Heregulin selectively upregulates vascular endothelial growth ... - Nature

ã

Oncogene (2000) 19, 3460 ± 3469 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc

Heregulin selectively upregulates vascular endothelial growth factor secretion in cancer cells and stimulates angiogenesis Lily Yen1, Xiao-Li You1, Ala-Eddin Al Moustafa1, Gerald Batist1,2, Nancy E Hynes3, Sylvie Mader4, Sylvain Meloche5 and Moulay A Alaoui-Jamali*,1,2 1

Department of Medicine, Oncology, Lady Davis Institute for Medical Research of the Sir Mortimer B. Davis Jewish General Hospital, McGill University, Montreal, Quebec H3T 1E2, Canada; 2Department of Pharmacology/Therapeutics and McGill Centre for Translational Research in Cancer, McGill University, Montreal, Quebec H3T 1E2, Canada; 3Friedrich Miescher-Institute, P.O. Box 2543, CH-4002 Basel, Switzerland; 4Department of Biochemistry, University of Montreal, Quebec H3T 1P1, Canada; 5Centre de Recherche, Hotel Dieu de Montreal and Department of Pharmacology, University of Montreal, Quebec H2W 1T8, Canada

The interaction between the erbB tyrosine kinase receptors and their ligands plays an important role in tumor growth via the regulation of autocrine and paracrine loops. We report the eect of heregulin b1, the ligand for erbB-3 and erbB-4 receptors, on the regulation of vascular endothelial growth factor (VEGF) expression, using a panel of breast and lung cancer cell lines with constitutive erbB-2 overexpression or engineered to stably overexpress the erbB-2 receptor. We demonstrate that heregulin b1 induces VEGF secretion in most cancer cell lines, while no signi®cant eect was observed in normal human mammary and bronchial primary cells. Overexpression of erbB-2 receptor results in induction of the basal level of VEGF and exposure to heregulin further enhances VEGF secretion. This is associated with increased VEGF mRNA expression. In contrast, VEGF induction is signi®cantly decreased in a T47D cell line where erbB-2 is functionally inactivated. Conditioned media from heregulin-treated cancer cells, but not from normal cells, stimulates endothelial cell proliferation; this paracrine stimulation is inhibited by co-exposure to a speci®c VEGF neutralizing antibody. Furthermore, heregulin-mediated angiogenesis is observed in the in vivo CAM assay. This study reports the ®rst evidence of VEGF regulation by heregulin in cancer cells. Oncogene (2000) 19, 3460 ± 3469. Keywords: ErbB receptors; heregulin b1; VEGF; cell proliferation; HUVEC Introduction The erbB family of tyrosine kinase receptors includes EGFR (HER-1/erbB-1), erbB-2 (HER-2), erbB-3 (HER3) and erbB-4 (HER-4). These receptors are characterized by a broad signaling diversity attributed to their propensity to form a speci®c network of heterodimers upon ligand stimulation. They have been implicated in the regulation of various physiological processes including cell cycle, cell dierentiation, cell ± cell interaction, cytokine signaling, and stress response (reviewed in Riese and Stern, 1998; Alroy and Yarden, 1997).

*Correspondence: MA Alaoui-Jamali, Lady Davis Institute for Medical Research, Room 523, 3755 Cote Ste-Catherine, Montreal, Quebec H3T 1E2, Canada Received 14 February 2000; revised 2 May 2000; accepted 16 May 2000

