I. INTRODUCTION

Top Previous Next References

Blood vessels respond to transmural pressure elevation with constriction and to pressure reduction with dilation. This behavior, termed the myogenic response, is inherent to smooth muscle and independent of neural, metabolic, and hormonal influences. It is most pronounced in arterioles but can be demonstrated occasionally in arteries, venules, veins, and lymphatics (186). When longitudinal comparisons are made among arterioles of a given vascular network, an inverse relationship between vessel size and myogenic responsiveness is consistently observed (55), although the cerebral circulation may be an exception to this rule (278).

Three examples of myogenic behavior are illustrated in Figure 1. 1) Figure 1A shows the prototypical myogenic response of a cannulated arteriole to a step increase in pressure. After the pressure step, an initial, passive distension is followed by two phases of constriction; upon release of the pressure step, the arteriole transiently collapses, then dilates. 2) In addition to this reaction, arterioles typically develop and maintain some degree of active force at their normal intravascular pressure. This is depicted in Figure 1B by the diameter response of an isolated arteriole that is maximally dilated after being cannulated and pressurized, but then spontaneously constricts to ~50% of its passive diameter when temperature is raised from 22 to 37°C. The constriction would typically be maintained for several hours. 3) A third way in which myogenic behavior is defined can be illustrated by a graph of diameter versus pressure for an arteriole with (active) or without Ca²⁺ (passive) in the bathing media (Fig. 1C). The range of pressures over which the active diameter curve has a less positive slope than the passive curve is termed the myogenic range. In arteries, this slope may be only slightly less positive than that of the passive curve (231), but in arterioles, it can be quite negative (55). Additional descriptions of myogenic behavior in isometric preparations are discussed in section IIIB.

View larger version (11K): [in this window] [in a new window]   Fig. 1. Examples of myogenic behavior in arterioles (for explanation, see text). Approximate duration of pressure step in A is 2 min. Quantitative scales for time, pressure, and diameter depend on vessel type, size, species, and method of study. For representative examples, see References 64 (A), 211 (B), and 55 (C).

Discovery of the myogenic response is credited to Bayliss in 1902, when he recorded large increases in the volume of the dog hindlimb following release of brief aortic occlusions (14). Bayliss considered this response too rapid to be mediated by accumulation of metabolites and thought it reflected the same mechanism by which isolated arteries constricted following sudden distension. It is not usually appreciated that Bayliss' experiments were predated by a number of other studies, for example, by Jones in 1852 (190), Ostroumoff in 1868 (283), and Gaskell in 1881 (104). It was Bayliss, however, who clearly formulated the idea that a significant component of vascular tone could be determined by intravascular pressure.

Bayliss' ideas were challenged by Anrep (8), who believed the hindlimb response could be explained by metabolic factors. Partly because of Anrep's persuasive arguments, relatively little work on the myogenic response was performed over the subsequent 45 years. Notable exceptions to this include the work of Fog (89), Forbes et al. (93), Wachholder (353), Klemensiewicz (204), and Bürgi (36). Despite these studies, which tended to confirm Bayliss' original findings, the majority of workers in this field attributed local vascular regulation primarily to chemical and neural mechanisms until Folkow demonstrated that denervated preparations developed pressure-dependent vascular tone (90) and that autoregulation of blood flow was due in part to a nonneural, pressure-dependent mechanism (91).

Due in large measure to Folkow's work, Selkurt and Johnson (316) in the 1950s, and Johnson (183) in the 1960s, studied the myogenic response using increasingly sophisticated whole organ techniques and concluded this mechanism could account for significant vascular resistance changes in vivo. Concurrently, Burnstock and Prosser (37) demonstrated that strips of nonvascular smooth muscle reacted to quick stretch with active force generation, and Sparks (331) described the same phenomenon in vascular smooth muscle (VSM). In the late 1960s, Johnson (184) and Wiederhielm (370) pioneered the application of techniques for quantitating the myogenic response in the microcirculation, which led to intense investigation over the subsequent decade (11, 33, 188). Development of isolated vessel techniques in 1981 (75) enabled more careful quantitation of the myogenic response and its underlying mechanisms, first in small arteries (132, 279) then in arterioles (175, 211), where the effects of pressure could be clearly distinguished from flow, metabolic, neural, and endothelial influences.

This review attempts to summarize and synthesize what is known regarding the cellular mechanisms underlying the vascular myogenic response. Because myogenic behavior is only one aspect of VSM mechanotransduction, thorough treatment of that topic would include a discussion of mechanical effects on secretion and growth as well as contractile function, which is beyond the scope of this review (see Ref. 278 for more general coverage). The reader is referred to Johnson's comprehensive review on the myogenic response for information on studies before 1979 (186), to reviews of relevant microcirculatory studies from 1979 to 1990 (55, 61), and to several shorter reviews on myogenic mechanisms by workers in this field (23, 24, 51, 137, 241, 278). The role of the endothelium has been reviewed previously (24, 241). Because isolated vessel preparations have provided the most definitive information regarding cellular mechanisms involved in the response, the present article emphasizes in vitro studies using blood vessels and single VSM cells performed mostly after 1980. Particular attention is given to recent work using biochemical and electrophysiological techniques and to data collected from arterioles; however, relevant data from conduit arteries are cited when specific information about microvessel function is missing.

	II. PHYSIOLOGICAL SIGNIFICANCE

Top Previous Next References

In the vascular system, the myogenic response has been proposed to participate in a number of physiologically important functions. The two most important of these are 1) establishment of basal vascular tone and 2) autoregulation of blood flow and capillary hydrostatic pressure. Other roles for the myogenic response have been discussed in detail elsewhere (186, 187).

Basal vascular tone is a prerequisite for dilator influences. It establishes an underlying arteriolar constriction, a "regional blood flow reserve" (92), upon which other control mechanisms produce vasodilation or vasoconstriction.

Both Bayliss and Folkow suggested that basal tone might result from myogenic mechanisms. This conclusion derived, in part, from the pressure-dependent resistance to flow observed in denervated whole organ preparations (91, 245, 361). A common finding in microcirculatory studies is that responsive and stable preparations are associated with the development of spontaneous tone in nearly all arterial vessels less than 150 µm ID; the tone is easily compromised by excessive levels of anesthesia, extensive surgical manipulation, or trauma (60, 75). In isolated artery and arteriole preparations, the level of tone is often comparable to that observed in the same vessels in vivo and rarely develops if the vessels are not pressurized to a physiological level (55, 60).

In addition to the effect of a static pressure head, another component of vascular tone may be related to pulsatile pressure. A classic study of isometric portal vein by Johansson and Mellander (180) demonstrated both static- and rate-sensitive components in the response to stretch, evident in the electrical and mechanical activity of the preparation. A similar effect was observed in the cerebral artery, although maximum sensitivity occurred at a much different rate of stretch than portal vein (259). In studies of isolated pump-perfused organs, switching from static to a pulsatile pressure produced an increase in calculated vascular resistance of the perfused organs (303, 319). Mellander (244) suggested these responses reflected a rate-sensitive myogenic component that is essential for the development of normal vascular tone. However, experimental support for this idea is weak because isolated arteries and arterioles usually develop tone comparable to that observed in vivo when connected to a static pressure head (55). Moreover, switching from static to pulsatile pressure produces no significant change in the diameter of cannulated arterioles (57) or small arteries (113). Thus the response of isolated organs to pulsatile perfusion may involve more than simply a pressure effect, possibly due to release of endothelium-derived vasoactive factors (91, 167).

The myogenic response has also been postulated to play a central role in the maintenance of constant blood flow and capillary hydrostatic pressure (P_c) during variations in systemic arterial pressure. Whole organ data collected by Johnson (185) suggested that changes in arterial inflow or venous outflow pressure produced changes in arterial resistance that would serve to minimize changes in capillary hydrostatic pressure. Mellander and colleagues (26, 177) demonstrated that tissue volume of cat hindlimb skeletal muscle was nearly constant over a wide range of systemic arterial pressures (30-170 mmHg). Under the assumption that a constant tissue volume reflected a constant P_c, it was concluded that "autoregulation of P_c" was achieved through myogenic adjustments of arteriolar tone. However, this conclusion assumed that other Starling forces were not involved in control of tissue volume and that other local regulatory mechanisms did not contribute significantly to the vascular resistance adjustments (31, 52, 112). Even though whole organ techniques are subject to significant limitations (61), direct measurements of P_c in microcirculatory preparations have been unable to completely resolve this issue (see Ref. 61 for review).

It is important to note that the contribution of myogenic mechanisms to P_c regulation might depend on whether a selective change in arterial or venous pressure occurs or whether both pressures change equally (61, 115). In the case of perfusion pressure reduction, microcirculatory data suggest that partial P_c regulation does occur in some tissues but that a significant fraction of that regulation may be contributed by factors other than the myogenic response (31, 53, 111, 327). However, when arterial and venous pressures are equally raised or lowered, as during postural changes, the contribution of the myogenic response to P_c regulation appears to be much greater (52, 92, 224); this may be related to the position at which an arteriole normally rests on its pressure-diameter curve (64) or to the fact that endothelial-derived nitric oxide (the release of which is altered if flow changes along with pressure) is a potent antagonist of myogenic tone (210).

	III. CONCEPTUAL BASIS FOR MYOGENIC BEHAVIOR

Top Previous Next References

Vertebrate muscle, including VSM, exhibits length-dependent regulation of force, such that peak force development due to contractile protein interaction is generated at an intermediate, optimal sarcomere length. A classic concept in cardiac and smooth muscle mechanics is the distinction between active force development due to initial length (preload) and that due to activation (inotropy, contractility) (179). However, over the past 20 years, experiments on cardiac muscle have clearly demonstrated that muscle length itself influences contractility (217), leading to the conclusion that preload and ionotropic state are not independent regulators of active force (5, 179). There is now general agreement that the relative steepness of the ascending limb of the length-active tension relationship in cardiac muscle reflects a progressive shift to increasing levels of activation with increasing length (97). That relationship broadens, as predicted, at high (fixed) levels of Ca²⁺ in skinned preparations of cardiac muscle (84), in contrast to skeletal muscle where the length-active tension relations of intact and skinned (at saturating Ca²⁺ concentration) preparations are superimposable (5).

In blood vessels, initial length is a well-known modulator of agonist sensitivity (110, 122, 133, 138, 231, 271, 293, 332, 341, 347). Conversely, agonists often potentiate myogenic responsiveness (83, 242, 299, 334, 335, 347). Thus it is likely that agonist- and stretch-activated signaling pathways overlap. Agonists such as norepinephrine (NE) are positive inotropic agents for smooth muscle. Likewise, a myogenic constriction is considered to represent an enhanced smooth muscle activation state (175, 186). Despite the plausibility of this idea, experimental support for it is mostly indirect. Johnson (184), and others (59), analyzed the behavior of in vivo arterioles following step changes in perfusion pressure and concluded that smooth muscle must shift to a higher active length-tension curve in response to elevated pressure. However, those studies were limited in that active and passive components of wall tension could not be distinguished. Nevertheless, a shift in activation state is supported by isolated arteriole experiments showing that maximal velocity of arteriolar muscle shortening increases with pressure over the myogenic range of the vessel (54).

The two experimental approaches typically used to quantitate the vascular myogenic response, isometric and isobaric protocols, have often led investigators to different conclusions with regard to mechanisms (74). Perhaps part of the reason for this is that the magnitude, time course, and direction of vascular wall tension changes in isometric contractions of vascular rings and strips are very different from those in isobaric contractions of cannulated arterioles and arteries. These differences are illustrated in Figure 2. In isometric preparations, stretch activation is represented by a slower, secondary increase in tension after stretch. By this definition, skeletal (301), cardiac (5), and smooth muscle (37) all exhibit stretch activation. In isobaric preparations, activation of the contractile apparatus following a pressure increase results in a constriction that secondarily reduces total wall tension. However, this reduction is achieved by a decrease in passive tension, which more than compensates for the increase in active tension due to activation of the contractile machinery. Because cannulated vessels often respond with sustained constrictions to pressure elevation, it has been suggested that wall tension, rather than smooth muscle cell length, may be regulated during a myogenic (isobaric) constriction (186, 208). If contractile and sensor elements were arranged in series, a tension-control system would require only modest gain to perfectly regulate diameter, and sustained constrictions could be achieved in the face of elevated transmural pressure (as shown in Fig. 2B). Although widely accepted, the wall tension hypothesis has been difficult to test experimentally, and support for it derives chiefly from correlative evidence (38) and logical arguments (347).

View larger version (11K): [in this window] [in a new window]   Fig. 2. Comparison of protocols for determining stretch activation of a blood vessel. A: in an isometric experiment, myogenic activity is identified by a secondary increase in tension observed in response to a rapid length increase. Dotted line shows estimated stress-relaxation curve in Ca2+-free solution; difference between tension trace and this curve (or, more conservatively, a horizontal line drawn from beginning of secondary response) is defined as myogenic tone (259). B: in an isobaric experiment, a rapid pressure increase elicits a sharp rise in tension and length, but as arteriole constricts (in this case below control diameter), tension is reduced. Myogenic tone is defined as difference between maximum (dotted line refers to Ca2+-free) and minimum diameter (194) [or, more conservatively, between control and minimum diameter (128, 231)].

There are a number of other differences between the behavior of isometric and isobaric preparations. Isometric preparations typically show maximal stretch activation in response to large, and perhaps unphysiological, changes in length. For example, secondary force production is maximal for length increases to 149% of control in rat mesenteric artery (347), 150% of control in rabbit basilar artery (259), and 140% of control in pig coronary artery (290). In contrast, isobaric preparations show maximal constrictions in response to much smaller length changes (<25% of control) (55, 347), even in the absence of detectable distension (64). In vessels of the same size and type, isobaric preparations exhibit different agonist sensitivity than isometric preparations (77, 174, 225, 313, 347), as well as differences in the magnitude of agonist-induced VSM depolarization (313). Interestingly, most of the evidence for stretch-activated Ca²⁺ entry through a non-voltage-dependent pathway comes from isometric preparations (23, 168, 169, 215, 216, 379) (see sect. IVA6).

One phenomenon confirmed by both isometric and isobaric preparations is shortening deactivation. In isometric protocols, shortening deactivation is the disproportionate decline in force relative to length observed in actively contracting muscle (5). In isotonic release protocols, it is a depression in shortening velocity in response to a step decrease in length (240). Shortening deactivation is observed in both large and small vessels with (175) or without myogenic tone (29, 126, 300) as well as in nonvascular smooth muscle (121), cardiac muscle (5), and skeletal muscle (343). Jackson and Duling (175) demonstrated this phenomenon in pressurized arterioles. The mechanism of shortening deactivation has not been elucidated but is thought, in other muscle types, to represent changes in mechanisms controlling intracellular Ca²⁺ concentration ([Ca²⁺]_i) as well as changes in myofilament Ca²⁺ sensitivity (84). In arterioles, shortening deactivation is more pronounced with intrinsic tone than with agonist-induced tone (175). It seems reasonable to conclude that the myogenic constrictions and dilations characteristic of pressurized vessels reflect the same underlying mechanisms represented in isometric rings by stretch activation and shortening deactivation, respectively.

In summary, although some authors have made distinctions between the terms myogenic response, stretch activation, pressure-dependent contraction, myogenic tone, basal tone, spontaneous tone, and intrinsic tone (24, 278, 280, 335, 350), the limited amount of quantitative information available in any one tissue restricts the usefulness of such an approach at the present time. For the purposes of this review, we assume that all of the terms above describe cellular processes with similar underlying mechanisms. Doubtless, some of the discrepancies in the literature regarding mechanisms will be resolved when this issue is addressed systematically.

	IV. TRANSDUCTION MECHANISMS

Top Previous Next References

Bohr and colleagues (345) have been credited (259, 278) with the initial suggestion that the myogenic response might reflect an improved excitation-contraction coupling resulting from membrane depolarization and increased Ca²⁺ permeability. This idea was based on simultaneous measurements of tension and membrane potential in taenia coli by Bülbring (35), coupled with the demonstration of Ca²⁺-dependent myogenic tone in resistance vessels (345). Currently, the prevailing thought is that a myogenic contraction is initiated by VSM depolarization (mechanisms not yet agreed upon) which then regulates Ca²⁺ entry through voltage-gated Ca²⁺ (VGC) channels (241). This basic mechanism, as proposed in Figure 3, is almost certainly modulated by a number of intracellular signaling mechanisms. The experimental evidence for each of these components is discussed in section IVA. These discussions will almost exclusively focus on mechanisms activated in response to pressure elevation (or increased stretch), but it is assumed that the same mechanisms are modulated in the opposite way in response to pressure reduction.

View larger version (19K): [in this window] [in a new window]   Fig. 3. Putative membrane mechanisms involved in vascular smooth muscle mechanotransduction (for details, see text).

A.  Electromechanical Coupling

1.  Depolarization of VSM

Since Bülbring's original study (35), much additional evidence is now available to suggest that membrane depolarization plays a central role in the response of smooth muscle to stretch (46, 135, 205, 250, 333, 347, 368). Resting potentials of VSM cells typically range from 60 to 75 mV in unpressurized small arteries and arterioles (150, 267), and graded depolarization is observed as pressure is increased [although membrane potential usually cannot be measured continuously as pressure is changed due to movement of the vessel wall (368)]. At physiological pressures, resting potentials range from 40 to 60 mV (78, 132, 266, 267, 394). The work of Harder and others has demonstrated that pressure- or stretch-induced depolarizations occur in a number of different vascular (132, 205, 328, 368) and nonvascular (46, 212, 363) smooth muscle preparations. Figure 4A summarizes data from cat cerebral artery myocytes as pressure was changed from 0 to 160 mmHg, showing that a graded, 20-mV depolarization occurred at pressures between 30 and 110 mmHg, which was the range associated with myogenic tone. Pressurization also increased the rate of action potential firing (132, 328). Both depolarization and constriction were attenuated when the extracellular Ca²⁺ concentration was reduced but were unaffected by tetrodotoxin (to block voltage-gated Na⁺ channels) or phentolamine (to block the action of NE released from nerve terminals) (132).

View larger version (11K): [in this window] [in a new window]   Fig. 4. Pressure- and stretch-induced depolarization of smooth muscle in pressurized arteries (A) and longitudinally stretched single cells (B). VSM, vascular smooth muscle; Em, membrane potential. [Data in A are redrawn from Harder (132) by computing averages of individual data points. Data in B are redrawn from Davis et al. (58) (open circles), Wellner and Isenberg (363) (solid circles), and Setoguchi et al. (317) (open square with length estimated).]

Single smooth muscle cells also exhibit graded depolarization when longitudinal stretch is applied (Fig. 4B). This observation was first recorded in pig coronary VSM (58), then in bladder myocytes (363), and more recently in mesenteric artery myocytes (317). When coronary artery myocytes were stretched 25% beyond their slack length, a 35-mV peak depolarization (from a resting potential of 52 mV) was recorded (58, 381). This degree of stretch was equivalent to that seen in isolated arterioles rapidly pressurized from the minimum to the maximum of their myogenic range (64). In single-cell preparations, stretch was also associated with initiation of action potentials or an increase in action potential firing rate (363).

Despite the above evidence, it has been difficult to establish a definitive cause-and-effect relationship between membrane depolarization and myogenic responsiveness. In preparations without inherent myogenic tone, KCl application is often used to mimic myogenic depolarization, yet the behavior of KCl-activated and spontaneously myogenic preparations is often different, leading to the conclusion that simple, electromechanical coupling cannot fully account for myogenic behavior and that other mechanisms, e.g., changes in Ca²⁺ sensitivity, must be involved (348, 367). Two types of experiments have been used to minimize pressure-induced depolarization. 1) Vessels permeabilized with saponin or -toxin, in which no membrane potential can be generated, fail to demonstrate myogenic tone or constrict to pressure elevation (80, 178, 239, 400). 2) Depolarization of normal arterioles with KCl should theoretically prevent stretch-induced membrane potential changes when sufficiently high concentrations of extracellular K⁺ ([K⁺]_o) are reached, because the K⁺ equilibrium potential approaches 0 mV. Yet, intermediate increases in [K⁺]_o could shift the VSM membrane potential to a more optimal point on the open probability versus membrane potential relationship for VGC channels, thereby enhancing Ca²⁺ entry through that pathway (this effect has been demonstrated in pial arteries; Ref. 108). When the sustained phases of myogenic contractions are analyzed, KCl consistently reduces myogenic responsiveness, as indicated by the increased values of calculated myogenic index in Table 1. However, when individual records are shown, it is clear that the initial constrictor phase of the response is retained (239, 369). This differential action of KCl probably reflects differences in the underlying mechanisms involved in the two components of the response. In addition, KCl substitution protocols may have unanticipated actions on ion transporters or contractile protein sensitivity to Ca²⁺ (276, 388).

Because the resting potential of smooth muscle is determined to a large extent by K⁺ (267), stretch-induced depolarization could be explained by activation of mechanosensitive (MS) ion channels promoting Na⁺ or Ca²⁺ influx, Cl efflux, or inhibiting K⁺ efflux (Fig. 5). Sodium-permeable MS channels were first described in cultured skeletal muscle cells (119) and have since been found in a number of cell types, including smooth muscle (Table 1). Likewise, MS K⁺ and Cl channels have also been described in several cell types (254). On the basis of both theoretical considerations and experimental evidence, it is thought that MS channel gating is controlled by forces transmitted through the cytoskeleton (see sect. IVE and Ref. 308). Mechanosensitive channels appear to be involved in many aspects of cell function, but it is not clear whether different mechanical stimuli activate different classes of MS channels. Stretch-activated currents, such as those recorded in muscle cells (119, 201), are often contributed by nonselective cation channels with characteristics similar to currents in mechanotransduction organs (274), whereas volume-activated currents are typically carried by Cl (125, 307). Some studies have distinguished between volume- and stretch-activated currents in the same cell type (163, 310, 351, 396).

View larger version (15K): [in this window] [in a new window]   Fig. 5. Putative ion channels involved in vascular smooth muscle mechanotransduction (for details, see text). PLA2, phospholipase A2; PLC, phospholipase C; IP3, inositol 1,4,5-trisphosphate; AA, arachidonic acid; 20-HETE, 20-hydroxyeicosatrienoic acid; EET, epoxyeicosatrienoic acid; VGC, voltage-gated Ca2+ channel; KCa, Ca2+-activated K+ channel; Kv, voltage-gated Ca2+ channel; NSC, nonselective cation.

Patch-clamp techniques have been used to identify and characterize a number of MS channels in muscle cells. Because cells in intact vessels are electrically coupled and surrounded by matrix proteins, single cells must be harvested for patch-clamp studies by enzymatic digestion of arteries and arterioles. Mechanosensitive channels relevant to vascular and visceral smooth muscle are summarized in Table 2. In each case, the channels were recorded from single cells using one of the three single-channel recording modes and were activated by suction applied to the rear of the patch pipette. It is apparent from this list that MS channels with a wide range of permeabilities have been identified in smooth muscle, although the most commonly reported type is a nonselective cation channel (NSC).

Are MS currents artifacts? In 1991, Morris and Horn (255) published a controversial paper in which a number of substantial mechanical stimuli failed to elicit whole cell MS currents from Aplysia neurons; because that cell type had been shown to contain a high density of stretch-activated K⁺ channels (322), it was predicted that activation of even a small fraction of the MS channel population would produce easily detectable, whole cell current. These negative results led to the conclusion that MS current might be an artifact of single-channel recording. However, most investigators in this field continue to accept MS current measurements as valid, for a number of reasons summarized previously (129, 241, 337, 338). A compelling argument against the artifact hypothesis derives from the observation that multiple studies in different muscle preparations have now demonstrated reversible, graded, whole cell, MS currents. In addition, MS channels from Escherichia coli have been cloned (337). Nevertheless, the physiological roles of these channels, particularly as they relate to force transduction in muscle, remain to be established.

The first recordings of a MS channel in smooth muscle were made by Kirber et al. (201) in myocytes isolated from toad stomach (Table 2). In cell-attached and inside-out patch recording modes, increases in membrane stretch activated a cation channel permeable to K⁺, Na⁺, and Ca²⁺. The channel exhibited a sigmoidal increase in open probability with increasing pipette suction (364). Channels with similar characteristics were recorded in myocytes isolated from coronary artery (58), mesenteric arterioles (56, 275), and urinary bladder (362). Single-channel conductances ranged from 30 to 40 pS (58, 201, 362) for monovalent cations. In the presence of Ca²⁺, the channels exhibited slight inward rectification (201, 362) and reduced monovalent cation conductance (201), suggesting a Ca²⁺-dependent inactivation mechanism. The channels were blocked by Gd³⁺ (275, 362). Although the opening of a cation channel by membrane stretch would conceivably depolarize a cell and recruit voltage-gated Ca²⁺ channels (58, 201), its physiological role cannot be determined from single-channel measurements alone.

To better address the issue of physiological relevance, a method was developed for recording whole cell currents during VSM stretch (58). With the use of two to three modified patch pipettes, single myocytes could be stretched in the longitudinal direction up to 30% above the slack length of the cell. This stimulus consistently elicited an inward current, whereas, in current-clamp mode, single-cell stretch produced depolarization. Subsequently, stretch-activated, whole cell currents (Table 3) and/or depolarizations (Fig. 3B) were confirmed by Wellner and Isenberg (363, 364) and Setoguchi and co-workers (275, 317) in smooth muscle as well as in other types of muscle (163, 310) and nonmuscle cells (123, 397). In all three smooth muscle studies, the reversal potential for whole cell current (after excluding the contribution of secondary K⁺ current) was between 0 and 20 mV (58, 275, 363), varied with intracellular Na⁺ concentration ([Na⁺]_o) (317), and was not altered by changes in extracellular Cl concentration (58, 317); these characteristics are consistent with activation of a nonselective cation channel rather than a Cl or Ca²⁺ conductance. A component of the whole cell current was carried by Ca²⁺ (58, 317), but whole cell, MS cation currents could still be recorded in the presence of nicardipine to block VGC channels (317). Gadolinium blocked stretch-activated, whole cell current (275, 317, 363) and blocked stretch-induced depolarization (317). In two preparations, both single-channel and whole cell currents were shown to be Gd³⁺ sensitive (275, 363, 364). Whole cell MS cation currents were inhibited by increases in extracellular Ca²⁺ concentration ([Ca²⁺]_o) and enhanced by decreases in [Ca²⁺]_o, consistent with the modulatory effects of intracellular Ca²⁺ (consequent to Ca²⁺ influx) known to occur with other Ca²⁺-permeable channels (201, 362, 390). Calcium entry through the MS cation channel produced a significant and sustained rise in [Ca²⁺]_i and contraction (64). Calcium influx caused inactivation of VGC channels and activation of tetraethylammonium (TEA)-sensitive K⁺ channels (363). The interaction of this cation channel with other channels and signaling pathways in smooth muscle remains to be completely elucidated, but evidence suggests that its mechanosensitivity can be modulated by cAMP-dependent protein kinase (364). It is likely that this channel will be found to be regulated by other kinase systems as well [e.g., protein kinase C (PKC) regulates agonist-activated cation channels in gastric smooth muscle (197)].

Is activation of a MS cation channel necessary for initiation of the vascular myogenic response? This question has been difficult to answer because selective blockers are not available (for review, see Ref. 130). Dihydropyridines abolish myogenic tone, but do so by acting on VGC channels that are presumably downstream from cation channels in the signaling pathway (see sect. IVA6). Gadolinium, often thought to be a specific MS cation channel blocker (127), inhibits stretch-induced depolarizations and MS cation currents in isolated mesenteric artery myocytes (317) and eliminates myogenic tone in arterioles (401). However, Gd³⁺ also blocks VGC channels in some VSM cells at severalfold lower concentrations than required to block MS channels (25, 330), even though it may be more selective for MS channels over VGC channels in heart (131, 213). Aminoglycoside antibiotics such as streptomycin and neomycin have also been used to inhibit MS channels in other tissues (103, 378); these compounds block myogenic tone in rat cerebral arteries but only at doses higher than those required to block VGC channels (219, 246). Other purported MS channel blockers, such as Grammostola spatulata venom (270) and amiloride derivatives (130), have not been thoroughly tested on VSM channels, although amiloride and one of its analogs (at high doses) have been shown to inhibit myogenic tone (142). At this time, however, the lack of selective pharmacological tools to block MS cation channels has prevented determination of their potential role in the myogenic response.

As mentioned above, MS cation channels have been proposed to initiate contraction by depolarizing VSM cells past the threshold for activation of VGC channels (58, 201) and allowing Ca²⁺ entry through VGC channels to activate contractile proteins (241). Consistent with this idea is the observation that focal activation of MS cation channels (using pipette suction applied to a small membrane patch of a smooth muscle cell) elicits depolarization of the entire cell along with increases in [Ca²⁺]_i (118). Several other lines of evidence also support this hypothesis: 1) stretch elevates [Ca²⁺]_i in single vascular muscle cells (62); 2) pressure elevates VSM cell [Ca²⁺]_i in isolated arterioles (243); and 3) VGC channel antagonists produce only a partial block of stretch-induced [Ca²⁺]_i increases in VSM (62, 402), whereas Gd³⁺ produces a complete block (402).

Sodium substitution experiments have been used to test the role of stretch-activated Na⁺ entry mechanisms, but data from these experiments have produced confusing results. For example, in rat cerebral arteries, complete substitution of Tris⁺ for Na⁺ has no effect on the myogenic response (266), whereas in rabbit cerebral arteries, substitution of sucrose or N-methyl-D-glucamine for Na⁺ inhibits the response (142). Sodium ionophores increase myogenic tone (142), but increasing [Na⁺]_i with oubain does not (257). In rabbit facial vein, decreases in [Na⁺]_o potentiate (rather than attenuate) myogenic tone (142), possibly by changing the sensitivity of the contractile system to Ca²⁺ (141). Although the latter studies are consistent with a regulatory effect of [Na⁺]_o on an extracellular stretch sensor, as proposed by their authors (141), interpretation of all these studies is complicated by the likelihood that Na⁺ substitution has a profound effect on VSM ion transporters, such as the Na⁺/Ca²⁺ exchanger and the Na⁺-K⁺ pump (see sect. IVB) (273, 296).

Stretch-induced depolarization could result from inhibition of any of the various K⁺ currents identified in smooth muscle (212), provided the channel was active when the vessel had basal vascular tone. A direct role for a K⁺ channel in initiating the myogenic response has not been shown (see Table 2), but there is evidence that K⁺ currents can and do counteract myogenic tone. Of the five major types of K⁺ currents identified in VSM (267), three appear to play no significant role in the myogenic response: the inward rectifier K⁺ channel, the ATP-sensitive (K_ATP) K⁺ channel (294), and a novel K⁺ channel (K_N) with kinetics similar to M-type neuronal current (81). Indirect evidence at first suggested a role for K_ATP channels in the response of the coronary microcirculation to a fall in perfusion pressure (209), but recently a more direct study failed to show a significant effect of K_ATP channel antagonists on isolated small arteries at any pressure (205). However, there is evidence that the other two types of channels, voltage-dependent K⁺ (K_v) channels and Ca²⁺-activated K⁺ (K_Ca) channels, can provide potentially powerful repolarizing mechanisms to counteract stimuli resulting from VSM stretch. The K_v channels exhibit exponential increases in open probability upon depolarization and likely serve an important role in the repolarization of excitable cells (267). K_v channel blockers depolarize VSM cells in pressurized arterioles and augment myogenic tone (205). K_Ca channels are activated both by increases in [Ca²⁺]_i and by depolarization. The function of large-conductance K_Ca (BK) channels is particularly important to determine because activation of only a few channels would be sufficient to effect large changes in the membrane potential of a VSM cell (due to the high input resistance). For this reason, there is a substantial amount of information concerning the role of BK channels in the myogenic response.

It has been argued that stretch-induced depolarization could not be maintained unless an endogenous inhibitor of K_Ca channels is produced (137). This is because BK channels are activated by Ca²⁺ influx and by Ca²⁺ sparks [bursts of Ca²⁺ release from sarcoplasmic reticulum (SR)], producing pulses of outward current that substantially hyperpolarize the cell (265). Because myogenic tone is associated with both Ca²⁺ influx (9, 132, 147, 170, 215, 369) and Ca²⁺ release (62, 259), K_Ca current should be tonically activated when a blood vessel is at its normal pressure. Interplay between Ca²⁺-permeable MS channels and K_Ca channels has been demonstrated in other cell types (162, 321), and several lines of evidence support a similar interaction in VSM. For example, a K⁺ conductance in rat saphenous arteries is activated by pressurization, enhanced by Ca²⁺ ionophores, and blocked by TEA (an antagonist with moderate specificity for K_Ca channels) (20); in dog basilar artery, there is a tight coupling between stretch-induced increases in Ca²⁺ influx and ⁸⁶Rb efflux (9); in longitudinally stretched smooth muscle cells, a TEA-sensitive voltage-gated K⁺ current is activated secondary to activation of a MS cation current (363).

Further support for an important role of K_Ca channels comes from the observation that myogenic tone in cerebral arteries is enhanced by BK channel inhibition: at physiological levels of pressure, charybdotoxin (CTX), a specific inhibitor of BK channels, causes VSM depolarization and contraction, whereas at low pressures, it has little effect (34, 368). A depolarizing effect of CTX was also observed in pial arteries studied isometrically (108). In cerebral and renal arteries, Harder, Roman, and colleagues (136, 233, 398) have identified 20-hydroxyeicosatrienoic acid (20-HETE), a metabolite of arachidonic acid (AA) produced through cytochrome P-450 -hydroxylation, as an endogenous and possibly tonic inhibitor of BK channels. Stretch-induced production of 20-HETE [possibly through phospholipase (PL) C or PLA₂ pathways] could thereby sustain or even initiate myogenic responses. Of course, epoxyeicosatrienoic acids are also produced from oxidative metabolism of AA by cytochrome P-450 epoxygenase, and several of these products have been shown to activate, rather than inhibit, BK channels (165). In support of a role for 20-HETE, Wesselman et al. (368) found that BK channels were necessary for the pressure-induced response of rat mesenteric arteries treated with NE (368): inhibition of BK channels with CTX caused a flattening of the pressure-diameter relationship. The authors concluded that pressure may induce depolarization and myogenic contraction by closure of BK channels. The use of (reputedly) selective inhibitors has also confirmed a role for 20-HETE in some (368) but not all (40, 324) preparations. However, a ubiquitous role for 20-HETE and other related molecules depends (in part) on the demonstration that BK channels are tonically active in arteries and arterioles with normal tone: this is observed in some (34, 108, 265, 268), but not all, blood vessels (176, 229, 287, 393), although these differences may be due to the membrane potential-dependent characteristics of CTX block (96) rather than differences in the relative importance or expression of the BK channel.

