Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Jan 20.
Published in final edited form as: Vascul Pharmacol. 2007 Jan 27;46(5):373–382. doi: 10.1016/j.vph.2007.01.003

Differential Contributions of Alpha-1 and Alpha-2 Adrenoceptors to Vasoconstriction in Mesenteric Arteries and Veins of Normal and Hypertensive Mice

Alex A Pérez-Rivera 1, Alexandra Hlavacova 1, Leonardo A Rosario-Colón 1, Gregory D Fink 1, James J Galligan 1
PMCID: PMC3549429  NIHMSID: NIHMS433244  PMID: 17329171

Abstract

Mesenteric veins are more sensitive than arteries to the constrictor effects of sympathetic nerve stimulation and α–adrenergic receptor agonists. In the present study, we tested the hypothesis that α2-adrenergic receptors (α2-ARs) contribute to in vitro agonist-induced constriction in veins but not arteries and that α2-AR function is down-regulated in mesenteric arteries and veins in deoxycorticosterone acetate-salt (DOCA-salt) hypertension. Norepinephrine (NE) concentration-response curves were similar in SHAM and DOCA-salt arteries and veins indicating that adrenergic reactivity of mesenteric blood vessels is not altered in murine DOCA-salt hypertension in vitro. Veins were 30-fold more sensitive to NE than arteries. The α1-AR antagonist, prazosin (.003–0.3μM), produced concentration-dependent rightward shifts of the NE concentration-response curves in arteries but not veins. The α2-AR agonists, clonidine and UK-14,304, did not constrict arteries or veins in the absence or presence of indomethacin (10 μM) and nitro-L-arginine (NLA; 100 μM). The α2-AR antagonists, yohimbine (.003–0.3 μM) and rauwolscine (0.1μM) did not affect NE responses in SHAM or DOCA-salt arteries but antagonized NE responses in veins. These data indicate that there are different α-AR contractile mechanisms in murine mesenteric arteries and veins. α1-ARs, but not α2-ARs, mediate direct contractile responses in arteries and veins while α2-ARs contribute indirectly to NE-induced constrictions in veins but not arteries in vitro. There may be direct protein-protein interactions between α1- and α2-ARs or between their signaling pathways in veins. This contribution of α2-ARs may account for the greater sensitivity of veins compared to arteries to the contractile effects of NE.

Keywords: alpha-1 adrenoceptors, alpha-2 adrenoceptors, mesenteric arteries, mesenteric veins, mouse, DOCA-salt hypertension

1. Introduction

α-Adrenergic receptors (α-ARs) mediate the vascular constrictor effects of endogenous and exogenous catecholamines. α1- and α2-ARs are expressed by vascular smooth muscle cells and the relative contribution of each receptor type to vasomotor responses is specific to the vascular bed studied. α1- and α2-ARs mediate contractions of the rabbit (Daly et al., 1988) and dog saphenous vein (Fowler et al., 1984). Itoh et al. (1987) found that phenylephrine (PE), a selective α1-AR agonist, was a more potent agonist in canine mesenteric arteries compared to veins while the selective α2-AR agonist UK-14,304 was a more potent constrictor of vein than arteries. These data indicate that arteries and veins express different α-ARs in the same vascular bed. There are also differences in relative expression of α1- and α2-ARs across vascular beds. For example, in the mesenteric, splenic, renal and femoral vascular beds noradrenergic constrictions are mediated by bothα1- and α2-ARs but the contribution of α2-ARs was most prominent in the mesenteric and femoral beds (Polonia et al., 1986).

We have previously shown that murine mesenteric veins are more sensitive than arteries to the constrictor effects of α-AR agonists and veins are resistant to desensitization by adrenergic agonists and to α-AR inactivation by phenoxybenzamine (Pérez-Rivera et al., 2004). In addition, phenoxybenzamine treatment in vitro increased the α-AR desensitization rate in veins to that seen in untreated arteries. These data provided functional evidence that murine mesenteric veins have an increased α-AR reserve compared to arteries (Pérez-Rivera et al., 2004). The greater α-AR receptor reserve in veins would contribute to their increased sensitivity to the constrictor effects of NE and to their relative resistance to desensitization compared to arteries.

Increased venous reactivity and resistance to desensitization by α-ARs would play a role in blood pressure control and overall hemodynamics. Decreases in venous capacitance (resulting from increased venomotor tone) shift blood to the heart and the arterial circulation thereby increasing arterial pressure (Pang, 2001). Therefore, changes in venous capacitance can contribute to the development of hypertension. For example, humans with borderline hypertension exhibit decreased venous distensibility and redistribution of blood volume towards thoracic veins and the heart (Mark, 1984; Takeshita and Mark, 1979). In addition, patients with salt-sensitive hypertension show a greater decrease in venous distensibility during salt-loading than do subjects with salt-resistant hypertension (Draaijer et al., 1993). These changes can also occur in animal models of salt sensitive hypertension. In deoxycorticosterone acetate (DOCA)-salt hypertensive rats, venomotor tone is increased (Fink et al., 2000). The primary determinant of venomotor tone is sympathetic nerve activity and sympathetic input to veins is elevated in DOCA-salt hypertensive rats (Fink et al., 2000). Asα2-ARs can play an important role in contractile responses to adrenergic agonists, a differential role of α2-ARs in arteries and veins could contribute to differences in adrenergic reactivity. Hypertension is also associated with changes in vascular α-AR constrictor mechanisms. For example α1-AR reactivity is increased in large mesenteric arteries from DOCA-salt hypertensive rats (Perry and Webb, 1988). In addition, nitro-L-arginine-induced hypertension in rats is associated with an upregulation of α2-AR receptors in the aorta and in mesenteric arteries (Kanagy, 1997). Finally, α2-AR agonists can relax arteries via an action at endothelial α2-ARs which couple to release of nitric oxide an other vasodilators (Vanhoutte, 2001). Therefore, our objective was to determine whether α2-AR stimulation contributes to the increased sensitivity of mesenteric veins to α-AR agonists by examining contractile responses of murine mesenteric arteries and veins in the presence or absence of α1 orα2-AR specific agonists and antagonists. As sympathetic nerve activity is elevated in DOCA-salt hypertension (Fink et al., 2000), we also examined whether or not α-AR function is down-regulated in arteries and veins from DOCA-salt hypertensive rats in response to elevated sympathetic input to venous smooth muscle cells.

