Alterations in sympathetic neurovascular transduction during acute ...

Am J Physiol Regul Integr Comp Physiol 304: R959–R965, 2013. First published April 10, 2013; doi:10.1152/ajpregu.00071.2013.

Alterations in sympathetic neurovascular transduction during acute hypoxia in humans Can Ozan Tan,1,2 Yu-Chieh Tzeng,3 Jason W. Hamner,1 Renaud Tamisier,4,5 and J. Andrew Taylor1,2 1

Cardiovascular Research Laboratory, Spaulding Rehabilitation Hospital, Boston, Massachusetts; 2Department of Physical Medicine and Rehabilitation, Harvard Medical School, Boston, Massachusetts; 3Cardiovascular Systems Laboratory, Centre for Translational Physiology, University of Otago, Wellington, New Zealand; 4Sleep Laboratory and Epreuve Fonctionnelle Cardio-Respiatoire, Department of Rehabilitation and Physiology, University Hospital Grenoble, France; and 5HP2 Laboratory (Hypoxia: Pathophysiology) Institut National de la Santé et de la Recherche Médicale U1042 Joseph Fourier University, Grenoble, France Submitted 4 February 2013; accepted in final form 2 April 2013

Tan CO, Tzeng YC, Hamner JW, Tamisier R, Taylor JA. Alterations in sympathetic neurovascular transduction during acute hypoxia in humans. Am J Physiol Regul Integr Comp Physiol 304: R959 –R965, 2013. First published April 10, 2013; doi:10.1152/ajpregu.00071.2013.— Resting vascular sympathetic outflow is significantly increased during and beyond exposure to acute hypoxia without a parallel increase in either resistance or pressure. This uncoupling may indicate a reduction in the ability of sympathetic outflow to effect vascular responses (sympathetic transduction). However, the effect of hypoxia on sympathetic transduction has not been explored. We hypothesized that transduction would either remain unchanged or be reduced by isocapnic hypoxia. In 11 young healthy individuals, we measured beat-by-beat pressure, multiunit sympathetic nerve activity, and popliteal blood flow velocity at rest and during isometric handgrip exercise to fatigue, before and during isocapnic hypoxia (⬃80% SpO2), and derived sympathetic transduction for each subject via a transfer function that reflects Poiseuille’s law of flow. During hypoxia, heart rate and sympathetic nerve activity increased, whereas pressure and flow remained unchanged. Both normoxic and hypoxic exercise elicited significant increases in heart rate, pressure, and sympathetic activity, although sympathetic responses to hypoxic exercise were blunted. Hypoxia slightly increased the gain relation between pressure and flow (0.062 ⫾ 0.006 vs. 0.074 ⫾ 0.004 cm·s⫺1·mmHg⫺1; P ⫽ 0.04), but markedly increased sympathetic transduction (⫺0.024 ⫾ 0.005 vs. ⫺0.042 ⫾ 0.007 cm·s⫺1·spike⫺1; P ⬍ 0.01). The pressor response to isometric handgrip was similar during normoxic and hypoxic exercise due to the balance of interactions among the tachycardia, sympathoexcitation, and transduction. This indicates that the ability of sympathetic activity to affect vasoconstriction is enhanced during brief exposure to isocapnic hypoxia, and this appears to offset the potent vasodilatory stimulus of hypoxia. isocapnic hypoxia; muscle sympathetic nerve activity; vascular control HYPOXIA IS A POTENT STIMULUS to the cardiopulmonary system in humans. Both resting heart rate and vascular sympathetic outflow are significantly increased during acute hypoxia (21, 25, 28, 43), and the elevation in sympathetic outflow can persist up to 1 h beyond the period of exposure (38, 44, 45). However, the elevated sympathetic activity does not elicit a parallel increase in either vascular resistance or arterial pressure (15, 25, 37, 38). This apparent uncoupling between steady-state sympathetic activity and vascular tone may indicate an alteration in the ability of sympathetic outflow to effect an increase in vascular resistance,

Address for reprint requests and other correspondence: C. O. Tan, SW052, Cardiovascular Research L aboratory, Spaulding Hospital Cambridge, 1575 Cambridge St., Cambridge, MA 02138 (e-mail: [email protected]). http://www.ajpregu.org

