Smooth Muscle α-Actin Deficiency in Myofibroblasts Leads to Enhanced Renal Tissue Fibrosis*

Myofibroblasts are a major source of proinflammatory cytokines and extracellular matrix in progressive tissue fibrosis leading to chronic organ failure. Myofibroblasts are characterized by de novo expression of smooth muscle α-actin (SMαA), which correlates with the extent of disease progression, although their exact role is unknown. In vitro cultured myofibroblasts from kidney of SMαA knock-out mice demonstrate significantly more prominent cell motility, proliferation, and type-I procollagen expression than those of wild-type myofibroblasts. These pro-fibrotic properties are suppressed by adenovirus-mediated SMαA re-expression, accompanied by down-regulation of focal adhesion proteins. In interstitial fibrosis model, tissue fibrosis area, proliferating interstitial cell number, and type-I procollagen expression are enhanced under SMαA deficiency. In mesangioproliferative glomerulonephritis model, cell proliferation in the mesangial area is also enhanced in SMαA knock-out mice. Adenoviral SMαA introduction into renal interstitium obviously ameliorates tissue fibrosis in interstitial fibrosis model. These results indicate that SMαA suppresses the pro-fibrotic properties of myofibroblasts, highlighting the significance of smooth muscle-related proteins in moderating chronic organ fibrosis under pathological conditions.

Tissue fibrosis, replacing functional normal parenchymal tissue, is a common feature of a variety of chronically failing organs, including interstitial pulmonary fibrosis, liver cirrhosis, and progressive renal disease. These are important problems in terms of both mortality and morbidity worldwide. Understanding regulatory mechanisms is necessary for investigation of therapeutic approaches to these diseases.
"Myofibroblasts" appear in fibrosing tissues and characteristically show de novo expression of smooth muscle (SM) 2 -related proteins, including smooth muscle ␣-actin (SM␣A) (1), which is physiologically expressed in vascular smooth muscle cells (VSMCs) and has the function of regulating vascular tone in cooperation with myosin. Myofibroblasts are a specific cell population that has both fibroblastic and SM-like properties, thought to be a source of inflammatory cytokines and extracellular matrix in diseased conditions of various organs. These properties of myofibroblasts are common regardless of the affected organs; however, origins of myofibroblasts are quite diverse: stellate cells in liver and pancreas (2,3), interstitial fibroblasts in lung (4), glomerular mesangial cells (5), and renal interstitial fibroblasts (6). Recently, bone marrow-derived cells have been implicated as a substantial source of myofibroblasts in some organs (7)(8)(9). Epithelial-to-mesenchymal transition from tubular epithelial cells is also supposed as a pathway for renal interstitial myofibroblast generation (10). SM␣A expression is the most well known characteristic and widely used marker of myofibroblasts. In vitro SM␣A molecules incorporated into actin filaments have functions in contracting collagen gel (11) and retarding motility by increasing cell adhesion onto extracellular matrix (12). These contractile functions serve wound repair processes of skin or connective tissues by generating a closing force; however, for other visceral organs, such as lung, liver, and kidney, the role of their myogenic properties is not understood in pathological states.
We have investigated the transcriptional mechanism of SM␣A gene (Acta2) in renal myofibroblasts (13,14). Rodent unilateral ureteral obstruction (UUO), characterized by a diffuse and extensive emergence of renal interstitial myofibroblasts in a short period, initiated by a simple procedure (15), is a good model for studying tissue fibrosis. Using the UUO model, this study aims to elucidate the exact function of SM␣A in myofibroblasts in the course of tissue fibrosis, using SM␣A knockout mice and the adenoviral SM␣A gene transfer method. Its unanticipated importance against progressive tissue fibrosis is reported.

EXPERIMENTAL PROCEDURES
SM␣A Ϫ/Ϫ Mice and Cultured Mesangial Cells-Wild-type (WT) and SM␣A knock-out (SM␣A Ϫ/Ϫ ) mice were produced by mating SM␣A heterozygous (SM␣A ϩ/Ϫ ), 129-background mice generously provided by Dr. R. J. Schwartz (Baylor College of Medicine) (16). Handling and surgical manipulation of all experimental animals were carried out according to the guidelines of the Committee on the Use of Live Animals in Teaching and Research of Osaka University. Primary culture of mesangial cells was established from kidneys of these mice as described previously (17), maintained in RPMI 1640 medium with 17% FCS unless otherwise mentioned, and utilized as cultured myofibroblasts.
