INTRODUCTION

TOP ABSTRACT INTRODUCTION PHENOTYPIC PLASTICITY AND... REGULATION OF CONTRACTILE... MECHANICAL PLASTICITY OF SMOOTH... SPECULATION ON THE MOLECULAR... FUTURE DIRECTIONS REFERENCES

MATURE SMOOTH MUSCLE CELLS (SMC) are distinct among myogenic lineages in that they retain a multifunctional capacity for contraction, migration, proliferation, synthesis of extracellular matrix (ECM) components, and secretion of growth factors and cytokines. These characteristics uniquely equip SMCs with the potential to regulate lumen diameter of hollow organs both transiently, via reversible contraction, and chronically, via structural remodeling due to fibrosis and muscle hypertrophy. Studies of arteries after injury and of primary cultured SMCs from various tissues have revealed that myocytes dynamically exhibit distinct contractile and synthetic phenotypes with unique morphological, biochemical, functional, and gene expression characteristics (16, 43, 46, 111, 119, 161). Furthermore, plasticity of the length-tension properties and organization of contractile filaments in smooth muscle has recently been described, which further extends functional diversity of SMCs and smooth muscle-containing tissues (41, 42, 49, 104, 117). Collectively, these observations indicate that the full spectrum of SMC function, dictated by phenotypic and mechanical plasticity, is controlled by molecular mechanisms that regulate processes as diverse as gene transcription and the transient remodeling of actin and myosin filaments. In this review, we summarize some of the themes emerging from recent studies in these areas and propose future research directions that may lead to a better understanding of the integrated control of smooth muscle plasticity.

	PHENOTYPIC PLASTICITY AND HETEROGENEITY OF SMC

TOP ABSTRACT INTRODUCTION PHENOTYPIC PLASTICITY AND... REGULATION OF CONTRACTILE... MECHANICAL PLASTICITY OF SMOOTH... SPECULATION ON THE MOLECULAR... FUTURE DIRECTIONS REFERENCES

Phenotypic plasticity of differentiated contractile SMC was first recognized using primary cultured myocytes derived from the medial layer of large elastic arteries (16). When seeded in serum-enriched culture media, mature myocytes acquire an "immature" phenotype that contains abundant organelles for protein and lipid synthesis and numerous mitochondria. They exhibit a high proliferative index but lose typical in vivo pharmacological responsiveness and lack contractile myofilaments and their associated proteins (16, 92). The phenotype switching from a contractile to a synthetic phenotype is defined as modulation (16) and has been reported as being a characteristic response of mature SMCs derived from all vascular and visceral organs studied (46, 51, 68, 118, 119). The reversion of primary cultured SMCs to the contractile state also occurs and is termed maturation. In culture, maturation occurs as cultures grow to confluence and/or as a result of withdrawal of serum and other mitogens (43, 45). Phenotypic maturation of cultured SMCs is marked by increased myofilament and contractile apparatus-associated protein content, reacquisition of typical pharmacological responsiveness, and decreased abundance of synthetic organelles (16, 43, 46, 51, 52, 92). Recent studies using serum-free culture conditions describe the induction of a distinct subset of airway SMCs to a functionally contractile phenotype with elongated morphology, fully reconstituted contractile apparatus, abundant contractile protein content, and cell surface recoupling of muscarinic M₃ receptors for acetylcholine. (43, 76, 92). Similar reconstitution of contractile phenotype arterial myocytes in culture has also been reported (73). Collectively, these observations illustrate that phenotypic plasticity is a feature of cells committed to the smooth muscle lineage and is manifest as reversible modulation and maturation of individual myocytes (45) (Fig. 1).

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Schematic representation showing the association of phenotypic and mechanical plasticity on smooth muscle. Phenotypic plasticity results from reversible modulation and maturation of smooth muscle cells (SMC) between a synthetic and contractile state. Mechanical plasticity occurs in contractile myocytes as the result of time-dependent subcellular reorganization of the contractile apparatus in response to changes in muscle length.

There are numerous reports describing functional and molecular heterogeneity of SMCs in vivo and in vitro (2, 10, 30, 31, 44, 45, 68). Phenotypic plasticity, in combination with a spatially and temporally diverse developmental and differentiation repertoire, appears to produce divergent SMC populations both within and between smooth muscle-associated organs (30, 45, 72, 91, 102, 130, 141). Broad diversity exists in ultrastructural, pharmacological, electrophysiological, and contractile properties of mature SMCs from different tissues and between spatially segregated segments of the same tissue (14, 23, 30, 43, 46, 47, 68, 75, 74). The molecular and biochemical factors that underlie heterogeneity between different smooth muscles include differential expression of receptor and ion channel proteins and/or contractile apparatus proteins and those associated with regulation of contraction (19, 38, 46, 66, 101, 121, 148, 165). Individual SMCs exhibit striking heterogeneity that is marked by dissimilarities in morphology, contractile function, electrophysiological properties, proliferative responsiveness, and expression of cytoskeleton, contractile apparatus, and ECM proteins (2, 4, 30, 31, 43-45, 68, 70, 71, 75, 82, 105, 158, 159). Furthermore, distinct "clonal" subpopulations of SMCs have been identified, each retaining its own characteristic capacity for modulating functional, biochemical, and gene-expression phenomena (10, 30, 31, 93, 162).

Phenotypic heterogeneity appears to be an important functional determinant of the response of smooth muscle in normal physiology and pathophysiology. Heterogeneity in shortening velocity of SMCs dissociated from the airways (75) or from systemic arterial beds (26, 84) has been reported and appears to be related to differences in expression of smooth muscle myosin heavy chain (smMHC) and/or myosin light-chain kinase. The presence of diverse electrophysiological properties based on differences in K⁺ channel distribution among myocytes from large and small pulmonary and systemic vascular beds has also been described (2, 80, 105, 158) and appears to account for differential hypoxia-induced contractile responses (2, 105, 158). On the basis of immunohistochemical analysis of muscle-specific contractile and cytoskeletal markers, at least four developmentally divergent subpopulations of pulmonary arterial SMCs have been described (30). These subpopulations are spatially segregated into different compartments of the vessel wall and exhibit site-specific diversity in proliferation and ECM protein synthesis in response to hypoxia in vivo and other mitogens in vitro (24, 162). Distinct subpopulations of human, mammalian, and rodent SMCs from different regions of the arterial wall have now been isolated and cultured (6, 10, 27, 30, 31, 59, 90, 96, 149). Dramatic stable differences in morphology, in the proliferative response to various mitogens, and in expression of ECM proteins and cytoskeletal markers have been described between isolated arterial subpopulations (30, 95, 141). Several studies also show evidence that transcriptional activity of promoters for smooth muscle-specific genes is differentially activated in isolated vascular SMC cultures and between different myocytes within the same vessel in vivo (29, 62, 64, 81, 89, 118). Similarly, individual SMCs from nonvascular sources have also been shown to be segregated on the basis of size, contractile protein content, nuclear ploidy, and proliferative potential (44, 47, 68).

SMC plasticity and heterogeneity have best been described in vivo during neointimal thickening and fibrosis associated with the pathogenesis of atherosclerotic plaques and postangioplasty restenosis. Neointimal thickening results from the abnormal accumulation of a myofibroblast-like synthetic phenotype. These synthetic SMCs may be derived from medial myocytes that undergo phenotypic modulation from a contractile state (15, 16), or they may arise from clonal expansion of a subpopulation of "immature" myocytes present in the wall of the blood vessel (59, 83, 93). Both views are well supported by multiple studies, and they do not appear to be mutually exclusive (166). In asthma, there is similar fibrosis and accumulation of synthetic phenotype myocytes in the submucosal region of the bronchial wall, and there is significant accumulation of smooth muscle mass through hypertrophy and hyperplasia of individual myocytes (25, 37, 58). Evidence for SMC heterogeneity and plasticity and its association with differences in smooth muscle function in normal physiology and disease clearly indicates the importance of understanding the molecular mechanisms that regulate transcriptional and posttranscriptional regulation of phenotype-specific gene expression.

	REGULATION OF CONTRACTILE PHENOTYPE MARKER EXPRESSION

TOP ABSTRACT INTRODUCTION PHENOTYPIC PLASTICITY AND... REGULATION OF CONTRACTILE... MECHANICAL PLASTICITY OF SMOOTH... SPECULATION ON THE MOLECULAR... FUTURE DIRECTIONS REFERENCES

Phenotypic plasticity of SMCs requires the differential expression of a repertoire of phenotype-specific genes and subsequent accumulation of the proteins that they encode. Recent studies have begun to reveal information about the nature of the extracellular cues, signaling pathways, and transcription factors that regulate smooth muscle-specific gene expression. The plasticity of SMC phenotype expression in primary culture has been exploited to identify phenotype-specific protein markers and gene expression patterns (13, 43, 44, 46, 47, 51, 52, 118, 119). A detailed delineation of the wide range of receptor, structural, contractile, and signaling protein markers that are expressed in a phenotype-specific manner in smooth muscle is beyond the scope of this review and has been reported previously (45, 102, 111). Some of the best-characterized markers for SMC of the contractile phenotype include smooth muscle -actin (34), smooth muscle -actin (112), smMHC (94), calponin (36), h-caldesmon (32), SM22 (36, 126), smoothelin (147), metavinculin (7), and ₁-integrin (100). We will focus on mechanisms that regulate genes encoding proteins that are clearly abundant and associated with the contractile function of mature myocytes.

It is well established that phenotypic expression is regulated by a complex array of environmental cues that include cytokines, ECM components, and mechanical stimuli (16, 51-53, 56, 119, 122, 123, 141). How such diverse signals are coordinated to effect specific changes in transcription of smooth muscle-specific genes is an area of intense current research interest. Several investigators have reported that ECM components of the basement membrane, chiefly laminin and collagen type IV, induce a delay in the spontaneous phenotypic modulation and proliferative index of contractile SMCs from a variety of tissues when seeded in primary culture (51, 52, 56, 139). Shuger's group (106, 115) has also shown that laminin ₁- and ₂-chains are required for normal morphogenesis and elongation of airway myocytes in the developing lung. Hayashi et al. (51) reported that prolonged maintenance of the contractile phenotype of chicken gizzard myocytes seeded on laminin matrix in serum-free conditions only occurred if culture medium was also supplemented with insulin-like growth factor (IGF)-1. Under these conditions IGF-1 activated phosphatidylinositide 3-kinase and was required for maintenance of the differentiated phenotype (52). Conversely, spontaneous phenotypic modulation and proliferative responses of primary cultured myocytes appear to be enhanced by seeding on fibronectin (51-53, 56, 139).

A number of the smooth muscle-specific genes encoding contractile apparatus-associated proteins have been cloned, and the functional nature of their promoter sequences has been characterized (8, 61, 126, 136, 160). Comparison of the 5'-flanking DNA promoter regions from smooth muscle-specific genes of different species, including murine SM22 and calponin, rat smooth muscle -actin and smMHC, chicken caldesmon, and rabbit telokin, has revealed a number of shared cis-acting elements known to bind nuclear transcription factors present in SMC (125). Some of the common motifs shared among all of the promoters included pairs of CArG box elements [CC(A/T)₆GG], which bind serum response factor (SRF), and a single transforming growth factor- control element ("TCE") located near the more 3' CArG box. These sites appear to be critical in conferring SMC-specific promoter activity, as mutation of these sites in the SM22 and smooth muscle -actin genes abolished transcription in transiently transfected cultured myocytes and transgenic mice (1, 64, 120, 126). Other binding motifs in the "stereotypical" smooth muscle contractile protein promoter include Sp1 and AP2 sites, as well as potential YY1 and Mhox binding elements. Mhox is a homeodomain transcription factor that is part of the homeobox gene family that specifies spatiotemporal gene expression patterns during development. Interestingly, in mature contractile SMCs, several homeobox genes are expressed at high levels, including Mhox, Gax, HoxA5, HoxA11, and HoxB1 (9, 22, 88). Moreover, in a recent study, angiotensin II-induced transcription of smooth muscle -actin in cultured vascular myocytes was partially mediated by an interaction between Mhox and the two CArG boxes resident in the promoter sequence (50). Nonetheless, a SMC-specific homeodomain transcription factor has not yet been identified.

No single transcription factor has been identified that can clearly determine SMC-specific gene expression. However, two of the binding motifs found in the "stereotypical" smooth muscle contractile gene promoter (125) have recently been investigated more thoroughly, and a critical role in transcriptional activation and phenotype-specific gene expression has been confirmed (1, 13, 78, 79). These two cis-acting elements include the pair of CArG boxes that bind SRF and the single TCE that is now known to bind both gut-enriched Kruppel-like factor (GKLF) and a related factor, BTEB2 (1). Both SRF and BTEB2 trans-activate the SM22 and smooth muscle -actin promoters, whereas GKLF appears to repress transcription (1). These results suggest that these transcription factors might coordinately regulate contractile gene expression in a phenotype-specific manner; however, this is speculative conjecture requiring experimental validation.

To date, the signaling pathways that control SRF-mediated transcriptional activation of smooth muscle-specific genes have been the most rigorously investigated. SRF is a 67-kDa protein of the MCM1-agamous-deficiens-SRF (MADS) family that was initially described for its role in serum-induced activation of the c-fos promoter (142). The c-fos promoter contains a CArG box (common to smooth muscle contractile genes) within its serum response element and which SRF binds as a dimer. The serum response element within the c-fos promoter also contains an adjacent Ets-binding site (C/AGGAA/T) that is important for formation of a stable complex between SRF dimers and members of the ternary complex factor family (e.g., Elk-1, SAP-1) (12). Importantly, no Ets-binding site exists in any smooth muscle contractile gene promoter (125). Substantial SRF expression has been localized to cells of myogenic lineages (5, 21) and binding of SRF to several skeletal, cardiac, and smooth muscle genes is essential for complete promoter activation and cell differentiation (120, 129, 131, 138). Some investigators have reported a primary regulatory role for the monomeric small GTP-binding protein, RhoA, in serum-activated c-fos transcription and in SRF-dependent transcription of skeletal myocyte genes (12, 142). Moreover, RhoA, which has a well-established effect on actin filament dynamics and bundling in cells that form stress fibers and contract (48, 107-109), was recently shown to regulate SRF-mediated transcription in both NIH 3T3 fibroblasts and cultured vascular SMC (78, 128). Positive regulation of both the SM22 and smooth muscle -actin promoter in vascular myocytes was mediated via increased actin polymerization. Importantly, recent studies from our own group have revealed that RhoA-mediated activation of smooth muscle gene transcription results, in part, through regulation of the cytoplasm vs. nuclear distribution of SRF (Ref. 13 and unpublished observations). The downstream signaling intermediate of RhoA for SRF-induced c-fos, SM22, and smooth muscle -actin transcriptional activation appears to be the serine/threonine kinase, ROCK-1 (Fig. 2) (Ref. 78 and unpublished observations). These recent results are important, as they are the first to elucidate a direct role for actin cytoskeletal dynamics in smooth muscle-specific gene expression.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Proposed model for integrated regulation of phenotypic and mechanical plasticity of smooth muscle via RhoA-mediated regulation of actin and myosin filament polymerization and stress fiber assembly. Activation of SMC-specific gene transcription is accomplished by binding serum response factor (SRF) dimers to CArG boxes in their 5' promoters. SRF is activated, and cytoplasm-to-nuclear translocation is induced both by ROCK (also known as Rho kinase) and by changes in stress fiber dynamics. ROCK also inhibits myosin light-chain (MLC) phosphatase, thereby indirectly leading to increased phosphorylation of myosin monomers by myosin light-chain kinase (MLCK), which leads to myosin filament polymerization. RhoA also activates mammalian Dia1/profilin, which triggers actin polymerization and the coordinated bundling of cables of contractile apparatus that underlies mechanical plasticity and regulates SMC-specific gene expression.

	MECHANICAL PLASTICITY OF SMOOTH MUSCLE

TOP ABSTRACT INTRODUCTION PHENOTYPIC PLASTICITY AND... REGULATION OF CONTRACTILE... MECHANICAL PLASTICITY OF SMOOTH... SPECULATION ON THE MOLECULAR... FUTURE DIRECTIONS REFERENCES

Beyond the broad functional potential that phenotypic plasticity affords smooth muscle, the contractile function of tissue and cells is strikingly adaptable and sensitive to both transient and chronic changes in length and load (17, 41, 49, 104, 117, 151, 152). Mechanical plasticity is rooted in seminal observations that the length-tension relationship of trachealis is dynamic, with the muscle being able to recover its ability to generate tension in a time-dependent manner after large static or oscillatory length changes (41, 104, 117). Recently, a rabbit carotid arterial preparation was reported to show the same plasticity phenomenon, albeit with significantly reduced capacity (116). Therefore, an in vitro length-tension curve for smooth muscle cannot be represented by a single curve nor can optimal length be determined precisely, since the function that describes the relationship changes depending on the initial length and contraction history of the muscle (35, 39, 41, 104, 156, 163). Numerous studies have demonstrated that maximal force generated by smooth muscle is nearly independent of length (40, 54, 133). This is a direct reflection of the unique mechanical plasticity inherent to smooth muscle. Empirically, mechanical plasticity appears to be a necessary functional characteristic for smooth muscles in vivo because the mechanical forces on muscle surrounding hollow organs, such as the airways and arteries, are chronically and variably oscillated through breathing, cardiac pumping, or both.

In contrast to smooth muscle, tension produced by skeletal muscle does not adapt to large changes in length; therefore, comparison of the length-tension properties of these two tissues provides some potential insight to understanding the subcellular mechanisms that determine mechanical plasticity in smooth muscle. Importantly, although qualitative similarities between the length-tension relationship of skeletal and smooth muscle support the notion that contraction in smooth muscle follows a sliding-filament/cross-bridge model, dissimilarities between the tissues point to possible mechanisms for mechanical plasticity (42, 132). The smooth muscle length-tension curve is much broader than that for skeletal muscle. The shape of the skeletal muscle length-tension curve can be accounted for on the basis of filament overlap; however, significant differences in contractile apparatus structure of smooth muscle complicates interpretation of its length-tension curve strictly on the basis of filament overlap. First, a highly ordered arrangement of multiple sarcomeric structures does not exist in smooth muscle. Second, in smooth muscle, cross bridges maintain the same polarity along the entire myosin thick filaments with bare zones at each end, whereas thick filaments in skeletal muscle are bipolar with a central bare zone (20, 164). Consequently, at short lengths, actin filaments are unlikely to abut and overlap a bare zone as they do in skeletal muscle. Third, unlike skeletal muscle, phosphorylation of the smooth muscle regulatory 20-kDa myosin light chain (MLC₂₀) is length dependent and is obligatory to initiate myosin-actin binding, cross-bridge cycling, and force generation (87, 135, 154). This fact further supports the notion that factors other than filament overlap regulate force generation in smooth muscle.

It is important to note that phosphorylation of MLC₂₀ does not appear to be affected by sudden changes in airway smooth muscle length that are large enough to effect acute reduced force generation (87). This means that transient loss of force-generating potential after a length change is not likely the result of deactivation of the contractile apparatus. Rather, it has been hypothesized that the loss of force generation after an abrupt change in length is due to disruption of the organization of the contractile filaments (104); thereafter, the recovery to a steady state, in which the muscle again generates maximal force, is dependent on plastic reorganization of the contractile apparatus as the muscle adapts to the newly imposed length (104, 117). This phenomenon has been demonstrated by in vitro studies that revealed force generated by trachealis strips between 0.5 and 1.5 of a reference length generated the same active force but muscles at longer lengths shortened at a significantly faster velocity (104). On the basis of these observations it is reasonable to speculate that, when a smooth muscle is stretched and held at a new longer length, over time there is mechanical plasticity due to contractile apparatus rearrangement such that more contractile units are placed in series. This is shown in Fig. 1. In contrast to the effects of stretching a muscle, imposing a shorter length thus leads to length adaptation and restructuring of the contractile apparatus through deletion of contractile units in series to a new steady state that can generate maximum force but at the expense of shortening velocity (Fig. 1). The mechanisms that regulate mechanical adaptation to new lengths are not clear, but there are several lines of evidence that suggest that plastic remodeling of the contractile apparatus and its association with the cytoskeleton may be responsible. First, mechanical adaptation to different muscle lengths is accelerated if the muscle is periodically stimulated to contract in the time immediately after the length change (104, 116, 117, 151). Recently, Seow et al. (117) provided mechanical evidence and a theoretical basis for thick filament evanescence during contraction and relaxation. They quantitatively confirmed that the decrease in velocity that occurs in the late phase of a single contraction could be completely ascribed to myosin filament lengthening during activation, resulting in a series-to-parallel filament transition without any need to postulate a change in cross-bridge cycling rate. These data suggest that repeated activation of a smooth muscle accelerates mechanical adaptation to a new length by inducing filament remodeling during each contraction-relaxation cycle. Anecdotal biochemical evidence that supports the filament evanescence hypothesis comes from studies in which agonist-induced contraction of smooth muscle significantly induced recruitment of globular actin to thin filament stores (85). Furthermore, phosphorylation of MLC₂₀ during active contraction may positively regulate assembly of thick filaments (163). Also, Gunst and colleagues (86, 137, 153) have reported that contractile activation of airway smooth muscle leads to RhoA-independent phosphorylation of the integrin-associated cytoskeletal proteins, focal adhesion kinase (FAK), and paxillin that may have some potential for modifying the association of actin filaments with focal membrane integrin complexes. These biochemical observations support the notion that repeated contractile activation of smooth muscle induces intracellular signaling pathways having protein targets that comprise specific domains of the cytoskeleton and/or contractile apparatus. Hypothetical molecular mechanisms that could regulate and integrate mechanical plasticity and phenotypic plasticity of smooth muscle are presented in the next section.

	SPECULATION ON THE MOLECULAR MECHANISMS OF MECHANICAL PLASTICITY

TOP ABSTRACT INTRODUCTION PHENOTYPIC PLASTICITY AND... REGULATION OF CONTRACTILE... MECHANICAL PLASTICITY OF SMOOTH... SPECULATION ON THE MOLECULAR... FUTURE DIRECTIONS REFERENCES

Recent advances in cell and molecular biology from a number of fields have revealed the cytoskeleton to be a complex and dynamic network that influences, and is modulated by, gene transcription, cell signaling, cell motility and spreading, protein trafficking and secretion, and cell division. The studies presented in this review clearly indicate a salient role for cytoskeletal dynamics as a fundamental mechanism that integrates regulation of both phenotypic and mechanical plasticity of smooth muscle. Recent and previously published evidence suggest a role for regulation of actin and coincident myosin filament polymerization and bundling by Rho GTP-binding proteins (48, 60, 108, 109, 157). We will focus on the potential role of RhoA and its downstream signaling intermediates as a principal candidate for integrating cytoskeleton-associated molecular events that mediate phenotypic and mechanical plasticity in smooth muscle. This paradigm is summarized schematically in Fig. 2.

A critical role for RhoA in positive regulation of actin stress fiber assembly, focal adhesion formation, and increased actomyosin-based motility was first established in Swiss 3T3 fibroblasts, and RhoA has since been shown to be ubiquitously expressed in vitro and in vivo (18, 48, 67, 97, 146). Actin stress fibers induced by RhoA are composed of actin and myosin-II filaments that are bundled together as cables that generally span the longitudinal axis of most cells, including SMC and fibroblasts, in culture (33, 43, 99). These structures contain many of the proteins associated with the contractile apparatus of smooth muscle, including smMHC, smooth muscle -actin, SM22, caldesmon, tropomyosin, and calponin (43, 57, 69, 77, 98, 150).

Numerous downstream targets of RhoA have been identified; these are activated through and interact with the active GTP-bound form of the GTPase (3). One such target is the serine/threonine kinase ROCK, also known as Rho kinase or ROK, which exists as two isoforms (ROCK-1 and ROCK-2). ROCK has been shown to phosphorylate and inhibit the myosin-binding subunit of smooth muscle myosin light-chain phosphatase, which ultimately leads to a Ca²⁺-independent increase in MLC₂₀ phosphorylation by disrupting the balance of myosin light-chain kinase to myosin light-chain phosphatase activity in favor of the kinase (Fig. 2) (65, 127, 146). Inhibition of RhoA activity with Clostridial C3 toxin completely blocked ROCK-induced Ca²⁺-independent MLC₂₀ phosphorylation, confirming that ROCK is activated by RhoA during smooth muscle contraction. Moreover, beyond the implications these observations have on the Ca²⁺ sensitivity of contraction, the capacity for ROCK to induce MLC₂₀ phosphorylation during a contraction has strong potential as one of the molecular mechanisms that regulate myosin filament polymerization and mechanical plasticity in smooth muscle. This contention is supported by compelling evidence that the assembly of smooth or nonmuscle myosin monomers into myosin filaments is positively regulated via MLC₂₀ phosphorylation in vitro (63, 114, 124, 143, 144). Disassembly and reassembly of myosin filaments in SMC have been known to occur for some time (28). From electron microscopy studies, there is also evidence in some muscles for a resident pool of monomeric myosin that is recruited to thick filaments during active contraction when MLC₂₀ is phosphorylated, thus resulting in an increased density of filaments (35, 39, 163). There is also strong evidence that constitutively active RhoA and ROCK mutants stimulate stress fiber and myosin filament formation (3, 146, 157). Importantly, there are reports that cholinergic contractile agonists induce RhoA/ROCK-mediated formation of myosin-containing stress fibers in Chinese hamster ovary cells transfected with human M₃ muscarinic receptor (134) and stress fibers in cultured human airway SMCs (55, 140). These observations indicate that RhoA/ROCK signaling has strong potential as a primary pathway for regulation of contractile apparatus reorganization in mechanical plasticity of SMCs (Fig. 2).

Although overexpression of constitutively active ROCK alone does induce stress fibers in cultured fibroblasts, the thickness, number, and distribution of the fibers is unlike that induced by active RhoA (110, 157). This suggests that ROCK is not the only downstream signaling protein required for "normal" stress fiber formation. A recent paper indicates that mammalian Dia1, a member of a family of proteins that contains formin-homology regions (FH1 and FH2), which interact with the actin-monomer-binding protein profilin, may be the only other Rho-activated intermediate required for the formation of fully functional stress fibers (107, 157). Dia1 has been proposed to localize profilin where RhoA is active. Profilin is a ubiquitous protein that promotes actin polymerization by targeting ATP-bound actin monomers to sites of actin assembly (155, 167). Truncated versions of Dia1 that contain only profilin-binding FH1 and FH2 domains are not compromised in their ability to induce stress fibers, confirming that RhoA/Dia1/profilin mediates polymerization of actin filaments that are subsequently bundled with myosin filaments to make intact stress fibers (157). We propose that similar pathways could be at work in smooth muscle to effect mechanical plasticity (Fig. 2). Actin polymerization is no doubt dynamic in steady-state SMC, as it is in other cells, with continual "treadmilling" of filaments in equilibrium with monomeric pools of G actin. Phosphoinositide- and Ca²⁺-dependent capping proteins such as gelsolin, tropomodulin and CapZ, which regulate the elongation and generation of actin filaments, along with G-actin-binding proteins like profilin are all expressed in smooth muscle (11, 103, 113). Significant recruitment of monomeric actin into filaments has been demonstrated in tracheal smooth muscle in which G actin was reduced by at least 30% following contractile stimulation (85). In addition, as noted above, actin stress fiber formation is greatly enhanced in cultured human airway myocytes after cholinergic agonist exposure (55). Smith and colleagues (122, 123) have shown that cyclic stretch of cultured airway SMCs greatly increases the number of actin stress fibers and focal adhesion complexes to which they are anchored, suggesting that chronic oscillation or contraction of myocytes leads to actin polymerization. Lastly, inhibition of actin polymerization with latruculin or cytochalasin D significantly decreases force generated by trachea smooth muscle and greatly flattened the length-tension curve, indicating that, in the absence of actin polymerization, optimal structural organization of the contractile apparatus cannot be maintained at any muscle length (85, 145).

Collectively, these data suggest that the molecular mechanisms that determine contractile apparatus organization could act in an integrated manner to regulate both phenotypic and mechanical plasticity in both a direct and indirect manner (Fig. 2). Both ROCK and mammalian Dia1/profilin appear to coordinate RhoA-induced stress fiber formation by targeting myosin and actin polymerization, respectively. We previously noted several reports that indicated that RhoA-induced stress fiber assembly positively regulates transcription activity of CArG box-containing genes (13, 78). Clearly, this highlights the potential for RhoA both to direct the differential expression of genes encoding smMHC, smooth muscle -actin, SM22, calponin, and caldesmon, which is associated with phenotypic plasticity, and to regulate the participation of these protein products in cytoskeletal remodeling associated with mechanical plasticity of smooth muscle.

	FUTURE DIRECTIONS

TOP ABSTRACT INTRODUCTION PHENOTYPIC PLASTICITY AND... REGULATION OF CONTRACTILE... MECHANICAL PLASTICITY OF SMOOTH... SPECULATION ON THE MOLECULAR... FUTURE DIRECTIONS REFERENCES

The contractile properties of smooth muscle are a fundamental determinant of the function of hollow organs, and alterations in contractile function have been linked to several diseases, including asthma and vasospastic sensitivity near the site of atherosclerotic lesions. In addition, the multifunctional potential of SMCs provides them with the capacity to effect chronic changes in organ function through fibrosis and hypertrophy of the muscle in the organ wall. It now appears that cytoskeletal remodeling mediated by Rho GTP-binding proteins may be a unifying pathway that plays a role in regulating contractile apparatus organization and transcriptional activity of muscle-specific genes during SMC differentiation. Several key issues need to be resolved before a full understanding of these pathways in normal physiology and disease can be appreciated. Some of these issues include identifying the factors and signaling mechanisms that lead to activation of RhoA, determining the role of cytoskeletal remodeling induced by other monomeric Rho-family GTPases, such as Rac and Cdc42, on smooth muscle-specific gene transcription and mechanical plasticity, identifying other proteins and their roles in signaling and regulation of RhoA-mediated smooth muscle plasticity, and determining the still unknown mechanosensitive pathways that might exist and coordinate functional consequences of phenotypic plasticity and mechanical plasticity during development and disease pathogenesis.