Dexmedetomidine is an α2-adrenoceptor agonist and anesthetic. The present study was designed to characterize the receptor subtypes and the downstream mechanisms of the vascular effects of dexmedetomidine in small (mesenteric artery) and large (aorta) arteries ex vivo. Isometric tension was measured in Sprague-Dawley rat mesenteric and aortic rings (with or without endothelium). To study relaxations, cumulative concentrations of dexmedetomidine, 5-bromo-N-(2-imidazolin-2-yl)-6-quinoxalinamine, (UK14304), or clonidine were added to rings contracted with 9,11-dideoxy-9α,11α-methanoepoxy prostaglandin F2α (U46619) in the presence or absence of indomethacin; Nω-nitro-l-arginine methyl ester hydrochloride (l-NAME); 2-[2H-(1-methyl-1,3-dihydroisoindole)methyl]-4,5-dihydroimidazole maleate (BRL44408); 2-[2-(4-(2-methoxyphenyl)piperazin-1-yl)ethyl]-4,4-dimethyl-1,3-(2H,4H)-isoquinolindione dihydrochloride (ARC239); l-657,743, (2S-trans)-1,3,4,5′,6,6′,7,12b-octahydro-1′,3′-dimethyl-spiro[2H-benzofuro[2,3-a]quinolizine-2,4′(1′H)-pyrimidin]-2′(3′H)-one hydrochloride hydrate (MK912); rauwolscine; prazosin; or pertussis toxin. To study contractions, dexmedetomidine was added to quiescent rings without endothelium or after incubation with l-NAME, rauwolscine, prazosin, indomethacin, or 3-[(6-amino-(4-chlorobenzensulfonyl)-2-methyl-5,6,7,8-tetrahydronaphth)-1-yl]propionic acid (S18886). Dexmedetomidine evoked relaxation at low concentrations (10 pM–30 nM) followed by contraction at higher concentrations (>30 nM) in the mesenteric artery. In the aorta, the relaxation was significantly smaller. The relaxation to dexmedetomidine depended on nitric oxide, endothelium, and Gi protein, and it was mediated by α2A/D-adrenoceptors and possibly α2B-adrenoceptors. The contraction was mediated mainly by α2B- and α1-adrenoceptors and involved the action of prostanoids. UK14304 and clonidine induced greater and smaller relaxations, respectively, than dexmedetomidine. In conclusion, depending on the concentration used and the presence of functional endothelium, dexmedetomidine may evoke both relaxation and contraction in isolated arteries. The vascular effects also vary depending on the blood vessel studied. Its vascular effect is different from that of clonidine and UK14304.
Dexmedetomidine is an α2-adrenoceptor (α2-AR) agonist used widely for sedation and as an adjunct in general anesthesia, in the operating room, in endoscopy suites, as well as in the intensive care unit. In general, the vascular effects of α2-AR agonists are complex, in that they may be the result of the interplay between a central sympatholytic effect (MacMillan et al., 1996), presynaptic α2-AR effect, and postsynaptic α2-AR effect (Figueroa et al., 2001; Guimarães and Moura, 2001). Joshi et al. (2007) and Kim et al. (2009) demonstrated that dexmedetomidine causes nitric oxide (NO) release from human umbilical vein endothelial cells. Kim et al. (2009) also reported that dexmedetomidine contracts isolated rat aortae without endothelium. The latter contraction is attenuated by phospholipase A2, lipoxygenase, and cyclooxygenase (COX) inhibitors (Kim et al., 2009). Hamasaki et al. (2002) and Yildiz et al. (2007) have also shown that dexmedetomidine potentiates the contraction to 40 mM KCl in the isolated human gastroepiploic and internal mammary arteries without endothelium and that this potentiation is attenuated by the α2-AR antagonists yohimbine and rauwolscine. Dexmedetomidine constricts human dorsal hand veins in vivo (Snapir et al., 2009; Muszkat et al., 2010). However, several questions concerning the vascular effects of dexmedetomidine remain unanswered. First, it is still unknown whether dexmedetomidine can relax isolated arteries with endothelium, given that it releases NO from human umbilical vein endothelial cells (Joshi et al., 2007; Kim et al., 2009). Second, it is not known whether endothelium-derived contracting factors (Félétou et al., 2009) are involved in the contraction mediated by dexmedetomidine, given the dependence of this contraction on the phospholipase A2-COX-lipoxygenase pathways. Third, it is not known whether these vascular effects of dexmedetomidine differ in large and small arteries, given the known different distribution of α2-AR in arteries of different sizes (Flavahan et al., 1987). Finally, α2-ARs are divided into the subtypes α2A/D, α2B, and α2C. Whether dexmedetomidine mediates its complex vascular effects via one subtype or another has not yet been determined.
