Contracting human skeletal muscle maintains the ability to blunt α₁-adrenergic vasoconstriction during K_IR channel and Na⁺/K⁺-ATPase inhibition

Abstract

Sympathetic vasoconstriction in contracting skeletal muscle is blunted relative to that which occurs in resting tissue; however, the mechanisms underlying this ‘functional sympatholysis’ remain unclear in humans. We tested the hypothesis that α₁-adrenergic vasoconstriction is augmented during exercise following inhibition of inwardly rectifying potassium (K_IR) channels and Na⁺/K⁺-ATPase (BaCl₂ + ouabain). In young healthy humans, we measured forearm blood flow (Doppler ultrasound) and calculated forearm vascular conductance (FVC) at rest, during steady-state stimulus conditions (pre-phenylephrine), and after 2 min of phenylephrine (PE; an α₁-adrenoceptor agonist) infusion via brachial artery catheter in response to two different stimuli: moderate (15% maximal voluntary contraction) rhythmic handgrip exercise or adenosine infusion. In Protocol 1 (n = 11 subjects) a total of six trials were performed in three conditions: control (saline), combined enzymatic inhibition of nitric oxide (NO) and prostaglandin (PG) synthesis (l-NMMA + ketorolac) and combined inhibition of NO, PGs, K_IR channels and Na⁺/K⁺-ATPase (l-NMMA + ketorolac + BaCl₂ + ouabain). In Protocol 2 (n = 6) a total of four trials were performed in two conditions: control (saline), and combined K_IR channel and Na⁺/K⁺-ATPase inhibition. All trials occurred after local β-adrenoceptor blockade (propranolol). PE-mediated vasoconstriction was calculated (%ΔFVC) in each condition. Contrary to our hypothesis, despite attenuated exercise hyperaemia of ∼30%, inhibition of K_IR channels and Na⁺/K⁺-ATPase, combined with inhibition of NO and PGs (Protocol 1) or alone (Protocol 2) did not enhance α₁-mediated vasoconstriction during exercise (Protocol 1: −27 ± 3%; P = 0.2 vs. control, P = 0.4 vs.l-NMMA + ketorolac; Protocol 2: −21 ± 7%; P = 0.9 vs. control). Thus, contracting human skeletal muscle maintains the ability to blunt α₁-adrenergic vasoconstriction during combined K_IR channel and Na⁺/K⁺-ATPase inhibition.

Key points

• During exercise there is a balance between vasoactive factors that facilitate increases in blood flow and oxygen delivery to the active tissue and the sympathetic nervous system, which acts to limit muscle blood flow for the purpose of blood pressure regulation.

• Functional sympatholysis describes the ability of contracting skeletal muscle to blunt the stimulus for vasoconstriction, yet the underlying signalling of this response in humans is not well understood.

• We tested the hypothesis that activation of inwardly rectifying potassium channels and the sodium–potassium ATPase pump, two potential vasodilator pathways within blood vessels, contributes to the ability to blunt α₁-adrenergic vasoconstriction.

• Our results show preserved blunting of α₁-adrenergic vasconstriction despite blockade of these vasoactive factors. Understanding this complex phenomenon is important as it is impaired in a variety of clinical populations.

Introduction

Muscle contractions stimulate vasodilatation of the resistance vasculature, leading to large increases in muscle blood flow to the active tissue. Due to the profound capacity of skeletal muscle blood flow to increase during exercise (Andersen & Saltin, 1985), some level of sympathetically mediated vasoconstriction is needed in order to maintain mean arterial blood pressure (Marshall et al. 1961; Buckwalter & Clifford, 2001). It has been shown that the degree to which sympathetically mediated vasoconstriction occurs in the active muscle is blunted relative to inactive muscle (Anderson & Faber, 1991; Thomas et al. 1994; Tschakovsky et al. 2002; Dinenno & Joyner, 2003; Kirby et al. 2008). This phenomenon, termed functional sympatholysis (Remensnyder et al. 1962), is thought to allow for optimal blood flow and therefore oxygen delivery to the active tissue (Joyner & Thomas, 2003) while some vasoconstriction occurs in order to limit the overall decrease in total peripheral resistance. Various laboratories have demonstrated the existence of functional sympatholysis (Tschakovsky et al. 2002; Rosenmeier et al. 2003a, 2004; Keller et al. 2004; Parker et al. 2007; Wray et al. 2007; Kirby et al. 2008; Fadel et al. 2012); however, to date, the signalling pathways that contribute to this response have not been well described in healthy humans.

Several studies have demonstrated that muscle contractions can limit vasoconstrictor responses within the active tissue to a variety of stimuli, including baroreflex increases in muscle sympathetic nerve activity and intra-arterial administration of tyramine, both of which evoke endogenous noradrenaline (NA) release from sympathetic nerve endings (Tschakovsky et al. 2002; Dinenno & Joyner, 2003; Keller et al. 2004). Additionally, it is clear that muscle contractions can limit direct α₁- and α₂-adrenoceptor-mediated vasoconstriction as assessed during intra-arterial infusions of selective agonists (Rosenmeier et al. 2003a; Kirby et al. 2008). These collective observations indicate that functional sympatholysis occurs post-junctionally at the level of the α-adrenoceptors (Thomas et al. 1994; Buckwalter et al. 1998; Rosenmeier et al. 2003a), and as such, has led to the idea that vasodilator signalling pathways within the resistance vessels are involved in this phenomenon.

While some studies have suggested there may be a role for nitric oxide (NO) in blunting sympathetic vasoconstriction during muscle contractions (Thomas & Victor, 1998; Grange et al. 2001; Chavoshan et al. 2002), several other investigations have demonstrated little involvement of NO either by inhibiting the synthesis via NO synthase during exercise (Dinenno & Joyner, 2003; Buckwalter et al. 2004) or direct NO donation (Rosenmeier et al. 2003b). Further, when NO is inhibited in combination with endothelial-derived prostaglandins (PGs), there is only a modest enhancement of α-adrenergic vasoconstriction during muscle contractions (Dinenno & Joyner, 2004). However, given that vasoconstrictor responses in a quiescent tissue were also slightly augmented in this latter study, it does not appear that NO and PGs are the primary mediators of functional sympatholysis in humans. Similar results were obtained in humans during inhibition of ATP-sensitive potassium (K_ATP) channels with oral glyburide (Keller et al. 2004). Thus, the mechanisms underlying this complex control of vascular tone in humans remain unclear.

To date, few exogenous vasodilator substances are capable of blunting sympathetically mediated vasoconstriction to an extent similar to that occurring during exercise. Specifically, in humans, exogenous ATP can attenuate vasoconstriction in response to both endogenous NA release and direct α-adrenoceptor stimulation (Rosenmeier et al. 2004; Kirby et al. 2008, 2011). We have shown ATP-mediated vasodilatation is largely independent of NO and PG synthesis (Crecelius et al. 2011a) and occurs primarily through inwardly rectifying potassium (K_IR) channel activation (Crecelius et al. 2012). Stimulation of K_IR channels leads to changes in vascular cell membrane potential and subsequent hyperpolarization, which can also occur with activation of Na⁺/K⁺-ATPase (Jackson, 2005). Additionally, in vitro evidence suggests a prominent role for K_IR channels in amplifying hyperpolarization of any origin and facilitating robust cell-to-cell communication and conducted vasodilatation (Jantzi et al. 2006). Importantly, Segal and colleagues have demonstrated that conducted vasodilatation, which can occur via hyperpolarization (Emerson & Segal, 2000), can overcome sympathetically induced vasoconstriction; at the same time, sympathetic stimulation attenuates conducted vasodilatation (Kurjiaka & Segal, 1995; VanTeeffelen & Segal, 2003; Haug & Segal, 2005). This reciprocal relationship is strikingly similar to the vascular control observed in contracting human muscle (Joyner & Thomas, 2003).