i.e., lesser sympathetic neurovascular transduction. For example, hypoxia can result in inhibition of norepinephrine release (18) and arteriolar desensitization to sympathetic stimulation (42). This suggests a reduction in sympathetic neurovascular transduction and may explain the lack of increased resistance in the face of marked sympathoexcitation. Conversely, it is possible that sympathetic transduction may not be altered during hypoxia. Vasodilatory ␤-adrenergic (5, 27) and nitric oxide (6, 26) pathways are stimulated by hypoxia, and so the lack of increases in vascular resistance and arterial pressure during hypoxia may simply reflect a shift to a new steady-state balance between local vasodilators and sympathetic vasoconstriction. Indeed, ␣-adrenergic blockade results in a pronounced reduction in regional resistance during hypoxia (43), suggesting that sympathetic vasoconstriction may simply be offset by hypoxia-mediated vasodilation. However, these data do not explicitly define the effect of hypoxia on sympathetic neurovascular transduction in humans. Therefore, we sought to quantify the effect of acute isocapnic hypoxia on sympathetic neurovascular transduction in young healthy individuals. To assess the functional effect of sympathetic activity on blood flow, we used isometric handgrip exercise as a physiological approach. Isometric handgrip exercise elicits a highly reproducible pressor response mediated by progressive increases in muscle sympathetic nerve activity (33) that are causally related to decreasing leg blood flow in a feed-forward manner, providing a nonpharmacological method to quantify regional sympathetic neurovascular transduction (16). During isometric handgrip exercise, sympathetic activity traverses a physiological range from resting to very high levels (16, 20, 31, 34), so that we could accurately determine the relation between sympathetic nervous outflow and blood flow (39). Moreover, handgrip exercise does not significantly change cutaneous vascular conductance (11, 19), and there is no functionally significant effect of hypoxia on skin blood flow responses to moderate exercise (29). Therefore, potential confounding effects of cutaneous vasodilation are avoided. We hypothesized that sympathetic neurovascular transduction would either remain unchanged or be reduced by isocapnic hypoxia. MATERIALS AND METHODS

Subjects. Eleven volunteers (25 ⫾ 3 yr old, 6 males) participated in this study. All subjects gave their written informed consent to participate. All studies were conducted between 8 AM and 12 PM, and all subjects had abstained from caffeine-containing beverages for at least

0363-6119/13 Copyright © 2013 the American Physiological Society

R959

R960

NEUROVASCULAR TRANSDUCTION DURING HYPOXIA

12 h and from alcoholic beverages and exercise for at least 48 h prior to the study. All subjects were nonsmokers and free of elevated resting blood pressures and overt autonomic or cardiovascular diseases. None of the participants were using prescription medications. This protocol was approved by institutional review board at Spaulding Rehabilitation Hospital (#2010-P-000960) and conformed to the Declaration of Helsinki. Protocol and measurements. Photoplethysmographic arterial blood pressure waveforms in a finger (Finapres, Ohmeda, Louisville, CO) and standard three-lead electrocardiogram were recorded continuously throughout the study. Oscillometric brachial pressures were recorded at 2-min intervals as a check for photoplethysmographic finger pressures throughout the study session. Multiunit postganglionic muscle sympathetic activity was recorded from the peroneal nerve (41). Neural activity was amplified, band pass filtered, rectified, and integrated to create sympathetic neurograms for real-time inspection. Raw unfiltered multifiber nerve activity was recorded at 20 kHz for off-line processing using the algorithm developed in our laboratory to quantify moment-by-moment sympathetic nerve activity (40). Popliteal artery blood flow velocity was recorded from a 4-MHz Doppler probe (Multi-Dop T2; Compumedics DWL, Singen, Germany) at the popliteal fossa of the leg contralateral to the nerve recording. End-tidal carbon dioxide concentrations and inspired volumes were recorded throughout all protocols via an infrared analyzer (VacuMed) connected to a mouthpiece with two-way respiratory valve. Oxyhemoglobin saturation was measured every minute via a pulse oximeter (GE Dash 2000). All signals were digitized at 20 kHz and stored (PowerLab, ADInstruments, Colorado Springs, CO) for subsequent off-line analysis. Following instrumentation, each subject’s maximal voluntary handgrip force was determined from at least three maximal contractions on a handgrip dynamometer. After a 5-min resting baseline, subjects performed sustained isometric handgrip exercise at 35% of their maximum voluntary contraction with the nondominant hand to fatigue. During exercise, target force was displayed on a computer monitor and set at 35% of the maximal voluntary handgrip force. Subjects were provided continuous visual and auditory feedback from the investigators to ensure maintenance of target force until exhaustion, which was defined as a decrease in handgrip force of ⬎10% below the target for ⬎2 s despite verbal encouragement and the attainment of a maximal perceived exertion. Following handgrip exercise, acute isocapnic hypoxia was induced using the partial rebreathing system described by Banzett et al. (3). Briefly, subjects breathed through a leak-free face-mask (8940 Series; Hans Rudolph, Kansas City, MO) attached to a large tube that acts as an expiratory alveolar gas reservoir. Subjects inspired from a bag reservoir containing premixed inspiratory gases regulated by a highimpedance constant flow source. The degree of hypoxia was adjusted by altering the inspiratory gas mixture while holding end-tidal concentration constant. When required, the length of the expiratory alveolar gas reservoir tube was altered to adjust end-tidal concentrations. Subjects were exposed to moderate hypoxia by breathing 100% oxygen blended with 100% nitrogen with end-tidal CO2 clamped at baseline levels to maintain oxyhemoglobin saturation (SpO2) at ⬃80% for 20 min. Twenty minutes of exposure to hypoxia ensured that a steady state (in terms of ventilation and cardiovascular responses) was consistently achieved by all subjects (Fig. 1). Following this 20-min period, subjects performed a second isometric handgrip exercise to fatigue during the isocapnic hypoxia. Data analysis. Raw multifiber muscle sympathetic nerve activity was analyzed using the algorithm recently developed and validated in our laboratory (40). This analysis involves explicit steps to remove noise and movement artifacts and to identify nerve firings on the basis of the statistical properties of background noise. The result is a sympathetic nerve firing with minimal noise that can be quantified as spikes (i.e., nerve firings) per second or per beat. We have previously shown that this algorithm provides an accurate representation