We believe that it is important to understand these various possible effects of dexmedetomidine to use the drug more safely and effectively under varied clinical situations. We have therefore conducted the present study to investigate these four issues.
Materials and Methods
Male Sprague-Dawley (SD) rats (10 weeks old) were studied. The animals were kept in a temperature-controlled room (21 ± 1°C) with a 12-h light/dark cycle and given free access to standard laboratory chow and tap water. They were anesthetized with pentobarbital sodium (70 mg/kg i.p.). Their superior mesenteric arteries and thoracic aortae were excised and put into Krebs-Henseleit solution of the following composition: 120 mM NaCl, 4.8 mM KCl, 1.25 mM CaCl2, 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, and 5.5 mM glucose (control solution). The blood vessels were dissected free of fat and connective tissue and cut into rings of 2 to 3 mm in length for the measurement of isometric force. Each harvested artery was cut into five or six such ring segments that were studied in parallel in organ chambers. Each ring was exposed to a different treatment, and no single ring was exposed to more than one treatment.
In some preparations, the endothelium of the aortic rings was removed mechanically by inserting the tip of a syringe needle and rolling it back and forth in a Sylgard-based container (Dow Corning, Midland, MI) filled with control solution, whereas in mesenteric arteries, the endothelium was removed by perfusion of 1 ml of 0.1% Triton X-100 solution for 30 s before cutting the rings. Integrity of the smooth muscle layer after endothelial removal was confirmed by normal contraction to 30 mM KCl. The present investigations were approved by the Committee for the Use of Laboratory Animals for Teaching and Research, the University of Hong Kong.
Aortic and mesenteric rings from SD rats were suspended in organ chambers containing the control solution (37°C) and were continuously aerated with 95% O2 and 5% CO2. The rings were connected to force transducers (ADInstruments Pty Ltd., Sydney, Australia) for isometric tension recording (PowerLab; ADInstruments). They were allowed to equilibrate for 1.5 h to reach their optimal resting tensions, which were ∼2.5 g for aortic rings and ∼1 g for mesenteric rings, respectively (data not shown).
After obtaining two reference contractions to 60 mM KCl, in some preparations, the presence of intact endothelium was confirmed by adding 1 μM acetylcholine to rings contracted with 1 μM phenylephrine (Furchgott and Vanhoutte, 1989). The endothelium was considered viable when relaxation to acetylcholine was 80% or greater. The rings were then washed thoroughly, incubated with different blockers, including l-NAME [nitric-oxide synthase (NOS) inhibitor, 100 μM] (Lüscher and Vanhoutte, 1986), rauwolscine (α2-AR antagonist; 10 nM–1 μM), BRL44408 (α2A/D-AR antagonist; 3–300 nM), 2-[2-(4-(2-methoxyphenyl)piperazin-1-yl)ethyl]-4,4-dimethyl-1,3-(2H,4H)-isoquinolindione dihydrochloride (ARC239, α2B-AR antagonist; 10 nM–1 μM), l-657,743, (2S-trans)-1,3,4,5′,6,6′,7,12b-octahydro-1′,3′-dimethyl-spiro[2H-benzofuro[2,3-a]quinolizine-2,4′(1′H)-pyrimidin]-2′(3′H)-one hydrochloride hydrate (MK912, α2C-AR antagonist; 300 pM–30 nM), and prazosin (α1-AR antagonist; 1 μM) (Görnemann et al., 2007) for 40 min, except for incubation with pertussis toxin (Gi protein inhibitor; 400 ng/ml), which lasted 2 h (Ng et al., 2008).
To study relaxations, the rings were first incubated with indomethacin (nonselective COX inhibitor, 10 μM); preliminary experiments had shown that this preparation provided the best relaxation condition (Fig. 1A) for studying the effects of different blockers. The rings were then contracted with 9,11-dideoxy-9α,11α-methanoepoxy prostaglandin F2α [U46619, synthetic thromboxane-prostanoid (TP) receptor agonist, 10–100 μM], targeting a final tension of approximately 50% of that of 60 mM KCl. After a stabilized contraction to U46619 was achieved, cumulative concentrations of dexmedetomidine (10 pM–10 μM) were added. To compare the vascular effects of dexmedetomidine with other α2-adrenergic agonists, cumulative concentrations of 5-bromo-N-(2-imidazolin-2-yl)-6-quinoxalinamine, 5-bromo-N-(4,5-dihydro-1H-imidazol-2-yl)-6-quinoxalinamine, brimonidine (UK14304; 100 pM–10 μM) or clonidine (1 nM–10 μM) were used in some experiments. Relaxations were expressed as percentages of the contractions to U46619.
To study contractions, cumulative concentrations of dexmedetomidine (10 nM–10 μM) were added to quiescent rings, after incubation with different blockers, including l-NAME, prazosin, rauwolscine, indomethacin, and 3-[(6-amino-(4-chlorobenzensulfonyl)-2-methyl-5,6,7,8-tetrahydronaphth)-1-yl]propionic acid (S18886, TP receptor antagonist; 100 nM) (Tang et al., 2005), for 40 min. Contractions were expressed as percentages of the reference contraction caused by 60 mM KCl.
Acetylcholine, BRL44408, clonidine, indomethacin, MK912, l-NAME, prazosin, phenylephrine, rauwolscine, and UK14304 were purchased from Sigma-Aldrich (St. Louis, MO). ARC239 was purchased from Tocris Bioscience (Bristol, UK). Dexmedetomidine was purchased from Abbott Laboratories (North Chicago, IL). Pertussis toxin was purchased from List Biological Laboratories Inc. (Campbell, CA). U46619 was purchased from Biomol International Inc. (Exeter, UK). S18886 was a kind gift of the Institut de Recherches Servier (Paris, France).
Stock solutions of dexmedetomidine and UK14304 were prepared in dimethyl sulfoxide. A stock solution of indomethacin was prepared in 5 mM sodium bicarbonate solution. A stock solution of pertussis toxin was prepared in 0.1 M sodium sulfate buffer. A stock solution of U46619 was prepared in absolute ethanol. All other compounds were prepared in deionized water. Concentrations are expressed as final molar concentrations in the bath solution.
The results are presented as means ± S.E.M., with n representing the number of rats. Data were analyzed and curve fitting performed using Prism statistical software (ver. 4.02; GraphPad Software Inc., San Diego, CA). The effects of different inhibitors on contractions and relaxations by dexmedetomidine were compared by two-way analysis of variance followed by a Bonferroni post test. P values less than 0.05 were considered to indicate statistically significant differences.
In all preparations, relaxations to dexmedetomidine were more pronounced in the presence of 10 μM indomethacin (Fig. 1A). Therefore, all subsequent relaxation experiments were performed in the presence of indomethacin.
Contracted Mesenteric Artery.