Under the proper conditions, Cl channel activation is another potential mechanism to explain stretch-induced depolarization of VSM. Chloride channels have been implicated in agonist-induced depolarization of VSM (284). In smooth muscle, the estimated equilibrium potential for Cl (E_Cl) is somewhere between 47 and 10 mV (106), with the variation probably reflecting differences in the activity or expression of different Cl transport systems in different vessels. If E_Cl were more positive than the resting potential of the cell (3), opening of a Cl-selective channel would allow Cl efflux, producing depolarization. Possible candidates mediating this effect would be a Ca²⁺-activated Cl current (156, 221) and a volume-activated Cl current (387), both of which have been described in VSM.

In view of this, Nelson (264) has proposed that activation of Cl channels may explain stretch-induced depolarization of VSM. Support for this idea derives from the observation that Cl channel inhibitors (DIDS and indanyloxyacetic acid) hyperpolarize rat cerebral artery myocytes and inhibit myogenic tone of pressurized cerebral arteries (266). In addition, reduction of [Cl]_o from ~120 to 60 mM (which shifts the calculated E_Cl to 2 mV) enhances pressure-induced myogenic tone in cerebral arteries.

Although this idea is intriguing, a subsequent and more thorough study casts doubt on these conclusions. Doughty et al. (74) tested the effects of Cl channel blockers on rat cerebral arteries using patch-clamp techniques in combination with isobaric and isometric measurements of mechanical activity. The Cl channel blockers flufenamic acid and 9-anthracine chloride, which are fairly specific for Ca²⁺-activated Cl channels, had no effect on myogenic tone, even at high doses. Likewise, glibenclamide, an inhibitor of the cystic fibrosis transmembrane conductance regulator channel (as well as the K_ATP channel) was without effect on myogenic tone. 5-Nitro-2-(3-phenylpropylamino)benzoic acid, another Cl channel blocker, reversibly inhibited both myogenic tone and KCl-induced tone, but these effects were shown to be mediated by inhibition of VGC channels (74). At this time, the lack of specific blockers does not permit definitive conclusions to be made regarding the role of Cl channels in the myogenic response, but the existing evidence suggests they do not play an initiating role.

Voltage-gated Ca²⁺ channels have been recorded in many types of VSM, exhibiting characteristics of both L-type (15, 17, 101, 237, 269, 380) and T-type (15, 101, 230) channels. The L-type channel (also referred to as the VGC channel) is thought to be more important in arterial smooth muscle (267). In bath solutions containing physiological concentrations of Ca²⁺, both the activation threshold (50 to 60 mV) and peak current (10 mV) for the L-type Ca²⁺ channel occur at negative potentials (1). Because resting membrane potentials of VSM cells are in this range (132, 151, 262), a significant fraction of current must normally be activated at rest (98, 267, 305).

A large body of evidence now suggests that VGC channels play a central, obligatory role in determining myogenic responsiveness. 1) Voltage dependence of the L-type channel predicts that the 20- to 35-mV depolarization associated with VSM stretch would increase the open probability of the VGC channel by 10- to 15-fold (267). 2) Dihydropyridines eliminate or dramatically attenuate myogenic responsiveness in all (9, 132, 147, 170, 215, 369) but a few vessel types (159, 289) (the voltage dependence of dihydropyridine block may explain the discrepancies). 3) Dihydropyridines attenuate pressure- or stretch-induced [Ca²⁺]_i increases in isolated arterioles (401) and VSM cells (64). 4) Activators of VGC channels (e.g., BAY K 8644) enhance myogenic responses (83, 147, 202, 369). 5) Elevated levels of [Ca²⁺]_o enhance both myogenic responsiveness and the degree of pressure-induced depolarization (132). This evidence does not rule out an upstream role for other types of channels that could regulate VGC channel gating by depolarization, but it indicates that Ca²⁺ influx through VGC channels is at least a common step downstream in the signaling pathway.

It should be pointed out that myogenic tone in a few vessel types, notably rabbit facial vein and ear artery, does not exhibit the same dependence on VGC channel-mediated Ca²⁺ entry as determined for other vessels (23). This has led to the conclusion that a unique Ca²⁺ entry pathway is activated by stretch (24, 379). The specific arguments for this are based on comparisons of stretch-dependent tone with KCl- and agonist-induced tone (the latter two presumably act through VGC channels). In facial vein, stretch-dependent tone 1) has a different sensitivity to vasodilators, 2) has a different sensitivity to Ca²⁺ channel blockers (379), 3) is more susceptible to temperature changes (23), and 4) is more susceptible to experimental trauma (23, 60). The reasons for these differences are not known, but it is possible that some vessel types rely more extensively on Ca²⁺ influx through MS cation channels than through VGC channels (also, the studies cited above were performed under isometric conditions).

There are at least three ways in which VGC channels might participate in myogenic responses: 1) by opening when an upstream depolarizing stimulus brings the VGC channel to threshold (discussed in sect. IVA3), 2) by a shift in the activation or inactivation curve of the VGC channel to a voltage range more favorable for opening, and 3) by a direct effect of stretch on gating of the VGC channel.

With regard to the second mechanism, plots of open probability versus membrane potential for VGC channels show an activation threshold at approximately 50 mV and 90% inactivation at approximately 5 mV (values quoted for tracheal myocytes in bath solution containing 1.8 mM Ca²⁺) (87). The relationship between the activation and inactivation curves predicts a voltage window (with a peak around ~30 mV) in which Ca²⁺ current can be sustained under physiological conditions (267, 305). This is confirmed by simultaneous measurements of [Ca²⁺]_i and current in voltage-clamped cells showing an excellent correlation between Ca²⁺ entry through VGC channels and depolarization-induced [Ca²⁺]_i increases (102, 191). Shifting the activation curve to more negative potentials would lead to increased VGC channel activation at rest, whereas shifting the inactivation curve to more positive potentials would result in less Ca²⁺-induced inactivation and thus more sustained Ca²⁺ entry at any given potential. This effect is known to occur with some agonists and antagonists (1, 18) and may account for at least some of the potentiating action of -adrenergic agonists on the myogenic response (160, 242).

In addition to the above mechanisms, VGC channels might be directly modulated by stretch. Current flow through VGC channels is unlikely to account for stretch-induced depolarization because the depolarization persists in the presence of Ca²⁺ channel blockade (205, 317). Also, estimates of channel density, cell size, and degree of steady-state inactivation make it unlikely that VGC channels contribute more than 2-5 pA of steady-state inward current at 40 mV (246). However, L-type Ca²⁺ currents in rat cerebral artery myocytes, as recorded using the conventional whole cell mode, are enhanced by inflating cells through the patch pipette (218, 239) and are enhanced in the perforated-patch recording mode by hyposmotic cell swelling (218). Similar findings have been reported in rabbit cardiac myocytes (238) and gastric myocytes (386). These results suggest L-type channels may be directly gated by membrane distension, although an alternative explanation is that cell volume changes following inflation or swelling lead to alterations in the concentration of intracellular second messengers that modulate channel activity (e.g., cAMP; Ref. 228). However, stretch-induced changes in L-type current occur whether or not ATP and GTP are added to the patch pipette (218), when a peptide inhibitor of cAMP-dependent protein kinase is present (238), and when intracellular Ca²⁺ is chelated with high concentrations of EGTA or BAPTA [1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid] to minimize Ca²⁺-induced inactivation (238). Although the possible involvement of other second messenger systems [e.g., PKC (314), diacylglycerol (137, 352), and 20-HETE (136)] cannot be ruled out at this time, these experiments provide compelling evidence that VGC channels can be activated by VSM membrane stretch.

It remains to be determined if the degree of membrane stretch in inflation and swelling experiments corresponds to the force experienced by VSM cells in an intact vessel wall during physiological changes in pressure. In this regard, direct modulation of whole cell VGC channel currents either could not be detected (58) or was reduced (due to Ca²⁺-dependent inactivation) (363) in smooth muscle cells that were stretched longitudinally within apparent physiological limits. Also, neuronal VGC channels have been shown to be activated by flow (19), which would seem to be an unlikely physiological stimulus. It will be important to test if direct gating of VGC channels can be reproduced in single-channel recording modes and to determine the mechanosensitivity of the channel at that level; true MS channels typically change their open probability by three to four orders of magnitude following a mechanical stimulus (164), as opposed to more modest levels of mechanosensitivity that, under some circumstances, can be exhibited by agonist- and voltage-gated channels (285, 350).

Vascular smooth muscle plasma membranes contain a number of carrier-mediated ion exchangers and transporters (273) whose function could potentially be modulated by membrane stretch. Active transport systems include a Na⁺-K⁺-ATPase, a Ca²⁺-ATPase, and a K⁺-H⁺-ATPase; facilitated diffusion systems include Na⁺/Ca²⁺ and Na⁺/H⁺ exchangers in addition to Na⁺-Cl-HCO₃, Na⁺-K⁺-2Cl, and Cl-HCO₃ cotransporters (273). Modulation of electrogenic pumps could directly stimulate membrane depolarization, whereas modulation of electroneutral exchangers could alter ion gradients; either of these mechanisms would impact myogenic tone if they altered Ca²⁺ availability to the contractile system.

The most likely candidates for relevant MS transporters would be the plasmalemmal Ca²⁺-ATPase and the Na⁺/Ca²⁺ exchanger. The Ca²⁺-ATPase in VSM is regulated by calmodulin, by cGMP-dependent protein kinase, and possibly by phosphatidylinositol kinase (273), but no direct mechanical effects on this electrogenic pump have been reported. The Na⁺/Ca²⁺ exchanger is regulated by PKC and by cGMP-dependent protein kinase, but again no direct effect of membrane stretch on this protein is known (273). In fact, no evidence points to direct MS regulatory control of any of the other transport systems in smooth muscle, with the possible exception of the Na⁺-K⁺-ATPase.

In VSM, the Na⁺-K⁺-ATPase is regulated by increases in [Na⁺]_i, [K⁺]_o, and by other ions including Ca²⁺, Cd²⁺, and vanadate. Protein kinase C and PKA also modulate Na⁺-K⁺ pump activity in smooth muscle (273). In cardiac myocytes, cell swelling induced by hypotonic solutions is associated with a 66% increase in Na⁺-K⁺-ATPase current (311), a response that is not secondary to accumulation of cytosolic Na⁺, suggesting a direct mechanical effect. Although there is no electrophysiological evidence for MS Na⁺-K⁺ pump currents in smooth muscle, at least two studies (142, 259) have examined the effects of cardiac glycosides such as ouabain (a Na⁺-K⁺-ATPase inhibitor) on myogenic tone. In rabbit facial vein, ouabain potentiates both myogenic- and agonist-induced tone, and a similar effect was recorded in cerebral artery strips using other cardiac glycosides (259). These results do not point to a role for the Na⁺-K⁺ pump in mediating the myogenic response but are consistent with the idea that the depolarizing action of ouabain (28) serves to enhance Ca²⁺ entry.

It has long been appreciated that Ca²⁺ plays a pivotal role in smooth muscle contraction and the setting of arteriolar tone (345). Removal of extracellular Ca²⁺ from isolated arterioles causes rapid relaxation and passive behavior. This effect of Ca²⁺ is assumed to be mediated through Ca²⁺-calmodulin activation of myosin light chain kinase (MLCK) (Fig. 1). The following section examines the involvement of Ca²⁺ during myogenic vasoconstriction with an emphasis on sources of Ca²⁺, modulation of Ca²⁺ sensitivity, and temporal aspects of signaling.

The importance of Ca²⁺ in arteriolar tone was first established by Uchida and Bohr (345) some 30 years ago. In this classic study, skeletal muscle small arteries developed an inherent level of tone that was abolished by perfusion with a Ca²⁺-free solution. Dependence of single arterioles on an extracellular Ca²⁺ source for contraction and myogenic tone was first demonstrated by Duling et al. (75) in studies describing the isolated arteriole technique. Laher et al. (216) later demonstrated that myogenic tone shown by the rabbit facial vein was dependent on entry of Ca²⁺ as demonstrated by ⁴⁵Ca²⁺ influx. This technique, however, lacks the sensitivity necessary for its application to single arterioles.

Although it is evident that arterioles possess functional intracellular Ca²⁺ stores (releasable by agonists, caffeine, and ryanodine), studies indicate that relative to conduit vessels, arterioles have a greater dependence on extracellular Ca²⁺ for contractile activity (42, 170). On the basis of studies of isolated intact hamster cheek pouch arterioles, this does not appear to reflect a fundamental difference in the sensitivity of the contractile proteins for Ca²⁺ (170) but may relate to factors such as 1) smaller vessels having a relatively smaller volume of SR than larger vessels (10) or 2) differences in the Ca²⁺ influx/efflux rates (170). In apparent contrast, Boels et al. (30) in studies of permeabilized mesenteric vessels have suggested that Ca²⁺ sensitivity of the contractile proteins is greater in arterioles than in conduit vessels.

