Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Clonidine is an ₂-adrenergic receptor (₂-AR) agonist that is thought to mediate its antihypertensive effects through sites in the medulla oblongata (Lipski et al., 1976; Bousquet et al., 1981; Punnen et al., 1987). Several specific sites of action have been proposed, including the nucleus of the solitary tract (NTS) and the sympathoexcitatory region of the rostral ventrolateral medulla (RVL) (Lipski et al., 1976; Bousquet et al., 1981; Sun and Guyenet, 1986). A recent study also indicates that clonidine microinjections into the vasodepressor region of the caudal ventrolateral medulla can evoke sympathoinhibitory responses in the cat (Orer et al., 1996). Finally, other studies have suggested that the medullary gigantocellular reticular formation may be a site of action for the antihypertensive effects of ₂-AR agonists (Chan and Koo, 1978; Chen and Chan, 1980; Lim and Chan, 1986; Lim et al., 1988).

We recently described a vasodepressor subregion within the medial medullary reticular formation, the gigantocellular depressor area (GiDA) (Aicher et al., 1994, 1995). Glutamate nanoinjections into the GiDA produce dose-dependent vasodepressor responses (Aicher et al., 1994). This region also contains reticulospinal neurons that project directly to sympathetic preganglionic neurons in the thoracic spinal cord and may be a substrate for sympathoinhibition (Aicher et al., 1995). Reticulospinal neurons in the GiDA also contain the _2A-AR (Guyenet et al., 1994). A recent study using gene-targeting techniques has implicated the 2A subtype of the -AR (Bylund, 1988) in mediating the antihypertensive effects of ₂-AR agonists such as clonidine (MacMillan et al., 1996). In the present study, we sought to determine whether clonidine evokes vasodepressor responses when applied to this region of the medulla oblongata and whether these responses can be blocked by the potent and selective ₂-AR antagonist 2-methoxyidazoxan (RX821002) (Langin et al., 1989; O'Rourke et al., 1994) injected into the same site. The results support the idea that the _2A-AR is the principal receptor involved in the antihypertensive actions of ₂-AR agonists (MacMillan et al., 1996) and indicate that the medullary sites of action for these drugs may be widely distributed, consistent with the wide distribution of ₂-ARs (Guyenet et al., 1994).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and General Surgical Methods

Male Sprague-Dawley rats (300-400 g) were used in these experiments. Before experiments, rats were housed in wire-mesh cages with food and water available ad libitum. All protocols used in these studies have been approved by the Institutional Animal Care and Use Committee of Cornell University Medical College. Rats were anesthetized with urethane (1.2 g/kg i.p.) followed by -chloralose (70 mg/kg i.v.). The trachea was intubated to permit artificial ventilation (70 strokes/min with 100% oxygen) before paralysis (0.8 mg/kg i.m. tubocurarine, with 0.2 mg/kg/h supplements), and the femoral artery and vein were catheterized to allow the measurement of arterial pressure and the administration of drugs, respectively. Rectal temperature was maintained at 37°C with a thermostatically controlled heating pad. Rats were placed in a stereotaxic frame for brain nanoinjections (see below). Arterial pressure was monitored continuously via a Cobe pressure transducer interfaced to a CyberAmp (Axon Instruments, Pacer Scientific Instruments, Los Angeles, CA) and recorded continuously on a Dell personal computer using a 1401 interface and Spike2 software (Cambridge Electronic Design, Cambridge, England).

Brain Nanoinjections

The dorsal surface of the medulla oblongata was exposed by removal of the atlanto-occipital membrane and partial removal of the occipital bone (Aicher and Reis, 1997). Nanoinjections were made into the brain through single- or double-barrel glass pipettes (30-40-µm tip) drawn from calibrated tubing. Pressure injections were made with a picospritzer (General Valve Inc., Fairfield, NJ), and the volume was determined by monitoring the lemniscus through a microscope. Stereotaxic coordinates for injection sites measured from the calamus scriptorius were as follows: GiDA (1.0-1.3 mm rostral, 1.0 mm lateral, 2.0-2.5 mm ventral) and RVL (2.4-2.7 mm rostral, 1.9-2.1 mm lateral, 2.4-2.7 mm ventral).

