Additionally, our studies support that, by impairing K_ATP channel-mediated vasodilation, sulphonylurea treatment may exacerbate disease-related exercise intolerance by contributing to the decreased supply of blood and therefore O₂ to the exercising muscles.

Effects of systemic GLI on Mean Arterial Pressure and Heart Rate

Systemic GLI increased MAP to a greater extent in SAD rats compared to rats with intact sinoaortic baroreceptors. As seen in Figures 1 and 2, the timing of the transient decrease of HR in baroreceptor-intact rats, and lack thereof in the SAD group, was coincident with the significantly greater elevation of MAP in SAD rats versus their intact counterparts. This suggests that cardiobaroreflex-mediated decreases in cardiac output and peripheral sympathetic constriction are counteracting increases in peripheral vascular resistance (originating within muscle, kidneys, and potentially elsewhere) in an attempt to constrain the elevation of MAP following K_ATP channel inhibition. This baroreceptor modulation potentially masks the direct effects of K_ATP channel inhibition on neural activity (see Effects of systemic GLI on lumbar and renal SND below). The increased MAP following K_ATP inhibition seen herein (~20 mmHg) agrees with previous studies in rats and swine (Duncker et al., 2001; Holdsworth et al., 2015; Moreau et al., 1994) and humans (Farougue et al., 2003). Therefore the evidence is consistent with K_ATP channels, at least at rest, playing a key role in blood pressure regulation and muscle O₂ delivery via vascular smooth muscle relaxation, decreased arteriolar vascular resistance (Lu et al., 2013) and increased blood flow (Figures 5 and 6).

Effects of systemic GLI on blood flow and vascular conductance

Total hindlimb and renal BF and VC were impaired following GLI-induced K_ATP channel inhibition. Importantly, GLI inhibits vascular K_ATP channels (Lu et al., 2013; Saito et al., 2016; Thomas et al., 1997) depolarizing the membrane potential of vascular smooth muscle cells resulting in vasoconstriction. These results, in conjunction with GLI-induced increases in MAP and altered HR responses, suggest that functioning sinoaortic baroreceptors take a commanding precedence over the sympathoexcitatory response to K_ATP channel inhibition, at least until workloads approach maximal intensity. Accordingly, GLI reduces HR at rest and whilst running at low speeds (i.e., 20 and 40 m/min) but no change in HR is present at a supra-critical speed of 60 m min⁻¹ (critical speed ~44-47 m min⁻¹; Copp et al., 2010; Craig et al., 2019) when sufficient cardiovascular and vasodilatory stimuli may overwhelm the GLI-induced bradycardia evident at slower speeds (i.e., greater increase in blood [lactate] at 20 and 40, but not 60 m min⁻¹; Holdsworth et al., 2015). Furthermore, the proportional contribution of K_ATP channels in supporting muscular O₂ transport may increase in diseases such as heart failure where nitric oxide bioavailability is compromised, submaximal exercise tolerance decreases (Craig et al., 2019) and muscle microvascular oxygenation exhibits greater impairments following GLI (see Figure 2 in Holdsworth et al., 2017).

Effects of systemic GLI on lumbar and renal sympathetic nerve discharge (SND)

The effect of GLI on SND herein was dependent on the presence of intact baroreceptors. It has been shown that K_ATP channels are active in both nervous tissue (i.e., sympathoinhibition, Guo et al., 2011, Guo et al., 2016; Xiao et al., 2006) and within skeletal muscle (i.e., vascular smooth muscle, Lu et al., 2013; Saito et al., 1996; Thomas et al., 1997). Previous studies demonstrate decreased blood flow to ischemic (Banitt et al., 1996; Bijlstra et al., 1996) and contracting muscle (Holdsworth et al., 2015; Keller et al., 2004; Saito et al., 1996) and to kidneys (Holdsworth et al., 2015) following K_ATP channel inhibition with equivocal effects on blood flow in resting tissue (Duncker et al., 2001; Farouque et al., 2003; Holdsworth et al., 2015; Keller et al., 2004; Nielsen et al., 2003). The balance between concurrent increases of MAP and decreases of SND may explain the absence of blood flow changes following K_ATP channel inhibition in those studies. As seen in Figures 3 and 4, K_ATP channel inhibition in anesthetized rats resulted in significantly decreased SND to both renal and lumbar regions. Importantly, the timing of the GLI-induced decrease in blood flow herein and during treadmill running (measured ~5-6 min following GLI infusion (Holdsworth et al., 2015)) coincides temporally with the decreased lumbar SND (Figure 4). This decrease in SND does not support the notion of increased sympathetic vasoconstriction decreasing blood flow to skeletal muscle. Therefore, it is likely that the decreases in O₂ delivery following K_ATP channel inhibition (Holdsworth et al., 2016, 2017) are peripheral in origin: specifically that GLI, via inhibition of K_ATP channels, reduces the vascular smooth muscle potential (i.e. depolarization) leading to vasoconstriction and decreased blood flow (Figure 5; Banitt et al., 1996; Bijlstra et al., 1993; Farouque and Meredith, 2003; Lu et al., 2013; Quast et al., 1994; Saito et al., 1995; Thomas et al., 1997).

Experimental Considerations

It is possible that neural activity measurements made under anesthesia (i.e. rest) without electrically-induced muscle contractions may not be indicative of those influencing vasomotor tone during conscious exercise. However, even though no muscle contractions were present in the current investigation, the effect of 5 mg kg⁻¹ GLI at rest is consistent with the following findings: GLI increased MAP under anesthesia, during conscious rest and treadmill running (Holdsworth et al., 2015, 2016) and decreased skeletal muscle blood flow at rest (Figure 5) and all running speeds (20-60 m min⁻¹; Holdsworth et al., 2015); decreased HR at rest and during slow running speeds (Holdsworth et al., 2015); and increased MAP during electrically-induced muscle contractions under anesthesia with simultaneously decreased O₂ delivery-utilization matching (Holdsworth et al., 2016). Thus, the authors contend that the presence of muscle contractions per se is not likely to alter significantly the SND responses to GLI seen herein nor the conclusions based upon those responses.

Additionally, although SND recordings were not made bilaterally in the current investigation, changes in renal and total hindlimb BF and VC were symmetrical between left and right sides of conscious rats. Therefore the effect of systemic K_ATP channel inhibition on hindlimb muscle BF and VC, and subsequent baroreflex-mediated changes in SND, are considered to elicit bilaterally uniform neural transduction to renal and lumbar regions.

Conclusions

The present investigation provides compelling evidence for the role of vascular K_ATP channels in the regulation of resting blood pressure, vascular conductance and thus O₂ delivery. Sympathetic nerve activity to kidneys and skeletal muscle, and thus the central regulation of vasomotor tone, actually decreases following K_ATP channel inhibition. Regarding the increased MAP produced by GLI, this finding highlights a vascular K_ATP channel mechanism present within renal and muscular systems that contributes importantly to maintaining adequate blood flow and O₂ delivery in health, and potentially in patient populations (diabetes, heart failure) in which other vasodilatory pathways, such as nitric oxide, are downregulated. Additionally, baroreflex regulation of MAP may result in an underestimation of the extent to which GLI decreases skeletal muscle blood flow during exercise. As such, impaired vascular K_ATP channel function consequent to this diabetes medication may, in part, explain the elevated blood pressure and impaired renal/muscle O₂ delivery that hinder daily physical activity, potentially constraining the performance and benefits of exercise rehabilitation, and generally contributing to adverse cardiovascular events.

Highlights.

• ATP-sensitive K⁺ (K_ATP) channels are active in neural, vascular and muscle tissue.

• K_ATP channels regulate mean arterial pressure, blood flow and oxygen delivery.

• Inhibiting K_ATP channels decreased sympathetic activity to lumbar and renal regions.

• Impaired blood flow via K_ATP inhibition is not due to sympathetic vasoconstriction.