The advent of Ca²⁺-sensitive fluorescent dyes together with video-based imaging and photometer techniques have allowed the study of Ca²⁺ dynamics in true resistance vessels in a way that was not possible with radiolabeled tracer studies. Using isolated and cannulated skeletal muscle arterioles, Meininger et al. (243) first demonstrated that such approaches could be used to define arteriolar smooth muscle intracellular Ca²⁺ signaling during agonist and myogenic stimulation. Care was taken to exclude significant involvement of the endothelium by loading of the Ca²⁺-sensitive dye from the abluminal surface, focusing on the outer cell layers of the vessel and demonstrating similar results in the presence and absence of a functional endothelial layer. An acute pressure step equivalent to 40 cmH₂O resulted in an increase in [Ca²⁺]_i of ~15% above baseline. The initial increase in Ca²⁺ appeared to parallel the pressure-induced distension of the vessel. In addition to examining changes in [Ca²⁺]_i during myogenic constriction, responses were also examined after stimulation of the arterioles with either NE or the PKC activator indolactam. Despite these agents causing a similar level of constriction, the adrenergic response occurred in the presence of a large increase in [Ca²⁺]_i, whereas the PKC-mediated response occurred without a change in [Ca²⁺]_i. These data therefore not only demonstrated pressure-induced increases in arteriolar wall [Ca²⁺]_i but also suggested that arterioles, like conduit vessels, possess mechanisms for modulating Ca²⁺ sensitivity (see also sects. IVC1C and IVC4). These basic results have been confirmed in a number of subsequent studies (50, 206, 348, 401).

Given the biphasic nature of the mechanical response of an arteriole to an increase in intraluminal pressure, it is clearly important to consider temporal aspects of [Ca²⁺]_i signaling if the role of this cation is to be understood. The schematic diagram shown in Figure 6 depicts the temporal aspects of the diameter response of an arteriole to an acute increase in intraluminal pressure together with possible intracellular Ca²⁺ signals.

View larger version (17K): [in this window] [in a new window]   Fig. 6. Temporal pattern of intracellular Ca2+ concentration ([Ca2+]i) changes in arteriolar vascular smooth muscle in response to a step elevation in transmural pressure (for details, see text).

Figure 6 illustrates that an arteriole passively distends in response to an acute pressure increase, followed by a constriction to a steady-state diameter that is typically smaller than that before the pressure step. This mechanical response has been shown to be associated with either 1) a monophasic increase in [Ca²⁺]_i where Ca²⁺ peaks following distension and remains at that level for the duration of the pressure increase or 2) a biphasic increase in [Ca²⁺]_i where an initial peak is followed by a decline to a steady-state [Ca²⁺]_i level that remains elevated relative to baseline. Although both [Ca²⁺]_i patterns have been reported, there are several explanations for this apparent inconsistency. First, the magnitude of the initial peak in [Ca²⁺]_i appears to be related to the extent of the pressure-induced distension, possibly reflecting a change in cell length or wall tension. Thus larger pressure steps may amplify the appearance of a biphasic pattern. The magnitude of the change in [Ca²⁺]_i associated with a pressure increase is small relative to that seen with agonists, so the study of [Ca²⁺]_i responses to small pressure steps is therefore more dependent on the sensitivity of the measurement techniques. As such, it may be difficult to resolve a biphasic [Ca²⁺]_i change in arterioles exposed to relatively small changes in pressure. A further consideration is that as the vessel constricts, reflecting shortening of the smooth muscle cells, the stimulus for Ca²⁺ mobilization presumably decreases; a biphasic [Ca²⁺]_i pattern might, therefore, be expected. However, it could be argued that this would be predicted by any model of Ca²⁺ availability and is not necessarily an indication of temporal variation in the contribution of Ca²⁺ pools or the participation of alternate regulatory mechanisms/Ca²⁺ sensitization in the steady state. An additional explanation for the biphasic change in [Ca²⁺]_i relates to the possibility that the initial increase in [Ca²⁺]_i activates an inhibitory process aimed at dampening the[Ca²⁺]_i rise and hence vasoconstriction (265). Again, such a process may be expected to be more evident following large pressure steps that are associated with relatively larger [Ca²⁺]_i peaks.

The above discussion has not considered whether the distension-induced [Ca²⁺]_i peak and the steady-state [Ca²⁺]_i level are related or the relative roles of these phases in the contractile response. In an effort to determine if the initial increase in [Ca²⁺]_i was necessary to elicit steady-state myogenic contraction, the responses of isolated arterioles to 30- to 120-mmHg pressure increases were compared when the pressure change was delivered either instantaneously or as a ramp function over 5 min (148). During the latter protocol, the rapid pressure-induced distension and the associated transient increase in [Ca²⁺]_i is avoided. Despite this, the steady-state diameter achieved is similar under both protocols (64), suggesting that the initial peak in [Ca²⁺]_i is not an absolute requirement for effective myogenic constriction. Similarly, D'Angelo et al. (50), in a study of isolated hamster cheek pouch arterioles, demonstrated that although the degree of distension was related to the peak change in [Ca²⁺]_i, steady-state [Ca²⁺]_i levels were similar regardless of the size of the applied pressure step. Steady-state constriction was greater, however, in vessels exposed to larger pressure steps. It was suggested that an excess of Ca²⁺ (relative to a required threshold level) was mobilized during the initial phase and that processes of Ca²⁺ sensitization were activated during the maintained or steady-state phase. With the consideration of both sets of data, however, it could be argued that more than one event occurs, for example, a purely mechanical or stretch-mediated response that occurs with distension and a second phase related to a variable other than overt cell length, such as wall tension. Interestingly, when cannulated skeletal muscle arterioles were subjected to acute longitudinal stretch, the vessels responded with a rapid increase in [Ca²⁺]_i which then returned to baseline levels despite maintenance of the stretch stimulus (S. Potocnik, M. J. Davis, H. Zou, S. Price, and M. A. Hill, unpublished observations).

A) INTERRELATIONSHIPS BETWEEN CA²⁺ SOURCES. The involvement and relative roles of specific Ca²⁺ sources in the myogenic response still remain uncertain. Although it is clear that there is a major dependency on extracellular Ca²⁺, questions remain as to the specific entry mechanisms (see sect. IVA6) and the involvement of release from compartments such as the SR. It is apparent that arteriolar smooth muscle possesses sarcoplasmic Ca²⁺ stores released by activation of either inositol trisphosphate (IP₃) or ryanodine receptors, with the latter being involved in Ca²⁺-induced Ca²⁺ release (192, 265). The involvement of IP₃-mediated Ca²⁺ release from the SR is supported by studies demonstrating accumulation of inositol phosphates following length or pressure changes (143, 261, 391). These changes are not necessarily associated with myogenic behavior (391). Furthermore, when briefly exposed to 0 mM extracellular Ca²⁺ solutions, arterioles respond to agonist stimulation with transient contractions consistent with the presence of a releasable intracellular Ca²⁺ store. This store appears to be more rapidly depleted than in larger vessels