INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE SYMPATHETIC NERVOUS system and specific endothelium-derived mediators regulate cardiovascular activity. It has become increasingly evident that these two regulatory systems are not independent of one another but interact to control cardiovascular function with notable precision. The sympathetic nervous system normally functions in the vasculature predominantly through local activation of -receptors (31), whereas the endothelium releases vasoactive hormones, including nitric oxide, endothelium-derived hyperpolarizing factor, endothelins, and prostaglandins (PGs), which have a large number of effects both locally and systemically (28). Although most studies have tended to focus on the former substances (particularly nitric oxide), a potential exists for endothelium-released mediators such as PGs to influence sympathetic nervous system-mediated vascular -receptor responsiveness (8, 11, 16, 27, 32, 37).

In vitro studies have demonstrated that stimulation of -receptors results in the activation of phospholipase A₂ [in addition to traditional G protein-linked activation of phospholipase C (₁) and inhibition of adenylyl cyclase (₂)] and consequently may lead to activation of the arachidonic acid cascade (7, 11, 16, 37) with subsequent release of vasoactive PGs such as prostacyclin (PGI₂), which is present in most blood vessels and is the principal product of arachidonic acid metabolism generated via cyclooxygenase activity (29, 33). Indeed, using an in vitro swine model, Bockman and colleagues (7) demonstrated an ₂-receptor-mediated release of PGI₂ from arterial endothelium. PGI₂ is considered the predominant venous endothelium-derived mediator involved in regulating venous tone (30, 35), as recent in vivo studies have downplayed the role of nitric oxide (18). Despite the importance of the venous system in the control of cardiac filling pressures and output, relatively few studies have focused on alterations in adrenoceptor and endothelial function in the venous system (3-5).

We hypothesized a functional association in vivo between -receptor (both ₁ and ₂) stimulation and endothelial liberation of vasodilatory PGs such as PGI₂ that would not occur with venoconstriction to a nonadrenergic agonist. We therefore assessed changes in human dorsal hand vein distension in response to the ₁-agonist phenylephrine, the ₂-agonist clonidine, and the nonadrenergic agonist PGF₂ with and without pretreatment with the cyclooxygenase inhibitor indomethacin. Modulation of sympathetic venoconstriction by release of PGs is important not only in understanding the normal control of the vasculature but may also be important in cardiovascular diseases with underlying endothelial dysfunction.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Subjects. This study was approved by the University Review Board for Health Sciences Research of the University of Western Ontario. Informed written consent was obtained from all volunteers.

Twenty-four young (23.0 ± 0.5 yr) normal subjects (18 male, 6 female) participated. All were healthy nonsmokers with no significant past medical history, had normal 12-lead electrocardiograms, and refrained from caffeine- and alcohol-containing beverages for at least 12 h before the study. Subjects on aspirin, other nonsteroidal anti-inflammatory drugs, or vasoactive medications, with high blood glucose or lipid levels, or with known sensitivity to acetylsalicylic acid or indomethacin were excluded.

Hand vein measurements. Hand vein studies were carried out postprandially 1-2 h after a light breakfast with subjects in the supine position in a quiet, temperature-controlled (23-24°C) environment as previously described (2). Subjects were covered with a blanket to avoid generalized cold-induced venoconstriction. In each session, the subject's hand was placed on a standard inclined platform angled at 30° to the horizontal. This positioning minimized resting venous tone and facilitated emptying of the hand veins.

An occlusion cuff was placed on the upper arm of the limb being studied and was connected to a manually activated Hokanson Rapid Cuff Inflator (Issaqua, WA). The occlusion cuff was inflated to a pressure of 45 mmHg for all measurements of venous distension unless otherwise stated. A suitable dorsal hand vein (long, straight section with no immediate tributaries) was then selected, and two small (27-gauge) butterfly needles (E-Z Infusion Set; Becton-Dickinson Vascular Access) were inserted in a proximal direction <1 cm apart. A 0.9% saline solution was infused at a rate of 0.2 ml/min through each intravenous line using a Harvard Infusion Pump (model 2400-003; Harvard Apparatus).

