Cutaneous vascular and sweating responses to intradermal ...

14 downloads 60531 Views 537KB Size Report
Mar 23, 2015 - Email: [email protected]. Abbreviations ... The heat loss responses of cutaneous active vasodilatation and sweating are important to maintain a .... disorders, hypertension, heart disease, diabetes, auto- nomic disorders and ...
2515

J Physiol 593.11 (2015) pp 2515–2525

Cutaneous vascular and sweating responses to intradermal administration of ATP: a role for nitric oxide synthase and cyclooxygenase? Naoto Fujii1 , Ryan McGinn1 , Lyra Halili1 , Maya Sarah Singh1 , Narihiko Kondo2 and Glen P. Kenny1 1 2

Human and Environmental Physiology Research Unit, School of Human Kinetics, University of Ottawa, Ottawa, Canada Faculty of Human Development, Kobe University, Kobe, Japan

Key points

r In humans in vivo, the mechanisms behind ATP-mediated cutaneous vasodilatation along with

The Journal of Physiology

whether and how ATP increases sweating remains uncertain.

r Recent work has implicated nitric oxide synthase (NOS), cyclooxygenase (COX) and/or r r r

adenosine in the modulation of cutaneous vasodilatation and sweat production during both local (i.e. localized heating) and whole-body heat stress (i.e. exercise-induced heat stress). We evaluated whether ATP-mediated cutaneous vasodilatation and sweating is mediated via NOS, COX and/or adenosine. We show that in humans in vivo, intradermal administration of ATP induces pronounced vasodilatation which is partially mediated by NOS, but neither COX nor adenosine influences ATP-mediated vasodilatation, and ATP alone does not induce an increase in sweating. These findings advance our basic physiological knowledge regarding control of skin blood flow and sweating, and provide insight into the mechanisms governing thermoeffector activity, which has major implications for whole-body heat exchange and therefore core temperature regulation in humans during heat stress.

Abstract In humans in vivo, the mechanisms behind ATP-mediated cutaneous vasodilatation and whether and how ATP increases sweating remain uncertain. We evaluated whether ATP-mediated cutaneous vasodilatation and sweating is mediated via nitric oxide synthase (NOS), cyclooxygenase (COX) and/or adenosine-dependent mechanisms. Cutaneous vascular conductance (CVC, laser Doppler perfusion units/mean arterial pressure) and sweat rate (ventilated capsule) were evaluated at intradermal microdialysis forearm skin sites, each receiving pharmacological agents (two separate protocols). In Protocol 1 (n = 12), sites were perfused with: (1) lactated Ringer solution (Control), (2) 10 mM Nω -nitro-L-arginine (L-NNA, a NOS inhibitor), (3) 10 mM ketorolac (Ketorolac, a COX inhibitor) or (4) a combination of 10 mM L-NNA + 10 mM ketorolac (L-NNA + Ketorolac). In Protocol 2 (n = 8), sites were perfused with: (1) lactated Ringer solution (Control) or (2) 4 mM theophylline (Theophylline, an adenosine receptor inhibitor). At all sites, ATP was simultaneously perfused at 0.12, 1.2, 12, 120 and 1200 nM min−1 (each for 20 min). Relative to CVC at the Control site with ATP infused at 120 nM min−1 (71 ± 9% of max CVC), CVC at the Ketorolac site was comparable (64 ± 13% of max CVC, P = 0.407), but lower at L-NNA (51 ± 15% of max CVC, P = 0.040) and L-NNA + Ketorolac (51 ± 13% of max CVC, P = 0.049) sites. Conversely, across the four skin sites at any other ATP infusion rate (all P > 0.174), no differences in CVC were observed. Theophylline did not influence CVC at any ATP infusion rate (all P > 0.234). Furthermore, no ATP infusion rate elicited an increase in sweating from baseline at any skin site (all P > 0.235). We show that NOS, but neither COX nor adenosine receptors, modulates ATP-mediated cutaneous vasodilatation, whereas ATP does not directly increase sweating.

 C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

DOI: 10.1113/JP270147

2516

N. Fujii and others

J Physiol 593.11

(Received 8 January 2015; accepted after revision 13 March 2015; first published online 23 March 2015) Corresponding author Glen P. Kenny: University of Ottawa, School of Human Kinetics, 125 University, Room 367, Montpetit Hall, Ottawa, Ontario, Canada, K1N 6N5. Email: [email protected] Abbreviations ACh, acetylcholine; COX, cyclooxygenase; CVC, cutaneous vascular conductance; EDHF, endothelium-derived hyperpolarizing factor; KCa, calcium-activated potassium; L-NNA, NG -nitro-L-arginine; NO, nitric oxide; NOS, nitric oxide synthase; SNP, sodium nitroprusside.

Introduction The heat loss responses of cutaneous active vasodilatation and sweating are important to maintain a stable core body temperature during heat stress in humans. Although some of the mechanisms underlying the heat loss responses have been elucidated during passive heating at rest (McCord et al. 2006; Kellogg et al. 2008; Machado-Moreira et al. 2012; Wong & Fieger, 2012) and during exercise in the heat (Welch et al. 2009; Buono et al. 2011; Fujii et al. 2014c; McGinn et al. 2014b; McNamara et al. 2014), the precise pathways involved remain equivocal. ATP may be involved in the control of cutaneous active vasodilatation and sweating during heat stress given that hyperthermia has been associated with endogenous increases in plasma ATP (Pearson et al. 2011; Kalsi & Gonzalez-Alonso, 2012). Furthermore, sympathetic cholinergic nerves are responsible for the heat loss responses (Kellogg et al. 1995; Machado-Moreira et al. 2012) and are known to co-release ATP with acetylcholine (ACh) (Rabasseda et al. 1987). However, whether and how increases in ATP can modulate thermoeffector organs (i.e. skin vessels and eccrine sweat glands) remain to be elucidated. Intradermal administration of ATP can directly induce cutaneous vasodilatation in humans in vivo (Wingo et al. 2010a), although the precise pathways have not been investigated. Previous studies in humans in vivo have shown a role for nitric oxide (NO) synthase (NOS) and/or cyclooxygenase (COX) in the regulation of cutaneous vasodilatation (Holowatz et al. 2005; Kellogg et al. 2005; McCord et al. 2006; Medow et al. 2008; Welch et al. 2009; Fujii et al. 2013, 2014c; McGinn et al. 2014a,b; McNamara et al. 2014). Furthermore, note that ATP is rapidly catabolized to adenosine by ectonucleotidases in humans in vivo (Zimmermann, 1996), which has recently been associated with cutaneous vascular regulation (Fieger & Wong, 2010; McGinn et al. 2014a,b; Swift et al. 2014). Collectively, it is unclear whether NOS, COX and/or adenosine receptors contribute to ATP-mediated cutaneous vasodilatation in humans in vivo. The presence of P2Y purinergic receptors has been confirmed in the human sweat gland (Lindsay, 2002). In fact, ATP has been shown to increase sweating in vitro (Sato et al. 1991); however, this has not been investigated in humans in vivo. While the response in vitro may indicate possible modulators, the sweating response observed is not always congruent with that reported in vivo,

