In this study, we have systematically evaluated the signaling mechanisms underlying stimulated nitric oxide (NO) synthesis by estrogen (E2) and other vasoactive agents at the level of a single endothelium-derived cell. To do so, we have characterized and contrasted rapid E2-evoked NO synthesis with that of ATP using single-cell microfluorimetry and patch-clamp recordings to monitor stimulated changes in cellular NO synthesis (via 4-amino-5-methylamino-2′,7′-difluorofluorescein), Ca2+ transients (via Fluo-3), and membrane hyperpolarization in cultured human EA.hy926 cells. E2-evoked NO synthesis in single cells (EC50 ∼0.3 nM) was blocked by the E2 receptor antagonist ICI 182,780 and the NO synthase inhibitor Nω-nitro-l-arginine methyl ester. Although both E2 and ATP stimulated comparable Ca2+ transients, E2-induced NO synthesis was insensitive to intracellular BAPTA-AM or removal of external Ca2+. In contrast, ATP-evoked NO production was abolished by either one of these treatments. ATP-evoked hyperpolarizations (∼20 mV) and NO production were both inhibited by the respective small-conductance and intermediate-conductance calcium- activated K+ channel blockers apamin and charybdotoxin. E2 minimally affected membrane potential, and stimulated NO synthesis was insensitive to calcium-activated K+ channel blockers. Exposure to either the phosphatidylinositol 3-kinase inhibitor LY-294001 or the MAP kinase inhibitor PD-98059 abolished the NO response to E2, but not that to ATP. Finally, the NO response evoked by a combined stimulus of E2 plus ATP was similar to that of ATP alone. In conclusion, our data directly demonstrate that an individual human EA.hy926 cell contains at least two distinct mechanisms for stimulated NO synthesis that depend on either calcium or protein kinase signaling events.
- nitric oxide
- endothelial function
- signal transduction
estrogen promotes a number of beneficial effects within the cardiovascular system that include vasodilatation, prevention of leukocyte adhesion, angiogenesis, and antiproliferation of vascular smooth muscle (34, 35). Clinically, the loss of estrogen-mediated vasorelaxation in postmenopausal women correlates closely with increased incidence of cardiovascular events such as coronary heart disease and stroke (24). A key action of estrogen is to enhance nitric oxide (NO) production in vascular endothelium, which may occur via transcriptional and nontranscriptional mechanisms (6). Along this line, several recent studies have highlighted the importance of the phosphatidylinositol 3-kinase (PI3-kinase)/Akt signaling cascade to acute NO synthesis evoked by endothelial stimuli such as estrogen and shear stress (9, 15, 20, 22, 43). Taken together with earlier work demonstrating that endothelial NO production can be readily stimulated by calcium-mobilizing agonists such as ATP, histamine, and acetylcholine (3), these studies strongly suggest the existence of multiple cellular pathways for stimulated NO synthesis in vascular endothelium. Biochemical studies demonstrating that endothelial NO synthase (eNOS) itself may be regulated via Ca2+/calmodulin or by direct phosphorylation at Ser1177 (16) have further implied that both these pathways may operate within the same cell. This assumption, however, has not been rigorously examined using direct experimental measurements.
Given the physiological importance of both estrogen and Ca2+-mobilizing agonists to endothelial NO production, the goal of our study was to compare and contrast the cellular mechanism(s) underlying the NO response to each type of stimulus in the same endothelial cell. To do so, we have developed experimental approaches to monitor in real time evoked changes in NO synthesis, cytosolic free Ca2+, and membrane potential in single human EA.hy926 cells. Our findings provide direct evidence demonstrating for the first time that a single indentified endothelial cell is capable of producing NO in response to both Ca2+-mobilizing agonists and stimuli that activate eNOS via the PI3-kinase/Akt and MAP kinase signaling cascades. Our data further suggest that these two distinct cellular mechanisms operate in parallel and converge at the level of eNOS itself. Vascular endothelial cells thus appear to have evolved at least two separate stimulatory pathways leading to de novo NO synthesis, which may reflect the critical importance of this multipotent mediator in the long-term regulation of vascular tone and overall function of the circulatory system.
MATERIALS AND METHODS
Cell culture and fluorescence measurements.