Drugs for brain nanoinjections were dissolved in artificial cerebrospinal fluid (aCSF) (Kiely and Gordon, 1994), and salt solution concentrations were: L-glutamate (20 mM; Sigma Chemical Co., St. Louis, MO); clonidine HCl (0.38, 3.8, and 38 mM; Research Biochemicals, Inc., Natick, MA); prazosin HCl (10 µM; Sigma Chemical Co.), and 2-methoxyidazoxan (0.2, 2, and 20 mM; Sigma Chemical Co.). Prazosin was warmed and stirred to mix the drug into the solution, which was made fresh daily. Injectate volumes for each drug were 10 nl of L-glutamate, 20 nl of clonidine, 30 nl of prazosin, and 30 nl of 2-methoxyidazoxan. Initial concentrations of adrenergic drugs were based on published efficacious concentrations for central nervous system injections for clonidine (Chen and Chan, 1980), prazosin (Stone et al., 1997), and 2-methoxyidazoxan (Huangfu et al., 1995). All brain nanoinjections were made unilaterally. L-Glutamate injections were made before drug injections to identify vasoactive sites.

Testing Procedures

Clonidine Nanoinjections. The vasodepressor and bradycardic effects of clonidine (20 nl for each site) were assessed after unilateral nanoinjections into either the GiDA or RVL (see coordinates given). The vasoactive sites were first identified by nanoinjections of L-glutamate (10 nl) (Aicher et al., 1994); 5 to 10 min later, clonidine was nanoinjected into the same site via a second barrel of a double-barrel pipette. Single injections of clonidine were administered to each rat, and arterial pressure and heart rate (HR) were monitored continuously for 2 h. Preliminary studies indicated that the response to a vasodepressor dose of clonidine was greatly attenuated if the injection was repeated in the same site even 3 h later (i.e., tachyphylaxis was occurring); therefore, only single injections were made in each animal. At the conclusion of each experiment, rats were sacrificed, and injection sites were verified as described below. Some animals (n = 2) received an injection of methyl atropine (100 µg/kg i.p.) in 0.5 ml of saline 30 min before clonidine to determine whether the fall in arterial pressure was secondary to the fall in HR.

Antagonist Preinjections. We sought to determine whether the cardiovascular responses to clonidine could be blocked by prior injections of an antagonist into the same site. The antagonist (60 nl) was injected into the site from one barrel of a double-barrel pipette. Three minutes later, clonidine was injected into the same site from the second barrel of the pipette. Arterial pressure was monitored continuously, and the change in arterial pressure produced by each nanoinjection was measured for 2 h. Control animals received an injection of vehicle (aCSF) instead of the antagonist and then were tested in a similar fashion. As a control for the site-specificity of the antagonist, some animals received injections of the antagonist into the RVL while the agonist, clonidine, was injected into the GiDA.

Histological Verification of Injection Sites

All nanoinjection sites were marked with rhodamine microbeads (0.1%) dissolved in the injectate. At the conclusion of each experiment, rats were administered an overdose of sodium pentobarbital (100 mg/kg i.p.) and perfused with 4% paraformaldehyde through the aorta. The brain was removed, placed in fixative for at least 24 h, and sectioned (30 µm) on a vibrating microtome. Adjacent sections were mounted in two sets: one for identification of the nanoinjection sites under fluorescent illumination, and the other for Nissl staining to identify brain regions (Aicher et al., 1994). Injection sites were plotted onto representative brain sections (Swanson, 1992) spaced at 0.5-mm intervals rostral to obex, which is the point of transition from the fourth ventricle to the central canal (i.e., the most rostral extension of the area postrema). The experimental group and response outcomes were unknown to the observer who was plotting injection sites.

