Vol.

166,

No.

May

16,

1990

3, 1990

BIOCHEMICAL

AND

BIOPHYSICAL

J. Resink,

Alfred

W.A.

Hahn, Timothy and Fritz

Department of Research, University *Center

for

Experimental

Therapeutics, **Hoffmann-La

Received

COMMUNICATIONS

Pages

INDUClBLEENDOTHELINmRNAEXPRESSIONANDPEPTIDESECRETIONINCULTURED VASCULAR SMOOTH MUSCLE

‘IN&se

RESEARCH

March

29,

1303-1310

HUMAN CELLS

Scott-Burden*,Jerry

Powell**,

Erika

Webe.r

R Biihler Hospital,

Baylor Roche,

College Basel,

4031

Basel,

Switzerland

of Medicine,

Houston,

TX 77030

Switzerland

1990

This study demonstrates the induction of endothelin (ET) mRNA expression and synthesis of functional ET -peptide iu cultured human vascular smooth muscle cells (hVSMC). Compounds eliciting such responses in hVSMC include the vasoconstrictor hormones angiotensin II and arainine-vasouressin and the growth factors transformiug growth factor l3, platelet derived growth factor AA &d epidermh growth facto; Induction of ET mRNA expression iu hVSMC exhibited transient kinetics (peak at 3-5 hrs. and return to basal within 7 hrs.) which differed from the more sustained ET transcript induction observed for porcine endothelial cells. ET’peptide (determined by both radioimmtmo-and radioreceptor assays) produced by stimulated hVSMC attained levels I- 120-160 ~e/lO~ cells/4 hrs.: concentration -3 x 10-r’ Mf within the biolouicallv effective concentration range of ET: Stimulated secretion of ET from hVSMC was’abolished in the presence of the protein synthesis inhibitor cycloheximide. Sep-pak C!8 extracts of medium from stimulated hVSMC elicited a concentration-dependent phosphoinositide catabohc response in myo-[2-3H]-iuositol-prelabelled hVSMC. Our findings invoke a role for ET which extends beyond the paracrine regulation by peptide synthesized and secreted by endothelial cells. We propose that VSMC-synthesized ET may function in an autocrine manner to regulate both tone and structural modelling of vasculature. Q 1990 Academic Press, Inc.

Although endotherm (ET) was originally characterized as a potent vasoconstrictor peptide synthesized and secreted by cultured porcine endothelial cells (l), evidence is accumulating to support ET production by nonendothelial

cells. In situ hybridization,

immunocytochemistry

and radioimmunoassay

techniques have

demonstrated constitutive expression of ET transcripts and peptide secretion in neurons of human spinal cord and dorsal root ganglia (2) in the paraventricular nucleus of porcine hypothalamus (3) as well as in cultured kidney cells (4) and cultured epithelial cells from human breast (5) and canine trachea (6). Levels of ETmRNA thrombii

and peptide production are, for endothelial cells, regulated by several substances including

transforming growth factor D (TGF,), arg-vasopressin (AVP) and angiotensin II (Ang II) (7 for

reviews,8,9), while prolactiu and TGF, increase ET transcripts and peptide secretion iu breast epithelial cells (5). ET elaborated by endothelial cells and epithelial cells has been proposed to function in a paracrine manner on adjacent vascular smooth muscle and stromal cells, respectively (1,5,7,10). Abbreviations; ET, endothelin, hVSMC, human vascular smooth muscle cc.&, TGF,, transforming growth factor 8; PDGF-w platelet derived growth factor-AA homodimer; EGF, epidermal growth factor; Ang II, augiotensin II; AVP, arginine-vasopressin. 0006-291X/90 1303

$1.50

Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

Vol.

168, No. 3, 1990

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

The paracrine regulation of smooth muscle cell function and growth by factors either derived from circulating blood platelets or synthesized by endothelial cells is well recognized (llJ2).

However, it is becoming

increasingly acknowledged that smooth muscle cells themselves secrete a variety of substances which can modulate their growth/function.

Such compounds include both normal extracellular matrix constituents

such as heparin and thrombospondin

(l3-15), and peptide growth factors such as PDGF-AA

(16,17). Using

cultured rat vascular smooth muscle cells (VSMC) we have recently demonstrated that Ang II can increase levels of transcripts for TGF, and platelet derived growth factor (PDGF)-A

chain, and that PDGF-AA

homodimer and TGF, can also increase the levels of both their own and each others transcripts (18). Therefore the abiity of VSMC to synthesize and secrete TGFa and PDGF-AA mechanism for the regulation of their growth. Furthermore,

supports an autocrine

studies in our laboratory (manuscript in

preparation) have shown that ET, like Ang II, can also increase PDGF-A chain and TGF, transcript levels in cultured VSMC from both rat and human vessels, and such effects may be relevant to the mitogenic properties of ET. In the light of such observations we considered the possibility that growth factors and vasoactive agonists might induce ETmRNA expression in, and release of ET peptide from cultured vascular smooth muscle cells. MATERIALS

AND METHODS

Materials: All tissue culture material and chemicals were from Gibco AG, Switzerland with the exception of fetal calf serum (Fakola AG, Switzerland). Synthetic ET-l (porcine, human) was from Nova Biochem, Switzerland. Angiotensin II, [Sar’, Ala81 angiotensin II, A$-Vasopressin and cycloheximide were from Calbiochem AG, Switzerland. TGF, and epidermal growth factor (EGF) were from Collaborative Research Inc., USA and PDGF-AA homodimer was provided by Hoffmann-La Roche Ltd, Switzerland. [l-(8mercapb-8,8-cyclopentamethylene proprionic acid), 2-O-methyltyrosine] (POMT-AVP) was from Peninsula Laboratories USA. myo-[2-3Hj inositol(23 Ci/mmol) was purchased from ACR Inc., USA. Unless otherwise stated all other chemicals were obtained from Sigma Co., USA, Fluka, Switxerland or Bio-Rad., Switzerland. Cell culture; Vascular smooth muscle cells (hVSMC) from human omental vessels were isolated, characterized and propagated as described previously (5,19). Porcine and bovine aortic endothelial cells were isolated and cultured as detailed before (20). Immunocytochemical characterization of both vascular and bovine endothelial cells was performed in primary culture using monoclonal antibodies against smooth muscle a-a&n ( a gift from Dr. G. Gabbiani (21) or clone lA4, Siia Chemical Co., USA) and Von Wiiebrand factor (Factor VIII, Dakopatts M616), and according to the methods described by Petersen and Van Deurs (22). hVSMC showed >90% positive staining for a-actin but undetectable staining for factor VIII, while bovine endothelial cells (used as immunocytochemical control) were negative for u-a&in and >95% positive for factor VIII (data not shown; see 5). Experiments in this study utilized hVSMC between passage 10-16. Confluent cultures of hVSMC were rendered quiescent by serum deprivation (19) for 48 hrs. (0.1% bovine serum albumin substituted for 10% serum; with one medium change after 24 hrs) prior to experimentation. Northern Blot analwis: Following stimulation (described in Frgure legends) and removal of medium, cell layers were rinsed with phosphate buffered saline and lysed with guanidine isothiocyanate buffer (23). Total RNA was collected by centrifugation of the lysate on a 5.7 M CsCl gradient as described previously (18,23). For Northern analysis 20 Irg of total RNA was electrophoresed through a 1.2% agarose gel containing 2.2 M formaldehyde at a constant voltage of 50 V for 6-8 hrs. in MOPS buffer. The gel was vacublotted to Hybond nylon membranes (Amersham Corp.) with 20 x SSC (sodium chloride, sodium citrate) as transfer buffer. Blotted RNA was fixed to membranes by UV irradiation for 3 min. at 302 nm. Blots were hybridized to a random primed 1.3 kb cDNA probe for endothelin according to the method of Church and Gilbert (24), washed at high stringency (2 x 20 min., 0.1 x SSC / 0.1% SDS at 65” C) and exposed to Kodak X-Omat fbs overnight at -70” C using one intensifying screen. To assess variabilities of RNA in samples, blots were rehybridized (after stripping previous signals by incubation for 10 min. in boiling hot 10 mM Tris-HCl (pH 7.5)/10 mM EDTA) to a 1.5 kb cDNA probe specific for MHC class 1 antigens (clone pMF 48) (25) and exposed to x-ray film for 6-10 hrs. 1304

Vol. 166, No. 3, 1990

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Measurement of endotheiin uroductioo; Endotheiin was measured in 200 11aliqu0t.s of me&m overlay (see Results for experimental protocol) using an ET-1 radioimmunoassay kit according to the procedure detailed by the manufacturer (Peninsuia Laboratories Inc., USA). In some experiments endotheiin was aiso measured using an ET-l radioreceptor assay kit as detailed by the manufacturer (ANAWA Laboratories AG, Switzerland). Stimulation of DhosDhoinositide catabolism; hVSMC were preiabeiied for 48 hrs. with myo[2-3H]-inositoi (25 sCi/mi) under serum - and inositoi-free conditions and then exposed (im the presence of 15 mM LiCi) to ET-l or extracts of medium overlay and other appropriate additions (see Results and legends to Figures) for 15 min. Thereafter incubations were terminated and phosphoinositide metaboiites extracted and resolved by column chromatography as detailed previously. (19). Extracts of medium overlay were prepared as follows: 20 ml of medium from appropriately stimuiated hVSMC cuitures (see legend to Figure 3) were applied to Seppak C,, cartridges (Waters Associates, USA) that had been sequentiaiIy prewashed with methanoi, acetonitriie/5 mM triiiuoracetic acid, Hz015 mM trifhroracetic acid; cartridges were washed with H,O/ 5mM tritiuoracetic acid and ET-l was eiuted with methanol. Eiuates were dried under vacuum and reconstituted in 400 ri minimal essential medium containing 1 mg bovine serum albumin/ml. Recovery was controlled by inclusion of tracer [‘zsI]-ET-l (ANAWA Laboratories AG, Switzerland) in samples of medium overlay and this was generally 8590%. ET-l levels in reconstituted extracts were determined by radioimmunoassay and for phosphoinositide catabolic experiments 25 ri of reconstituted extracts were added per 250 rrl of hVSMC incubation medium overlay. RESULTS AND DISCUSSION Anuiotensin II increases exnression of ETmRNA in human smooth muscle ceils Immunocytochemicai characterization of our hVSMC isolates used in this and other studies (5,18) verified both their smooth muscle phenotype and the absence of endotheiiai ceii contamination (see Materials and Methods section). In cuitured hVSMC, rendered quiescent by 48 hrs. serum deprivation, expression of ETmRNA

was increased by Ang II in a time-dependent

manner (Fig. lA). Peak levels of ETmRNA

in

hVSMC were obtained within 3-5 hrs. exposure to Ang II and thereafter rapidly declined to basal levels (by 7 hrs.). The induction of ETmRNA

by Ang II was dose-dependent over the range lo-” - l@’ M (data not

shown). The transience of inducible ETmRNA

expression

in hVSMC exposed to Ang II differed from that

observed for porcine aortic endotheiiai ceiis in which levels of ET transcripts remained elevated even after exposure to thrombm for 10 hrs. (Fig. 1B) or to TGFs for 12 hrs (8; see aiso hVSMC response to TGF, in Fig. 4A). A

I 005

1

3 5

B

7hrs

0

2

L

6

8

1Ohrs

ET

ET

MHC

Mm

Induction of ETmRNA expression in human smooth muade cells and porcine eadothelial cells. ?iiF Co uenr,qukscenr hVSMC were exposed to la’ M Ang II (Panel A) and porcine endothelial cells to 2 U/ml thrombin (Panel B) for the times indicated. Northern blot analysis with hybridization to a random primed ET qeci6c cDNA probe was performed on 20 rg of total RNA, with rehybridization to MHC cDNA to quantify amounts of total RNA loaded onto geis. All procedures are descrii in Materials and Methods. Typical autoradiographs (from three separate experiments) are presentcxi. 1305

Vol.

BIOCHEMICAL

168, No. 3, 1990

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

20-

lo-

1 I

3

6

9

OL 0

3

Time (hr) 2.

6

9

C

Ang II

AVP

Time (hr)

Aagtoteasta

II and aq-vasopressta

TFCo uent, quiescent hVSMC were expcxed to 15 min. pretreatment) and 1U’ h4 Ang II (0).

promote &ease of ET pepttde ti-om WSMC. 10-’ M Ang II (@?, AVP ( A ) or to cycIolmximide (lo4 M, ET peptide levels III 200 rl aliquots of medium overlay were

determined using raclioinmmnoassay(Panel A) and radioreceptorassay (Panel B) procedures. In Panel C hVSMC were incubated for 2 hrs, without additions (C, open bar), and in the presence of 10’ M Ang II or la’ M AVP either without (solid bars) or with (hatched bars, +) inclusion of their respective spxific receptor antagonists [Sar’, Alas#Ang II (10” M) and POMT-AVF’ (104 M). Data are given as mean 2 SD (where n = at least 3). The cell number/vol. medium proportion in all experiments was -0.6 - 0.8 x la”/2 ml. Experimental detaih are given ia Materials and Methods.

hVSMC

smte

ET-tide

in

response

to aneiotensio

II and anwasowwsin

Induction of ETmRNA by Ang II in hVSMC was accompanied by their secretion of ET peptide into medium overlay as assessed using radioimmunoassay (Fig. 2A) and radioreceptorassay (Fig. 2B) procedures, both of which yielded qualitatively and quantitatively comparable results. The accumulation of ET peptide into conditioned media was most rapid within the fust 3-5 hrs. of Ang II stimulation, and thereafter gradually plateaued (Fig. 2) which is in agreement with the kinetics for induction of ETmRNA

expression by Ang II

(Fii. 1). Cycloheximide completely inhibited Ang II promoted production of ET peptide by hVSMC (Fig. 2), which indicates the requirement for de novo protein synthesis. Regulation of ET secretion at the level of peptide synthesis has previously been proposed for endothelial cells (1,7,10) and cycloheximide was also demonstrated to inhibit thrombm-stimulated

ET release in isolated perfused blood vessels (26). Secretion

of ET peptide from hVSMC was also stimulated by arg-vasopressin (AVP) in a time-dependent

manner

which was similar to that for Ang II (Fii LX). The stimulatory effects of Ang II and AVP on ET peptide production by hVSMC could be abolished by their respective specific receptor antagonists [Sar’, AlaslAng II and POMT-AVP

(Fig. 2C). Both vasoconstrictive hormone have previously been demonstrated to elicit

release of immunoactive ET from bovine endothelial cells (9). When considered with respect to cell number, the amounts of ET secreted by Ang II (10.’ M) and AVP (lo” M) stimulated hVSMC (-120-160

pg/106

cells/4 brs.) are -40% lower than those reported (9) for similarly stimulated endothelial cells (-uw)-m pg/lO’ cells/4 hrs.). Nevertheless, the concentrations of ET in the media of Ang II and AVP stimulated

1306

Vol.

BIOCHEMICAL

166, No. 3, 1990

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

hVSMC do attain levels (-3 x lo’” M) that are within the biologically effective range for this peptide (1,7W. Biolasdcal activitv of WSMC secreted wutide For a variety of cell types including smooth muscle cells (27) the activity of phospholipase C increases rapidly following activation of ET-specific receptors (7,lO). Therefore to determine whether ET peptide produced by hVSMC possesses biological activity concentrated Sep-pak C,, extracts (see Materials and Methods) of medium from cells exposed to Ang II (lo” M, 5 min. - 7 hrs.) were examined with respect to their ability to stimulate phosphoinositide

catabolism in myo[2-3H]-inositol prelabelled hVSMC. Since the conditioned

media from which extracts were prepared contained Ang II, [Sar’, Alay-Ang II (lo” M) was included during stimulation of prelabelled cells to exclude that a phosphoinositide catabolic response might be due to Ang II present in extracts. Figure 3B illustrates that this concentration of [Sar’, Ala?-Ang II completely abolished the phosphoinositide

response to lo4 M Ang II. Media extracts from Ang II exposed hVSMC were indeed

able to induce phospholipase C-mediated catabolism of phosphoinositides as evidenced by the accumulation of inositol phosphates in prelabelled hVSMC (Fig. 3A). This response was determined to be concentrationdependent in as much as levels of immunoreactive ET in the extracts differed and increased with the duration of exposure to hormone (Fig. 3A). The magnitude of increases in inositol phosphates elicited by the different extracts were also comparable to those elicited by standard ET-l

added at concentrations similar to

immunoactive ET in extracts (Fig. 3A). These tindings indicate that the phosphoinositide catabolic response to extracts is indeed due to ET peptide produced by hVSMC. Furthermore any contributing stimulatory effect

[.%I’. Aldl 0

100

200

300

ml

500

- Angll(lO.s

AngII(

1

M)

I o-6 M)

ET Wml)

Pipwe 3. ET peptide released by bVSMC is biologically acttve. hVSMC were exposed to NJ’ M Ang II for the tics (hours) indicated in parentheses. For the zero (a) time period hVSMC were exposed to Ang II for less thaa 5 min. Thereafter medium overlay (20 ml) was

withdrawn and Seppak C,, extracts prepared. Immunoreactive ET ia these extracts was quantitatcd. myo[fH]-inositol prelabekd hVSMC were exposed (in the presence of 15 mM LiCI and l@ M [sar’, Ala’]-Ang ll) for 15 min. to extracts (0) or to standard ET-1 ( A ) ( final concentrations of ET for both l and A indicated on x-axis of Panel A). In Panel B @II-inositol prelabekd lob aug II either without ((-), open bar) or with inclusion of l@

hVSMC

were exposed for 2 min. to

M [Sar’, AlaTAng II (f+), closed bar). Data (meen 2 SD, n = 3) expressthe tritium content of in&o1 phosphates (mono - plus bii - plus tris phosphate) relative to that (100%) in control samples (hVSMC exposed only to vehicle). 1307

Vol.

168,

No.

BIOCHEMICAL

3, 1990

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

of Ang II can be excluded since extracts from hVSMC exposed to Ang II for 5 min. (zero-time extract in Fig. 3A) did not change inositol phosphate levels; Ang II might be assumed to be at a constant concentration in extracts. Preliminary studies have indicated that extracts of media from Ang II exposed hVSMC are also capable of causing a contraction of endothelium-denuded

perfused rat mesenteric resistance arteries in a

manner typical for ET (i.e. difficult to wash out). Further studies are currently in progress to verify this physiological response. Induction of ETmRNA and release of DEwtide bv erowth factors The growth factors TGF,, PDGF-AA ETmFWA

as well as epidermal growth factor (EGF) were all found to induce

expression in quiescent hVSMC

(Fig. 4A), and this was accompanied

by secretion of

immunoreactive ET (Fii. 4B). The kinetics for growth factor stimulated ET transcript formation and peptide production, as well as levels of the latter were similar to those observed for Ang II and AVP (compare Figs. 1, 2 and 4). Growth factor promoted production of ET peptide was also completely inhibited (data not shown) by cycloheximide. When adult VSMC’s are isolated by enzymatic disaggregation, put into culture and passaged, they rapidly undergo a process of phenotypic modulation from the contractile to the proliferative / secretory form (28). The latter phenotype is capable of synthesizing and secreting not only extracellular matrix macromolecules (13-15) but also growth factors such as PDGF-AA

and TGF, (16-18, 29). To our knowledge, there is

presently no data available to indicate whether ETmRNA expression, either constitutive or inducible, occurs

A 0 0.5 1

3 5 7 hrs

PIXiF-AA EGF

I

Time (hr)

Growth factor induction of E’IWWA expression and ET peptide secretion in WSMC. Confluent, quiescent hVSMC were exposed to PDGF-AA (5 ag/ml), TGF, (4 OS/ml) or EGF (5 a&d) Figure 4.

the times indicated. Northern

in Materials and Methods medium overlay as described ia Materials aad Methods and legend to FI

determined ia Autoradiographs

for

blot analysis of total RNA to determine ETmRNA was performed as dmcribed and legend to Fve 1 (Panel A). Secretion of immunoreactive ET was

in Panel A represent

typical ETmRNA

expression

profiles

for hVSMC

2 (Panel 8).

m

to the

growth factors. Data in Panel B are given as mesa f SD, n = 3; hVSMC stimulated by TGF, ( A ) or PDGF-AA (0). 1308

Vol. 168, No. 3, 1990

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

within the smooth muscle cell layer in vivo. However, vascular smooth muscle cells in viva are also capable of undergoing reversible phenotypic conversion (l2,30), a property which is importantin the.pat,hogen& of hypertension and atherosclerosis (11,12,28,30). Compounds which have been implicatedin b&h the development and maintenance of the pathological proliferative / secretory smooth muscle cell phenotype include Ang II, PDGF-AA of ETmRNA

and TGF, (11,12,17-1828-31). Therefore our present observation of the induction

in cultured quiescent hVSMC by growth factors and vasoactive hormones suggest that

ETmRNA

expression by VSMC is likely to be of physiological/pathophysiological

relevance. Detection of

ETmRNA

expression by smooth muscle cells in vivo may depend upon the existing and/or past integrity

of the adjacent endothelial cell layer. Detection of the in situ release from isolated endothelium-denuded blood vessels may depend upon whether the vessels are in a state of health or disease (i.e. hypertensive or atherosclerotic) and also upon application of an appropriate stimuhis (e.g. Ang II, PDGF-AA,

TGF,).

In this study, we have demonstrated hVSMC ETmRNA expression and peptide production to be elicited by substances hnown to stimulate the proliferative behaviour of VSMC and/or vascular reactivity, both of which are also influenced by ET (reviewed in 7,lO). Our findings suggest a role for ET in the vasculature which extends beyond the known paracrine function of ET synthesized and released by the endothelium. We propose that VSMC - derived ET may act in an autocrine fashion on regulating both vascular tone and structure. Establishment of the physiological and/or pathophysiological reality of such an autocrine regulatory mechanism of action for ET would seem fundamental to our understanding of the role of this peptide in control of the cardiovascular system.

Achmn&&menta

Evi Miinger is thanked for the preparation of the manuscript as is Bernadette Libsig for

artistic help. Financial support was provided by the Swiss National Foundation Grant No. 3.827.087.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Yanagisawa M., Kurihara H., Tomobe Y, Kobayashi M., Mitsui Y, Yazaki Y., Goto K and Masaki T. (1988) Nature 332,411-415 Giid A., Gibson S.J., Ibrahim N.b.N., Legon S., Bloom S.R., Yanagisawa M., Masaki T., Vamdell I.M. and Polak J.M. (1989) Proc Natl Acad Sci USA 86,7634-7638 Yoshizawa T., Shinmi~O., &aid A., Yanagisawa M., Gibson S.J., Kimura S., Uchiyama Y, Polak J.M., Masaki T. and Kanazawa I. (1990) Science 247.462464 Kosaka T., Suzuki N., Matstkotd H.. Itoh Y..‘Yasuhara T., Onda H and Fuiino M. (1989) . _ FEBS L&s. 249;42-46 . Baley PA., Resink T.J., Eppenberger U. and Hahn A.WA. (1990) J Clin Invest (m press) Black P.N.. Ghalei MA.. Takahashi K.. Bretherton-Watt D., Krausz Dollerv C.T. and Bloom SJ. (1989) FEBS Letts. 2.55, ‘l29-132 . Yanagisawa M. and Masaki T. (1989) Trends Pharmacol Sci lo,374378 Kurihara H., Yoshizmni M., Sugiyama T., Takaku F., Yanagisawa M., Masaki T., Hamaoki M., Kato H. and Yazaki Y. (1989) Biochem Biophys Res Commun 159, 14351440 Emori T., Hirata Y., Ohta K., Shichiri M. and Marumo F. (1989) Biochem Biophys Res Commun 160, 93-100 Yanagisawa M. and Masaki T. (1989) Biochem Pharmacol38,1877-1883 Ross R. (1986) N Engl J Med 314,488-500 Schwartz S.M., Campbell G.R. and Campbell J.H. (1986) Ciic Res 58,427-444 1309

Vol. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

168, No. 3, 1990

BIOCHEMICAL

AND BIOPHYSICAL RESEARCHCOMMUNICATIONS

Herman I.M. and CastelIot JJJr. (1987) Arteriosclerosis 7,463469 Majack MA., Cook SC. and Bornstein P. (1985) J Cell Biol 101, 1059-1070 Scott-Burden T. and Btihler F.R. (1988) Trends Pharmacol Sci 9,94-98 Hosang M and Rouge M (1989) J Cardiovasc Pharmacoll4 (Suppl.6) 22-26 Majesky M.W., Bend&t E.P and Schwartz.8.M. (1988) Proc Natl Acad Sci USA 85,1524-1X28 Scott-Burden T., Resink T.J., Hahn A.WA and Biihler F.R. (1990) J Cardiovasc Pharmacol (m press) Scott-Burden T., Resink T.J., Hahn A.W& Baur U., Box R.J. and Biihler F.R. (1989) J Biol Chem 264,l2582-12589 Boutanger C., Hendrickson H., Lorenz R.R. and Vanhoutte P.P (1989) Ciic Res 64, 1070-1078 Kocher O., Skulli O., Bloom W.S. and Gabbiani G. (1984) Lab Invest 50,646-652 Petersen O.W. and Van Deurs B. (1988) Differentiation 39, 197-215 Ulhich A., Shine J., Chingwin J., Pichet R., Tischer E., Rutter W.J. and Goodman H.M. (1977) Science 196, 1313-1315 Church G. and Gilbert W. (1984) Proc Nat1 Acad Sci USA 81,1991-1995 Pohla H., Kuon W., Tabaaewski P., Dorner C. and Weiss H.E. (1989) Immunogenetics 29,297X17 Boulanger C. and Liischer T.F. (1990) J Clin Invest 85,587-590 Resink T.J., Scott-Burden T. and Biihler F.R. (1988) Biochem Biophys Res Commun l57,l360-1368 Charnley-Campbell J.H., Campbell G.R. and Ross R. (1979) Physiol Rev 59, l-55 Sejersen T., Betsholtz C., Sjolund M., Heldin C.-H., Westermark B. and Thyberg J. (1986) Proc Nat1 Acad Sci USA 83,6844-6846 Campbell J.H. and Campbell G.R. (1986) Arm Rev Physiol48,295-306 Powell J.S., Clozel J.-P., Miiller R.K.M., Kuhn M., Hefti F., Hosang M. and Baumgartner H.R. (1989) sci 245,186-188

1310