Vascular endothelial growth factor (VEGF) expression during arterial repair in the pig

Eur J Vasc Endovasc Surg 15, 225-230 (1998)

Vascular Endothelial Growth Factor (VEGF) Expression During Arterial Repair in the Pig S. J. WysockP*, M.-H. Zheng 2, A. Smith 2 and P. E. Norman 1

1University Department of Surgery, Fremantle Hospital, Fremantle 6160 and 2University Department of Orthopaedic Surgery, Queen Elizabeth II Medical Centre, Nedlands 6009, Western Australia Objectives: Vascular endothelial growth factor (VEGF) is reported to be a potent and specific mitogen for endothelial cells (EC) and an inducer of angiogenesis in vivo. Originally called vascular permeability factor (VPF), VEGF also increases permeability of microvessels to circulating macromolecules. The aim of this study was to examine whether the VEGF gene was expressed in porcine arteriesfollowing denudation of EC. Design: Experimental animal model with mechanical injury to large arteries. Methods: The right iliac artery of juvenile pigs was de-endothelialised using an inflated balloon catheter. At a number of time-points after injury, these arteries were harvested together with uninjured contralateral arteries. Sections of arteries were used for RNA analysis by Northern blots and for protein localisation studies by immunohistochemistry. Results: Two VEGF transcripts (2.0 kb, 4.5 kb) were markedly elevated in pig arteries soon after injury. Newly synthesised VEGF protein was located in smooth muscle cells (SMC) throughout the media of injured arteries. Conclusions: The elevated expression of VEGF by SMC in denuded porcine arteries is evidence that this cytokine plays a role in the injury response oflarge arteries. Since several biological activities have been identifiedfor VEGF, the function of this cytokine in the arterial repair process remains to be determined. Keywords: Pig arteries; Balloon injury; Gene expression; Vascular endothelial growth factor; Smooth muscle cells.

Introduction Angioplasty and other forms of surgical intervention for occlusive arterial disease are followed b y a repair process which can lead to restenosis in m a n y patients. The arterial injury response is difficult to s t u d y in h u m a n s and nearly all of our k n o w l e d g e of intimal hyperplasia rests on studies carried out on laboratory animals. 1 The most c o m m o n experimental injury is induced using an inflated balloon embolectomy catheter to r e m o v e the endothelial layer from a major artery. The resultant lesion contains primarily smooth muscle cells (SMC) resulting from synchronous proliferation of SMC in the media beginning at 24 h, 1'2 and followed b y migration of SMC to the luminal surface to form a neointima several days later. 3 Over the next few weeks, the accumulation of extracellular matrix proteins also contributes to the increased volu m e of the arterial intima. 2 In animals, the size of the lesion is generally self-limiting and it appears that * Please address all correspondence to: S. Wysocki, University Department of Surgery, Fremantle Hospital, PO Box 480, Fremantle, Western Australia 6160.

regeneration of an impervious, non-thrombogenic endothelium or p s e u d o e n d o t h e l i u m 4 results in cessation of processes leading to intimal hyperplasia. 5'6 The unique properties of vascular endothelial growth factor (VEGF) have focussed attention on this cytokine as a potential agent whose administration could accelerate re-endothelialisation and therefore enhance remodelling of injured arteries and diminish the extent of intimal hyperplasia. 7 VEGF is a secreted glycoprotein believed to be a multifunctional regulator of endothelial cell (EC) growth 8 whose biological activities are mediated via receptors which are expressed predominantly, if not exclusively, on vascular endothelial cells. 9 In addition to its reported abilities to stimulate EC division in vitro and angiogenesis in vivo, VEGF was originally called vascular permeability factor because it increased microvascular permeability to circulating macromolecules. 9 Attempts to use VEGF as a therapeutic agent to hasten re-endothelialisation of injured arteries have led to contested findings. Several groups have reported that administration of VEGF to rats and rabbits with balloon-injured arteries accelerated EC growth in these vessels. 7'1° By contrast, other researchers were unable to demonstrate any

1078-5884/98/030225+06 $12.00/0 © 1998 W.B. Saunders Company Ltd.

S.J. Wysocki et al.

226

stimulation of re-endothelialisation in denuded rat arteries using a number of doses of VEGF and various treatment protocols. ~I In this most recent study, the authors could not detect VEGF mRNA in injured arteries, ~ casting some doubt on any function for this cytokine in the repair of large arteries; however, there may have been problems with sufficient cross hybridisation of mouse VEGF antisense probes used in this investigation. In the present study we have examined arterial VEGF expression in the pig, an animal model where restenosis after angioplasty appears to be more closely related to findings obtained for humans, n

Methods

Animals This study used male juvenile (9-11 weeks) domestic cross-bred pigs (Landrace and Large White) weighing about 20 kg. All pigs were fed normal chow and were treated in accordance with the National Health and Medical Research Council guidelines for animal research. A total of 10 pigs were included in this investigation.

Balloon catheter injury to iliac arteries Under general anaesthesia, the right iliac was denuded of endothelial cells using an inflated 5-French Fogarty embolectomy catheter (Baxter). Following an arteriotomy in the right profunda femoris artery, the catheter was passed up to the aortic bifurcation and inflated. 13 The inflated balloon catheter was pulled down the iliac artery over a 20 s period and this procedure was repeated two more times. A consistent removal of the endothelium by this injury was demonstrated in an earlier study. 2 The left iliac artery was uninjured.

adventitia (later verified by histology), the arteries were dissected into two smaller pieces using transverse cuts. Part of the specimen was used for mRNA analysis and was frozen in liquid nitrogen before storage at - 8 5 °C. The remainder was used for detection of VEGF protein using immunohistochemistry and was gradually frozen in OCT medium by gradual submersion in liquid nitrogen. These preparations were also stored at - 8 5 °C prior to sectioning.

Northern blot analysis Frozen arteries were ground to a fine consistency under liquid nitrogen using a mortar and pestle and the resultant powder was homogenised in Ultraspec reagent (Biotecx Laboratories, Houston, TX) by four high speed bursts (15 s) using a Janke and Kunkle tissue disintegrator (Ultra-Turrax T25). Total RNA was then prepared using the Biotecx isolation protocol. Aliquots of total RNA were run on formaldehyde gels prior to overnight capillary transfer to Hybond N + nylon membrane (Amersham, U.K.) and alkali fixation. 2 Membranes were hybridised overnight (42 °C) to 32P-labelled DNA probes in a hybridisation solution containing 50% formamide (Amersham Hybond N + protocol). The VEGF probe was labelled by random primer extension of the insert and the 36B4 plasmid was labelled by nick translation, both using kit protocols (Amersham, U.K.).

DNA probes The DNA probes used were as follows: VEGF, a 583 bp BamH1/EcoR1 fragment of pVEGF18 and the plasmid 36B4 originally called pN1.14 36B4 is the cDNA for human acidic ribosomal phosphoprotein PO 15and represents an ubiquitously expressed mRNA. 16

Immunohistochemistry Harvesting of arteries Arteries were harvested at 2 h, 6 h, 8 h, 16 h and 24 h after balloon injury. Pigs were given a general anaesthetic and underwent a laparotomy. Isotonic saline was perfused into the infra-renal aorta and through distal arteries to drain by the inferior vena cava. The right and left common iliac arteries were then harvested. Following the surgical removal of Eur J Vasc Endovasc Surg Vol 15, March 1998

Frozen arterial sections (10 ~tm thick) were air-dried and fixed in cold acetone for 2 min. Slides were then rinsed five times in 0.2 M Tris-buffered saline (TBS) for 2 rain before blocking of endogenous peroxidase by a 20 rain incubation at room temperature with horse serum (20%). Incubation with a 1 in 4 dilution (25 ~tg/ ml) of a rabbit polyclonal antibody raised against a human VEGF fusion protein construct (sc-507, Santa Cruz Biotechnology) was carried out at 4 °C overnight.

VEGF Expression in Arterial Repair

This p r i m a r y antibody to VEGF was omitted for sections serving as negative controls. Slides were w a s h e d twice in TBS for 5 rain and incubated for 30 rain at 37 °C with a 1 in 200 dilution of fluorescein isothiocyanateconjugated swine anti-rabbit i m m u n o g l o b u l i n (Dako). Following five rinses in TBS for 2 min, slides were m o u n t e d in glycerol. In separate cell identification studies, frozen arterial sections were also incubated with m o u s e anti-human alpha-smooth muscle actin (Sigma) as the p r i m a r y antibody (data not shown).

Assessment of fluorescence staining by confocaI laser scanning microscopy (CLSM) The detection of fluorescence staining in arterial sections was p e r f o r m e d using a Bio-Rad CLSM (MRC1000 UV) e q u i p p e d with a 250 m W argon ion laser coupled to an epifluorescence Nikon Diaphot 300 inverted microscope. A 20X glycerine immersion lens (Nikon UV-F20, N A = 0 . 8 ) was employed. The excitation wavelength was 488 n m attenuated to 3%, and an emission b a n d pass filter (522/35 bp) was utilised. Serial optical sections were taken at 2 gm intervals and the images presented are maximal projections.

227

Un Un 2h 2h 2h 6h 6h 8h 8h 16h24h24h 4.5kb 2-Okb

VEGF

Fig. 1. Expression of the VEGF gene in pig iliac arteries after a balloon catheter injury. Total RNA was extracted from right iliac arteries at specified times after injury and also from the uninjured contralateral arteries. Aliquots were subjected to Northern blot analysis using 32p-labelled cDNA probes. 36B4 mRNA is ubiquitously expressed and has been used as an internal control to reflect the evenness of loading. Note that the VEGFtranscripts were just detectable in uninjured arteries and were rapidly elevated in injured arteries at 2 h before reaching maximal levels at 6-8 h after balloon injury. The mRNA in lane 9 (8 h) was partially degraded so that the 4.5 kb transcript was not obvious; however the 2.0 kb transcript was clearly elevated.

detectable in uninjured arteries and were rapidly elevated at 2 h in injured arteries before reaching maximal levels at 6-8 h after balloon injury (Fig. 1). This elevation was not due to a generalised increase in mRNAs as the level of the reference 36B4 m R N A remained fairly constant t h r o u g h o u t early arterial repair. By 16-24 h after balloon injury, the levels of VEGF mRNAs were m u c h reduced, although transcript levels were still slightly higher than observed in uninjured arteries (Fig. 1).

Results

Expression of the VEGF gene in pig arteriesfollowing balloon injury

Localisation of VEGF protein in injured pig arteries

To determine whether VEGF is expressed during repair of arteries, as it is in other regenerating tissues, we carried out N o r t h e r n blot analysis of total RNA extracted from injured pig arteries using a 32p-labelled h u m a n cDNA probe. We observed that two VEGF mRNAs were clearly identifiable in pig iliac arteries, the larger VEGF transcript being estimated to be about 4.5 kb and the second, m u c h smaller transcript at about 2.0 kb (Fig. 1). Similar sized VEGF transcripts have been previously detected in h u m a n tissues and tissues and cultured cells from the guinea pig, including S M C . ~7 Additional VEGF mRNAs of 5.5 kb and 3.7 kb, which had been demonstrated in h u m a n foetal vascular SMC, 18 were not obvious in pig arteries. The transient expression of VEGF during dev e l o p m e n t and repair has been interpreted as evidence of a role for this cytokine in the regulation of endothelial cell growth. Accordingly, we examined the time course of VEGF gene expression during early arterial repair in the pig. The VEGF transcripts were just

Immunohistochemistry was utilised to determine the cellular origin of VEGF expression in frozen sections from pig arteries e m p l o y i n g a rabbit polyclonal antib o d y to h u m a n VEGF which recognises the 165, 189 and 121 amino acid splice variants of this cytokine. Strong immunoreactivity for VEGF was demonstrated t h r o u g h o u t the inner two thirds of the media of pig iliac artery at 6 h after injury (Fig. 2b). Staining was uniform t h r o u g h o u t this part of the media and the distribution was identical to that obtained with alphasmooth muscle actin (not shown) indicating that SMC were the p r e d o m i n a n t source of the immunoreactivity. Macrophages are located in pig arteries at 6 h after balloon injury 12 but any potential contribution from these i m m u n e cells w o u l d have been masked b y the overwhelmingly more n u m e r o u s SMC. VEGF immunoreactivity in the outer one third of the media of injured arteries was scattered and not as intense (Fig. 2b), and there was a minimal VEGF reactivity in the media of uninjured artery (Fig. 2a). Eur J Vasc Endovasc Surg Vol 15, March 1998

S.J. Wysocki et al.

228

Fig. 2. Detection of VEGF protein-positive cells in pig iliac arteries 6h after balloon injury. Frozen sections of artery were processed by standard immunohistochemical methodology using a polyclonal antibody to human VEGF and a FITC-labelled secondary antibody. Detection of fluorescein staining was performed by confocal laser scanning microscopy. Sections were counter-stained with haematoxylin. (A) VEGF protein immunostaining of a transverse section of uninjured artery. (B) VEGF protein immunostaining of a transverse section of injured artery collected at 6 h after balloon injury. (C) Immunostaining of an adjacent serial section of injured artery in the absence of primary antibody. For all sections the arterial lumen is in the top right hand corner. The size bar is 100 #m. Note the presence of VEGF protein-staining cells throughout the inner two thirds of the media in iliac artery at 6 h after injury (B). Cells within this part of the artery stained strongly positive for alpha~smooth muscle actin (~SMA) (not shown).

Discussion Current u n d e r s t a n d i n g of arterial intimal h y p e r p l a s i a is b a s e d largely on experimental investigations using small laboratory animals. F i s h m a n a n d colleagues dev e l o p e d a m o d e l in w h i c h clearly delineated endothelial d e n u d a t i o n w a s achieved b y brief air-drying of the l u m e n of rat carotid artery. ~ These researchers established that endothelial regeneration resulted f r o m r e g r o w t h of EC f r o m each end of the d e n u d e d s e g m e n t a n d that the duration of endothelial d e n u d a t i o n correlated with degree of intimal thickening, s This latter observation s u p p o r t s the hypothesis that a functional endothelial layer is required to d a m p e n d o w n arterial intimal h y p e r p l a s i a following surgery a n d has p r o m p ted investigations into w h e t h e r application of the EC m i t o g e n VEGF could stimulate re-endothelialisation in d e n u d e d arteries and thereby eventually attenuate intimal thickening. Eur J Vasc Endovasc Surg Vol 15, March 1998

Two recent studies r e p o r t e d data suggesting that this strategy w o u l d be successful. Callow a n d colleagues d e m o n s t r a t e d accelerated endothelial r e g r o w t h in carotid arteries f r o m rabbits w h i c h h a d received intrav e n o u s infusion of vascular p e r m e a b i l i t y factor (VPF/ VEGF) c o m p a r e d to controls receiving vehicle. 1° Ashara and co-workers r e p o r t e d similar enhanced reendothelialisation of rat carotid arteries w h i c h h a d b e e n incubated with VEGF in situ and, significantly, there w a s also a reduction in intimal thickening in these arteries. 7 By contrast, Lindner a n d Reidy w e r e unable to s h o w m u c h stimulation of endothelial reg r o w t h in carotid arteries f r o m rats receiving an infusion of VEGF. 11These authors reported that d e n u d e d rat carotid arteries did not express VEGF message, a finding w h i c h a p p e a r e d to explain their observation of a lack of function for VEGF in arterial re-endothelialisation. In the present s t u d y w e were able to d e m o n s t r a t e

VEGF Expression in Arterial Repair

that VEGF gene expression was detectable in pig iliac arteries and that it was rapidly elevated following a balloon injury, suggesting that this multifunctional cytokine does play a role in arterial repair in the pig. Transient VEGF expression has been previously observed in several models of wound repair in other tissues including liver 19and skin, 2° and also in a model of pressure-induced ventricular hypertrophy. 21We feel confident that we were in fact detecting VEGF transcripts because we used stringent washing conditions in our hybridisation protocol and the sizes of the transcripts detected were identical to those reported for tissues and cell lines by other researchers) 7 When studying cytokine gene expression we routinely run as a positive control total RNA extracted from cells cultured from a human giant cell tumour of bone. These cells strongly express a number of cytokines including transforming growth factor-[~122 and VEGF (Zheng et al., unpublished observations). We cannot explain w h y Lindner and Reidy were not able to detect VEGF gene expression in denuded rodent vessels and it is unclear whether these authors used a positive control in their investigations. ~1It is possible that there is a species difference in expression of VEGF during arterial repair. In our investigation we have provided evidence that VEGF protein was located in SMC throughout the media of injured pig arteries. Similar findings have been reported for human myometrium where VEGF mRNA and VEGF protein were uniformly distributed in SMC, also the predominant cell type in this tissue. 23 In both studies the spatial distribution of VEGF expression did not clearly coincide with areas of rapidly proliferating endothelial cells, unlike the localisation of the VEGF receptor, fit-l, which was rapidly upregulated at the leading edge of regenerating endothelium in injured rat arteries. 1~ The human gene for VEGF is split among eight exons and various transcripts arise from alternative splicing) s Three VEGF transcripts corresponding to VEGF189, VEGF165 and VEGF12~ have been detected in human uterine tissues 24 but the techniques used in our study of pig arteries do not recognise splice variants. In previous studies we have reported orchestrated changes in cytokine gene expressions in pig arteries following balloon injury including TGF-[31 and MCP12'12 and we can now add VEGF to the list of cytokines whose expression is upregulated during arterial repair. In doing so we have further characterised molecular events in a useful animal model which has similarities to human arterial intimal hyperplasia. VEGF is a multifunctional-cytokine which acts as a specific regulator of endothelial cell growth s and also increases microvascular permeability to circulating macromolecules. 9

229

Two high affinity receptors for VEGF, fit-l, and ilk1, are localised to vascular endothelium 24'25 and are believed to mediate the effects of VEGF on the vasculature. 9 A recent study has reported that VEGF binding toflk-1 ellicited EC proliferation whereas binding to fit-1 did not. 26 Bearing in mind that fit-1 expression was upregulated at the leading edge of regenerating endothelium in injured rat arteries, the biological function of the increased VEGF expression in injured pig arteries observed by us remains unclear. Experimentally, the loss of the endothelial layer in large arteries causes striking intimal hyperplasia 5 and may impair remodelling. Inadequate re-endothelialisation may also contribute to restenosis following angioplasty. 1° It was hoped therefore that acceleration of re-endothelialisation by administration of a specific mitogen such as VEGF would diminish the degree of arterial intimal hyperplasia. This experimental approach would not appear valid if this cytokine does not normally play a role in arterial repair. 1: In the present study we have demonstrated that increased VEGF expression was associated with the arterial response to balloon injury in pigs, a finding which is suggestive of a role for this cytokine in the arterial repair process in this species.

Acknowledgments We wish to thank Drs G. Breieir and P. Chambon for donation of VEGF and 36B4 cDNA probes and Mr D. Heliams for assistance with animal surgery. We are also grateful for the expertise of Mr S. Cody at the Biomedical Confocal Microscopy Research Centre (BCMRC) located within the University of Western Australia Department of Pharmacology at the Queen Elizabeth II Medical centre, Nedlands, Western Australia. The BCMRC facility was funded by a grant from the Western Australia Lotteries Commission. We also thank the Royal Australasian College of Surgeons for funding of our research.

References 1 CLOWES AW, REIDY MA. Prevention of stenosis after vascular injury: pharmacological control of intimal hyperplasia - a review. J Vasc Surg 1991; 13: 885-891. 2 WYsocKI SJ, ZHENG M-H, FAN Y, LAMAWANSAMD, HOUSE AK, NORMAN PE. Expression of transforming growth factor-[~l (TGF[31) and urokinase-type plasminogen activator (u-PA) genes during arterial repair in the pig. Cardiovasc Res 1996; 31" 28-36. 3 JACKSONCL, RAINESEW, Ross R, REIDYMA. Role of endogenous platelet-derived growth factor in arterial smooth muscle cell Eur J Vasc Endovasc Surg Vol 15, March 1998

230

4 5

6

7

8

9

10 11

12

13

14

15

S.J.

Wysocki et

migration after balloon catheter injury. Arterioscler Thromb 1993; 13: 1218-1226. MAJESKY1V[W, REIDY MA, BOWEN-PoPE DE HART CE, WILCOX JN, SCHWARTZSM. PDGF ligand and receptor gene expression during repair of arterial injury. J Cell Biol 1990; 111: 2149-2158. FISHMANJA, RYANGB, KARNOVSKYMJ. Endothelial regeneration in the rat carotid artery and the significance of endothelial denudation in the pathogenesis of myointimal thickening. Lab Invest 1975; 32: 339-351. MERRILEES MJ, SCOTT LJ. Effects of endothelial removal and regeneration on smooth muscle glycosaminoglycan synthesis and growth in rat carotid artery in organ culture. Lab Invest 1985; 52: 409-419. ASHARAT, BAUTERSC, PASTOREC et aI. Local delivery of vascular endothelial growth factor accelerates re-endothelialisation and attenuates intimal hyperplasia in balloon-injured rat carotid artery. Circulation 1995; 91: 2793-2801. BREIRERG, ALBRECHTU, STERRERS, RISAU W. Expression of vascular endothelial growth factor during embryonic anglogenesis and endothelial cell differentiation. Development 1992; 114: 521-532. DVORAK HF, BROWN LF, DETMAR M, DVORAK AM. Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis. Am f Path 1995; 146: 1029-1039. CALLOW AD, CHOI ET, TRACHTENBERGJD et aI. Vascular permeability factor accelerates endothelial regrowth following balloon angioplasty. Growth Factors 1994; 10: 223-228. LINDERV, REIDYMA. Expression of VEGF receptors in arteries after endothelial injury and lack of increased endothelial regrowth in response to VEGE. Arterioscler Thromb Vasc Biol 1996; 16: 1399-1405. WYsocKI SJ, Zh~ENG MH, SMITH A et al. Monocyte chemoattractant protein-1 gene expression in injured pig artery coincides with early appearance of infiltrating monocyte/ macrophages, f Cell Biochem 1996; 62: 303-313. LAMAWANSAMD, WYSOCKI SJ, HOUSE AK, NORMAN PE. Morphometric changes seen in balloon-injured porcine iliac arteries: The influence of sympathectomy on intimal hyperplasia and remodelling. Eur J Vasc Endovasc Surg 1997; 13: 43-47. MASIOKOWSKIP, BREATHNACHR, BLOCHJ, GANON F, KRUST A, CHAMBON P. Cloning of cDNA sequences of hormone-related genes from the MCF-7 human breast cancer cell line. Nucleic Acids Res 1982; 10: 7895-7903. LABORDAJ. 36B4 cDNA used as an estradiol-independent mRNA control is the cDNA for human acidic ribosomal phosphoprotein PO. Nucleic Acid Res 1991; 19: 3998.

Eur J Vasc Endovasc Surg Vol 15, March 1998

al.

16 RUSSELME, ADAMSDH, WXNERLR, YAMASHITAY, HALNONNJ, KARNOVSKY MJ. Early and persistent induction of monocyte chemoattractant protein 1 in cardiac allografts. Proc Natl Acad Sci U.S.A. 1993; 90: 6086-6090. 17 BERSEB, BROWNLF, VAN DE WATERL, DVORAKHE, SENGERDR. Vascular permeability factor (vascular endothelial growth factor) gene is expressed differentially in normal tissues, macrophages, and tumors. MoIec Biol Cell 1992; 3: 211-220. 18 TISHERE, MITCHELLR, HARTMANT et aI. The human gene for vascular endothelial growth factor: Multiple protein forms are encoded through alternative exon splicing. J Biol Chem 1991; 266: 11947-11954. 19 MOCHIDA S, ISHIKAWAK, INAO M, SHIBUYAM, FUJIWARAK. Increased expressions of vascular endothelial growth factor and its receptors,fit-1 and KDR-flk-1, in regenerating rat liver. Biochem Biophys Res Comm 1996; 226: 176-179. 20 BROWN LF, Y~o KT, BERSE B et aI. Expression of vascular permeability factor (vascular endothelial growth factor) by epidermal keratinocytes during wound healing. J Exp Med 1992; 176: 1375-1379. 21 CARROLL SM, NIMMO LE, KNOEPTLER PS, WHITE FC, BLOOR CM. Gene expression in a swine model of right ventricular hypertrophy: intercellular adhesion molecule, vascular endothelial growth factor and plasminogen activators are upregulated during pressure overload. J Mol Ceil Cardiol 1995; 27: 1427-1441. 22 ZHENG MH, FAN Y, WYSOCKISJ et al. Gene expression of transforming growth factor-B1 and its type II receptor in giant cell tumors of bone: possible involvement in osteoclast-like cell migration. Am J Path 1994; 145: 1095-1104. 23 HARRISON-WOOLRYCHML, SHARKEYAM, CHARNOCKqONESDS, SMITHSK. Localisation and quantification of vascular endothelial growth factor messenger ribonucleic acid in human myometrium and leiomyomata. J Clin endocrinol Metab 1995; 80: 1853-1858. 24 MILLAUER B, WIZIGMANN-Voos S, SCHNIYRCH H et al. High affinity VEGF binding and developmental expression suggest Hk-1 as a major regulator of vasculogenesis and angiogenesis. Ceil 1993; 72: 835-846. 25 PETERS KG, DE VRIES C, WILLIAMS LT. Vascular endothelial growth factor receptor expression during embryogenesis and tissue repair suggests a role in endothelial differentiation and blood vessel growth. Proc Natl Acad Sci U.S.A. 1993; 90: 8915-8919. 26 KEYTBA, NGUYENHV, BERLEAULT et al. Identification of vascular endothelial growth factor determinants for binding KDR and FLT-1 receptors. Generation of receptor-selective VEGF variants by site-directed mutagenesis. J BioI Chem 1996; 271: 5638-5646. Accepted 24 September 1997