2. Methods

2.1. Animals

C57/BL male mice (25–30g) were obtained from Charles River Labs (Portage, MI). Upon arrival at the animal care facility, mice were maintained according to the standards approved by the Michigan State University All-University Committee on Animal Care and Use. Mice were individually housed in clear plastic cages with free access to standard pelleted chow (Harlan/Teklad 8640 Rodent Diet) and tap water. Mice were housed in temperature and humidity-controlled rooms with a 12 hours on/12 hours off light cycle. Animals were allowed a period of 2–3 days of acclimatization prior to entry into any experimental protocol.

2.2. Preparation of DOCA-salt hypertensive animals

Mice were unilaterally nephrectomized under anesthesia using a solution containing ketamine (500 mg/ml) and xylazine (20 mg/ml) in a 9:1 ratio, respectively. Animals within the weight range used (25 – 30g) received about 80 μL of the anesthetic. Standard surgical procedures were used to prepare DOCA-salt hypertensive mice (Perez-Rivera et al., 2005a; 2005b). SHAM mice were unilaterally nephrectomized, but they received no DOCA pellet implantation and were given tap water. Both groups of mice were placed on standard pelleted rodent chow. After recovery, the mice were housed under standard conditions for 4 weeks after which systolic BP was determined by the tail-cuff method.

2.3. In-vitro preparation of mesenteric vessels

Mice were euthanized with a lethal dose of pentobarbital (50 mg/kg i.p.). The small intestine and associated mesenteric vessels were removed and placed in oxygenated (95% O2, 5% CO2) Krebs’ solution of the following composition (mmol): NaCl 117, KCl 4.7, CaCl2 2.5, MgCl2 1.2, NaHCO3 25, glucose 11. A segment of the intestine with the associated vessels was removed and pinned flat in a silicone elastomer-lined (Sylgard, Dow Corning, Midland, MI) Petri dish. A section of mesentery containing vessels close to the mesenteric border was cut out using fine scissors and forceps. The preparation was transferred to a smaller silicone elastomer-lined recording bath and pinned flat. Second or third-order mesenteric veins or arteries (200 – 300 μm outside diameter) were isolated for study by carefully clearing away the surrounding fat tissue. The recording bath containing the preparation was mounted on the stage of an inverted microscope (Olympus CK-2) and superfused with warm (37°C) Krebs’ solution at a flow rate of 7 ml min−1. All preparations were allowed a 20 min equilibration period during which the vessels relaxed to a stable resting diameter.

2.4. Video monitoring of vessel diameter

Video monitoring of vessel diameter

The output of a black-white video camera (DAGE-MTI 100, IN) attached mounted on the inverted microscope (Olympus CK-2) was fed to a frame grabber card (Picolo, Euresys Inc., TX) mounted in a personal computer. Video images were analyzed online using DiamtrakR edge-tracking software (version 3.5, http://www.diamtrak.com) which provided a continuous measure of artery or vein outside diameter during experiments. Drug induced responses were measured in unpressurized blood vessels. Changes in diameter of 0.5 μm could be resolved.

2.5. Concentration-response studies

All drugs were added in known concentrations to the superfusing Krebs’ solution. Control concentration-responsecurves were obtained in arteries (0.1 – 30 μM) and veins (0.1 nM – 3 μM) after application of NE (Sigma-Aldrich, St. Louis, MO). Each agonist concentration was applied for 3 min and there was a 20-minute interval between successive applications. The contribution of α1-ARs to NE constrictor responses was studied by comparing curves in the absence and in the presence of the selective α1-AR antagonist prazosin (.003 – 0.3 μM; Sigma-Aldrich). A role for α2-ARs in mediating contractile responses to NE was studied by comparing curves in the absence and presence of the selective α2-AR antagonists yohimbine (.003 – 0.3 μM; Sigma-Aldrich) and rauwolscine (0.1 μM; Sigma Aldrich). We also directly tested for the presence of contractile α2-ARs in arteries and veins by challenging blood vessels with the α2-AR agonists clonidine (0.1 – 10μM; Sigma-Aldrich) and UK-14,304 (0.1 – 10μM; Sigma-Aldrich, St. Louis, MO) in the absence or the presence of the cyclooxygenase inhibitor, indomethacin (10 μM), or the nitric oxide synthase inhibitor N-nitro-L-arginine (NLA; 100 μM). All antagonists were applied for 20 minutes prior to agonist application. A single concentration-response curve for each agonist was obtained in each preparation.

2.6. Data analysis

Agonist-induced constrictions are expressed as a percent reduction of the resting diameter). Half maximal effective agonist concentration (EC50) and maximum response (Emax) were calculated from a least-squares fit of individual agonist concentration response curves using the following logistic function from Origin 7.0 (Microcal Software, Inc, Northampton, MA):

Y={(Emin-Emax)/[1+(x/EC50)n]}+Emax

where Emin is the minimum response (set at 0), n is the slope factor. Data are expressed as mean ± SEM.

Agonist concentration-response curves in the presence or absence of antagonists were analyzed by plotting the negative logarithm of the ratio of eqieffective agonist concentrations (50% maximal response) in the presence and absence of the antagonist minus 1 [log dose ratio (DR) – 1)] vs. the negative logarithm of the antagonist concentration (Arunlakshana and Schild, 1959). The X-intercept yields the pA2 value).

Statistical differences between groups were assessed by Student’s two-tailed unpaired t-test. When more than two groups were compared, analysis of variance (ANOVA) was used with Student-Newman-Keuls multiple comparison as a post test. P < 0.05 was considered statistically significant. All statistical analyses and 95% confidence interval (CI) calculations were performed using GraphPad InStat (GraphPad Software, San Diego, CA).

3. Results

3.1. General

Four weeks after the start of DOCA-salt treatment, systolic blood pressure in DOCA-salt (n=93) mice was significantly higher than in SHAM (n=93) mice (122 ± 1 mmHg vs 92 ± 1 mmHg, respectively; P < 0.05). The initial resting diameter of mesenteric arteries from SHAM and DOCA-salt mice was 138 ± 4 μm and 135 ± 4 μm, respectively (P > 0.05). The initial diameter of mesenteric veins from SHAM and DOCA-salt mice was 185 ± 5 μm and 190 ± 7 μm, respectively (P > 0.05).

