Evidence for Synergy Between α2-Adrenergic and Nonadrenergic Mechanisms in Central Blood Pressure Regulation
Background— Both α2-adrenergic and non–α2-adrenergic mechanisms seem to be involved in the hypotensive effect of imidazoline-like drugs. This study aimed at investigating how these 2 mechanisms work together to modify blood pressure (BP).
Methods and Results— LNP 509, which appeared in this study to be devoid of α2A-adrenergic activity, was administered to anesthetized rabbits and wild-type (WT) mice into the cisterna magna and into the fourth ventricle, respectively. Mean arterial pressure decreased by a maximum of 46±4% and 16±2%, respectively. In D79N mice, which lack functional α2A-adrenergic receptors, LNP 509 also reduced mean arterial pressure by 17±2%. The hypotension induced by LNP 509 (100 μg/kg intracisternally) was prevented by S23757 (1 mg/kg intracisternally), an antagonist highly selective for I1-imidazoline binding sites (I1BS). A synergy between LNP 509 and the α2-adrenergic agonist α-methylnoradrenaline (α-MNA) was observed in rabbits (cisterna magna injection) and in WT mice (fourth ventricle injection) but not, as expected, in D79N mice. Similar to LNP 509 alone, rilmenidine (fourth ventricle injection), which binds both to α2-adrenergic receptors and to I1BS, decreased BP in D79N mice. In WT animals, rilmenidine had a significantly greater effect. Microinjections performed in rabbits showed that the synergism occurred at least in part in the nucleus reticularis lateralis of the brain stem.
Conclusions— These results demonstrate that a central imidazoline-sensitive, but non–α2-adrenergic, mechanism can modify BP by itself. This mechanism, which may involve I1BS, interacts synergistically with an α2-adrenergic mechanism to decrease BP.
Received October 19, 2001; revision received December 13, 2001; accepted December 21, 2001.
The central nervous system (CNS) plays a key role in the regulation of cardiovascular function. The nucleus tractus solitarii (NTS) and the nucleus reticularis lateralis of the rostral ventrolateral medulla (NRL/RVLM) are 2 nuclei that are highly involved in this control.1
α2-Adrenergic receptors (α2ARs) are distributed throughout the CNS, including the NTS and the NRL/RVLM.2,3⇓ They are known to mediate many effects, such as sedation, anesthetic-sparing response, analgesia, and sympathoinhibition.4 Thus, α2AR agonists administered centrally cause hypotension.5 α2ARs have been classified into 3 subtypes: α2AAR, α2BAR, and α2CAR.6 All of them are present in the CNS.7 Gene-targeted mice have been engineered for each α2AR subtype to determine the functional relevance of subtype diversity. For example, D79N mice have a point mutation of the α2aAR gene. The mutation of Asp79 to Asn uncouples the α2AAR from its transduction pathways in vivo.8 Thus, D79N mice constitute functional α2A-knockout animals.9 Experiments performed in D79N and α2A-knockout mice have confirmed that α2A is the α2AR subtype that mediates the central effects of α2AR agonists, including their hypotensive action.9
One unique mechanism of action explains the central hypotensive effect of the α2-adrenergic agonist, α-methylnoradrenaline (α-MNA). This catecholamine is the active metabolite of α-methyldopa, and it reduces blood pressure (BP) in inhibiting the vasomotor tone by activating α2ARs in the NTS.10
The mechanism of the central antihypertensive effects of imidazoline-related compounds appears more complex. As a matter of fact, a large body of experimental data indicates that these drugs also activate α2AARs to reduce BP.5,11,12⇓⇓ Nevertheless, several groups have reported that a nonadrenergic mechanism also contributes to their hypotensive action. This assumption was based on structure-activity relationship studies and on antagonism experiments.13–15⇓⇓ The nonadrenergic component of the central hypotension induced by imidazoline-like drugs was afterward attributed to their ability to act at specific binding sites, insensitive to catecholamines, called imidazoline binding sites (IBS) and located in the NRL/RVLM region.16,17⇓ Because relevant animal models and ligands selective for these nonadrenergic sites over α2ARs were lacking, it was impossible to clarify how (in the case that these 2 mechanisms should work together) they might affect BP regulation and also how they might account for the overall hypotensive action of imidazoline-like drugs. It may be interesting to note that after central injection, the onset of the hypotensive effect of these drugs is much faster than that of α-MNA. Their maximal effect is also greater. These observations suggest that (imidazoline-specific) nonadrenergic and adrenergic mechanisms may cooperate when they are triggered by hybrid imidazoline drugs (which are capable of binding on both types of sites), such as clonidine, rilmenidine, and moxonidine.13,18⇓ Taking advantage of the recent availability of selective tools and of genetically engineered mice with impaired α2AARs, the present study was designed to clarify all these aspects.
It has recently been reported that dicyclopropylmethyl-(4,5-dimethyl-4,5-dihydro-3H-pyrrol-2-yl)-amine hydrochloride (LNP 509), a new pharmacological tool, had neither affinity for α2ARs and α1ARs nor activity at α2ARs.19 This drug has been shown to bind selectively to the I1 subtype of IBS (I1BS) over the I2 subtype of IBS (I2BS).19,20⇓ In the present study, we used LNP 509 and α-MNA to trigger nonadrenergic and α2-adrenergic mechanisms, respectively. We looked at their effects on BP when given alone and when given in combination in rabbits and mice. We also studied the response to rilmenidine after central injection in mice.
[35S]GTPγS Binding Assay
The method of Jasper et al21 was modified as described previously.22 Membranes of CHO cells (clone 1E5) transfected with human α2AARs were incubated with 0.2 nmol/L [35S]GTPγS and drugs for 1 hour at 25°C. Nonspecific binding was defined by 10 μmol/L cold GTPγS. Reactions were stopped by vacuum filtration over Whatman GF/B filters. Filters were washed with ice-cold buffer (20 mmol/L HEPES/NaOH, 100 mmol/L NaCl, and 3 mmol/L MgCl2, pH 7.4 at 4°C), and incorporated radioactivity was determined by using liquid scintillation counting.
Animals and Hemodynamic Measurements
The present study was conducted in compliance with institutional guidelines and those formulated by the European community for use of experimental animals (L358-86/609/EEC). Studies were performed on normotensive male rabbits (Zika strain, Wendling, Wasselonne, France), weighing 2.5 to 3.5 kg, prepared as described previously,22 and on 10- to 20-week-old B6.129S2-Adra2atm1Lel mice, ie, D79N mice. These mice were kindly provided by Dr L. Limbird (Vanderbilt University, Medical Center, Nashville, Tenn). C57BL/6N mice were the wild-type (WT) control strain (Charles River, St Aubin les Elbeuf, France). Mice were anesthetized with pentobarbitone (50 mg/kg IP) and tracheotomized. The right carotid artery and the left external jugular vein were catheterized to measure arterial pressure and to administer anesthetic and curare, respectively. The mice were then ventilated with room air (tidal volume 170 μL, 120 cycles/min) by means of a MiniVent type 845 mouse ventilator (Hugo Sachs Elektronik) and paralyzed with pancuronium bromide (250 μg/kg IV).
Mean arterial pressure (MAP) was calculated as diastolic pressure plus one third of the differential pressure. The heart rate (HR) was continuously monitored from the pressure signal with a Gould Biotach amplifier (model 13-4615-66).
Central Administration of Drugs
The head of each rabbit was placed after surgery in a stereotaxic frame (La Précision Cinématographique Française). Drugs were administered either into the cisterna magna (intracisternal injections) or microinjected into the NRL/RVLM (2 mm lateral to the midline, 4 mm rostral to the obex, and 8 mm from the dorsal surface), as described previously.23 At the beginning of each intracisternal experiment, a volume of cerebrospinal fluid, equal to the volume injected during the experiment, was withdrawn. The volume of injection (per dose) was 100 μL for intracisternal experiments and 0.5 μL for microinjections. However, for intracisternal experiments, (±)-2-(2-fluoro-5-methylphenyl)-4,5-dihydro-1H-imidazole (S23757) was given at 400 μL. When synergy was investigated, 10 minutes separated the 2 intracisternal injections. For microinjections, both drugs (each in a volume of 0.5 μL) were drawn up successively in the same needle and were administered very slowly in 1 injection. The site of injection was confirmed histologically as described previously.23
The mice were anesthetized as described above. The head of each mouse was placed in a mouse stereotaxic frame (Cunningham Mouse Adaptor, Stoelting). A stainless-steel guide cannula (external diameter 0.4 mm, internal diameter 0.3 mm; Cortat SA) was placed in the fourth ventricle (4thV, 5.88 mm caudal to the bregma and 2.8 mm from the surface of the skull; coordinates are according to Franklin and Paxinos24) and fixed with dental cement. The catheterization of vessels and the central drug injections (1 to 2 μL) were performed 48 hours after the stereotaxic procedures. After each experiment, the injection site was checked by an injection of 1 μL Evans blue through the guide cannula, so that to validate the experiment, the walls of the 4thV and the ventral surface of the medulla had to be stained in blue. Experiments in which the 4thV alone was stained were not considered. Intracerebroventricular injections were not conclusive in our experimental conditions. The Evans blue injected into the lateral ventricle was confined to this cavity and did not diffuse to the 4thV or to the medullary ventral surface.