Fig. 1. Resting ventilatory and cardiovascular responses to hypoxic exposure. Upper panel shows minute ventilation (l per minute) and lower panel shows mean arterial pressure (mmHg,
) and heart rate (beats per minute, Œ) averaged over each minute of 5 min of normoxia (gray area) and the 20 min of hypoxic exposure.

of sympathetic activity with greater temporal resolution than sympathetic burst identification. After application of the algorithm to the sympathetic activity, we derived neurovascular transduction for each subject as previously described (39). This approach uses a transfer function in the time domain that is equivalent to an autoregressive model reflecting Poiseuille’s law of flow, specifically: F ⫽ 共␲ ⁄ 8兲⌬P共r4 ⁄ ␩L兲

(1)

where ␲/8 is constant, ⌬P is the pressure gradient and the term r /(␩L) is directly proportional to vessel conductance (4). The approach assumes that pressure drives flow within the same beat, whereas sympathetic activity affects vessel conductance over a longer period. Thus, considering the time course of sympathetic effect, Eq. 1 is rewritten as 4

F ⫽ k⌬P

T ct S关T⫺t兴 兺t⫽0

(2)

That is, blood pressure (⌬P) affects blood flow (F) linearly within the same beat, and sympathetic activity affects vessel conductance via a time-dependent transfer function equivalent to an autoregressive model (⌺Tt⫽0 ct S[T ⫺ t]) (8). The parameter k provides an estimate for the gain of blood pressure; the parameters (ct) for each time point t provides a transfer function in the time domain between sympathetic activity and blood flow, and the sum of all parameters provides an overall sympathetic neurovascular transduction estimate. Although changes in diameter over time would confound the estimate of vessel conductance, previous work has shown that acute hypoxia does not alter femoral or popliteal artery diameter at rest or during moderate exercise (22, 35). Moreover, femoral and popliteal artery diameter

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00071.2013 • www.ajpregu.org

R961

NEUROVASCULAR TRANSDUCTION DURING HYPOXIA

Table 1. Resting hemodynamics and responses to handgrip exercise during normoxia and isocapnic hypoxia Normoxia

Hypoxia

Condition

Rest

End Exercise

Rest

End Exercise

Heart rate, beats/min Systolic blood pressure, mmHg Diastolic blood pressure, mmHg Mean blood pressure, mmHg Flow velocity, cm/s Vascular resistance, mmHg 䡠 s 䡠 cm⫺1 Muscle sympathetic nerve activity Spikes/min Spikes/100 heartbeats

60.9 ⫾ 2.4 110.8 ⫾ 2.1 61.5 ⫾ 2.3 76.8 ⫾ 1.9 12.0 ⫾ 1.9 8.1 ⫾ 0.4

90.6 ⫾ 3.8* 160.3 ⫾ 4.6* 95.8 ⫾ 3.9* 117.3 ⫾ 3.9* 11.2 ⫾ 2.4 15.2 ⫾ 0.9*

76.4 ⫾ 3.2† 111.7 ⫾ 2.3 58.7 ⫾ 3.2 77.9 ⫾ 2.9 11.8 ⫾ 1.7 8.2 ⫾ 0.4

102.7 ⫾ 4.8*† 158.4 ⫾ 4.4* 91.2 ⫾ 3.3* 113.7 ⫾ 3.2* 15.4 ⫾ 2.3†‡ 9.8 ⫾ 0.5*†‡

450 ⫾ 49 747 ⫾ 79

1424 ⫾ 157* 1583 ⫾ 185*

727 ⫾ 62† 957 ⫾ 75†

1423 ⫾ 181* 1434 ⫾ 218*

*P ⬍ 0.05 vs. rest. †P ⬍ 0.05 vs. normoxia. ‡P ⬍ 0.05 exercise ⫻ hypoxia interaction.