3.2. Prazosin blocksα1-ARs in mesenteric arteries and veins

In the first set of experiments, we assessed the relative contribution of α1-AR activation to NE-induced constrictions of mesenteric arteries and veins from SHAM and DOCA-salt hypertensive rats. We examined NE-induced constrictions arteries and veins in the absence and presence of prazosin, anα1-AR antagonist. Prazosin alone did not change blood vessel diameter. NE produced a concentration-dependent constriction of arteries but there were no differences in the NE concentration-response curves obtained in SHAM and DOCA-salt arteries (Fig. 1A,B and Table 1). In SHAM arteries, prazosin produced parallel rightward shifts of the NE concentration-response curve without changing the Emax (Fig. 1A, Table 1). Schild analysis of these rightward shifts (Fig. 1C) produced a Schild plot with a slope of 0.9 ± 0.1 (95% CI: 0.6 – 1.1) that was not different from unity. Similar data were obtained in arteries from DOCA-salt rats (Fig. 1B, Table 1) where the Schild plot (Fig. 1D) yielded a line with a slope of 0.8 ± 0.08 (95% CI: 0.6 – 1.0) that was not different from unity.

Figure 1.

Figure 1

Open in a new tab

Effect of prazosin on NE-induced constrictions of SHAM (A) and DOCA-salt (B) mesenteric arteries. Prazosin produced concentration-dependent and parallel rightward shifts in the NE-concentration-response curves with no changes in maximal response. (C) Schild plot for prazosin antagonism of NE-induced contractile responses in SHAM mesenteric arteries. (D) Schild plot for prazosin antagonism of NE-induced contractile responses in DOCA-salt mesenteric arteries. In both cases, the slope of the plot was not different from 1.0 (see text). Data are mean ± SEM. N indicates the number of animals from which preparations were obtained.

Table 1. Properties of NE concentration response curves in arteries and veins from SHAM and DOCA-salt mice in the absence and presence of prazosin.

Data are expressed as mean ± SEM. Numbers in parentheses are the number of animals from which the data were obtained. Emax is the maximum constriction. EC50 is the negative logarithm of the agonist concentration producing half maximal constriction.

Emax (%) EC50 (- log M)
ARTERY VEIN ARTERY VEIN
SHAM
NE (control) 25.7 ± 3.1 (5) 38.8 ± 4.7 (8) 5.7 ± 0.08 (5) 7.2 ± 0.2 (8)
NE/Prazosin (3 nM) 23.9 ± 2.9 (4) 32.6 ± 2.4 (6) 5.1 ± 0.2* (4) 6.3 ± 0.3* (6)
NE/Prazosin (30 nM) 24.0 ± 7.0 (4) 32.0 ± 3.2 (7) 4.5 ± 0.1* (4) 6.1 ± 0.3* (7)
NE/Prazosin (300 nM) 23.4 ± 2.3 (4) 28.8 ± 3.0 (7) 3.5 ± 0.2* (4) 5.7 ± 0.3* (7)
DOCA-salt
NE (control) 28.1 ± 2.9 (5) 33.3 ± 2.8 (12) 5.8 ± 0.06 (5) 7.4 ± 0.2 (12)
NE/Prazosin (3 nM) 19.4 ± 2.5 (4) 26.2 ± 3.8 (8) 5.1 ± 0.09* (4) 6.3 ± 0.2* (8)
NE/Prazosin (30 nM) 18.6 ± 3.8 (4) 28.5 ± 3.0 (7) 4.5 ± 0.06* (4) 5.8 ± 0.3* (7)
NE/Prazosin (300 nM) 23.8 ± 1.3 (4) 29.8 ± 8.5 (5) 3.7 ± 0.1* (4) 5.4 ± 0.2* (5)
Open in a new tab
*

indicates P < 0.05 –vs- control.

NE constricted veins in a concentration-dependent manner (Fig. 2A, Fig. 2B, and there were no differences between concentration response curves obtained in veins from SHAM and DOCA-salt rats (Table 1). However, veins were up to 30-fold more sensitive (determined by comparisons of EC50 values) than arteries to the constricting effects of NE (Table 1). Prazosin inhibited NE-induced constrictions in SHAM (Fig. 2A, Table 1) and DOCA-salt (Fig. 2B, Table 1) veins. However, prazosin did not produce concentration-dependent rightward shifts in the NE concentration-response curves obtained in veins (Fig. 2A, 2B). Schild plots for SHAM veins (Fig. 2C) had a slope of 0.3 ± 0.2 significantly less than 1 (95% CI: 0.2 – 0.8). Similarly, the Schild plot from data obtained in DOCA-salt veins also had a slope (0.5 ± 0.2) that was significantly less than 1 (95% CI: 0.1 – 0.9).

Figure 2.

Figure 2

Open in a new tab

Effect of prazosin on NE- induced constriction of SHAM (A) and DOCA-salt (B) mesenteric veins. Prazosin (3 nM) produced a significant rightward shift in NE concentration-response curves in SHAM and DOCA-salt veins with no change in maximal response among treatment groups. Higher concentrations of prazosin did not did not produce further rightward shifts in NE concentration response curves. (C) Schild plot for prazosin antagonism of NE-induced contractile responses in SHAM mesenteric arteries. (D) Schild plot for prazosin antagonism of NE-induced contractile responses in DOCA-salt mesenteric arteries. In both cases, the slope of the plot was significantly less than 1.0 (see text). Data are mean ± SEM. N indicates the number of animals from which preparations were obtained.

3.3. Yohimbine and rauwolscine inhibit α2-ARs in veins but not arteries

We studied α2-AR mediated contractile responses in mesenteric vessels by measuring NE-induced constrictions in the presence or absence of selective α2-AR antagonists. Yohimbine did not alter the resting diameter or NE-induced constrictions of SHAM or DOCA-salt arteries (Fig. 3A, 3B, Table 2). However, yohimbine antagonized NE-induced constrictions of SHAM and DOCA-salt veins (Fig. 4A, 4B, Table 2) rats. Nevertheless, yohimbine did not cause concentration-dependent and parallel rightward shifts in the NE concentration response curve. As a consequence, this resulted in nonlinear Schild plots in SHAM (Fig. 4C) and DOCA-salt veins (Fig. 4D). Control experiments that looked at concentration-response curves to PE, a selective α1-AR agonist, in the absence and presence of yohimbine were performed in veins to rule out the possibility that yohimbine was acting at α1-ARs. Yohimbine (30 nM) did not antagonize contractile responses to PE in SHAM (Fig. 5A, Table 3) or DOCA-salt (Fig. 5B, Table 3) veins.