A linear variable differential transformer (Schaevitz, type 025 MHR), an electromechanical device used to monitor changes in venous distension (3, 4), was vertically placed over the summit of the vein ~10 mm from the most proximal needle. The transducer was connected to a physiograph and, to avoid local cold-induced venoconstriction, a cloth was placed over the fingers in such a way that it did not interfere with measurements, and the fingers were loosely taped to minimize small finger twitches. The transformer consists of a central movable core surrounded by a primary coil energized by an alternating current and two secondary coils connected in serial opposition. Therefore, with the core in its central position, the resultant voltage in the coils is zero. However, when the core is displaced from its central position (by venous distension), the resultant voltage in the secondary coils becomes different and will be either positive or negative, depending on the direction of movement of the core (2). The magnitude of this voltage reflects the distance by which the core has been moved from its central position and thus when calibrated gives an accurate measurement of venous distension (2).

In 16 subjects, a small temperature probe (YSI 409B; VWR Scientific of Canada) was placed on the dorsum of the hand on each study day and attached to a thermometer with digital readout display to monitor skin temperature. Arterial pressure and heart rate were monitored noninvasively (Dinamap 846SX; Critikon, Tampa, FL) in the contralateral arm.

Study design. For each participant, the appropriate study was carried out in two sessions <10 days apart. Before any drug infusions, at least two recordings of hand vein distension were obtained to ensure a stable baseline. During the first session, the distal intravenous line was connected to a syringe that was randomly assigned to deliver either 0.9% saline (control) or indomethacin (in saline; Merck Sharpe & Dohme, Quebec, Canada), a PG synthase inhibitor, at a rate of 0.2 ml/min. Indomethacin was administered at a dose of 3 µg/min to obtain an estimated local venous concentration of 3 µg/ml. This concentration is at the upper end of the effective dose range for indomethacin (17), and the infusion rate was calculated from previous studies assessing effective steady-state drug concentrations in the human hand vein model (26). The proximal intravenous line was used for administration of either phenylephrine (Sabex, Quebec, Canada), clonidine (Boehringer Ingelheim, Ontario, Canada), or PGF₂ (Upjohn, Ontario, Canada) in separate subjects on both study days.

Eight subjects received phenylephrine. Indomethacin or saline (randomly determined) was infused for 15 min before phenylephrine administration, and control venous distension was determined. Sequential graded infusions of the ₁-agonist phenylephrine (8, 25, 75, 225, 675, 1,500, 3,000, 6,000, and 12,000 ng/min) were administered via the proximal intravenous line (0.2 ml/min) for 5 min at each dose level. The occlusion cuff was inflated at the 3rd min and deflated at the 5th min of each 5-min interval. Resultant venous distension was expressed as the percent change from control distension. On the 2nd day, the above protocol was repeated with 0.9% saline or indomethacin, whichever was not used on the 1st day.

Ten subjects received clonidine. After measurement of control venous distension, sequential graded infusions of the ₂-agonist clonidine (3, 10, 30, 90, 270, 750, 2,250, and 7,000 ng/min) were administered via the proximal intravenous line (0.2 ml/min) for 5 min at each dose level. Resultant venous distension was measured as described previously for phenylephrine.

Six subjects received the nonadrenergic venoconstrictor, PGF₂. Sequential graded infusions of PGF₂ (1, 4, 16, 64, 256, 512, 1,024, and 2,048 ng/min) were administered via the proximal intravenous line (0.2 ml/min) for 15 min at each dose level (34). Three measurements of venous distension were taken during the 3rd-5th min, the 8th-10th min, and the 13th-15th min of each 15-min infusion, and, in each case, the maximum constricting effect of PGF₂ was achieved at the 13th- to 15th-min interval, which was used to construct the dose-response curve as described previously for phenylephrine.

To assess the effect of indomethacin on control venous distension, the mean of three measurements of baseline venous distension was obtained during indomethacin and saline preinfusion before phenylephrine infusions on the days of the phenylephrine study. On the days of the clonidine study, sequential graded increases in venous occlusion pressure (10, 20, 30, and 45 mmHg) were applied (2 min inflation, 3 min recovery) to record a baseline distension-pressure curve during indomethacin and saline preinfusion. For both sets of paired data, indomethacin and saline measurements were compared to determine if indomethacin alone influenced baseline venous distension or distension-pressure curves.