Dexmedetomidine affected mesenteric rings contracted with U46619 in a biphasic manner. It caused a concentration-dependent relaxation from 10 pM to 30 nM, followed by a rebound in tension at concentrations between 30 nM and 10 μM (Figs. 1B and 2A). Treatment with the NOS inhibitor l-NAME or removal of the endothelium abolished the relaxation and, at the same time, potentiated the secondary contraction (Fig. 2A).
α2-AR antagonist rauwolscine at 1 μM significantly reduced the dexmedetomidine-induced endothelium-dependent relaxation (P < 0.05; Fig. 2B). α2A/D-AR antagonist BRL44408 and, to a lesser extent, α2B-AR antagonist ARC239 caused a rightward shift of the concentration-response curves compared with control (Fig. 2, C and D). ARC239, at the same time, significantly reduced the secondary contraction to dexmedetomidine (P < 0.05), whereas α2C-AR antagonist MK912 (30 nM) had virtually no effect on either relaxation or contraction (Fig. 2E).
The Gi protein inhibitor pertussis toxin (400 ng/ml) led to a significant reduction in endothelium-dependent relaxations. It did not significantly alter the levels of the secondary contraction. The contractile response induced by dexmedetomidine was abolished by α1-AR antagonist prazosin (1 μM), whereas the relaxation was not affected (Fig. 2F). Analyzing the concentration-relaxation curve in the presence of prazosin, the Emax and EC50 value of dexmedetomidine averaged 60.84 ± 3.48 and −8.43 ± 0.22, respectively (compared with acetylcholine-induced relaxation in the same preparations, Emax = 109 ± 1.25 and EC50 = −8.0 ± 0.05).
Dexmedetomidine induced negligible relaxations in rat aortic rings compared with those in mesenteric arteries (Fig. 1A). Small relaxations of the aortic rings were observed after prior treatment with ARC239 and prazosin (data not shown).
Cumulative concentrations of dexmedetomidine between 10 nM and 10 μM were added to mesenteric rings without previous exposure to U46619. A concentration-dependent contraction resulted in rings without endothelium and in those treated with l-NAME (100 μM). No significant contractions were observed in rings with endothelium (Fig. 3A). Rauwolscine (100 nM; Fig. 3B) and indomethacin (10 μM; Fig. 3C) reduced the contraction induced by dexmedetomidine, and prazosin (1 μM; Fig. 3B) completely blocked the contraction. TP receptor antagonist S18886 (100 nM) did not significantly affect the level of contraction (Fig. 3C). Similar data were obtained in the quiescent aorta (data not shown).
Comparison with Other α2-Adrenergic Agonists.
In mesenteric rings contracted with U46619, dexmedetomidine caused a biphasic response, with concentration-dependent relaxation followed by contraction. UK14304 (100 pM–10 μM) evoked concentration-dependent relaxations only, whereas clonidine (1 nM–10 μM) did not have significant effect (Fig. 4A). Prazosin (1 μM) significantly unmasked a relaxation to clonidine (Fig. 4B). In aortic rings, minimal relaxations were induced by UK14304 and clonidine (data not shown).
This study demonstrates the entire spectrum of direct effects of dexmedetomidine in isolated blood vessels, and we determined that the vascular effects of dexmedetomidine depend on the concentration of dexmedetomidine and the type of blood vessel studied. In small arteries, dexmedetomidine evoked endothelium-dependent relaxations at low concentrations (10 pM–30 nM) followed by secondary contraction (30 nM–10 μM). l-NAME, the eNOS inhibitor, and rauwolscine, the α2-AR blocker, abolished endothelium-dependent relaxation because of dexmedetomidine. These findings show that dexmedetomidine causes endothelium-dependent relaxation by activating endothelial α2-AR. Activated endothelial α2-ARs increases the formation of NO by eNOS (Kim et al., 2009) and endothelium-dependent relaxation results (Vanhoutte, 2001). In addition, by using several selective α2-subtype adrenoceptor antagonists (Görnemann et al., 2007), this study suggests that the relaxant effects of dexmedetomidine are mediated mostly by the subtype α2A/D-AR and, to a lesser extent, α2B-AR.