Figure 3.

Figure 3

Open in a new tab

Yohimbine did not affect NE concentration response curves in SHAM (A) or DOCA-salt (B) mesenteric arteries. Data are expressed as mean ± SEM. N indicates the number of animals from which preparations were obtained.

Table 2. Response of mesenteric arteries and veins from SHAM and DOCA-salt mice to NE in the absence or presence of yohimbine.

Data are mean ± SEM. Numbers in parentheses are the number of animals from which the data were obtained. Emax is the maximum constriction. EC50 is the negative logarithm of the agonist concentration producing half maximal constriction.

Emax (%) EC50 (-log M)
ARTERY VEIN ARTERY VEIN
SHAM
NE (control) 25.7 ± 3.1 (5) 38.8 ± 4.7 (8) 5.7 ± 0.08 (5) 7.2 ± 0.2 (8)
NE/Yohimbine (3 nM) 26.8 ± 2.6 (5) 33.9 ± 2.6 (4) 5.8 ± 0.05 (5) 6.4 ± 0.2* (4)
NE/Yohimbine (30 nM) 20.5 ± 2.8 (4) 30.4 ± 4.0 (5) 5.6 ± 0.05 (4) 5.3 ± 0.2* (5)
NE/Yohimbine (300 nM) 28.8 ± 1.1 (4) 31.1 ± 3.1 (6) 5.6 ± 0.06 (4) 5.1 ± 0.3* (6)
DOCA-salt
NE (control) 28.1 ± 2.9 (5) 33.3 ± 2.8 (12) 5.8 ± 0.06 (5) 7.4 ± 0.2 (12)
NE/Yohimbine (3 nM) 29.5 ± 5.3 (4) 32.8 ± 4.5 (4) 5.7 ± 0.1 (4) 6.4 ± 0.2 (4)
NE/Yohimbine (30 nM) 33.6 ± 6.6 (4) 30.4 ± 2.7 (6) 5.8 ± 0.1 (4) 6.4 ± 0.5 (6)
NE/Yohimbine (300 nM) 35.6 ± 5.1 (4) 26.4 ± 6.1 (5) 5.7 ± 0.09 (4) 5.4 ± 0.7* (5)
Open in a new tab
*

indicates P < 0.05 –vs- control.

Figure 4.

Figure 4

Open in a new tab

Effect of yohimbine on NE-induced constriction of SHAM (A) and DOCA-salt (B) mesenteric veins. Yohimbine produced a significant rightward shift in the concentration-response curve of SHAM and DOCA-salt veins. Agonist responses are expressed as percent change from resting diameter. (C) Schild plot for yohimbine antagonism of NE-induced contractile responses in SHAM mesenteric veins (D) Schild plot for yohimbine antagonism of NE-induced contractile responses in DOCA-salt mesenteric veins. In both cases, the slope of the plot was significantly different from 1.0 (see text). Data are mean ± SEM. N indicates the number of animals from which preparations were obtained.

Figure 5.

Figure 5

Open in a new tab

Yohimbine did not affect constrictions induced by phenylephrine (PE) in SHAM (A) or DOCA-salt (B) mesenteric veins. PE responses are expressed as percent change from resting diameter. Data are mean ± SEM. N indicates the number of animals from which preparations were obtained.

Table 3. Response of mesenteric arteries and veins from SHAM and DOCA-salt mice to NE in the absence or presence of rauwolscine and to PE in the absence and presence of yohimbine.

Data are mean ± SEM. Numbers in parentheses are the number of animals from which the data were obtained. Emax is the maximum constriction. EC50 is the negative logarithm of the agonist concentration producing half maximal constriction.

Emax (%) EC50 (-log M)
ARTERY VEIN ARTERY VEIN
SHAM
NE (control) 25.7 ± 3.1 (5) 40.0 ± 3.0 (5) 5.7 ± 0.08 (5) 7.5 ± 0.2 (5)
NE/Rauwolscine (100 nM) 26.1 ± 2.5 (6) 28.7 ± 2.6* (6) 5.7 ± 0.08 (6) 7.0 ± 0.05* (6)
PE (control) n.d. 36.5 ± 2.8 (9) n.d. 6.5 ± 0.1 (9)
PE/Yohimbine (30 nM) n.d. 39.7 ± 3.1 (9) n.d. 6.6 ± 0.1 (9)
DOCA-salt
NE (control) 28.1 ± 2.9 (5) 38.9 ± 1.2 (8) 5.7 ± 0.07 (5) 7.2 ± 0.2 (8)
NE/Rauwolscine (100 nM) 25.3 ± 1.4 (5) 24.7 ± 4.0* (7) 5.8 ± 0.04 (5) 6.7 ± 0.1* (7)
PE(control) n.d. 34.9 ± 5.4 (6) n.d. 6.6 ± 0.1 (6)
PE/Yohimbine (30 nM) n.d. 42.2 ± 6.4 (5) n.d. 6.4 ± 0.03 (5)
Open in a new tab
*

indicates P < 0.05 –vs- control.

To determine whether the inhibition of NE-induced contraction in veins but not arteries was specific for yohimbine, we examined NE-induced contractile responses in arteries and veins in the absence and presence of rauwolscine. Rauwolscine (100 nM) did not antagonize NE-induced constrictions in SHAM (Fig. 6A, Table 3) or DOCA-salt (Fig. 6B, Table 3) arteries. However, rauwolscine did cause a rightward shift of NE concentration-response curves in veins from SHAM (Fig. 7A, Table 3) and DOCA-salt (Fig. 7B, Table 3) rats.

Figure 6.

Figure 6

Open in a new tab

Rauwolscine did not affect NE-induced constriction of SHAM (A) and DOCA-salt (B) mesenteric arteries. NE-induced responses are expressed as percent change from resting diameter. Data are mean ± SEM. N indicates the number of animals from which preparations were obtained.

Figure 7.