warranting investigation. Furthermore, we have recently demonstrated that both NOS and COX exhibit a role in modulating sweating during exercise in the heat (Fujii et al. 2014c), albeit not during methacholine-induced sweating (Fujii et al. 2014b). Thus, NOS- and/or COX-dependent sweating may be mediated through ACh-independent mechanisms during exercise in the heat (Fujii et al. 2014b). Given that cholinergic nerves co-release ATP, as mentioned above, it is plausible to suggest that ATP can directly increase sweat production through NOS- and/or COX-dependent mechanisms in humans in vivo. Thus, the purpose of the present study was to evaluate the separate and combined roles of NOS and COX in the cutaneous vascular and sweating responses to increasing levels of intradermal ATP in humans in vivo. We also aimed to elucidate whether the cutaneous vascular responses associated with ATP administration were mediated by adenosine or exclusively by ATP. Our hypotheses were three-fold: (1) both NOS and COX would contribute to ATP-mediated cutaneous vasodilatation, (2) ATP-mediated cutaneous vasodilatation would be mediated through ATP and not by adenosine, and (3) ATP administration would increase sweating, which would be mediated by both NOS- and COX-dependent mechanisms. Methods Ethical approval

The University of Ottawa Health Sciences and Science Research Ethics Board approved the current study, which conforms to the guidelines set out by the Declaration of Helsinki. Verbal and written informed consent was obtained from all volunteers prior to their participation in the study. Subjects

Twenty-seven, habitually active (2–5 days per week, 30 min of exercise per day), young adults (15 males and 12 females) participated in at least one of the experimental protocols for this study (three protocols in total). In Protocol 1 (7 males and 5 females), mean (± SD) body mass, height and age for the participants were: 75.0 ± 16.8 kg, 1.73 ± 0.10 m and 22 ± 3 years, respectively. In Protocol  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

J Physiol 593.11

ATP and thermoeffector activity

2 (5 males and 4 females), the respective values were: 68.9 ± 14.2 kg, 1.67 ± 0.10 m and 24 ± 4 years. In Protocol 3 (4 males and 4 females), mean (± SD) body mass, height and age were: 69.0 ± 10.8 kg, 1.71 ± 0.07 m and 23 ± 5 years, respectively. The exclusion criteria for participating in this study were having a history of cystic fibrosis transmembrane conductance regulator mutations, skin disorders, hypertension, heart disease, diabetes, autonomic disorders and smoking. All subjects were currently not taking prescription medications with the exception of contraceptives. Eight of the 12 females were using a contraceptive (2 females were using an intrauterine device and 6 were using oral contraceptives). All female subjects participated in the experimental session during the early follicular phase (within 6 days of starting menstruation) or during the placebo phase if using contraceptives. This was necessary to minimize the reported effects of female sex hormones on the regulation of cutaneous vasodilatation and sweating (Kuwahara et al. 2005; Brunt et al. 2011). Experimental design

All subjects were instructed to abstain from taking over-the-counter medications (including non-steroidal anti-inflammatory agents and vitamins) for at least 48 h before the study, as well as alcohol and caffeine consumption at least 12 h before the study. Subjects did not perform heavy exercise the day prior to the study. On the day of the study, subjects did not consume any food 2 h before and throughout the study. Upon arrival, the subjects provided a urine sample (only Protocols 1 and 3) and voided their bladder, after which a measurement of body mass was taken using a digital weight scale platform (Model CBU150X, Mettler Toledo Inc., Columbus, OH, USA) with a weighing terminal (Model IND560, Mettler Toledo). Thereafter, the subject’s body height was measured using an eye-level physician stadiometer (Model 2391, Detecto Scale Company, Webb City, MO, USA). Subjects were then seated in a semi-recumbent position in a thermoneutral room (26 °C) and instrumented with four (Protocol 1) or two (Protocols 2 and 3) intradermal microdialysis fibres (30 kDa cutoff, 10 mm membrane) (MD2000, Bioanalytical Systems, West Lafayette, IN, USA) on the dorsal side of the left forearm within the dermal layer of the skin. Using an aseptic technique, a 25 gauge needle was first inserted into the unanaesthetized skin. We did not employ ice or local anaesthetic cream as previously used (Hodges et al. 2009). The entry and exit points were 2.5 cm apart. A microdialysis fibre was then threaded through the lumen of the needle, after which the needle was withdrawn leaving only the fibre in place. Microdialysis fibres were secured with surgical tape and each fibre was separated by at least 4.0 cm.  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

2517

Protocol 1

This protocol was designed to elucidate whether NOS and COX are involved in ATP-mediated cutaneous vasodilatation and/or sweating. Approximately 20 min after placement of the microdialysis fibres, perfusion of pharmacological agents began via microdialysis. Fibres were assigned in a counterbalanced manner to receive (1) lactated Ringer solution (Control; Baxter, Deerfield, IL, USA), (2) 10 mM NG -nitro-L-arginine (L-NNA; Sigma-Aldrich, St Louis, MO, USA) to non-selectively inhibit NOS and thus NO production, (3) 10 mM ketorolac, a non-selective COX inhibitor (Ketorolac; Sigma-Aldrich), or (4) a combination of 10 mM L-NNA and 10 mM ketorolac (L-NNA + Ketorolac). These concentrations were determined from previous studies in which intradermal microdialysis was employed in human skin for L-NNA (Medow et al. 2008; Fujii et al. 2014a) and ketorolac (Holowatz et al. 2005; Kellogg et al. 2005; McCord et al. 2006; Medow et al. 2008; Fujii et al. 2013). Each drug was continuously perfused at a rate of 4.0 μl min−1 using a micro infusion pump (Model 4004, CMA Microdialysis, Solna, Sweden) for at least 75 min to ensure the establishment of each blockade (Holowatz et al. 2005; Fujii et al. 2013, 2014b). Perfusion continued for the entire experimental protocol until the procedure for maximal cutaneous vasodilatation began (see below). Following the 75 min period of perfusion, 10 min of baseline data were collected. Subsequently, each microdialysis fibre was perfused with ATP (Cayman Chemical, Ann Arbor, MI, USA) in an incremental manner with five different concentrations of 0.03, 0.3, 3, 30 and 300 mM ATP at a rate of 4.0 μl min−1 (five infusion rates: 0.12, 1.2, 12, 120 and 1200 nM min−1 , each for 20 min) in combination with the site-specific pharmacological agents. These infusion rates were based on extensive pilot work and a previous report showing that 30 mM ATP can increase cutaneous vascular conductance (CVC) up to 55% of maximum vascular conductance (%max) (Wingo et al. 2010a). Based on this study, we developed our protocol to obtain CVC values below and above 55%max. In our pilot work, we confirmed that ATP at 1200 nM min−1 can yield maximal vasodilatation in response to ATP administration as a higher rate of 2400 nM min−1 showed no further increases in CVC. After completing the last infusion of ATP (1200 nM min−1 ), 50 mM sodium nitroprusside (SNP; Sigma-Aldrich) administration was initiated at each microdialysis site at a rate of 6.0 μl min−1 . Our pilot work indicated that 50 mM SNP induces maximal cutaneous vasodilatation to a similar extent as would be observed during local heating to 44 °C. The SNP administration lasted for approximately 25–30 min until a stable plateau for a minimum of 2 min was observed. At this point, blood pressure was measured to quantify maximal CVC.