Data analysis and statistics. Dose-response curves (semilogarithmic) were constructed for phenylephrine, clonidine, and PGF₂ during both indomethacin and saline infusions using a nonlinear curve-fitting program (GraphPad Inplot 4.0 software package; H. J. Motulsky, San Diego, CA). The concentration of phenylephrine and PGF₂ required to elicit a 50% constriction of the control hand vein distension at 45 mmHg (ED₅₀) during indomethacin and saline infusion was computed as the geometric mean. Because venoconstriction to clonidine in most cases did not reach 50%, particularly during saline infusions, maximum venoconstriction to clonidine was assessed. Repeated-measures two-way ANOVA was used to compare dose-response curves with indomethacin versus saline. The effects of indomethacin versus saline on maximum venoconstriction to clonidine and the phenylephrine and PGF₂ ED₅₀ values were then compared using paired Student's t-test. Results are given as means ± SE. A two-tailed P value <0.05 was considered significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Skin temperature remained constant during each study (32.4 ± 0.2°C) and did not vary between subjects or between study days. Resting arterial pressure and heart rate were not significantly different on the two study days (Table 1) and were not significantly altered over the duration of the studies. Before -agonist or PGF₂ infusions, indomethacin did not alter control venous distension (Table 1) or the distension-pressure curve (Fig. 1).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Hand vein distension upon graded increases in applied distending pressure during coinfusion of saline and indomethacin in 10 young, normal subjects. Basal hand vein distension measured at 45 mmHg. NS, not significant.

Graded infusions of phenylephrine induced dose-dependent venoconstriction with both saline and indomethacin in all eight subjects. The average maximum constriction obtained with phenylephrine was unaltered during indomethacin infusion (Table 2). However, indomethacin caused a significant parallel shift of the dose-response curve to the left compared with saline (P = 0.003, Fig. 2). Thus the phenylephrine ED₅₀ was significantly decreased in the presence of indomethacin compared with saline (Table 2). This change was consistent in seven of the eight subjects studied (Fig. 3).

View this table: [in this window] [in a new window]   Table 2.   Summary of results for phenylephrine, clonidine, and PGF2

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Average superficial hand vein response to graded infusions of phenylephrine during coinfusion of saline and indomethacin in eight young, normal subjects. Venoconstriction is expressed as %change from control distension.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Individual phenylephrine 50% effective dose (ED50) values during either indomethacin or saline coadministration in 8 young, normal subjects.

Similarly, indomethacin significantly displaced the dose-response curve to clonidine compared with saline (P = 0.005, Fig. 4), and the peak constriction was also significantly increased (Table 2). Furthermore, venoconstriction to clonidine without indomethacin coinfusion was variable among subjects. Some subjects elicited no venoconstriction, whereas most elicited minimal venoconstriction with attenuation of this constriction at higher doses of clonidine (Fig. 5A). Only one subject showed substantial dose-dependent venoconstriction during saline. This subject did not exhibit a shift in the clonidine dose-response curve with indomethacin, suggesting little vasodilator PG release to ₂-receptor stimulation. In the presence of indomethacin, graded local infusions of clonidine induced a dose-dependent venoconstriction in all subjects (Fig. 5B), and no subject elicited a biphasic response.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Average superficial hand vein response to graded infusions of clonidine during coinfusion of saline and indomethacin in 10 young, normal subjects. Venoconstriction is expressed as %change from control distension.

View larger version (36K): [in this window] [in a new window]   Fig. 5.   Original tracings of hand vein distension in a young, normal male subject showing a biphasic response to clonidine during saline (A) and a sustained constriction during indomethacin (B) coinfusion.

Graded infusions of the nonadrenergic agonist PGF₂ induced dose-dependent venoconstriction during coinfusion of both saline and indomethacin in all subjects studied. However, in contrast to the results obtained with -receptor stimulation, the average PGF₂ dose-response curve obtained during indomethacin infusion was not significantly different from that obtained with saline (Fig. 6). Thus the average PGF₂ ED₅₀ was not significantly different in the presence of indomethacin compared with saline (Table 2).

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Average superficial hand vein response to graded infusions of PGF2 during coinfusion of saline and indomethacin in 6 young, normal subjects. Venoconstriction is expressed as %change from control distension.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Previously, it was thought that the exclusive mechanism associated with ₁-receptor-mediated actions was a G protein-linked activation of phospholipase C, resulting in the production of inositol 1,4,5-trisphosphate and diacylglycerol (19). Other data (8, 11, 16, 20, 37) now suggest that additional mechanisms are involved, including activation of phospholipase A₂, which liberates arachidonic acid, the precursor of PGs. A recent in vitro study has shown that phenylephrine produces concentration-related increases in arachidonic acid release although without decreasing the maximum constriction achieved with phenylephrine (6). Similarly, stimulation of ₂-receptors in vitro has been shown to result in the release of both nitric oxide and PGs from vascular smooth muscle and/or endothelium, which may modulate ₂-receptor-mediated contraction of vascular smooth muscle (7, 10). However, recent hand vein studies in healthy subjects have shown that pretreatment with methylene blue (an inhibitor of nitric oxide production) did not augment the observed ₂-receptor-mediated venoconstriction in response to clonidine (18). This suggests that the role of nitric oxide in ₂-receptor-mediated venoconstriction is minimal. Although blood vessels produce many vasodilatory (PGE₁, PGE₂, and PGI₂) and vasoconstricting (PGF₂) PGs, early studies by Moncada and colleagues (30) have shown that PGI₂ is the predominant vasoactive PG produced in the venous system. Work by Vane and Botting (39) suggests that the endothelium produces substantially more PGI₂ than vascular smooth muscle. The role of venodilatory PGs has not previously been defined in vivo.

The findings of our study suggest an association between  (both ₁ and ₂)-receptor stimulation and the release of vasodilatory autacoids (most likely PGI₂) in normal human dorsal hand veins. This proposed mechanism is supported by the observed enhancement of phenylephrine- and clonidine-induced venoconstriction by indomethacin and correlates well with studies done in vitro. These results are consistent with the existence of a population of ₁- and ₂-receptors located within the vascular endothelium and mediating the release of PGI₂ when stimulated. However, we cannot exclude the possibility that stimulation of smooth muscle (nonendothelial) -receptors results in the direct release of PGI₂ or the release of an intermediary that promotes endothelial PG release.

Because the baseline measurements of hand vein distension and distension-pressure curves were identical between indomethacin and saline, it is unlikely that indomethacin has direct venoconstrictor properties. Furthermore, the lack of vasodilation during indomethacin preinfusion suggests that basal PG release is low and does not contribute to venous tone under the conditions of our experiment. Rather, it is more plausible that indomethacin enhances the effects of both phenylephrine and clonidine by attenuating the proposed liberation of PGI₂ induced by -receptor stimulation.

Venoconstriction elicited by ₂-receptor agonism is considered to be intrinsically weaker than that obtained from ₁-agonism. This is considered to be a result of only partial agonist activity of ₂-agonists such as clonidine (5). However, in the present study, clonidine-induced venoconstriction was significantly increased with indomethacin, suggesting that the in vivo potency of clonidine as an ₂-agonist may be underestimated since the influence of vasodilatory PGs has not been previously considered or has been underestimated. Furthermore, venoconstriction elicited by clonidine during saline coinfusion was somewhat variable, with most subjects eliciting a biphasic (venoconstriction followed by attenuation) response. This finding cannot be adequately explained with traditional pharmacological arguments such as 1) the development of tachyphylaxis, since clonidine-induced constriction is stable for up to 2 h (5); 2) presynaptic ₂-receptor activation, as basal neuronal release of norepinephrine appears minimal under the conditions of our experiment (24); 3) diminished ₂-receptor selectivity at high doses (as discussed below); or 4) central ₂-receptor activation, since clonidine was infused locally, and no changes in blood pressure were observed, indicating no systemic effects of clonidine occurred. A recent study by Blochl-Daum and co-workers (5) showed that clonidine produces a dose-dependent venoconstriction of dorsal hand veins in selected young male subjects. In the present study, coinfusion of indomethacin amplified this venoconstrictor response in unselected subjects, and all subjects exhibited dose-dependent venoconstriction to clonidine during indomethacin administration. In contrast, all subjects exhibited dose-dependent venoconstriction to phenylephrine during either indomethacin or saline infusion. Thus it would appear that ₂-mediated venoconstriction to clonidine is counteracted by a gradual increase in the production of potent vasodilatory PGs, such as PGI₂, which may become sufficient to markedly attenuate the ₂-mediated smooth muscle venoconstriction in many young normal subjects. Although PG release may attenuate venoconstriction at lower doses of phenylephrine, at higher pharmacological doses, such attenuation can be overcome. Such results have several implications. First, it may suggest that ₂-receptor-mediated venoconstriction is more sensitive than ₁-mediated venoconstriction to functional antagonists such as vasodilator PGs. Second, these results might be attributed to a greater density of ₂-receptors versus ₁-receptors on the endothelium or enhanced intracellular coupling mechanisms between ₂-receptors and PG production.

Although clonidine is a classic ₂-agonist, its specificity for the ₂-receptor has also been questioned. Although its selectivity for ₂-receptors is less than BHT-933 and UK-14304, clonidine is more readily available for human use. It has been suggested that, at high doses (6,975 ng/min), clonidine may lose its selectivity for the ₂-receptor and consequently stimulate ₁-adrenoceptors (5). However, the results of the present study do not suggest that this likely occurred under the conditions of our experiment, which involved administration of clonidine infusions over a similar dose range. Because the maximum venoconstrictor response obtained with clonidine was relatively small during saline coinfusion (27.2 ± 5.3%) compared with that commonly obtained with ₁-agonists such as phenylephrine (>75%), significant activation of ₁-receptors by clonidine is unlikely, as one would expect such activation to result in an enhanced venoconstriction rather than the attenuation observed with the biphasic response. Furthermore, although clonidine-induced constriction is decreased with prazosin, suggesting clonidine has ₁-effects (5), Bylund (12) demonstrated that the _2B-receptor subtype is also susceptible to blockade by prazosin, providing an alternative explanation for the reduced venoconstriction to clonidine in the presence of prazosin. Moreover, the fact that, within the same study, the dorsal hand vein constriction to clonidine is completely inhibited by the specific ₂-receptor antagonist yohimbine (5) indicates that the response in the dorsal hand vein is mainly mediated through ₂-adrenoceptors.

PGF₂ causes vascular smooth muscle constriction independent of the -receptor, as demonstrated by Ducharme and Weeks (15) in the unanesthetized rat after ganglion blockade and after pretreatment with reserpine, and causes dose-dependent venoconstriction in superficial hand veins (34), consistent with the current results. We have shown that the venoconstriction elicited by PGF₂ was not significantly increased by indomethacin. This not only confirms that basal release of vasodilatory PGs is very low or negligible and stimulation by PGF₂ does not elicit their release but also supports the hypothesis that the release of vasodilatory PGs is specific to stimulation of ₁- and ₂-receptors rather than a nonspecific response to vasoconstriction.

The findings of the present study may have significant implications in conditions with increased sympathetic nervous system activity and/or endothelial dysfunction, such as congestive heart failure or hypertension. It is probable that endothelial dysfunction in patients with severe heart failure (13, 22, 23, 38) may attenuate the production of vasodilator PGs to -receptor stimulation with consequent exaggerated peripheral vasoconstriction. Angiotensin-converting enzyme inhibitors improve endothelial responsiveness in heart failure (14, 21) and may improve PG release through this mechanism in addition to their effects on bradykinin activity (9, 25, 36, 40). Whether low-dose aspirin therapy diminishes sympathetically mediated PG release requires further study. It is also important to determine if similar results are obtained with endogenous norepinephrine release.

We have shown that the venoconstriction induced by stimulation of vascular smooth muscle ₁- and ₂-receptors is significantly increased by indomethacin in young, normal subjects but is unchanged upon PGF₂ stimulation. This increased responsiveness provides evidence for vasodilatory PG release, most likely PGI₂, secondary to -receptor stimulation in sufficient quantities to significantly antagonize the agonist-induced venoconstriction. The ability of PGs to modulate responses to -receptor stimulation in disease states with endothelial dysfunction such as heart failure, hypercholesterolemia, diabetes, atherosclerosis, and hypertension requires further study.