Data Analysis

Baseline (resting) levels of mean arterial pressure (MAP) and HR, as well as the onset (change of at least 10 mm Hg) and peak responses produced by clonidine, were compared. Statistical comparisons were made using ANOVAs or t tests as appropriate (Winer, 1971). For experiments comparing the two brain sites, the peak magnitude of the clonidine vasodepressor response at each of three doses was compared using a two-way ANOVA. For the antagonist studies, the three doses of 2-methoxyidazoxan were compared using a one-way ANOVA, and a t test was used to compare the prazosin responses to the vehicle control group. Significant effects were further compared using Newman-Keuls post hoc analyses (Winer, 1971). For all tests,  = .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Vasodepressor Effects of Nanoinjections of Clonidine into GiDA and RVL. The nanoinjection of clonidine (20 nl; 750 pmol total dose) into GiDA evoked a large and sustained decrease in arterial pressure (Fig. 1) and HR, which slowly returned to baseline values over a 2-h period. Arterial pressure usually began to fall within 1 min of the completion of the nanoinjection and reached a peak (61 ± 16 mm Hg; n = 3) approximately 10 min later (10.6 ± 1.2 min; n = 3). We observed that although arterial pressure did return to baseline values, a second injection of the same dose of clonidine into the same site evoked a much smaller response; therefore, only a single dose of clonidine was tested in each animal. In two animals with atropine pretreatment, there was no significant attenuation of the fall in pressure (peak = 37 ± 13 mm Hg; n = 2).

View larger version (6K): [in this window] [in a new window]   Fig. 1.   Nanoinjection of clonidine into the GiDA evokes a large, sustained hypotension. Arterial pressure drops after the nanoinjection of 20 nl of clonidine (750 pmol total dose; arrow indicates completion of injection) unilaterally into the GiDA. The fall in arterial pressure began approximately 1 min after the injection and reached its lowest point about 4 min later. Middle and right, arterial pressure recovery at 30-min intervals beginning approximately 1 and 2 h after the injection (hr Post). Each of the three segments of the arterial pressure trace is approximately 33 min long. Scale bar indicates 5 min.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   MAP decreases in a dose-dependent fashion after clonidine nanoinjections into either the GiDA (filled bars) or the RVL (shaded bars). Three doses of clonidine were tested (7.5, 75, and 750 pmol, n = 3 per dose). All injections were 20 nl in volume and administered unilaterally into a vasoactive site that was verified by prior injection of glutamate (10 nl, 200 pmol total dose). Two-way ANOVA revealed a significant effect of dose [F(2,18) = 16.6, p < .0001] but not of site [F(1,18) = 0.099, p = .75] and no site × dose interaction [F(2,18) = 0.08, p = .92]. Post hoc pairwise multiple comparisons (Student-Newman-Keuls) tests revealed significant differences between the responses to each of the three doses of clonidine.

Effects of 2-Methoxyidazoxan and Prazosin on Clonidine Vasodepressor Responses Injection of the potent and selective ₂-AR antagonist 2-methoxyidazoxan into the GiDA before clonidine injection into the same site blocked both the vasodepressor and bradycardic responses to clonidine (750 pmol dose) in a dose-dependent manner (Fig. 3). Control animals received nanoinjections of the vehicle (aCSF) before nanoinjections of clonidine into the GiDA. aCSF injections into the GiDA did not themselves evoke a change in arterial pressure or HR (Fig. 4A) or alter the responses to clonidine (Figs. 3 and 4A). In addition, injection of 2-methoxyidazoxan into the GiDA did not itself evoke changes in arterial pressure or HR (Table 1, Fig. 4B), although it did completely block the response to clonidine from the same site (Fig. 4B). This effect was specific to the ₂-AR because prior injections of the ₁-AR antagonist prazosin (10 µM concentration, 0.06 pmol total dose) failed to block the responses to clonidine (peak decrease in MAP = 45 ± 6 mm Hg, n = 4).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   2-Methoxyidazoxan preinjections dose dependently blocked the responses to clonidine nanoinjections into the GiDA. Control animals received injections of aCSF (60 nl) into the GiDA 3 min before nanoinjections of clonidine (20 nl, 750 pmol total dose) into the same site. In control animals (aCSF), clonidine evoked large falls in MAP (filled bars) and HR (shaded bars). Nanoinjections of 2-methoxyidazoxan into the GiDA 3 min before clonidine attenuated both the vasodepressor and the bradycardic responses in a dose-dependent fashion (0.012, 0.12, and 1.2 nmol total doses, 60 nl volume). A one-way ANOVA of the MAP responses for the three doses of 2-methoxyidazoxan and aCSF controls revealed a significant effect of dose [F(3,11) = 46.7, p < .0001]. Post hoc comparisons revealed significant differences between all of the doses and the aCSF control group, as well as between each of the doses of 2-methoxyidazoxan.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   The blockade of clonidine responses in GiDA by 2-methoxyidazoxan is drug specific and site selective. A, response to clonidine (750 pmol) nanoinjection in the GiDA was not altered by prior injection of vehicle (aCSF) into the same site. B, injection of 60 nl of 2-methoxyidazoxan (1.2 nmol total dose) into the GiDA completely blocked subsequent responses to clonidine nanoinjection, although the nanoinjection of 2-methoxyidazoxan into GiDA did not change arterial pressure or HR (see also Table 1 for group data). C, vasodepressor and bradycardic responses to clonidine nanoinjection into the GiDA were not reduced by prior injection of 2-methoxyidazoxan (60 nl, 1.2 nmol total dose) into the RVL [n = 4; compared with animals that received vehicle injection of aCSF into GiDA before clonidine; t test (t = 1.08, 5 df, p = .3200)]. However, nanoinjection of the antagonist into the RVL evoked a transient (<30 s) fall in arterial pressure in three of four animals tested.

Histology of Injection Sites. The locations of injection sites in the GiDA and the RVL are illustrated in Fig. 5. Clonidine injection sites in GiDA () and those in RVL () were equally effective in lowering MAP. Sites within the GiDA at which clonidine injections lowered MAP were all located less than 1 mm rostral to obex (i.e., 4.30 to 3.80 mm caudal to the interaural line) and centered approximately 1 mm lateral to the midline (Fig. 5, B and C). Sites within RVL at which clonidine lowered MAP () were located 1.5 to 2.0 mm rostral to obex (2.80 mm caudal to the interaural line) and centered approximately 2 mm lateral to the midline (Fig. 5A). Injection sites for the antagonists into the GiDA and RVL were located in overlapping positions with the injection sites illustrated in Fig. 5.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Schematic summary of the histological location of the clonidine injection sites in GiDA () and RVL (). RVL sites () were 1.5 to 2.0 mm rostral to obex (A) and were 1.8 to 2.0 mm lateral to the midline. GiDA sites () were 0.0 to 0.7 mm rostral to obex (B and C) and were 0.9 to 1.1 mm lateral to the midline. The spread of the 20 nl injection volume extended 180 to 240 µm from the center of the injectate. The histological sections were modified from the computer rat brain atlas by Swanson (1992), and the symbols were applied using Adobe Illustrator software. XII, hypoglossal nucleus; AMB, nucleus ambiguus; IO, inferior olive; LRN, lateral reticular nucleus; RPA, raphe pallidus; RO, raphe obscurus; SP5, spinal trigeminal nucleus.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present data demonstrate that clonidine produces potent vasodepressor responses when applied to two functionally distinct regions of the medullary reticular formation. These data further show that these sites are equipotent in their vasodepressor efficacy and that the local application of the selective ₂-AR antagonist 2-methoxyidazoxan (Langin et al., 1989; O'Rourke et al., 1994) completely blocks the effects of clonidine in GiDA. These data support the previous suggestion that the _2A-AR plays a critical role in mediating the antihypertensive actions of drugs such as clonidine (Stornetta et al., 1995; MacMillan et al., 1996), but they challenge the notion that these actions can be attributed to a single locus in the medulla oblongata. We argue, rather, that the existence of numerous sites of action is more consistent with the widespread localization of the _2A-AR in the medulla oblongata (Guyenet et al., 1994).

Clonidine Is Equipotent in GiDA and RVL. Several studies have suggested that the antihypertensive actions of clonidine and other ₂-AR agonists are mediated primarily or solely through specific medullary sites, including the NTS, the gigantocellular reticular formation, and particularly the RVL (Lipski et al., 1976; Chan and Koo, 1978; Bousquet et al., 1981; Punnen et al., 1987; Lim et al., 1988). However, our data indicate that clonidine is equipotent in two functionally distinct regions of the medulla oblongata: the GiDA, which is a tonically active vasodepressor, sympathoinhibitory region (Aicher et al., 1994; 1995), and the RVL, which is a tonically active vasopressor, sympathoexcitatory region (Morrison and Reis, 1991; Schreihofer and Guyenet, 1997). Some of the differences between our studies and earlier localizations of active sites may be partly related to the larger volumes of drugs used in other studies (100-500 nl) (Bousquet et al., 1981; Lim et al., 1988), which could spread to several different functional sites in the medulla oblongata. Alternatively, the attribution of the antihypertensive effects of clonidine to a single brain region, particularly RVL, may be related to the potent tonic influence of RVL on arterial pressure and sympathetic outflow, such that any manipulation of this area would cause changes in baseline measures that may have confounded the results of prior studies. Regardless of the reasons for these previous outcomes, in the present study, our use of nanoliter volumes and picomolar doses of drugs supports the idea that clonidine acts with equivalent potency and efficacy at multiple sites of action in the medulla oblongata.

Effects of Clonidine Are Blocked by a Selective ₂-AR Antagonist. Previous studies have demonstrated that microinjections of the ₂-AR agonist guanabenz produce dose-dependent vasodepressor responses (Lim et al., 1985). It has also been shown that the systemic vasodepressor effects of guanabenz could be blocked by local application of the ₂-AR antagonist yohimbine into the gigantocelluar reticular formation (Lim et al., 1988), supporting the idea that ₂-ARs in the gigantocelluar reticular formation participate in the vasodepressor actions of ₂-AR agonists (Chen and Chan, 1980; Lim and Chan, 1986). The present results extend these findings to show that local application of an even more potent antihypertensive agent, clonidine, can produce dose-dependent falls in arterial pressure after the application of volumes and doses (20 nl volume, 750 pmol total dose) that are a fraction of the volumes used in many other studies. These results also show that the vasodepressor and bradycardic responses to clonidine can be completely blocked in a dose-dependent fashion by the local application of a selective ₂-AR antagonist, 2-methoxyidazoxan. The local application of prazosin, which is an antagonist at both the ₁-AR and the 2B and 2C subtypes of the -AR (Bylund et al., 1988; Hieble and Ruffolo, 1996), failed to block the response to clonidine. Together with evidence from genetic studies involving receptor subtype-selective point mutations (MacMillan et al., 1996), these data support the idea that it is the 2A subtype of the -AR that is specifically involved in mediating hypotension from medullary autonomic sites.

Correlation Between Sites of Action and _2A-AR Localization. As discussed, there appear to be several potential sites of action for clonidine and other ₂-AR agonists in the medulla oblongata, including the RVL (Bousquet et al., 1981; Punnen et al., 1987), the NTS (Lipski et al., 1976), the caudal ventrolateral medulla (Orer et al., 1996), and the gigantocellular reticular formation (Lim et al., 1988; present study). Although this list includes both sympathoexcitatory and sympathoinhibitory regions, all of these medullary areas play a role in the tonic regulation of vasomotor tone, and they all contain immunocytochemical evidence for _2A-AR (Guyenet et al., 1994). Therefore, the localization of vasoactive sites after clonidine microinjections is consistent with the widespread localization of the _2A-AR in these regions, particularly in reticulospinal neurons.

Functional Significance of GiDA. The GiDA is a relatively unexplored region from which potent vasodepressor responses can be evoked by nanoinjections of excitatory amino acids (Aicher et al., 1994). Chemical lesions of this area evoked fulminating hypertension and blocked the baroreceptor reflex (Aicher and Reis, 1997). This area also is one of only a few regions with efferent projections to multiple sympathetic targets, as shown by dual retrograde viral transport from the adrenal gland and stellate ganglion (Jansen et al., 1995). We have also shown that GiDA efferents monosynaptically innervate sympathoadrenal preganglionic neurons in the spinal cord (Aicher et al., 1995). Together, these studies have established that the GiDA contains tonically active sympathoinhibitory neurons with direct input to sympathetic output neurons. The present evidence indicates that in addition, these sympathoinhibitory neurons in the GiDA are potential targets for the demonstrated hypotensive actions of clonidine and other ₂-AR agonists (Chan and Chan, 1983), although these ₂-ARs probably are not tonically active because ₂-AR blockade in GiDA did not increase blood pressure or HR.