2518

N. Fujii and others

In three subjects, maximal CVC was obtained during the last infusion rate of ATP (1200 nM min−1 ). Protocol 2

This protocol was employed as a sub-study to verify whether theophylline, known as a competitive non-specific adenosine receptor blocker, can effectively inhibit adenosine-mediated cutaneous vasodilatation. This protocol was used to validate our study design for Protocol 3 (see below). After placement of the two microdialysis fibres (see above), both were perfused with lactated Ringer solution (Baxter) at a rate of 4.0 μl min−1 via a micro perfusion pump (model 400, CMA Microdialysis) for 60–90 min to allow for the resolution of the local hyperemic response. At this point, one microdialysis fibre was perfused with 4 mM theophylline (Theophylline; Sigma-Aldrich) while the other site continued to be perfused with lactated Ringer solution (Control). The concentration of theophylline was determined based on previous studies in human skin (Fieger & Wong, 2012; McGinn et al. 2014a,b). Drug perfusion was maintained for 45 min to ensure establishment of the blockade (Fieger & Wong, 2012; McGinn et al. 2014a,b). At the end of the 45 min blockade period, 15 min of baseline resting data were collected. Subsequently, 5.6 mM adenosine (Sigma-Aldrich) was administered at a rate of 4.0 μl min−1 (22.4 nM min−1 ) through the Control (i.e. Adenosine) and Theophylline (i.e. Adenosine + Theophylline) sites. The concentration of adenosine was chosen based on previous work (Shibasaki et al. 2007) and its perfusion continued for 45 min, which was sufficient to induce a stable plateau in cutaneous blood flow for at least 2 min in each subject. At the end of the adenosine perfusion, we evaluated maximal CVC. SNP was administered (50 mM at 6.0 μl min−1 ) while the local skin heaters were set to 42 °C for 20 min and to 44 °C for an additional 25 min. This protocol was sufficient to induce a stable plateau in cutaneous blood flow for at least 2 min in each subject. Subsequently, blood pressure was measured via manual auscultation for the assessment of maximal CVC. Protocol 3

This protocol evaluated whether adenosine is involved in ATP-mediated cutaneous vasodilatation. Approximately 20 min following the placement of the two microdialysis fibres (see above), both sites continuously received either (1) lactated Ringer solution (Control) or (2) 4 mM theophylline (Theophylline). Drug perfusion was maintained for 60 min to ensure establishment of the blockade. Thereafter, ATP perfusion commenced in the same way as was described for Protocol 1. Following ATP perfusion, maximal CVC was measured as described in

J Physiol 593.11

Protocol 1. Maximal CVC was obtained for three subjects during administration of 1200 nM min−1 of ATP. Measurements

Cutaneous red blood cell flux (expressed in perfusion units) was locally measured as an index of cutaneous blood flow at a sampling rate of 32 Hz with laser Doppler flowmetry (PeriFlux System 5000, Perimed, Stockholm, Sweden). In Protocols 1 and 3, integrated laser Doppler flowmetry probes with a seven-laser array (Model 413, Perimed) were housed in the centre of each sweat capsule over each microdialysis fibre for simultaneous measurement of both local forearm sweat rate and cutaneous red blood cell flux at each skin site. In Protocols 1 and 3, manual auscultation was performed using a validated mercury column sphygmomanometer (Baumanometer Standby Model, WA Baum Co., Copiague, NY, USA) to obtain blood pressures every 5 min. For Protocol 2, we did not measure sweat rate such that the integrated laser Doppler flowmetry probes were housed in local heating units. Furthermore, blood pressure was assessed using an automated blood pressure monitor (Tango+, SunTech Medical Inc., Morrisville, NC, USA) and verified by an auditory inspection every 15 min throughout the experimental protocol. For all three protocols, mean arterial pressure was calculated from diastolic arterial pressure plus one-third the difference between systolic and diastolic pressures (i.e. pulse pressure). CVC was evaluated as cutaneous red blood cell flux divided by mean arterial pressure. CVC data are presented as %max. Expressing CVC data as %max minimizes the effect of site-to-site heterogeneity at the level of cutaneous blood flow (Minson, 2010). Sweat capsules each covered an area of 3.8 cm2 during Protocols 1 and 3 and were placed directly over the centre of each microdialysis membrane. The sweat capsules were secured to the skin with adhesive rings and topical skin glue (Collodion HV, Mavidon Medical products, Lake Worth, FL, USA). Dry compressed air in a gas tank located in the thermoneutral room (26 °C) was supplied to each capsule at a rate of 0.5 l min−1 , while water content of the effluent air from the sweat capsule was measured with high precision dew point mirrors (Model 473, RH systems, Albuquerque, NM, USA). Long vinyl tubes were used for connections between the gas tank and the sweat capsule, and between the sweat capsule and the dew point mirror. This ensured that the internal gas temperature was equilibrated to near room temperature (26 °C) before reaching the sweat capsule (inlet) and the dew point mirror (outlet). Local forearm sweat rate was calculated every 5 s from the difference in water content between influent and effluent air multiplied by the flow rate and normalized for the skin surface area under the capsule (mg min−1 cm−2 ).  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

J Physiol 593.11

ATP and thermoeffector activity

For Protocols 1 and 3, urine-specific gravity was assessed from the urine samples obtained prior to starting the experimental protocol using a handheld total solids refractometer (Model TS400, Reichter Inc., Depew, NY, USA). Urine-specific gravity prior to the experiment was 1.014 ± 0.005 and 1.017 ± 0.007 in Protocols 1 and 3, respectively, indicating the subjects were similarly euhydrated before each protocol. Data analysis