Recently, our laboratory demonstrated that combined inhibition of K_IR channels and Na⁺/K⁺-ATPase reduces exercise hyperaemia ∼30% in the forearm and this was predominantly attributed to activation of K_IR channels (Crecelius et al. 2014). These studies were performed during a mild exercise intensity which does not engage the sympathetic nervous system, and thus whether these specific pathways are obligatory to observe functional sympatholysis is unknown. With this information as background, we directly tested the hypothesis that α₁-adrenergic vasoconstriction is augmented during exercise following inhibition of K_IR channels and Na⁺/K⁺-ATPase in healthy humans.

Methods

Subjects

With Institutional Review Board approval and after written informed consent, a total of 17 young healthy adults (Protocol 1: 8 men and 3 women, age 23 ± 1 years old, weight 67.6 ± 2.9 kg, height 173 ± 3 cm, body mass index 22.4 ± 1.1 kg m⁻², forearm volume (FAV) 913 ± 58 ml; Protocol 2: 4 men and 2 women, age 23 ± 2 years old, weight 70.7 ± 4.6 kg, height 175 ± 3 cm, body mass index 23.0 ± 1.3 kg m⁻², FAV 968 ± 99 ml (means ± SEM)) participated in the present study. All subjects were sedentary to moderately active, non-smokers, non-obese, normotensive (resting blood pressure <140/90 mmHg), and not taking any medications. Studies were performed after an overnight fast and 24 h abstention from caffeine and exercise. The subjects were in the supine position with the experimental arm abducted to 90° and slightly elevated above heart level upon a tilt-adjustable table. Female subjects were studied during the early follicular phase of their menstrual cycle or placebo phase of oral contraceptive use to minimize any potential cardiovascular effects of sex-specific hormones. All studies were performed according to the Declaration of Helsinki.

Arterial and venous catheterization, arterial blood pressure and heart rate

A 20 gauge, 7.6 cm catheter was placed in the brachial artery of the non-dominant (experimental) arm under aseptic conditions after local anaesthesia (2% lidocaine (lignocaine)) for local administration of study drugs and blood sampling. The catheter was connected to a 3-port connector as well as a pressure transducer for mean arterial pressure (MAP) measurement and continuously flushed at 3 ml h⁻¹ with heparinized saline. The two side ports were used for drug infusions of vasoactive drugs (Kirby et al. 2008; Crecelius et al. 2010). In addition, in Protocol 1, an 18 or 20 gauge (depending on visual inspection of vein size), 5.1 cm catheter was inserted in retrograde fashion into an antecubital vein of the experimental arm for deep venous blood samples. Saline was continuously infused through this catheter at a rate of approximately 3 ml min⁻¹ for the duration of the study to keep it patent (Crecelius et al. 2011b). Heart rate (HR) was determined using a 3-lead electrocardiogram (Cardiocap/5, Datex-Ohmeda, Louisville, CO, USA).

Blood gas analysis

In Protocol 1, brachial artery and deep venous blood samples (∼2 ml) were sampled anaerobically and immediately (<1 min) analysed with a clinical blood gas analyser (Rapid Point 400 Series Automatic Blood Gas System, Siemens Healthcare Diagnostics, Deerfield, IL, USA) for partial pressures of oxygen and carbon dioxide ( and ), fraction of oxygenated haemoglobin (), oxygen content, pH and [K⁺].

Forearm blood flow and vascular conductance

A 12 MHz linear-array ultrasound probe (Vivid 7, General Electric, Milwaukee, WI, USA) was used to determine brachial artery mean blood velocity (MBV) and brachial artery diameter. The probe was securely fixed to the skin over the brachial artery proximal to the catheter insertion site as previously described (Crecelius et al. 2010). For blood velocity measurements, the probe insonation angle was maintained at <60 deg and the frequency used was 5 MHz. The Doppler shift frequency spectrum was analysed via a Multigon 500M TCD (Multigon Industries, Mt Vernon, NY, USA) spectral analyser from which mean velocity was determined as a weighted mean of the spectrum of Doppler shift frequencies. Brachial artery diameter was measured in triplicate at the end of rest, steady-state conditions pre-constriction, and end of constrictor effect (see General Timeline below). Forearm blood flow (FBF) was calculated as: FBF = MBV × π × (brachial artery diameter/2)² × 60, where the FBF is in ml min⁻¹, the MBV is in cm s⁻¹, the brachial diameter is in cm, and 60 is used to convert from ml s⁻¹ to ml min⁻¹. Forearm vascular conductance (FVC) was calculated as (FBF/MAP) × 100, and expressed as ml min⁻¹ (100 mmHg)⁻¹. All studies were performed in a cool (20–22°C) temperature-controlled environment with a fan directed toward the forearm to minimize the contribution of skin blood flow to forearm haemodynamics.

Rhythmic handgrip exercise

Maximal voluntary contraction (MVC; mean 41 ± 3 kg, range 19–67 kg) was determined for the experimental arm as the average of three maximal squeezes of a handgrip dynamometer (Stoelting, Chicago, IL, USA) that were within 3% of each other. Forearm exercise during the trials was performed with weight corresponding to 15% MVC (mean 6.0 ± 0.5 kg, range 2.9–10.1 kg) attached to a pulley system and lifted 4–5 cm over the pulley at a duty cycle of 1 s contraction–2 s relaxation (20 contractions per minute). Visual and auditory feedback were used to ensure the correct timing as described previously (Kirby et al. 2008; Crecelius et al. 2010). We chose this moderate intensity rhythmic handgrip exercise to limit the contribution of systemic haemodynamics to forearm hyperaemic responses and eliminate reflex activation of the sympathetic nervous system (Seals & Victor, 1991; Limberg et al. 2014). Previous studies in our laboratory have determined that MVC is not affected by any of the vasoactive substances administered in the present study, particularly barium chloride and ouabain (Crecelius et al. 2013).

Vasoactive drug infusion

All drug infusions occurred via a brachial artery catheter to create a local effect in the forearm. Phenylephrine (PE; see below) is capable of stimulating β-adrenoceptor-mediated vasodilatation (Torp et al. 2001). In order to avoid this potentially confounding effect, we infused propranolol (a non-selective β-adrenoceptor antagonist; West-Ward Pharmaceutical Corp., Eatontown, NJ, USA) at a loading dose of 200 μg min⁻¹ for 5 min (total, 1000 μg) and continued infusions at a maintenance dose of 50 μg min⁻¹ prior to any experimental trials and for the duration of the protocol. This dose of propranolol has been documented to block forearm vasodilatation to isoproterenol (isoprenaline) (Johnsson, 1967) and we have used this approach in similar studies where PE-mediated vasoconstriction has been investigated (Dinenno et al. 2002a,b). Regarding any potential effects of β-adrenoceptor inhibition during exercise, evidence indicates that β-mediated vasodilatation does not contribute to exercise hyperaemia in a similar experimental model (Hartling et al. 1980).

Given the large increase in blood flow that occurs as a result of muscle contractions, we infused adenosine (Akorn, Lake Forest, IL, USA) at variable doses to elevate forearm blood flow of a quiescent tissue to levels similar to those observed during exercise and this served as a ‘high flow control’ in each experimental condition. We have previously demonstrated that contracting muscle blunts direct α₁- and α₂-adrenoceptor-mediated vasoconstriction, whereas vasoconstrictor responses are preserved during adenosine infusion (Dinenno & Joyner, 2004; Kirby et al. 2008, 2011).

The selective α₁-adrenoceptor agonist PE (Sandoz, Princeton, NJ, USA) was infused at 0.125 μg (dl FAV)⁻¹ min⁻¹ to experimentally mimic the vasoconstriction that would occur during exercise-induced activation of the sympathetic nervous system (Buckwalter & Clifford, 1999). The absolute dose of PE was adjusted to the appropriate hyperaemic condition as previously described (Kirby et al. 2008, 2011) in order to normalize concentrations to this relative dose based on the observed level of forearm blood flow during adenosine infusion and exercise (see below). This relative dose of PE is twice that used in male subjects (Kirby et al. 2008, 2011) and this was done in order to ensure that all subjects would demonstrate vasoconstriction during exercise in control conditions, and thus provide an appropriate control for any potential augmentation of this response due to our pharmacological inhibitors.

A number of vasodilator pathways were inhibited in these experimental protocols. N^G-Monomethyl-l-arginine (l-NMMA; Bachem, Weil am Rhein, Germany) was administered to inhibit nitric oxide synthase-mediated production of NO, and ketorolac (Hospira, Lake Forest, IL, USA) was administered to inhibit cyclooxygenase-mediated synthesis of PGs. Loading doses of l-NMMA and ketorolac were 25 mg (5 mg min⁻¹ for 5 min) and 6 mg (600 μg min⁻¹ for 10 min), respectively (Crecelius et al. 2011b, 2012). In subsequent trials, l-NMMA and ketorolac were infused at maintenance doses of 1.25 mg min⁻¹ and 150 μg min⁻¹, respectively. To inhibit vascular hyperpolarization via Na⁺/K⁺-ATPase and K_IR channel activation, ouabain octahydrate (Na⁺/K⁺-ATPase inhibitor, Sigma-O3125; Sigma-Aldrich, St Louis, MO, USA) was infused at 2.7 nmol min⁻¹ in combination with BaCl₂ (K_IR channel inhibitor; 10% w/v BDH3238, EMD Chemicals, Gibbstown, NJ, USA) at 0.9 μmol (dl FAV)⁻¹ min⁻¹ with an absolute dose of 8–10 μmol min⁻¹ (Crecelius et al. 2012, 2014). Ouabain and BaCl₂ were loaded for 15 and 3 min, respectively, prior to the first trial (exercise or adenosine) of the inhibited condition and were continued at these same doses for the duration of the trial. In the second trial of the blocked condition, infusion of all inhibitors began 3 min before the start of muscle contractions or adenosine infusion. Ouabain and BaCl₂ were prepared in saline and confirmed sterile and free of fungus/endotoxin and particulate matter with a standard microbiology report (JCB-Analytical Research Labs, Wichita, KS, USA) prior to use. The forearm volume used for normalization for specific vasoactive drugs was determined from regional analysis of whole-body dual-energy X-ray absorptiometry scans (QDR series software, Hologic, Inc., Bedford, MA, USA).

General timeline

Figure1 presents the overall experimental timelines (A and B) as well as the protocol for each individual trial (C). After instrumentation and β-adrenoceptor inhibition, subjects performed a bout of handgrip exercise or received intra-arterial adenosine infusion (final average dose: Protocol 1, 46.2 ± 5.6 μg (dl FAV)⁻¹ min⁻¹; Protocol 2, 46.1 ± 9.3 μg (dl FAV)⁻¹ min⁻¹) matched to the predicted/observed hyperaemic response to exercise. The total length of each trial was 9 min, consisting of 3 min of baseline conditions and 6 min total of exercise or adenosine infusion. Steady-state conditions were achieved after 3 min of hyperaemia (minute 6 of Fig.1) and the dose of PE was calculated on the basis of FAV and FBF. Vasoconstrictor infusion then began after this (4 min into the stimulus) and lasted for 2 min. In the first exercise and adenosine trials, saline was used as a control infusate. Exercise and adenosine trials were counterbalanced in each condition between subjects. In cases where the adenosine trial preceded exercise, for the purposes of adjusting the adenosine dose, we predicted the level of hyperaemia based on our experience with exercise responses in control and inhibited conditions (Dinenno & Joyner, 2004; Schrage et al. 2004; Crecelius et al. 2011b, 2012, 2014). All trials were separated by 15 min of rest.

Figure 1.

Experimental timeline

A, in Protocol 1, after instrumentation of the brachial artery catheter, subjects received propranolol to inhibit β-adrenoceptors and then underwent exercise and adenosine (ADO) trials (see C) in control (saline) conditions. l-NMMA and ketorolac were then administered in order in inhibit the synthesis of nitric oxide (NO) and prostaglandins (PGs), respectively, and exercise and ADO trials were repeated. Finally, BaCl₂ and ouabain were infused in order to inhibit vascular hyperpolarization via activation of inwardly rectifying potassium (K_IR) channels and Na⁺/K⁺-ATPase, respectively, in combination with NO and PGs. B, in Protocol 2, the experimental approach was similar; however, only two conditions were examined: control (saline) and inhibition of K_IR channels and Na⁺/K⁺-ATPase (BaCl₂ + ouabain). In each individual trial for Protocol 1 (C), forearm blood flow (FBF) was measured for 3 min of baseline. Subjects then performed exercise (15% maximal voluntary contraction (MVC) rhythmic handgrip exercise (RHG)) or received variable-dose ADO infusion to match hyperaemia observed with exercise. After 4 min of the stimulus (adenosine or exercise) ‘steady-state’ FBF was calculated and a dose of phenylephrine (PE; α₁-adrenoceptor agonist) based on this FBF was then infused for 2 min and vasoconstrictor responses were determined. Arterial blood samples were taken during Protocol 1 at rest and venous (retrograde antecubital catheter) samples were taken at baseline, steady-state pre-PE and end-PE. Similar trials were performed in Protocol 2.

Protocol 1 (n = 11)

Given evidence for interactions of hyperpolarization and endothelium-dependent vasodilators, particularly NO (Bauersachs et al. 1996; Taddei et al. 1999), prior to investigating the role of vascular hyperpolarization via K_IR and Na⁺/K⁺-ATPase activation, we inhibited the production of NO via l-NMMA. Redundancy and compensation has also been observed between NO and PGs during physiological stressors (Schrage et al. 2004; Markwald et al. 2011) and thus to eliminate this, we also inhibited PG synthesis via ketorolac. Combined inhibition of NO and PGs also served to support the previous findings that these pathways do not contribute significantly to functional sympatholysis (Dinenno & Joyner, 2004). Following control trials, l-NMMA and ketorolac were administered in loading doses, and exercise and adenosine (final average dose: 59.5 ± 6.9 μg (dl FAV)⁻¹ min⁻¹) trials were repeated in this condition of combined inhibition of NO and PGs. After these third and fourth trials, BaCl₂ and ouabain were infused and exercise and adenosine (final average dose: 45.2 ± 7.5 μg (dl FAV)⁻¹ min⁻¹) trials were performed with combined inhibition of NO, PGs, K_IR channels and Na⁺/K⁺-ATPase.

Venous blood samples were taken at rest, in steady-state hyperaemic conditions (pre-PE) and at end-PE. Arterial blood samples were only taken at baseline as these samples interfere with the ability to measure FBF and we have shown arterial blood gases to be unchanged with mild–moderate intensity rhythmic handgrip exercise (Crecelius et al. 2011b).