Figure 7

Open in a new tab

Effect of rauwolscine on NE-induced constriction of SHAM (A) and DOCA-salt (B) mesenteric veins. Rauwolscine produced a significant rightward shift in the concentration-response curve in veins from both groups. Data are mean ± SEM. N indicates the number of animals from which preparations were obtained.

3.4. α2-AR agonists do not constrict arteries or veins: lack of effect of inhibitors of nitric oxide synthase or cyclooxygenase

Our previous work showed that neither clonidine or UK 14,304 (α2-AR agonists) constriction mesenteric arteries or veins from rats (Luo et al., 2003) or mice (Perez-Rivera et al., 2004). Similar results were obtained in the present study where it was found that neither clonidine (0.1–10 μM) or UK-14,304 (0.1–10μM) caused more than 10% constriction of arteries or veins (data not shown). As activation of endothelial α2-ARs results in endothelium-dependent vasorelaxation (Bockman et al., 1996; Figueroa et al., 2001), it is possible that α2-AR mediated release of vasodilators could antagonize any contractile effects of α2-AR agonists on vascular smooth muscle. For this reason, clonidine and UK-14,304 were studied in the presence of the cyclooxygenase inhibitor, indomethacin (10 μM), or the nitric oxide synthase inhibitor N-nitro-L-arginine (NLA; 100 μM) which inhibit cyclooxygenase and nitric oxide synthase, respectively. Even in the presence of these inhibitors, clonidine and UK-14,304 caused < 10% constriction of arteries or veins (data not shown).

3.5 Interaction between α1- and α2-AR agonists in veins

The results obtained using α-AR receptor antagonists indicate that both α1 and α2-AR contribute to NE-induced constrictions of mesenteric veins. However, while PE, caused direct venoconstriction, α2-AR agonists alone did not constrict veins. It is possible that α2-ARs do not mediate direct venoconstriction but instead they may sensitize the α1-AR contractile mechanism. To test this hypothesis, PE concentration-response curves were obtained in the absence and presence of UK 14,304 (0.1 μM). It was found that UK 14,304 shifted the PE curve to the left without changing the maximum response (Figure 8). The PE EC50 in the absence of UK 14,304 was 127 ± 26 nM and in the presence of UK 14,304 this value was 42 ± 20 nM (n = 5, P < 0.05). UK 14,304 also changed the steepness of the PE concentration-response curve. The slope factor for the PE alone concentration-response curve was 0.95 ± 0.1 while in the presence of UK 14,304 the slope factor for the PE concentration-response curve was reduced to 0.5 ± 0.1 (P < 0.05).

Figure 8.

Figure 8

Open in a new tab

UK 14,304 potentiates PE-induced constrictions of mesenteric veins. Concentration response curves for PE-induced constriction of mesenteric veins were obtained in the absence and presence of UK 14,304 at a concentration (0.1 μM) that did not cause direct venous constriction. UK 14,304 caused a leftward shift in the PE concentration response curve without changing the maximum response. UK 14,304 also caused a decrease in the slope factor of the PE concentration-response curve (see text). Data are mean ± SEM (n = 5).

4. Discussion

4.1. Contribution of α2-ARs to constriction in veins but not arteries

In vitro studies have consistently shown that α1-ARs are the predominant receptor mediating NE-induced contraction of mesenteric arteries in the rat (Hussain and Marshall, 2000) and mouse (Yamamoto and Koike, 2001). In the present study, Schild analysis revealed that prazosin acted as a simple competitive antagonist vs. NE in arteries confirming previous data that α1-ARs mediate NE-induced constrictions of mesenteric arteries in the mouse. This conclusion is also supported by our data showing that yohimbine and rauwolscine (α2-AR antagonists) do not affect NE-induced arterial constrictions. Furthermore, clonidine and UK 14,304 (α2-AR agonists) did not constrict mesenteric arteries. Similar data using UK 14,304 have also been obtained in pressurized mesenteric arteries from the mouse (Flavahan, 2005). We conclude that α2-ARs are not involved in NE-induced constrictions of murine mesenteric arteries.

Schild analysis revealed that, in veins, prazosin did not act as a simple competitive antagonist. This result suggests that NE may act at more than one receptor site. It is possible that α2-ARs contribute to the prazosin-resistant response in mesenteric veins. However, we found that neither clonidine nor UK 14,304 constricted arteries or veins suggesting that α2-ARs do not mediate direct NE-induced constriction of mesenteric blood vessels. We further explored the contribution of α2-ARs to NE-induced constrictions of mesenteric blood vessels by examining NE concentration-response curves in the presence of yohimbine or rauwolscine. Both of these α2-AR antagonists competitively antagonized NE-induced constrictions in veins suggesting that α2-ARs contribute via an indirect mechanism to these responses.

The suggestion that α2-ARs indirectly contribute to NE-induced constrictions of mesenteric veins is also supported by our data showing that UK 14,304 potentiates PE-induced constrictions of veins. UK 14,304 did not cause venous constriction but it did cause a leftward shift in the PE concentration-response curve. Similar data have been obtained in rat tail artery where it was shown that UK 14,304 casued a leftward shift in the concentration constriction curve for methoxamine, an α1-AR agonist. In this same study it was shown that UK 14,304 alone did not cause arterial constriction (Chen et al., 1999). Therefore, the proposed interaction between α1-ARand α2-ARs is not unique to murine mesenteric veins. Taken together data suggest that activation of α2-ARs sensitizes veins to the constrictor effects of α1-AR activation. In addition, the slope factor for the PE concentration-curve was reduced in the presence of UK 14304. The decline in slope factor indicates that UK 14,304 alters the interaction of PE with the α1-AR or with the associated contractile mechanism.

4.3. Veins are more sensitive than arteries to NE: contribution of α2-ARs

Veins are more sensitive than arteries to the constrictor effects of sympathetic nerve stimulation (Hottenstein and Kreulen, 1987; Luo et al., 2003). In the present studies we confirmed that are more sensitive than arteries to the contractile effects of NE (Luo et al., 2003). Data from the present study suggest that a contribution of α2-ARs to NE-induced constriction of veins but not arteries contributes to greater noradrenergic reactivity of veins compared to arteries. It should also be pointed out that there is also likely to be differential expressionα1-AR receptor subtypes or post-receptor signaling in arteries and veins as veins were also more sensitive to the contractile effects of PE, a selective α1-AR agonist. This latter result agrees with previous studies in rats (Luo et al., 2003) and mice (Pérez-Rivera et al., 2004) that showed an increased reactivity of mesenteric veins over arteries to stimulation by PE.

