Introduction

Top Abstract Introduction Methods Results Discussion References

There is considerable evidence to support a role for endogenous opioid peptides in the brain as regulators of cardiovascular system activities (Feuerstein, 1985; Feuerstein and Sirén, 1987; Holaday, 1983; Sirén and Feuerstein, 1992). Opioid peptides and their receptors have been found in brain areas with an established role in cardiovascular regulation (Atweh and Kuhar, 1977; Desjardins et al., 1990; Fallon and Leslie, 1986; Goodman et al., 1980; Hokfelt et al., 1977; Mansour et al., 1988), and potent cardiovascular effects have been reported after central administration of opioid peptides (Appel et al., 1986; Hassen et al., 1983; Jin and Rockhold, 1991; Kiritsy-Roy et al., 1986; Marson et al., 1989a,b; May et al., 1989; Pfeiffer et al., 1983a,b; Sirén and Feuerstein, 1991; Sirén et al., 1989). The PVN, which contains numerous enkephalin-containing neurons and nerve terminals, as well as opioid binding sites (Desjardins et al., 1990; Fallon and Leslie, 1986; Goodman et al., 1980; Sar et al., 1979; Sawchenko and Swanson, 1982a; Wamsley, 1983), is well known as an important integrating site for modulating autonomic and neuroendocrine cardiovascular responses (Jin and Rockhold, 1989; Kannan et al., 1989; Sawchenko and Swanson, 1982a,b; Swanson and Sawchenko, 1983). Earlier studies have shown that activation of a µ-opioid receptor-mediated pathway in this nucleus, using the µ-selective enkephalin analog DAMGO or dermorphin, causes dose-related increases in blood pressure, heart rate and sympathetic outflow, as measured by a rise in the plasma level of catecholamines in conscious rats (Appel et al., 1986; Kiritsy-Roy et al., 1986). Previous work from this laboratory has confirmed that microinjection into the PVN of increasing doses (0.01-1.0 nmol) of DAMGO increases blood pressure and heart rate in conscious unrestrained rats (Bachelard and Pître, 1995). Moreover, we demonstrated that the pressor and heart rate responses were associated with substantial vasodilation in the hindquarter vascular bed and, at the highest dose (1 nmol of DAMGO), vasoconstrictions in renal and superior mesenteric vascular beds. These effects were not observed after administration of increasing doses (0.01-5.0 nmol) of the -selective enkephalin analog [D-Phe^2,5]enkephalin or the -selective enkephalin analog U50488H into the PVN (Bachelard and Pître, 1995).

Although the effects of PVN µ-opioid receptor activation on blood pressure and heart rate are well known, the mechanisms involved in opioid-mediated cardiovascular responses have not been characterized fully. Therefore, as a continuation of our previous work, it was of interest to further examine the mechanisms by which the selective µ-opioid receptor agonist DAMGO, microinjected into the PVN, causes changes in blood pressure, heart rate and regional hemodynamics in conscious, unrestrained, Wistar Kyoto rats. In the present study, the dose of 1 nmol of DAMGO was used, because we were interested in characterizing the mechanism of the hindquarter vasodilation as well as the renal and superior mesenteric vasoconstrictions. At the lowest doses of DAMGO (0.01 and 0.1 nmol), the renal and superior mesenteric vasoconstrictions did not reach a level of significance (Bachelard and Pître, 1995). Therefore, the experiments were carried out to determine the overall role of alpha adrenoceptor-mediated sympatho-vasoconstriction in the pressor response observed and to investigate whether the vasoactive hormones vasopressin and angiotensin II also play a role in this pressor response. The contribution of adrenoceptors to the cardiovascular responses to DAMGO injected into the PVN was examined by treating the rats with the alpha adrenoceptor antagonist phentolamine or the beta adrenoceptor antagonist propranolol, or with a combination of both antagonists, infused i.v. before PVN administration of DAMGO. Moreover, to determine the involvement of vasopressin or angiotensin II in the cardiovascular responses to DAMGO, a specific vasopressin V₁ receptor antagonist or an angiotensin-converting enzyme inhibitor, captopril, was administered i.v. before DAMGO administration. Thus, blood pressure, heart rate and regional hemodynamic responses to PVN administration of DAMGO in untreated rats were compared with those elicited in rats receiving one of the previously mentioned treatments. To avoid the complicating effects of anesthesia, these studies were carried out in conscious unrestrained rats (Van Loon, 1984).

