The A-Smooth Muscle Actin-Positive Cells in Healing Human Myocardial Scars

Home , Healing

Ingrid E. M. G. Willems, Michael G. Havenith, shaped cells have been observed. These cells were Jo G. R. De Mey, and Mat J. A. P. Daemen called myofibroblasts (MFs) and exhibit contractile properties. 1-2 MFs share characteristics of both fibro- From the Departments ofPathology and Pharmacology, blasts and smooth muscle cells and their original defi- Cardiovascular Research Institute Maastrncht, nition is based on ultrastructural criteria.2 Character- Maastricht, The Netherlands istic ultrastructural features of MFs are both the presence of stress fibers with subplasmalemmal attachment plaques and abundant rough endoplasmic reticulum next to a discontinuous deposition of basal Interstitial ceUs in the scars of human myocar- lamina-like material and intracytoplasmic filaments.3 dial infarctions ofdifferent postinfarction times Immunohistochemical analysis of MF revealed a het- hours to 1 7years old) were characterized by (6 erogeneous cytoskeletal composition under different antibodies to cv-smooth muscle actin (ASMA), vi- pathological conditions. Healed dermal scars ex- mentin, and desmin. Basal lamina deposition pressed only vimentin, whereas co-expression of was studied with antibodies to the basal lamina a-smooth muscle actin (ASMA) and desmin was ob- protein type IV coUagen. Nonvascular spindle- served in hypertrophic dermal scars and fibromato- shaped cells expressing ASMA were present SiS.4 During experimental dermal wound healing in within 4 to 6 days after infarction. These cells rats the appearance of MF is transient. The first co-expressed vimentin but no desmin and ASMA- and vimentin-positive cells appear at day 6 showed discontinuous basal lamina deposition. but can no longer be detected after week 4. Further- In electron microscopy these ceUs showed fea- more, they are oriented parallel to the wound surface tures characteristic of myo.ffbroblasts. The and perpendicular to the vessel.5 spindle-shaped ceUs persistedfor a long period In experimentally induced myocardial infarctions of time and could even be identified 17 years in rats the first ASMA expressing spindle-shaped postinfarction. In transmural infarctions they cells were identified between days 2 and 4.6 In con- were orientated parallel to the endocardium trast to dermal scars, they persisted up to 10 weeks and epicardiun. In nontransmural patchy in- in the rat and showed a parallel orientation to surviv- farctions they showed an orientation adjacent ing cardiomyocytes present at the border of the in- to the cardiomyocytes and appeared to be less farction. In this and other models of wound healing dense than in the transmural infarctions. In con- basal lamina deposition was also detected, espe- clusion, myofibroblasts expressing ASMA per- cially at the interface of the scar and the surrounding sist within human scars and show a myocardial stromal tissue.76 preferential alignment that may be the result of Because no information is available on the pres- the continuous mechanical stress caused by the ence of ASMA expressing spindle-shaped cells in hu- ongoing contraction and relaxation of the sur- man myocardial scars, we investigated the appear- viable Pathol rounding myocardiunt. (Am J in 1994, 145:868-875) ance of these cells human myocardial infarctions with different postinfarction survival times, varying from less than 2 days to several years.

After necrotizing injuries, like an infarction, the myo- Supported by grant 90282 from the Netherlands Heart Foundation. cardial tissue responds by inflammation and repair. Accepted for publication July 7, 1994. During this process necrotic tissue is replaced by Address reprint requests to Dr. Ingrid E.M.G. Willems, Depart- granulation tissue that matures into scar tissue. In ment of Pathology, University of Limburg, P.O. Box 5800, 6202 AZ granulation tissue characteristic interstitial spindle- Maastricht, The Netherlands.

868 MFs in Human Myocardial Scars 869 AJP October 1994, Vol. 145, No. 4

Materials and Methods tibodies to a-smooth muscle actin (ASMA monoclonal antibody, dilution 1:4000; Sigma), desmin (mono- Source of Tissue clonal antibody, dilution 1:100; Organon), vimentin Tissue blocks of human myocardial infarctions ob- (monoclonal antibody, dilution 1:100; Organon), or type IV collagen (polyclonal antibody, dilution 1:500). tained during autopsy were retrieved from the ar- chives of the Department of Pathology of the Univer- The specificity of these antibodies has been de- sity Hospital of Maastricht. For inclusion in this study scribed elsewhere.1012 the duration from the first signs of myocardial infarc- To expose the antigenic sites for antibodies di- tion to the demise of the patient had to be known. rected against vimentin and type IV collagen, prein- cubation with 0.1% pepsin (Boehringer Mannheim, Furthermore, infarct duration was estimated histologi- Mannheim, Germany) in 0.1 N HCI, pH 2, for 30 min- cally.9 A patient was included in the study when the clinical and histological postinfarction survival time utes at room temperature was required. After incu- bation with the primary antibody the sections were were in agreement. Based on these criteria a total washed 3 x 5 minutes with TBS and incubated with number of 35 infarct cases and an additional nonin- horseradish peroxidase-conjugated swine anti-rabbit farcted control group with no cardial history of any (Dako; dilution 1: 150) for the antitype IV collagen an- kind (group 0, n = 5) were studied. The infarct group tibody or rabbit anti-mouse antibodies (Dako; dilution was subdivided in five groups. Group 1 included 5 1:200) for anti-ASMA, desmin, and vimentin antibod- cases with a postinfarction time of less than 2 days, ies, respectively, during 45 minutes at room tempera- group 2 included 6 cases with a postinfarction time of ture. After final washing with TBS, a diaminoben- 2 to 7 days, group 3 included 10 cases with a postin- zidine-H202 substrate was used to visualize the im- farction time of 1 to 2 weeks, group 4 included 6 cases munoreactivity. To enhance the precipitation of the with a postinfarction time of 2 to 4 weeks, and group chromogen, 0.1 M imidazole was added to the sub- 5 included 8 cases with a postinfarction time exceed- ing 4 weeks. strate. When the presence of pigment-containing macrophages obscured the interpretation of the im- munohistological findings, aminoethylcarbazole was Electron Microscopy used as the substrate. Before mounting, the sections were weakly counterstained with Mayer's hematoxy- Small tissue cubes of infarcted areas (postinfarction lin. time more than 4 weeks) were obtained during autopsy (postmortem time less than 4 hours) and fixed in 2.5% glutaraldehyde for 4 hours; postfixed in 1% Results osmium tetroxide in 0.1 M phosphate buffer (pH 7.2) for 1 hour, dehydrated, and embedded in epon 812. In the control and noninfarcted myocardial tissue of all Semithin sections were cut and stained with methyl- other groups, ASMA immunoreactivity could only be ene blue to select representative areas. These were detected in the medium of arteries and veins and in collected on a copper grid, contrasted with saturated pericytes surrounding arterioles and venules. ASMA uranyl acetate in 50% ethanol and 10% lead citrate, expressing spindle-shaped cells not belonging to the and examined with a Philips CM 12 transmission elec- vascular smooth muscle cell compartment were ex- tron microscope. amined with regard to their alignment, immunophe- notype, and presence in time. Immunohistochemistry Group 0: Controls Tissue blocks from the infarcted myocardium were Interstitial cells located in the subendocardial space fixed in 4% neutral-buffered formaldehyde and em- showed a parallel orientation to the endothelial lining bedded in paraffin, according to routine procedures. and expressed ASMA. The cells expressing ASMA Adjacent sections of each block were stained with the showed a discontinuous pericellular immunoreactiv- respective antibodies. After deparaffinization, 4-p- ity with type IV collagen and a weak cytoplasmic im- thick sections were washed 3 x 5 minutes in Tris- munoreactivity with vimentin. However, only very few buffered saline (TBS). Thereafter, endogenous per- of these subendocardial cells showed immunoreac- oxidase activity was blocked by 0.1% H202 in tivity for desmin. methanol for 20 minutes at room temperature and the Interstitial cells in the myocardium and cardio- sections were incubated during 45 minutes with an- myocytes and endothelial cells lacked any immuno- 870 Willems et al AJP October 1994, Vol. 145, No. 4

