MOLECULAR REPRODUCTION AND DEVELOPMENT 32~121-126 (1992)

Modulation of Vascular Cell Behavior by Transforming Growth Factors p JOSEPH A. MADRIl.’, LEONARD BELL’.’, AND JUNE RAE MERWIN’ ‘Departments of Pathology, ‘Internal Medicine-Cardiology, Yale University School New Haven, Connecticut

ABSTRACT The vascular cell responses to the type 1, 2, and 3 isoforms of transforming growth factor-p (TGF-p1, TGF-p2,TGF-p3) were studied using bovine aortic endothelial (BAECs)and smooth muscle cells (BASMC3) as well as rat epididymal fat pad microvascular endothelia (RFCs). Three distinct bioassays indicated that TGF-p3 elicits results that do not differ significantly from those of the TGF-p1 isoform in all three cell populations. These assays are: inhibition of proliferation, cell migration, and neovascularization. By contrast the cellular responses to TGF-p1 and TGF-p3 differed from those to TGF-p2. Three distinct receptor assays revealed the preesnce of type 1 and type II TGF-p1 cell surface binding proteins on BAECs, BASMCs, and RFCs. Experimentation to decipher cell surface binding by the different isoforms revealed that iodinated TGF-p1 bound to the surface of all three vascular cell types can be competed off in similar fashion by either TGF-p1 or TGFp3; however, competition with TGF-p2 produced unique binding profiles dependent on the cell type examined. The ratios of type I to type II TGF-p receptors in these three vascular cell types vary from 1:l in BAECs to 1.5:l in RFCs to 3:l in BASMCs and can be correlated with the differences noted in cellular responses to TGF-p1 and TGF-p2 in proliferation, migration, and in vitro angiogenic assays. In summary, both the TGF-p1 and TGF-p3 isoforms of the transforming growth factor-p family evoke comparable responses in proliferation, migration, angiogenic and cell surface bindinga ssays using three distinct vascular cell types, while the biofunctions of TGF-p2 on these cells are distinct. These findings support the hypothesis that there are different responses to the TGF-ps depending on the cell type and experimental conditions as well as the TGF-P concentrationand isoform. 0 1992 Wiley-Liss, Inc. Key Words: Vascular cells, TGF-p, Cell surface binding

INTRODUCTION

of

Medicine,

two-thirds of saphenous vein bypass grafts will be either totally occluded or severely compromised after 7-11 years (Bourassa et al., 1985).Therefore it’s of major importance to better understand the responses of vascular cells to injury (endothelial and smooth muscle proliferation, migration and matrix metabolic profiles) if we wish to beneficially modulate these responses to affect maximal healing with minimal luminal restenosis. The two major cell types of the vessel wall are the endothelial cell and the smooth muscle cell. They respond to injury (angioplasty, endarterectomy, synthetic and saphenous vein grafting) in very different ways and influence each other’s responses through multiple, complex pathways. The vascular system is lined by mitotically quiescent, metabolically active endothelial cells, which provide a nonthrombogenic surface for blood flow, as well as having a broad range of metabolic activities. Smooth muscle cells, found beneath the endothelium in the media of large vessels, play significant roles in maintaining vessel wall integrity, including controlling vascular tone, influencing endothelial cell behavior and maintaining the connective tissues of the vessel wall (Madri et al., 1991,1992).Vascular cells (large vessel-derived and microvascular endothelial and smooth muscle cells) have been found t o respond to injury in specific ways, depending upon the vascular bed and the cell type(s) injured. Here we will address vascular cell responses to iatrogenic injuries such as those occurring to vessel walls during angioplasty, autologous and synthetic grafting, endarterectomy and soft tissue injury. Following the denudation injury, large vessel endothelial cells adjacent to affected areas exhibit rapid sheet migration over the exposed subendothelial extracellular matrix and proliferate in an attempt to reconstitute a continuous endothelial cell lining (Madri et al., 1987, 1988b). Medial smooth muscle cells of large and medium-sized vessels respond to denudation injury by proliferation and migration into the intima, where they synthesize matrix components, which results in the formation of an expanded neointimal compartment that narrows and can eventually occlude the vessel lumen (ROSS, 1988).

Atherosclerosis is a major health concern in the Western world. In 1986, more than 250 million patient discharges for angina were reported, and 900,000 patients were discharged with the diagnosis of acute myocardial infarction (Feinleib et al., 1989). In addition, approximately 400,000 patients required coronary revascularization in 1986; and it’s predicted that 3 0 4 0 % Address reprint requests to Dr. Joseph A. Madri, Department of Paof lesions treated with balloon angioplasty will reste- thology, Yale University School of Medicine, 310 Cedar Street, New nose in 3-6 months (Beatt et al., 1990), while up to Haven, CT 06510.

0 1992 WILEY-LISS, INC.

122

J.A. MADRI ET AL.

Here we will review the responses to injury of large vessel and microvascular endothelial and smooth muscle cells and their modulation by the presence of selected isoforms of transforming growth factor p.

SOLUBLE FACTORS MODULATE AORTIC ENDOTHELIAL AND SMOOTH MUSCLE CELL MIGRATION IN VITRO AND IN VIVO Angioplasty, autologous and synthetic grafting, endarterectomy, atherectomy, and laser ablation) all result in de-endothelialization and medial injury of vascular segments and subsequent platelet adhesion, aggregation, and release of soluble factors. Platelet releasate has been proposed to play a n important role in the development of the restenotic, occlusive lesions observed to occur following these therapeutic interventions (Bell and Madri, 1989; Madri et al., 1991, 19921, since it contains growth factors and vasoactive agents, such as PDGF, TGF-P1, serotonin, norepinephrine and histamine, which modulate vascular endothelial and smooth muscle cell chemotaxis, migration, and proliferation.

Soluble Factor Effects on Endothelial Cell Behavior In recent studies we have explored the effects and the mechanisms of action of TGF-p1 on large vessel endothelial cell migration in vitro. TGF-P1, in addition to being a major component of platelet releasates is also synthesized and secreted by endothelial and smooth muscle cells and therefore is likely involved in autocrine and paracrine regulatory modulation of vascular cell behavior (Antonelli-Orlidge et al., 1989; Roberts and Sporn, 1990). TGF-pl-treated bovine aortic endothelial cells have been found to exhibit transiently increased fibronectin mRNA levels and also exhibit increased synthesis and deposition of fibronectin (Madri et al., 1989b). Furthermore, using differential library screening techniques, Southern and Northern blots and PCR analyses comparing cultured bovine aortic endothelial cells grown in the absence and presence of TGFPl, we documented differences in the mRNA levels for selected alternatively spliced isoforms of this molecule (Kocher et al., 1990). Specifically, changes in the relative amounts of mRNA species coding for isoforms spliced in the IIICS region of fibronectin were observed. In untreated, control cultures, mRNA isoforms of fibronectin contained predominately both ED domains and either the CS1 or the CS1 and CS5 spliced domains of the IIICS domain. In the presence of TGF-P1, however, cultures produced increased amounts of fibronectin mRNAs containing both ED domains and the complete unspliced IIICS domain with lesser amounts of the CSl or CS1 and CS5 spliced regions of the IIICS domain (Fig. 1)(Kocher et al., 1990). Additionally, we have shown that the proteins coded for by these mRNA species modulate bovine aortic endothelial cell behavior. Namely, endothelial migration is decreased and spreading is increased in the presence of fibronectin molecules containing the CS5 and CS1 regions of the

TGF-P1 Fig. 1. Southern blot analysis of selected cDNA clones after digestion with EcoRI, hybridized with radiolabeled cDNAs from (A) untreated ($1 or (B) TGF-pl treated (TGF-p1)bovine aortic endothelial cells. The first lane (starting at the left) illustrates an example of an as-yet-unidentified nonmodulated clone as a control. The next lane (labeled #lo) illustrates the dramatic up-regulation of the mRNA isoform coding for fibronectin containing the complete IIICS region following TGF-P1 stimulation. The next two lanes (labeled #14 and #54) illustrate moderate up-regulation of the mRNA isoforms coding for fibronectin containing either the IIICS region spliced in the CS1 domain (#14)or both in the CS1 and CS5 domains (#54) in response to TGF-p1 treatment.

IIICS domain. This is consistent with the observed biological effects of TGF-P1 on bovine aortic endothelial cells, which include decreased migration rates, increased cell spreading and increased fibronectin mRNA isoforms coding for fibronectin species containing the complete IIICS region and increased fibronectin protein deposition (Madri et al., 198913; Kocher et al., 1990). The effects of TGF-P1 on bovine aortic endothelial cell migration could be mimicked by either using fibronectin as a substratum or adding soluble fibronectin to the migrating cultures suggesting that the TGF-pl effect on migration is mediated, in part, by modulation of the matrix synthetic profile of the vascular endothelial cells. Recently, we have demonstrated that TGF-P1 also elicits increases in the mRNA and cell surface protein expression of the a5Pl and P3 integrin chains and plasminogen activator inhibitor and decreases in plasminogen activator levels (Basson et al., 1990,1992; Bell and Madri, 1992) consistent with the notion that the changes in endothelial cell behavior elicited by TGF-P1 are due to complex integrative modulation of extracellular matrix components, adhesion proteins, cytoskeletal components and proteases.

MODULATION OF VASCULAR CELL BEHAVIOR BY TGF Soluble Factor Effects on Smooth Muscle Cell Behavior Endothelial loss following denudation injury and its subsequent reconstitution is a critical factor in the maintenance of luminal patency. The responses of vascular smooth muscle cells following denudation injury is equally important (Fishman, 1982; Munro and Cotran, 1988; Ross, 1986). Stimulated by platelet releasate (TGF-P1, PDGF, serotonin, norepinephrine, heparin), and possibly by endothelial-drived factors (angiotensin 11,TGF-p1, PDGF), as well as by autocrine and paracrine factors derived from local smooth muscle cell populations in the area of injury, smooth muscle cells in proximity to the lesion respond by migrating into the injured area, synthesizing and depositing extracellular matrix components and proliferating. This response leads to intimal thickening and subsequently restenosis and occlusion (Fishman, 1982; Munro and Cotran, 1988; Ross, 1986; Hedin et al., 1988; Madri et al., 1990).We have found that many of the same soluble factors that inhibit endothelial cell migration enhance vascular smooth muscle migration (Madri et al., 1989a,b, 1991; Bell and Madri, 1990; Basson et al., 1992). Changes in extracellular matrix component synthesis and deposition, integrin expression, and protease activity have been documented following stimulation with TGF-P1 and likely contribute to the changes in migratory rate (Basson et al., 1992). Specifically, our studies have revealed that in addition to increasing bovine aortic smooth muscle cells migratory rate, TGF-P1 treatment increases fibronectin mRNA and protein deposition in confluent and migrating bovine aortic smooth muscle cells cultures. In addition we have noted that TGF-P1 treatment of cultured bovine aortic smooth muscle cells causes an increase in the mRNA levels and surface expression of p3 (but not pl) integrin chains, supporting the concept that soluble autocrine and paracrine factors such as TGF-P1 may elicit their effects on smooth muscle cell migration, in part, uia the selective modulation of matrix component(s) and cell surface matrix receptors including the integrins (Madri et al., 1990; Basson et al., 1992). In contrast to bovine aortic endothelial cells, bovine aortic smooth muscle cell plasminogen activator levels are increased in response to TGF-p1 (Basson et al., 1992; Bell and Madri, 1992).These data, support the concept that the changes in vascular smooth muscle cell behavior elicited by TGF-p1 are due to complex integrative modulation of extracellular matrix components, adhesion proteins, cytoskeletal components, and proteases as noted in the case of aortic endothelial cells (Fig. 2). In Vivo Studies: The Roles of TGF-p1 in Modulating Vascular Cell Responses to Injury In Vivo We have also investigated vascular cell-extracellular matrix, vascular cell-TGF-p1 and cell-cell interactions in a rat carotid balloon de-endothelialization

123

DENUDATION INJURY

4

+

EXPOSURE OF SUBENDOTHELiAL MATRIX PLATELET ADHESION, AGGREGATION AND RELEASE of TGF-beta ENWTHELIAL AND S M L T H MUSCLE AUTOCRiNE AND PARACRiNE PRODUCTION Of TGF-beta

A EC Altered ECM synthesis. Fn EC Altered lntegrin Expression

4

EC Altered Platminogen Activator ' Expression + EC Proliferation EC Migration

t

EC Activation

4

4 b -

SMC Altered lntegrin Expression

wc

SMC Prollferation SMC Migration

4

4

SMC Activation

J

Monocyte Adhesion and Migration via: a401NCAM-1 and a4pl/IIICS of Fn

Fig. 2. Schematic representation of the modulation of medial smooth muscle cell (SMC) and endothelial cell (EC) behavior following denudation injury by TGF-P1. In this scheme, TGF-P1 elicits changes in matrix production, integrin expression, and plasminogen activator activity, which in turn modulate endothelial and smooth muscle cell migration and proliferation.

model in which there is incomplete re-endothelialization of the initially de-endothelialized area, and increased fibronectin and TGF-p1 staining throughout the media and luminal surface of the chronically deendothelialized region of the injured vessel (Madri et al., 1989). More recent studies using this in vivo model have also revealed distinct differences in the localization of p l and P3 integrins in the neointimal regions and luminal lining cells of both the chronically de-endothelialized and re-endothelialized areas (Basson et al., 1992). Namely, in chronically de-endothelialized areas, the neointimal smooth muscle cells exhibited increased staining intensity for P3 integrins with no changes in p l integrins compared with uninjured and post-injury medial smooth muscle cells. These findings are consistent with the effects of TGF-P1 on bovine aortic smooth muscle cells noted in vitro (Basson et al., 1992). In contrast, in re-endothelialized areas, the neointimal smooth muscle cells nearer the lumen displayed much less intense staining for p3 integrins (and for fibrinogen) than those deeper in the neointima, supporting the concept of endothelial cell modulation of smooth muscle phenotype (Basson et al., 1992). Thus, both in vivo and in vitro, the presence of TGF-p1 appears t o have profound effects on the migratory and matrix synthetic behavior of large vessel endothelial and medial smooth muscle cells.

TGF-pl: A MODULATOR OF MICROVASCULAR ENDOTHELIAL CELL DIFFERENTIATION Large vessel endothelial cells undergo sheet migration in response to denudation injury, in contrast, mi-

124

J.A. MADRI ET AL.

crovascular endothelial cells respond to injury by disrupting their investing basement membranes, migrating into and proliferating in the surrounding three-dimensional interstitial stroma, forming new microvessels, stabilizing them and ultimately dismantling the newly formed microvascular bed following the conclusion of the repair response (Madri and Pratt, 1988). In studying the complex process of angiogenesis and microvascular endothelial cell differentiation several in vitro culture systems have been employed (Madri and Williams, 1983; Madri and Pratt, 1986; Madri et al., 1988a; Nicosia and Madri, 1987); including the culture of microvascular endothelial cells from the rat epididymal fat pad and bovine calf adrenal cortex on selected extracellular matrix components in two-dimensional culture, on the interstitial and basement membrane aspects of the amnion, and in three-dimensional type I collagen gels. Under all conditions studied these microvascular endothelial cells retain vonWillebrand factor positivity. Microvascular endothelial cells, in their three-dimensional environment, have a significant arc of curvature and intimate associations with specific matrix basement membrane components and pericytes. When they are placed in two-dimensional culture, they are placed in a n environment in which they have no arc of curvature, dramatically different associations with the substratum, and no contacts or associations with pericytes. Microvascular endothelial cells cultured in this manner display a high proliferative rate and a relatively undifferentiated phenotype, including the expression of a-smooth muscle actin mRNA and protein and PDGF receptors and PDGF AB and AA isoform responsiveness consistent with pericytic as well as endothelial differentiation (Madri et al., 1988a, 1991; Madri and Marx, 1992). The addition of TGF-61 elicits a dramatic inhibition of proliferation and a n eight-fold increase in a-smooth muscle actin mRNA and protein. The change in environment from their normal three-dimensional one to a two-dimensional one contributes to the loss of their differentiated phenotype in culture. Placement of cultured microvascular endothelial cells into a three-dimensional type I collagen gel, with the addition of TGF-p1, p3, or -p2 a t higher concentrations, elicits a very low proliferate rate which is unchanged by the addition of TGF-P isoforms, loss of a-smooth muscle actin expression and loss of PDGF receptor expression and responsiveness. In addition, under these culture conditions the formation of tubelike structures having junctional complexes and abluminal basal lamina formation consistent with the re-institution of a differentiated microvascular endothelial phenotype occurs (Madri and Williams, 1983; Pratt et al., 1985; Merwin et al., 1990a, 1991a-c). These changes are consistent with a modulation of phenotype, driven, in part, by the organization of the extracellular matrix (Kocher and Madri, 1989; Madri and Williams, 1983; Madri et al., 1989a, 1990; Merwin e t al., 1990a; Pratt et al., 1985). These data also are consistent with

the concept that the organization (architecture) of the surrounding matrix drives the cells to a differentiated endothelial phenotype having markedly different responses (and cell surface receptor repertoire) to cytokines produced during inflammatory, repair, and angiogenic processes. The induction of in vitro angiogenesis by TGF-P isoforms in three-dimensional cultures provided us with the opportunity to investigate the process of angiogenesis (tube formation) and the roleb) of TGF-P isoforms in modulating this process. While in two-dimensional culture TGF-P isoforms function a s modulators of mitotic rate, dramatically affecting the proliferative rates of cultured microvascular endothelial cells; in three-dimensional culture TGF-P isoforms appear to have a morphogenic function, eliciting the multicellular organization of microvascular endothelial cells in complex, branching tube-like structures having luminal and abluminal specializations (Madri et al., 1988a; Merwin et al., 1990a,b, 1992). In three-dimensional cultures TGF-P1 elicits profound morphogenic effects including rapid cell-cell contact mediated by platelet/endothelial cell adhesion molecule-1 (PECAM-1) and rapid tube formation apparently mediated by PECAM-PECAM and p l integrin-extracellular matrix interactions (Merwin et al., 1991a, 1992). This early phase of the angiogenic process can be inhibited completely by functional antibodies directed against PECAM-1 and p l integrin and RGDcontaining peptides. Later in the angiogenic process, basement membrane components are secreted and deposited abluminally, P l integrins organize in areas of cell-cell contact and tight junctions form. At this later stage in the angiogenic process antibodies directed against PECAM are ineffective in altering tight junction formation, while antibodies directed against P l integrin and RGD-containing peptides are effective in preventing tight junction formation, but do not cause disruption of already-formed tubes (Merwin et al., 1991a, 1992) (Fig. 3). Thus, TGF-P1 appears to initiate a complex, hierarchical, multistep process, involving the assembly and surface organization of selected cell adhesion molecules, substrate adhesion molecules and junction associated molecules as well as the abluminal deposition of selected matrix molecules. The mechanism(s) involved in these processes are currently unknown.

DIFFERENTIAL EFFECTS OF TGF-PI, TGF-P2, AND TGF-P3 ON VASCULAR CELL BEHAVIOR TGF-P1 is one member of a family of polypeptides, including the homodimers TGF-P2 and -p3, found in mammals. Their differential effect(s1 on vascular cell behavior (including proliferation, migration, extracellular matrix synthesis, protease and protease inhibitor synthesis, and tube formation) may prove to be important in modulating the processes of vascular development, wound healing and atherosclerosis (Merwin et al., 1990a,b, 1991a-q Bell and Madri, 1992). Previous

MODULATION OF VASCULAR CELL BEHAVIOR BY TGF TGF-P1 Induced Angiogenesis: The Roles of CAMS, SAMs and JAMS

-

A

Cell-cell Interactions

Cell-Matrix Interactions

I

03 lntearins ~ E ~ ~ ~ r Tight i n Junction s Formation

6

zo1

ntegrin anization

c---

Fig. 3. Model for cell adhesion molecule (CAM),substrate adhesion molecule (SAM), and junction associated molecule (JAM) hierarchical expression and organization during TGF-pl induced in vitro angiogenesis. This figure illustrates putative temporospatial relationships of selected CAMS(PECAM-11,SAMS ( p l Integrins), and JAMS (ZO-1) in three-dimensional cultures of microvascular endothelial cells. TGF-p1 induces cell-cell contact initially via PECAM-1 homotypic interactions and integrin-xtracellular matrix interactions. Following this stage of the in vitro angiogenesis process pl integrins exhibit abluminal organization in areas of cell-cell contact which likely serve to stabilize the newly formed tube-like structures. Following stabilization of the tube-like structures by the p l integrins, tight junction formation occurs, which correlates with ZO-1 assembly and organization. (Based on data from Merwin et al., 1990b, 1992.)

studies on large vessel endothelial cells have revealed that TGF-p2 had no effect on proliferation of large vessel endothelial cells (Merwin et al., 1991a) and that TGF-02 did not compete with TGF-P1 in functional assays (proliferation). TGF-p2 did compete effectively for TGF-P1 receptor binding, which was assayed using morphological (competition binding with biotinylated TGF-P1 and avidin gold localization) and FACS (competition binding with biotinylated TGF-P1 and fluorescein-conjugated avidin) assays. However, TGF-P2, a t high concentrations (25.0 ng/ml) was found to modestly inhibit bovine aortic endothelial cell migration. In contrast, microvascular endothelial cell proliferation in two-dimensional cultures is inhibited by TGF-P2 a t all concentrations tested (0.005-5.0 ng/ml), but at levels lower than those observed with TGF-p1. We have also observed that TGF-P2 enhances in vitro angiogenesis (tube formation) in three-dimensional cultures of microvascular endothelial cells, again a t concentrations ten-fold higher than that observed for TGF-PI. Lastly, the proliferation of cultured bovine aortic medial smooth muscle cells is inhibited by TGF-P2 at all concentrations (0.5-5.0 ng/ml) tested to a degree similar to TGF-P1, while smooth muscle cell migration was unaffected by TGF-p2 at all concentrations tested (0.055.0 ng/ml) (Merwin et al., 1991a). These data suggest that vascular cells derived from diverse vascular beds exhibit individual differential sensitivities to various isoforms of TGF-P and this differential sensitivity and the broad range of effects may have significance in the

125

regulation of vascular development, inflammatory and healing responses of the vascular system. In contrast to TGF-P2, TGF-p3 was identical to TGF-P1 in its ability to bind to bovine aortic endothelial, smooth muscle cells, and microvascular endothelial cells derived from the rat epididymal fat pad and to compete off TGF-61 in these cell types. In addition, TGF-P3 was identical to TGF-P1 in its ability to modulate proliferation, migration, and tube formation in these cell types. To develop a n understanding of the mechanisms involved in producing this complex differential response, further investigation is needed to isolate and characterize the TGF-p receptors and address the TGF-P signaling pathways, which are currently incompletely understood. The vascular cells we have studied each express different ratios of type I to type I1 TGF-P receptors (Merwin et al., 1991a). Whether these observed differences in receptor type ratios mediate the behavioral differences or just correlate with the different cellular responses is still unknown. The successful cloning and characterization of the type I and type I1 TGF-P receptors will provide powerful tools in helping the scientific community better understand the complex cellular responses to the various TGF-p isoforms.

ACKNOWLEDGMENTS This work was supported in part by USPHS grants R01-HL-28373 (to J.A.M.) and Physician Scientist Award Kll-HL-02351 (to L.B.). REFERENCES Antonelli-Orlidge A, Saunders KB, Smith SR, D’Amore PA (1989):An activated form of transforming growth factor beta is produced by cocultures of endothelial cells and pericytes. Proc Natl Acad Sci USA 86:4544-4548. Basson CT, Knowles WJ, Abelda S, Bell L, Castronovo V, Liotta LA, Madri J A (1990): Spatiotemporal segregation of endothelial cell integrin and non-integrin extracellular matrix binding proteins during adhesion events. J Cell Biol 110789-802. Basson CT, Kocher 0, Basson MD, Asis A, Madri J A (1992): Modulation of vascular cell integrin expression in vitro by TGF-p1 and PDGF. Submitted. Beatt KJ, Erruys P, Hugenholtz PJ (1990): Restenosis after coronary angioplasty: New standards for clinical studies: Am Coll Cardiol 15:491-498. Bell L, Madri J A (1989): Effect of platelet factors on migration of cultured bovine aortic endothelial and smooth muscle cells: Circ Res 65:1057-1065. Bell L, Madri J A (1990): Influence of the angiotensin system on endothelial and smooth muscle cell migration in vitro. Am J Pathol 137:7-1 2. Bell L, Madri J A (1992): Differential modulation of bovine aortic endothelial cell fibronectin and plasminogen activator expression by TGF-pl and TGF-62. Submitted. Bourassa M, Fisher L, Campeau L, Gillespie M, McConney M, Lesperance N (1985): Long-term fate of bypass grafts: the coronary artery surgery study (CASS) and Montreal Heart Institute experiences. Circulation 72:V71-V78. Feinleib M, Havlik R, Gillum R, Pokras R, McCarthy E, Moien M (1989): Coronary heart disease and related procedures. National hospital discharge survey data. Circulation 79:113-118. Fishman AP (ed) (1982): “Endothelium.” New York: The New York Academy of Sciences. Hedin U, Bottger BA, Forsberg E, Johansson S, Tyberg J (1988): Diverse effects of fibronectin and laminin on phenotypic properties of cultures arterial smooth muscle cells. J Cell Biol 107:307320.

126

J.A. MADRI ET AL.

Kocher 0, Madri J A (1989): Modulation of actin mRNAs in cultured vascular cells by matrix components and TGF-p1. In Vitro 25424434. Kocher 0, Kennedy S, Madri J A (1990): Alternative splicing of fibronectin mRNA in the IIICS region in endothelial cells: Functional significance. Am J Pathol 137:1509-1524. Madri J A (1987): The extracellular matrix as a modulator of neovascularization. In M Gallo (ed): “Cardiovascular Disease: Molecular and Cellular Mechanisms, Prevention, Treatment.” New York: Plenum Press, p 177. Madri JA, Marx M (1992): Matrix composition, organization and soluble factors: Modulators of microvascular cell differentiation in vitro. Kidney Int 41560-565. Madri JA, Pratt BM (1986): Endothelial cell-matrix interactions: In vitro models of angiogenesis, J Histochem Cytochem 3435-91. Madri JA, Pratt BM (1988): Angiogenesis. In RF Clark, P Henson (eds): “The Molecular and Cellular Biology of Wound Repair.” New York: Plenum Press, pp 337-358. Madri JA, Williams SK (1983): Capillary endothelial cell cultures: Phenotypic modulation by matrix components. J Cell Biol 97:152165. Madri JA, Pratt BM, Tucker AM (1988a): Phenotypic modulation of endothelial cells by transforming growth factor-p depends upon the composition and organization of the extracellular matrix. J Cell Biol 106:13751384. Madri JA, Pratt BM, Yanniarello-Brown J (1988b): Matrix-driven cell size changes modulate aortic endothelial cell proliferation and sheet migration. Am J Pathol 132:1&27. Madri JA, Kocher 0, Merwin JR, Bell L, Yannariello-Brown J (1989a): The interactions of vascular cells with solid phase (matrix) and soluble factors. J Cardiovasc Pharmacol14S7oS75. Madri JA, Reidy MA, Kocher 0, Bell L (1989b): Endothelial cell behavior following denudation injury is modulated by TGF-P1 and fibronectin. Lab Invest 60:75&765. Madri JA, Kocher 0, Merwin JR, Basson CT, Bell L (1990): The interactions of vascular cells with transforming growth factor p. In KA Piez, MB Sporn (eds): “Transforming Growth Factor-ps: Chemistry, Biology and Therapeutics.” Ann NY Acad Sci 593:243-258. Madri JA, Bell L, Marx M, Menvin JR, Basson CT, Prinz C (1991):The effects of soluble factors and extracellular matrix components on vascular cell behavior in vitro and in vivo: Models of de-endothelialization and repair. J Cell Biochem 45:l-8. Madri JA, Merwin JR, Bell L, Basson CT, Kocher 0, Perlmutter R, Prinz C (1992): Interactions of matrix components and soluble factors in vascular cell response to injury: Modulation of cell phenotype. In N Simionescu, M Simionescu (eds): “Endothelial Cell Dysfunction.” New York: Plenum Press, pp. 11-30. Menvin JR, Anderson J, Kocher 0, van Itallie C, Madri J A (1990a): Transforming growth factor p l modulates extracellular matrix organization and cell-cell junctional complex formation during in vitro angiogenesis. J Cell Physiol 142:117-128. Merwin JR, Tucker A, Albelda SM, Madri J A (1990b): CAMS, JAMS and SAMs-Expression in microvascular endothelial cells. J Cell Biol lll:157a (abst). Menvin JR, Newman W, Beall D, Tucker A, Madri J A (1991a):Vascular cells respond differentially to transforming growth factors-beta, and beta,. Am J Pathol13837-51. Merwin JR, Tucker A, Madisen L, Purchio A, Madri J A (1991b):Vascular cell responses to a hybrid transforming growth factor beta molecule. Biochem Biophys Res Commun 175589495. Merwin JR, Tucker A, Roberts A, Kondaiah P, Madri J A (1991~): Vascular cell responses to transforming growth factor beta, mimic those of transforming growth factor beta,. In Vitro Growth Factors 5149-158. Merwin JR, Tucker A, Albelda SM, Madri J A (1992):The cell adhesion molecule PECAM-1 mediates early adhesive events during in vitro angiogenesis. In Preparation. Munro JM, Cotran RS (1988): The pathogenesis of Atherosclerosis: Atherogenesis and inflammation. Lab Invest 58249-253.

Nicosia RF, Madri J A (1987): The microvascular extracellular matrix: Developmental changes during angiogenesis in the aortic ringplasma clot model. Am J Pathol 128:7&90. Pratt BM, Form D, Madri J A (1985): Endothelial cell-extracellular matrix interactions. In R Fleishmajer, B Olsen, K Kuhn (eds):“Biology, Chemistry and Pathology of Collagen.” Ann NY Acad Sci 460:274-288. Roberts AB, Sporn MB (1990): The transforming growth factor-betas. Peptide growth factors and their receptors. In MB Sporn, AB Roberts (eds):“Handbook of Experimental Pharmacology.” Heidelberg: Springer-Verlag, 95:419472. Ross R (1986): Medical progress: The pathogenesis of atherosclerosis-An update. N Engl J Med 314:488-500. Ross R (1988): “Endothelial injury and atherosclerosis. In N Simionescu, M Simionescu (eds): “Endothelial Cell Biology in Health and Disease.” New York: Plenum Press, pp 371384.

QUESTIONS AND ANSWERS Q: Do you find any differences in the response of endothelial cells to the three isoforms of TGF-P? A: TGF-P2 does not alter fibronectin mRNA levels or the deposition of fibronectin in endothelial cultures. It also does not affect proliferation or migration of cultured large vessel endothelial cells. However, TGF-P2 does down-regulate plasminogen activator activity in endothelial cells. TGF-63 mimics TGF-P1 in all functional assays performed on endothelial cells. In smooth muscle cell cultures TGF-P2 has a n equipotent effect on proliferation as TGF-P1. However, it does not affect migration and does not affect fibronectin levels. Q: Could you elaborate on the observation of TGF-ps in the in vivo lesion, and did you look for TGF+2 as well as TGF-PI? A: In collaboration with Dr. Michael Reidy, we found TGF-P1 (using immunofluorescence) in the smooth muscle cells of the neo-intima in our balloon de-endothelialization model. We did not perform in situ hybridization studies for assessing TGF-P1 or TGF-P2 in tissue samples. In other studies, Dr. Michael Reidy has presented data in this area. Q: Have you investigated the receptors for the various TGF-P isoforms in the endothelial cells? A: Using lZ5ITGF-P1 crosslinking methods, we have determined the type I to type I1 receptor ratios for our cultured large vessel endothelial and smooth muscle cells and microvascular endothelial cells and have found them to be different. At this time we do not know the functional significance (if any) of these different ratios. Q: Have you tested the effect of different matrices on the differentiation in the three-dimensional culture system? A: We have doped our type I collagen gels with varying amounts of fibronectin and did not note any appreciable differences in the amount of tube formation, the branching or the time that tube formation takes to occur. We have not tested any other matrix components in our type I collagen gel culture system at this time.