	Methods

Top Abstract Introduction Methods Results Discussion References

Surgical procedures. The experiments were carried out in male Wistar Kyoto rats (250-300 g; Charles River). At least 14 days before investigation, the animals were anesthetized with a mixture of ketamine and xylazine (100 and 10 mg/kg, respectively, i.p., supplemented as required) and positioned in a stereotaxic frame with the incisor bar set at 3.3 mm below the interaural line. The skull was exposed and cleaned, and two 23-gauge stainless steel guide cannulae targeted 2 mm dorsal to the PVN were obliquely implanted (at an angle of 10 degrees, relative to the vertical), according to the following coordinates: 1.90 mm caudal and 1.75 mm lateral to the bregma and 6.3 mm ventral to the surface of the skull. The cannulae, sealed with 31-gauge stainless steel obturators, were secured to the skull with screws and dental cement. The reflected muscles and skin were replaced and sutured. After surgery the animals were treated with ampicillin (Polyflex, 7 mg/kg i.m.; Ayerst) and flunixin (Banamine, 1 mg/kg i.m.; Schering), housed in individual cages and allowed to recover.

View larger version (136K): [in this window] [in a new window]   Fig. 1.   Representative micrograph showing the locations of injection sites in the PVN. Incisor bar, 3.3 mm below the interaural line; anteroposteriorly, 1.9 mm posterior to the bregma; laterally, 1.75 mm, relative to bregma; ventrally, 8.3 mm ventral to the surface of the skull.

Experimental protocols. Animals were used on 3 consecutive days, during which they received randomized PVN bilateral injections of vehicle (aCSF, pH 7.4) and DAMGO (1.0 nmol) in the presence of an i.v. infusion of specific antagonists or saline (0.9% NaCl), which was used as a control infusion. Each animal received no more than two bilateral injections on the same day, separated by at least 120 min. The composition of the aCSF (in mM) was as follows: NaCl, 125; NaHCO₃, 27; KCl, 2.5; NaH₂PO₄, 0.5; Na₂HPO₄, 1.2; Na₂SO₄, 0.5; CaCl₂, 1.0; MgCl₂, 1.0; glucose, 5.0. On the first day, the animals were given a continuous i.v. infusion of saline (0.3 ml/hr), followed after 15 min by bilateral microinjection of aCSF (0.2 µl on each side) or DAMGO (1 nmol on each side) into the PVN (n = 26). Cardiovascular variables were recorded for 60 min after each injection. On the next day, the animals were given one of the treatments described below.

Effect of pretreatment with phentolamine or a mixture of phentolamine and propranolol on cardiovascular responses to DAMGO bilaterally injected into the PVN. The alpha adrenoceptor antagonist phentolamine was administered i.v. as a bolus (1 mg/kg, 0.1 ml), followed by continuous infusion of 1 mg/kg/hr (0.3 ml/hr) (Winn et al., 1985), to a group of eight rats. Fifteen minutes after the onset of phentolamine infusion, the animals received a bilateral microinjection of DAMGO (1 nmol) or aCSF (0.2 µl) into the PVN. The next day, the same animals were given a mixture of phentolamine (1 mg/kg i.v. bolus, 1 mg/kg/hr infusion) and the beta adrenoceptor antagonist propranolol (1 mg/kg i.v. bolus, 0.5 mg/kg/hr infusion) (Gardiner and Bennett, 1988), followed after 15 min by bilateral microinjection of DAMGO (1 nmol) or aCSF (0.2 µl) into the PVN.

Effect of pretreatment with propranolol or a mixture of phentolamine and propranolol on cardiovascular responses to DAMGO bilaterally injected into the PVN. Propranolol was administered i.v. as a bolus (1 mg/kg, 0.1 ml), followed by continuous infusion of 0.5 mg/kg/hr (0.3 ml/hr), to a group of 10 rats. Fifteen minutes after the onset of propranolol infusion, the animals received a bilateral microinjection of DAMGO (1 nmol) or aCSF (0.2 µl) into the PVN. The next day, these animals were given a mixture of phentolamine (1 mg/kg i.v. bolus, 1 mg/kg/hr infusion) and propranolol (1 mg/kg i.v. bolus, 0.5 mg/kg/hr infusion), followed after 15 min by bilateral microinjection of DAMGO (1 nmol) or aCSF (0.2 µl) into the PVN.