reactivity for ASMA. Cytoplasmic immunoreactivity for Group 1: 6 to 48 hours of Infarction vimentin could be observed next to pericytes, in myocardial interstitial cells, endothelial cells, fat cells, in- In infarctions with a postinfarction time of less than 2 flammatory cells, and vascular smooth muscle cells. days, cardiomyocytes in the infarcted area clearly The antidesmin antibody stained vascular smooth showed depletion of desmin. The expression of muscle cells and cardiomyocytes. Immunoreactivity ASMA in infarcted areas was restricted to vascular for type IV collagen was absent around myocardial structures (Figure 1, a). Vimentin was expressed interstitial cells but the antibody stained all basal lami- weakly in vessel walls but stained densely in the in- nas around individual cardiomyocytes, vascular flammatory infiltrate. The scaffold of the basal lamina smooth muscle cells, pericytes, Schwann cells, fat around cardiomyocytes was almost completely intact cells, and the basal lamina underlying endothelial in the infarcted tissue, especially in the central areas cells. that showed no infiltration by inflammatory cells.

...... "' ' ..*:'..r._Fe:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.....

Figure 1. Pbotomicrograph ofhuman myocardalscar tissue ofdifferent postinfarction time stainedv'~~~~~~~~~~~~~~~~~~~~~~~with anti-ASMA...., (magnification x200). A: 6 to 48 hourspostinfarction: ASMA expression is restricted to vascular structures (arrow). B: 2 to 7 dayspostinfarction: ASMA expression in the granulation tissue at the interface ofvital myocardium and infarcted area. Nopreferred orientation. Arrowhead, vital cardiomyocyte; arrow, ASMfA-positive spindle-sbaped cells in the granulation tissue. C: 1 to 2 weeks postinfarction: ASMA expression in the granulation tissue. Note theparallel orientation to the sunrvitng cardiomyocyte. Arrowhead, vital cardiomyocyte; arrow, ASMA-positive spindle- shaped cells. D: 2 to 4 weeks postinfarction: ASMA expression is abundant. Note the alignment ofthe ASMA-positive cells. E: >1 month postinfarction: ASMA expression in a 1 7-year-old infarction, witb per- sisting alignment of ASMA-positive cells. Arrowhead, remaining cardiomyocyte; arrow, ASMA-positive spindle-shaped cells. MFs in Human Myocardial Scars 871 AJP October 1994, Vol. 145, No. 4

Group 2: 2 to 7 Days of Infarction mural infarctions they were oriented parallel to the endocardium and epicardium. Again, these cells did not The first spindle-shaped cells expressing cytoplas- show immunoreactivity with desmin (Figure 2, b). mic ASMA appeared between days 4 and 6 in the ASMA and vimentin were present in the cytoplasm granulation tissue at the interface of vital myocardium and type IV collagen was present in a discontinuous and infarct area (Figure 1, b). These spindle-shaped pattern around the spindle-shaped cells. Also, rem- cells co-expressed vimentin (Figure 2, a), which was nants of cardiomyocytic basal laminas still expressed also expressed by inflammatory cells, that were collagen type IV (Figure 2, c). present in high numbers at the margin of the infarction. In the infarcted area only smooth muscle cells of Group 4: 2 to 4 Weeks of Infarction remaining vessel walls expressed desmin, whereas no desmin expression was found in cardiomyocytes In the period from week 2 to 4 the staining pattern was or ASMA-positive nonvascular spindle-shaped cells. identical to the pattern described in group 3 and the Type IV collagen was expressed in remnants of car- parallel orientation of the spindle-shaped cells to sur- diomyocytic basal laminas and in a discontinuous viving cardiomyocytes or endocardium and epicar- pattern around the spindle-shaped cells in the in- dium was even more pronounced (Figure 1, d). Apart farcted area. from vessel walls, desmin could be detected in a few spindle-shaped cells in the scar tissue (Figure 2, d). Group 3: 1 to 2 Weeks of Infarction Group 5: More Than 4 Weeks of Infarction In 1- to 2-week-old nontransmural infarctions, the spindle-shaped cells were oriented parallel to adja- In more than 4-week-old infarctions the same pattern cent surviving myocardial cells (Figure 1, c). In trans- persisted, even in scar tissue with a postinfarction

2 ., -

Figure 2. Photomicrograph of human myocardial scar tissue, stained with different antibodies (vimentin, desmin, and collagen IV) (magnification X200). A: Vimentin: 2 to 7 dayspostinfarction. Interstitial spindle-shaped cells in the scar tissue expressing vimentin. Arrowhead, cardiomyocyte. B: Desmin: 1 to 2 weeks postinfarction. Vital cardiomyocytes remain to express desmin (arrowhead). Cardiomyocytes in the infarcted area show no desmin expression (arrow). ASMA-positive cells are desmin negative (center of the photomicrograph). C: Collagen type IV: 1 to 2 weeks postinfarction. Subepicardial scar tissue shows remnants of basal laminas of cardiomyocytes expressing collagen type IV(arrowhead). Arrow, fat cells. D: Desmin: 2 to 4 weeks postinfarction. In the infarcted area some spindle-shaped cells expressing ASMA are desmin positive (arrow). 872 Willems et al AJP October 1994, Vol. 145, No. 4

time of 17 years (Figure 1, e). ASMA-positive spindle- very few MFs showed desmin expression, and, if so, shaped cells in nontransmural infarctions showed an only from 2 to 4 weeks postinfarction. The vast ma- orientation parallel to the adjacent cardiomyocytes jority of MFs in human myocardial scars would there- and appeared to be less dense than in transmural fore be VA cells, according to a proposed immuno- infarctions. The spindle-shaped cells in transmural in- phenotypic division of MF.5 To further illustrate their farctions showed a striking parallel orientation to the heterogeneity, we summarized characteristics of MF endocardium and pericardium. Ultrastructurally, the in myocardial scars in humans and rats, as well as in ASMA expressing spindle-shaped cells were dermal wounds in rats (Table 1). MFs in all these stud- multipolar, with abundant dilated rough endo- ies expressed both vimentin and ASMA. In both rat plasmic reticulum, a Golgi vesicular transport system, and human myocardial studies desmin expression and microfilaments extending into subplasma- was found in the spindle-shaped cells. We found sus- lemmal attachment plaques, characteristic for MFs tained expression of the basal lamina protein colla- (Figure 3). gen IV in early human myocardial scars. From ultrastructural studies in the rat it was already known that myocytes and vascular- and Schwann cell basal lami- Discussion nae remain intact after induction of myocardial infarc- Many authors have shown the presence of MF in heart tion.28 In older human myocardial infarctions collagen valves, lung interstitium,13-15 and pathological con- IV was present in a discontinuous pattern in the basal ditions as wound healing, fibromatosis, tumor- lamina of MF, which coincides with the findings in the surrounding stroma, and intima formation in aortic rat myocardial infarction study. The presence of col- grafts. 14'16-22 MFs were also found in the infarcted lagen type IV in dermal wounds is not well described myocardium of the rat and in the pressure overloaded in the literature. right ventricle in the rabbit6'23 but the immunophe- The appearance in time of MFs in human myocar- notype of MF is heterogeneous. In this study MFs dial scars is comparable to the finding in the rat myo- were identified in human myocardial scars5'624-26 cardium where MFs appeared between day 2 and 4 and found to co-express vimentin and ASMA. These after coronary artery ligation, and persisted for at cells showed the ultrastructural features of MF.27 Only least 4 weeks (Table 1). Normal rat29 and normal and

Figure 3. Electron photomicrograph of human myocardial scar tissue (magnification X20, 400) detailing the cytoplasm of an interstitial cell with the ultrastructural characteristics of a MF; abundant dilated rough endoplasmic reticulum (arrow) and intracytoplasmic filaments uwth dense bodies (arrowhead). MFs in Human Myocardial Scars 873 AJP October 1994, Vol. 145, No. 4

Table 1. Characteristics ofMFs in Human and Rat Myocardial Scars and Dermal Scars Human Myocardial Scars Rat Myocardial Scars Dermal Wound Vimentin + + + Actin + + + Desmin -/+ - + Collagen type IV + + ? Time course Continuous Continuous Transient Orientation Parallel to surviving myocardium Parallel to surface Data obtained from this study (human myocardial scars): Vracko et a16 (rat myocardial scars) and Darby et a15 (dermal wounds).

hypertrophied human myocardium (this study) lack large, transmural infarctions as a consequence of the MFs. The persistence of ASMA-positive MFs in myo- different mechanical forces. cardial scars (in our study up to 17 years) is in contrast The subendocardial arranged spindle-shaped with the findings in dermal wound healing where ex- cells expressing ASMA could actually be MFs ap- pression of ASMA is only transient:5 the first ASMA- pearing in the context of wound healing after small positive MFs appeared 6 days after experimental der- endocardial injury, analogous to endothelial injury in mal wound healing in rats, reached a peak at day 15, arteries30 or myocardial infarction (this study). but could no 1 longer be identified month after in- In the last years several molecules have been iden- duction of the wound. tified that show the ability to induce ASMA Another striking observation made in this study is expression in fibroblasts,31'32 endothelial cells,33 and the orientation of MFs in the scars. They were always fat-storing oriented parallel to surviving adjacent cardiomyo- cells in the liver,34 including heparin,31 endothelin,3 cytes in nontransmural, patchy infarctions but parallel tumornecrosis factor-a, platelet-derived growth fac- to the endocardium and epicardium in transmural in- tor, and granulocyte macrophage-colony-stimulating farction. In addition, it was noted that ASMA immu- factor.36 Recently, Desmouliere et al,37 proposed noreactivity was more intense in transmural than transforming growth factor-p1l (TGF-1l) as the key patchy infarctions. MFs in small myocardial scars in molecule for the induction of ASMA expression in sub- the rat also showed a parallel orientation to surviving cutaneous granulation tissue in rats and in cultured cardiomyocytes6 (Table 1). In experimental dermal human and rat fibroblasts.37 The effect of other fac- wound healing in rats MFs were oriented parallel to tors like endothelin is possible indirectly and also in- the epidermal surface and perpendicular to vascular volves TGF-f31.35,38 TGF-j1 is present in the devel- structures 8 days after inducing the wound. It is oping and adult heart,39'40 its immunoreactivity known from experimental dermal wound healing in disappears in the central necrotic part of an experi- mice that MFs were aligned circumferentially to the mentally induced myocardial infarction but a strong wound margin to close the wound in a purse string immunoreactivity and increased mRNA amounts effect.1 This alignment disappeared within 4 to 5 were found at the margins of the infarcted area.41 days. Thus, TGF-p1 may co-localize with MFs in the infarct. The reason for the persistence and apparently de- TGF-f31 also stimulates the production of interstitial termined orientation of MFs in myocardial scars is not collagen, which is increased after myocardial infarc- clear. The expression of the cytoskeletal protein actin tions in rats and humans.4243 Although no data are provides the myocardial scar with the possibility to available on the expression of contract and counteract the mechanical forces of the TGF-,B1 in human myo- surrounding myocardium. Preliminary studies using cardial scars, the available data do certainly suggest the Mulvany-Halpern myograph indicate that rat and a role for the molecule in inducing MFs in human myo- human myocardial scars are indeed able to contract cardial scars. but with a unique pharmacological profile (J. De Mey, The findings in this study illustrate that the healing personal communication). A possible explanation of processes in a myocardial scar differs from that in a the persistence of MFs and their preferential align- dermal wound. ASMA-positive MFs persist in myo- ment in myocardial scars, in contrast to dermal scars, cardial scars and have a preferred orientation, is the continuous mechanical stress caused by the whereas the same cells are only transiently present in ongoing contraction and relaxation of the surrounding dermal wounds and without a preferred orientation. viable myocardium. This might also explain the ob- The difference is probably due to the different envi- served differences between patchy infarctions and ronments in which the healing process take place. 874 Willems et al AJP October 1994, Vol. 145, No. 4

Acknowledgment ventilation/perfusion ratio. J Cell Biol 1974, 60:375- 392 We thank Birgit Engels-Kuypers for excellent techni- 16. Stout LC, Boor PJ, Whorton EB Jr: Myofibroblastic pro- cal assistance. liferation on mitral valve chordae tendineae: a distinc- tive lesion associated with alcoholic liver disease. Hum Pathol 1988, 19:720-725 References 17. Pourreau-Schneider N, Ahmed A, Soudry M, Jacque- mier J, Kopp F, Franquin JC, Martin PM: Helium-neon 1. Baur PS Jr, Parks DH, Hudson JD: Epithelial mediated laser treatment transforms fibroblasts into myofibro- wound contraction in experimental wounds: the purse- blasts. Am J Pathol 1990, 137:171-178 string effect. J Trauma 1984, 24:713-720 18. Tomasek JJ, Schultz RJ, Episalla CW, Newman SA: 2. Gabbiani G, Ryan GB, Majno G: Presence of modified The cytoskeleton and extracellular matrix of the Du- fibroblasts in granulation tissue and their possible role puytren's disease "myofibroblast": an immunofluores- in wound contraction. Experientia 1971, 27:549-550 cence study of a nonmuscle cell type. J Hand Surg 3. Lipper S, Kahn LB, Reddick RL: The myofibroblast. Am 1986, 11:365-371 Pathol Ann 1980:409-441 19. Matsuoka LY, Uitto J, Wortsman J, Abergel RP, Dietrich 4. Skalli 0, Schurch W, Seemayer T, Lagace R, Montan- J: Ultrastructural characteristics of keloid fibroblasts. don D, Pittet B, Gabbiani G: Myofibroblasts from di- Am J Dermatopathol 1988, 10:505-508 in verse pathologic settings their content of actin iso- 20. Battersby S, Anderson TJ: Myofibroblast activity of ra- forms and intermediate filament proteins. Lab Invest dial scars. Am J Pathol 1985, 147:33-40 1989, 60:275-285 21. Vande-Berg JS, Rudolph R, Woodward M: Growth dy- 5. Darby I, Skalli 0, Gabbiani G: Alpha-smooth muscle namics of cultured myofibroblasts from human breast actin is transiently expressed by myofibroblasts during cancer and nonmalignant contracting tissues. Plastic experimental wound healing. Lab Invest 1990, 63: Reconstr Surg 1984, 73:605-618 21-29 22. Sottiurai VS, Batson RC: Role of myofibroblasts in 6. Vracko R, Thorning D: Contractile cells in rat myocar- formation. Surgery 1983, 94:792-801 dial scar tissue. Lab Invest 1991, 65:214-227 pseudointima 7. Seemayer TA, Lagace R, Schurch W, Thelmo WL: The 23. Leslie KO, Taatjes DJ, Schwarz J, von Turkovich M, myofibroblast: biologic, pathologic and theoretical Low RB: Cardiac myofibroblasts express alpha considerations. Pathol Ann 1980:443-470 smooth muscle actin during right ventricular pressure 8. Schurch W, Seemayer TA, Lagace R, Gabbiani G: The overload in the rabbit. Am J Pathol 1991, 139:207-216 intermediate filament cytoskeleton of myofibroblasts: 24. El-Labban NG, Lee KW: Myofibroblasts in central gi- an immunofluorescence and ultrastructural study. Vir- ant cell granuloma of the jaws: an ultrastructural study. chows Arch A 1984, 403:323-336 Histopathology 1983, 7:907-918 9. Bouchardy B, Majno G: Histopathology of early myo- 25. Foo ITH, Naylor IL, Timmons MJ, Trejdosiewicz LK: In- cardial infarcts: a new approach. Am J Pathol 1974, tracellular actin as a marker for myofibroblasts in vitro. 74:301-330 Lab Invest 1992, 67:727-733 10. Skalli 0, Ropraz P, Trzeciak A, Benzonana G, Gil- 26. Eddy RJ, Petro JA, Tomasek JJ: Evidence for the non- lessen D, Gabbiani G: A monoclonal antibody against muscle nature of the "myofibroblast" of granulation tis- alpha-smooth muscle actin: a new probe for smooth sue and hypertropic scar: an immunofluorescence muscle differentiation. J Cell Biol 1986, 103:2787-2796 study. Am J Pathol 1988, 130.252-260 11. van Muyen, Ruiter DJ, Warnaar SO: Coexpression of 27. Ghadially FN: Ultrastructural Pathology of the Cell and intermediate filament polypeptides in fetal and adult Matrix, vol 2, ed 3. Edited by FN Ghadially. London, tissue. Lab Invest 1987, 57:359-369 Butterworths, 1988, pp 872-877 12. Havenith MG, Cleutjens JPM, Beek C, vd Linden E, 28. Vracko R, Cunningham D, Frederickson RG, Thorning De Goeij AFPM, Bosman FT: Human specific anti-type D: Basal lamina of rat myocardium: its fate after death IV collagen monoclonal antibody, characterisation and of the cardiac myocytes. Lab Invest 1988, 58:77-87 immunohistochemical application. Histochemistry 29. Vracko R, Thorning D, Fredrickson RG: Connective tis- 1987, 87:123-128 sue cells in healing rat myocardium: a study of cell re- 13. Filip DA, Radu A, Simionescu M: Interstitial cells of the actions in rhythmically contracting environment. Am J heart valves possess characteristics similar to smooth Pathol 1989, 134:993-1006 muscle cells. Circ Res 1986, 59:310-320 30. Daemen MJAP, Lombardi DM, Bosman FT Schwartz 14. Adler KB, Low RB, Leslie KO, Mitchell J, Evans JN: SM: Angiotensin 11 induces smooth muscle cell prolif- Contractile cells in normal and fibrotic lung. Lab In- eration in the normal and injured rat arterial wall. Circ vest 1989, 60:473-485 Res 1991, 68:450-456 15. Kapanci Y, Assimacopoulos A, Irle C, Zwahlen A, 31. Desmouliere A, Rubbia-Brandt L, Grau G, Gabbiani G: Gabbiani G: Contractile interstitial cells in pulmonary Heparin induces alpha-smooth muscle actin expres- alveolar septa. Ultrastructural, immunofluorescence sion in cultured fibroblasts and in granulation tissue and in vitro studies: a possible regulation of myofibroblasts. Lab Invest 1992, 67:716-726 MFs in Human Myocardial Scars 875 AJP October 1994, Vol. 145, No. 4

32. Pierce GF, Vande Berg J, Rudolph R, Tarpley J, Mus- smooth muscle actin expression in granulation tissue toe TA: Platelet-derived growth factor-BB and trans- myofibroblasts and in quiescent and growing cultured forming growth factor-beta 1 selectively modulate gly- fibroblasts. J Cell Biol 1993, 122:103-111 cosaminoglycans, collagen, and myofibroblasts in 38. MacLellan WR, Brand T, Schneider MD: Transforming excisional wounds. Am J Pathol 1991, 138:629-646 growth factor-beta in cardiac ontogeny and adapta- 33. Amberger A, Bauer H, Tontsch U, Gabbiani G, Kocher tion. Circ Res 1993, 73:783-791 0, Bauer HC: Reversible expression of sm alpha actin 39. Engelman GL, Boehm KD, Birchenall-Roberts MC, protein and sm alpha actin mRNA in cloned cerebral Ruscetti FW: Transforming growth factor-beta 1 in endothelial cells. FEBS Lett 1991, 287:223-225 heart development. Mech Dev 1992, 38:85-97 34. Bachem MG, Sell KM, Melchior R, Kropf J, Eller T, 40. Heine Ul, Burmester JK, Flanders KC: localization of Gressner AM: Tumor necrosis factor alpha (TNF-al- transforming growth factor-beta 1 in mitochondria of pha) and transforming growth factor beta-1 (TGF- murine heart and liver. Cell Regul 1991, 2:467-477 beta-1) stimulate fibronectin synthesis and the trans- 41. Thompson NL, Bazoberry F, Speir BH: Transforming differentiation of fat-storing cells in the rat liver into growth factor-beta 1 in acute myocardial infarction in myofibroblasts. Virchows Arch B Cell Pathol 1993, 63: rats. Growth Fact 1988, 1:91-99 123-130 42. van-Krimpen C, Schoemaker RG, Cleutjens JP, Smits 35. Thiemermann C, Corder R: Is endothelin-1 the regula- JF, Struyker-Boudier HA, Bosman FT, Daemen MJ: An- tor of myofibroblast contraction during woundhealing? giotensin converting enzyme inhibitors and cardiac Lab Invest 1992, 67:677-679 remodeling. Basic Res Cardiol 1991, 86 (Suppl 1): 36. Vyalov S, Desmouliere A, Gabbiani G: GM-CSF- 149-155 induced granulation tissue formation: relationships 43. Volders PGA, Willems IEMG, Cleutjens JPM, Arends between macrophage and myofibroblast accumula- JW, Havenith MG, Daemen MJAP: Interstitial collagen tion. Virchows Arch B Cell Pathol 1993, 63:231-239 is increased in the non-infarcted human myocardium 37. Desmouliere A, Geinoz A, Gabbiani F, Gabbiani G: after myocardial infarction. J Mol Cell Cardiol 1993, Transforming growth factor-beta 1 induces alpha- 25:1317-1323