Gi protein inhibitor pertussis toxin (Flavahan et al., 1989; Miller et al., 1991; Ng et al., 2008) significantly reduced the endothelium-dependent relaxation to dexmedetomidine, further confirming that this relaxation is α2A/D-mediated because activation of this receptor subtype in endothelial cells is transduced via Gi protein (Vanhoutte, 2003). By contrast, the contraction to dexmedetomidine is little affected. This is in line with previous findings that Gi proteins are coupled to α2-AR but not α1-AR (Miller et al., 1991). The dependence of dexmedetomidine-induced relaxation on NO and Gi protein function suggests that this relaxation is vulnerable in arteries with atherosclerosis or other causes of endothelial dysfunction, because Gi-dependent responses are impaired in atherosclerotic and regenerated endothelium (Flavahan, 1993; Shibano et al., 1994). This implies that the vascular effects of dexmedetomidine in patients with atherosclerosis may be quite different from those in healthy subjects.
Unlike in the mesenteric artery, dexmedetomidine did not induce endothelium-dependent relaxations in the rat aorta, although a small relaxation develops after incubation with ARC239 or prazosin. These observations suggest that any relaxant effect of dexmedetomidine in the aorta is masked by the contraction it causes by activating α2B- and α1-ARs. This conclusion is in line with the findings that α1-ARs predominate in the rat aorta (Aboud et al., 1993) and that, in general, α2-ARs are more prominent in smaller arteries and α1-ARs in larger ones (Flavahan et al., 1987). The mesenteric artery may possess a higher density of α2-AR than the aorta, resulting in greater relaxations to dexmedetomidine.
In contracted mesenteric arteries, dexmedetomidine caused a secondary increase in tension at high concentrations. These contractions are abolished in the presence of either α2B or α1 antagonists. Likewise, in quiescent rings, dexmedetomidine induces a concentration-dependent contraction after removal of the endothelium or incubation with l-NAME. Preliminary findings suggest that, in actuality, the NOS inhibitor also augments the contraction to dexmedetomidine in endothelium-denuded arteries (unpublished observations), a surprising finding that is currently under investigation. The present results in isolated animal arteries concur with those by Kim et al. (2009), Yildiz et al. (2007), and Hamasaki et al. (2002), as well as those obtained in human small arteries in vivo, where NOS inhibition enhanced the dexmedetomidine-induced peripheral vasoconstriction (Snapir et al., 2009). These contractions are abolished by prazosin and significantly reduced by the presence of rauwolscine, suggesting the involvement of both the α1-and α2-ARs in the response. Similar to findings by Kim et al. (2009), indomethacin also reduces the contractile levels of dexmedetomidine significantly, but these contractions are unaffected by the TP receptor antagonist S18886, which argues against an involvement of endothelium-derived contracting factors (Félétou et al., 2009).