Although we showed that α-ARs contribute to NE induced venous constriction, α2-AR agonists did not cause direct venoconstriction. This contrasts with previously published work showing that both α1-AR and α2-AR agonists constrict veins (Fowler et al., 1984; Itoh et al., 1987; Daly et al., 1988; Daniel et al., 1997). While α2-AR agonists can constrict some veins, these responses are generally smaller than those caused by α1-AR agonists (Fowler et al., 1984; Daniel et al., 1997). The smaller constriction caused by α2-AR compared to α1-AR agonists in veins might be due to lower concentrations of α2-AR compared to α1-ARs. However, binding studies revealed that in canine mesenteric vein, α2-AR concentrations are higher than those for α1-ARs (Daniel et al., 1997). An alternative explanation is that α2-AR receptors may couple very inefficiently to the signaling mechanisms producing constriction in veins. This suggestion is consistent with the observation that canine mesenteric veins require pre-depolarization with elevated extracellular potassium concentration in order to potentiate responses caused by α2-AR agonists (Daniel et al., 1987). Murine mesenteric veins may express fewer α2-ARs than mesenteric veins in other species and these receptors may couple very poorly to the intracellular signaling pathways leading to venoconstriction. However, even modest activation of these signaling mechanisms by α2-AR may be sufficient to sensitize venous smooth muscle to the constrictor effects of α1-AR activation.

4.4. Functional interaction between α1- and α2-ARs

The data present here indicate that in order to see a contribution of α2-ARs to contractile responses in veins, co-activation of both α1- and α2-ARs is necessary. Similar receptor interactions occur in chinese hamster lung fibroblasts expressing α1-ARs and α2A-ARs (Reynen et al., 2000). In those studies, NE did not increase [Ca2+]i in cells expressing onlyα1-ARs but NE did cause an increase in [Ca2+]i in cells co-expressing α1-AR and α2A-AR. The effects of NE were blocked by subtype selective concentrations of α1- and α2-AR antagonists. Selective agonists of α1-AR and of α2-ARs agonists applied alone did not have any effect on [Ca2+]i release but when added together induced a robust stimulation of [Ca2+]i. These data are similar to our own where we showed that concentrations of PE that normally do not constrict veins do elicit a constriction when the same concentrations are applied in the presence of UK 14,304. As Reynen et al. (2000) stated, this phenomenon could be of physiological importance in vascular smooth muscle cells (in particular venous smooth muscle cells) that express functional α1-ARs and α2-ARs. Our data provide evidence that this functional interaction occurs in venous smooth muscle cells that normally co-express α1-ARs and α2-ARs at physiologically relevant levels. This is not always the case in heterologous expression systems. Previous studies in veins have also suggested that a functional interaction occurs between α1-ARs and α2-ARs (Daly et al., 1987) and this interaction may be important to the hemodynamic function of veins (see below)

An important question is how this receptor interaction might occur. Is it due to a direct physical interaction between α1- and α2-ARs or an interaction involving the signaling cascades activated by both receptors? It is known that α-ARs could interact with other receptor systems in ways that are receptor-specific. For example, in mouse atria, angiotensin and bradykinin receptors interact with α2-AR (Cox et al., 2000; Trendelenburg et al., 2003) and protein kinase C activation plays a role in this functional interaction (Mota and Guimaraes, 2003). More detailed experiments are necessary to determine whether or not there is direct coupling of α1- and α2-ARs or if post receptor events are responsible for the functional interaction in veins.

4.5 DOCA-salt hypertension does not alter adrenergic reactivity in arteries or veins

We did not find any differences in adrenergic reactivity between DOCA-salt arteries and veins compared to their SHAM counterparts. This is in contrast to the studies by Luo et al. (2003) who showed a decreased reactivity of DOCA-salt veins but no difference in reactivity between SHAM and DOCA-salt arteries. Other studies performed in DOCA-salt rats have found that mesenteric arterial adrenergic reactivity is enhanced (Longhurst et al., 1988; Perry and Webb, 1988). Potential reasons for the discrepancies seen are differences in size of the vessels studied or the different methods used to assess vascular reactivity. It should also be noted that despite significant increases in blood pressure in DOCA-salt mice, the degree of hypertension in mice is much less than that reported for rats (Johns et al., 1996). Therefore, increases in blood pressure seen in DOCA-salt mice may have be sufficient to stimulate the signaling mechanisms leading to changes in adrenergic reactivity in the mesenteric vasculature.

4.6. Conclusions

There are different α-AR contractile mechanisms in murine mesenteric arteries and veins: α1-ARs mediate constriction in arteries and veins whereas α2-ARs contribute to responses in veins but not arteries. This difference in adrenoceptor function and pharmacology contributes to artery vs. vein differences in adrenergic reactivity: veins are more sensitive than arteries to the constrictor effects of NE. Our data also indicate that there is a functional interaction between α1- and α2-ARs or their signaling mechanisms in veins but not arteries. Finally, adrenergic reactivity in murine mesenteric arteries and veins is not altered in DOCA-salt hypertension.

Acknowledgments

This work has been supported by National Institutes of Health Minority Student Predoctoral Fellowship 5 F31 HL072732-02 (AAP) and by National Institutes of Health Program Project Grant PO1 HL70687 (GDF, JJG). LAR was a participant in an American Society of Pharmacology and Experimental Therapeutics Summer Undergraduate Research Fellowship (SURF) Program.

References

  1. Arunlakshana O, Schild HO. Some quantitative uses of drug antagonists. Br J Pharmacol. 1959;14:48–58. doi: 10.1111/j.1476-5381.1959.tb00928.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bockman CS, Gonzalez-Cabrera I, Abel PW. Alpha-2 adrenoceptor subtype causing nitric oxide-mediated vascular relaxation in rats. J Pharmacol Exp Ther. 1996;278:1235–1243. [PubMed] [Google Scholar]
  3. Chen DC, Duckles SP, Krause DN. Postjunctional alpha2-adrenoceptors in the rat tail artery: effect of sex and castration. Eur J Pharmacol. 1999;372:247–252. doi: 10.1016/s0014-2999(99)00226-5. [DOI] [PubMed] [Google Scholar]
  4. Cox SL, Schelb V, Trendelenburg AU, Starke K. Enhancement of noradrenaline release by angiotensin II and bradykinin in mouse atria: evidence for cross-talk between G(q/11) protein- and G(i/o) protein-coupled receptors. Br J Pharmacol. 2000;129:1095–1102. doi: 10.1038/sj.bjp.0703167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Daly CJ, McGrath JC, Wilson VG. Pharmacological analysis of postjunctional alpha-adrenoceptors mediating contractions to (−)-noradrenaline in the rabbit isolated lateral saphenous vein can be explained by interacting responses to simultaneous activation of alpha 1- and alpha 2-adrenoceptors. Br J Pharmacol. 1988;95:485–500. doi: 10.1111/j.1476-5381.1988.tb11669.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Daniel EE, Low AM, Lu-Chao H, Gaspar V, Kwan CY. Characterization of alpha-adrenoceptors in canine mesenteric vein. J Cardiovasc Pharmacol. 1997;30:591–598. doi: 10.1097/00005344-199711000-00009. [DOI] [PubMed] [Google Scholar]
  7. de Champlain J. Pre- and postsynaptic adrenergic dysfunctions in hypertension. J Hypertens Suppl. 1990;8:S77–S85. [PubMed] [Google Scholar]
  8. Draaijer P, Kool MJ, Maessen JM, van Bortel LM, de Leeuw PW, van Hooff JP, Leunissen KM. Vascular distensibility and compliance in salt-sensitive and salt-resistant borderline hypertension. J Hypertens. 1993;11:1199–207. [PubMed] [Google Scholar]
  9. Figueroa XF, Poblete MI, Boric MP, Mendizabal VE, Adler-Graschinsky E, Huidobro-Toro JP. Clonidine-induced nitric oxide-dependent vasorelaxation mediated by endothelial alpha(2)-adrenoceptor activation. Br J Pharmacol. 2001;134:957–968. doi: 10.1038/sj.bjp.0704320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fink GD, Johnson RJ, Galligan JJ. Mechanisms of increased venous smooth muscle tone in desoxycorticosterone acetate-salt hypertension. Hypertension. 2000;35(Pt 2):464–9. doi: 10.1161/01.hyp.35.1.464. [DOI] [PubMed] [Google Scholar]
  11. Flavahan NA. Phenylpropanolamine constricts mouse and human blood vessels by preferentially activating alpha2-adrenoceptors. J Pharmacol Exp Ther. 2005;313:432–439. doi: 10.1124/jpet.104.076653. [DOI] [PubMed] [Google Scholar]
  12. Fowler PJ, Grous M, Price W, Mathews WD. Pharmacological differentiation of postsynaptic alpha adrenoceptors in the dog saphenous vein. J Pharmacol Exp Ther. 1984;229:712–718. [PubMed] [Google Scholar]
  13. Guimaraes S, Moura D. Vascular adrenoceptors: an update. Pharmacol Rev. 2001;53:319–356. [PubMed] [Google Scholar]
  14. Hottenstein OD, Kreulen DL. Comparison of the frequency dependence of venous and arterial responses to sympathetic nerve stimulation in guinea-pigs. J Physiol. 1987;384:153–67. doi: 10.1113/jphysiol.1987.sp016448. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hussain MB, Marshall I. α1-Adrenoceptor subtypes mediating contractions of the rat mesenteric arteries. Eur J Pharmacol. 2000;395:69–76. doi: 10.1016/s0014-2999(00)00220-x. [DOI] [PubMed] [Google Scholar]
  16. Itoh H, Kohli JD, Rajfer SI. Pharmacological characterization of the postsynaptic alpha-adrenoceptors in isolated canine mesenteric arteries and veins. Naunyn Schmiedebergs Arch Pharmacol. 1987;335:44–49. doi: 10.1007/BF00165034. [DOI] [PubMed] [Google Scholar]
  17. Johns C, Havras I, Handy DE, Salomao A, Gavras H. Models of experimental hypertension in mice. Hypertension. 1996;28:1064–1069. doi: 10.1161/01.hyp.28.6.1064. [DOI] [PubMed] [Google Scholar]
  18. Kanagy NL. Increased vascular responsiveness to alpha2-adrenergic stimulation during NOS inhibition induced hypertension. Am J Physiol. 1997;273:H2765–2764. doi: 10.1152/ajpheart.1997.273.6.H2756. [DOI] [PubMed] [Google Scholar]
  19. Langer SZ. Presynaptic regulation of catecholamine release. Biochem Pharmacol. 1974;23:1793–1800. doi: 10.1016/0006-2952(74)90187-7. [DOI] [PubMed] [Google Scholar]
  20. Longhurst PA, Rice PJ, Taylor DA, Fleming WW. Sensitivity of caudal arteries and the mesenteric vascular bed to norepinephrine in DOCA-salt hypertension. Hypertension. 1988;12:133–142. doi: 10.1161/01.hyp.12.2.133. [DOI] [PubMed] [Google Scholar]
  21. Luo M, Hess MC, Fink GD, Olson LK, Rogers J, Kreulen DL, Dai X, Galligan JJ. Differential alterations in sympathetic neurotransmission in mesenteric arteries and veins in DOCA-salt hypertensive rats. Auton Neurosci. 2003;104:47–57. doi: 10.1016/S1566-0702(02)00287-4. [DOI] [PubMed] [Google Scholar]
  22. Mark AL. Structural changes in resistance and capacitance vessels in borderline hypertension. Hypertension. 1984;6(Pt 2):69–73. doi: 10.1161/01.hyp.6.6_pt_2.iii69. [DOI] [PubMed] [Google Scholar]
  23. Mo FM, Kwan CY. Alpha-adrenoceptor subtypes of dog saphenous vein: another unusual property. Life Sci. 1998;62:1455–1459. doi: 10.1016/s0024-3205(98)00089-7. [DOI] [PubMed] [Google Scholar]
  24. Minneman KP. Alpha 1-adrenergic receptor subtypes, inositol phosphates, and sources of cell Ca2+ Pharmacol Rev. 1988;40:87–119. [PubMed] [Google Scholar]
  25. Mota A, Guimaraes S. Influence of alpha2-autoreceptor stimulation on the facilitation by angiotensin II and bradykinin of noradrenaline release. Naunyn Schmiedebergs Arch Pharmacol. 2003;368:443–447. doi: 10.1007/s00210-003-0840-2. [DOI] [PubMed] [Google Scholar]
  26. Oparil S. The sympathetic nervous system in clinical and experimental hypertension. Kidney Int. 1986;30:437–452. doi: 10.1038/ki.1986.204. [DOI] [PubMed] [Google Scholar]
  27. Pang CC. Autonomic control of the venous system in health and disease: effects of drugs. Pharmacol Ther. 2001;90:179–230. doi: 10.1016/s0163-7258(01)00138-3. [DOI] [PubMed] [Google Scholar]
  28. Perez DM, DeYoung MB, Graham RM. Coupling of expressed alpha1B- and alpha1D-adrenergic receptor to multiple signaling pathways is both G protein and cell type specific. Mol Pharmacol. 1993;44:784–795. [PubMed] [Google Scholar]
  29. Pérez-Rivera AA, Fink GD, Galligan JJ. Increased reactivity of murine mesenteric veins to adrenergic agonists: functional evidence supporting increased α1-adrenoceptor reserve in veins compared with arteries. J Pharmacol Exp Ther. 2004;308:350–357. doi: 10.1124/jpet.103.056184. [DOI] [PubMed] [Google Scholar]
  30. Perez-Rivera AA, Fink GD, Galligan JJ. Alpha-1B adrenoceptors mediate neurogenic constriction in mesenteric arteries of normotensive and DOCA-salt hypertensive mice. Auton Neurosci. 2005a;121:64–73. doi: 10.1016/j.autneu.2005.07.004. [DOI] [PubMed] [Google Scholar]
  31. Perez-Rivera AA, Fink GD, Galligan JJ. Vascular reactivity of mesenteric arteries and veins to endothelin-1 in a murine model of high blood pressure. Vascul Pharmacol. 2005b;43:1–10. doi: 10.1016/j.vph.2005.02.014. [DOI] [PubMed] [Google Scholar]
  32. Perry PA, Webb RC. Sensitivity and adrenoceptor affinity in the mesenteric artery of the deoxycorticosterone acetate hypertensive rat. Can J Physiol Pharmacol. 1988;66:1095–1099. doi: 10.1139/y88-179. [DOI] [PubMed] [Google Scholar]
  33. Piascik MT, Perez DM. α1-Adrenergic receptors: new insights and directions. J Pharmacol Exp Ther. 2001;298:403–410. [PubMed] [Google Scholar]
  34. Piascik MT, Soltis EE, Piascik MM, Macmillan LB. α-Adrenoceptors and vascular regulation: molecular, pharmacologic and clinical correlates. Pharmacol Ther. 1996;72:215–241. doi: 10.1016/s0163-7258(96)00117-9. [DOI] [PubMed] [Google Scholar]
  35. Polonia JJ, Guerreiro M, Guimaraes S, Garrett J. Postsynaptic alpha-adrenoceptor subtypes in the internal carotid, mesenteric, splenic, renal and femoral vascular beds of the dog. Arch Int Pharmacodyn Ther. 1986;279:5–16. [PubMed] [Google Scholar]
  36. Reynen PH, Martin GR, Eglen RM, MacLennan SJ. Characterization of human recombinant α2A-adrenoceptors expressed in chinese hamster lung cells using intracellular Ca2+ changes: evidence for cross-talk between recombinant α2A- and native α1-adrenoceptors. Br J Pharmacol. 2000;129:1339–1346. doi: 10.1038/sj.bjp.0703184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Starke K, Borowski E, Endo T. Preferential blockade of presynaptic alpha-adrenoceptors by yohimbine. Eur J Pharmacol. 1975a;34:385–388. doi: 10.1016/0014-2999(75)90268-x. [DOI] [PubMed] [Google Scholar]
  38. Starke K, Endo T, Taube HD. Relative pre- and postsynaptic potencies of alpha-adrenoceptor agonists in the rabbit pulmonary artery. Naunyn Scmiedebergs Arch Pharmacol. 1975b;291:55–78. doi: 10.1007/BF00510821. [DOI] [PubMed] [Google Scholar]
  39. Starke K, Montel H, Gayk W, Merker R. Comparison of the effects of clonidine on pre- and postsynaptic adrenoceptors in the rabbit pulmonary artery. Alpha-sympathomimetic inhibition of neurogenic vasoconstriction. Naunyn Schmiedebergs Arch Pharmacol. 1974;285:133–150. doi: 10.1007/BF00501149. [DOI] [PubMed] [Google Scholar]
  40. Takeshita A, Mark AL. Decreased venous distensibility in borderline hypertension. Hypertension. 1979;1:202–6. doi: 10.1161/01.hyp.1.3.202. [DOI] [PubMed] [Google Scholar]
  41. Trendelenburg AU, Meyer A, Klebroff W, Guimaraes S, Starke K. Crosstalk between presynaptic angiotensin, bradykinin receptors and alpha 2-autoreceptors in sympathetic neurons: a study in alpha 2-adrenoceptor-deficient mice. Br J Pharmacol. 2003;138:1389–1402. doi: 10.1038/sj.bjp.0705223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Vanhoutte PM. Endothelial adrenoceptors. J Cardiovasc Pharmacol. 2001;38:796–808. doi: 10.1097/00005344-200111000-00016. [DOI] [PubMed] [Google Scholar]
  43. van Meel JC, Towart R, Kazda S, Timmermans PB, van Zwieten PA. Correlation between the inhibitory activities of calcium entry blockers on vascular smooth muscle contraction in vitro after K+-depolarization and in vivo after alpha 2-adrenoceptor stimulation. Naunyn Schmiedebergs Arch Pharmacol. 1983;322:34–37. doi: 10.1007/BF00649349. [DOI] [PubMed] [Google Scholar]
  44. van Zwieten PA, van Meel JC, Timmermans PB. Functional interaction between calcium antagonists and the vasoconstriction induced by the stimulation of postsynaptic alpha 2-adrenoceptors. Circ Res. 1983;52 (2 Pt 2):I77–I80. [PubMed] [Google Scholar]
  45. Yamamoto Y, Koike K. Alpha(1)-adrenoceptor subtypes in the mouse mesenteric artery and abdominal aorta. Br J Pharmacol. 2001;134:1045–1054. doi: 10.1038/sj.bjp.0704350. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES