Intravascular ATP and the Regulation of Blood Flow and Oxygen Delivery in Humans : Exercise and Sport Sciences Reviews

Secondary Logo

Journal Logo


Intravascular ATP and the Regulation of Blood Flow and Oxygen Delivery in Humans

Crecelius, Anne R.1; Kirby, Brett S.2; Dinenno, Frank A.3,4

Author Information
Exercise and Sport Sciences Reviews 43(1):p 5-13, January 2015. | DOI: 10.1249/JES.0000000000000031
  • Free
  • Journal Club


Editor’s note: Go online to view the Journal Club questions in the Supplemental Digital Content: see


Appropriate matching of oxygen delivery to tissue metabolic demand occurs largely via adjustments in convective oxygen delivery or, more simply, changes in blood flow. In healthy humans, blood flow is correlated linearly with oxygen consumption in tissue beds such as cardiac and skeletal muscle, and augmentation of oxygen demand (e.g., muscle contractions) is met with a synchronized increase in blood flow and oxygen supply (20). In the case of challenged oxygen supply, as in systemic hypoxia, where hemoglobin oxygen saturation is lower than normal, blood flow increases to normalize oxygen delivery (i.e., blood flow × arterial blood oxygen content) to the given demand (6,27,30). Accordingly, any oxygen mismatch evoked by an acute elevation in oxygen demand, reduction in oxygen delivery, or an exacerbated combination of both stimuli results in an elevation in tissue blood flow.

Absolute tissue blood flow is determined by perfusion pressure and arteriolar resistance. During physiological stresses, both determinants of blood flow are modulated and influential; however, it is vascular resistance that has a greater impact because changes in vessel diameter are magnified to the fourth power (Poiseuille’s law). Arteriolar vascular caliber is controlled by multiple input signals including sympathetic vasoconstrictor tone, circulating vasoactive hormones or neurotransmitters, transmural pressure, endothelial-derived substances, and metabolic factors acting on the vessel extraluminally and intraluminally. At the regional and microcirculatory level, a great deal of redundancy is observed with regard to regulating net vascular tone and this is often evidenced by the failure of pharmacological inhibition of one or more vasoactive factors to impair metabolic autoregulation (20).

Metabolic autoregulation has been the topic of many investigations in both animal and human models, and the compilation of findings clearly demonstrates an extremely complex control of vascular tone. Specific vasoactive candidates important in the control of vascular tone during physiological stress often are framed against a number of criteria. First, regarding the endogenous molecule, it must be measureable, have inactivation mechanisms, and release should occur at the required location and be stimulated during the stress. Secondly, if the candidate molecule is administered exogenously, the vasoactive response should mimic that which occurs during the stress. Finally, inhibiting the vascular action of the molecule should be consistent with the proposed hypothesis of vascular regulation (12). Recent insights gained from our laboratory and others demonstrate that intravascular adenosine triphosphate (ATP) is a candidate molecule that largely satisfies these criteria. Therefore, the purpose of this review is to present the most significant and recent data related to the role of intravascular ATP in vascular control of humans including the potential sources and stimuli for endogenous release and the signaling pathways underlying ATP’s powerful vasoregulatory action. Finally, we present the postulate that changes in the regulation of intravascular ATP and associated vasomotor signaling may contribute to observed impairments in vascular control of aged or diseased humans during conditions of oxygen mismatch.


In the late 1960s, Forrester and Lind (13) reported an increase in venous plasma ATP concentrations during exercise in humans. Since that time, our laboratory and others have advanced understanding through investigations demonstrating an increase in plasma [ATP] during graded-intensity whole-body or isolated limb models of exercise in humans (4,14,16,25,30). In general, the observed elevations in intravascular [ATP] are greatest in venous blood draining the active muscle and occur in an exercise intensity (and tissue oxygen consumption)–dependent manner (Fig. 1A, B). With regard to hypoxia, when humans are exposed to low–oxygen content air (∼10% fractional inspired O2; O2 saturations, ∼80%), venous plasma [ATP] draining muscle also is elevated in young healthy humans (Fig. 1C, D) (16,25,30). Along with increases occurring primarily in the venous circulation, the degradation of ATP via cell-surface ectonucleotideases seems to be rapid (half-life of intravascular ATP, <1 s) and, thus, it is likely that it is a local release of ATP within or in close proximity to the microcirculation that drives elevations in plasma [ATP] draining skeletal muscle during exercise and hypoxia (25,30).