Baseline values used for data analysis were obtained by averaging measurements made over the last 5 min of the 10–15 min baseline period in all protocols. For Protocols 1 and 3, 1 min of peak CVC at each ATP infusion rate was used for data analysis. Using the peak value was necessary because we occasionally observed a gradual decrease in CVC after the peak response. For Protocols 1 and 3, the averaged sweat rate data over the last 1 min of each infusion rate of ATP were used for analysis. For Protocol 2, CVC values were obtained from an average of the final minute at each 15 min interval throughout the adenosine perfusion. In all three protocols, maximal CVC was obtained during SNP administration (with the exception of three subjects in Protocols 1 and 3 who exhibited maximal CVC during ATP administration at 1200 nM min−1 ). To obtain insight into potential sex differences, we pooled data from Protocols 1 and 3 to compare CVC responses to ATP administration evaluated at the Control site between 9 males and 9 females. Subjects who completed Protocols 1 and 3 were included only once for the between-sex comparisons. Statistical analysis

In Protocols 1 and 3, CVC and sweat rate were analysed using a two-way repeated-measures ANOVA. The two factors were: ATP infusion rate (six levels: baseline, ATP infusion rate at 0.12, 1.2, 12, 120 and 1200 nM min−1 ) and treatment site (four levels in Protocol 1: Control, L-NNA, Ketorolac, L-NNA + Ketorolac; two levels in Protocol 3: Control, Theophylline). To clarify a potential sex difference, we further analysed CVC and sweat rate at the Control site using a two-way, mixed-model, repeated-measures ANOVA with the factor of sex (two levels: male, female) and of ATP infusion rate (six levels). In Protocol 2, CVC was analysed using a two-way repeated-measures ANOVA with the factor of time (four levels: baseline, 15, 30 and 45 min of adenosine administration at 22.4 nM min−1 ) and of treatment site (two levels: Adenosine, Adenosine + Theophylline). In Protocol 1, maximal absolute CVC (perfusion units mmHg−1 ) was analysed with a one-way repeated-measures ANOVA with a single factor of treatment site (four levels: Control, L-NNA, Ketorolac,  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

2519

+ Ketorolac). When a significant interaction or main effect was detected, post hoc comparisons were carried out using Student’s paired t-tests corrected for multiple comparisons using the Holm–Bonferroni procedure. Student’s paired t-tests were also used to compare maximal absolute CVC between the Adenosine and Adenosine + Theophylline sites in Protocol 2, between the Control and Theophylline sites in Protocol 3, and between males and females. For the between-sex comparison, as a post hoc test, we employed Student’s unpaired t-tests. The level of significance for all analyses was set at P  0.05. All data used for parametric statistical analyses were normally distributed. All values are reported with a mean ± 95% confidence interval. The confidence intervals at 95% were calculated as 1.96 × SEM.

L-NNA

Results Cutaneous vascular response Mechanisms of ATP-mediated cutaneous vasodilatation.

Cutaneous blood flow response during incremental ATP perfusion in a representative subject is depicted in Fig. 1. There was a main effect of ATP infusion rate (P < 0.001) for CVC in Protocol 1 as an increase from baseline was noted at 120 and 1200 nM min−1 of ATP irrespective of treatment sites (all P < 0.018). We also found a main effect of treatment site (P = 0.011) for CVC. At 120 nM min−1 of ATP, L-NNA and L-NNA + Ketorolac reduced CVC relative to the Control, but Ketorolac did not (Fig. 2). CVC did not differ across the treatment sites at baseline or at 0.12, 1.2 and 12 nM min−1 of ATP (all P > 0.174, Fig. 2). A main effect of time and treatment site and an interaction of time and treatment site for CVC in Protocol 2 were all significant (all P < 0.001). Although CVC at baseline was similar between the treatment sites (P = 0.294), CVC values at 15, 30 and 45 min of adenosine administration at 22.4 nM min−1 were all lower at the Adenosine + Theophylline site relative to the Adenosine site (Fig. 3, all P < 0.008). Furthermore, an interaction of ATP infusion rate and treatment site was measured for CVC in Protocol 3 (P = 0.009). However, CVC was similar between the Control and Theophylline site at baseline and at all infusion rates of ATP (Fig. 4, all P > 0.094). Mean arterial pressure at baseline (Protocol 1: 90 ± 4 mmHg; Protocol 2: 88 ± 3 mmHg; Protocol 3: 83 ± 2 mmHg) did not change throughout any of the protocols over time. Sex difference. We detected a main effect of sex for CVC measured at the Control site during ATP administration (P = 0.024). Specifically, females exhibited a higher CVC at 120 nM min−1 (78 ± 9%max, P = 0.025) and 1200 nM min−1 (95 ± 6%max, P = 0.007) of ATP relative to their male counterparts (60 ± 11 and

2520

N. Fujii and others

75 ± 11%max, respectively). By contrast, there was no difference in CVC between males and females at baseline (18 ± 8 vs. 23 ± 8%max, P = 0.400) or during ATP administration at 0.12 nM min−1 (23 ± 10 vs. 25 ± 10%max, P = 0.696), 1.2 nM min−1 (24 ± 12 vs. 26 ± 6%max, P = 0.744) and 12 nM min−1 (27 ± 9 vs. 36 ± 9%max, P = 0.193). Maximal absolute CVC. In Protocol 1, there was no

main effect of treatment site for maximal absolute CVC (P = 0.916) such that maximal CVC was comparable between the four sites (perfusion units mmHg−1 : Control, 1.60 ± 0.29; L-NNA, 1.58 ± 0.28; Ketorolac, 1.71 ± 0.27; L-NNA + Ketorolac, 1.62 + 0.26). Likewise, maximal absolute CVC did not differ between the Adenosine and Adenosine + Theophylline sites in Protocol 2 (2.77 ± 0.33 vs. 2.81 ± 0.23 perfusion units mmHg−1 , P = 0.747) or between the Control and Theophylline sites in Protocol 3 (1.85 ± 0.36 vs. 1.66 ± 0.41 perfusion units mmHg−1 , P = 0.331). Moreover, maximal absolute CVC at the Control site did not differ between males and females (1.44 ± 0.29 vs. 1.86 ± 0.34 perfusion units mmHg−1 , P = 0.087).