The EA.hy926 cell line (10), originally derived from human umbilical vein endothelium, was cultured and loaded with the membrane-permeable forms of the fluorescent dyes 4-amino-5-methylamino-2′,7′-difluorofluorescein (DAF-FM) or Fluo-3, as recently described (41). Fluorescence measurements were performed in a ∼0.3-ml bath chamber mounted on the stage of a Nikon TE300 inverted microscope equipped with a 75 W xenon arc lamp and SFX-1 microfluorimeter (Solamere Technology Group). Both DAF-FM and Fluo-3 fluorescence signals were measured using excitation and emission band-pass filters centered on 488 and 515 nm, respectively; data were acquired using AxoScope software and analyzed with pClamp 7 and SigmaPlot 2000 software suites. Because the fluorescent intensity of the triazole or NO-bound form of DAF-FM originating from a single cell was typically quite modest, the strong excitation light needed to observe reliable fluorescent signals often resulted in some photobleaching of the NO-modified form of DAF-FM during continuous cell illumination. Exposure of the cell to intermittent illumination through the use of a timer-driven, optical shutter reduced, but did not completely eliminate, photobleaching of NO-modified DAF-FM. A manually controlled diaphragm was used to restrict the region of light collection to the single cell of interest.
Current-clamp recordings of membrane voltage were performed using perforated patch-clamp methodology in combination with an Axopatch 200B amplifier, Digidata 1200B analog/digital interface, and Clampex 8 software; electrical signals were typically sampled at 1 Hz. Borosilicate glass micropipettes (2–4 MΩ tip resistance) were first briefly dipped into standard filling solution (final concentrations in mM: 100 K-aspartate, 30 KCl, 1 MgCl2, 2 Na2-ATP, and 10 HEPES, pH 7.2 with 1 M KOH) and then back-filled with the same filling solution containing nystatin (50 μg/ml final). Following nystatin-mediated perforation, measured cell membrane capacitance and intracellular access resistance ranged from 14 to 20 pF and 17 to 25 MΩ, respectively. The bath solution for both fluorescence and electrophysiological recordings contained (in mM) 135 NaCl, 5 KCl, 1 MgCl2, 1.5 CaCl2, and 10 HEPES, pH 7.4 with 1 M NaOH. For the Ca-free solution, CaCl2 was omitted and replaced by 2 mM EGTA. Cells in the bath chamber were constantly superfused at ∼1 ml/min, and solution changes were performed by gravity flow from a series of elevated solution reservoirs using manually controlled solenoid valves. All fluorescence and electrophysiological recordings were performed at 35°C.
Apamin, charybdotoxin, estrogen (17β-estradiol), ICI 182,780, LY-294001, LY-303511, PD-98059, histamine dihydrochloride, Na2-ATP, along with the chemicals used to prepare physiological saline solutions (American Chemical Society grade or higher) were purchased from Sigma-Aldrich (St. Louis, MO). DAF-FM diacetate and Fluo-3 AM were obtained from Molecular Probes/Invitrogen (Eugene, OR).
Data are presented as means ± SE. Mean values calculated from independent experiments were statistically analyzed using Student's t-test, and differences among means were considered to be significant when P < 0.05.
Estrogen induces acute NO production in single EA.hy926l cells.
In the present study, we have used the cultured human EA.hy926 cell line as a model vascular endothelial cell to examine stimulus-driven NO production. Earlier studies have shown that these cells display Ca2+-dependent signaling events and membrane currents similar to those observed in other types of isolated vascular endothelial cells. Upon loading these cells with the NO-sensitive fluorescent dye DAF-FM diacetate (26), acute exposure to estrogen (17β-estradiol) evoked concentration-dependent increases in cellular fluorescence that were detectable in single cells within seconds of estradiol application (Fig. 1A). The rapidity of this response is consistent with a nongenomic action of estrogen and agrees with earlier studies using alternate NO detection methods that report acute estrogen-evoked NO production within minutes of exposure (5, 8, 27, 46). As depicted by the concentration-response curve in Fig. 1B, estrogen-induced increases in cellular fluorescence occurred over the range of 0.01–10 nM, with a half-maximal fluorescence rise observed at an estrogen concentration of ∼0.3 nM (EC50 value). Importantly, this observed sensitivity to estrogen in our human endothelial cell model agrees closely with the range of circulating plasma levels of estrogen observed during the menstrual cycle in premenopausal women (∼30 to 200 pg/ml or ∼0.1 to 0.8 nM), suggesting that these cells have a physiologically appropriate sensitivity to this hormone. In the presence of ICI 182,780, a competitive antagonist of the steroidal estrogen receptor α- and β-isoforms (48, 50), this observed estrogen-evoked NO synthesis was largely abolished (Fig. 1, C and D). As shown in the same experiment, however, ATP-induced NO synthesis was unaffected by ICI 182,780, demonstrating the pharmacologic selectivity of this inhibitor. Importantly, estrogen-induced increases in DAF-FM fluorescence could be blocked by brief treatment of DAF-FM-loaded cells with the NOS inhibitor Nω-nitro-l-arginine methyl ester [l-NAME (25)], which effectively interfered with the stimulated rises in NO production in response to either estrogen or the calcium-mobilizing agonist ATP (Fig. 2, A and B). Moreover, this l-NAME-mediated inhibition was readily reversible, because ATP-evoked NO synthesis could be restored following a 2- to 3-min washout of the NOS inhibitor from the bath. ATP is known to act via endothelial purinergic P2Y receptors to induce l-NAME-sensitive vasodilation in intact blood vessels (2), and we have recently described acute ATP-stimulated NO synthesis in these same human cells (41). Taken together, these observations demonstrate that estrogen acts via ICI-182,780-sensitive receptors to elicit rapid de novo synthesis of NO in single EA.hy926 cells that also display ATP-evoked NO synthesis.
Estrogen-induced NO production occurs in a Ca2+-independent manner.
It is now well recognized that elevations in cytosolic free Ca2+ concentration play a critical signaling role in the activation of eNOS by calcium-mobilizing agonists (Ref. 40 and references therein). To examine the nature of free Ca2+ transients evoked by estrogen and ATP in the same cell, we monitored changes in single-cell fluorescence in Fluo-3-loaded EA.hy926 cells in response to both agonists. As shown in Fig. 3A, estrogen evoked dose-dependent elevations in Fluo-3 fluorescence that were kinetically similar to those induced by a maximally effective concentration of ATP but did not reach the same magnitude (Fig. 3B).
A number of studies, including recent data from our own lab (40), have demonstrated that the endothelial production of NO/endothelium-derived relaxing factor (EDRF) evoked by Ca2+-mobilizing agonists, such as acetylcholine, ATP, and histamine, is strictly dependent on the presence of extracellular Ca2+ (23, 28, 29, 32). Because both estrogen and ATP were observed to produce clear Ca2+ transients in single EA.hy926 cells, we examined further the potential involvement of internal and external Ca2+ pools in estrogen-mediated NO synthesis. To elucidate a role for extracellular Ca2+, we first monitored estrogen- and ATP-evoked NO responses under control conditions and then repeated the agonist exposures in the presence of a physiological saline solution with no added Ca2+ that also contained the Ca2+ chelator EGTA (2 mM). As shown in Fig. 4A, estrogen-mediated NO production was unaffected by the rapid removal of external Ca2+, whereas the NO response to ATP was virtually abolished, in agreement with our previous observations (40). These data thus highlight a major difference in the signaling mechanisms underlying the NO response evoked by each agonist. To determine whether estrogen-stimulated changes in intracellular free Ca2+ levels per se may be more physiologically relevant to the estrogen-mediated NO production, EA.hy926 cells were simultaneously loaded with both the Ca2+ chelator BAPTA-AM (20 μM) and the NO-sensitive dye DAF-FM. Under such conditions, we observed that estrogen-evoked increases in NO production were unaffected by intracellular BAPTA loading, whereas the ATP-stimulated NO response monitored in the same cell was largely inhibited by BAPTA pretreatment (Fig. 4B). The effects of external Ca2+ removal and intracellular BAPTA loading on estrogen- and ATP-stimulated NO synthesis are summarized in Fig. 4C.
Estrogen- and ATP-induced NO synthesis display differential sensitivity to inhibition of PI3-kinase, MAP kinase, and small- and intermediate-conductance calcium-activated potassium channels.
Recently, we and others have directly demonstrated that membrane hyperpolarization mediated by endothelial small- and intermediate-conductance Ca-activated K+ channels (SKCa and IKCa, respectively) plays an essential role in the cellular mechanism underlying stimulated NO synthesis by Ca2+-mobilizing agonists, such as ATP, histamine, and acetylcholine (40, 45). Given the estrogen-induced Ca2+ transients observed in our isolated EA.hy926 cells (Fig. 3), we examined whether these same KCa channels also contribute to estrogen-mediated NO synthesis. To do so, EA.hy926 cells were exposed to apamin and charybdotoxin, selective inhibitors of SKCa and IKCa channels, respectively, following control NO responses to estrogen and ATP. In agreement with our earlier study (40), the combination of apamin and charybdotoxin abolished the ATP-evoked increase in NO production; however, the estrogen-stimulated NO response was unaffected by the presence of these two channel blockers (Fig. 5). To determine whether estrogen was able to evoke membrane potential changes via other types of ion channels, we performed current-clamp recordings of membrane voltage in single EA.hy926 cells using perforated patch methodology under conditions of estrogen and ATP application. As shown in Fig. 6A, increasing concentrations of estrogen (0.01–1 nM) produced minimal changes in membrane potential, whereas ATP and histamine evoked robust membrane hyperpolarizations in the same cell. The magnitudes of membrane potential changes evoked by estrogen, ATP, and histamine are quantified in the histogram displayed in Fig. 6B.
To examine whether the weak estrogen-evoked membrane hyperpolarizations described in Fig. 6 were related to differences in the amplitudes of estrogen- versus ATP-evoked Ca2+ transients (see Fig. 3), we simultaneously recorded evoked changes in membrane potential and Fluo-3 fluorescence in single EA.hy926 cells in response to equipotent concentrations of estrogen and ATP. As shown in Fig. 7, 1 nM estrogen and 1 μM ATP produced similar elevations in cytosolic free Ca2+, as judged by Fluo-3 fluorescence. Despite this similarity, ATP produced a considerably larger membrane-hyperpolarizing response compared with estrogen. These data thus rule out the possibility that the estrogen-induced Ca2+ elevation was below threshold for activation of membrane ion channels (i.e., SKCa and IKCa) and point to a fundamental difference in the calcium-signaling pathways stimulated by estrogen and ATP in these cells.
In vascular endothelium, protein kinase signaling cascades are known to participate in the regulation of NO synthesis evoked by a variety of stimuli, including shear stress (9, 17), growth factors (38, 51), insulin (37), and estrogen (43, 44). Moreover, phosphorylation of eNOS at Ser1177 appears to be an important activation mechanism used by several agonists to stimulate de novo NO production (7, 15, 21, 33), and modification of this site has been documented in both isolated endothelial cells and intact arterial tissue (9, 15, 17, 47). To examine the potential contributions of two major protein kinase signaling cascades, the PI3-kinase/Akt pathway and MAP kinase pathway, to both estrogen- and ATP-evoked NO production, EA.hy926 cells were exposed to either the selective PI3-kinase inhibitor LY-294002 (49) or PD-98059, a blocker of MAP kinase signaling (1). As shown in Fig. 8A, brief exposure to LY-294002 (10 μM) abolished estrogen-induced NO synthesis in a single EA.hy926 cell but did not significantly alter the NO response to ATP in the same cell. Importantly, exposure of cells to the inactive control compound LY-303511 (10 μM) had no effect on either the estrogen- or ATP-evoked increases in NO production in the same single-cell preparation. The effects of both LY-294002 and LY-303511 on stimulated NO synthesis in single EA.hy926 cells are quantified in Fig. 8B.
Because acute activation of the Raf-1/ERK1/2 signaling pathway by estrogen has also been reported to increase NO synthesis (7, 8, 21), we used the selective MEK inhibitor PD-98059 (1) to examine the potential contribution of MAP kinase signaling to estrogen-induced NO production. As shown in Fig. 9, pretreatment of EA.hy926 cells with 20 μM PD-98059 abolished estrogen-stimulated NO synthesis but did not significantly affect the ATP-induced response under the same exposure conditions (Fig. 9C).
Finally, the clear differences noted above in estrogen- and ATP-evoked cellular actions begged the question of whether these agonists induced NO production in an additive or synergistic manner. Figure 10A shows simultaneous recordings of membrane potential and DAF-FM fluorescence in a single EA.hy926 cell exposed to pharmacologically maximal concentrations of estrogen and ATP first separately and then together. As can be seen, the combination of 1 nM estrogen and 10 μM ATP did not evoke a larger increase in NO production compared with 10 μM ATP alone. We did, however, note a modest but statistically insignificant increase in the amplitude of membrane hyperpolarization evoked by estrogen plus ATP compared with ATP alone (Fig. 10B).
It is now well recognized that estrogen exerts important functional and protective effects on the human cardiovascular system (34, 35), and one of the major mechanisms underlying this protection involves enhanced production of NO by the vascular endothelium. NO is a major physiological regulator of circulatory function and health as a result of its ability to modulate blood vessel tone, vascular permeability, leukocyte and platelet adhesion, smooth muscle proliferation/migration, and atherogenesis. As part of our efforts to establish a more comprehensive picture of vascular NO synthesis and actions, we have used direct, real-time fluorescence- and electrophysiology-based measurements to compare and contrast fundamental aspects of stimulated NO production in single human EA.hy926 cells in response to estrogen and ATP, a Ca2+-mobilizing purinergic receptor agonist. EA.hy926 is a permanent cell line derived from human umbilical vein endothelium (10) that has been used in a variety of studies as a model endothelial cell. Using this novel strategy, we report for the first time direct evidence showing that at least two distinct and seemingly parallel signaling mechanisms exist for stimulated NO production within the same single EA.hy926 cell. As we have recently described (40), one such mechanism involves the mobilization of intracellular Ca2+, membrane hyperpolarization, and the Ca2+-dependent activation of eNOS via external Ca2+ entry. A second pathway, triggered by stimuli such as estrogen, shear stress, and growth factors, involves the PI3-kinase/Akt and MAP kinase signaling cascades and appears to culminate in the phosphorylation-dependent activation of eNOS.
Mechanistically, Busse and colleagues (31, 32) were the first to identify the importance of endothelial Ca2+ signaling in the production of NO or EDRF. While we observed that estrogen was capable of producing rapid intracellular Ca2+ transients in single Fluo-3-loaded EA.hy926 cells (Fig. 3), estrogen-evoked NO synthesis was not significantly altered following the acute removal of external Ca2+ by EGTA or buffering of cytosolic free Ca2+ by BAPTA (Fig. 4, A and B, respectively). This apparent lack of Ca2+ dependence of estrogen-evoked NO synthesis is thus in agreement with earlier findings (5, 20) and is consistent with an alternative (i.e., phosphorylation-dependent) signaling mechanism for eNOS activation by estrogen. Under some experimental conditions, however, intracellular Ca2+ buffering has been reported to interfere with the estrogen-evoked NO response (27, 46). In contrast, ATP-induced NO production monitored in the same cells following estrogen removal was abolished by either external Ca2+ removal or introduction of an intracellular Ca2+ chelator, in agreement with earlier reports and observations from our own lab (23, 28, 29, 32, 40). Moreover, Ca2+ transients evoked by either ATP or histamine were associated with large membrane hyperpolarizations, whereas estrogen stimulation evoked very modest changes in membrane potential, despite producing clear elevations in cytosolic free Ca2+. This observation held true even under conditions in which estrogen- and ATP-stimulated Ca2+ transients were of equal magnitude (see Fig. 7), indicating that the estrogen-evoked Ca2+ elevation could not be below the calcium threshold required to activate membrane ion channels. One likely explanation for this somewhat paradoxical set of observations is the fact that second messenger signaling may be “compartmentalized” within a cell, occurring within restricted compartments or microdomains. For example, the compartmentalization of agonist-stimulated cyclic AMP production in cardiac myocytes has been shown to underlie the differential effects of agonists on cellular contractility and metabolism (4, 30). More recently, several calcium-mobilizing hormones have been shown to induce differential activation of Ca2+-activated K+ currents and membrane potential responses in isolated vascular endothelial cells (11–13, 19), data that are consistent with our own observations presented in Figs. 6 and 7. Collectively, these findings support the concept of functionally distinct microdomains of intracellular Ca2+ signaling. Such compartmentalization of Ca2+ signals (39) could thus readily explain how estrogen-evoked Ca2+ transients in one cellular compartment may be unable to activate membrane ion channels and alter membrane potential compared with Ca2+ elevations produced by ATP and histamine that likely occur in a distinct compartment (i.e., subplasmalemmal domain). This last point remains quite speculative, because Fluo-3 typically reports global changes in cytosolic free Ca2+, and we did not have the necessary spatial resolution in our single-cell measurements to detect the absence or presence of Ca2+ microdomains within defined regions of a cell. The lack of association between stimulated Ca2+ transients and membrane hyperpolarization in the case of estrogen, but not ATP, argues nonetheless for differential Ca2+ signaling by each agonist. If estrogen-induced Ca2+ transients do not lead to changes in membrane potential, could they modulate a different physiological response? One possibility is that such Ca2+ elevations underlie rapid estrogen-stimulated prostacyclin synthesis (42) or the Ca2+-dependent cellular redistribution of eNOS induced by acute estrogen exposure (18).
Our finding that estrogen-evoked NO production could be prevented by the PI3-kinase inhibitor LY-294002, but not by the inactive control compound LY-303511 (Fig. 8), is consistent with studies demonstrating that the PI3-kinase/Akt signaling cascade is capable of directly activating eNOS via phosphorylation at Ser1177 in both isolated endothelial cells (9, 15, 17) and intact arterial vessels (47). Similarly, the observed inhibition of the estrogen-induced NO response by the selective MEK inhibitor PD-98059 (Fig. 9) is in agreement with earlier reports (7, 8, 21) and further indicates that MAP kinase signaling contributes to stimulated NO production in human EA.hy926 cells. The fact that inhibition of either the PI3-kinase/Akt pathway or MAP kinase cascade prevents acute estrogen-evoked NO production implies that both pathways participate in the estrogen-mediated response and may possibly converge at the level of eNOS. Similar effects of LY-294002 and PD-98059 on estrogen- and high-density lipoprotein-mediated eNOS activation have been recently described in other endothelial cell preparations (21, 36).
Interestingly, ATP-evoked NO synthesis was unaffected by exposure to either LY-294002 or PD-98059, demonstrating that intracellular Ca2+ mobilization leading to external Ca2+ entry (40) is the primary mechanism underlying eNOS activation by this and similar agonists (i.e., histamine). Furthermore, our observation that estrogen and ATP stimulated NO production in a nonadditive manner (Fig. 10) suggests that EA.hy926 cells may contain only a single “functional pool” of eNOS that is sensitive to both calcium- and phosphorylation-dependent signaling pathways. On the basis of recent work from Fulton and colleagues (14, 52), this functional pool of eNOS may be spatially distributed between the plasma membrane and Golgi complex, with plasma membrane-bound eNOS somewhat more sensitive to calcium- versus kinase-dependent signaling and Golgi-bound eNOS displaying similar sensitivities. Whether the existence of spatially distinct pools of eNOS gives rise to physiologically distinct effects of NO remains unclear. A cartoon summarizing the cellular mechanisms underlying endothelial NO production in response to stimulation by estrogen and Ca2+-mobilizing agonists is shown in Fig. 11.
In summary, the novel findings of our study provide the first direct experimental evidence for the existence of at least two distinct hormone-driven signaling pathways for NO production within a single identified human EA.hy926 cell. On the basis of biochemical data, it has been largely assumed that eNOS activation by both Ca2+/calmodulin and protein phosphorylation events may occur within a single cell. By using direct experimental measurements, the results of our study thus serve as the crucial functional correlate to these earlier reports. Given the widespread sensitivity of the vasculature to the relaxant effects of estrogen and Ca2+-mobilizing hormones, we speculate that these two separate mechanisms will function in parallel in individual endothelial cells present in a variety of vascular beds. Moreover, it is possible that such pathways could result in spatial and/or temporal differences in NO synthesis and signaling that differentially affect vascular tone and health both acutely and in the long term. Finally, although human EA.hy926 cells are reported to display a number of features in common with primary vascular endothelial cells, the extent to which data derived from such a model endothelial cell can be extrapolated to native endothelium remains a consideration.
This work was supported by operating grants to A. P. Braun from the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Alberta, Northwest Territories, and Nunavut. A senior research scholarship to A. P. Braun from the Alberta Heritage Foundation for Medical Research is gratefully acknowledged.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society