Protocol 2 (n = 6)

In order to directly address the roles of K_IR channel and Na⁺/K⁺-ATPase activation, a second protocol occurred where these pathways were inhibited alone rather than in the presence of prior NO and PG inhibition. Similar to Protocol 1, control (saline) trials of exercise and adenosine infusion were followed by BaCl₂ and ouabain infusion during exercise and adenosine (final average dose: 41.3 ± 8.1 μg (dl FAV)⁻¹ min⁻¹) trials.

Data acquisition and analysis

Data were collected and stored on computer at 250 Hz and analysed off-line with signal-processing software (WinDaq; DATAQ Instruments, Akron, OH, USA). FBF, FVC, MAP and HR represent an average of the last 30 s of the corresponding time point within a trial (Fig.1; baseline, pre-PE, PE). To quantify the vasoconstrictor effect of PE, the percentage change (%Δ) was calculated as: (FVC_PE − FVC_Pre-PE)/(FVC_Pre-PE) × 100. We use %Δ in FVC as our standard index to compare vasoconstrictor responses as this appears to be the most appropriate way to compare vasoconstrictor responsiveness under conditions where there might be differences in vascular tone and blood flow, rather than pressure, is the main variable changing (Lautt, 1989; Buckwalter & Clifford, 2001). Forearm oxygen consumption () was calculated as: FBF × (O₂ content_arterial – O₂ content_venous) and presented in ml min⁻¹.

Statistics

All values are reported as means ± SEM. Given the complexity of our experimental design, many comparisons are possible. We have used a statistical approach that focuses on the comparisons that are most relevant to our experimental questions. Absolute hyperaemic forearm haemodynamics were assessed by two-way (time point (pre-PE or PE) × condition (control, l-NMMA + ketorolac, BaCl₂ + ouabain, or l-NMMA + ketorolac + BaCl₂ + ouabain)) repeated measures (RM) ANOVA inclusive of both adenosine and exercise trials. Due to the predictably large change from rest to exercise/adenosine, resting forearm haemodynamics were analysed separately by one-way RM ANOVA. Systemic haemodynamics were evaluated by two-way RM ANOVA (time point × condition) with all time points (rest, pre-PE, PE) included. Vasoconstrictor responses (%Δ) were compared with a two-way (stimulus (exercise or adenosine) × condition (control, l-NMMA + ketorolac, BaCl₂ + ouabain or l-NMMA + ketorolac + BaCl₂ + ouabain)) RM ANOVA. Blood gas variables for the pre-PE and PE time points of exercise trials were analysed with two-way ANOVA while a one-way ANOVA was used for baseline values. Student–Newman–Keuls post hoc testing was performed to make pairwise comparisons. Changes in the haemodynamic response to exercise were compared with paired Student’s t tests. Significance was set at P < 0.05.

Results

Protocol 1

Systemic haemodynamics in all experimental conditions are presented in Table1. Small increases in MAP and HR occurred with exercise and throughout the course of the experiment and these are detailed in Table1.

Table 1.

Protocol 1: Systemic haemodynamics and forearm vascular conductance in all conditions

					FVC (ml min−1
Condition	Time point	Trial	MAP (mmHg)a	HR (beats min−1)b	(100 mmHg)−1)
Control	Baseline	Adenosine	94 ± 2	57 ± 2	31 ± 3
		Exercise	92 ± 2	60 ± 2	35 ± 2
	Pre-PE	Adenosine	97 ± 2	59 ± 2	232 ± 21
		Exercise	96 ± 2	66 ± 2	239 ± 19
	PE	Adenosine	101 ± 3	59 ± 2	117 ± 11
		Exercise	99 ± 3	64 ± 1	187 ± 12*
+ l-NMMA + ketorolac	Baseline	Adenosine	96 ± 2	54 ± 2	25 ± 3†
		Exercise	98 ± 2	57 ± 2	25 ± 2†
	Pre-PE	Adenosine	98 ± 3	56 ± 2	204 ± 11
		Exercise	100 ± 2	62 ± 2	219 ± 19
	PE	Adenosine	103 ± 2	57 ± 2	111 ± 14
		Exercise	104 ± 3	61 ± 2	164 ± 13*
+ l-NMMA + ketorolac	Baseline	Adenosine	100 ± 3	56 ± 2	28 ± 3
+ BaCl2 + ouabain		Exercise	101 ± 2	58 ± 2	26 ± 3†
	Pre-PE	Adenosine	106 ± 2	57 ± 2	142 ± 18†‡
		Exercise	110 ± 3	63 ± 3	137 ± 18†‡
	PE	Adenosine	109 ± 3	55 ± 2	39 ± 5†‡
		Exercise	113 ± 4	62 ± 2	97 ± 12*†‡

PE, phenylephrine; n = 11.

^*P < 0.05 vs. adenosine (within condition)

^†P < 0.05 vs. control

^‡P < 0.05 vs. + l-NMMA + ketorolac.

^aP < 0.05, main effects of Time point and Condition; P < 0.05 Time point × Condition.

^bTime points: within all adenosine trials P = NS (not significant); within all exercise trials: P < 0.05 Baseline vs. Pre-PE; P = NS Pre-PE vs. PE; Condition: P < 0.05 main effect.

As intended by experimental design, FBF during adenosine infusion and exercise was matched prior to PE infusion in all experimental conditions (Fig.2A) as was FVC (Table1). In all trials, PE reduced FBF from these matched pre-PE levels. FBF at the end of PE infusion was always significantly greater in exercise conditions, as compared to control.

Figure 2.

Absolute forearm blood flow responses across all experimental conditions

Forearm blood flow (FBF) responses are presented at rest (black bars), pre-phenylephrine (pre-PE) vasoconstrictor stimulus (grey bars) and end-PE (open bars) in all experimental conditions. In Protocol 1 (A) as intended, FBF was matched between adenosine (ADO) infusion and 15% maximal voluntary contraction (MVC) rhythmic handgrip exercise. PE elicited a significant reduction in FBF in all conditions. In control (saline), FBF after PE was significantly greater during exercise than ADO infusion. l-NMMA and ketorolac, to inhibit nitric oxide (NO) and prostaglandin (PG) synthesis, respectively, reduced baseline FBF but had no effect on pre-PE or PE FBF for either ADO or exercise. The addition of BaCl₂ and ouabain to inhibit vascular hyperpolarization via inwardly rectifying potassium channels and Na⁺/K⁺-ATPase significantly attenuated exercise hyperaemia, and FBF during PE was lower than in control and + l-NMMA + ketorolac conditions, but still greater than during the ADO trial within this condition. In Protocol 2 (B), similar results were obtained. Infusion of BaCl₂ + ouabain significantly attenuated exercise hyperaemia and FBF during PE was lower than in control but still greater than during the ADO trial within this condition. *P < 0.05 vs. ADO (within condition); ^†P < 0.05 vs. control; ^‡P < 0.05 vs. + l-NMMA + ketorolac.

Infusion of l-NMMA + ketorolac reduced baseline FBF in both the adenosine and exercise trial but had no effect on steady-state FBF pre-PE (Fig.2A). When quantified as a percentage change from control, there was no significant change in FBF and a small but significant decrease in FVC (Fig.3A). For both adenosine and exercise, there was no impact of l-NMMA + ketorolac infusion on the vasoconstrictor responses (%ΔFVC) to PE (Fig.4A).

Figure 3.