Dysregulation of these receptors is commonly found in various types of human cancers including lung, breast, ovary, prostate, gastric, bladder and kidney carcinomas, as well as in osteosarcomas and neuroblastomas (reviewed in Dougall et al., 1994). Several clinical studies have reported that overexpression of the erbB receptors, particularly erbB-2, predicts poor prognosis (reviewed in Dougall et al., 1994). Although the molecular mechanism(s) by which erbB-2 aects clinical prognosis has yet to be established, there is experimental evidence indicating that overexpression of the erbB-2 receptor can confer a cell growth advantage and enhance metastatic potential (Yu et al., 1994; Tan et al., 1997). The worsened prognosis associated with tumors overexpressing erbB receptors has been attributed to dysregulation of growth factor-mediated paracrine loops, particularly in relation to tumor angiogenesis (reviewed in Pluda, 1997). This hypothesis was based on the observation that EGF, the ligand for the erbB member EGFR, upregulates vascular endothelial growth factor (VEGF) (Tsai et al., 1995; Enholm et al., 1997), while neutralizing antibodies against EGFR and erbB-2 receptors resulted in downregulation of VEGF and tumor growth inhibition (Perrotte et al., 1999; Viloria-Petit et al., 1997). Accumulating evidence supports an association between angiogenesis and the processes of tumor invasion and metastasis (Pluda, 1997). Several angiogenic factors and their receptors have been identi®ed as important mediators for angiogenesis. These include VEGF (Ferrara et al., 1991; Plate et al., 1992), VEGFrelated protein (vrp) (Lee et al., 1996), VEGF-B, -C, -D and -E (Meyer et al., 1999; Yamada et al., 1997; Olofsson et al., 1996; Joukov et al., 1996), the placenta growth factor (PlGF) (Maglione et al., 1991), and basic and acidic ®broblast growth factor (Friesel and Maciag, 1995). Among these angiogenic factors, VEGF has been shown to be a principal tumor angiogenic factor, and play a role in several endothelial-speci®c functions (Dvorak et al., 1995). It is a potent angiogenic factor that can stimulate endothelial cellmitogenesis (Dvorak et al., 1995). Excessive production of VEGF can occur from tumor cells as well as the in®ltrating immune cells that penetrate solid tumors. In several studies, overexpression of VEGF has been detected in tumor cells (Plate et al., 1992; Ferrara et al., 1992; Rossler et al., 1999; Brown et al., 1997), and was associated with high metastatic potential (Balbay et al., 1999; Pluda, 1997; Claey et al., 1996). Additionally, inhibition of VEGF signaling has been

Regulation of VEGF by heregulin b1 L Yen et al

shown to impair tumor growth (Im et al., 1999; Cheng et al. 1996; Lin et al., 1998) and suppress tumor metastasis (Rofstad and Danielsen, 1999; Melnyk et al., 1999; Skobe et al., 1997). Seven isoforms of VEGF have been isolated to date, as a result of alternative splicing one gene: VEGF206, VEGF189, VEGF183, VEGF165, VEGF148, VEGF145 and VEGF121, (Jingjing et al., 1999; Whittle et al., 1999; Poltorak et al., 1997; Tischer et al., 1989; Ferrara et al., 1991). VEGF189 and VEGF206 isoforms were shown to be associated with heparin-containing proteoglycans on the cell surface or extracellular matrix, while VEGF121, VEGF145 and VEGF165 are soluble secretory proteins (Park et al., 1993). VEGF is secreted as disul®de-linked homodimers of approximately 46 kDa which bind to at least two receptortyrosine kinases, the fms-like tyrosine kinase VEGFR-1 (¯t-1) and VEGFR-2 (¯k-1/KDR) which are speci®cally expressed in tumor endothelial cells (Matthews et al., 1991; de Vries et al., 1992; Terman et al., 1992; Brown et al. 1997; Dvorak et al., 1995; Pluda, 1997), supporting a paracrine mechanism of tumor angiogenesis. Both VEGFR-1 and VEGFR-2 were shown to be phosphorylated in endothelial cells upon ligand stimulation (Waltenberger et al., 1994). Several oncogenes, growth factors, hormones and hypoxia have been shown to upregulate angiogenic factors including VEGF, also called vascular permeability factor or vasculotropin (reviewed in Claey and Robinson, 1996). However, there is no data regarding VEGF regulation by the erbB-3 and erbB-4 ligand, heregulin. Given the wide biological diversity and the implication of erbB signaling following stimulation by heregulin, we examined the regulation of the VEGF secretory isoform, VEGF165, in relation to the level of erbB-2. In this study, we report the ®rst evidence that heregulin b1 (HRG) stimulates VEGF expression in cancer cells but not in normal primary cells established from breast and bronchial tissues. This stimulation results in endothelial cell proliferation and increased angiogenesis. Our study also supports a contribution of erbB-2 in heregulin-mediated VEGF regulation in cancer cells.