Figure 1:
Increased venous plasma (adenosine triphosphate) [ATP] draining skeletal muscle during exercise and hypoxia. Progressive rhythmic handgrip (% of maximal voluntary contraction (MVC)) (A) or single-leg knee extensor (B) exercise significantly increases venous plasma [ATP]. Exposure to systemic hypoxia (inspired O2 fraction, ∼10%; hemoglobin O2 saturation, ∼80%) results in significantly increased plasma [ATP] as measured by blood sample from deep forearm vein (C) and via intravascular microdialysis in the femoral vein (D) of young healthy humans. Differences in measurement technique contribute to discrepant absolute values between studies (see text for details). *P < 0.05 versus normoxia/rest; ‡P < 0.05 versus 5% MVC exercise. (Panels A and C adapted from (25). Copyright © 2012 American Heart Association. Used with permission. Panels B and D adapted from (30). Copyright © 2011 The Physiological Society. Used with permission).

It is important to note that careful consideration must be made toward reported absolute values of plasma [ATP] because of differences in technical measurement (luciferin-luciferase assay vs high-performance liquid chromatography), sample location (e.g., intravascular microdialysis, large and small vessels), and processing (e.g., preservation or “stop” solutions, immediate vs delayed measurement) (17). These discrepant absolute values are apparent in the data reproduced in Figure 1. Also, in these in vivo human studies, samples were obtained from vessels either upstream (arterial) or downstream (venous) of the skeletal muscle microcirculation where intravascular [ATP] would be greatest and impact vascular tone. Despite these considerations, within a given study, it is well established that muscle contractions and systemic hypoxia increase venous plasma [ATP] draining skeletal muscle in humans (4,13–16,25,30).


For intravascular ATP to have a role in vascular control during mismatches in oxygen delivery and demand, cellular sources of ATP must be able to release ATP during these stimuli at the required location of the microcirculation. Potential candidates for the source(s) of increased intravascular [ATP] during exercise and hypoxia are discussed later.

Extravascular Sources: Sympathetic Nerves and Skeletal Muscle Cells

During systemic hypoxia and moderate- to high-intensity exercise, sympathetic nervous system activity is increased (20,27). ATP can be coreleased from sympathetic nerves along with norepinephrine and, thus, investigators have questioned whether the nerves may contribute to the observed increased plasma [ATP] under these conditions. In addition, interstitial [ATP] increases during exercise, and skeletal muscle itself also has been proposed to be a potential source of plasma ATP during muscle contractions (28). However, the size of ATP molecules, its rapid degradation by ectonucleotidases located on cell surfaces, and previous findings that intravascular infusion of exogenous ATP does not elevate interstitial concentrations suggest that vascular smooth muscle and endothelial cells may provide an effective barrier that prevents nerve- and/or muscle-released ATP from reaching the intravascular space (28).

Recently, we determined whether acute elevations in sympathetic nervous system activity increases venous plasma [ATP] draining skeletal muscle in humans. Lower-body negative pressure to elicit baroreflex-mediated activation of the sympathetic nervous system failed to increase venous plasma [ATP] both at rest (Fig. 2A) and during exercise (Fig. 2B) (22). Regarding skeletal muscle as a potential source, muscle contractions fail to increase [ATP] independently when blood flow is occluded (Fig. 2C, D). Thus, the collective data indicate that sympathetic nerves and skeletal muscle cells likely do not contribute to the observed increases in plasma [ATP] during exercise or other sympathoexcitatory conditions such as systemic hypoxia.

Figure 2:
Source of and stimulus for increased venous plasma (adenosine triphosphate) [ATP] during exercise. Regarding the source of intravascular ATP, 2 min of lower-body negative pressure (LBNP) at −40 mm Hg to engage the baroreflex and stimulate increased sympathetic nervous system activity did not alter venous plasma [ATP] significantly at rest (A) or during moderate-intensity (15% maximal voluntary contraction) rhythmic handgrip exercise (Ex; B). When forearm blood flow is occluded (Occl) during exercise and muscle contractions continue, [ATP] is reduced significantly from steady-state exercise (SS Ex) (C). When perfusion is prevented with occlusion before exercise, resting [ATP] does not change and exercise fails to increase [ATP] (D). Taken together, these findings suggest that the increase in venous plasma [ATP] during exercise arises from cells within or in contact with the blood. n.s. indicates not significant; *P < 0.05 versus rest; †P < 0.05 versus steady-state exercise. [Adapted from (22). Copyright © 2013 The Physiological Society. Used with permission]. Regarding the stimulus for ATP release, studies in whole-blood samples (E) and plasma measures (F) demonstrate a strong correlation between hemoglobin deoxygenation and increased ATP release. In panel E, whole-blood samples were drawn from Sprague-Dawley rats and then exposed to varying gas concentrations via solution chamber. In Panel F, blood samples were obtained from the human femoral vein at rest and during incremental knee-extensor exercise and plasma was separated. Subjects were exposed to differing systemic atmospheric conditions, including normoxia (inspiratory oxygen fraction FIO2, 21%), hypoxia (FIO2, ∼10%), hyperoxia (FIO2, 100%), and carbon monoxide ((CO) circulating FCOHb, ∼21%) breathing with normoxia to vary oxygen saturations. (Panel E reprinted from (18). Copyright © 2001 American Physiological Society. Used with permission. Panel F reprinted from (16). Copyright © 2002 American Heart Association. Used with permission).

Intravascular and Vascular Sources: Blood Cells and Endothelial Cells

Manipulation of forearm blood flow in the aforementioned study not only allowed for determination of whether the sympathetic nerves and/or skeletal muscle may be a source of ATP but also provided insight as to whether perfusion itself and, therefore, supply of erythrocytes, other blood cells, and shear stress along endothelial cells, are obligatory to observe increased venous plasma [ATP] draining skeletal muscle during exercise (22). When perfusion to actively contracting muscle is occluded, plasma [ATP] levels decline to resting values (Fig. 2C). In addition, when blood flow to a resting tissue is occluded and then muscle contractions commence, plasma [ATP] does not increase (Fig. 2D). Taken together, our findings suggest that the source of increased plasma [ATP] during exercise is dependent on perfusion and, thus, from cells within or in contact with the blood. In this context, isolated erythrocytes release significant amounts of ATP in response to a variety of stimuli, which we discuss in further detail in the following section (1,11). Similarly, vascular endothelial cells, the monolayer of epithelial cells capable of releasing a variety of vasoactive substances in isolation, can release ATP under similar conditions (2). Although the specific cell source of increased plasma [ATP] draining skeletal muscle remains somewhat elusive, collective evidence suggests that it is within or in contact with the intravascular space.


For ATP to be a candidate molecule involved in the regulation of vascular tone during muscle contractions or systemic hypoxia, specific stimuli for ATP release must be present during these physiological stresses. To date, a number of stimuli have been proposed to increase intravascular [ATP] under these conditions.

Changes in Blood Milieu

Repeated muscle contractions evoke an increase in metabolism, which results in greater oxygen consumption, the production of CO2, and acidosis (4,6,28). Classic physiology studies manipulating hemoglobin concentrations demonstrated that muscle blood flow is more directly dependent on changes in oxygen content rather than changes in the partial pressure of oxygen (PO2) (31). In this context, it is important to note that both in vitro data in whole-blood samples (Fig. 2E) (18) and plasma measurements of [ATP] from humans (Fig. 2 F) (16) show a strong relationship between increased ATP release and hemoglobin deoxygenation. Although declines in PO2 that accompany exercise and hypoxia may stimulate endothelial cell release of ATP independent of erythrocytes (2), there is evidence demonstrating that isolated resistance vessels fail to dilate in response to hypoxia without the presence of erythrocytes (8), thus it is unlikely that the endothelium is the primary site of ATP release.

In addition to deoxygenation that occurs with muscle contractions or systemic hypoxia exposure, changes in pH, particularly acidosis, also can stimulate ATP release from red blood cells (11). Furthermore, as a result of repeated muscle contractions, increases in blood temperature can occur and this also may contribute to ATP release during exercise (21). The specific ATP release pathways from erythrocytes and other cell sources and the various stimuli for these processes continue to be a topic of interest. Taken together, the combined local metabolic milieu resultant from exercise or hypoxia is a stimulus for increased ATP release from intravascular sources and would appropriately increase with greater exercise intensity or duration of these conditions of oxygen mismatch, as do ATP levels (Fig. 1A, B).

Mechanical Stresses

In addition to the changes in metabolic milieu associated with exercise that have been shown to stimulate ATP release, mechanical stimuli during muscle contraction also have been associated with increased ATP release. It is well known that, as erythrocytes traverse the microcirculation, they undergo mechanical deformation and this has been shown experimentally to stimulate ATP release (11). As skeletal muscles contract, the elevation in extravascular pressure causes compression or distortion of the resistance vessels, thus exposing the erythrocytes to even greater mechanical stress during exercise. Endothelial cells also experience increased mechanical stimulation during exercise in the form of greater shear stress because of elevated blood flow and mechanical distortion as a result of contracting tissue. Studies in vitro demonstrate that these factors increase endothelial cell ATP release (2). Along these lines, we and others have shown that mechanical stimulation to mimic the compressive forces of a muscle contraction (via rhythmic inflation and deflation of a blood pressure cuff) increases venous plasma [ATP] (4,15). It should be noted that in vivo human models cannot determine the cell-specific source of ATP in these conditions because it is not possible to cause mechanical distortion of erythrocytes or endothelial cells selectively. However, when vasodilators are infused into the brachial or femoral artery and increase blood flow and shear stress along endothelial cells without changing oxygenation, [ATP] does not increase in the venous effluent (25,30). Thus, it does not appear that shear-mediated endothelial cell release largely contributes to the increase in plasma [ATP] draining skeletal muscle observed during exercise and/or hypoxia.


Historically, when attempting to discern the role for a given substance or signaling cascade in vascular control, pharmacological antagonists or physiological maneuvers are used to inhibit the source or action of the substance in question and determine the possible impact on the regulation being studied, for instance, exercise hyperemia. One of the significant challenges that we and others have faced in our investigations of ATP is the lack of an appropriate pharmacological antagonist to inhibit ATP binding to its respective purinergic receptors (P2). Even in animal models, specific pharmacology is limited and, thus, the data in support of our overall hypothesis of vascular regulation are derived from physiological and experimental manipulations of the previously discussed stimuli, as well as the unique signaling pathways and vasomotor properties of ATP as described later.

Potent Vasodilation

The potential role for extracellular ATP as a vasoactive molecule in humans was described initially in the mid–20th century when observations of increased blood flow were made after exogenous ATP intra-arterial infusion (9). As compared with other purine compounds such as adenosine, the potency of ATP is robust, causing significant dilation that mimics levels achieved during maximal exercise (Fig. 3A) (32).

Figure 3:
Vasomotor responses of intravascular adenosine triphosphate (ATP) in humans. (A) Intra-arterial (femoral) infusion of exogenous ATP (black line) is capable of causing profound vasodilation (increased vascular conductance), similar to that which occurs with maximal exercise (gray line; incremental cycling exercise). *P < 0.05 versus rest. [Adapted from (32). Copyright © 2004 The Physiological Society. Used with permission]. (B) Exogenous ATP is capable of modulating sympathetically mediated (phenylephrine (PE), α1-adrenergic agonist) vasoconstriction (decreased vascular conductance), similar to that which occurs during moderate-intensity handgrip exercise (15% maximum voluntary contraction). Here, adenosine (white bars) is used as a high-flow control vasodilator in quiescent tissue and is not sympatholytic. *P < 0.05 versus Pre-PE; †P < 0.05 versus adenosine. (Data from (26). Copyright © 2008 The Physiological Society. Used with permission).

Modulation of Sympathetic Vasoconstriction

In addition to the vasodilator properties of ATP, this nucleotide also is able to modulate postjunctional sympathetically mediated vasoconstriction, a property unique among exogenous vasodilator substances in humans (Fig. 3B) (24,26,32). Our working hypothesis is that ATP binds to P2 receptors on the endothelium, and this leads to hyperpolarization of endothelial and vascular smooth muscle cells that, in turn, limits sympathetic vasoconstriction. Importantly, the ability to limit postjunctional sympathetic vasoconstriction is a significant phenomenon that occurs in actively contracting skeletal muscle and is known as functional sympatholysis. Given the profound vasodilator capacity of the skeletal muscle vasculature, vasoconstriction even within the active muscle is needed to prevent a decline in total peripheral resistance and thus maintain arterial blood pressure. In this manner, functional sympatholysis permits increased blood flow and oxygen delivery to the active tissue to support increased metabolism. It is, therefore, critical that intravascular ATP possesses dual vasomotor properties in that it can facilitate hyperemia by causing direct vasodilation during exercise and also act to limit the amount of sympathetic vasoconstriction, thereby preserving adequate blood flow to the active tissue during conditions of sympathoexcitation.


Early in vitro data demonstrated that ATP stimulated vasodilation via an endothelium-dependent mechanism; however, downstream obligatory signaling pathways in humans remained uncertain for some time. In humans, a variety of investigations explored whether the endothelial-derived autocoids nitric oxide (NO) and prostaglandins (PG) explained ATP vasodilation, with equivocal results being found, even in our own laboratory, depending on the timing of inhibition and method of blood flow measurement (5). On critical review of the collective data, it seems that up to 20% of the vasodilation stimulated by ATP may be caused by NO and PG in humans.

In contrast, intra-arterial infusion of barium chloride to inhibit inwardly rectifying potassium (KIR) channels reduces ATP-mediated dilation approximately 50% (Fig. 4A) (3). Activation of KIR channels is understood to hyperpolarize vascular cells (endothelial and/or smooth muscle) directly (10) as well as amplify hyperpolarization signals originating from adjacent cells (19). Importantly, vascular hyperpolarization is the essential underpinning for conducted vasodilation or, specifically, the ability for electrical signals to spread rapidly throughout the vasculature and cause profound vasodilation (33). Taken together, these data are consistent with the hypothesis that intravascular ATP and the ensuing signaling cascade is a robust regulator of vascular tone.

Figure 4:
Similar downstream signaling pathways of exogenous adenosine triphosphate (ATP) and exercise in humans. (A) Exogenous ATP-mediated vasodilation (increased vascular conductance) in the forearm primarily occurs (∼50%) via activation of inwardly rectifying potassium (KIR) channels as determined via intra-arterial barium chloride (BaCl2) to inhibit KIR channels *P < 0.05 versus control. (Reprinted from (3). Copyright © 2012 The Physiological Society. Used with permission). Similarly, as seen in panel B, a significant portion (∼30%) of the hyperemic response to mild-intensity (10% maximal voluntary contraction) rhythmic handgrip exercise is attenuated with infusion of BaCl2, indicating a significant role for KIR channels in this vasomotor response. This inhibition occurs both in the rapid response to muscle contractions at the onset of exercise as well as during steady-state hyperemic conditions. Additional inhibition of Na+-K+-ATPase via intra-arterial ouabain does not further impact exercise hyperemia. (Reprinted from (7). Copyright © 2014 American Physiological Society. Used with permission).

Recently, we have inhibited KIR channels during muscle contractions in humans (7). In our model of forearm exercise, there is no impact on hyperemia during steady-state exercise when the synthesis of NO and PG are antagonized; however, inhibiting KIR channels significantly reduces exercise hyperemia by approximately 30% (Fig. 4B). The magnitude of this effect is profound and to date represents the largest impact on forearm exercise hyperemia via pharmacological antagonists of single or multiple vasodilator pathways (20).

Consistent with the significant impact on exercise hyperemia of inhibiting KIR channels, hyperpolarization and resultant conducted dilation may be crucial to the robust vasodilation and modulation of sympathetic vasoconstriction that occurs in the microvasculature during muscle contraction (33). Taken together, activation of KIR channels resulting in hyperpolarization of the vasculature seems to be crucial to vasomotor regulation during exercise and, importantly, ATP signals via this mechanism.

Regarding hypoxic vasodilation, we have explained local mechanisms of hypoxic vasodilation in humans via inhibition of NO and PG previously (27). Our current studies are attempting to determine whether hyperpolarizing pathways also are involved in the hypoxic response, particularly that which occurs when hypoxia is combined with exercise, which cannot be attributed completely to NO derived from NO synthase and PG (6). Given that some data indicate that a portion of ATP-mediated vasodilation can signal via NO and PG, it still remains a possibility that ATP may mediate the rise in blood flow that allows for perfusion matching in circumstances of decreased oxygen supply.


Thus far, we have built support for the hypothesis that ATP is involved in vascular control during mismatches of oxygen delivery and demand in young healthy humans. As depicted in Figure 5A, exercise increases tissue oxygen demand, whereas hypoxia decreases tissue oxygen supply. In addition, during exercise, CO2 increases and acidosis occurs. These stimuli, along with exercise-induced mechanical factors and increases in blood temperature can serve as stimuli for ATP release from intravascular cell sources. Circulating extracellular ATP then signals for increased vasodilation and blunts sympathetically mediated vasoconstriction. Both actions increase red blood cell supply and therefore oxygen delivery to the tissue in need. The net hyperemic response works in a negative feedback manner to limit the original oxygen deficit and, thus, a steady-state homeostasis is reached until exercise or hypoxic exposure changes or ceases.

Figure 5:
Intravascular adenosine triphosphate (ATP) as a regulator of vascular tone during conditions of mismatched oxygen supply and demand. Overall schematic of our working hypothesis on the role of ATP in vascular control. A. In young healthy individuals, during exercise, oxygen demand is increased, whereas during hypoxia, oxygen supply is diminished. Both instances result in a mismatch that decreases the fraction of oxyhemoglobin (FO2Hb), and exercise increases CO2, resulting in acidosis. These conditions, along with exercise-induced mechanical factors and increases in blood temperature, serve as stimuli for ATP release from cell sources within or in contact with the blood. Circulating plasma ATP then signals for increased vasodilation and blunts sympathetically mediated vasoconstriction. Both actions increase red blood cell supply and therefore oxygen delivery to the tissue in need. The net hyperemic response thus works in a negative feedback manner (red dashed line) to limit the original oxygen deficit and reach a steady-state homeostasis until exercise or hypoxic exposure changes or ceases. B. In older or diseased populations, failure to increase [ATP] adequately compromises this regulation, leading to exacerbated mismatches (positive feedback indicated by solid blue line) in oxygen delivery to demand, which can result in relative ischemia.

Investigations to date have focused largely on the role of extracellular ATP in vascular control of young healthy humans. As such, studies in older individuals (older than 60 yr) and patient populations to determine vasomotor responsiveness to ATP and/or the level of extracellular ATP release into the bloodstream are limited, despite these populations being at an increased risk for cardiovascular morbidity and mortality. Nevertheless, evidence from our laboratory demonstrates an intact vasodilator responsiveness and sympatholytic capacity of exogenous ATP at rest in the forearm of aged humans despite the presence of classic endothelial dysfunction (23,24). Similarly, in the leg vasculature of diabetic humans, both dilator responsiveness and sympatholytic capacity of exogenous ATP seem to be largely preserved (35). It should be noted that preserved vasodilator capacity to ATP with advanced age is not a universal finding (29); however, at present, we interpret the existing data to indicate that the net vasomotor responses to intravascular ATP generally remain intact with age and in certain disease populations.

With regard to measuring endogenous plasma [ATP] in high-risk humans at rest or during mismatched oxygen supply and demand conditions, few studies have been published. In a recent study from our laboratory, we demonstrated that older healthy humans have reduced blood flow because of impaired local vasodilation during both graded handgrip exercise and systemic hypoxia relative to young adults, and this was associated with impaired increases in plasma [ATP] with age (Fig. 6) (25). Furthermore, we demonstrated that elevated ATP catabolism during the stimulus was not responsible for the low plasma [ATP] values in older adults, but rather that isolated erythrocytes from this population fail to release a significant quantity of ATP in response to hemoglobin deoxygenation at levels observed during muscle contraction and systemic hypoxia (Fig. 7A). Similarly, erythrocytes obtained from type II diabetic patients fail to release ATP in response to deoxygenation (Fig. 7B) (34). Impaired erythrocyte release of ATP also has been observed in pulmonary hypertension, cystic fibrosis, and sickle cell patients, all conditions typically associated with dysfunctional vasomotor control at rest and presumably during oxygen mismatch, although the latter has not been determined completely (11).

Figure 6:
Older individuals demonstrate impaired vascular control and are unable to increase plasma (adenosine triphosphate) [ATP] during conditions of oxygen mismatch. Relative to young individuals, older individuals exhibit attenuated hyperemia (A) and fail to increase [ATP] (B) during progressive rhythmic handgrip exercise (% maximal voluntary contraction (%MVC)). Similarly, older adults have attenuated hyperemic responses (C) and fail to increase [ATP] significantly (D) in response to systemic hypoxic exposure (inspired O2 fraction, ∼10%; hemoglobin O2 saturation, ∼80%). *P < 0.05 versus rest/normoxia; †P <0.05 versus young; ‡P < 0.05 versus 5% and 15% within young. [Adapted from (25). Copyright © 2002 American Heart Association. Used with permission).
Figure 7:
Isolated erythrocytes from older and diseased populations fail to release adenosine triphosphate (ATP) in low oxygen conditions. A. Isolated erythrocytes obtained from young individuals release ATP on exposure to low oxygen gas, whereas this does not occur in cells obtained from older individuals. *P < 0.05 versus normoxia (partial pressure of oxygen (PO2), ∼110 mm Hg); †P < 0.05 versus young. (Reprinted from (25). Copyright © 2012 American Heart Association. Used with permission). B. Similarly, hypoxia-induced ATP release is attenuated in erythrocytes obtained from individuals with type II diabetes mellitus as compared with healthy control subjects. *P < 0.05 versus 15% O2; †P <0.05 versus 15% and 4.5% O2. [Adapted from (34). Copyright © 2012 American Physiological Society.] In both studies, tonometry was used to expose isolated erythrocytes to varied O2 concentrations, 6% CO2, and N2 balance.

Collectively, we propose the overall scheme (Fig. 5B) that, in older and diseased individuals, the failure to increase intravascular [ATP] adequately partially explains attenuated exercise and hypoxic vasodilation (24,25) as well as an impaired ability to modulate sympathetic vasoconstriction (24) in these populations. Whereas, in young healthy adults, ATP functions to assist matching of oxygen delivery to demand in a homeostatic function, this feedback control system is defective in aged or diseased individuals and fails to regulate blood flow and oxygen supply, thus allowing the oxygen mismatch to persist or become exacerbated.


The regulation of muscle blood flow during conditions of mismatched oxygen delivery and demand is a complex interaction of a variety of factors, including neuronal signals and local chemical and mechanical stimuli. Here, we have reviewed recent literature that suggests that ATP may be an important local signaling molecule in this regard as it fits the aforementioned criteria for likely candidates of vascular control during physiological stress. First, exercise and hypoxia evoke measureable elevations in skeletal muscle plasma [ATP]. This ATP is expected to be of intravascular origin and is located appropriately to increase blood flow and oxygen delivery to tissues of metabolic need. Although not discussed extensively in this review, several inactivation mechanisms are in place to regulate plasma ATP concentrations (e.g., ectonucleotidases) and, thus, finely control the resulting vasoactive action. Second, exogenous administration of ATP mimics the predicted responses of exercise specifically in terms of robust vasodilation and the ability to modulate sympathetic vasoconstriction. Third, although specific and selective inhibition of intravascular ATP and the concomitant vascular signaling during exercise remains difficult, pharmacology known to blunt ATP-induced hyperemia significantly in quiescent muscle correspondingly results in profound attenuation of exercise hyperemia in humans (∼30%). Furthermore, we propose that because of their impaired ability to release ATP, older healthy and diseased humans exhibit compromised vascular control during cases of mismatches in oxygenation and, in turn, this leads to further impairments in oxygen delivery. Future research should be aimed at determining therapeutic interventions to improve ATP release and increase intravascular [ATP] in these at-risk populations. The resulting normalized vascular control may mitigate the current elevated risk of cardiovascular mortality and acute cardiovascular events in these populations.

We thank the current and past members of the Human Cardiovascular Physiology Laboratory at Colorado State University for their efforts on these projects as well as the subjects who volunteered to participate. We also thank the countless investigators who have contributed to the ideas presented here.

Funding: National Institutes of Health Grants: AG027150, HL087952, HL095573, AG022337, and HL102720 (F.A. Dinenno) and Colorado State University Monfort Professorship (F.A. Dinenno).

Conflicts of Interest: None.


1. Bergfeld GR, Forrester T. Release of ATP from human erythrocytes in response to a brief period of hypoxia and hypercapnia. Cardiovasc. Res. 1992; 26 (1): 40–7.
2. Bodin P, Burnstock G. Synergistic effect of acute hypoxia on flow-induced release of ATP from cultured endothelial cells. Experientia. 1995; 51 (3): 256–9.
3. Crecelius AR, Kirby BS, Luckasen GJ, Larson DG, Dinenno FA. ATP-mediated vasodilatation occurs via activation of inwardly-rectifying potassium channels in humans. J. Physiol. 2012; 590 (21): 5349–59.
4. Crecelius AR, Kirby BS, Richards JC, Dinenno FA. Mechanical effects of muscle contraction increase intravascular ATP draining quiescent and active skeletal muscle in humans. J. Appl. Physiol. 2013; 114 (8): 1085–93.
5. Crecelius AR, Kirby BS, Richards JC, et al. Mechanisms of ATP-mediated vasodilation in humans: modest role for nitric oxide and vasodilating prostaglandins. Am. J. Physiol. Heart. Circ. Physiol. 2011; 301 (4): H1302–10.
6. Crecelius AR, Kirby BS, Voyles WF, Dinenno FA. Augmented skeletal muscle hyperaemia during hypoxic exercise in humans is blunted by combined inhibition of nitric oxide and vasodilating prostaglandins. J. Physiol. 2011; 589 (Pt.14): 3671–83.
7. Crecelius AR, Luckasen GJ, Larson DG, Dinenno FA. KIR channel activation contributes to onset and steady-state exercise hyperemia in humans. Am. J. Physiol. Heart. Circ. Physiol. 307 (5): H782–91.
8. Dietrich HH, Ellsworth ML, Sprague RS, Dacey RG Jr. Red blood cell regulation of microvascular tone through adenosine triphosphate. Am. J. Physiol. Heart. Circ. Physiol. 2000; 278 (4): H1294–8.
9. Duff F, Patterson GC, Shepherd JT. A quantitative study of the response to adenosine triphosphate of the blood vessels of the human hand and forearm. J. Physiol. 1954; 125 (3): 581–9.
10. Edwards G, Dora KA, Gardener MJ, Garland CJ, Weston AH. K+ is an endothelium-derived hyperpolarizing factor in rat arteries. Nature. 1998; 396 (6708): 269–72.
11. Ellsworth ML, Ellis CG, Goldman D, Stephenson AH, Dietrich HH, Sprague RS. Erythrocytes: oxygen sensors and modulators of vascular tone. Physiology (Bethesda). 2009; 24: 107–16.
12. Feigl EO. Coronary physiology. Physiol. Rev.. 1983; 63 (1): 1–205.
13. Forrester T, Lind AR. Identification of adenosine triphosphate in human plasma and the concentration in the venous effluent of forearm muscles before, during and after sustained contractions. J. Physiol. 1969; 204 (2): 347–64.
14. Gonzalez-Alonso J, Mortensen SP, Dawson EA, Secher NH, Damsgaard R. Erythrocytes and the regulation of human skeletal muscle blood flow and oxygen delivery: role of erythrocyte count and oxygenation state of haemoglobin. J. Physiol. 2006; 572 (Pt 1): 295–305.
15. Gonzalez-Alonso J, Mortensen SP, Jeppesen TD, et al. Haemodynamic responses to exercise, ATP infusion and thigh compression in humans: insight into the role of muscle mechanisms on cardiovascular function. J. Physiol. 2008; 586 (9): 2405–17.
16. Gonzalez-Alonso J, Olsen DB, Saltin B. Erythrocyte and the regulation of human skeletal muscle blood flow and oxygen delivery: role of circulating ATP. Circ. Res. 2002; 91 (11): 1046–55.
17. Gorman MW, Feigl EO, Buffington CW. Human plasma ATP concentration. Clin. Chem. 2007; 53 (2): 318–25.
18. Jagger JE, Bateman RM, Ellsworth ML, Ellis CG. Role of erythrocyte in regulating local O2 delivery mediated by hemoglobin oxygenation. Am. J. Physiol. Heart Circ. Physiol. 2001; 280 (6): H2833–9.
19. Jantzi MC, Brett SE, Jackson WF, Corteling R, Vigmond EJ, Welsh DG. Inward rectifying potassium channels facilitate cell-to-cell communication in hamster retractor muscle feed arteries. Am. J. Physiol. Heart Circ. Physiol. 2006; 291 (3): H1319–28.
20. Joyner MJ, Wilkins BW. Exercise hyperaemia: is anything obligatory but the hyperaemia? J. Physiol. 2007; 583 (Pt 3): 855–60.
21. Kalsi KK, Gonzalez-Alonso J. Temperature-dependent release of ATP from human erythrocytes: mechanism for the control of local tissue perfusion. Exp. Physiol. 2012; 97 (3): 419–32.
22. Kirby BS, Crecelius AR, Richards JC, Dinenno FA. Sources of intravascular ATP during exercise in humans: critical role for skeletal muscle perfusion. Exp. Physiol. 2013; 98 (5): 988–98.
23. Kirby BS, Crecelius AR, Voyles WF, Dinenno FA. Vasodilatory responsiveness to adenosine triphosphate in ageing humans. J. Physiol. 2010; 588 (Pt 20): 4017–27.
24. Kirby BS, Crecelius AR, Voyles WF, Dinenno FA. Modulation of postjunctional alpha-adrenergic vasoconstriction during exercise and exogenous ATP infusions in ageing humans. J. Physiol. 2011; 589 (Pt 10): 2641–53.
25. Kirby BS, Crecelius AR, Voyles WF, Dinenno FA. Impaired skeletal muscle blood flow control with advancing age in humans: attenuated ATP release and local vasodilation during erythrocyte deoxygenation. Circ. Res. 2012; 111 (2): 220–30.
26. Kirby BS, Voyles WF, Carlson RE, Dinenno FA. Graded sympatholytic effect of exogenous ATP on postjunctional alpha-adrenergic vasoconstriction in the human forearm: implications for vascular control in contracting muscle. J. Physiol. 2008; 586 (Pt. 17): 4305–16.
27. Markwald RR, Kirby BS, Crecelius AR, Carlson RE, Voyles WF, Dinenno FA. Combined inhibition of nitric oxide and vasodilating prostaglandins abolishes forearm vasodilatation to systemic hypoxia in healthy humans. J. Physiol. 2011; 589 (Pt. 8): 1979–90.
28. Mortensen SP, Gonzalez-Alonso J, Nielsen JJ, Saltin B, Hellsten Y. Muscle interstitial ATP and norepinephrine concentrations in the human leg during exercise and ATP infusion. J. Appl. Physiol. 2009; 107 (6): 1757–62.
29. Mortensen SP, Nyberg M, Winding K, Saltin B. Lifelong physical activity preserves functional sympatholysis and purinergic signalling in the ageing human leg. J. Physiol. 2012; 590 (Pt. 23): 6227–36.
30. Mortensen SP, Thaning P, Nyberg M, Saltin B, Hellsten Y. Local release of ATP into the arterial inflow and venous drainage of human skeletal muscle: insight from ATP determination with the intravascular microdialysis technique. J. Physiol. 2011; 589 (Pt. 7): 1847–57.
31. Roach RC, Koskolou MD, Calbet JA, Saltin B. Arterial O2 content and tension in regulation of cardiac output and leg blood flow during exercise in humans. Am. J. Physiol. 1999; 276 (2 Pt. 2): H438–45.
32. Rosenmeier JB, Hansen J, Gonzalez-Alonso J. Circulating ATP-induced vasodilatation overrides sympathetic vasoconstrictor activity in human skeletal muscle. J. Physiol. 2004; 558 (Pt. 1): 351–65.
33. Segal SS. Regulation of blood flow in the microcirculation. Microcirculation. 2005; 12 (1): 33–45.
34. Sprague RS, Goldman D, Bowles EA, et al. Divergent effects of low-O(2) tension and iloprost on ATP release from erythrocytes of humans with type 2 diabetes: implications for O(2) supply to skeletal muscle. Am. J. Physiol. Heart Circ. Physiol. 2010; 299 (2): H566–73.
35. Thaning P, Bune LT, Zaar M, Saltin B, Rosenmeier JB. Functional sympatholysis during exercise in patients with type 2 diabetes with intact response to acetylcholine. Diabetes Care. 2011; 34 (5): 1186–91.

blood flow; hypoxia; exercise; purine; hyperemia

Supplemental Digital Content

© 2015 American College of Sports Medicine