remains unchanged with stimuli that increase flow (calf exercise) (32) and that decrease flow (cold pressor test) (23). We normalized the blood flow velocity for each individual to the first 30 s of (normoxic or hypoxic) handgrip exercise [blood flow velocities during the first 30 s of exercise were not different compared with that at rest (11.9 ⫾ 1.2 cm/s at rest vs. 12.9 ⫾ 1.4 cm/s during the first 30 s of handgrip exercise)]. In this way, we were able to accurately and reliably capture changes in blood flow, while accounting for intersubject differences in basal flow velocity. Although the changes in arterial pressure, sympathetic nerve activity, and blood flow during handgrip exercise are progressive, we have previously shown that their relation is unrelated to the overall duration of handgrip exercise (39). Therefore, estimation of the regional neurovascular transduction via this method is not impacted by nonstationarity in the signals themselves. All models were estimated using beat-by-beat signals down-sampled to 4 Hz via custom-written software in MatLab (Mathworks, Natick, MA) and R-Language Environment for Statistical Computing (R Foundation). We used Akaike’s information criterion (1) to maximize the goodness-of-fit of the models. We have also calculated the resistance (mmHg·s·cm⫺1) at rest and at the end (last 30 s) of handgrip exercise during normoxia and hypoxia as the ratio of mean blood pressure to popliteal blood flow. Statistics. All variables were normally distributed (Shapiro-Wilk test), and so parametric statistics were used for all comparisons. Hemodynamic responses to handgrip exercise and during normoxia and hypoxia were compared using a two-way repeated-measures ANOVA using hypoxia and handgrip exercise as independent factors

(normoxia vs. hypoxia and baseline vs. last 30 s of exercise). Blood pressure gain and sympathetic neurovascular transduction during hypoxia were compared between hypoxia and control (normoxia) conditions using Student’s paired t-test. The level of significance was set at P ⬍ 0.05. All data are presented as means ⫾ SE. RESULTS

By design, exposure to isocapnic hypoxia resulted in a significant reduction in arterial oxygen saturation (98.7 ⫾ 0.3% vs. 78.9 ⫾ 0.8%, P ⬍ 0.01) without any change in end-tidal carbon dioxide (45.3 ⫾ 1.6 vs. 45.3 ⫾ 1.5 mmHg) at rest. As expected, inspired volume (liters: 0.49 ⫾ 0.04 vs. 0.87 ⫾ 0.19, P ⫽ 0.07), minute ventilation (l/min: 7.0 ⫾ 0.6 vs. 14.6 ⫹ 3.5, P ⫽ 0.06), heart rate, and sympathetic nerve activity increased (spikes/min: 450 ⫾ 49 vs. 727 ⫾ 62, bursts/min: 11 ⫾ 2 vs. 14 ⫾ 2, and total activity, arbitrary integration units: 477 ⫾ 62 to 794 ⫾ 144, all P ⬍ 0.05), whereas arterial pressure and popliteal flow remained unchanged from normoxia (Table 1). Average maximum voluntary contraction (MVC) force was 34 ⫾ 3 kg. All subjects sustained isometric handgrip exercise at 35% of their MVC to fatigue for a similar duration during normoxic and hypoxic exercise (290 ⫾ 17 vs. 247 ⫾ 14 s; P ⬎ 0.1). Handgrip exercise elicited significant increases in heart rate, blood pressure, and sympathetic activity under both con-

Fig. 2. Box-and-whisker plots of changes in mean pressure, heart rate, sympathetic activity and popliteal blood flow with isometric handgrip exercise (exercise–rest) during normoxia and isocapnic hypoxia. The boundaries indicate 25th and 75th percentiles; the line and the point within the box mark the median and the mean, while the whiskers show the 10th and 90th percentiles. AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00071.2013 • www.ajpregu.org

R962

NEUROVASCULAR TRANSDUCTION DURING HYPOXIA

Fig. 3. The effect of isocapnic hypoxia on blood pressure gain and sympathetic neurovascular transduction.

ditions (Table 1 and Fig. 2). Sympathetic responses to handgrip exercise were lower, although not statistically different during hypoxic exercise (Table 1 and Fig. 2; handgrip ⫻ hypoxia interaction P ⫽ 0.19 for spikes/min and P ⫽ 0.12 for spikes/ 100 heartbeats). However, the difference in sympathetic responses to handgrip exercise was statistically significant when hypoxia-mediated changes in resting sympathetic activity were taken into account (percent change from rest: 328 ⫾ 27% vs. 204 ⫾ 28%, paired t-test, P ⫽ 0.002). At the end of handgrip exercise, absolute arterial pressure and sympathetic nerve activity were similar between the two conditions, whereas the absolute level of heart rate and blood flow velocity achieved were higher during hypoxic exercise (Table 1; P ⬍ 0.01 for both). The sympathetic neurovascular transduction assessment captured the majority of the relationship between changes in blood flow, arterial pressure, and sympathetic nerve activity. The percent variance explained (R2) was similar during normoxia and hypoxia, averaging more than 80% for both. The gain relation between arterial pressure and blood flow was slightly higher during hypoxia than normoxia (0.062 ⫾ 0.006 vs. 0.074 ⫾ 0.004 cm·s⫺1·mmHg⫺1; P ⫽ 0.04), with an increase in 7 out of 11 individuals (Fig. 3). However, sympathetic neurovascular transduction increased in all individuals during hypoxia, almost doubling from ⫺0.024 ⫾ 0.005 to ⫺0.042 ⫾ 0.007 cm·s⫺1·spike⫺1 (P ⬍ 0.01; Fig. 3). (The increase in transduction was not related to the handgrip duration [linear regression R2 ⫽ 0.09, P ⫽ 0.17), or a change in inspiratory volume (R2 ⫽ 0.005, P ⫽ 0.34), ventilatory rate (R2 ⫽ 0.26, p ⫽ 0.10), or minute ventilation (R2 ⫽ 0.07, P ⫽ 0.24)]. Change in neurovascular transduction was not related to the change in the gain relation between arterial pressure and blood flow (R2 ⫽ 0.001, P ⫽ 0.92). The time course of sympathetic effect was similar in terms of the time to maximal sympathetic effect (normoxia: 5.5 ⫾ 0.6, and hypoxia: 5.4 ⫾ 0.5 s), but the total duration of the sympathetic effect tended to be shorter during hypoxia (normoxia: 11.4 ⫾ 0.3 vs. hypoxia: 10.5 ⫾ 0.3 s; P ⫽ 0.11). There were no sex differences in the gain relation between arterial pressure and blood flow, sympathetic neurovascular transduction, or the time course of sym-

pathetic effect (P ⬎ 0.1 for main effect or interaction for all comparisons). Given the almost doubling of sympathetic transduction but only modestly lower sympathoexcitation with isometric exercise, we were surprised to find that the average resistance response to handgrip was significantly less (Table 1). Moreover, despite this, the blood pressure responses were similar. However, there was a much more marked variability in the heart rate and sympathoexcitatory responses to hypoxic handgrip exercise (see Fig. 2). Therefore, we explored whether alterations in the tachycardia and sympathoexcitation coupled with the changes in transduction provide insight into the blood pressure response during hypoxic exercise. A multiple linear regression of blood pressure differences as a function of differences in tachycardia, sympathoexcitation, and transduction was derived. The combination of these three factors were highly predictive (R2 ⫽ 0.74, P ⬍ 0.05; Fig. 4). This suggests that a balance between tachycardia and sympathoexcitatory responses with respect to the change in transduction accounts for differences in how the pressor response was mediated during normoxic vs. hypoxic exercise. DISCUSSION

In contrast to expectations, we observed an increase in sympathetic neurovascular transduction during isocapnic hypoxia. We have previously shown that assessment of neurovascular transduction via Poiseuille’s law is unrelated to the overall duration of handgrip exercise and that it is highly reproducible across repeated isometric handgrip exercises (39). Therefore, the consistent increase in sympathetic neurovascu-

Fig. 4. Predicted vs. actual differences (normoxia– hypoxia) in pressor response (mmHg) to isometric handgrip exercise. The x-axis shows the differences predicted from the multiple linear regression using differences in tachycardia, sympathoexcitation, and transduction as independent variables. Percent variance explained (R2) is adjusted to account for the number of parameters and sample size.

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00071.2013 • www.ajpregu.org

NEUROVASCULAR TRANSDUCTION DURING HYPOXIA

lar transduction during isocapnic hypoxia is unlikely to be related to the duration or order of handgrip exercises. The increases in neurovascular transduction were not qualitatively impacted by sex. We did not control for the menstrual phase in female subjects. However, unless all women were, by chance, at the very same phase of their menstrual cycle (an improbable coincidence), it is unlikely that menstrual phase was responsible for the consistent increase in the female subjects. Although the magnitude of response could be impacted by hormonal status in women, our findings are broadly representative of the effects of hypoxia in young healthy humans. The increase in neurovascular transduction was not only consistent, occurring in every subject, but also large, with the majority of subjects increasing by at least 50%. Thus, our results show that the overall ability of sympathetic activity to effect vasoconstriction is enhanced by hypoxia and is not diminished despite the hypoxic vasodilatory effects. Moreover, the increase in transduction is balanced by the integrated responses to isometric handgrip exercise, such that the same absolute arterial pressure is achieved at the end of hypoxic exercise. Acute hypoxia stimulates vasodilatory ␤-adrenergic pathways (5, 27) and endothelial nitric oxide release (6, 24, 36), but coupled with an increase in sympathetic outflow, vascular resistance is maintained (38). Nonetheless, the presence of potent vasodilators during hypoxia might suggest less effective sympathetic neurovascular transduction, and this was one of our primary hypotheses. However, arterioles in a dilated state may possess greater capacity to vasoconstrict. For example, vascular responses to a given level of sympathetic stimulation is higher after drug-induced vasodilation, partly due to increased smooth muscle fiber length (30). The augmented sympathetic transduction during hypoxia could simply reflect increased responsiveness of the vasculature due to the vasodilatory effects of ␤-adrenergic receptor stimulation and/or endothelial nitric oxide release. However, although responses varied across subjects, on average, there was no change in either basal blood pressure or popliteal blood flow with hypoxia. Thus, hypoxia did not consistently result in arterial vasodilation, and so this cannot fully explain the augmentation of sympathetic neurovascular transduction seen in every single subject. Many physiological input-output systems have characteristic sigmoidal relations, such that small-output responses are seen at very low (subthreshold) or very high (saturation) input levels, while larger responses are seen between these regions (2). One might surmise that greater tonic sympathetic outflow during hypoxia simply moves the relationship into the region with the greatest vascular response. However, despite increases in resting sympathetic activity with hypoxia, the average increase in activity with handgrip exercise traversed a similar range and reached a similar absolute level by end-exercise compared with normoxia. Thus, it does not seem likely that greater tonic sympathetic firing can account for the increase in sympathetic transduction during hypoxia. Lastly, acute alterations in adrenergic responsiveness could occur. For example, even 10 min of hypoxic exposure results in greater ␣1-adrenoreceptor density and increased production of second messengers in cardiomyocytes, leading to enhanced adrenergic responsiveness (17). If these hypoxia-induced changes also apply to vascular smooth muscle, they may underlie the augmented sympathetic transduction. On the other hand, vasoconstrictor responses to pharmacologically provoked norepinephrine release do not appear to change during

R963

exposure to mild to moderate hypoxia (12, 13). However, sympathetic neurovascular control is a dynamic process, and vasoconstrictor responses to pharmacologically evoked norepinephrine release represent only a snapshot rather than the range of responses. Therefore, it remains possible that alterations in noradrenergic function play a role in augmented sympathetic transduction during acute hypoxia. We also observed a small but significant increase in the effect of blood pressure on blood flow (i.e., stronger blood pressure gain) during hypoxia. Given the lack of a relation between hypoxia-mediated increase in neurovascular transduction and blood pressure gain, it is unlikely that the former underlies the latter. However, greater tonic sympathetic activity could underlie the hypoxia-mediated increase in blood pressure gain. Pharmacological, electrically stimulated, and reflex activation of sympathetic outflow reduces distensibility of small- and medium-sized arteries in both animals (9, 10, 14) and humans (7). Thus, higher sympathetic activity during hypoxia may reduce arterial distensibility and, thereby, enhance the “driving force” of blood pressure into flow. There are a myriad of potential physiological contributors to augmented sympathetic neurovascular transduction during acute hypoxia. Nonetheless, our aim was not to tease apart these various contributors but to determine whether the relationship between sympathetic activity and blood flow reflected a change in neurovascular transduction. Our work lays a foundation for future work to consider the role of various contributors to the hypoxia-mediated differences in sympathetic neurovascular transduction. To our surprise, despite an apparent dissociation between sympathetic activity and vascular tone, sympathetic neurovascular transduction was markedly and consistently enhanced during hypoxia. Thus, merely observing the steady state balance between sympathetic outflow and vascular resistance does not capture the underlying regulatory physiology. Despite the increased neurovascular transduction, limb blood flow decreases and resistance increases were not augmented by hypoxia. However, it should be noted that the increase in sympathetic activity with handgrip was ⬃25% less, in part, because of the increase in transduction (see Fig. 4). In addition, the gain relation between arterial pressure and blood flow was roughly 30% greater. Hence, the balance of these factors determining regional flow responses to handgrip exercise would appear to explain the effect of hypoxia. Nonetheless, the physiological relevance of our sympathetic neurovascular transduction estimate is evidenced by the relationship the changes in transduction had to the altered tachycardia and sympathoexcitation during hypoxic exercise. Across individuals, these three factors were strongly determinative of differences in the isometric handgrip pressor response. The magnitude of change in transduction was balanced by changes in tachycardic and/or sympathoexcitatory responses to result in the same absolute arterial pressure at the end of hypoxic exercise. Future work should examine the roles of vasodilatory substances, sympathetic nerve firing patterns, and/or noradrenergic function in mediating the increased sympathetic neurovascular transduction consequent to isocapnic hypoxia. Perspectives and Significance In contrast to our initial hypothesis, we observed a large and consistent increase in sympathetic neurovascular transduction

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00071.2013 • www.ajpregu.org

R964

NEUROVASCULAR TRANSDUCTION DURING HYPOXIA

during isocapnic hypoxia. We further found that hypoxiamediated changes in chronotropic and sympathoexcitatory response to exercise, and those in sympathetic neurovascular transduction together account for differences in the pressor response to normoxic and hypoxic exercise. These results show that the overall ability of sympathetic activity to effect vasoconstriction is enhanced by hypoxia, and they lay a foundation for future work to understand hypoxia-mediated alterations in the integrated balance between chronotropic, sympathetic, and neurovascular mechanisms. GRANTS Y. C. Tzeng was funded by AMP Foundation Visiting Fellowship. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the authors. AUTHOR CONTRIBUTIONS Author contributions: C.O.T., Y.-C.T., and J.W.H. performed experiments; C.O.T. and J.W.H. analyzed data; C.O.T., Y.-C.T., J.W.H., R.T., and J.A.T. interpreted results of experiments; C.O.T. prepared figures; C.O.T. drafted manuscript; C.O.T., Y.-C.T., J.W.H., R.T., and J.A.T. approved final version of manuscript; R.T. and J.A.T. edited and revised manuscript; J.A.T. conceptualized and designed research. REFERENCES 1. Akaike H. On the use of a linear model for the identification of feedback systems. Ann Inst Stat Math 20: 425, 1968. 2. Andersson PO. Comparative vascular effects of stimulation continuously and in bursts of the sympathetic nerves to cat skeletal muscle. Acta Physiol Scand 118: 343–348, 1983. 3. Banzett RB, Garcia RT, Moosavi SH. Simple contrivance “clamps” end-tidal PCO2 and PO2 despite rapid changes in ventilation. J Appl Physiol 88: 1597–1600, 2000. 4. Berne RM, Levy MN. Cardiovascular Physiology. St. Louis, MO: Mosby, 2001. 5. Blauw GJ, Westendorp RG, Simons M, Chang PC, Frolich M, Meinders AE. ␤-Adrenergic receptors contribute to hypoxaemia induced vasodilation in man. Br J Clin Pharmacol 40: 453–458, 1995. 6. Blitzer ML, Lee SD, Creager MA. Endothelium-derived nitric oxide mediates hypoxic vasodilation of resistance vessels in humans. Am J Physiol Heart Circ Physiol 271: H1182–H1185, 1996. 7. Boutouyrie P, Lacolley P, Girerd X, Beck L, Safar M, Laurent S. Sympathetic activation decreases medium-sized arterial compliance in humans. Am J Physiol Heart Circ Physiol 267: H1368 –H1376, 1994. 8. Box GEP, Jenkins GM, Reinsel GC. Time Series Analysis: Forecasting and Control. Englewood Cliffs, NJ: Prentice Hall, 1994, p. 598. 9. Cox RH. Effects of norepinephrine on mechanics of arteries in vitro. Am J Physiol 231: 420 –425, 1976. 10. Cox RH, Bagshaw RJ. Effects of pulsations on carotid sinus control of regional arterial hemodynamics. Am J Physiol Heart Circ Physiol 238: H182–H190, 1980. 11. Crandall CG, Musick J, Hatch JP, Kellogg DL Jr., Johnson JM. Cutaneous vascular and sudomotor responses to isometric exercise in humans. J Appl Physiol 79: 1946 –1950, 1995. 12. Dinenno FA. Hypoxic regulation of blood flow in humans. ␣-adrenergic receptors and functional sympatholysis in skeletal muscle. Adv Exp Med Biol 543: 237–248, 2003. 13. Dinenno FA, Joyner MJ, Halliwill JR. Failure of systemic hypoxia to blunt alpha-adrenergic vasoconstriction in the human forearm. J Physiol 549: 985–994, 2003. 14. Gerova M. Gero J: Range of the sympathetic control of the dog femoral artery. Circ Res 24: 349 –359, 1969. 15. Gilmartin G, Tamisier R, Anand A, Cunnington D, Weiss JW. Evidence of impaired hypoxic vasodilation after intermediate-duration hypoxic exposure in humans. Am J Physiol Heart Circ Physiol 291: H2173– H2180, 2006. 16. Halliwill JR, Taylor JA, Eckberg DL. Impaired sympathetic vascular regulation in humans after acute dynamic exercise. J Physiol 495: 279 – 288, 1996.

17. Heathers GP, Evers AS, Corr PB. Enhanced inositol trisphosphate response to alpha 1-adrenergic stimulation in cardiac myocytes exposed to hypoxia. J Clin Invest 83: 1409 –1413, 1989. 18. Leuenberger U, Gleeson K, Wroblewski K, Prophet S, Zelis R, Zwillich C, Sinoway L. Norepinephrine clearance is increased during acute hypoxemia in humans. Am J Physiol Heart Circ Physiol 261: H1659 – H1664, 1991. 19. McCord GR, Minson CT. Cutaneous vascular responses to isometric handgrip exercise during local heating and hyperthermia. J Appl Physiol 98: 2011–2018, 2005. 20. Minson CT, Halliwill JR, Young TM, Joyner MJ. Influence of the menstrual cycle on sympathetic activity, baroreflex sensitivity, and vascular transduction in young women. Circulation 101: 862–868, 2000. 21. Moradkhan R, Spitnale B, McQuillan P, Hogeman C, Gray KS, Leuenberger UA. Hypoxia-induced vasodilation and effects of regional phentolamine in awake patients with sleep apnea. J Appl Physiol 108: 1234 –1240, 2010. 22. Nishiwaki M, Kawakami R, Saito K, Tamaki H, Takekura H, Ogita F. Vascular adaptations to hypobaric hypoxic training in postmenopausal women. J Physiol Sci 61: 83–91, 2011. 23. Parker BA, Smithmyer SL, Jarvis SS, Ridout SJ, Pawelczyk JA, Proctor DN. Evidence for reduced sympatholysis in leg resistance vasculature of healthy older women. Am J Physiol Heart Circ Physiol 292: H1148 –H1156, 2007. 24. Pohl U, Busse R. Hypoxia stimulates release of endothelium-derived relaxant factor. Am J Physiol Heart Circ Physiol 256: H1595–H1600, 1989. 25. Querido JS, Kennedy PM, Sheel AW. Hyperoxia attenuates muscle sympathetic nerve activity following isocapnic hypoxia in humans. J Appl Physiol 108: 906 –912, 2010. 26. Reed AS, Tschakovsky ME, Minson CT, Halliwill JR, Torp KD, Nauss LA, Joyner MJ. Skeletal muscle vasodilatation during sympathoexcitation is not neurally mediated in humans. J Physiol 525: 253–262, 2000. 27. Richardson DW, Kontos HA, Raper AJ, Patterson JL Jr. Modification by beta-adrenergic blockade of the circulatory respones to acute hypoxia in man. J Clin Invest 46: 77–85, 1967. 28. Rowell LB, Blackmon JR. Lack of sympathetic vasoconstriction in hypoxemic humans at rest. Am J Physiol Heart Circ Physiol 251: H562– H570, 1986. 29. Rowell LB, Freund PR, Brengelmann GL. Cutaneous vascular response to exercise and acute hypoxia. J Appl Physiol 53: 920 –924, 1982. 30. Rowlands DJ, Donald DE. Sympathetic vasoconstrictive responses during exercise- or drug-induced vasodilatation. A time-dependent response. Circ Res 23: 45–60, 1968. 31. Saito M, Mano T, Iwase S. Changes in muscle sympathetic nerve activity and calf blood flow during static handgrip exercise. Eur J Appl Physiol Occup Physiol 60: 277–281, 1990. 32. Schlager O, Giurgea A, Margeta C, Seidinger D, Steiner-Boeker S, van der LB, Koppensteiner R. Wall shear stress in the superficial femoral artery of healthy adults and its response to postural changes and exercise. Eur J Vasc Endovasc Surg 41: 821–827, 2011. 33. Seals DR. Influence of muscle mass on sympathetic neural activation during isometric exercise. J Appl Physiol 67: 1801–1806, 1989. 34. Seals DR. Sympathetic neural discharge and vascular resistance during exercise in humans. J Appl Physiol 66: 2472–2478, 1989. 35. Stickland MK, Fuhr DP, Haykowsky MJ, Jones KE, Paterson DI, Ezekowitz JA, McMurtry MS. Carotid chemoreceptor modulation of blood flow during exercise in healthy humans. J Physiol 589: 6219 –6230, 2011. 36. Sun MK, Reis DJ. Evidence nitric oxide mediates the vasodepressor response to hypoxia in sino-denervated rats. Life Sci 50: 555–565, 1992. 37. Tamisier R, Nieto L, Anand A, Cunnington D, Weiss JW. Sustained muscle sympathetic activity after hypercapnic but not hypocapnic hypoxia in normal humans. Respir Physiol Neurobiol 141: 145–155, 2004. 38. Tamisier R, Norman D, Anand A, Choi Y, Weiss JW. Evidence of sustained forearm vasodilatation after brief isocapnic hypoxia. J Appl Physiol 96: 1782–1787, 2004. 39. Tan CO, Tamisier R, Hamner JW, Taylor JA. Characterizing sympathetic neurovascular transduction in humans. PLoS One 8: e53769, 2013. 40. Tan CO, Taylor JA, Ler AS, Cohen MA. Detection of multifiber neuronal firings: a mixture separation model applied to sympathetic recordings. IEEE Trans Biomed Eng 56: 147–158, 2009. 41. Vallbo AB, Hagbarth KE. Impulses recorded with micro-electrodes in human muscle nerves during stimulation of mechanoreceptors and voluntary contractions. Electroencephalogr Clin Neurophysiol 23: 392, 1967.

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00071.2013 • www.ajpregu.org

NEUROVASCULAR TRANSDUCTION DURING HYPOXIA 42. Vanhoutte PM, Verbeuren TJ, Webb RC. Local modulation of adrenergic neuroeffector interaction in the blood vessel well. Physiol Rev 61: 151–247, 1981. 43. Weisbrod CJ, Minson CT, Joyner MJ, Halliwill JR. Effects of regional phentolamine on hypoxic vasodilatation in healthy humans. J Physiol 537: 613–621, 2001.

R965

44. Xie A, Skatrud JB, Crabtree DC, Puleo DS, Goodman BM, Morgan BJ. Neurocirculatory consequences of intermittent asphyxia in humans. J Appl Physiol 89: 1333–1339, 2000. 45. Xie A, Skatrud JB, Puleo DS, Morgan BJ. Exposure to hypoxia produces long-lasting sympathetic activation in humans. J Appl Physiol 91: 1555–1562, 2001.

AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00071.2013 • www.ajpregu.org