Sweat rate

In both Protocols 1 and 3, we found a main effect of ATP infusion rate for sweat rate (both P < 0.002). However, after post hoc analysis and correcting for multiple comparisons, sweat rate did not increase significantly from baseline at any of the skin sites (all P > 0.235, Fig. 5 for Protocol 1). No main effect of treatment site was measured for sweat rate in Protocols 1 and 3 (both P > 0.329). Therefore, sweat rate did not differ between the skin sites at baseline or at any infusion rate of ATP (Fig. 5 for Protocol 1). Furthermore, there was no main effect of sex for sweat rate (P = 0.515). Thus, there was no difference in sweat 1200 nM min–1 ATP

Index of cutaneous blood flow (perfusion units)

200

J Physiol 593.11

rate between males and females at baseline (0.05 ± 0.01 vs. 0.05 ± 0.01 mg min−1 cm−2 ) or at any infusion rate of ATP (1200 nM min−1 : 0.06 ± 0.02 vs. 0.06 ± 0.03 mg min−1 cm−2 ).

Discussion The present study examined the roles of NOS and COX in cutaneous vascular and sweating responses to ATP administered in a rate-dependent manner from 0.12 to 1200 nM min−1 in humans in vivo. We found that ATP infused at 120 and 1200 nM min−1 caused a robust and pronounced cutaneous vasodilatation. Inconsistent with our hypothesis, the cutaneous vasodilatation evoked by ATP was not affected by an inhibition of COX; however, as per our hypothesis, it was partially diminished by an inhibition of NOS with an infusion of ATP at a rate of 120 nM min−1 . We also demonstrated that cutaneous vasodilatation induced by the infusion of ATP at a rate of 120 and 1200 nM min−1 is not modulated by adenosine receptor activation. In contrast to our hypothesis, no infusion rates of ATP resulted in any increase in sweat rate. Collectively, we show that NOS, but not COX or adenosine, can partly explain ATP-mediated cutaneous vasodilatation, and that ATP is not a direct activator that can mediate sweat secretion in humans in vivo.

NOS and COX in ATP-mediated cutaneous vasodilatation

In the present study, ATP infused at low rates (0.12, 1.2 and 12 nM min−1 ) did not increase CVC at the Control site. However, a high rate (120 nM min−1 ) induced a pronounced increase in CVC, reaching 71%max, which is in accordance with a previous study (Wingo et al. Maximal vasodilation with SNP

120 nM min–1 ATP

150

100

12 nM min–1 ATP 1.2 nM min–1 ATP

50

0.12 nM min–1 ATP

0 0

20

40

60

80

Time (min)

100

120

140

Figure 1. Time-course changes of cutaneous blood flow Time-course changes of cutaneous blood flow from a representative individual at the lactated Ringer solutrion site (Control) during ATP and sodium nitroprusside (SNP) infusions. Arrows indicate each infusion period.

 C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

ATP and thermoeffector activity

J Physiol 593.11

2010a). This response pattern may be explained by the rapid break down and clearance of lower concentrations of ATP by ectonucleotidases (Zimmermann, 1996), although this effect may become saturated by higher infusion rates, ultimately allowing ATP to enter the intravascular space and cause vasodilatation via P2 receptors on endothelial cells. Importantly, the increase in CVC in our study was partially attenuated by L-NNA and L-NNA + Ketorolac (both sites by 20%max) (Fig. 2). Thus, ATP can induce cutaneous vasodilatation at higher concentrations, and NOS activity can modulate the response. It is unclear from the present data how NOS is involved in the ATP-mediated cutaneous vasodilatation. ATP can increase calcium influx into endothelium cells (Yamamoto et al. 2000), which can activate NOS (Busse & Mulsch, 1990). Hence it is plausible that increased intracellular calcium levels during ATP administration invoke NOS-dependent cutaneous vasodilatation. On the other hand, most of the subjects in this study reported slight discomfort (i.e. stinging sensation) during administration of ATP at the higher infusion rates of 120 and 1200 nM min−1 , which may indicate that greater elevation in interstitial ATP activates sensory nerves. Indeed, P2 receptors are known to exist on cutaneous sensory neurons (Aoki et al. 2003) and sensory nerve-mediated cutaneous

Control 80

vasodilatation has a component that is NOS-dependent (Houghton et al. 2006). It is noteworthy that the contribution of NOS to cutaneous vasodilatation associated with an ATP infusion rate of 120 nM−1 was diminished at the highest rate of ATP infusion (i.e. 1200 nM min−1 ) (Fig. 2). Consequently, our results indicate a modulation of cutaneous vasodilatation that is dependent upon the concentration of the agonist. This effect has previously been shown as the contribution of NOS to ACh-induced cutaneous vasodilatation decreases as the concentration of ACh increases (Medow et al. 2008) and this may extend to ATP-mediated cutaneous vasodilatation. Based on this information, it is plausible that an overwhelming activation of one signalling pathway (i.e. ATP signalling) causes the recruitment of several redundant vasodilator mechanisms, which can serve to compensate for the diminished NOS component induced by L-NNA, thereby allowing for a sustained level of vasodilatation. Despite the dual inhibition of NOS and COX in combination, most of the ATP-mediated increases in cutaneous blood flow (60% of the increase from baseline) remained intact (Fig. 2). Thus, ATP-mediated cutaneous vasodilatation probably involves mechanisms beyond NOS- and COX-dependent pathways, such as endothelium-derived hyperpolarizing factors (EDHFs). Calcium-activated potassium (KCa) channels are the key component of EDHF-mediated vasodilatation and are involved in cutaneous vascular regulation in humans in

L-NNA

Ketorolac

80

L-NNA + Ketorolac

60

* *

40

20

0 Baseline

0.12

1.2 ATP (nM

12

120

1200

min –1 )

Figure 2. Cutaneous vascular conductance at baseline and during ATP administration Cutaneous vascular conductance at baseline and during ATP administration from 0.12 to 1200 nM min−1 (five levels) at four intradermal microdialysis skin sites with (1) lactated Ringer solution (Control, open circles), (2) 10 mM NG-nitro-L-arginine (L-NNA, squares), a non-specific nitric oxide synthase inhibitor, (3) 10 mM ketorolac (Ketorolac, triangles), a non-selective cyclooxygenase inhibitor, or (4) a combination of 10 mM L-NNA + 10 mM ketorolac (L-NNA + Ketorolac, diamonds). Values are means ± 95% confidential interval (n = 12). ∗ Significantly different vs. Control site (P < 0.05).  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

Cutaneous vascular conductance (%max)

Cutaneous vascular conductance (%max)

100

2521

Adenosine Adenosine + Theophylline

60

40

*

*

*

20

0 Baseline

15

30

45

Time (min)

Figure 3. Cutaneous vascular conductance at baseline and during adenosine administration Cutaneous vascular conductance at two skin sites continuously receiving (1) adenosine at 22.4 nM min−1 (Adenosine, circles) or (2) adenosine at 22.4 nM min−1 + 4 mM theophylline, known as a non-selective adenosine receptor inhibitor (Adenosine + Theophylline, squares). Values are means ± 95% confidential interval (n = 9). ∗ Significantly different vs. Adenosine site (P < 0.05).

2522

N. Fujii and others

vivo (Lorenzo & Minson, 2007; Brunt & Minson, 2012; Cracowski et al. 2013). Therefore, future studies are warranted to explore whether KCa channels are involved in ATP-mediated cutaneous vasodilatation. Adenosine receptors in ATP-mediated cutaneous vasodilatation

ATP can be rapidly catabolized to adenosine in humans in vivo (Zimmermann, 1996), which in itself can cause potent cutaneous vasodilatation as demonstrated in the present study (Fig. 3). However, our results indicate that theophylline, which we show to effectively block adenosine-mediated cutaneous vasodilatation (Fig. 3), did not influence ATP-mediated increases in cutaneous blood flow (Fig. 4). This extends previous work in a human conduit artery in vivo (Mortensen et al. 2009; Kirby et al. 2010) indicating that there is no adenosine-mediated component to ATP-induced vasodilatation. ATP-mediated sweating

ATP can directly evoke secretion from human sweat glands in vitro (Sato et al. 1991). In fact, ATP can increase intracellular calcium levels in the sweat gland epithelial cells, which contributes to sweat secretion elicited by ACh in humans in vivo (Metzler-Wilson et al. 2014). However, the present findings demonstrate no sweating response at

any infusion rate of ATP, which indicates that activation of P2 receptors by ATP does not independently modulate human sweating in vivo (Fig. 5). It was previously reported that the suppression of cutaneous blood flow at rest simultaneously reduced the sweat rate during passive heat stress at rest (Wingo et al. 2010b), indicating that the sweating response is in part related to local cutaneous blood flow. However, we observed the perfusion of ATP at 120 and 1200 nM min−1 to be associated with pronounced cutaneous vasodilatation (i.e. up to 85%max) and therefore increased cutaneous blood flow, whereas no change was detected in the level of sweating. Consequently, these findings indicate that increases in cutaneous blood flow do not affect sweating under normothermic conditions, albeit there may be a link during heat stress wherein the sweat gland is already activated (Wingo et al. 2010b). Sex difference

Although not our main focus, it is noteworthy that females showed modestly greater cutaneous vasodilatation in response to ATP infused at 120 and 1200 nM min−1 in comparison with males. Because we tested all female subjects during the early follicular phase or during the placebo phase, if taking contraceptives, the differences in hormonal status between males and females were 0.5

Control L-NNA

cm –2 )

Control Theophylline

0.4

Ketorolac L-NNA +

Ketorolac

–1

80

Sweat rate (mg min

Cutaneous vascular conductance (%max)

100

J Physiol 593.11

60

40

0.3

0.2

0.1

20 0 Baseline

0 Baseline

0.12

1.2

12

120

1200

ATP (nM min –1 ) Figure 4. Cutaneous vascular conductance at baseline and during ATP administration Cutaneous vascular conductance at baseline and during ATP administration from 0.12 to 1200 nM min−1 (five levels) at two intradermal microdialysis skin sites with (1) lactated Ringer solution (Control, open circles), or (2) 4 mM theophylline, known as a non-selective adenosine receptor inhibitor (Theophylline, squares). Values are means ± 95% confidential interval (n = 8).

0.12

1.2

12

120

1200

ATP (nM min –1)

Figure 5. Sweat rate at baseline and during ATP administration Sweat rate at baseline and during ATP administration from 0.12 to 1200 nM min−1 (five levels) at four intradermal microdialysis skin sites with (1) lactated Ringer solution (Control, open circles), (2) 10 mM NG-nitro-L-arginine (L-NNA, squares), a non-specific nitric oxide synthase inhibitor, (3) 10 mM ketorolac (Ketorolac, triangles), a non-selective cyclooxygenase inhibitor, or (4) a combination of 10 mM L-NNA + 10 mM ketorolac (L-NNA + Ketorolac, diamonds). Values are means ± 95% confidential interval (n = 12).  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

J Physiol 593.11

ATP and thermoeffector activity

minimized. As such, our results may reflect a female hormone-independent sex difference. Interestingly, the concentration of P2 receptors is increased in female neuronal cells compared with males (Crain et al. 2009), and the greater vasodilatatory response to ATP may be explained by a similar upregulated expression of P2 receptors on female skin vessels; however, further exploration of this difference is warranted. Limitations

There are three limitations of the present study. First, we did not evaluate whether interstitial ATP increases during heat stress. Thus, we do not know whether heat stress increases ATP around the sweat glands. On the other hand, it was shown that heat stress increases plasma ATP (Pearson et al. 2011; Kalsi & Gonzalez-Alonso, 2012) that can in turn induce the activation of P2 receptors located on the endothelial cells and thus vasodilatation. However, as discussed above, sensory nerves may be involved in ATP-mediated cutaneous vasodilatation perhaps due to increases in interstitial ATP. Hence future work should evaluate whether heat stress can increase interstitial ATP and the potential implications on cutaneous vascular regulation (i.e. endothelium-dependent, sensory nerve-dependent or a combination of both). Second, given the pharmacological nature of the current study, our data cannot be simply applied to a relevant physiological situation such as exercise in the heat. However, the level of cutaneous vasodilatation to ATP infused at 120 nM min−1 (71%max) is similar to that observed during exercise in the heat (60–70%max) (Fujii et al. 2014c). Furthermore, in the present study we show that NOS is an important modulator for ATP-mediated cutaneous vasodilatation and this is similar to the pattern of response observed in the regulation of cutaneous vasodilatation during exercise in the heat (Fujii et al. 2014c). Hence our results may serve as a surrogate for the control of cutaneous vasodilatation during exercise in the heat. Finally, we did not employ ice to numb the skin before placing the microdialysis fibres. As such, it is possible that the placement of microdialysis fibres may have influenced cutaneous blood flow responses as previously observed during passive heating (Hodges et al. 2009). Perspectives

Hyperthermia can increase plasma ATP levels (Pearson et al. 2011; Kalsi & Gonzalez-Alonso, 2012), and ATP can be co-released with ACh from cholinergic nerves (Rabasseda et al. 1987). Additionally, given that ATP can directly induce cutaneous vasodilatation as shown in the present study, ATP may be required for full expression of cutaneous active vasodilatation during heat stress. On  C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

2523

another front, it is well established that NOS contributes to cutaneous active vasodilatation elicited during passive heating at rest (McCord et al. 2006; Kellogg et al. 2008; Brunt et al. 2013; Wong, 2013) and during exercise (Welch et al. 2009; Fujii et al. 2014c; McGinn et al. 2014a,b; McNamara et al. 2014). Based on our results, which implicate a role for NOS in ATP-mediated cutaneous vasodilatation, ATP may participate in cutaneous active vasodilatation through NOS-dependent mechanisms. With respect to sweating, it is known that cholinergic mechanisms explain most if not all of its production during heat stress (Kellogg et al. 1995; Machado-Moreira et al. 2012). However, sweat rate during administration of highly concentrated cholinergic agents (0.4–0.6 mg min−1 cm−2 ) (Lee & Mack, 2006; Fujii et al. 2014b; Metzler-Wilson et al. 2014) is much lower than that observed during moderate to vigorous intensity exercise in the heat (1.0–2.0 mg min−1 cm−2 ) (Welch et al. 2009; McGinn et al. 2014b). This may indicate the influence of other non-cholinergic mechanisms that may be independently recruited to stimulate sweat production during heat stress. The current study excluded the possibility that ATP directly increases sweating. However, given that ATP and ACh are released from cholinergic nerves, future studies should aim to delineate the possibility of an amplificatory role for ATP with respect to cholinergic sweating.

Conclusion

We show that in humans in vivo, ATP-induced cutaneous vasodilatation is in part explained by NOS- but not through COX- or adenosine-dependent mechanisms. We also show that ATP does not directly increase sweating in vivo. We verified that theophylline can be effective as a non-specific competitive adenosine receptor inhibitor in human skin.

References Aoki Y, Ohtori S, Takahashi K, Ino H, Ozawa T, Douya H, Chiba T & Moriya H (2003). P2X3-immunoreactive primary sensory neurons innervating lumbar intervertebral disc in rats. Brain Res 989, 214–220. Brunt VE, Fujii N & Minson CT (2013). No independent, but an interactive, role of calcium-activated potassium channels in human cutaneous active vasodilation. J Appl Physiol 115, 1290–1296. Brunt VE, Miner JA, Meendering JR, Kaplan PF & Minson CT (2011). 17-β estradiol and progesterone independently augment cutaneous thermal hyperemia but not reactive hyperemia. Microcirculation 18, 347–355. Brunt VE & Minson CT (2012). KCa channels and epoxyeicosatrienoic acids: major contributors to thermal hyperaemia in human skin. J Physiol 590, 3523–3534.

2524

N. Fujii and others

Buono MJ, Tabor B & White A (2011). Localized β-adrenergic receptor blockade does not affect sweating during exercise. Am J Physiol Regul Integr Comp Physiol 300, R1148– R1151. Busse R & Mulsch A (1990). Calcium-dependent nitric-oxide synthesis in endothelial cytosol is mediated by calmodulin. FEBS Lett 265, 133–136. Cracowski JL, Gaillard-Bigot F, Cracowski C, Sors C, Roustit M & Millet C (2013). Involvement of cytochrome epoxygenase metabolites in cutaneous postocclusive hyperemia in humans. J Appl Physiol 114, 245–251. Crain JM, Nikodemova M & Watters JJ (2009). Expression of P2 nucleotide receptors varies with age and sex in murine brain microglia. J Neuroinflammation 6, 24. Fieger SM & Wong BJ (2010). Adenosine receptor inhibition with theophylline attenuates the skin blood flow response to local heating in humans. Exp Physiol 95, 946–954. Fieger SM & Wong BJ (2012). No direct role for A1/A2 adenosine receptor activation to reflex cutaneous vasodilatation during whole-body heat stress in humans. Acta Physiol (Oxf) 205, 403–410. Fujii N, Brunt VE & Minson CT (2014a). Tempol improves cutaneous thermal hyperemia through increasing nitric oxide bioavailability in young smokers. Am J Physiol Heart Circ Physiol 306, H1507–H1511. Fujii N, McGinn R, Paull G, Stapleton JM, Meade RD & Kenny GP (2014b). Cyclooxygenase inhibition does not alter methacholine-induced sweating. J Appl Physiol (1985) 117, 1055–1062. Fujii N, McGinn R, Stapleton J, Paull G, Meade R & Kenny G (2014c). Evidence for cyclooxygenase-dependent sweating in young males during intermittent exercise in the heat. J Physiol 592, 5327–5339. Fujii N, Reinke MC, Brunt VE & Minson CT (2013). Impaired acetylcholine-induced cutaneous vasodilation in young smokers: roles of nitric oxide and prostanoids. Am J Physiol Heart Circ Physiol 304, H667–H673. Hodges GJ, Chiu C, Kosiba WA, Zhao K & Johnson JM (2009). The effect of microdialysis needle trauma on cutaneous vascular responses in humans. J Appl Physiol (1985) 106, 1112–1118. Holowatz LA, Thompson CS, Minson CT & Kenney WL (2005). Mechanisms of acetylcholine-mediated vasodilatation in young and aged human skin. J Physiol 563, 965– 973. Houghton BL, Meendering JR, Wong BJ & Minson CT (2006). Nitric oxide and noradrenaline contribute to the temperature threshold of the axon reflex response to gradual local heating in human skin. J Physiol 572, 811–820. Kalsi KK & Gonzalez-Alonso J (2012). Temperature-dependent release of ATP from human erythrocytes: mechanism for the control of local tissue perfusion. Exp Physiol 97, 419–432. Kellogg DL, Jr, Pergola PE, Piest KL, Kosiba WA, Crandall CG, Grossmann M & Johnson JM (1995). Cutaneous active vasodilation in humans is mediated by cholinergic nerve cotransmission. Circ Res 77, 1222–1228. Kellogg DL, Jr, Zhao JL & Wu Y (2008). Neuronal nitric oxide synthase control mechanisms in the cutaneous vasculature of humans in vivo. J Physiol 586, 847–857.

J Physiol 593.11

Kellogg DL, Zhao JL, Coey U & Green JV (2005). Acetylcholine-induced vasodilation is mediated by nitric oxide and prostaglandins in human skin. J Appl Physiol 98, 629–632. Kirby BS, Crecelius AR, Voyles WF & Dinenno FA (2010). Vasodilatory responsiveness to adenosine triphosphate in ageing humans. J Physiol 588, 4017–4027. Kuwahara T, Inoue Y, Abe M, Sato Y & Kondo N (2005). Effects of menstrual cycle and physical training on heat loss responses during dynamic exercise at moderate intensity in a temperate environment. Am J Physiol Regul Integr Comp Physiol 288, R1347–R1353. Lee K & Mack GW (2006). Role of nitric oxide in methacholine-induced sweating and vasodilation in human skin. J Appl Physiol 100, 1355–1360. Lindsay SL, Elder HY, Tavadia S, Bovell DL (2002). Immunolocalisation of the P2Y1, P2Y2 and P2Y4 receptor subtypes in human eccrine sweat glands. J Physiol 543P, S210. Lorenzo S & Minson CT (2007). Human cutaneous reactive hyperaemia: role of BKCa channels and sensory nerves. J Physiol 585, 295–303. Machado-Moreira CA, McLennan PL, Lillioja S, vanDijk W, Caldwell JN & Taylor NAS (2012). The cholinergic blockade of both thermally and non-thermally induced human eccrine sweating. Exp Physiol 97, 930–942. McCord GR, Cracowski JL & Minson CT (2006). Prostanoids contribute to cutaneous active vasodilation in humans. Am J Physiol Regul Integr Comp Physiol 291, R596– R602. McGinn R, Fujii N, Swift B, Lamarche DT & Kenny GP (2014a). Adenosine receptor inhibition attenuates the suppression of postexercise cutaneous blood flow. J Physiol 592, 2667– 2678. McGinn R, Paull G, Meade R, Fujii N & Kenny G (2014b). Mechanisms underlying the postexercise baroreceptor-mediated suppression of heat loss. Physiol Rep 2, e12168. McNamara TC, Keen JT, Simmons GH, Alexander LM & Wong BJ (2014). Endothelial nitric oxide synthase mediates the nitric oxide component of reflex cutaneous vasodilatation during dynamic exercise in humans. J Physiol 592, 5317–5326. Medow MS, Glover JL & Stewart JM (2008). Nitric oxide and prostaglandin inhibition during acetylcholine-mediated cutaneous vasodilation in humans. Microcirculation 15, 569–579. Metzler-Wilson K, Sammons DL, Ossim MA, Metzger NR, Jurovcik AJ, Krause BA & Wilson TE (2014). Extracellular calcium chelation and attenuation of calcium entry decrease in vivo cholinergic-induced eccrine sweating sensitivity in humans. Exp Physiol 99, 393–402. Minson CT (2010). Thermal provocation to evaluate microvascular reactivity in human skin. J Appl Physiol 109, 1239–1246. Mortensen SP, Gonzalez-Alonso J, Bune LT, Saltin B, Pilegaard H & Hellsten Y (2009). ATP-induced vasodilation and purinergic receptors in the human leg: roles of nitric oxide, prostaglandins, and adenosine. Am J Physiol Regul Integr Comp Physiol 296, R1140–R1148.

 C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

J Physiol 593.11

ATP and thermoeffector activity

Pearson J, Low DA, Stohr E, Kalsi K, Ali L, Barker H & Gonzalez-Alonso J (2011). Hemodynamic responses to heat stress in the resting and exercising human leg: insight into the effect of temperature on skeletal muscle blood flow. Am J Physiol Regul Integr Comp Physiol 300, R663–R673. Rabasseda X, Solsona C, Marsal J, Egea G & Bizzini B (1987). ATP release from pure cholinergic synaptosomes is not blocked by tetanus toxin. FEBS Lett 213, 337–340. Sato K, Ohtsuyama M, Suzuki Y & Sato F (1991). Normal and abnormal eccrine sweat gland function (2nd ed.). In Pathophysiology of Dermatologic Diseases, ed. Baden HP, pp. 211–234. McGraw-Hill, New York. Shibasaki M, Durand S, Davis SL, Cui J, Low DA, Keller DM & Crandall CG (2007). Endogenous nitric oxide attenuates neutrally mediated cutaneous vasoconstriction. J Physiol 585, 627–634. Swift B, McGinn R, Gagnon D, Crandall CG & Kenny GP (2014). Adenosine receptor inhibition attenuates the decrease in cutaneous vascular conductance during whole-body cooling from hyperthermia. Exp Physiol 99, 196–204. Welch G, Foote KM, Hansen C & Mack GW (2009). Nonselective NOS inhibition blunts the sweat response to exercise in a warm environment. J Appl Physiol 106, 796–803. Wingo JE, Brothers RM, Del Coso J & Crandall CG (2010a). Intradermal administration of ATP does not mitigate tyramine-stimulated vasoconstriction in human skin. Am J Physiol Regul Integr Comp Physiol 298, R1417–R1420. Wingo JE, Low DA, Keller DM, Brothers RM, Shibasaki M & Crandall CG (2010b). Skin blood flow and local temperature independently modify sweat rate during passive heat stress in humans. J Appl Physiol 109, 1301–1306. Wong BJ (2013). Sensory nerves and nitric oxide contribute to reflex cutaneous vasodilation in humans. Am J Physiol Regul Integr Comp Physiol 304, R651–R656. Wong BJ & Fieger SM (2012). Transient receptor potential vanilloid type 1 channels contribute to reflex cutaneous vasodilation in humans. J Appl Physiol 112, 2037–2042. Yamamoto K, Korenaga R, Kamiya A, Qi Z, Sokabe M & Ando J (2000). P2X4 receptors mediate ATP-induced calcium influx in human vascular endothelial cells. Am J Physiol Heart Circ Physiol 279, H285–H292.

 C 2015 The Authors. The Journal of Physiology  C 2015 The Physiological Society

2525

Zimmermann H (1996). Biochemistry, localization and functional roles of ecto-nucleotidases in the nervous system. Prog Neurobiol 49, 589–618.

Additional information Competing interests None. Author contributions N.F. and G.P.K. conceived and designed the experiments. N.F., R.M., L.H. and M.S.S. contributed to data collection. N.F., R.M., L.H. and M.S.S. performed data analysis and assembly. N.F., R.M., L.H., M.S.S., N.K. and G.P.K. interpreted the experimental results. N.F. prepared figures and drafted the manuscript. N.F., R.M., L.H., M.S.S., N.K. and G.P.K. edited and revised the manuscript. All authors approved the final version of the manuscript. All experiments took place at the Human and Environmental Physiology Research Unit located at the University of Ottawa. Funding This study was supported by the Natural Sciences and Engineering Research Council Discovery Grant (RGPIN298159-2009 and RGPIN-06313-2014), Discovery Grants Program – Accelerator Supplements (RGPAS-462252-2014), and by Leaders Opportunity Fund from the Canada Foundation for Innovation (Grant 22529) (Funds held by G.P.K.). G.P.K. was supported by a University of Ottawa Research Chair Award. N.F. was supported by the Human and Environmental Physiology Research Unit. R.M. was supported by a Queen Elizabeth II Graduate Scholarship in Science and Technology. Acknowledgements We thank all of the volunteers for taking their time to participate in this study.