In vivo dexmedetomidine induces constriction of human dorsal hand veins (Snapir et al., 2009; Muszkat et al., 2010). Although the constriction is consistent, there is substantial variability in sensitivity to dexmedetomidine, and the mechanism underlying this variability remains elusive (Kurnik et al., 2008; Muszkat et al., 2010). It is noteworthy that α2-AR-mediated effects are different in venous and arterial blood vessels (Wright et al., 1995; Muszkat et al., 2010). For instance, in both the present study and that by Snapir et al. (2009), eNOS inhibition is required for the manifestation of dexmedetomidine-induced contractions of arteries, whereas venous constriction does not require such treatment (Muszkat et al., 2005a,b, 2010). Taking these and the present findings in conjunction, in terms of clinical relevance, the biphasic vascular response of dexmedetomidine observed in contracted small arteries, the contribution of α1-AR activity to the contraction, and the contraction observed in quiescent arteries that lack endothelium or NOS activity suggests that the effect of dexmedetomidine on systemic vascular resistance and blood pressure may depend on the dose, context, and patient. Vasoconstriction will probably predominate at plasma concentrations above 10 nM (approximately 2.7 ng/ml, which corresponds to moderate to deep sedation in healthy volunteers) (Ebert et al., 2000) and in patients with endothelial dysfunction. The biphasic response in the present study is mirrored by a similar change in mean arterial pressure when dexmedetomidine is infused systemically at increasing doses in healthy subjects (Ebert et al., 2000). The vasoconstriction may be attenuated and the vasodilation at low concentrations may be amplified in the presence of indomethacin, which suggests that these responses may be different in patients taking nonselective COX inhibitors (nonsteroidal anti-inflammatory drugs). Although in the present study the NO-mediated relaxation caused by dexmedetomidine was more pronounced in preparations incubated with indomethacin, which casts some doubt on the clinical relevance of the observation, patients receiving COX inhibitors such as aspirin and nonsteroidal anti-inflammatory drugs are frequently encountered. However, we have to emphasize that these in vivo variations are only postulations and have not been directly demonstrated in the present study. Nevertheless, they may be plausible mechanisms underlying the high individual variability in dexmedetomidine sensitivity.
The present study demonstrates that the vascular responses to dexmedetomidine differ from those of UK14304 and clonidine. Differences in α2- to α1-AR selectivity may contribute to these differences. The order of α2- to α1-adrenoceptor selectivity is UK14304 > dexmedetomidine > clonidine (Bhana et al., 2000; Görnemann et al., 2007). These results imply that different α2-AR agonists have rather different vascular effects, and caution should be exercised in extrapolating the known vascular effects of other α2-AR agonists, in particular that of clonidine, to predict that of dexmedetomidine.
In conclusion, dexmedetomidine has complex direct vascular effects. It induces endothelium-dependent relaxation at low concentrations in contracted small arteries and contraction at higher concentrations. In quiescent arteries lacking NOS activity or endothelium, it induces contraction, whereas in quiescent arteries with endothelium, the relaxing and contracting effects of dexmedetomidine oppose each other, resulting in little change in net tension. Different α2-AR subtypes as well as α1-AR are involved in these vascular effects (Fig. 5). Therefore, vasoconstriction or vasodilation may be possible after administration of dexmedetomidine clinically, depending on the dose and context of administration. Last, caution should be taken when extrapolating the known vascular responses to other drugs (e.g., clonidine) to that of dexmedetomidine, as the present findings demonstrate there are significant differences between different α2-AR agonists.
This work was supported in part by the general research fund of the Research Grants Council of Hong Kong [Grant HKU 773509M] and a University of Hong Kong Seed Funding for Basic Research Grant [Grant 200710159019] and partly by departmental funds.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- nitric oxide
- Sprague Dawley
- l- name
- Nω-nitro-l-arginine methyl ester hydrochloride
- nitric-oxide synthase
- 2-[2H-(1-methyl-1,3-dihydroisoindole)methyl]-4,5-dihydroimidazole maleate
- 2-[2-(4-(2-methoxyphenyl)piperazin-1-yl)ethyl]-4,4-dimethyl-1,3-(2H,4H)-isoquinolindione dihydrochloride
- l-657,743, (2S-trans)-1,3,4,5′,6,6′,7,12b-octahydro-1′,3′-dimethyl-spiro[2H-benzofuro[2,3-a]quinolizine-2,4′(1′H)-pyrimidin]-2′(3′H)-one hydrochloride hydrate
- 9,11-dideoxy-9α,11α-methanoepoxy prostaglandin F2α
- 3-[(6-amino-(4-chlorobenzensulfonyl)-2-methyl-5,6,7,8-tetrahydronaphth)-1-yl]propionic acid
- endothelial NO synthase.
- Received June 22, 2010.
- Accepted September 9, 2010.
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics