General anesthetics threaten cardiovascular stability by causing changes in cardiac function, vascular reactivity, and cardiovascular reflexes and significantly alter distribution of cardiac output to various organs. Their overall impact is often systemic hypotension, which is attributable to myocardial depression, peripheral vasodilation, and attenuated sympathetic nervous system activity. However, one could be more causative than the others, depending on anesthetic agents and cardiovascular factors inherent in patients (e.g., coexisting heart disease). It is generally believed that most general anesthetics attenuate sympathetic nervous system outflow from the central nervous system, thereby decreasing vascular resistance in peripheral circulations. Indeed, in previous in vivo studies, during administration of various general anesthetics, vascular resistance was decreased in most peripheral circulations; however, it was unaffected or increased in some peripheral circulations. General anesthetics may act directly on vascular smooth muscle and/or endothelial cells in various vascular beds, influencing total peripheral and/or regional vascular resistance, and hence organ blood flow. This article reviews previously reported direct (i.e., nonneural) vascular actions of general anesthetics and discusses their underlying mechanisms, their in vivo relevance, and the future of research for general anesthetic vascular pharmacology.
VASCULAR tone is regulated by mechanisms that are intrinsic to the organ, as well as by extrinsic mechanisms, such as the autonomic nervous system and circulating hormones. The vascular endothelium releases a variety of vasoactive mediators in response to various neurohumoral and physicochemical stimuli, contributing to both intrinsic and extrinsic regulation. The intrinsic mechanisms—i.e. , metabolic and myogenic mechanisms—are considered essential for fine adjustment of regional vascular resistance to meet the ever-changing metabolic needs in regional tissues. On the other hand, extrinsic mechanisms actively regulate vascular tone to maintain arterial pressure and to alter distribution of cardiac output in response to various physiologic and pathologic stresses (e.g. , hemorrhage). In particular, the sympathetic nervous system plays a major role in regulating vascular tone of both resistance and capacitance vessels in peripheral circulations to sustain systemic vascular resistance and venous return, respectively, and hence systemic arterial pressure. Cardiovascular reflexes (e.g. , baroreceptor reflex) are essential for short-term regulation of the circulation; namely, in response to moment-to-moment changes in hemodynamics, they rapidly control myocardial contractility, peripheral vascular tone, and heart rate, using the autonomic nervous system. The relative importance of each mechanism varies in different vascular beds.
General anesthetics threaten cardiovascular stability by causing changes in cardiac function, vascular reactivity, and cardiovascular reflexes. Their overall impact is often systemic hypotension, which is attributable to myocardial depression, peripheral vasodilation, and decreased sympathetic nervous system activity; however, one could be more causative than the others, depending on anesthetic agents and cardiovascular factors inherent in patients (e.g. , heart disease).1–9General anesthetics also significantly alter blood flow to various organs.1,5,6,10–16Most general anesthetics are believed to act on multiple sites within the sympathetic nervous system, decreasing its influence on the peripheral vasculature and hence peripheral vascular resistance.17Indeed, in previous in vivo studies, during administration of various general anesthetics, vascular resistance or tone was decreased in most peripheral vascular beds, although it was unaffected or increased in some peripheral vascular beds.1,10–13,18–20General anesthetics may directly influence cellular mechanisms regulating vascular reactivity in various vascular beds, thereby altering total peripheral and/or regional vascular resistance. To gain access to the validity of these hypotheses, to date, numerous studies have investigated direct (i.e. , nonneural) actions of general anesthetics on vascular smooth muscle cells (VSMCs) and endothelial cells from various vascular beds under in vitro or in situ conditions.
This article reviews previous literature evaluating the direct vascular actions of general anesthetics and discusses their underlying mechanisms, their in vivo relevance, and the future of research for general anesthetic vascular pharmacology. General anesthetics dealt with in this article include halogenated volatile anesthetics, intravenous nonopioid anesthetics (i.e. , barbiturates, ketamine, propofol, etomidate, benzodiazepines), and intravenous opioid anesthetics (i.e. , morphine, fentanyl), all of which are currently available for clinical practice.
Cellular Mechanisms Regulating the Contractile State of Vascular Smooth Muscle Cells
This section briefly reviews cellular mechanisms regulating vascular tone, which have been reported to be influenced by various types of general anesthetics as described later.
Role of Calcium
Changes in cytosolic free Ca2+concentration ([Ca2+]c) are the principal mechanisms that regulate contractile state of VSMCs (i.e. , vascular tone). Namely, an increase and a decrease in [Ca2+]c result in vasoconstriction and vasorelaxation, respectively.21The Ca2+-dependent activation of myosin light chain kinase (MLCK) and its phosphorylation of 20-kd myosin light chain (MLC20) are the primary mechanisms responsible for the initial development of contractile force.21,22However, during its subsequent maintenance, the Ca2+sensitivity of MLC20phosphorylation can be secondarily modulated by other signaling pathways.21,23–27In addition, some regulatory mechanisms that maintain high contractile force at low energy (i.e. , adenosine 5′-triphosphate [ATP]) cost—i.e. , low levels of MLC20phosphorylation—may also contribute to its maintenance.28,29
In response to vasoconstrictor stimuli, Ca2+is mobilized from intracellular stores (i.e. , sarcoplasmic reticulum [SR]) and/or the extracellular space to increase the [Ca2+]c in VSMCs (fig. 1). The increase in [Ca2+]c, in turn, stimulates the binding of Ca2+to calmodulin (CaM). The Ca2+–CaM complex then activates MLCK to phosphorylate myosin at MLC20on serine 19 (Ser19), allowing myosin ATPase to be activated by actin and the muscle to contract as a result of cyclic interactions between myosin and actin.21,22,30Ser19of MLC20also can be phosphorylated by Ca2+/CaM-dependent protein kinase II (CaMKII); however, it occurs only at a very slow rate and probably does not contribute to the initiation of contraction.21
In response to vasodilator stimuli or removal of the vasoconstrictor stimuli, the [Ca2+]c decreases mainly as a result of plasmalemmal extrusion and/or uptake into the SR. When the [Ca2+]c decreases to lower than 1 μm, CaM dissociates from MLCK to inactivate MLCK. Under these conditions, myosin light chain phosphatase (MLCP), the activity of which is independent of Ca2+, dephosphorylates MLC20and thereby causes relaxation by inactivating the actomyosin ATPase (i.e. , actin-activated, ATPase activity of myosin).21,31
Role of the Phosphatidylinositol Cascade
The majority of vasoconstrictor agonists (e.g. , norepinephrine, angiotensin II [AT-II], ATP, endothelin 1 [ET-1]), acting on receptors coupled to heterotrimeric guanosine-5′-triphosphate (GTP)–binding protein (G protein), activate phospholipase C (PLC) to hydrolyze the membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) into two messengers, i.e. , inositol 1,4,5-triphosphate (IP3) and diacylglycerol (fig. 1).21,32–34IP3stimulates the SR to release Ca2+(i.e. , IP3-induced Ca2+release [IICR]) that in turn activates contractile proteins and initiates contraction. This increase in [Ca2+]c may induce further Ca2+release from SR via the Ca2+-induced Ca2+release (CICR) mechanism.35In response to the increase in [Ca2+]c and phosphoinositide levels, protein kinase C (PKC) migrates from the cytosol to the cell membrane, where it interacts with diacylglycerol to become activated, possibly increasing the sensitivity of contractile myofilaments to Ca2+(i.e. , myofilament Ca2+sensitivity).34,36,37After the initial increase in [Ca2+]c caused by the PLC-mediated Ca2+release, a tonic increase in [Ca2+]c, which is largely dependent on extracellular Ca2+, is normally observed in contractile response to receptor agonists.25–27,34
Role of Intracellular Calcium Stores
In VSMCs, Ca2+is stored intracellularly in the SR, which contains at least two types of Ca2+-release channels, i.e. , those sensitive to IP3(IP3receptor [IP3R] channels) and those sensitive to the plant alkaloid ryanodine and caffeine (ryanodine receptor [RyR] channel).38–40The SR plays a pivotal role in the [Ca2+]c regulation not only as a supply of the activator Ca2+but also as a buffer against VSM activation by Ca2+.21,41The IP3R channels are believed to play a primary physiologic role in Ca2+mobilization,21,34whereas the physiologic roles of RyR channels are not fully understood.38,39,42,43The CICR would occur through the RyR channels and contribute to the amplification of agonist-induced Ca2+signals, cytosolic Ca2+oscillations, and Ca2+waves. In addition, the CICR may play a key role in the superficial buffer barrier mechanism, which is considered essential for cellular Ca2+homeostasis in VSMCs (see review articles40,44for details).
Role of Plasmalemmal Calcium Influxes
Plasmalemmal Ca2+channels are the major routes by which Ca2+enters the VSMCs. Several types exist in VSMCs, including voltage-operated Ca2+channels (VOCCs), receptor-operated Ca2+channels (ROCCs), and store-operated Ca2+channels (SOCCs) (fig. 1).21,34,45The open probability of VOCCs increases as the membrane potential becomes more depolarized. The membrane potential of VSMCs measured in vivo (i.e. , −40 to −55 mV) falls within the range over which VOCCs are activated.17,46,47In addition, membrane depolarization occurs during stimulation with norepinephrine, the sympathetic neurotransmitter.46,48Therefore, the VOCCs are believed to play a crucial role in the physiologic regulation of VSM tone.34,46
The ROCCs, the opening of which is primarily controlled by receptor stimulation, are further subdivided into ligand-gated Ca2+channels and second messenger–operated Ca2+channels (SMOCCs).34,45The ligand-gated Ca2+channels are directly activatable by receptor agonists, and have been suggested to be nonselective cation channels, with some degree of selectivity for divalent cations.21,45On the other hand, the SMOCCs are indirectly activated by diffusible second messengers such as IP3or IP4after receptor activation; however, there is limited evidence for their existence in VSMCs.34,45
Depletion of the SR activates the SOCCs without continued receptor occupation or generation of related second messengers.49,50Unlike the nonselective cation channels, the SOCCs are highly selective for Ca2+over other cations.51,52The cellular mechanisms linking SR depletion to the opening of SOCCs might involve the generation of a diffusible messenger (i.e. , calcium influx factor) or a direct protein–protein interaction between the SR and the plasmalemma.49,50Ca2+entry via SOCCs is considered important not only for the refilling of the depleted SR but also as a source of the activator Ca2+.49
Mechanisms Reducing the Cytosolic Calcium Level
In VSMCs, there exist several mechanisms that reduce the [Ca2+]c, including plasma membrane Ca2+ATPase (PMCA), SR Ca2+ATPase (SERCA), Na+/Ca2+exchanger, and cytosolic Ca2+-binding proteins. Under physiologic conditions, decreases in the [Ca2+]c result mainly from either PMCA-mediated Ca2+extrusion or SERCA-mediated Ca2+uptake into the SR.53However, the Na+–Ca2+exchanger may also play a role in the [Ca2+]c regulation, i.e. , in the superficial buffer barrier mechanism, which has been proposed to maintain the [Ca2+]c at lower levels.40,44The precise roles of cytosolic Ca2+-binding proteins (e.g. , S100 proteins) in [Ca2+]c regulation are unknown.34,36
Cytosolic Calcium Oscillations
Oscillatory changes in [Ca2+]c have been demonstrated in VSMCs during stimulation with receptor agonists; however, its physiologic roles are not clear.54–58A variety of different mechanisms have been proposed to explain cytosolic Ca2+oscillations, including a periodic release of Ca2+from the SR,58,59and oscillatory changes in membrane potential due to an interplay between VOCCs and K+channels.58–61The endothelium, Na+–K+pump, and Cl−channels may also play some important roles in the generation of cytosolic Ca2+oscillations.54–58
Mechanisms Regulating Myofilament Calcium Sensitivity
During receptor stimulation, myofilament Ca2+sensitivity is greatly enhanced through the activation of G proteins21,37; however, its mechanisms are not fully understood (fig. 1). It is currently believed that during receptor stimulation, MLCP activity is inhibited by Rho kinase and/or PKC, leading to the enhanced myofilament Ca2+sensitivity.36,37,62The activated Rho kinase is believed to phosphorylate MLCP and inhibit its catalytic activity,36,37,62whereas the activated PKC presumably phosphorylates and activates CPI-17 (an inhibitory protein of MLCP), thereby inhibiting MLCP activity.37,62The activated Rho kinase is also capable of phosphorylating CPI-17.62In addition, phosphorylation of the thin filament-associated proteins (e.g. , caldesmon, calponin), and resultant reversal of their capability to inhibit the actomyosin ATPase activity, may also contribute to this enhancement.31,36,63PKC, CaMKII, Rho kinase, and mitogen-activated protein kinases (MAPKs) have been reported to have the ability to phosphorylate those proteins, although their physiologic significance remains uncertain.21,31,36,62
In addition to PLC, phospholipase D (PLD) and cytosolic phospholipase A2are activated after stimulation with receptor agonists (e.g. , norepinephrine) in VSMCs.64–69Therefore, after receptor activation, diacylglycerol would be generated not only by PLC-mediated PIP2hydrolysis but also by PLD-mediated hydrolysis of phosphatidylcholine, leading to PKC activation. In addition, arachidonic acid, generated by the phospholipase A2–mediated hydrolysis of arachidonyl phospholipids, has been reported to inhibit MLCP activity, or activate PKC, Rho kinase, and MAPKs, suggesting its involvement in Ca2+sensitization (fig. 1).21,36,37,62,70Furthermore, arachidonic acid may also be released through PLD activation during receptor stimulation.69
Myosin light chain kinase activity is primarily regulated by the Ca2+–CaM complex. Phosphorylation of MLCK at a specific serine residue (i.e., site A ), however, decreases its affinity for the Ca2+–CaM complex.37,71CaMKII, capable of phosphorylating MLCK at site A at higher [Ca2+]c, thus may down-regulate Ca2+signals by decreasing myofilament Ca2+sensitivity.36,37,71
Myosin light chain phosphatase inhibitors, PKC activators, and sphingosylphosphorylcholine (a sphingolipid) have been shown to evoke contraction without an increase in [Ca2+]c in VSMCs.72–75Depending on the stimulants, the Ca2+-independent contraction was associated with or without an increase in MLC20phosphorylation.73,74,76Its underlying mechanisms and physiologic roles remain to be clarified.73–75
Other Important Regulatory Mechanisms
Increases in cyclic guanosine 3′,5′-monophosphate (cGMP) or cyclic adenosine 3′,5′-monophosphate (cAMP) in VSMCs mediates vasodilator response to various mediators such as nitric oxide, carbon monoxide, or natriuretic peptides.77The cytosolic cGMP level is increased by activation of soluble or membrane-bound guanylyl cyclase,77–81or inhibition of various phosphodiesterase subtypes,82,83leading to activation of cGMP-dependent protein kinase (protein kinase G). Activated protein kinase G causes vasodilation via multiple mechanisms including activation of SERCA, PMCA, K+channels, and MLCP, as well as inhibition of VOCCs and IP3R channels.53,77Similarly, increased cytosolic cAMP levels lead to activation of cAMP-dependent protein kinase (protein kinase A), which in turn causes vasodilation via activation of SERCA, PMCA, and K+channels, as well as inhibition of PLC activation, VOCCs, and MLCK activity.23,53,84–86Protein kinase G can be cross-activated by cAMP,23,77whereas protein kinase A can be cross-activated by cGMP77,85; either cross-activation would lead to vasodilation.
Tyrosine Kinases, Mitogen-activated Protein Kinases.
Tyrosine kinases, present abundantly in VSMCs, have been suggested to influence both Ca2+mobilization and myofilament Ca2+sensitivity in VSMCs.87,88Specifically, tyrosine kinase–catalyzed protein tyrosine phosphorylation has been reported to regulate the activity of VOCCs, SOCCs, K+channels, PLC, PLD, and MAPKs in VSMCs.69,87,89PLD hydrolyzes phosphatidylcholine to generate diacylglycerol, leading to the activation of PKC, whereas MAPKs phosphorylate caldesmon to attenuate its inhibitory action on actomyosin ATPase activity.36Therefore, tyrosine phosphorylation might be involved in the regulation of myofilament Ca2+sensitivity through the effects on PLD and/or MAPKs.87The precise roles of tyrosine kinases and MAPKs in the regulation of vascular tone are not fully understood.
Chloride (Cl−) channel, abundantly distributed in the VSMC membrane, would play an important role in the regulation of [Ca2+]c.90There exist at least two types of Cl−channels in VSMCs, i.e. , Ca2+-dependent Cl−(ClCa) and volume-regulated Cl−(Clv) channels,90and their activation leads to the activation of VOCCs and membrane depolarization (ECl=−26 mV). The ClCachannels, activatable by an increase in [Ca2+]c,90,91may contribute to the contractile response to receptor agonists such as norepinephrine in some vascular beds (e.g. , portal vein,91small mesenteric arteries27), whereas the Clv channels, activatable by vascular distension, may play a role in maintaining tissue integrity against mechanical stretch.90
In VSMCs, potassium (K+) channels play a fundamental role in maintaining the membrane potential, a major determinant of vascular tone.92,93The blockade of K+channels results in membrane depolarization, leading to increased Ca2+influx through VOCCs and possibly enhanced myofilament Ca2+sensitivity,94,95whereas their activation induces membrane hyperpolarization, leading to reduced Ca2+influx through VOCCs,92,93inhibited phosphoinositide metabolism,96,97and suppressed myofilament Ca2+sensitivity.98,99The activity of K+channels can be altered by various factors, including intracellular Ca2+, ATP, pH, extracellular K+, G proteins, cyclic nucleotides, and protein kinases (e.g. , protein kinase A, protein kinase G).86,92The activation of K+channels has been shown to mediate vasodilator responses to various endogenous mediators such as nitric oxide, prostacyclin (PGI2), endothelium-derived hyperpolarizing factor (EDHF), calcitonin gene–related peptide, or β-adrenergic agonists, as well as vasodilator responses to changes in metabolic activity (i.e. , hypoxia, reactive hyperemia, acidosis, or shock),92,93,100–104whereas their inhibition may underlie contractile responses to receptor agonists (e.g. , phenylephrine, arginine vasopressin [AVP], AT-II, thromboxane A2).86,92,105At least five distinct types of K+channels exist in VSMCs, including voltage-gated K+(KV), Ca2+-activated K+(KCa), ATP-sensitive K+(KATP), inward rectifying K+(Kir), and cGMP-gated K+channels (see review articles92,93,105,106for details on each type of K+channels).92
In response to various neurohumoral, physical, and chemical stimuli, vascular endothelial cells produce and release a variety of vasoactive factors, including both relaxing factors (e.g. , nitric oxide, EDHF, PGI2, adenosine) and contracting factors (e.g. , prostanoids, ET-1).107,108Physical force exerted by blood flow (i.e. , shear stress, pulsatile stretch) serves as a stimulus for the endothelial release of nitric oxide, PGI2, and EDHF.107,109–111Therefore, continual, basal release of those relaxing factors would keep the vasculature in a dilated state, serving as a primary determinant of resting vascular tone.110,112In addition, they mediate vasodilator responses to various endogenous substances.109,112,113
Endothelin 1 is the most potent endogenous vasoconstrictor identified to date.114–116Its synthesis is enhanced by numerous physicochemical factors such as hypoxia, ischemia, shear stress, or inflammatory cytokines and inhibited by atrial natriuretic peptide, nitric oxide, nitroglycerin, PGE2, PGI2, and heparin.114–116Besides the vasoconstrictor action, ET-1 exerts vasodilator action by stimulating the release of prostanoids, nitric oxide, and/or EDHF from endothelial cells.114,115,117The endothelium-derived ET-1 would serve as a local regulator of vascular tone by acting on underlying VSMCs to increase vascular tone in a paracrine manner, as well as on endothelial cells to release those vasodilator factors in an autocrine manner.115In addition, ET-1 would play some key roles in various disease processes including hypertension, myocardial ischemia, heart failure, cerebral vasospasm, endotoxin shock, and liver failure.114,115
Anesthetic Actions on Vascular Reactivity or Tone
In previous in vitro or in situ studies, most general anesthetics affected vasoconstrictor and vasodilator responses to a wide range of both physiologic and pharmacologic stimuli, as well as basal vascular tone (tables 1 and 2). This section reviews such previously observed direct vascular actions of general anesthetics and discusses their in vivo relevance.
In the following sections, the effective concentrations of anesthetics were not necessarily described if they were clinically relevant. Regarding volatile anesthetics, concentrations up to 1.5 minimum alveolar concentration were considered clinically relevant.
Anesthetic Actions on Vasoconstrictor Response
Depending on blood vessels, concentrations, or experimental conditions (e.g. , the presence of endothelium), the contractile responses were enhanced, inhibited, or unaffected by the anesthetics. However, some tendencies exist in the observed anesthetic actions on vasoconstrictor responses in isolated vessels.
Most volatile and intravenous anesthetics inhibited the contractile response to KCl or norepinephrine in either the presence or the absence of endothelium (KCl response : halothane,118–130enflurane,131isoflurane,27,132–134sevoflurane,25,135barbiturates,136,137ketamine,26,138–142propofol,143–150etomidate,151diazepam,152morphine153,154; norepinephrine response : halothane,123,125,128,155–157enflurane,157–160isoflurane,27,123,125,128,132,133,157,160–162sevoflurane,25,135,160,163,164ketamine,26,138,139,141,165propofol,144,145,147,166–168etomidate,151diazepam,152midazolam,169,170morphine,153,154fentanyl171) (table 1). However, all previously investigated intravenous anesthetics inhibited these responses only at concentrations higher than the clinically relevant free anesthetic concentrations (KCl response : barbiturates,136,137≥ 100 μm; ketamine,26,138–142≥ 100 μm; propofol,143–150≥ 10 μm; etomidate,151≥ 100 μm; diazepam,152≥ 30 μm; morphine,153,154≥ 30 μm; norepinephrine response : ketamine,26,138,139,141,165≥ 10 μm; propofol,144,145,147,166–168≥ 1 μm; etomidate,151≥ 100 μm; diazepam,152≥ 100 μm; midazolam,169,170≥ 1 μm; morphine,153,154≥ 10 μm; fentanyl,171≥ 1 μm).
Previous in vitro studies123,133–135,156,172–175have yielded conflicting results regarding endothelium dependence of the anesthetic actions on vasoconstrictor responses, possibly reflecting species or regional differences, or differences in experimental conditions (e.g. , test stimulants). However, it has been clearly shown that in some vessels, halothane, enflurane, isoflurane, and sevoflurane all enhanced contractile responses to norepinephrine or phenylephrine in an endothelium-dependent manner,133,135,174,176although the underlying mechanisms have not yet been clarified. In rat small mesenteric arteries, the enhancement by halothane, isoflurane, or sevoflurane was still observed after inhibition of the nitric oxide, EDHF, cyclooxygenase, and lipoxygenase pathways or after blockade of ET-1, AT-II, or serotonin receptors.133,135,176Therefore, nitric oxide, EDHF, cyclooxygenase products, lipoxygenase products, ET-1, AT-II, and serotonin all would not be involved in the enhancement by those anesthetics.
In isolated rat small mesenteric arteries exposed to halothane, isoflurane, or sevoflurane, the contractile response to norepinephrine was enhanced in the presence of endothelium, but inhibited or unaffected in its absence.133,135,176However, regardless of the presence or absence of endothelium, the norepinephrine response was inhibited after the removal of either anesthetic from the extracellular space.133,135,176Therefore, those anesthetics presumably have opposing actions on the contractile response to norepinephrine, i.e. , endothelium-dependent enhancing and endothelium-independent inhibitory actions. Namely, the former probably predominates or counteracts the latter during exposure to those anesthetics, whereas only the latter seems to persist after their removal. The experiments using fura-2, a fluorescent Ca2+indicator, further suggested that such prolonged inhibitory action of isoflurane on the norepinephrine response is associated with a reduction in the intracellular Ca2+concentration ([Ca2+]i) in VSMCs, possibly due to inhibition of ClCachannels.27
In Vivo Relevance.
Plasmalemmal depolarization of VSMCs by KCl activates the VOCCs, which are active in vivo ,46presumably contributing to the maintenance of basal vascular tone as the primary source of Ca2+.91Norepinephrine plays a central role in the sympathetic maintenance of vascular tone. Thus, the inhibitory action of volatile anesthetics on contractile response to KCl or norepinephrine is consistent with their well-recognized vasodilatory properties. However, it was not consistently observed in systemic resistance arteries, particularly in the presence of endothelium.133,135,160,161,163,176Therefore, it seems unclear whether the direct action of volatile anesthetics on VSMCs contributes to systemic hypotension during their administration. However, prolonged systemic hypotension observed after anesthesia with halothane, isoflurane, or sevoflurane10,47might be due, in part, to their direct inhibitory action on the norepinephrine response, i.e. , the persistent hyporesponsiveness to norepinephrine after their removal from the extracellular space.133,135,176In addition, the enhanced contractile response to norepinephrine and the inhibited endothelial nitric oxide or EDHF-mediated vasodilator response during exposure to halothane observed in isolated mesenteric resistance arteries176might contribute to a decrease in intestinal blood flow associated with an increase in intestinal vascular resistance during anesthesia with halothane.177
The in vivo relevance of the above-described direct vasodilator actions of intravenous anesthetics is unclear because of their effective concentrations.
Anesthetic Effects on Vasodilator Response
In previous in vitro or in situ studies, all currently used volatile anesthetics and some intravenous anesthetics inhibited various vasodilator responses, such as those mediated by nitric oxide, EDHF, carbon monoxide, β-adrenoceptor agonists, and KATPchannels (table 2).
In most previous studies, halothane,156,178–183enflurane,160,178isoflurane,133,160,178,184desflurane,177sevoflurane,135,160,163,180,185,186thiopental,187ketamine,188propofol,189–191and etomidate188,192–194inhibited endothelium-dependent, presumed nitric oxide–mediated vasodilator responses (table 2). However, most of those anesthetics did not inhibit the endothelium-independent vasodilator response to nitrovasodilators (table 2) and the associated increases in cGMP level in VSMCs.156,160,163,177,178,180,183,185–187,192,194–196In addition, most of them inhibited the endothelial nitric oxide–mediated vasodilator response to both receptor agonists and A23187 (i.e. , Ca2+ionophore) (table 2).178,185,187,194Therefore, those anesthetics presumably inhibit the nitric oxide–guanylyl cyclase signaling distal to receptor activation in the endothelial cell and proximal to the nitric oxide activation of guanylyl cyclase.78,178,196This idea is consistent with recent observations that they inhibited Ca2+mobilization induced by nitric oxide–releasing agonists in endothelial cells.188,197–202However, in some previous studies using isolated aorta,179,203halothane inhibited the endothelium-independent vasodilator response to nitric oxide and the associated increases in cGMP level in VSMCs. Therefore, halothane may interfere with the nitric oxide activation of guanylyl cyclase, depending on the experimental condition.
In experiments with isolated rat aorta,181,204halothane inhibited the endothelial nitric oxide–mediated vasodilator response to acetylcholine181,204and the associated increases in cGMP level in VSMCs,204but it did not affect the endothelial nitric oxide–mediated vasodilator response to isoproterenol and the associated increases in cGMP level in VSMCs.181Therefore, halothane may inhibit some mechanisms specifically involved in the cholinergic receptor–mediated nitric oxide production but not in the β-adrenoceptor–mediated nitric oxide production in endothelial cells.181
Besides the endothelial nitric oxide–mediated vasodilator response, some general anesthetics (halothane,177,182enflurane,160,177isoflurane,160,177,192desflurane,177sevoflurane,160,177thiopental,192,193ketamine,188etomidate188,192,193) inhibited the EDHF-mediated vasodilator response. However, it has not yet been shown that they inhibit the EDHF-mediated hyperpolarization of the VSMC membrane. Conversely, in rabbit mesenteric arteries, propofol (10 μm) did not influence acetylcholine-induced, EDHF-mediated hyperpolarization, but inhibited acetylcholine-induced, prostanoid-mediated hyperpolarization.190
Again, the intravenous anesthetics exerted their inhibitory actions on either the nitric oxide–mediated or EDHF-mediated vasodilator response only at the supraclinical concentrations (i.e. , thiopental,187,192,193≥ 30 μm; ketamine,188≥ 100 μm; propofol,189≥ 1 μm; etomidate,188,192,193≥ 30 μm).
Halothane and isoflurane also inhibited vasodilator responses to β-adrenoceptor agonists and the associated increase in cAMP level in VSMCs.205,206However, both anesthetics had little effect on the decreases in force and [Ca2+]i induced by forskolin (i.e. , adenylyl cyclase activator) or dibutyryl cAMP (i.e. , membrane-permeable cAMP analog). They also did not affect the forskolin-induced increases in cAMP levels. Furthermore, they did not affect β-adrenoceptor binding characteristics and affinity for agonists. Therefore, halothane and isoflurane presumably inhibit the β-adrenoceptor–mediated vasodilator response distal to agonist–receptor binding and proximal to adenylyl cyclase activation.206
The KATPchannels, activatable by depletion of intracellular ATP, mediate vasodilator responses to changes in metabolic activity (i.e. , hypoxia) and those to various endogenous mediators such as adenosine, nitric oxide, or β agonists.92,100–104In coronary circulation, the KATPchannels are active under resting conditions, contributing to the maintenance of resting vascular tone.92Previous in vivo studies using beating hearts have proposed that volatile anesthetics (i.e. , halothane,207,208isoflurane,208,209enflurane,208desflurane,210sevoflurane210) cause coronary vasodilation through activation of KATPchannels. In addition, previous in situ studies using superperfused mesenteric vascular beds211,212have proposed that isoflurane hyperpolarizes VSMCs by activating KATPchannels. By contrast, in previous in vivo and in vitro studies using pulmonary vessels, halothane,213,214isoflurane,215enflurane,213and desflurane,216but not sevoflurane,216attenuated vasodilator responses to lemakalim, a KATPchannel activator. Ketamine (≥ 10 μm) and etomidate (≥ 50 μm) also inhibited the vasodilator response to lemakalim in isolated endothelium-denuded pulmonary arteries.217However, to my knowledge, there is currently no electrophysiologic evidence indicating that those anesthetics directly influence KATPchannel activity in VSMCs.
In Vivo Relevance.
Vasodilator responses to nitric oxide, EDHF, carbon monoxide, β-adrenoceptor agonists, and KATPchannels are believed to play important roles in the physiologic regulation of vascular tone. However, in vivo relevance of the above-described inhibitory actions of general anesthetics on those vasodilator responses remains to be elucidated, as discussed later.
Anesthetic Effects on Basal Vascular Tone
Some general anesthetics increased basal vascular tone in vitro (halothane,125,130,218–220enflurane,160,218,220thiopental,221,222propofol,223,224morphine225). In isolated small mesenteric arteries, halothane and enflurane caused transient contractions, which were eliminated by ryanodine, which depletes the SR.130,219,220Therefore, in those arteries, both the anesthetics presumably stimulate Ca2+release from the ryanodine-sensitive SR and cause contraction.130,219,220However, in conduit arteries, the volatile anesthetic–induced contractions were only partially inhibited by ryanodine.125,218Some differences may exist in vascular responsiveness to volatile anesthetics between conduit and resistance arteries.
The thiopental (≥ 30 μm)221,222– and morphine (≥ 30 μm)225–induced contractions observed in isolated rat aorta and basilar artery, respectively, seemed to be due to increases in [Ca2+]c resulting from either Ca2+release from the SR221or plasmalemmal Ca2+influx.222,225
In isolated canine epicardial coronary arteries, propofol increased basal vascular tone at clinical concentrations (i.e. , < 1 μm) but decreased at higher concentrations (≥ 100 μm).224The propofol-induced sustained increase in force was independent of endothelium and inhibited by the removal of extracellular Ca2+or verapamil, suggesting that propofol activates VOCCs and thereby increases vascular tone.224By contrast, in human omental arteries and veins, propofol caused sustained increases in basal vascular tone at a high concentration (1 mm).223Species or regional differences may exist in vascular sensitivity to propofol.
In Vivo Relevance.
The vascular tone in vivo is determined by the net balance between various vasoconstrictor and vasodilator stimuli. Even in the resting condition, the sympathetic vasoconstrictor system is continually active, maintaining a partial state of contraction in the blood vessels. Therefore, in vivo relevance of the above-described anesthetic actions on basal vascular tone observed in vitro is unclear. However, they would suggest that besides the vasodilatory actions, some general anesthetics possess vasoconstrictor actions, possibly contributing to the anesthetic-induced changes in vascular tone in vivo .
Mechanisms behind Direct Actions of General Anesthetics on Vascular Smooth Muscle: Overview
To investigate the mechanisms behind direct vasodilator action of general anesthetics, previous studies25–27,125,140,145,170,220,226–232evaluated their actions on receptor agonist- or KCl-induced increases in force and [Ca2+]i in VSMCs loaded with Ca2+-sensitive fluorescent dyes (e.g. , fura-2). Previous studies118,127,159,162,165,230,233,234also evaluated the anesthetic actions on Ca2+-induced contraction in VSMCs permeabilized with chemical detergents (e.g. , saponin, β-escin). To summarize the results of those previous studies, the vasodilator action of volatile anesthetics including halothane,118,125,127,226,228–230,232–234enflurane,127,159isoflurane,27,162,228,234,235and sevoflurane25is presumably due to both a reduction in [Ca2+]c and inhibition of myofilament Ca2+sensitivity, whereas that of ketamine26,140or midazolam170is due largely to a reduction in [Ca2+]c. Interestingly, in isolated pulmonary arteries depolarized with KCl, propofol reduced [Ca2+]i while increasing myofilament Ca2+sensitivity, leading to a depressed contractile response to KCl.143Less information is available regarding the mechanisms behind previously observed direct vasodilator actions of barbiturates, etomidate, diazepam, morphine, and fentanyl (table 1).
Recent studies utilizing fura-2 fluorometry25,27,129have suggested that the underlying mechanisms may depend on the concentrations of volatile anesthetics. Namely, in isolated small mesenteric arteries, the depressed contractile response to KCl by lower concentrations of halothane (≤ 2.5%, approximately 0.7 mm), isoflurane (3%, approximately 0.6 mm), and sevoflurane (3%, approximately 0.4 mm) seemed to be due to inhibition of the myofilament Ca2+sensitivity, whereas that by higher concentrations of halothane (4.5%, approximately 1.4 mm), isoflurane (5%, approximately 1.0 mm), and sevoflurane (5%, approximately 0.7 mm) seem to arise by both a reduction in [Ca2+]c and inhibition of myofilament Ca2+sensitivity.25,27,129
The following two sections review the previously observed general anesthetic actions on Ca2+mobilization and myofilament Ca2+sensitivity in VSMCs and discuss the mechanisms underlying the direct inhibitory or excitatory actions of general anesthetics on VSMCs.
Mechanisms behind General Anesthetic–induced Changes in Calcium Mobilization
Previous experiments using VSMCs loaded with fluorescent Ca2+-indicators or 45Ca2+have provided convincing evidence to indicate that various general anesthetics reduce the [Ca2+]c by inhibiting Ca2+release from the SR and/or plasmalemmal Ca2+influx (halothane,125,132,220,226,228,229,236isoflurane,27,125,132,227,228,232,236sevoflurane,25,220pentobarbital,137ketamine,26,140,237propofol,143,145,231,238,239midazolam170) (table 3).
Anesthetic Effects on the Phosphatidylinositol Cascade
Some general anesthetics have been suggested to reduce the [Ca2+]c in VSMCs by inhibiting the phosphatidylinositol cascade.139,140,145,226,227,238–240
Halothane,226isoflurane,227and propofol (≥10 μm)238–240inhibited inositol phosphate (IP) production in cultured rat aortic VSMCs (A7r5, A10) stimulated with AVP or ET-1. Ketamine (≥ 100 μm) and propofol (≥ 30 μm) also inhibited the IP production in arterial VSM tissues stimulated with phenylephrine or norepinephrine.139,140,145However, in rat aortic VSM tissue, both halothane and isoflurane did not affect the norepinephrine-induced, presumed IP3-induced Ca2+release from SR.132
During receptor stimulation, the above anesthetics may interfere with PLC-mediated PIP2hydrolysis (i.e. , synthesis of IP3and diacylglycerol), thereby inhibiting the IICR from SR and possibly the receptor-operated Ca2+influx through SMOCCs. However, the inability of halothane and isoflurane to inhibit norepinephrine-induced Ca2+release132suggests that they may inhibit some mechanisms specifically involved in AVP-induced PIP2breakdown but not in norepinephrine-induced PIP2breakdown. Alternatively, the difference may reflect the difference in vascular responsiveness to those anesthetics between cultured VSMCs and intact VSM tissues, or the difference in the experimental condition.
Anesthetic Effects on the Intracellular Calcium Stores (SR)
General anesthetics have been proposed to influence Ca2+mobilization by impairing functional integrity of the SR in VSMCs.118,132,159,162,219,220,230,236In most previous studies evaluating the anesthetic effects on Ca2+uptake into the SR in VSMCs, the amount of Ca2+in the SR was estimated by vascular responses to caffeine after removal of extracellular Ca2+.
The relative importance of intracellular and extracellular Ca2+pools in the excitation–contraction coupling in VSMCs seems to vary in different vascular beds.91For example, Ca2+release from the SR seems to be of considerably less importance than plasmalemmal Ca2+influx for development of contractile force in small mesenteric arteries,25–27,91whereas it presumably plays a central role in contractile responses to agonists in portal veins.91Therefore, pharmacologic significance of the previously observed general anesthetic actions on SR would depend on the type of blood vessels.
In previous contraction and Ca2+measurement experiments with isolated conduit or resistance arteries, halothane118,127,218–220,230,236and enflurane127,159,218,220caused transient increases in [Ca2+]i and/or force in the absence of extracellular Ca2+in VSMCs with the SR sensitive to ryanodine and/or caffeine. Similar results were obtained with isoflurane, although the effect was smaller than halothane or enflurane, and observed only at low temperatures (22°–23°C)127,162but not at a higher temperature (35°C).220In mesenteric arterial VSMCs,219,220if the Ca2+-releasing action of halothane or enflurane was blocked by procaine, either anesthetic stimulated Ca2+uptake by the ryanodine-sensitive SR; their overall effects, when applied during Ca2+loading, were to reduce the amount of Ca2+in the SR.219,220In membrane-permeabilized VSMCs,219the Ca2+-releasing action of halothane was not blocked by heparin, a specific inhibitor of the IICR.241
Previous studies also evaluated the effects of various anesthetics on Ca2+-releasing mechanisms (e.g. , Ca2+release induced by caffeine, receptor agonists, or IP3) in VSMCs. In both conduit or resistance arteries, halothane, enflurane, and isoflurane all enhanced the increase in [Ca2+]i and/or the force induced by caffeine.118,132,159,162,220,230In addition, in previous contraction studies with conduit arteries, these anesthetics seemed to facilitate the ryanodine depletion of SR.118,159,162Interestingly, sevoflurane attenuated the caffeine-induced Ca2+release from the SR, whereas it enhanced the norepinephrine-induced Ca2+release from the SR.220
The above results118,127,159,218–220,230,236suggest that both halothane and enflurane stimulate Ca2+release from the ryanodine-sensitive SR in VSMCs. In addition, their ability to stimulate Ca2+uptake into the SR after blockade of their Ca2+releasing action219,220indicates that both the anesthetics actually have opposing actions on the amount of Ca2+in the SR, i.e. , the Ca2+-releasing action and a stimulating action on Ca2+uptake. Their overall effects were to reduce the amount of Ca2+in the SR,219,220which would influence vascular reactivity by decreasing Ca2+availability and enhancing the Ca2+-buffering capacity.
The inability of isoflurane or sevoflurane to stimulate Ca2+release from the SR or Ca2+uptake by the SR at 35°C220suggests that they would not alter the amount of Ca2+stored in the SR under physiologic conditions. The ability of isoflurane to stimulate Ca2+release from the ryanodine-sensitive SR at 22°–23°C127,162could be due to the increased aqueous concentrations or increased vascular sensitivity to isoflurane at the low temperatures.
The ability of halothane, enflurane, and isoflurane to stimulate Ca2+release from the SR in membrane-permeabilized VSMCs18,127,159,162,219,230suggests that their Ca2+releasing actions are, at least in part, independent of the intact plasma membrane. Although the Ca2+releasing actions of halothane and enflurane were blocked by procaine,219,220the nonspecific nature of its blocking action did not allow us to definitely characterize their Ca2+releasing action. However, the lack of effect of heparin on halothane-induced Ca2+release219indicates that the halothane-induced Ca2+release can occur independently of IP3production.
Because ryanodine depletes the SR by binding to the RYR/Ca2+-release channels in an open state and then locking them open,38the ability of halothane, enflurane, and isoflurane to facilitate the ryanodine depletion of SR118,159,162suggests that those anesthetics stimulate the opening of RYR/Ca2+-release channels (i.e. , CICR). Therefore, their ability to enhance caffeine-induced Ca2+release (i.e. , Ca2+release from the ryanodine-sensitive SR)118,132,159,162,220,230could be a result of the activation of the RYR/Ca2+-release channels. By contrast, the ability of sevoflurane to attenuate caffeine-induced Ca2+release and to enhance norepinephrine-induced Ca2+release suggests that it inhibits the CICR, whereas it enhances the IICR.220There is no definitive evidence to date that those anesthetics directly act on the SR and alter its CICR or IICR channel activity in VSMCs. Further studies using isolated SR vesicles from VSMCs or membrane-permeabilized VSMCs would be necessary to clarify the underlying mechanisms.
In conclusion, halogenated volatile anesthetics would alter the [Ca2+]c and hence vascular reactivity through direct actions on the SR in VSMCs. Indeed, the depletion of SR altered their actions on vascular reactivity (e.g. , contractile response to norepinephrine).132,176
Ketamine (≥ 300 μm) inhibited α-adrenergic agonist–induced synthesis of IP or IP3in femoral and mesenteric arterial VSMCs.139,140However, ketamine (1 mm) did not affect the IICR in membrane-permeabilized VSMCs.165In addition, in isolated mesenteric arteries, ketamine (1 mm) did not affect caffeine-induced Ca2+release.26,165
In cultured aortic VSMCs, propofol (≥ 10 μm) also inhibited IP production during stimulation with receptor agonists (i.e. , ET-1, AVP).239,240In support of this finding, in VSMCs of isolated mesenteric arteries, propofol (≥ 30 μm) inhibited norepinephrine-induced Ca2+release from the SR; however, it did not affect caffeine-induced Ca2+release even at 100 μm.145Similarly, in VSMCs of isolated mesenteric arteries, midazolam (≥ 1 μm) inhibited norepinephrine-induced, but not caffeine-induced, Ca2+release from the SR.170
The ability of ketamine to inhibit the synthesis of IP or IP3139,140suggests that ketamine attenuates the IICR by inhibiting PLC-mediated PIP2breakdown. However, its inability to influence IICR in membrane-permeabilized VSMCs165suggests that ketamine does not directly influence IICR. In addition, its inability to influence the caffeine-induced Ca2+release26,165indicates the lack of effects of ketamine on the CICR.
The ability of propofol to inhibit IP production239,240and the norepinephrine-induced Ca2+release145suggests that propofol also attenuates IICR by inhibiting PLC-mediated IP3production. However, its inability to influence the caffeine-induced Ca2+release suggests that propofol does not affect the CICR.145The previous result on midazolam170suggests that midazolam also inhibits IICR but not CICR.
Previous studies have failed to demonstrate that any of the intravenous anesthetics influence functional integrity of the SR at clinically relevant free concentrations, and in vivo relevance of the above-described intravenous anesthetic actions on SR is currently unclear.
Anesthetic Effects on the Plasmalemmal Calcium Influx through VOCCs
Both volatile and intravenous anesthetics have been suggested to inhibit plasmalemmal Ca2+influx through VOCCs, thereby reducing [Ca2+]c in VSMCs.
In earlier studies118,119,121,124,242,243using isolated aorta or coronary arteries, halothane, enflurane, isoflurane, and sevoflurane all inhibited the contractile response to KCl, suggesting that they inhibit plasmalemmal Ca2+influx through VOCCs. In direct proof of this proposal, by using the whole cell mode of single cell patch clamp technique, Buljubasic et al. 244,245showed that both halothane and isoflurane inhibited the L-type voltage-dependent Ca2+currents in VSMCs dispersed from cerebral or coronary arteries. This finding was subsequently confirmed in portal venous VSMCs.246,247In addition, in fura-2–loaded VSMCs of aorta and small mesenteric arteries, halothane, isoflurane, and sevoflurane inhibited KCl-induced increases in [Ca2+]i, mimicking the effects of VOCC blockers.25,27,125,132
The above results suggest that most volatile anesthetics inhibit VOCC activity. However, the underlying mechanisms are currently unknown. Because the anesthetic effects have not yet been examined on single-channel activity, it is unclear whether they directly inhibit VOCC activity or indirectly via diffusible second messengers. Both halothane and isoflurane were previously reported to increase basal cytosolic cAMP or cGMP levels in aortic VSMCs.120,248Therefore, they may indirectly inhibit VOCC activity by increasing cytosolic cAMP and/or cGMP levels. However, sevoflurane did not increase basal cytosolic cGMP levels in aortic VSMCs.180
Pentobarbital (≥ 100 μm),137ketamine (≥ 100 μm),138,141,142,165,249,250propofol (≥ 10 μm),147,148diazepam (≥ 30 μm),152and midazolam (≥ 10 μm)169inhibited the contractile response to KCl. In patch clamp experiments using VSMCs dispersed from the portal vein,249,251ketamine (≥ 10 μm) inhibited the whole cell L-type voltage-dependent Ca2+currents. In addition, in experiments with 45Ca2+or Ca2+-sensitive fluorescent dyes, ketamine (≥ 100 μm),26,140propofol (≥ 10 μm),145,238and midazolam (≥ 10 μm)170inhibited plasmalemmal Ca2+influx induced by KCl26,145,170or that sensitive to VOCC blockers in VSMCs.238
These results suggest that some intravenous anesthetics may also inhibit VOCC activity. However, to my knowledge, to date, electrophysiologic studies have not yet been conducted on the above intravenous anesthetics except ketamine to directly prove their inhibitory action on VOCC activity in VSMCs.
Anesthetic Effects on the Plasmalemmal Calcium Influx through ROCCs
General anesthetics have also been suggested to influence plasmalemmal Ca2+influx through ROCCs, thereby altering the [Ca2+]c in VSMCs.
In cultured aortic VSMCs loaded with indo-1 or fura-2, both halothane and isoflurane inhibited plasmalemmal Ca2+influx induced by receptor agonists (i.e. , AVP, platelet-derived growth factor, AT-II).226,227,229,232Isoflurane and sevoflurane also inhibited norepinephrine-induced plasmalemmal Ca2+influx in fura-2–loaded VSMCs of isolated mesenteric arteries.25,27By contrast, in contraction and 45Ca2+studies using isolated aorta,252,253both halothane and isoflurane stimulated plasmalemmal Ca2+influx sensitive to SKF-96365, a putative inhibitor of ROCCs.254,255
The ability of halothane, isoflurane, and sevoflurane to inhibit receptor agonist–induced Ca2+influx suggests that they inhibit ROCC activity. However, plasmalemmal Ca2+influx induced by some receptor agonists (e.g. , norepinephrine) has been shown to be eliminated by VOCC blockers,25,27,170apparently suggesting that it is mediated exclusively by VOCCs. Therefore, the above results might have simply reflected the anesthetic effects on VOCCs. However, activation of the ligand-gated nonselective cation channels, a type of ROCCs, and resultant membrane depolarization may precede the activation of VOCCs during receptor stimulation.91Thus, there remains a possibility that the volatile anesthetics inhibit receptor agonist–induced Ca2+influx through VOCCs by inhibiting activity of nonselective cation channels.
On the basis of sensitivity to SKF-96365, both halothane and isoflurane have been proposed to stimulate the Ca2+influx through ROCCs.252,253However, recent evidence indicates that SKF-96365 does not serve as a selective inhibitor of the ROCCs in VSMCs.27
At this time, there is no electrophysiologic evidence to indicate that volatile anesthetics alter the ROCC activity in VSMCs. This topic should be the subject of future studies.
In earlier contraction studies, thiamylal (1 mm),136pentobarbital (≥ 300 μm),136ketamine (≥ 30 μm),138,165propofol (≥ 1 μm),147and midazolam (≥ 30 μm)169inhibited extracellular Ca2+-dependent responses to receptor agonists (i.e. , prostaglandin F2α, norepinephrine, histamine). In recent Ca2+measurement studies, pentobarbital (≥ 100 μm),137ketamine (≥ 300 μm),26propofol (≥ 10 μm),145,231and midazolam (≥ 30 μm)170inhibited the agonist (e.g. , serotonin, norepinephrine, AT-II)–induced plasmalemmal Ca2+influx.
These results suggest that intravenous anesthetics may also inhibit ROCC activity, thereby reducing the [Ca2+]c. However, again, to date, electrophysiologic experiments have not yet been conducted definitively to prove these hypotheses.
Anesthetic Effects on the Plasmalemmal Calcium Influx through SOCCs
Less information is available regarding general anesthetic actions on plasmalemmal Ca2+influx through SOCCs. However, some general anesthetics have been suggested to influence activity of SOCCs, thereby altering the [Ca2+]c in VSMCs.
In cultured aortic VSMCs loaded with indo-1 or fura-2, isoflurane (≥ 1.5%) inhibited plasmalemmal Ca2+influx activated after depletion of the SR by thapsigargin.227,232In a previous study with cultured aortic VSMCs loaded with fura-2, propofol (56 μm) did not affect the thapsigargin-induced Ca2+influx.238However, in recent studies using cultured aortic or pulmonary arterial VSMCs loaded with fura-2, propofol (aorta, ≥ 56 μm; pulmonary artery, ≥ 1 μm) inhibited the thapsigargin-induced Ca2+influx.89,231By contrast, thiopental (approximately 0.42 mm) was recently proposed to deplete the SR and thereby activate SOCCs, increasing vascular tone in isolated aorta.222
These results suggest that the SOCCs are potential targets for general anesthetic agents. However, there is currently no electrophysiologic evidence to support this hypothesis, and this topic also should be the subject to future investigations.
Anesthetic Effects on Mechanisms Reducing the Cytosolic Calcium Level
Stimulation and inhibition of the [Ca2+]c-reducing mechanisms (i.e. , SERCA-mediated Ca2+uptake into the SR, PMCA-mediated plasmalemmal Ca2+extrusion, Na+–Ca2+exchanger) in VSMCs would lead to vasodilation and vasoconstriction, respectively. Less information is available regarding general anesthetic actions on those mechanisms.
In both conduit and resistance arteries, halothane and enflurane, applied during Ca2+loading, decreased the amount of Ca2+loaded in the SR (as judged by the response to caffeine).118,159,162,219,220,256However, as mentioned above, these anesthetics, applied during Ca2+loading after blockade of their Ca2+-releasing action, conversely increased the amount of Ca2+in SR.219,220By contrast, in previous experiments performed at 35°–37°C,220,256neither isoflurane nor sevoflurane, applied during Ca2+loading, affected the amount of Ca2+in SR.
In our recent studies with isolated small mesenteric arteries,176halothane prolonged vasorelaxation following cessation of stimulation with norepinephrine, and the prolongation was eliminated by treatment with vanadate, a putative inhibitor of the PMCA.
The ability of halothane and enflurane to inhibit Ca2+loading into the SR118,159,162,219,220,256apparently suggests that these anesthetics inhibit SERCA activity. However, their ability to conversely increase the Ca2+loading after blockade of their Ca2+-releasing action219,220indicates that both halothane and enflurane actually stimulate SERCA activity, and that the inhibition of Ca2+uptake by those anesthetics (observed before blockade of their Ca2+-releasing action) was due to their Ca2+-releasing action118,127,159,218–220,230,236but not due to inhibition of SERCA activity.219,220The stimulated SERCA-mediated Ca2+uptake into the SR may underlie vascular reactivity during exposure to halothane176or enflurane. By contrast, the previous results220,256suggest that neither isoflurane nor sevoflurane affects the SERCA activity at physiologic temperatures. To my knowledge, no direct evidence is currently available to indicate that volatile anesthetics modulate SERCA activity, and further studies using isolated SR vesicles from VSMCs would be essential to clarify this issue.
Volatile anesthetics inhibit PMCA activity in erythrocytes or neuronal cells,257,258and the Na+–Ca2+exchanger in cardiac cells.259However, little information is available regarding volatile anesthetic action on those plasmalemmal Ca2+extrusion mechanisms in VSMCs. On the basis of the sensitivity to vanadate, we recently proposed that halothane inhibits PMCA activity in VSMCs.176However, because of vanadate’s nonspecific nature as a pharmacologic tool,260–262further investigations are necessary to prove our proposal definitively.
In previous contraction and 45Ca2+studies using isolated mesenteric arteries165or cultured aortic VSMCs,237ketamine (1 mm) did not affect Ca2+uptake into the SR under membrane-permeabilized conditions. In addition, in a 45Ca2+study with membrane-intact cultured aortic VSMCs,237ketamine did not affect plasmalemmal Ca2+extrusion. However, in a contraction study with isolated pulmonary arteries, the vasodilator action of ketamine was attenuated by both ruthenium red (100 μm) and La3+(0.1–10 mm), a putative inhibitor of Ca2+ATPase and ATP-dependent Ca2+extrusion, respectively.250
In fura-2–loaded mesenteric arterial VSMCs, propofol (≥ 10 μm) increased the resting [Ca2+]i level after depletion of the SR and inhibition of plasmalemmal Ca2+influx.145
On the basis of sensitivity to ruthenium red and La3+, ketamine was proposed to enhance PMCA activity in pulmonary arterial VSMCs,250which is inconsistent with the results obtained in cultured aortic VSMCs.237However, the nonspecific nature of both ruthenium red and La3+(e.g. , inhibitions of plasmalemmal Ca2+influx263,264and RYR/Ca2+release channel activity265) would limit the interpretation of the data obtained with those inhibitors. The ability of propofol to increase the resting [Ca2+]i level in the absence of a functional SR and plasmalemmal Ca2+influx145suggests that propofol inhibits PMCA activity. Further studies using plasma membrane fractions from VSMCs would be necessary to directly evaluate the anesthetic actions on PMCA activity and definitively prove those previous proposals.145,176,237,250
Anesthetic Effects on Cytosolic Calcium Oscillations
Rhythmic oscillations have been observed in contractile responses to a wide variety of agonists in isolated vessels.54,55,57,157,266,267Correspondingly, oscillatory changes in vessel diameter, blood flow, or even oxygen tension have been observed in vivo .268–272The oscillatory activity is especially evident in small arteries and arterioles55,56,161; however, its functional importance is still a matter of debate. It might be effective in regulating vascular resistance without disturbing tissue perfusion, minimizing fluid filtration into the extravascular space by reducing hydrostatic pressure, and enhancing lymphatic drainage through the pumping action of closely adjacent arterioles.273–275The cellular mechanisms behind oscillatory contractile responses have not been fully clarified. However, they would correlate with cytosolic Ca2+oscillations in VSMCs.58,276,277
In previous contraction studies with isolated small mesenteric arteries and veins,130,133,135,157,161volatile anesthetics (i.e. , halothane, enflurane, isoflurane, sevoflurane) potently inhibited rhythmic oscillations during the contractile response to norepinephrine. Ketamine (≥ 10 μm), thiopental (≥ 3 μm), lorazepam (≥1 μm), diazepam (0.3 μm), and midazolam (≥100 μm) inhibited the cytosolic Ca2+oscillations in cultured pulmonary arterial VSMCs loaded with fura-2.278,279Under the same experimental conditions, propofol (≥10 μm), applied in its commercially available 10% Intralipid® emulsion (Kabi Pharmacia AB, Stockholm, Sweden), also inhibited the Ca2+oscillations.278
The anesthetic-induced inhibition of oscillatory vasomotion could result in alterations of vascular homeostasis such as fine control of tissue blood flow or vascular permeability. To my knowledge, volatile anesthetics have not yet been shown to inhibit cytosolic Ca2+oscillations in VSMCs. They may inhibit cytosolic Ca2+oscillations through effects on the SR, VOCCs, K+channels, ClCachannels, endothelium, and/or gap junctions, all of which have been proposed to be involved in the generation of cytosolic Ca2+oscillations,54–61and to be targets for volatile anesthetics.27,118,220,244,245,280,281The cellular mechanisms of anesthetic-induced inhibition of oscillatory vasomotion or cytosolic Ca2+oscillations in VSMCs remain to be elucidated.
Mechanisms behind General Anesthetic–induced Changes in Myofilament Calcium Sensitivity
General anesthetics have been proposed to significantly influence (either inhibit or enhance) the myofilament Ca2+sensitivity of VSMCs in either absence or presence of receptor stimulation (table 3). However, the underlying mechanisms are not well understood.
Anesthetic Effects on the Myofilament Calcium Sensitivity in the Absence of Receptor Stimulation
There is increasing evidence that volatile anesthetics affect Ca2+activation of contractile proteins (i.e. , basal myofilament Ca2+sensitivity) in VSMCs. However, data are limited regarding the effects of intravenous anesthetics on the Ca2+activation of contractile proteins.
In isolated aorta, pulmonary arteries, and small mesenteric arteries, halothane inhibited the [Ca2+]i–force relation in VSMCs loaded with fura-2, or the Ca2+-induced contractions in VSMCs permeabilized with saponin or β-escin.118,125,127,129,233,234,282However, in pulmonary arterial, saponin-permeabilized VSMCs, halothane enhanced the Ca2+-induced contraction (at 21°–23°C); however, it was transient and followed by sustained inhibition.230,233
In rat mesenteric arterial, fura-2–loaded VSMCs, during KCl depolarization, halothane (0.5–4.5%) markedly shifted the [Ca2+]i–force relation as well as the MLC phosphorylation–force relation to the right without altering the MLC phosphorylation level or the [Ca2+]i–MLC phosphorylation relation.129However, in rat aortic VSMCs loaded with fura PE3 (a fura-2 derivative), halothane did not alter the [Ca2+]i–force relation during KCl depolarization.282
In isolated small mesenteric arteries, enflurane inhibited the Ca2+-induced contraction in VSMCs permeabilized with β-escin.127
In isolated aorta, pulmonary arteries, and small mesenteric arteries, isoflurane also inhibited the [Ca2+]i–force relation in fura-2–loaded VSMCs,27,125or Ca2+-induced contraction in saponin-permeabilized VSMCs.234,235However, in femoral or pulmonary arterial VSMCs permeabilized with saponin, isoflurane conversely enhanced the Ca2+-induced contraction (at 21°–23°C); it was sustained in the femoral arterial VSMCs,283whereas it was transient and followed by sustained inhibition in pulmonary arterial VSMCs.235
In isolated rat small mesenteric arteries, isoflurane and sevoflurane inhibited the [Ca2+]i–force relation only in the fura-2–loaded, membrane-intact VSMCs, but not in the β-escin–membrane-permeabilized VSMCs.25,27
The above results suggest that halothane, enflurane, isoflurane, and sevoflurane inhibit the Ca2+activation of contractile proteins. However, the underlying mechanisms are not well understood. In earlier studies, halothane and isoflurane stimulated formation of either cAMP or cGMP in aortic VSMCs,120,248suggesting that increased cAMP or cGMP levels, which reduce the Ca2+/MLCK-dependent phosphorylation of MLC20,21,23,37,77,85might be involved in their inhibition of Ca2+activation of contractile proteins. However, more recent studies have failed to confirm their ability (halothane,179,196,204,206,228isoflurane,172,196,204,206,228sevoflurane180) to stimulate the formation of either cyclic nucleotide in aortic VSMCs. Because all those results were obtained in rat aortic VSMCs, the discrepancies might be explained by differences in experimental conditions (e.g. , endothelial intactness, cultured VSMCs vs. fresh VSM tissue).
Su et al. 230,233,235,283reported the ability of halothane or isoflurane to enhance the contractile response to Ca2+in pulmonary or femoral arterial, membrane-permeabilized VSMCs. In pulmonary arterial VSMCs, the enhancement was followed by sustained inhibition,230,233,235but not in femoral arterial VSMCs,283suggesting differences in vascular responsiveness to volatile anesthetics between vascular beds. On the basis of sensitivity to PKC inhibitors (i.e. , bisindolylmaleimide 1 [0.1–10 μm], Gö-6976 [0.1–10 μm]) and a CaMKII inhibitor (i.e. , CKIINtide [0.01–1 μm]), they have proposed involvement of those signaling pathways in either the enhancement or inhibition.230,233,235,283However, bisindolylmaleimide 1 and Gö-6976 are both capable of inhibiting many other protein kinases with similar potency to PKC isoforms.284,285In addition, bisindolylmaleimide 1 may exert some other nonspecific actions at micromolar or higher concentrations.286,287Therefore, further studies would be necessary to prove their proposals on PKC involvement.233–235,283
By comparing the effects of halothane on [Ca2+]i–force relation, MLC20phosphorylation–force relation, and [Ca2+]i–MLC20phosphorylation relation during KCl depolarization, Tsuneyoshi et al. 129proposed that the inhibited contractile response to KCl by halothane is due to suppressed myofilament sensitivity to both Ca2+and MLC20phosphorylation, and independent of the Ca2+–CaM–MLCK pathway. Evidence is accumulating that changes in membrane potential alter myofilament Ca2+sensitivity94,95,98,99or activity of membrane-associated enzymes (e.g. , PLC,96,97,288–291adenyl cyclase292,293). Specifically, depolarization and hyperpolarization of VSMC membrane have been shown to increase and decrease myofilament Ca2+sensitivity, respectively.94,95,98,99Thus, there is a possibility that in the study,129halothane inhibited the Ca2+sensitizing mechanisms specifically involved in the KCl response. The above-described difference in the effects of halothane on Ca2+–force relation during KCl depolarization between rat aortic282and mesenteric129VSMCs might reflect a regional difference in vascular responsiveness to halothane.
The above-mentioned ability of isoflurane and sevoflurane to inhibit the Ca2+activation of contractile proteins only in membrane-intact, but not in membrane-permeabilized, VSMCs25,27suggests that their inhibitions are mediated by some membrane-associated or intracellular regulatory mechanism of contraction that is impaired as a result of the membrane permeabilization. By contrast, the ability of halothane and enflurane to inhibit the Ca2+activation of contractile proteins under membrane-permeabilized conditions118,127suggests that their inhibitions would be at least in part independent of the intact plasma membrane.
The above-discussed data suggest a possibility that in spite of their vasodilatory properties in vivo (as recognized by the dilated cutaneous veins), halothane, enflurane, and isoflurane may constrict vessels by enhancing the Ca2+activation of contractile proteins and/or by stimulating Ca2+release from the SR in VSMCs. However, they also have been shown to have inhibitory actions on VSMCs (tables 3–5), and their overall effects on VSMCs would be determined by the net balance between the excitatory and inhibitory actions on VSMCs. In addition, some of the vasoconstrictor actions were observed only at 21°–23°C after treatment with the chemical detergents.230,233,235,283Therefore, their in vivo relevance is currently unclear. However, previous studies1,10–13,18–20have reported decreases in blood flow associated with increases in vascular resistance or tone in several vascular beds (e.g. , heart, liver, intestine, and skeletal muscle) during administration of those three anesthetics, possibly reflecting their direct vasoconstrictor actions.
In VSMCs from systemic resistance arteries,26,165ketamine (0.3–1 mm) did not affect the [Ca2+]i–force relation in VSMCs from systemic resistance arteries under both fura-2–loaded and membrane-permeabilized conditions.26,165Similarly, ketamine (0.1 mm) did not affect the [Ca2+]i–force relation in pulmonary venous, fura-2–loaded VSMCs.294Midazolam (10–30 μm) also did not affect the [Ca2+]i–force relation in mesenteric arterial, fura-2–loaded VSMCs.170By contrast, propofol (100 μm) modestly, although significantly, shifted the [Ca2+]i–force relation to the left in pulmonary arterial, fura-2–loaded VSMCs.143
These results26,143,165,170,294suggest that both ketamine and midazolam do not affect, whereas propofol enhances, the Ca2+activation of contractile proteins in VSMCs.
Anesthetic Effects on Myofilament Calcium Sensitivity during Stimulation with Receptor Agonists
Volatile anesthetics, in clinical concentrations, have been reported to affect myofilament Ca2+sensitivity in VSMCs stimulated with receptor agonists,25,27,128,282or some signaling pathways presumed to be involved in the agonist-induced increase in myofilament Ca2+sensitivity (e.g. , PKC pathway, Rho–Rho kinase pathway). However, information is also limited regarding intravenous anesthetic actions on the myofilament Ca2+sensitivity during receptor stimulation.
Halothane and isoflurane inhibited the IP formation induced by receptor agonists (i.e. , AVP, platelet-derived growth factor) in cultured aortic VSMCs.227,295Isoflurane also inhibited IP formation induced by acetylcholine in VSMCs of isolated coronary arteries.295Sevoflurane was recently reported to inhibit phosphorylation of the cPKC-α, an indicator of its activity, in rat aortic VSMCs stimulated with AT-II.296
In isolated porcine coronary arteries, halothane (1–3%) and isoflurane (1–3%) little affected the contractile response to phorbol ester (PBE), an analog of diacylglycerol.228,295By contrast, in isolated rat coronary arteries, halothane (0.8–1.5%) inhibited, whereas isoflurane (1.2–2.3%) enhanced, contractile response to PBE.297In isolated rat aorta, halothane (2–3%) and isoflurane (4%) both inhibited the contractile response to PBE.128In membrane-permeabilized pulmonary arterial VSMCs, both of them initially enhanced but later inhibited the contractile response to PBE (at 21°C).233,235
In isolated rat aortic VSMCs, sevoflurane inhibited guanosine-5′-(3-O -thio) triphosphate (GTPγS, a potent and stable activator of G proteins including RhoA)–induced contraction that was sensitive to 3 μm Y27632 (a Rho-kinase inhibitor), as well as GTPγS-stimulated membrane translocation of RhoA and Rho kinase from the cytosol.298
The ability of halothane and isoflurane to inhibit IP formation227,295suggests that during receptor stimulation, those anesthetics inhibit myofilament Ca2+sensitivity by inhibiting the PLC-mediated formation of diacylglycerol and thereby preventing PKC activation. However, as described above, previous studies128,228,235,295,297examining the effects of halothane and isoflurane on vascular responses to the physiologic PKC activator PBE have yielded conflicting results. The differences might be due to those in animal species, experimental conditions, and/or vascular bed. Because PBE slightly increases the [Ca2+]c in porcine coronary arterial and rat aortic VSMCs,128,299the PBE contraction observed in those VSMCs128,228,295could be mediated not exclusively by increased myofilament sensitivity to the basal [Ca2+]c but in part by an increase in [Ca2+]c. Therefore, the interpretation of results obtained in some of the previous studies128,228,295would not be straightforward, and it is currently uncertain whether volatile anesthetics interfere with the PKC pathway at steps after the production of diacylglycerol.
The ability of sevoflurane to inhibit the AT-II–induced phosphorylation of PKC296would suggest that sevoflurane inhibits the activation of PKC during stimulation with AT-II.
Recent evidence suggests that during receptor stimulation, the inhibitory signal to MLCP is communicated by RhoA, a monomeric G protein, to a Rho kinase that phosphorylates MLCP and inhibits its catalytic activity, increasing MLC20phosphorylation (fig. 1).62Upon activation, both RhoA and Rho kinase are translocated from the cytosol to the membrane in VSMCs.62The recently reported ability of sevoflurane to inhibit GTPγS-stimulated membrane translocation of RhoA and Rho kinase from the cytosol298thus suggests that during receptor stimulation, sevoflurane inhibits the Rho–Rho kinase pathway, thereby reducing the myofilament Ca2+sensitivity and causing vasodilation. However, in that study,298its mechanistic link to the observed vasodilator action seemed unclear because of the nonspecific nature of Y-27632 (3 μm) as a Rho-kinase inhibitor.285
Volatile anesthetics have been reported to either activate207–210,212or inhibit244,245,300,301the K+channel activity in VSMCs. Therefore, they may modulate myofilament Ca2+sensitivity by altering the K+channel activity and thereby causing changes in the membrane potential, which have been suggested to alter the myofilament Ca2+sensitivity in either presence or absence of receptor stimulation.94,95,98,99
Ketamine (0.3–0.32 mm) inhibited PLC activity in femoral arterial VSMCs,140as well as phenylephrine (α1-adrenergic agonist)–stimulated IP3production in mesenteric arterial VSMCs139; however, ketamine (0.3–1 mm) did not influence the [Ca2+]i–force relation during stimulation with norepinephrine in mesenteric arterial, fura-2–loaded VSMCs.26In pulmonary venous, fura-2–loaded VSMCs,294ketamine (0.1 mm) inhibited the [Ca2+]i–force relation only during stimulation with acetylcholine but not in its absence, mimicking the effect of bisindolylmaleimide 1 (3 μm, the PKC inhibitor). Midazolam (10–30 μm) did not influence the [Ca2+]i–force relation during stimulation with norepinephrine in mesenteric arterial, fura-2–loaded VSMCs.170
The ability of ketamine to inhibit PLC activity140and phenylephrine-stimulated IP3production139suggests that ketamine may inhibit the agonist-induced increase in myofilament Ca2+sensitivity by inhibiting PLC-mediated PIP2breakdown and preventing subsequent PKC activation. However, its inability to inhibit myofilament Ca2+sensitivity during stimulation with norepinephrine26suggests that PKC may not play a significant role in the norepinephrine-induced increase in myofilament Ca2+sensitivity. In pulmonary venous VSMCs, ketamine seems to inhibit the acetylcholine-induced, presumed PKC-mediated increase in myofilament Ca2+sensitivity. Midazolam probably does not influence the norepinephrine-induced increase in myofilament Ca2+sensitivity in mesenteric arterial VSMCs.
Investigations with Human Vessels
In previous studies using isolated vascular preparations from various organs, most general anesthetics affected vascular reactivity through direct (i.e. , nonneural) actions on VSMCs and/or endothelial cells. However, as discussed above, the underlying mechanisms are not entirely clear. In addition, the obtained results on some of them have not been consistent, but conflicting, apparently depending on vascular beds, size of blood vessels, experimental condition, and animal species. General anesthetics, because of their high lipophilicity, would be able to gain access to numerous membranous structures in VSMCs and endothelial cells. Then, they would perturb the functional integrity of membrane proteins by acting either directly on their hydrophobic sites or indirectly through the lipid bilayer surrounding them, and thereby exert multiple actions on VSM reactivity. In addition, general anesthetics would act on various (i.e. , metabolic, myogenic, neural, humoral, or endothelial) vasoregulatory mechanisms. Therefore, their overall vascular action would represent a balance among their effects on those mechanisms, the relative importance of which varies in different vascular beds. Such a nonspecific nature of general anesthetics would readily lead to the differences in vascular responsiveness between vascular beds, type of blood vessels, experimental conditions, and animal species. Therefore, investigations using human vessels under physiologic conditions would be particularly important to clarify the relevance of direct vascular actions of general anesthetics to their circulatory action observed in the clinical setting. The number of studies using isolated human vessels122,144,146,151,168,183,193,202,223,302–308is increasing, and they have reported both similarities and differences in vascular responsiveness to general anesthetics between humans and animals. However, our knowledge is still limited.
Investigations with Vessels from Subjects with Cardiovascular Dysfunction
General anesthesia rarely causes cardiovascular collapse in healthy younger patients, but often in patients with cardiovascular dysfunction (e.g. , aged patients; neonates; patients with heart diseases, atherosclerosis, hypertension, diabetes mellitus, and sepsis). Many of the cellular mechanisms regulating vascular tone—i.e. , potential targets for general anesthetics—are known to be altered in those susceptible populations (e.g. , endothelial function, ion channel activity, enzyme activity).309–314Vascular responsiveness to general anesthetics thus could be altered in those pathologic conditions. Some investigators166,212,231,315,316have already demonstrated that vascular sensitivity to general anesthetics (i.e. , propofol, isoflurane, sevoflurane) is indeed altered (either enhanced or attenuated) in hypertensive subjects. However, little information is available regarding direct actions of general anesthetics on VSMCs from atherosclerotic, diabetic, and aged subjects. These topics should be the subjects of future studies.
Investigations on Intravenous Anesthetics
In most previous studies investigating the action of intravenous anesthetics, as mentioned above, their effective concentrations were higher than their clinically relevant free concentrations (table 6). Therefore, intravenous anesthetics may not directly influence VSM reactivity in the normal clinical setting. However, vascular sensitivity could be altered in the in vitro or in situ conditions, particularly in cultured VSMCs. In addition, in many previous in vitro studies, their actions have been evaluated on maximal or near-maximal contractile responses to agonists. However, it must be rare that VSMCs are being maximally contracted under physiologic conditions, and therefore, the anesthetic actions, assessed by percent changes from the control response, could have been underestimated in those studies. Further studies thus would be necessary regarding the intravenous anesthetic actions on submaximal responses to lower concentrations (i.e. , ≤ EC50) of agonists that are involved in the physiologic regulation of vascular tone.
Investigations with Systemic Resistance and Capacitance Vessels
In previous studies with systemic resistance arteries from rats or rabbits, some general anesthetics (i.e. , halothane, isoflurane, sevoflurane), despite their hypotensive action in vivo , did not consistently decrease vascular tone but occasionally increased it.26,130,133,135,160,161,176Depending on experimental conditions (e.g. , endothelial intactness) or species (e.g. , rat vs. rabbit), they enhanced the contractile response to norepinephrine133,135,176that plays a central role in the sympathetic maintenance of vascular tone in vivo . In addition, they attenuated both the nitric oxide–mediated and EDHF-mediated vasodilator responses133,135,160,176that would be essential in keeping the vasculature in an appropriately dilated state. Therefore, at this time, no definite evidence seems available for any general anesthetic indicating that its direct action on arterial VSMCs contributes to systemic hypotension during its administration. However, as discussed above, the direct action of volatile anesthetics on arterial VSMCs may contribute to the prolonged systemic hypotension observed during the postanesthesia period.10,47
It has been recognized that venodilation, particularly in the splanchnic region, and a resultant decrease in venous return also underlies systemic hypotension during general anesthesia.2In previous studies with isolated mesenteric veins,157,158,317halothane, enflurane, and isoflurane inhibited the contractile response to norepinephrine. In addition, in recent in situ experiments, propofol hyperpolarized mesenteric venous VSMCs by activating K+channels.318,319Therefore, systemic hypotension during anesthesia with those anesthetics would be due, in part, to their direct action on venous VSMCs in the splanchnic circulation. Presumably because of the technical difficulty, information about the direct action of general anesthetics on venous VSMCs is still limited.157,158,294,303,305,307,308,317–319Again, on consideration of possible species differences, further studies using human capacitance vessels, as well as human systemic resistance vessels, would be essential to evaluate a possibility that the direct (i.e. , nonneural) actions of general anesthetics on VSMCs contribute to systemic hypotension during their administration.
Investigations on Microcirculation
Vasodilator responses to nitric oxide, EDHF, and KATPchannels play a key role in the endothelial and/or metabolic regulation of vascular tone in microcirculation.320Vasodilator responses via the β adrenoceptors also would contribute to resting tone of coronary arterial microvessels.321,322General anesthetics have been shown to inhibit all these vasodilator responses (table 2). In addition, in coronary microvessels, some general anesthetics have been reported to influence myogenic response to changes in intravascular pressure or flow (i.e. , shear stress)–induced vasomotion,323–325both of which are essential for fine regulation of blood flow (e.g. , autoregulation).326Therefore, general anesthetics may perturb microcirculatory homeostasis, possibly affecting oxygen delivery to the tissues. However, the clinical significance of such anesthetic actions seems unclear at this time, and further studies evaluating anesthetic effects on venular blood oxygen content under various critical conditions (e.g. , hypoxia, shock) would be helpful for clarification of this issue.
Conducted vasomotor responses—i.e. , propagation of local vasomotor responses to upstream and downstream along the microvessels—are believed to contribute to functional distribution of blood flow in microcirculation of various organs including the brain and heart.327–329Specifically, conducted vasodilation could be an important mechanism for increasing blood supply to meet the metabolic demands (i.e. , functional hyperemia), whereas conducted vasoconstriction would underlie blood flow autoregulation.327However, data are limited regarding the anesthetic actions on the conducted vasomotor response,330and this topic should be the subject to future investigations as well.
In previous in vitro or in situ experiments using blood vessels from various vascular beds, most general anesthetics affected a wide variety of cellular and molecular mechanisms regulating vascular reactivity. However, in most of them, intravenous anesthetics exerted their vascular actions only at supraclinical concentrations, whereas volatile anesthetics exerted their vascular actions at clinical concentrations. Therefore, direct (i.e. , nonneural) actions of volatile anesthetics on VSMCs and/or endothelial cells would contribute to the alterations in hemodynamics and organ blood flow during their administration in the clinical setting. On the other hand, most of the previously observed direct vascular actions of intravenous anesthetics might be irrelevant to normal clinical practice.
The previous results on the direct vascular action of general anesthetics have not necessarily been consistent even in experiments with the same blood vessels, possibly reflecting the differences in vascular responsiveness between experimental conditions or species. It is conceivable that such differences exist, because general anesthetics presumably act on multiple sites within VSMCs and endothelial cells and thereby exert multiple actions on vascular reactivity. In addition, general anesthetics would act on multiple vasoregulatory mechanisms, the relative importance of which differs in different vascular beds, leading to the differences in vascular responsiveness to them among vascular beds. On consideration of such a nonspecific nature of general anesthetics, it would be particularly important to evaluate their actions on human vessels under physiologic conditions, as well as on blood vessels from subjects susceptible to their circulatory depressant effects. In addition, electrophysiologic, biochemical, and molecular-biologic techniques would be essential for future studies to clarify the mechanisms behind the direct vascular actions of general anesthetics. Better understanding of the vascular actions of general anesthetics, as well as vascular physiology, would lead to better circulatory management during general anesthesia.
The author thanks Junichi Yoshitake, M.D., Ph.D. (deceased, former Professor of Anesthesiology, Faculty of Medicine, Kyushu University, Fukuoka, Japan), and Shosuke Takahashi, M.D., Ph.D. (Professor of Anesthesiology, Graduate School of Medical Sciences, Kyushu University), for their encouragement, and Kaoru Izumi, M.D., Ph.D. (Staff Anesthesiologist, Iizuka Hospital, Iizuka, Japan), Jun Yoshino, M.D., Ph.D. (Staff Anesthesiologist, Kyushu Medical Center, Fukuoka, Japan), Kazuhiro Shirozu, M.D. (Postgraduate of Anesthesiology, Graduate School of Medical Sciences, Kyushu University), and Tomoo Kanna, M.D. (Research Associate, Department of Advanced Medical Initiatives, Faculty of Medical Sciences, Kyushu University), for their enthusiastic cooperation as a postgraduate or research associate.