Impact of experimental inhibition on steady-state forearm haemodynamics during exercise

The relative change in steady-state forearm blood flow (FBF; black bars) and forearm vascular conductance (FVC; grey bars) due to pharmacological inhibition compared to the control (saline) condition is shown. In Protocol 1 (A), inhibition of the synthesis of nitric oxide (NO) and prostaglandins (PGs) via l-NMMA and ketorolac, respectively, had no impact on FBF and reduced FVC. The additional inhibition of inwardly rectifying potassium (K_IR) channels (BaCl₂) and Na⁺/K⁺-ATPase (ouabain) had a substantial effect on both FBF and FVC, attenuating exercise hyperaemia and vasodilatation ∼40% from control. In Protocol 2 (B), BaCl₂ + ouabain significantly reduced both exercise hyperaemia and vasodilatation ∼30% compared to control. *P < 0.05 vs. zero; ^†P < 0.05 vs. + l-NMMA +ketorolac.

Figure 4.

Phenylephrine-mediated vasoconstriction in all experimental trials

The vasoconstriction elicited by phenylephrine (PE) infusion is presented for adenosine (ADO; black bars) and 15% maximal voluntary contraction rhythmic handgrip exercise (15% MVC Ex; grey bars) trials. Protocol 1 (A) included the following conditions: control (saline), combined inhibition of nitric oxide (NO) and prostaglandins (PGs) via l-NMMA and ketorolac, respectively, and combined inhibition of NO, PGs and vascular hyperpolarization via activation of inwardly rectifying potassium (K_IR) channels (BaCl₂) and Na⁺/K⁺-ATPase (ouabain). In all conditions, exercise blunted PE-mediated vasoconstriction. Combined inhibition of NO, PGs, K_IR channels and Na⁺/K⁺-ATPase augmented PE-mediated vasoconstriction during ADO infusions but did not alter the vasoconstrictor response during exercise. In Protocol 2 (B), exercise significantly blunted PE-mediated vasoconstriction. The inhibition of K_IR channels and Na⁺/K⁺-ATPase augmented PE-mediated vasoconstriction during ADO infusions but did not alter the attenuated vasoconstrictor response during exercise.*P < 0.05 vs. ADO (within condition); ^†P < 0.05 vs. control; ^‡P < 0.05 vs. + l-NMMA + ketorolac.

With infusion of l-NMMA + ketorolac + BaCl₂+ ouabain, baseline FBF was still significantly reduced from control in the exercise trial; however, this was not significant in the adenosine trial (P = 0.27). Steady-state pre-PE FBF was significantly lower during exercise and by design in the adenosine trial (Fig.2A). As shown in Fig.3A, the magnitude of the effect on the exercise response was rather profound (FBF: ∼40% reduction; FVC: ∼45% reduction). Inhibition of NO, PG, K_IR channels and Na⁺/K⁺-ATPase significantly augmented PE-mediated vasoconstriction during adenosine. In contrast, the vasoconstrictor response during exercise was unchanged (Fig.4A). Accordingly, FBF with PE was still greater during the exercise trial as compared to during adenosine (Fig.2A).

Blood gases from the arterial and deep venous samples obtained prior to and during the exercise trials are presented in Table2. For simplicity, we have omitted presenting samples obtained during adenosine trials, as any observed differences were predictable based on changes in blood flow to a quiescent tissue and we were most interested in the exercise responses. A venous catheter was unsuccessful in one subject and due to equipment malfunction some variables were not analysed for all samples, resulting in a range of n values (7–10) for these data. The significant changes observed were largely predictable based on the vasoconstriction induced by PE in each condition. Of note, oxygen extraction increased (increased arterial–venous O₂ difference (a-), decreased and ) pre-PE with infusion of l-NMMA + ketorolac + BaCl₂ + ouabain as compared to control and l-NMMA + ketorolac conditions. This coincided with decreased blood flow and resulted in being maintained near previous levels. In control and l-NMMA + ketorolac conditions, a- significantly increased with PE infusion. In contrast, increased oxygen extraction did not occur with l-NMMA + ketorolac + BaCl₂ + ouabain and combined with attenuated FBF resulted in a significantly lower as compared to control levels. Reduced was paralleled with lower pH and greater and [K⁺] in this condition of combined inhibition of NO, PG and vascular hyperpolarization via K_IR channels and Na⁺/K⁺-ATPase.

Table 2.

Resting arterial and deep venous blood gases in exercise trials

							[K+]	a-
Condition			(mmHg)	(%)	pH	(mmHg)	(mmol l−1)	(ml dl−1)	(ml min−1)
Control	Arterial blood		82.0 ± 1.0	94.3 ± 0.1	7.402 ± 0.007	37.4 ± 1.1	3.86 ± 0.03	—	—
		Baseline	27.9 ± 2.1	46.2 ± 4.9	7.347 ± 0.008	47.5 ± 1.1	4.04 ± 0.09	11.1 ± 1.1	3.6 ± 0.4
	Deep venous	Ex pre-PE	23.9 ± 0.9	33.9 ± 1.6	7.271 ± 0.011	59.8 ± 1.9	4.75 ± 0.07	13.6 ± 0.7	32.6 ± 2.9
	blood	Ex PE	20.4 ± 1.6	24.5 ± 3.1*	7.249 ± 0.010	62.6 ± 1.5	4.55 ± 0.06	15.5 ± 0.8*	30.0 ± 3.0
+ l-NMMA	Arterial blood		84.1 ± 1.6	94.7 ± 0.2	7.408 ± 0.009	34.8 ± 1.1	3.90 ± 0.05	—	—
+ ketorolac		Baseline	23.1 ± 1.7	34.2 ± 4.0†	7.321 ± 0.009	49.2 ± 2.0	3.93 ± 0.12	13.5 ± 1.0	3.5 ± 0.3
	Deep venous	Ex pre-PE	22.7 ± 1.0	31.3 ± 1.8	7.278 ± 0.014	56.0 ± 1.9	4.68 ± 0.13	14.1 ± 0.6	32.3 ± 4.1
	blood	Ex PE	19.7 ± 1.3	22.9 ± 2.7*	7.243 ± 0.013*	61.6 ± 3.2	4.62 ± 0.09	16.1 ± 0.7*	26.8 ± 3.7
+ l-NMMA	Arterial blood		81.9 ± 1.8	94.1 ± 0.3	7.396 ± 0.003	36.4 ± 0.5	3.94 ± 0.09	—	—
+ ketorolac		Baseline	26.1 ± 2.7	44.1 ± 5.7‡	7.328 ± 0.009	44.7 ± 1.4	3.30 ± 0.11†‡	12.3 ± 1.1	3.5 ± 0.3
+ BaCl2	Deep venous	Ex pre-PE	18.3 ± 1.3†‡	19.9 ± 1.6†‡	7.239 ± 0.016	61.4 ± 3.2	5.10 ± 0.15†‡	16.4 ± 0.5†‡	25.4 ± 3.7
+ ouabain	blood	Ex PE	17.3 ± 2.0	15.8 ± 3.6*†	7.196 ± 0.019*†‡	68.4 ± 3.7*	5.00 ± 0.16†‡	17.1 ± 0.7	19.0 ± 2.8†

a-, arterial–venous O₂ content difference; Ex, Exercise; PE, phenylephrine. Pre-PE and PE were statistically analysed separately from baseline given the predictable large changes due to exercise. Due to technical error, n = 7–10.

^*P < 0.05 vs. Pre-PE (within condition)

^†P < 0.05 vs. control

^‡P < 0.05 vs. + l-NMMA + ketorolac.

Protocol 2

Systemic haemodynamics in all experimental conditions are presented in Table3. Similar to Protocol 1, small increases in MAP occurred with exercise and throughout the course of the experiment.

Table 3.

Protocol 2: Systemic haemodynamics and forearm vascular conductance

					FVC (ml min−1
Condition	Time point	Trial	MAP (mmHg)a	HR (beats min−1)	(100 mmHg)−1)
Control	Baseline	Adenosine	93 ± 4	54 ± 3	37 ± 4
		Exercise	93 ± 2	56 ± 2	38 ± 2
	Pre-PE	Adenosine	97 ± 4	55 ± 3	252 ± 26
		Exercise	93 ± 2	56 ± 2	254 ± 28
	PE	Adenosine	100 ± 4	54 ± 3	135 ± 19
		Exercise	98 ± 2	57 ± 1	197 ± 18*
+ BaCl2 + ouabain	Baseline	Adenosine	96 ± 3	50 ± 3	29 ± 2†
		Exercise	92 ± 3	53 ± 2	31 ± 3†
	Pre-PE	Adenosine	98 ± 3	52 ± 3	192 ± 31†
		Exercise	97 ± 3	54 ± 1	179 ± 18†
	PE	Adenosine	102 ± 3	54 ± 4	63 ± 7†
		Exercise	103 ± 4	55 ± 2	135 ± 4*†

PE, phenylephrine; n = 6.

^*P < 0.05 vs. adenosine (within condition)

^†P < 0.05 vs. control.

^aTime points: within all trials, P < 0.05 Baseline vs. PE; within control: Adenosine, P < 0.05 Baseline vs. Pre-PE; within BaCl₂ + ouabain: Exercise, P < 0.05 Baseline vs. Pre-PE, P < 0.05 Pre-PE vs. PE.

Absolute FBF and FVC were well matched between adenosine and exercise conditions prior to PE infusions (Fig.2B and Table3). In all trials, PE significantly reduced FBF from these pre-matched PE levels. FBF at the end of PE infusion was always significantly greater in exercise conditions, as compared to control.

Infusion of BaCl₂ + ouabain reduced baseline FBF in both the adenosine and exercise trial and attenuated steady-state FBF pre-PE (Fig.2B). When quantified as a percentage change from control steady-state levels, there was a significant effect on the exercise response (FBF, ∼30% reduction; FVC, ∼30% reduction). This effect was slightly less than the impact of l-NMMA + ketorolac + BaCl₂ + ouabain in Protocol 1 (Fig.3), consistent with earlier observations from our laboratory (Crecelius et al. 2014). Inhibition of K_IR channels and Na⁺/K⁺-ATPase significantly augmented PE-mediated vasoconstriction during adenosine. In contrast, the vasoconstrictor response during exercise was unchanged (Fig.4B). Accordingly, FBF with PE was still greater during the exercise trial as compared to during adenosine (Fig.2B). The magnitude of the changes in vasoconstrictor responsiveness to phenylephrine was similar to that observed in Protocol 1.

Discussion

The purpose of the present study was to determine whether activation of K_IR channels and Na⁺/K⁺-ATPase mediates functional sympatholysis in healthy humans. We specifically tested the hypothesis that inhibition of K_IR channels and Na⁺/K⁺-ATPase would significantly impair the ability of contracting skeletal muscle to blunt direct post-junctional α₁-adrenergic vasoconstriction, as compared to the vasoconstrictor response observed in resting muscle. The primary novel finding of the present study is that a preserved blunting of α₁-adrenergic vasoconstriction persists in contracting skeletal muscle during combined inhibition of K_IR channels and Na⁺/K⁺-ATPase, alone or combined with inhibition of NO and PGs. This occurs despite augmented constrictor responsiveness during a control vasodilator condition (Fig.4A and B). The present results also support our previous findings that activation of K_IR channels contributes to exercise hyperaemia in the human forearm (Fig.3) (Crecelius et al. 2014) and that there is a minimal role for NO and PGs in mediating exercise hyperaemia (Schrage et al. 2004) and sympatholysis (Dinenno & Joyner, 2003, 2004) in this model of exercise. Collectively, these results suggest that vascular signalling beyond the activation of K_IR channels and Na⁺/K⁺-ATPase as well as NO and PGs is requisite for the modulation of α₁-adrenergic vasoconstriction observed during moderate, rhythmic muscle contractions of the human forearm.

Vascular signalling pathways underlying functional sympatholysis

Identifying the sympatholytic ‘factor’ during muscle contractions has been the aim of a number of different investigators, and findings across a range of experimental models and using various approaches have yielded conflicting results. Previous studies in the rat hindlimb clearly demonstrated a role for NO in functional sympatholysis (Thomas et al. 1998; Thomas & Victor, 1998), and these findings were supported by the same investigative team in humans utilizing changes in forearm tissue oxygenation as an index of vasoconstriction in response to reflex activation of the sympathetic nervous system (Chavoshan et al. 2002). In contrast, other studies in the human forearm have demonstrated that direct α₁- and α₂-adrenoceptor responsiveness during exercise is unchanged by the inhibition of NO synthase, and infusion of a NO donor is incapable of modulating sympathetic vasoconstriction to an extent similar to that occurring in the vasculature of contracting skeletal muscle (Dinenno & Joyner, 2003; Rosenmeier et al. 2003b). These findings are similar to those in experimental dogs with respect to inhibition of NO synthase not affecting direct α₂-mediated vasoconstriction during exercise (Buckwalter et al. 2004), and those in the hamster retractor muscle with respect to the failure of a NO donor to modulate sympathetic constriction (VanTeeffelen & Segal, 2003). Prostaglandins, another group of endothelium-derived vasodilators, have also been shown to be capable of blunting sympathetic vasoconstriction in certain conditions (Lippton et al. 1981; Faber et al. 1982); however, when synthesis of PGs was inhibited in the human forearm, similar to NO, vasoconstrictor responsiveness during exercise was unaltered (Dinenno & Joyner, 2004). In this same study, synthesis of NO and PGs was inhibited in combination, given the potential compensatory crossover of these pathways (Barker et al. 1996; Osanai et al. 2000). The observed sympathetic vasoconstrictor responses during exercise were slightly augmented compared to control conditions; however, they were substantially blunted compared with those at rest (Dinenno & Joyner, 2004). Given this range of findings, as well as evidence for interaction amongst NO, PGs and vascular hyperpolarization (Bauersachs et al. 1996; Taddei et al. 1999), before addressing a potential role of hyperpolarization via K_IR channel and Na⁺/K⁺-ATPase activation in the present study, our primary approach was to inhibit these pathways to: (1) eliminate any potential compensatory effects, and (2) confirm a minimal role of NO and PGs in sympatholysis in humans.

The present findings support the previous work (Dinenno & Joyner, 2004) that α₁-adrenergic vasoconstriction is significantly blunted during muscle contractions as compared to a resting condition under combined inhibition of NO and PG synthesis (Fig.4A). Our current findings differ slightly from the previous work in that we did not see a significant augmentation of vasoconstriction during both rest and exercise conditions in the inhibited condition. This may be due to our use of propranolol in the current study to inhibit any potential PE-induced β-mediated vasodilatation, which has been shown to have a substantial NO-dependent component (Dawes et al. 1997). The current study is also in agreement with previous reports that combined inhibition of NO and PGs does not impair the magnitude of steady-state exercise hyperaemia attained during mild-to-moderate intensity rhythmic handgrip exercise (Fig.3A) (Dinenno & Joyner, 2004; Schrage et al. 2004; Crecelius et al. 2011b). The collective data suggest that vascular control during rhythmic exercise and particularly the interaction with sympathetic vasoconstriction extends beyond NO and PGs in the human forearm.

The majority of non-NO and non-PG mechanisms of vasodilatation, particularly those that are endothelium dependent, are thought to elicit hyperpolarization of endothelial and vascular smooth muscle cells, and there is significant interest in the role of electrical communication within and between both the endothelial and smooth muscle cells (Emerson & Segal, 2000). Limited animal studies have pursued whether or not hyperpolarizing mechanisms can modulate sympathetic vasoconstriction and therefore may be involved in functional sympatholysis. Investigations from Segal and colleagues have demonstrated that conducted vasodilatation can overcome sympathetically induced vasoconstriction, while at the same time sympathetic stimulation can attenuate conducted vasodilatation (Kurjiaka & Segal, 1995; Segal, 2000; Haug & Segal, 2005). This reciprocal relationship between conducted vasodilatation and sympathetic vasoconstriction is strikingly similar to vascular control observed in contracting human muscle (Joyner & Thomas, 2003). Taken together, these data suggest that conducted vasodilatation and possible hyperpolarization may be able to oppose sympathetically mediated depolarization and thus underlie functional sympatholysis.

We have recently demonstrated that inhibition of K_IR channels and Na⁺/K⁺-ATPase via BaCl₂ and ouabain, respectively, abolishes a mild hyperpolarizing stimulus (low dose KCl infusion) in the human forearm (Crecelius et al. 2012). Activation of either K_IR channels or Na⁺/K⁺-ATPase leads to vascular hyperpolarization and evidence suggests that K_IR activation is important in ‘amplifying’ the spread of hyperpolarization through the vasculature (Jantzi et al. 2006). Thus, given the previous evidence to suggest that hyperpolarization may be involved in functional sympatholysis and that activation of K_IR channels contributes to exercise hyperaemia in humans (Crecelius et al. 2014), we utilized our established pharmacological approach to test whether these pathways contributed to the ability of muscle contractions to blunt α₁-mediated vasoconstriction. Contrary to our hypothesis, in the presence of K_IR channels and Na⁺/K⁺-ATPase inhibition, alone or combined with inhibition of NO and PGs, vasoconstrictor responses during exercise were similar compared to control (uninhibited) conditions and still significantly attenuated compared to resting (adenosine) conditions (Fig.4A and B). Absolute changes in FVC reflect a similar result (Tables1 and 3). Interestingly, vasoconstriction during adenosine infusion, which served as a high-flow control condition, was significantly augmented with combined infusion of l-NMMA + ketorolac + BaCl₂ + ouabain as compared to control and l-NMMA + ketorolac conditions, and also with infusion of BaCl₂ + ouabain compared to control. We are not certain as to why this occurred, particularly to the large magnitude that we observed. It is possible, although speculative, that inhibition of K_IR channels limited any adenosine-mediated conducted vasodilatation (Rivers et al. 2001) and this may have contributed to the enhanced vasoconstriction. Regardless, the fact that vasoconstrictor responsiveness in resting tissue was increased but its attenuation during exercise was unaltered highlights the unique signalling that occurs during muscle contractions versus other hyperaemic conditions.

Although not studied extensively, few studies have tested whether K_ATP channels may be involved in functional sympatholysis, as these are metabolically sensitive channels and K⁺ efflux via these channels could also evoke hyperpolarization upon activation (Jackson, 2005). In the rat hindlimb, Thomas and colleagues demonstrated that pharmacological activation of K_ATP channels was able to blunt sympathetically stimulated vasoconstriction and inhibition of K_ATP channels during muscle contractions augmented sympathetic vasoconstriction (Thomas et al. 1997), although this latter response was not fully restored to that observed in quiescent muscle. In humans, inhibition of K_ATP channels with oral glyburide augmented carotid baroreflex-induced sympathetically mediated vasoconstriction during leg exercise (Keller et al. 2004); however, this also occurred at rest and thus functional sympatholysis per se may have been unaltered (i.e. muscle contractions still robustly inhibited the vasoconstriction). Interestingly, Thomas and Victor also demonstrated that K_ATP channel activation is downstream of NO (Thomas & Victor, 1998), and given findings in humans that a NO donor at rest and NOS inhibition during muscle contractions do not appear to impact sympathetic vasoconstriction (Dinenno & Joyner, 2003; Rosenmeier et al. 2003b), the minimal effect of K_ATP channel inhibition during exercise in humans is generally consistent with this.

Effects of inhibition of K_IR channel and Na⁺/K⁺-ATPase activity on blood parameters

In the present study, we sampled blood from a deep antecubital vein and the brachial artery catheter in order to measure a variety of potential muscle metabolites that may be contributing to our observed responses, as well as determine the oxygen consumption of the forearm tissue during exercise. These data (Table2) strengthen the findings of our haemodynamic measures and add additional insight to the metabolic consequences of our inhibitors. Rhythmic forearm contractions stimulated predictable changes reflective of increased metabolism and, as would be anticipated, in control and l-NMMA + ketorolac conditions PE-mediated vasoconstriction and reduced forearm blood flow was coupled with increased extraction of oxygen, and thus preserved skeletal muscle oxygen consumption.

Of interest, given the significant (∼40%) reduction in exercise hyperaemia and therefore oxygen delivery that occurred during l-NMMA + ketorolac + BaCl₂+ ouabain, oxygen extraction during exercise was significantly increased from the previous experimental conditions (control and l-NMMA + ketorolac). It appears that this was near maximal levels of oxygen extraction as there was no significant increase with PE infusion despite a reduction in blood flow. Thus, oxygen consumption in the condition of combined inhibition of NO, PGs, K_IR channels and Na⁺/K⁺-ATPase was attenuated during α₁-adrenergic vasoconstriction. This observation may have important relevance for populations that demonstrate impaired exercise hyperaemia and sympatholysis, such as older healthy adults and hypertensive humans (Dinenno et al. 2005; Kirby et al. 2009; Vongpatanasin et al. 2011). Given that all subjects were able to continue contractions, and thus muscle work was maintained, it is likely that non-oxidative pathways compensated for the loss in muscle ATP resynthesis via oxidative mechanisms, and this is reflected in significantly lower pH and greater values. Whether this shift in metabolism would impair larger muscle mass exercise tolerance over prolonged periods of time and during concomitant sympathetic activation remains unknown.

Experimental considerations

Sympathetic activation, as occurs during exercise, involves the release of several neurotransmitters, including NA, ATP and neuropeptide Y (von Kugelgen & Starke, 1985; Pernow, 1988). It is generally thought that NA is the primary neurotransmitter involved in mediating exercise-related sympathetic vasoconstriction (Buckwalter & Clifford, 1999; Richards et al. 2014). While some studies have used tyramine to stimulate the endogenous release of NA when studying sympatholysis in humans (Dinenno et al. 2002a; Tschakovsky et al. 2002; Rosenmeier et al. 2003a, 2004), this approach does not isolate post-junctional responsiveness; experimental control of this nature is possible with direct α-adrenoceptor stimulation. Importantly, muscle contractions blunt sympathetic vasoconstriction similarly when using tyramine or direct α₁- and α₂-agonists to stimulate α-adrenoceptors (Rosenmeier et al. 2003a; Kirby et al. 2008). In the current study we used a direct α₁-adrenergic agonist similar to previous studies by our group (Kirby et al. 2008, 2011) and others (Buckwalter et al. 1998; Rosenmeier et al. 2003a) as this can be given safely and repeatedly over the course of the experiments performed. Further, functional sympatholysis does not differ based on α-adrenoceptor subtypes (Rosenmeier et al. 2003a; Kirby et al. 2008). Thus, we do not anticipate that a lack of a role for K_IR channels and Na⁺/K⁺-ATPase is dependent on our current methodology of utilizing PE as an α₁-adrenoceptor agonist to stimulate vasoconstriction. In other words, we do not believe there would be a difference in our primary findings had we used α₂-adrenergic agonists or stimulated endogenous NA release through pharmacological (e.g. tyramine) or physiological (e.g. activation of the sympathetic nervous system) means.

Given that our primary findings regarding the role of K_IR channels and Na⁺/K⁺-ATPase in sympatholysis were contrary to our hypothesis, the efficacy of our pharmacological inhibition of these pathways could be questioned. We do not believe this is the case as we used doses of BaCl₂ + ouabain that had previously abolished vasodilatation to increasing doses of intra-arterial KCl (Crecelius et al. 2012). Further, in the present study, BaCl₂ + ouabain had a number of vasoactive effects: (1) they significantly augmented vasoconstriction to PE during adenosine infusion; (2) they significantly attenuated the hyperaemic and vasodilator response to forearm muscle contractions; and (3) they attenuated resting FBF in Protocol 2. However, it is important to acknowledge that our inhibition of K_IR channels and Na⁺/K⁺-ATPase blocks only K⁺-stimulated hyperpolarization which could be due to efflux from endothelial cells via calcium-activated potassium (K_Ca) channel activation or skeletal muscle cells during repolarization, and potentially the amplification of any hyperpolarizing stimulus (Jackson, 2005; Jantzi et al. 2006). In humans, it is not possible to fully prevent the hyperpolarization of endothelial or smooth muscle cells, nor the ability of electrical signals to spread among adjacent endothelial cells or underlying smooth muscle cells that could occur independently of activation of K_IR channels and Na⁺/K⁺-ATPase.

Further related to the efficacy of our pharmacological inhibition, we observed no effect of l-NMMA + ketorolac on vasoconstrictor responses during exercise and adenosine infusion. We did, however, observe a significant reduction in resting blood flow and vascular conductance as would be expected with sufficient inhibition of NO and PGs and used standard doses of l-NMMA and ketorolac which have previously been shown to be effective (Dinenno & Joyner, 2003; Schrage et al. 2004; Crecelius et al. 2011a). Taken together, we have used established pharmacology to directly test our hypothesis, and given the effects on various local haemodynamic parameters at rest and/or during adenosine infusions or exercise, lack of inhibitor effectiveness cannot explain the findings from the present study.

Perspectives

The impetus to investigate the role of K_IR channels (and Na⁺/K⁺-ATPase) in functional sympatholysis was based on our previous findings regarding the contributions of K_IR channels to ATP-mediated vasodilatation (∼50%) (Crecelius et al. 2012), the unique ability of ATP to blunt vasoconstriction to an extent similar to that occurring during exercise (Rosenmeier et al. 2004; Kirby et al. 2008), and the profound effect of K_IR channel inhibition on exercise hyperaemia (Crecelius et al. 2014). There is significant interest in intravascular ATP and its potential role in vascular control during exercise; however, this remains an unanswered question given limited available pharmacology to specifically inhibit purinergic receptors (Crecelius et al. 2015). Data from our laboratory demonstrate the ability of exogenous ATP to cause vasodilatation and modulate α-adrenergic vasoconstriction (Kirby et al. 2008), and K_IR channel activation explains ∼50% of ATP-mediated vasodilatation, whereas NO and PGs have only a modest role in this vasodilatory response (Crecelius et al. 2011a, 2012). In the present study, the same combination of pharmacological inhibition that greatly reduces ATP-mediated vasodilatation (l-NMMA + ketorolac + BaCl₂ + Na⁺/K⁺-ATPase) does not augment α₁-mediated vasoconstriction during exercise. Taken together, these data could be interpreted to suggest ATP plays a limited role in the modulation of sympathetic vasoconstriction during exercise. Alternatively, we propose it is possible that the remaining ‘unexplained’ vascular signalling via ATP may be what allows exercise to blunt sympathetic vasoconstriction.

What then are these remaining signalling pathways that could explain our observed responses? We propose that electrical communication (Segal, 2000), probably through activation of endothelial small- and intermediate-calcium-activated potassium channels (sK_Ca and iK_Ca) and direct transfer of electrical charge through the physical connections of endothelial and vascular smooth muscle cells (Emerson & Segal, 2000), causes hyperpolarization evoking a conducted vasomotor response of the vasculature which results in both profound vasodilatation as well as opposition to sympathetically induced depolarization (i.e. vasoconstriction) (Kurjiaka & Segal, 1995). Interestingly, ATP is capable of evoking conducted vasodilatation via activating sK_Ca and iK_Ca channels (Winter & Dora, 2007). Additionally, slight augmentations in endothelial signalling, via infusion of acetylcholine or ATP, significantly enhances the ability of mild handgrip exercise to blunt sympathetic vasoconstriction and this occurs independently of NO and PG signalling (Kirby et al. 2013b; Hearon et al. 2014). Thus, it is possible that initiation of endothelial cell hyperpolarization is the key event that spreads to adjacent endothelial and underlying smooth muscle cells and ultimately limits sympathetic vasoconstriction in active muscle. During exercise, ATP is released from intravascular sources (e.g. red blood cells) (Kirby et al. 2013a) and might provide this stimulus for endothelial cell hyperpolarization. Further development of specific pharmacological inhibitors that can attenuate this cell-to-cell communication and are safe for human administration is necessary before this hypothesis can be directly tested.

Conclusions

Muscle contractions stimulate a large increase in blood flow and can also blunt sympathetically mediated vasoconstriction relative to what occurs in resting tissue. This functional sympatholysis is thought to preserve oxygen delivery to active tissue in the face of the increased sympathetic activity that is necessary for appropriate blood pressure regulation.

In the present investigation, we show that acute pharmacological inhibition of NO and PGs does not significantly reduce exercise hyperaemia nor impact α₁-mediated vasoconstriction during handgrip exercise. While inhibition of K_IR channels and Na⁺/K⁺-ATPase reduces exercise hyperaemia by ∼30–40% it does not impact the normal ability of muscle contractions to blunt α₁-adrenergic vasoconstriction. The combined impact of attenuated hyperaemia and preserved vasoconstriction during exercise is metabolically significant and results in attenuated oxygen consumption and increased non-oxidative metabolism in the contracting tissue. Our results indicate that blunting of α₁-adrenergic vasoconstriction during rhythmic forearm exercise is not governed by the activity of K_IR channels and Na⁺/K⁺-ATPase in humans.

Glossary

ADO

adenosine

arterial–venous O₂ content difference

FAV

forearm volume

FBF

forearm blood flow

fraction of oxygenated haemoglobin

FVC

forearm vascular conductance

HR

heart rate

K_ATP

ATP-sensitive potassium channels

K_Ca

calcium-activated potassium channels

K_IR

inwardly rectifying potassium channel

l-NMMA

N^G-monomethyl-l-arginine

MAP

mean arterial pressure

MBV

mean blood velocity

MVC

maximal voluntary contraction

NA

noradrenaline

NO

nitric oxide

partial pressure of carbon dioxide

PE

phenylephrine

PG

prostaglandin

partial pressure of oxygen

oxygen consumption

Additional information

Author contributions

A.R.C. contributed to the design of the experiment, collection, analysis and interpretation of the data, and the writing of this article. B.S.K. contributed to the design of the experiment, interpretation of the data and critical revision of this article. C.M.H. Jr contributed to experimental design, collection and analysis of the data, and critical revision of this article. G.J.L and D.G.L. contributed to the experimental design, provided invasive methodology for data collection and contributed to critical revision of this article. F.A.D. contributed to the conception and design of the experiment, collection, analysis and interpretation of the data and the writing of this article. All authors gave final approval of the article. All experiments were performed in the Human Cardiovascular Physiology Laboratory at Colorado State University.