Effect of pretreatment with a vasopressin V₁ receptor antagonist on cardiovascular responses to DAMGO bilaterally injected into the PVN. The vasopressin V₁ receptor antagonist [1-(-mercapto-,-cyclopentamethylenepropionyl)-2-(O-ethyl)-tyrosine]-Arg⁸-vasopressin was administered i.v. as a bolus (10 µg/kg, 0.1 ml), followed by continuous infusion (10 µg/kg/hr, 0.3 ml/hr) (Gardiner et al., 1989), to a group of 13 rats. Fifteen minutes after onset of the infusion, the animals were given a bilateral microinjection of DAMGO (1 nmol) or aCSF (0.2 µl) into the PVN.

Effect of pretreatment with captopril on cardiovascular responses to DAMGO bilaterally injected into the PVN. The angiotensin-converting enzyme inhibitor captopril was administered i.v. as a bolus (3 mg/kg, 0.1 ml), followed by continuous infusion (1 mg/kg/hr, 0.3 ml/hr), to a group of 13 rats. This dose of captopril completely blocked the cardiovascular effects of bolus doses of 100 ng of angiotensin I (H. Bachelard, M. Pître and A. Lessard, unpublished observations). Fifteen minutes after onset of the infusion, the animals were given a bilateral microinjection of DAMGO (1 nmol) or aCSF (0.2 µl) into the PVN.

Drugs. The drugs used were DAMGO (Bachem), phentolamine mesylate (RBI), dl-propranolol hydrochloride (Sigma), captopril (Sigma) and the vasopressin V₁ receptor antagonist [1-(-mercapto-,-cyclopentamethylenepropionyl)-2-(O-ethyl)-tyrosine]-Arg⁸-vasopressin (Bachem). The vasopressin V₁ receptor antagonist was dissolved in 0.5 ml of glacial acetic acid and diluted to a working concentration with isotonic saline; phentolamine, propranolol, captopril, noradrenaline, angiotensin I and vasopressin were dissolved in saline.

Statistical analysis. Values are expressed as the mean ± S.E.M.; n is the number of animals. Results were analyzed for statistical significance by analysis of variance, and individual comparisons were made by Fisher's test. The cardiovascular changes caused by the infusion of antagonists were analyzed by Student's t test for unpaired comparisons. A P value of <.05 was taken to indicate a significant difference.

	Results

Top Abstract Introduction Methods Results Discussion References

Hemodynamic responses to PVN injections of DAMGO. The base-line values (before any drug administration into the PVN) for cardiovascular variables for all treatment groups are given in table 1. Control bilateral injection of aCSF (0.2 µl) into the PVN of rats receiving i.v. infusion of saline alone had no significant effect on any measured or calculated variables (figs. 2 and 3). Bilateral injection of DAMGO (1 nmol) into the PVN in these same rats produced cardiovascular effects, characterized by a long-lasting increase in blood pressure (significant at 2-60 min) and marked increases in heart rate (significant at 15-60 min) (fig. 2), compared with measurements after aCSF. The maximum rises in blood pressure and heart rate were +24 ± 3 mm Hg and +98 ± 10 bpm, respectively, achieved 45 min after the injection of DAMGO. Furthermore, there were substantial and long-lasting falls in renal (significant at 3-60 min) and superior mesenteric (significant at 1-60 min) flows, whereas hindquarter flow increased (significant at 1-60 min) (fig. 2). The maximum decreases in renal (14 ± 3%) and superior mesenteric (32 ± 2%) flows and the maximum increase in hindquarter flow (+99 ± 9%) were observed 30 min after the administration of DAMGO. These cardiovascular responses to DAMGO were associated with falls in renal (significant at 2-60 min) and superior mesenteric (significant at 2-60 min) vascular conductances and increases in hindquarter vascular conductance (significant at 1-60 min) (fig. 3). The maximum decrease in renal vascular conductance (27 ± 4%) occurred 45 min after the administration of DAMGO, and that in superior mesenteric vascular conductance (41 ± 3%) occurred 30 min after the injection of DAMGO. The maximum increase in hindquarter vascular conductance (+78 ± 9%) was reached 15 min after the injection of DAMGO.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Cardiovascular responses to DAMGO (1 nmol) (, n = 26; , n = 8) or aCSF (, n = 22; , n = 7) bilaterally injected into the PVN in conscious, unrestrained, Wistar Kyoto rats, in the absence (, ) or presence (, ) of i.v. treatment with phentolamine (1 mg/kg i.v. bolus, 1 mg/kg/hr infusion). Values are mean ± S.E.M. (vertical lines). §P < .05, when the cardiovascular responses to DAMGO are compared with those produced by the control injection of aCSF in the absence of antagonist; analysis of variance followed by Fischer's test. P < .05, when the cardiovascular responses to DAMGO are compared with those produced by the control injection of aCSF in the presence of antagonist; analysis of variance followed by Fischer's test. *P < .05, when the cardiovascular responses to DAMGO alone are compared with those elicited in the presence of antagonist; analysis of variance followed by Fischer's test. HR, heart rate; MAP, mean arterial blood pressure.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Changes in regional vascular conductances elicited by DAMGO (1 nmol) (, n = 26; , n = 8) or aCSF (, n = 22; , n = 7) bilaterally injected into the PVN in conscious, unrestrained, Wistar Kyoto rats, in the absence (, ) or presence (, ) of i.v. treatment with phentolamine (1 mg/kg i.v. bolus, 1 mg/kg/hr infusion). These data were derived from the data shown in figure 2. Values are mean ± S.E.M. (vertical lines). §P < .05, when the cardiovascular responses to DAMGO are compared with those produced by the control injection of aCSF in the absence of antagonist; analysis of variance followed by Fischer's test. P < .05, when the cardiovascular responses to DAMGO are compared with those produced by the control injection of aCSF in the presence of antagonist; analysis of variance followed by Fischer's test. *P < .05, when the cardiovascular responses to DAMGO alone are compared with those elicited in the presence of antagonist; analysis of variance followed by Fischer's test.

Hemodynamic responses to PVN injection of DAMGO (1 nmol) in the presence of phentolamine. The base-line values (before any drug administration into the PVN) for cardiovascular variables for all treatment groups are given in table 1. Fifteen minutes after the onset of pretreatment with phentolamine (i.v.), blood pressure decreased significantly, persistent tachycardia occurred and hindquarter flow and vascular conductance increased (all significant, P < .05). No significant changes in renal or superior mesenteric flows and vascular conductances were seen. The cardiovascular changes caused by the infusion of the antagonist are indicated in table 1.

Hemodynamic responses to PVN injection of DAMGO (1 nmol) in the presence of propranolol. Fifteen minutes after the onset of pretreatment with propranolol, no significant changes were seen in blood pressure, heart rate, renal superior mesenteric or hindquarter flows or vascular conductances (table 1). In the presence of propranolol, control injection of aCSF (0.2 µl) into the PVN had no significant effect on any measured or calculated variables (figs. 4 and 5), whereas DAMGO caused a pressor response (significant at 3-15 and 45-60 min) and bradycardia (significant at 3-15 min), followed by a significant increase in heart rate 60 min after the injection (fig. 4). The pressor response to DAMGO was not different from that evoked by DAMGO in the absence of propranolol. However, the heart rate responses differed significantly from those in untreated rats, in which PVN injection of DAMGO alone produced marked tachycardia. There were decreases in renal (significant at 30-60 min) and superior mesenteric (significant at 10-45 min) flows, but no significant change in hindquarter flow, compared with measurements after aCSF. The decrease in renal flow was similar to that in rats not receiving propranolol, whereas the decrease in superior mesenteric flow was less than that in rats not treated with propranolol (fig. 4). The increase in hindquarter flow observed in untreated rats was completely inhibited by propranolol (fig. 4). The cardiovascular responses to DAMGO in the presence of propranolol were associated with a fall in renal vascular conductance (significant at 5-60 min), which did not differ from that in the absence of propranolol, and a fall in superior mesenteric vascular conductance (significant at 4-60 min), which was less than that in rats not treated with propranolol. Furthermore, the hindquarter vasodilation observed in untreated animals was abolished by the treatment (fig. 5).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Cardiovascular responses to DAMGO (1 nmol) (, n = 26; , n = 10) or aCSF (, n = 10) bilaterally injected into the PVN in conscious, unrestrained, Wistar Kyoto rats, in the absence () or presence (, ) of i.v. treatment with propranolol (1 mg/kg i.v. bolus, 0.5 mg/kg/hr infusion). Values are mean ± S.E.M. (vertical lines). P < .05, when the cardiovascular responses to DAMGO are compared with those produced by the control injection of aCSF in the presence of antagonist; analysis of variance followed by Fischer's test. *P < .05, when the cardiovascular responses to DAMGO alone are compared with those elicited in the presence of antagonist; analysis of variance followed by Fischer's test. The cardiovascular responses produced by the control injection of aCSF in the absence of antagonist have been omitted for the sake of clarity. HR, heart rate; MAP, mean arterial blood pressure.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Changes in regional vascular conductances elicited by DAMGO (1 nmol) (, n = 26; , n = 10) or aCSF (, n = 10) bilaterally injected into the PVN in conscious, unrestrained, Wistar Kyoto rats, in the absence () or presence (, ) of i.v. treatment with propranolol (1 mg/kg i.v. bolus, 0.5 mg/kg/hr infusion). These data were derived from the data shown in figure 4. Values are mean ± S.E.M. (vertical lines). P < .05, when the cardiovascular responses to DAMGO are compared with those produced by the control injection of aCSF in the presence of antagonist; analysis of variance followed by Fischer's test. *P < .05, when the cardiovascular responses to DAMGO alone are compared with those elicited in the presence of antagonist; analysis of variance followed by Fischer's test. The cardiovascular responses produced by the control injection of aCSF in the absence of antagonist have been omitted for the sake of clarity.

Hemodynamic responses to PVN injection of DAMGO (1 nmol) in the presence of phentolamine and propranolol. Fifteen minutes after the onset of pretreatment with phentolamine and propranolol, a slight but persistent decrease in blood pressure and a reduction in renal vascular conductance (all significant, P < .05) occurred. However, no significant changes were seen in heart rate, renal, superior mesenteric or hindquarter flows or superior mesenteric or hindquarter vascular conductances. Table 1 lists the cardiovascular changes caused by the infusion of the antagonists.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Cardiovascular responses to DAMGO (1 nmol) (, n = 26; , n = 12) or aCSF (, n = 8) bilaterally injected into the PVN in conscious, unrestrained, Wistar Kyoto rats, in the absence () or presence (, ) of i.v. treatment with phentolamine (1 mg/kg i.v. bolus, 1 mg/kg/hr infusion) and propranolol (1 mg/kg i.v. bolus, 0.5 mg/kg/hr infusion). Values are mean ± S.E.M. (vertical lines). P < .05, when the cardiovascular responses to DAMGO are compared with those produced by control injection of aCSF in the presence of antagonist; analysis of variance followed by Fischer's test. *P < .05, when the cardiovascular responses to DAMGO alone are compared with those elicited in the presence of antagonist; analysis of variance followed by Fischer's test. The cardiovascular responses produced by the control injection of aCSF in the absence of antagonist have been omitted for the sake of clarity. HR, heart rate; MAP, mean arterial blood pressure.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Changes in regional vascular conductances elicited by DAMGO (1 nmol) (, n = 26; , n = 12) or aCSF (, n = 8) bilaterally injected into the PVN in conscious, unrestrained, Wistar Kyoto rats, in the absence () or presence (, ) of i.v. treatment with phentolamine (1 mg/kg i.v. bolus, 1 mg/kg/hr infusion) and propranolol (1 mg/kg i.v. bolus, 0.5 mg/kg/hr infusion). These data were derived from the data shown in figure 6. Values are mean ± S.E.M. (vertical lines). P < .05, when the cardiovascular responses to DAMGO are compared with those produced by control injection of aCSF in the presence of antagonist; analysis of variance followed by Fischer's test. *P < .05, when the cardiovascular responses to DAMGO alone are compared with those elicited in the presence of antagonist; analysis of variance followed by Fischer's test. The cardiovascular responses produced by the control injection of aCSF in the absence of antagonist have been omitted for the sake of clarity.

Hemodynamic responses to PVN injection of DAMGO (1 nmol) in the presence of a vasopressin V₁ receptor antagonist. Fifteen minutes after the beginning of pretreatment with the vasopressin V₁ receptor antagonist, no significant changes were seen in blood pressure, heart rate or renal, superior mesenteric or hindquarter flows or vascular conductances (table 1). In the presence of the vasopressin V₁ receptor antagonist, control injection of aCSF (0.2 µl) into the PVN had no significant effect on any measured or calculated variables, whereas DAMGO caused cardiovascular changes that were not different from those evoked by DAMGO in the absence of treatment (data not shown). Thus, there were increases in blood pressure (significant at 2-5 and 30-60 min) and heart rate (significant at 15-60 min), reductions in superior mesenteric flow (significant at 1-45 min) and increases in hindquarter flow (significant at 2-60 min). The reduction in renal flow was not different from that in rats receiving DAMGO alone or from the response in the control animals receiving aCSF alone. These cardiovascular responses were associated with renal (significant at 2-45 min) and superior mesenteric (significant at 1-60 min) vasoconstrictions and hindquarter vasodilation (significant at 5-60 min).

Hemodynamic responses to DAMGO (1 nmol) in the presence of captopril. Fifteen minutes after the onset of pretreatment with captopril, blood pressure decreased and renal flow and vascular conductance increased (all significant, P < .05). However, no significant changes occurred in heart rate, superior mesenteric or hindquarter flows or vascular conductances. Table 1 lists the effects of the inhibitor.

	Discussion

Top Abstract Introduction Methods Results Discussion References

This study focused on the pharmacological characterization of the mechanisms contributing to the regional hemodynamic effects of a highly selective and potent µ-opioid receptor agonist, DAMGO, bilaterally microinjected into the PVN of conscious unrestrained rats. Our results are in continuation of our previous work, which showed that the pressor and heart rate responses to bilateral administration of DAMGO into the PVN were accompanied by marked decreases in renal and superior mesenteric flows and vascular conductances and increases in hindquarter flow and vascular conductance in conscious, freely moving rats (Bachelard and Pître, 1995). The differential changes in regional blood flows elicited by centrally injected DAMGO further suggest that the opioid µ-receptor in the PVN might be involved in the integration of peripheral blood flow in the hypothalamus during affective behavior. Indeed, striking similarities exist between the pattern of cardiovascular changes elicited by PVN microinjection of DAMGO and the classic defense response produced by environmental stressors.

In a previous study, we found that i.v. injections of DAMGO (2 nmol) caused no cardiovascular effects, indicating that the cardiovascular responses to bilateral administration of DAMGO (1 nmol) into the PVN were not due to leakage of DAMGO from its central site of injection into the periphery (Bachelard and Pître, 1995). Moreover, injection of aCSF (0.2 µl) into the same site in all treatment groups had no effect on blood pressure, heart rate or regional blood flows, indicating that displacement of the tissue by the volume injected was not a cause of the observed effects. On the other hand, considering the volume of injection (0.2 µl) and the dose of DAMGO (1 nmol) we used, we cannot exclude the possibility that the injections could have diffused to opioid-sensitive sites outside the PVN in concentrations sufficiently high to affect circulatory regulation by adjacent brain networks.

Although the cardiovascular responses observed in the present study are comparable to those reported previously by this laboratory (Bachelard and Pître, 1995) and by others (Kiritsy-Roy et al., 1986) using selective µ-agonists, further studies are required to determine whether the hemodynamic changes elicited by DAMGO were due to an action restricted to the region of the PVN. A previous study using tritiated enkephalin injected in a volume of 1 µl into a specific hypothalamic nucleus, the preoptic nucleus, showed that approximately 90% of the injected radioactivity was confined to an area with a radius of 1 mm from the site of injection (Feuerstein and Faden, 1982).

The PVN is rich in opioid binding sites (Goodman et al., 1980) and enkephalin-containing neurons (Rossier et al., 1979; Sar et al., 1979; Sawchenko and Swanson, 1982a,b). Moreover, opioid peptide-containing connections have been demonstrated from the hypothalamus to preganglionic sympathetic nerve roots in the intermediolateral cell column of the spinal cord (Khachaturian et al., 1985; Pasternak and Wood, 1986). The results reported here indicate that the tachycardic, pressor and regional hemodynamic responses to central microinjection of DAMGO into the PVN are likely to be due to activation of the sympatho-adrenomedullary system, because these responses were significantly inhibited by the alpha adrenergic antagonist phentolamine, the beta adrenergic antagonist propranolol or a combination of the alpha and beta adrenergic antagonists phentolamine and propranolol. These results are supported by previous studies showing that the pressor and heart rate responses to intrahypothalamic injection of DAMGO in conscious rats were attenuated or even reversed in adrenal demedullated rats treated with the sympathetic blocker bretylium (Pfeiffer et al., 1982, 1983b) and in rats treated with the ganglion blocker chlorisondamine (Sirén and Feuerstein, 1991). Moreover, central administration of -endorphin, DAMGO or dermorphin in rats produced a naloxone-reversible increase in plasma catecholamine concentration (Appel et al., 1986; Kiritsy-Roy et al., 1986; Pfeiffer et al., 1983b; Sirén et al., 1989; Van Loon et al., 1981) and an increase in sympathetic outflow in peripheral postganglionic sympathetic nerves (Sirén and Feuerstein, 1991). These stimulatory cardiovascular responses were blocked after treatment with adrenergic neuronal blocking drugs (Jin and Rockhold, 1991; Pfeiffer et al., 1983b; Sirén and Feuerstein, 1991).

Our results indicate that the hypertension seen after DAMGO injection into the PVN was secondary to vasoconstriction in selective vascular beds and probably also to an increase in cardiac output, although the rise in mean arterial blood pressure did not parallel the rise in heart rate. However, the reason the tachycardia occurred about 10 min after the onset of the pressor response might be that some degree of parasympathetic (vagal) activation (secondary to a baroreceptor-mediated reflex response to the rise in blood pressure) was acting to oppose the beta-1 adrenoceptor-mediated increases in heart rate. Indeed, previous studies reported that the pressor response to a low dose (0.1 nmol) of centrally injected DAMGO was associated with tachycardia but the response to a high dose (0.3 nmol) of DAMGO was associated with bradycardia (Kiritsy-Roy et al., 1986; Pfeiffer et al., 1982, 1983a,b). High doses of DAMGO have been shown to activate parasympathetic outflow when injected into the PVN (Kiritsy-Roy et al., 1986) or the anterior hypothalamus (Pfeiffer et al., 1983b). Moreover, the results of experiments carried out in the presence of propranolol reveal bradycardia of short duration after PVN injection of DAMGO, which accounts for the late tachycardia we observed in untreated rats.

By combining the results of our various antagonist studies, we found that the hypertension seen after DAMGO injection into the PVN was, at least in part, secondary to alpha adrenoceptor-mediated vasoconstriction in the renal and superior mesenteric vascular beds. This finding is consistent with the observation that, in the presence of phentolamine alone, the renal and superior mesenteric vasoconstrictions after PVN administration of DAMGO were attenuated. The reason the renal and superior mesenteric vasoconstrictions in response to PVN administration of DAMGO were attenuated by phentolamine but not by the combination of phentolamine and propranolol may have been that PVN µ-receptor stimulation caused the release of another, unidentified, vasoconstrictor component. Such an influence might be mediated by neural and/or circulating vasoactive substances, and a possible candidate is neuropeptide Y, because this neuropeptide is coreleased with noradrenaline after stimulation of postganglionic noradrenergic neurons (Pernow, 1988) and it was found to exert vasoconstrictor effects after i.v. injection in conscious rats (Gardiner et al., 1988). On the other hand, because the vasopressin and renin-angiotensin systems can compensate effectively in the absence of the sympathetic nervous system, we cannot exclude the possibility that, under autonomic blockade, both systems were involved in the renal and superior mesenteric vasoconstrictions after PVN administration of DAMGO. Another possibility could be that PVN µ-receptor stimulation caused the release of some endothelium-derived constricting factors from the renal and superior mesenteric blood vessels. During the past decade it has been widely recognized that the vascular endothelium can modulate the tone of underlying vascular smooth muscle by the synthesis/release of potent vasorelaxant and vasoconstrictor substances in response to a variety of stimuli. The endothelium can mediate vasoconstriction by the release of a variety of endothelium-derived constricting factors. Candidates for endothelium-derived constricting factors include superoxide anions, arachidonic acid metabolites and the peptide endothelin. These substances cause contraction of vessels and thus might have contributed to the residual renal and superior mesenteric constrictor responses to PVN administration of DAMGO in rats pretreated with both phentolamine and propranolol. However, additional studies are required to determine the exact mechanisms mediating the residual renal and superior mesenteric constrictor responses.

Although we found that in the presence of both phentolamine and propranolol the renal and superior mesenteric vasoconstrictor responses to DAMGO remained unchanged, whereas the pressor effect of DAMGO was completely inhibited, it is possible that the previously reported DAMGO-mediated increases in cardiac output (Sirén and Feuerstein, 1991) were inhibited by the treatment. Thus, in those animals, the renal and superior mesenteric vasoconstrictor responses associated with the hindquarter opposing residual vasodilator effect of DAMGO and the absence or reduction of its effect on cardiac output may result in no change in blood pressure.

The factors underlying the enormous hindquarter vasodilator response to PVN injection of DAMGO were probably based on relatively selective activation of beta-2 adrenoceptor-mediated vasodilator mechanisms in this vascular bed. This is consistent with the observation that, in the presence of propranolol alone, which is a nonselective beta adrenoceptor antagonist, the hindquarter vasodilator response after PVN injection of DAMGO was abolished. Moreover, the vascular bed of the hindquarter is particularly well endowed with beta-2 adrenoceptors, which mediate vasodilation (Gardiner and Bennett, 1988). Circulating adrenaline, which is released from the adrenal medulla during PVN stimulation, might contribute to the hindquarter vasodilator effect. This proposition is supported by earlier studies showing that, concomitantly with their peak effects on blood pressure and heart rate, microinjection of selective µ-agonists, like DAMGO and dermorphin, into the PVN produces dose-related increases of adrenaline and noradrenaline in plasma in conscious rats (Appel et al., 1986; Kiritsy-Roy et al., 1986). Indeed, Kiritsy-Roy et al. (1986) found that DAMGO injected into the PVN selectively increased adrenaline release over noradrenaline plasma, suggesting that the treatment produced a relatively selective activation of the adrenal medulla. The reason the hindquarter vasodilator response to DAMGO was potentiated in the presence of phentolamine alone but was not completely abolished in the presence of both phentolamine and propranolol may have been that phentolamine caused prejunctional disinhibition of noradrenaline release (Saeed et al., 1982), thereby enhancing the hindquarter beta adrenoceptor-mediated vasodilator response and overcoming the effects of propranolol when propranolol was used in combination with phentolamine. However, it seem unlikely that noradrenaline, which is a low-efficacy agonist at beta-2 adrenoceptors, would overcome a substantial block of these receptors by propranolol. Thus, a more likely explanation would be that the vasodilation observed after PVN injection of DAMGO in the presence of phentolamine and propranolol is due to antagonism of alpha-1 adrenoceptor-mediated vasoconstriction.

The PVN is a complex nucleus consisting in part of parvocellular elements projecting to the intermediolateral cell column of the spinal cord, the origin of the preganglionic sympathetic neurons, and magnocellular vasopressinergic neurons projecting to the posterior pituitary (Swanson and Sawchenko, 1983). In the present study with the type of cannula we used, it is likely that DAMGO administered into the PVN was acting on both parvocellular and magnocellular divisions of the PVN. Therefore, it is possible that both increases in sympathoadrenal outflow and vasopressin secretion contributed to the cardiovascular responses to DAMGO. However, in the presence of a vasopressin V₁ receptor antagonist, the cardiovascular responses to PVN injection of DAMGO were similar to those in untreated rats. Thus, it seems unlikely that the release of vasopressin into the systemic circulation plays an important role in the cardiovascular responses elicited by bilateral PVN administration of DAMGO. These results are in agreement with the previous findings that vasopressin does not contribute to pressor action of enkephalin centrally administered in SHR rats (Rockhold et al., 1981). Similarly, i.c.v. injection of DAMGO increased plasma adrenaline and noradrenaline levels but did not cause any response in plasma vasopressin concentration (Matsumura et al., 1992). On the other hand, it has been suggested that opioid peptides exert an inhibitory action on the release of vasopressin from the hypothalamo-neurohypophysial system into the plasma (Arnauld et al., 1983; Grossman et al., 1980; Otake et al., 1991; Van de Heijning et al., 1991; Yamada et al., 1989).

In the presence of the angiotensin-converting enzyme inhibitor captopril, the observed cardiovascular responses to PVN injection of DAMGO were similar to those seen in untreated rats. Thus, it seems unlikely that plasma angiotensin plays a primary role in the cardiovascular responses elicited by bilateral PVN administration of DAMGO. These results are in agreement with earlier studies showing that i.c.v. injection of DAMGO did not cause any responses in plasma renin activity and plasma vasopressin concentrations (Matsumura et al., 1992).

In summary, the results of the present study are consistent with those of previous reports indicating that the hypertensive response to central opioidergic stimulation, in conscious rats, is mediated by an increase in the sympathetic outflow to the adrenal medulla and sympathetic nerve terminals (Appel et al., 1986; Feuerstein and Sirén, 1987; Jin and Rockhold, 1991; Kiritsy-Roy et al., 1986; Marson et al., 1989a,b; Pfeiffer et al., 1983b). By combining the results of our various antagonist studies, we extend previous studies by presenting evidence that DAMGO bilaterally injected into PVN exerts its hypertensive effect through alpha adrenoceptor-mediated vasoconstrictor actions in the renal and superior mesenteric vascular beds and substantial beta adrenoceptor-mediated vasodilation in the hindquarter vascular bed. Moreover, in the renal and superior vascular bed there may also be an involvement of nonadrenergic, nonvasopressinergic and nonangiotensinergic vasoconstrictor mechanisms, which requires further study. Taken together, the present results are further evidence of a role for opioid peptides and µ-opioid receptors in central regulation of cardiovascular function.