LNP 509 was synthesized in our laboratory by Schann et al.19 S23757 was kindly provided by the Institut de Recherches Internationales Servier (Courbevoie, France). The drugs injected into the animals were dissolved in saline solution, and pH was adjusted to 7.4. CHO cells were kindly provided by Dr A.D. Strosberg (Paris, France).
Statistics and Calculations
Data are given as mean±SEM or mean±SE. Homogeneity of the initial cardiovascular parameters between groups was checked with an ANOVA. Mean values were compared by ANOVA (for repeated measures or intergroup) or by Student t tests (for paired or unpaired observations) as appropriate. Significance was set at P<0.05, and n indicates the number of experiments.
[35S]GTPγS Binding Assay
In CHO cells transfected with human α2AARs, noradrenaline (NA) induced [35S]GTPγS binding to G proteins (pEC50 6.18±0.03, Figure 1A). MK912, an α2AR antagonist,25 had no effect by itself but antagonized the effect of NA (10−5 mol/L) dose-dependently, with a pIC50 value of 7.56±0.08 (Figure 1A and 1B). LNP 509 had neither agonist nor antagonist effects, inasmuch as it never induced [35S]GTPγS binding and did not prevent the effect of NA (10−5 mol/L) on [35S]GTPγS binding (Figure 1A and 1B).
Cardiovascular Effects of LNP 509
Given intracisternally in cumulative doses (10 to 1000 μg/kg) to rabbits, LNP 509 dose-dependently decreased MAP (55±2 versus 102±3 mm Hg, ED50 184±18 μg/kg) and HR (221±9 versus 310±7 bpm) (values at dose of 1000 μg/kg). These effects were significant for doses ranging from 100 to 1000 μg/kg (P<0.05, n=5). Repeated intracisternal injections of vehicle did not significantly modify the hemodynamic parameters (Figure 2). The hypotensive response to 100 μg/kg LNP 509 was prevented by an intracisternal injection of 1 mg/kg S23757, an antagonist highly selective for I1BS.22 MAP decreased from 102±3 to 90±3 mm Hg in the control group (P<0.05, n=5), whereas it was stable at 97±4 mm Hg in the animals pretreated with S23757 (n=5).
Administered into the 4thV of mice at a dose of 1 mg/kg, LNP 509 induced a significant change in MAP in WT as well as D79N mice (59±2 versus 71±2 mm Hg, P<0.05 [n=13], and 60±2 versus 72±2 mm Hg, P<0.05 [n=13], respectively; Figure 3). The HR was not affected in either strain: 382±22 versus 407±25 bpm (WT mice) and 423±25 versus 433±24 bpm (D79N mice). In both WT and D79N mice, the 4thV injection of vehicle did not affect MAP (Figure 3). The HR was not modified either (data not shown).
Synergy Between LNP 509 and α-MNA Effects
To examine a possible synergism between the nonadrenergic and the α2-adrenergic mechanisms, we selected a subthreshold dose of LNP 509 and a threshold dose of α-MNA. At an intracisternal dose of 30 μg/kg, LNP 509 did not affect MAP in rabbits (95±2 versus 96±2 mm Hg, n=6) (Figure 4A). In 6 other rabbits, an intracisternal dose of 0.5 μg/kg α-MNA decreased MAP maximally from 100±2 to 90±1 mm Hg (P<0.05) (Figure 4A). In 6 other animals, α-MNA (0.5 μg/kg intracisternally) was given 10 minutes after LNP 509 (30 μg/kg intracisternally). MAP decreased immediately, and the maximal effect was reached within 5 minutes of injection (73±3 versus 101±2 mm Hg, P<0.05) (Figure 4A). This reduction in MAP was significantly different from that obtained with α-MNA or LNP 509 given alone (P<0.05). Moreover, LNP 509 injected twice at the same dose (30 μg/kg intracisternally) had no effect on MAP (−1±2%, n=4). Finally, we switched the order of injections; ie, α-MNA was injected first. The same maximal effect as when LNP 509 was administered first was observed, but it was delayed: it appeared 15 minutes after the injection of LNP 509 (Figure 4A). We obtained similar results when both drugs were microinjected together into the NRL/RVLM of the rabbits. LNP 509 (4 μg/kg) was ineffective on MAP (1±2%, P>0.05; n=5); 1 μg/kg α-MNA was also ineffective (0.5±2%, P>0.05; n=9). However, when LNP 509 was injected first and α-MNA was subsequently injected into the NRL/RVLM, MAP decreased by 15±2% (77±6 versus 90±5 mm Hg, P<0.05; n=5) (Figure 4B). Interestingly, when α-MNA was microinjected first, we did not observe any MAP variation.
This interaction was further investigated in mice. Administered to 5 WT mice, LNP 509 (1 mg/kg in 4thV) decreased MAP from 72±1 to 61±2 mm Hg (P<0.05) (Figure 4C). In 5 other WT animals, the 4thV injection of LNP 509 was followed by the 4thV injection of 10 μg/kg α-MNA (a dose that was ineffective on MAP, 70±4 versus 68±3 mm Hg; n=5). In that case, MAP decreased significantly from 72±5 to 49±4 mm Hg (P<0.05) (Figure 4C). This hypotension was significantly different from that induced by LNP 509 alone, whereas such a potentiation between the effects of LNP 509 and α-MNA was not observed when these drugs were administered in combination to D79N mice according to the same protocol: MAP changed from 71±1 to 60±2 mm Hg with LNP 509 alone (P<0.05, n=5) and from 74±3 to 65±3 mm Hg with LNP 509 plus α-MNA (P<0.05, n=5) (Figure 4C). These variations of MAP obtained in D79N mice were similar to those obtained in WT animals when LNP 509 was given alone.
Cardiovascular Effects of Rilmenidine in Mice
The effects of increasing doses of rilmenidine (3 μg/kg to 1 mg/kg, 4thV injection) were investigated in WT and D79N mice. Each animal received a single dose (5 animals per dose). In WT mice, rilmenidine lowered MAP in a dose-dependent manner, and the effect became significant from the dose of 30 μg/kg (Figure 5A): MAP decreased from 71±4 to 55±4 mm Hg at the highest dose (n=9). In transgenic mice, rilmenidine did not affect MAP, except at the dose of 1 mg/kg (65±3 versus 75±3 mm Hg, n=9) (Figure 5B). Although the hypotensive effect of 1 mg/kg rilmenidine was significantly weaker in D79N than in WT mice (P<0.05), it was very similar to the effect of LNP 509 given alone in both WT and transgenic mice.
We used α-MNA to mimic the α2-adrenergic activity of hybrid drugs. This catecholamine binds to α2ARs with high affinity13,21⇓ but does not recognize I1BS at all.13 It is a full agonist at human and murine α2AARs.21 Therefore, α-MNA can be considered a “pure” α2AAR agonist.
LNP 509 was used in the present study to reproduce the non–α2-adrenergic component of imidazoline-like drugs for 2 reasons. First, it has been shown recently (Schann et al19) that LNP 509 binds to I1BS with a pKi value of 5.98±0.15, whereas it has no detectable affinity at α1ARs, α2ARs, and I2BS; pKi values are >5. Since that study,19 the binding characterization of LNP 509 has been extended to a large variety of other receptors (serotonin, dopamine, β1 and β2, histamine, adenosine, endothelin, bradykinin, neuropeptide Y, cannabinoid, muscarinic, and nicotinic receptors), enzymes (monoamine oxidase A and B), and transporters (NA and serotonin). In all these systems, pKi values are >5.
Second, LNP 509 is devoid of activity at α2AARs, because in the present study, it was unable to induce or prevent [35S]GTPγS binding to the G proteins associated to human α2AARs expressed in CHO cells. Our experiments were performed on the α2A subtype because it has been shown that unlike the α2B and α2C subtypes, it is the subtype involved in the central hypotensive action of imidazoline-related compounds.11,12,26⇓⇓
Although LNP 509 was devoid of α2-adrenergic activity, after intracisternal injection to rabbits, it lowered BP as much as hybrid drugs do. This hypotensive action was prevented by S23757, an antagonist highly selective for I1BS.22 LNP 509 also caused hypotension when administered in the 4thV of WT and D79N mice. These in vivo data demonstrate (1) that LNP 509 is capable of reducing BP in 2 different species and (2) that this activity is independent of any action at α2ARs. The recurrent and justified objection against previous studies concerning imidazoline-induced hypotension was that drugs selective for I1BS over α2ARs could lose their selectivity at high doses and stimulate α2ARs to lower BP in animals with functional α2AARs. In the present work, thanks to the use of D79N mice, this objection is definitely forestalled. Therefore, although the involvement of other receptors cannot be definitively ruled out, the functional results reported in the present study, together with previous binding data, strongly suggest that the hypotensive action of LNP 509 involves I1BS.
It remains that rather high doses of LNP 509 were required to lower BP. This might be interpreted in at least 2 ways: (1) LNP 509 may be weakly efficient on I1BS, or (2) the α2-adrenergic and non–α2-adrenergic systems may work synergistically. Consequently, when one of them is impaired, the other is only weakly effective. It will not be possible to check the first hypothesis before a series of selective I1BS ligands with various efficacy levels is available, but with LNP 509, we had the appropriate pharmacological tool to test the second hypothesis. A marked potentiation of the effect of α-MNA by LNP 509 was indeed observed both in rabbits and in WT mice. This synergy was nonexistent in D79N mice. In these animals, the effect of LNP 509 only was obtained, and its amplitude was similar to that of the effect induced by the injection of LNP 509 alone in WT and D79N mice. Thus, a synergy between an α2-adrenergic and a non–α2-adrenergic mechanisms occurs. If one assumes that LNP 509 and α-MNA are full agonists at α2ARs, the effect of the combination of the 2 drugs should be additive. If one compound were a partial agonist and the other one were a full agonist at α2ARs, one should observe a subadditive response.27 In the present study, a supra-additive effect is obtained, supporting the assumption that LNP 509 and α-MNA act by different mechanisms.
Rilmenidine decreased MAP in a dose-dependent manner when administered centrally to WT mice. It also lowered BP in D79N mice, but only at the highest dose. This effect was comparable to the effect triggered by LNP 509 when given alone both in WT mice and in D79N mice. Therefore, we assume that the effect induced by rilmenidine in transgenic mice is representative of the non–α2-adrenergic mechanism that contributes to its hypotensive action. These data corroborate previous results showing the hypotensive action of moxonidine in D79N mice after central injection,28 but the data are in contrast to the data of the study reporting the lack of effect on BP of imidazoline-like drugs injected intravenously to such mice.11 The difference in the routes of drug administration or in the state of the animal (anesthetized versus conscious) might account for such a discrepancy. However, because rilmenidine is effective in WT mice whatever the route of injection (Zhu et al11 and present study), another hypothesis might be put forward. Considering the synergy aforementioned and the fact that rilmenidine is a hybrid drug,18,29⇓ one could assume that in this strain, rilmenidine triggers a synergy between α2-adrenergic and non–α2-adrenergic mechanisms. Thus, a lower dose of rilmenidine might be sufficient to decrease BP in WT mice than in transgenic mice, in which no synergy can occur. This synergy could explain the rightward shift of the dose-response curve of LNP 509 compared with that of rilmenidine in rabbits (Feldman et al23 and present study). Thus, rilmenidine may trigger the synergy by itself, whereas LNP 509, which only binds to I1BS, cannot do so.
In synergistic experiments, the sequence of injections was critical. In both species, the onset of hypotension was much faster when LNP 509 was given before α-MNA. This was even more striking when the drugs were microinjected into the NRL/RVLM of the rabbits. Hypotension was observed only when LNP 509 was injected before α-MNA. To explain this observation, one could assume that α-MNA is cleared rapidly after microinjection into the NRL/RVLM. We cannot rule out such a metabolic explanation, despite the fact that nothing is known about the NRL/RVLM region being capable of clearing catecholamines faster than other medullary regions, such as the NTS, where α-MNA has a marked effect on BP after microinjection.10 One can rather hypothesize that catecholamines are ineffective in the NRL/RVLM, unless I1BSs are previously occupied. Imidazoline-related compounds would make the hypotensive effects of catecholamines possible, at least in the NRL/RVLM. This assumption might explain why reasonable doses of catecholamines had no effect on BP when microinjected alone into this nucleus14 and why huge doses of α-MNA had to be injected to induce hypotension.30 Further study of the effect of treatment with imidazoline-like drugs on the expression of α2ARs on neuronal cells should further highlight our hypothesis.
The cell type on which such an interaction occurs remains unknown. It has recently been suggested that the sympatholytic action of imidazoline-like drugs is partly mediated by non-C1 bulbospinal neurons of the NRL/RVLM rather than by C1 adrenergic cells.31,32⇓ Whether these non-C1 neurons are the site of the interaction between α2-adrenergic and non–α2-adrenergic (I1 subtype) mechanisms remains to be investigated.
In conclusion, (1) the combined use of LNP 509, a compound with no activity at α2AARs, and of mice with impaired α2AARs definitely confirmed that a central imidazoline-sensitive, but non–α2-adrenergic, mechanism is involved in the BP control; (2) this mechanism might be mediated by I1BS; and (3) α2A-adrenergic and nonadrenergic mechanisms interact synergistically in the control of BP.
This study was supported by the Institut de Recherches Internationales Servier, by the Société Française de Pharmacologie, and by National Institutes of Health grant HL-43671 (which funded the development of D79N mice). We are very grateful to Drs Jean Boutin and Nelly Fabry for their help and advice.
↵*Drs Bruban and Estato contributed equally to this work.
- ↵Sun MK. Central neural organization and control of sympathetic nervous system in mammals. Prog Neurobiol. 1995; 47: 157–233.
- ↵Scheinin M, Lomasney JW, Hayden-Hixson DM, et al. Distribution of alpha-2 adrenergic receptor subtype gene expression in rat brain. Mol Brain Res. 1994; 21: 133–149.
- ↵Rosin DL, Zeng D, Stornetta RL, et al. Immunohistochemical localization of alpha 2A-adrenergic receptors in catecholaminergic and other brainstem neurons in the rat. Neuroscience. 1993; 56: 139–155.
- ↵Ruffolo RR Jr, Nichols AJ, Stadel JM, et al. Pharmacologic and therapeutic applications of α2-adrenoceptor subtypes. Annu Rev Pharmacol Toxicol. 1993; 32: 243–279.
- ↵Timmermans PBMWM, Schoop AMC, Kwa HY, et al. Characterization of α-adrenoceptors participating in the central hypotensive and sedative effects of clonidine using yohimbine, rauwolscine and corynanthine. Eur J Pharmacol. 1981; 70: 7–15.
- ↵Bylund DB, Eikenberg DC, Hieble JP, et al. International Union of Pharmacology nomenclature of adrenoceptors. Pharmacol Rev. 1994; 46: 121–136.
- ↵MacDonald E, Kobilka BK, Scheinin M. Gene targeting-homing in on α2-adrenoceptor-subtype function. Trends Pharmacol Sci. 1997; 18: 211–219.
- ↵Lakhlani PP, MacMillan LB, Guo TZ, et al. Substitution of a mutant α2a-adrenergic receptor via “hit and run” gene targeting reveals the role of this subtype in sedative, analgesic, and anesthetic-sparing responses in vivo. Proc Natl Acad Sci U S A. 1997; 94: 9950–9955.
- ↵Kable JW, Murrin LC, Bylund DB. In vivo gene modification elucidates subtype-specific functions of α2-adrenergic receptors. J Pharmacol Exp Ther. 2000; 293: 1–7.
- ↵Zandberg P, De Jong W, De Wied D. Effect of catecholamine-receptor stimulating agents on blood pressure after local application in the nucleus tractus solitarii of the medulla oblongata. Eur J Pharmacol. 1979; 55: 43–56.
- ↵Zhu QM, Lesnick JD, Jasper JR, et al. Cardiovascular effects of rilmenidine, moxonidine and clonidine in conscious wild-type and D79N α2A-adrenoceptor transgenic mice. Br J Pharmacol. 1999; 126: 1522–1530.
- ↵MacMillan LB, Hein L, Smith MS, et al. Central hypotensive effects of the α2A-adrenergic receptor subtype. Science. 1996; 273: 801–803.
- ↵Ernsberger P, Giuliano R, Willette RN, et al. Role of imidazole receptors in the vasodepressor response to clonidine analogs in the rostral ventrolateral medulla. J Pharmacol Exp Ther. 1990; 253: 408–418.
- ↵Bousquet P, Feldman J, Schwartz J. Central cardiovascular effects of α-adrenergic drugs: differences between catecholamines and imidazolines. J Pharmacol Exp Ther. 1984; 230: 232–236.
- ↵Bousquet P, Feldman J, Bloch R, et al. The nucleus reticularis lateralis: a region highly sensitive to clonidine. Eur J Pharmacol. 1981; 69: 389–392.
- ↵Bousquet P, Dontenwill M, Greney H, et al. I1-Imidazoline receptors: an update. J Hypertens. 1998; 16 (suppl 3): S1–S5.
- ↵Ernsberger P, Haxhiu MA. The I1-imidazoline-binding site is a functional receptor mediating vasodepression via the ventral medulla. Am J Physiol. 1997; 273: R1572–R1579.
- ↵Bricca G, Dontenwill M, Molines A, et al. Rilmenidine selectivity for imidazoline receptors in human brain. Eur J Pharmacol. 1989; 163: 373–377.
- ↵Schann S, Bruban V, Pompermayer K, et al. Synthesis and biological evaluation of pyrrolinic isosteres of rilmenidine: discovery of cis-/trans-dicyclopropylmethyl-(4,5-dimethyl-4,5-dihydro-3H-pyrrol-2-yl)-amine (LNP 509), an I1 imidazoline receptor selective ligand with hypotensive activity. J Med Chem. 2001; 44: 1588–1593.
- ↵Michel MC, Insel PA. Are there multiple imidazoline binding sites? Trends Pharmacol Sci. 1989; 10: 342–344.
- ↵Jasper JR, Lesnick JD, Chang LK, et al. Ligand efficacy and potency at recombinant α2 adrenergic receptors: agonist-mediated [35S]GTPγS binding. Biochem Pharmacol. 1998; 55: 1035–1043.
- ↵Bruban V, Feldman J, Greney H, et al. Respective contributions of α-adrenergic and non-adrenergic mechanisms in the hypotensive effect of imidazoline-like drugs. Br J Pharmacol. 2001; 133: 261–266.
- ↵Feldman J, Tibiriça E, Bricca G, et al. Evidence for the involvement of imidazoline receptors in the central hypotensive effect of rilmenidine in the rabbit. Br J Pharmacol. 1990; 100: 600–604.
- ↵Franklin KBJ, Paxinos G. The Mouse Brain in Stereotaxic Coordinates. San Diego: Academic Press; 1997.
- ↵Pettibone DJ, Clineschmidt BV, Lotti VJ, et al. Pharmacological profile of a new potent and specific α2-adrenoceptor antagonist, L-657,743. Naunyn Schmiedebergs Arch Pharmacol. 1987; 336: 169–175.
- ↵Link RE, Desai K, Hein L, et al. Cardiovascular regulation in mice lacking α2-adrenergic receptor subtypes b and c. Science. 1996; 273: 803–805.
- ↵Leff P, Dougall IG, Harper D. Estimation of partial agonist affinity by interaction with a full agonist: a direct operational model-fitting approach. Br J Pharmacol. 1993; 110: 239–244.
- ↵Tolentino-Silva FP, Haxhiu MA, Waldbaum S, et al. α2-Adrenergic receptors are not required for central anti-hypertensive action of moxonidine in mice. Brain Res. 2000; 862: 26–35.
- ↵Bousquet P, Feldman J. Drugs acting on imidazoline receptors: a review of their pharmacology, their use in blood pressure control and their potential interest in cardioprotection. Drugs. 1999; 58: 799–812.
- ↵Head GA, Burke SL. Relative importance of medullary brain nuclei for the sympatho-inhibitory actions of rilmenidine in the anaesthetized rabbit. J Hypertens. 1998; 16: 503–517.
- ↵Schreihofer AM, Guyenet PG. Role of presympathetic C1 neurons in the sympatholytic and hypotensive effects of clonidine in rats. Am J Physiol. 2000; 279: R1753–R1762.
- ↵Schreihofer AM, Stornetta RL, Guyenet PG. Regulation of sympathetic tone and arterial pressure by rostral ventrolateral medulla after depletion of C1 cells in rat. J Physiol. 2000; 529: 221–236.