Cell Proliferation and Migration Assay-Cell proliferation was evaluated by determining viable cell number colorimetrically, using CellTiter 96 AQueous One Solution Cell Proliferation Assay kit (Promega Corp., Madison, WI), generally according to the product instructions. 2 ϫ 10 2 cells were seeded in each well of 96-well plates and incubated in 17% FCS-RPMI 1640 medium for 24 h, followed by 0.5% FCS incubation to stop their proliferation for 24 h. Medium was then replaced by fresh 10% FCS or 0.5% FCS RPMI 1640, and 24 h later absorbance at 490 nm was measured. Cell migration was evaluated by scrape wounding assay. Subconfluent cells on a 35-mm-diameter dish were scraped to make a straight-edged wound using a plastic tip, incubated in RPMI 1640 medium 0.5% FCS to avoid the effect of proliferation, and monitored periodically by light microscope.
Mouse Models for Renal Interstitial Fibrosis and Mesangioproliferative Glomerulonephritis-As a model for progressive renal interstitial fibrosis, UUO was performed in 8-week-old male WT and SM␣A Ϫ/Ϫ mice. The operation procedure was as described previously (13). Briefly, mice were anesthetized by an intraperitoneal injection of 50 mg/kg pentobarbital and inhalation of diethyl ether. After making a small ventral incision, the left ureter was identified and ligated. Kidney samples were harvested 2, 3, 7, or 14 days after operation (n ϭ 5-7 each). Mesangioproliferative glomerulonephritis was made by Habu-snake (Trimeresurus flavoviridis) venom injection (Habu-venom glomerulonephritis; HVGN) as described previously (14). Briefly, 8-week-old male mice were hemi-nephrectomized and received an intravenous injection of lyophilized venom from Habu-snake (Wako, Japan) dissolved in saline at 1.5 mg/kg body weight. Kidney samples were harvested 7 days after injection.
Immunostaining and Histopathological Analysis-Cultured cells on fibronectin-coated culture slides (BD Biosciences) were fixed with an ice-cold 1:1 mixture of acetone and methanol for 20 min. Kidney tissues were fixed with 4% paraformaldehyde overnight or for SM␣A staining by methacarn (10% acetic acid and 30% chloroform in methanol) for 3 h. Tissues were then embedded in paraffin and sectioned to 2-mm thickness. Stress fiber was stained by fluorescein isothiocyanate-conjugated phalloidin (Sigma). The proportion of fibrosis area in UUO kidney section was calculated from the blue-stained area in Masson's trichrome (MT) staining using MacSCOPE software (Mitani Corp.). Proliferating cell number was counted by proliferating cell nuclear antigen (PCNA)-positive cells per high magnification field, stained by anti-PCNA antibody (Immunotech). Mesangial proliferation was evaluated by counting the number of mesangial cell nucleoli in periodic acid-Schiffstained sections.
Adenovirus-mediated Gene Transfer-Ad-mSM␣A was developed generally according to the method of Graham and Prevec (18). Briefly, we first amplified murine SM␣A cDNA (from bp ϩ63 to bp ϩ1268, fully including the translated region between bp ϩ72 and bp ϩ1205) using pCR-Blunt II-TOPO cloning vector (Invitrogen) and cloned downstream of the CMV promoter region of pACCMV.pLpA shuttle vector. Next, pACCMV.pLpA with SM␣A cDNA inserted and pJM17 vector (a replication-defective adenoviral genome vector with deleted E1 and mutated E3 promoter regions) were co-transfected into human embryonic kidney 293 cells to induce recombination to develop CMV promoter-driven SM␣A cDNA-expressing adenoviral vector. The vector was then amplified and purified, and its infectivity was measured by plaque assay. When Ad-mSM␣A was transfected into SM␣A Ϫ/Ϫ -cultured myofibroblasts at a multiplicity of infection of 300 PFU/cell, more than 90% of cells expressed SM␣A protein with no apparent cytotoxicity. Another adenoviral vector, Ad-LacZ, was a generous gift from Dr. T. Mano (Osaka University) (19), used as a control for nonspecific effects of adenoviral infection. To transfer genes in vitro, subconfluent mesangial cells were rinsed twice with PBS(ϩ), incubated with serum-free medium containing adenovector for 1 h at room temperature. After removing the adenoviral medium, cells were cultured in medium with 0.5% FCS for 7 days, and then RNA and cell lysates were collected. For in vivo gene transfer when operating for UUO, adenoviral vector solution (6 ϫ 10 7 PFU in 50 l of saline/ individual kidney) was injected intraparenchymally with a 30-gauge needle separately at two points while clamping the renal artery and renal vein for 3 min to minimize adenoviral vector flowing immediately out of the kidney (n ϭ 5-6 each).
RNA Preparation and RT-PCR-RNA from mouse kidneys and cultured cells was collected as described previously (14). Reverse transcription was performed using 1 g of RNA, MuLV reverse transcriptase (Applied Biosystems, Foster City, CA), dNTP, and random hexamer. Sequences of primers for semiquantitative PCR are indicated in Table 1A. Semiquantitative PCR was carried out with 1 l of template cDNA, primers (10 pmol each), dNTP, and ExTaq DNA polymerase (Takara Bio Inc.) in a final volume of 20 l. Thermal cycler conditions were as follows: denaturation at 95°C for 45 s, annealing at 56 -59°C for 45 s, and extension at 72°C for 1 min with appropriate cycle numbers. To perform real-time quantitative PCR, ABI PRISM 7700 Sequence Detection System (Applied Biosystems) was used with 50 l of final volume containing 5 l of template cDNA, solution, 10 pmol primers, labeled probe, 25 l of Platinum Quantitative PCR Supermix-UDG (Invitrogen). PCR primer and probe sequences for real-time PCR are shown in Table 1, B and C. Thermal cycler conditions were 2 min at 50°C, 10 min at 95°C, and 50 cycles of 15 s at 95°C, followed by 1 min at 60°C.

RESULTS
Myogenic Gene Expression in SM␣A Ϫ/Ϫ Myofibroblasts-SM␣A Ϫ/Ϫ mice had no leaky expression of SM␣A mRNA in all checked tissues, and primary kidney cells cultured by our method demonstrated positive staining for desmin, which is regarded as one of the markers for mesangial cells (data not shown). Cultured mesangial cells show general characteristics of myofibroblasts; therefore, they were used as cultured myofibroblasts for the following experiments. Cultured myofibroblasts from SM␣A Ϫ/Ϫ mouse kidney had more sparse stress fibers in cytoplasm than WT, which seemed due to lack of SM␣A (Fig. 1A). SM␣A Ϫ/Ϫ myofibroblasts showed complete disappearance of SM␣A expression but enhancement of other SM-related gene expression, including SM ␥-actin (SM␥A), SM myosin heavy chain isoform 1 (SM1), and another actin isoform, skeletal muscle ␣-actin (Sk␣A) (Fig. 1B). This result suggests the existence of some compensatory mechanism for maintaining SM phenotype and cellular actin content. Cysteine-rich LIM-only protein (CRP) 1 and myocardin, SM-specific transcriptional cofactors (20,21), were not detected in either WT or SM␣A Ϫ/Ϫ myofibroblasts by RT-PCR, whereas expression of CRP2 was observed in WT myofibroblasts but was down-regulated under SM␣A deficiency (Fig. 1C).

Enhanced Interstitial Fibrosis and Mesangial Proliferation in
Renal Disease Models of SM␣A Ϫ/Ϫ Mice-There were no overt macroscopic or microscopic differences between WT and SM␣A Ϫ/Ϫ kidney. Therefore, UUO, which causes artificial hydronephrosis and is the most common experimental model for renal interstitial fibrosis (15), was performed on both WT and SM␣A Ϫ/Ϫ mice. For WT kidney under physiological conditions, SM␣A immunostaining was only seen in VSMCs of blood vessels but was newly observed in renal interstitium on UUO day 7, indicating the emergence of myofibroblasts ( Fig.  2A). In contrast, no SM␣A immunostaining was seen in either  physiological or UUO kidneys of SM␣A Ϫ/Ϫ mice. The proportion of fibrosis area calculated from the blue-stained area of MT staining was more extensive in SM␣A Ϫ/Ϫ mice than WT mice (Fig. 2, A and B). Interstitial proliferating cells estimated from proliferating cell nuclear antigen-positive cell count in UUO kidney on day 7 was significantly higher in SM␣A Ϫ/Ϫ mice than WT mice (Fig. 2C). In fibrotic tissue, type-I collagen is the main extracellular matrix protein, composed of procollagen ␣1(I) and ␣2(I) chains processed from the products of genes Col1a1 and Col1a2. In the UUO model, expression of type-I procollagen (␣2) mRNA increased with time. Type-I procollagen (␣2) mRNA in SM␣A Ϫ/Ϫ mouse UUO kidney on days 7 and 14 was raised to about twice the level found in WT UUO kidney (Fig.  2D). These results demonstrate that SM␣A deficiency promotes the increase in myofibroblasts and progression of renal interstitial fibrosis. Next, we performed HVGN to see the difference in the progression of glomerulonephritis between SM␣A Ϫ/Ϫ and WT mice. In HVGN, normal mesangial cells transdifferentiate into SM␣A-positive myofibroblasts and exhibit prominent cell proliferation (14). Glomerular mesangial cell number was significantly more increased in SM␣A Ϫ/Ϫ mice than WT mice (39.3 Ϯ 5.0 versus 31.7 Ϯ 5.3) (Fig. 3). Therefore, SM␣A deficiency also promotes progression of mesangioproliferative glomerulonephritis.
Myofibroblast Activity Is Diminished by Forced SM␣A Re-expression-Both WT and SM␣A Ϫ/Ϫ cultured myofibroblasts show no evidence of proliferation in 0.5% FCS RPMI  medium. SM␣A Ϫ/Ϫ myofibroblasts showed more vigorous proliferating activity than WT cells (153 Ϯ 28% versus 117 Ϯ 23%) after 10% FCS stimulation for 24 h compared with incubation without stimulation, as measured by formazan colorimetric assay. Their enhanced proliferation was blunted to a degree comparable with WT cells with forced SM␣A re-expression by transfection of adenoviral vector (Ad-mSM␣A), which effectively expresses murine SM␣A downstream of the CMV promoter (Fig. 4A). Type-I procollagen (␣2) mRNA expression was enhanced in SM␣A Ϫ/Ϫ cultured myofibroblasts 2.2-fold more than in WT cells (Fig. 4B), consistent with in vivo UUO data. When SM␣A was re-expressed in SM␣A Ϫ/Ϫ myofibroblasts, type-I procollagen mRNA expression was reduced to a similar extent as in WT myofibroblasts.
Migration is an important property of active myofibroblast. In wound scrape assay, SM␣A Ϫ/Ϫ cultured myofibroblasts also demonstrated more enhanced migration than quite slow migration of WT cells in minimum (0.5%) FCS medium, used to avoid effects from cell division. Forced re-expression of SM␣A in SM␣A Ϫ/Ϫ myofibroblasts slowed their enhanced migration (Fig. 4C). Enhanced myofibroblast migration is related to accelerated tissue fibrosis in several organs (22,23) although the relation is still not shown in kidney UUO model. It seems that enhanced migration accelerates the accumulation of myofibroblasts to the inflammation site in kidney. Ad-LacZ, an adenoviral vector expressing ␤-galactosidase from the same CMV promoter, was used as a control for adenoviral infection. Transfection of Ad-LacZ into myofibroblasts at the same multiplicity of infection as Ad-mSM␣A had no significant influence on their proliferation, type-I procollagen mRNA expression, or migration. These results indicate that expression of SM␣A affects suppression of proliferation, procollagen synthesis, and migration of myofibroblasts.
Down-regulation of Focal Adhesion Proteins in Myofibroblasts by SM␣A Introduction-Focal adhesion (FA) cell membrane protein complex, which links intracellular cytoskeleton and extracellular matrix, mediates not only motility force but also extracellular signals. FA includes FAK, which is a nonreceptor tyrosine kinase. FAK is involved in integrin-related and other cell surface receptor-related signal transduction and is reported to be regulated by the status of the actin cytoskeleton (24,25). Compared with WT myofibroblasts, SM␣A Ϫ/Ϫ cells had prominently up-regulated FA complex proteins, FAK, phosphorylated FAK at Tyr 397 , which is the key autophosphorylation site for its activity, paxillin, and vinculin (Fig. 5). When SM␣A was forcibly re-expressed in SM␣A Ϫ/Ϫ myofibroblasts, FAK, Tyr 397 -phosphorylated FAK, paxillin, and vinculin were  DECEMBER 29, 2006 • VOLUME 281 • NUMBER 52 clearly diminished. In these conditions, there was not a large difference in total actin proteins. These findings suggest that SM␣A expression in myofibroblasts down-regulates cellular FA protein contents.

Smooth Muscle ␣-Actin Inactivates Myofibroblasts
In Vivo Adenoviral SM␣A Gene Transfer Ameliorates Renal Fibrosis-As the above in vitro results suggested the possibility of ameliorating tissue fibrosis by forced SM␣A expression, Ad-mSM␣A was introduced into mouse kidney and its effect on interstitial fibrosis was examined. Intraparenchymal Ad-mSM␣A injection induced SM␣A expression in interstitial fibroblasts of SM␣A Ϫ/Ϫ UUO kidneys as early as day 2, although WT non-injected UUO kidneys had no SM␣A staining in the interstitium on the same day (Fig. 6A). SM␣A expression induced by Ad-mSM␣A was observed to persist until day 7 in SM␣A Ϫ/Ϫ UUO kidneys (Fig. 6B).
Analysis of the extent of renal disease progression for SM␣A Ϫ/Ϫ UUO kidneys with forced SM␣A re-expression demonstrated a significantly diminished interstitial fibrosis area and type-I procollagen mRNA expression in renal tissue to almost half the level of non-injected UUO kidneys. Also, SM␣A overexpression in WT UUO kidneys ameliorated interstitial fibrosis, and type-I procollagen mRNA expression in renal tissue was reduced to about one-third of noninjected UUO kidneys (Fig. 6, D and E). Collectively, this amelioration of tissue fibrosis appears attributable to suppression of myofibroblast expansion and collagen synthesis by forced SM␣A expression.
Interestingly, Ad-mSM␣A-transfected WT kidneys 7 days after UUO showed paradoxically less total SM␣A-positive area and SM␣A mRNA expression compared with both non-injected and Ad-LacZ-injected WT UUO kidneys (Fig. 6, B and  C). The CMV promoter in Ad-mSM␣A vector drives SM␣A gene expression strongly and constitutively, whereas the proper promoter in SM␣A gene shows spatiotemporally regulated gene activation. The earlier expression of SM␣A forced by Ad-mSM␣A was observed compared with non-transfected WT UUO kidneys (Fig. 6A) that may have suppressed expansion and pro-fibrotic activity in interstitial fibroblasts.

DISCUSSION
Myofibroblasts, which express SM␣A and other SM-related proteins, appear in various pathological states, including connective tissue granulation, fibrotic diseases, and stroma reaction to neoplasia (1); however, little is known about the role of SM␣A in these conditions. In fibrotic diseases of several organs, SM␣A expression of myofibroblasts is recognized as a hallmark of their emergence and an indicator of disease severity (1,26). However, the role of SM-related proteins in the pathogenesis of fibrosis has been poorly elucidated. The present study demonstrates that SM␣A deficiency enhances the progression of kidney diseases in two models, interstitial fibrosis (UUO) and mesangioproliferative glomerulonephritis (HVGN) (Figs. 2 and  3), suggesting the role of SM␣A in myofibroblast is common in both mesangial and interstitial lesions. Tissue fibrosis is ameliorated by forced expression of SM␣A in renal interstitial myofibroblasts.
Recently, reports have shown that contractile, plus proliferative and/or collagen-producing, properties do not coincide or are differentially exhibited in myofibroblasts under certain conditions (27,28). VSMCs, which physiologically express abundant SM contraction-related proteins including SM␣A, modulate their phenotype from "contractile type" to "synthetic type" under pathological conditions such as atherosclerotic lesions (29). To date, several transcription factors related to VSMC differentiation have been demonstrated to inhibit VSMC proliferation, migration, and progression of arteriosclerosis (19,30), suggesting that the contractile properties are distinct from proliferative, migratory, and synthetic properties. In addition, SM␣A promoter has a target sequence of tumor suppressor protein p53 (31), and SM␣A gene introduction into transformed fibroblasts reduces their proliferative properties (32), although the underlying mechanism has not been elucidated. Taking these findings together, we propose that the SM-like phenotype and productive phenotype are mutually opposed within myofibroblasts and that SM␣A per se has the function of maintaining a cell in a static state.
FA complex consists of paxillin, vinculin, talin, plasma membrane integrins, and FAK and is linked to cytoplasmic actin fibers (24). Several stimuli affecting integrin or changes in actin fiber status induce phosphorylation at several tyrosine residues of FAK, inducing active cell motility, cell cycle promotion, and protein (25,33). Recent reports indicate the importance of FAK in the regulation of myofibroblast activation and collagen synthesis (34,35). Our results, decreasing of FA complex in myofibroblasts by SM␣A expression, suggest the relationship between SM␣A in actin stress fibers and FA. We think additional investigations about protein-protein interaction between SM␣A and FA complex are necessary.
Our results show no evident difference in other important receptor kinases for myofibroblast activity such as transforming growth factor ␤ receptor type-I and platelet-derived growth factor ␤-receptor (data not shown). In addition, Rho-GTPases are important downstream effectors of FAK on cell motility (36). Our previous report demonstrated that inhibition of Rhoassociated kinase attenuated renal interstitial fibrosis in the UUO model (37), suggesting the importance of this pathway in the pathogenesis of fibrosis.
Recent evidence has emerged indicating that actin molecules are abundant in the nucleus and perform functions in chromatin remodeling and gene activation (38,39). SM␣A is reported to be a binding target of antioxidant-responsive element present in some genes (40), although it is still unclear whether SM␣A exists and has a specific function in the nucleus. This mechanism might also take part in the type of gene expression and proliferation reported here.
SM␣A Ϫ/Ϫ myofibroblasts interestingly demonstrated upregulation of mRNA of other SM-related proteins and another myogenic isoform of actin, such as SM␥A, SM1, and Sk␣A, presumably to compensate for the lack of SM␣A. Similar compensatory expression systems have been reported in cardiomyocytes of cardiac ␣-actin knock-out mice (41) and in skeletal muscle of Sk␣A knock-out mice (42), although the underlying transcriptional mechanisms have not been clarified. In addition, from the present findings expression of CRP1, CRP2, and myocardin, transcriptional co-factors for SM-specific gene expression in VSMCs (20,21), was not detected or up-regulated in SM␣A Ϫ/Ϫ myofibroblasts (Fig. 1C); therefore, they do not seem responsible for the compensatory SM gene expression in these cells. These results suggest that different signaling systems from VSMCs predominate for other SM-related protein expression in myofibroblasts.
The limitation in this experiment is the usage of somewhat conventional knock-out mice and an adenoviral gene transfer system. SM␣A Ϫ/Ϫ mice physiologically exhibit lower blood pressure and poor responsiveness of arteries to vasoconstrictor agents (16). As fibrosis in the UUO kidney is triggered by vasoconstriction and a decrease in renal blood flow (15), plus is ameliorated by several blood pressure-lowering agents (43,44), these characteristics appear unlikely to contribute to renal fibrosis progression. It is also possible that the effect of SM␣A deficiency is modulated by compensatory up-regulation of other SM-related proteins. However, this up-regulation seems not to contribute to the modulation of disease course, as SM␣A gene transfer into UUO kidneys significantly ameliorated interstitial fibrosis in both WT and SM␣A Ϫ/Ϫ mice. These results suggest a pivotal role for SM␣A but not for other SM-related proteins. Inflammation evoked by adenovirus infection is reported as an important problem (45); however, there have been no reports on adenoviral vector-mediated fibrosis in mouse kidney and no apparent differences in mRNA expression levels of pro-inflammatory mediators such as transforming growth factor ␤1 and monocyte chemoattractant protein-1 between non-injected and Ad-LacZ-injected UUO kidneys (data not shown). Of note, the present study showed that forced SM␣A expression ameliorates tissue fibrosis in a kidney UUO model. However, in some pathological conditions such as contracture of skin scars, contraction force generated by SM␣A plays a certain role in disease deterioration. Those may be beneficially influenced by SM␣A deficiency. Appropriateness of SM␣A introduction needs to be considered on an individual disease basis.
In conclusion, SM␣A in myofibroblasts appears to have a suppressing role in tissue fibrosis progression, demonstrated by both loss-of-function and gain-of-function analyses. These findings suggest several novel therapeutic approaches to myofibroblast-related fibrotic diseases: for example, enhancement of SM␣A expression by gene transfer methods or by up-regulation of SM-related transcription factors.