Results Stimulation of receptor phosphorylation in cell lines expressing erbB-2 Figure 1 shows the basal level of expression of erbB-2 in the cell lines used in this study. As indicated in Figure 1a, the human breast adenocarcinoma cell line SKBR3 and the Non-Small Cell Lung Cancer (NSCLC) cell lines H322 and H661 constitutively overexpress erbB-2 which is autophosphorylated (Figure 1b). Addition of HRG leads to erbB-2 activation, as evidenced by further enhancement of the level of receptor tyrosine phosphorylation (Figure 1b). Next, we looked at cell lines that normally express low levels of erbB-2, and the same cells engineered to overexpress erbB-2 (Figure 1c). The level of erbB-2 receptor expression in control and transfected cells is shown in Figure 1c, and its phosphorylation status, in the absence and presence of HRG, is shown in Figure 1d. Treatment with HRG resulted in enhanced receptor

3461

Figure 1 Expression and phosphorylation of the erbB-2 receptor in the cell lines used in this study. Cell extracts were collected from subcon¯uent cells, unstimulated or stimulated with 100 ng/ ml HRG, and then subjected to Western blot analysis to detect erbB-2 (a and c, upper panels). Western blot analysis of cell lines with constitutive overexpression of erbB-2 (SKBR3, H661 and H322) (a), erbB-2 transfectants (c), and T47D-puro and T47D-5R (c). Blots were stripped and subsequently reprobed with antiGAPDH. Aliquots of total protein were subjected to immunoprecipitation using a polyclonal antibody against erbB-2, then immunoprecipitates were subjected to immunoblotting with antiphosphotyrosine (b and d, upper panels) as described in Materials and methods. Phosphotyrosine expression was determined in cell lines with constitutive overexpression of erbB-2 (SKBR3, H322, and H661) (b), erbB-2 transfectants (d), and T47D-puro and T47D-5R (d). Membranes were subsequently stripped and reblotted with antibody to detect erbB-2

tyrosine phosphorylation in the erbB-2 transfectants. The phosphorylation status of B2BN1 and B2BE6 cells was reported earlier (You et al., 1998; Noguchi et al., 1993). Figure 1c also shows the expression of erbB-2 in the control T47D-puro and the T47D-5R cells stably expressing an ectopic single chain monoclonal antibody to erbB-2, which abrogates the expression of erbB-2 on the cell surface (Beerli et al., 1994). This is re¯ected by the fast electrophoretic migration of erbB-2 because of its retention in the endoplasmic reticulum, in agreement Oncogene


3462

with earlier reports (Graus-Porta et al., 1995; Beerli et al., 1994). While HRG induced erbB-2 tyrosine phosphorylation in T47D-puro cells, no phosphorylation was detected in T47D-5R cells (Figure 1d). Heregulin stimulates VEGF secretion independent of its mitogenic activity Figure 2a demonstrates the eect of HRG on the basal expression of VEGF in SKBR3, H322 and H661 cells, all of which constitutively overexpress erbB-2. VEGF stimulation by HRG was dose related (Figure 2a). The maximum stimulation was observed on cells treated with 100 ng/ml HRG for 48 h. This condition was used for subsequent experiments. Figure 2b shows the mitogenic eect of HRG on the SKBR3, H322 and H661 cell lines. Treatment with HRG resulted in a dose-dependent stimulation of cell growth in H322, while a slight stimulation or no

stimulation was observed in SKBR3 and H661 cells, respectively (Figure 2b). The highest mitogenic activity was observed in the H322 cell line. Compared to 1 and 10 ng/ml HRG of medium, a concentration of 100 ng/ ml HRG induced a potent stimulation of proliferation. A concentration of 200 ng/ml HRG also stimulated proliferation of these cells, although no signi®cant dierence was observed compared to 100 ng/ml HRG (not shown). Figure 3a shows the eect of HRG on VEGF secretion in various erbB transfectant cell lines and in T47D-5R cells, where erbB-2 cell surface expression was abolished. Treatment with HRG resulted in approximately two- to over ®vefold increase in VEGF secretion in the control MCF7-neo22 and T47D-puro cells. In the absence of HRG, the basal level of VEGF was signi®cantly higher in the erbB-2 transfectants (MCF7-HER218 and B2BE6) compared to the control cell lines; upon treatment of the MCF7-HER218 and

Figure 2 Mitogenic activity of heregulin and its eect on the secretion of VEGF in erbB-2 overexpressing cell lines. (a) Cell lines with constitutive overexpression of erbB-2 (SKBR3, H322, and H661) were serum starved for 24 h, then treated with increasing doses of HRG for 48 h. The level of VEGF in the supernatant was determined by the ELISA assay, and normalized to the amount of protein in the cell extracts. (b) Cells were serum-starved for 24 h, and then treated continuously with HRG for 96 h. Cell proliferation was determined by the MTT assay. The per cent cell growth was calculated in comparison to unstimulated cells. The dotted line on each graph indicates an arbitrary value of 100% growth assigned for control unstimulated cells. Values were calculated based on the average of at least ®ve independent experiments, each condition in quadruplicate. * Indicates a signi®cant dierence at P50.05, compared to unstimulated cells Oncogene


3463

Figure 3 Mitogenic activity of heregulin and its eect on the secretion of VEGF in transfected cell lines. (a) Cells stably transfected to overexpress erbB-2 (MCF7-HER218, BT20-erbB2, and B2BE6), cells with abolished erbB-2 cell surface expression (T47D-5R) and their control cell lines (MCF7-neo22, BT20-neo, B2BN1, and T47D-puro) were treated with 100 ng/ml HRG and analysis of VEGF secretion was performed as described in Figure 2. (b) Cells were serum-starved for 24 h, and then treated continuously with HRG for 96 h. Cell proliferation was determined by the MTT assay. Note that only the per cent growth rates of HRG-treated cells are shown, in comparison to unstimulated cells assigned 100% (dotted line)

BT20-erbB2 transfectants with HRG, VEGF secretion was greatly enhanced. Interestingly, blockade of erbB-2 surface expression in T47D-5R cells led to decreases in the basal expression of VEGF and HRG-induced VEGF stimulation. As shown in Figure 3b, HRG stimulated cellular proliferation in MCF7- and BT20-derived cells, although no signi®cant dierence was observed between erbB-2 transfectants and their respective neomycin-transfected controls. In B2BN1 and B2BE6 cells, HRG induced growth inhibition (Figure 3b, lanes 5 and 6). In primary normal human mammary (NHMEC) and bronchial (NHBEC) epithelial cells, HRG failed to induce VEGF secretion (Figure 4a). This was con®rmed in several primary cells established from

dierent individuals (not shown). No mitogenic activity was seen following HRG treatment of NHMEC and NHBEC normal primary cells (Figure 4b). Heregulin-mediated enhancement of VEGF secretion leads to stimulation of endothelial cell proliferation and in vivo neovascularization To examine the functional activity of the secreted VEGF, we have examined human umbilical vein epithelial cell (HUVEC) proliferation in the presence of conditioned media from HRG-stimulated cells. As shown in Figure 5, no mitogenic activity was seen following HRG treatment of HUVEC cells (Figure 5, lane 3). However, recombinant hVEGF treatment signi®cantly induced HUVEC proliferation (Figure 5, Oncogene


3464

lane 2). Compared to supernates from unstimulated cells, supernates from HRG-stimulated cancer cells induced HUVEC cell proliferation (Figure 5, lanes 5, 8, 11, 14 and 17). Furthermore, addition of VEGF165 neutralizing antibody Ab293A to the conditioned medium from HRG-stimulated cells prevented HUVEC cell proliferation. Consistent with the ELISA data in normal primary cells, which showed lack of VEGF induction following

HRG treatment (Figure 4a), conditioned media from primary normal mammary and bronchial epithelial cells failed to stimulate HUVEC endothelial cell proliferation (Figure 5, lane 20). The involvement of HRG in mediating angiogenesis was assessed in vivo (Figure 6) by implantation of a monolayer of MCF7-neo22 or MCF7-HER218 cells onto chick embryo chorioallantoic membrane (CAM). Implantation of CAMs for 7 days with HRG-treated cells resulted in a marked increase in vascularization (Figure 6, second and fourth panels), in comparison to unstimulated cells (Figure 6, ®rst and third panels). A dense vascular network following HRG stimulation was observed with both MCF7-neo22 and MCF7HER218. It should be noted that unstimulated MCF7HER218 cells shows more branched neovascularization compared to MCF7-neo22 controls, supporting our data regarding the paracrine stimulation of HUVEC cell growth. Induction of VEGF gene expression by heregulin occurs at the mRNA level

Figure 4 Mitogenic activity of heregulin and its eect on the secretion of VEGF in normal mammary and bronchial primary cells. Normal primary human mammary (NHMEC) and bronchial (NHBEC) epithelial cells were treated with 100 ng/ml HRG and assayed for VEGF production (a) and cell proliferation (b) as described in Figure 2. The per cent growth rates of HRG-treated cells are shown, in comparison to unstimulated cells de®ned as 100% (dotted line)

Figure 7a demonstrates a time course induction of VEGF mRNA expression by HRG in MCF7 cells. An increase in the steady-state levels of VEGF mRNA was seen as early as 1 h (Figure 7a) following exposure to HRG. Figure 7b shows that HRG stimulation leads to enhanced VEGF mRNA expression (MCF7-neo22, lane 2) and that overexpression of erbB-2 (MCF7HER218, lane 4) further enhances VEGF mRNA expression. The VEGF mRNA induction by HRG

Figure 5 Eect of conditioned medium from untreated and HRG-treated cells on HUVEC cell proliferation. HUVEC cells were grown in a mixture of 30% conditioned medium and 70% regular serum-free medium. Conditioned medium was obtained from precon¯uent cultures of cancer cells, unstimulated or stimulated with 100 ng/ml HRG. HUVEC cells incubated with conditioned medium were kept undisturbed for an additional 72 h. When indicated, VEGF in the conditioned medium was neutralized using a polyclonal neutralizing antibody against human VEGF165 or normal goat IgG as a negative control. Cell growth was examined using the MTT assay. Recombinant human VEGF165, at a concentration of 5 ng/ml, was used as a positive control. * Indicates a signi®cant dierence at P50.05, compared to unstimulated cells Oncogene


3465

Figure 6 HRG induces an angiogenic response in vivo. Chick embryos were monitored following implantation of the chorioallantoic membrane (CAM) with coverslips containing unstimulated cells (panels 1 and 3) or cells stimulated with 100 ng/ ml HRG (panels 2 and 4), as described in Materials and methods. Photographs of CAM angiogenesis in the region implanted with coverslips were taken at day 7 post-implantation. Note the extent of neovessel formation in HRG-treated cells

Figure 7 Heregulin induces VEGF mRNA expression. (a) Total RNA from MCF cells was isolated at the indicated time points following treatment with 100 ng/ml HRG and used for Northern blot analysis of VEGF165 mRNA. (b) Cells were treated with 100 ng/ml HRG and total RNA was isolated for 24 h posttreatment. RNA was analysed by Northern blotting using a 419 bp VEGF165-speci®c probe. Loading was controlled by staining the RNA gel with ethidium bromide

was also very high in the H322 cell line, which constitutively overexpresses the erbB-2 receptor. Measurement of mRNA stability over a period of 4 h did not reveal any signi®cant dierences in VEGF mRNA stability between unstimulated and HRG-stimulated MCF7-neo22 and MCF7-HER218 cells. In both cell lines, the half-life of VEGF mRNA was between 2 and 3 h (data not shown). Discussion Aberrant growth factor and growth factor receptor expression/function is a common feature in many cancers, often resulting in abnormal autocrine and

paracrine regulation of cell growth (Cross and Dexter, 1991; Kumar and Mendelsohn, 1991). The tyrosine kinase receptors of the erbB family are overexpressed and/or ampli®ed in several types of cancers including breast and lung carcinomas (reviewed in Dougall et al., 1994). Several studies have reported that overexpression of members of the erbB family, particularly erbB2, is a predictor of poor prognosis and shorter survival in the clinical setting (Dougall et al., 1994). While no speci®c ligand has yet been identi®ed for the erbB-2 receptor, the ligand for erbB-3 and erbB-4 receptors has been characterized as a group of peptides collectively referred to as heregulins, also known as neuregulins (NRG), neu dierentiation factors (NDF), and glial growth factors (Holmes et al., 1992; Chang et al., 1997; Carraway, et al., 1997). Two families of heregulins, NRG1 and NRG2, are each subdivided into two major isoforms, a and b, based on their EGFR-like domains (Chang et al. 1997). Heregulin b has a higher anity for erbB receptors than the a isoform (Tzahar et al., 1994). Additionally, new members of the heregulin family, namely NRG3 and NRG4, have been identi®ed (Harari et al., 1999; Zhang et al., 1997). Heregulins bind to erbB-3 and erbB-4 with low and high anity, respectively, and also interact with heterodimers between EGFR/erbB-3, EGFR/ erbB-4, erbB-2/erbB-3 and erbB-2/erbB-4 (reviewed in Riese and Stern, 1998; Alroy and Yarden, 1997). Although heregulin does not bind directly to erbB-2, it can activate erbB-2 via transmodulation and transphosphorylation through heterodimer formation with ligand-bound erbB-3 and erbB-4. This was re¯ected by enhanced erbB-2 phosphorylation upon heregulin treatment in most of the erbB-2-overexpressing cancer cell lines that we tested. The erbB-2 receptor has been shown to play an essential role in the activation of erbB signaling by heregulin. This is especially true for the kinase defective erbB-3, which is devoid of biochemical activity when expressed alone (Chen et al., 1996); erbB-3 is activated through preferential heterodimer formation with erbB-2 (reviewed in Dougall et al., 1994; Riese and Stern, 1998; Alroy and Yarden, 1997). Additionally, erbB-2 overexpression enhances the binding anities of heregulin Oncogene


3466

Oncogene

and EGF to their receptors by diminishing the rates of ligand dissociation (Karunagaran et al., 1996), and the abolishment of erbB-2 expression at the cell surface impairs ligand binding and erbB signaling (GrausPorta et al., 1995). Consistently, the relatively low ligand binding anity of erbB-3 is augmented by coexpression of erbB-2 (Sliwkowski et al., 1994). These observations led to the conclusion that EGFR, erbB-3, and erbB-4 receptors compete for interaction with erbB-2 to form heterodimers (Graus-Porta et al., 1997; Riese and Stern, 1998; Alroy and Yarden, 1997). This study provides evidence for the regulation of the angiogenic factor, VEGF, by heregulin. Using a panel of cancer cell lines, we demonstrate that stimulation of erbB receptors by heregulin signi®cantly enhances VEGF secretion, compared to unstimulated controls. Consistent with this ®nding, we show that conditioned media from heregulin-stimulated cells promotes the paracrine stimulation of HUVEC endothelial cells. HUVEC cell stimulation by conditioned medium can not be attributed to heregulin present in conditioned medium, since exposure of HUVEC cells to heregulin alone did not stimulate HUVEC proliferation. Furthermore, addition of speci®c anti-VEGF neutralizing antibody prevents the HUVEC cell stimulation by conditioned medium. Further evidence implicating heregulin-mediated regulation of angiogenesis was shown using the in vivo CAM assay. CAMs implanted with MCF7-neo22 or MCF7-HER218 cells stimulated with heregulin showed increased vascularization, compared to CAMs implanted with untreated cells. We observed that the overexpression of erbB-2 enhances the basal expression of VEGF (in the absence of heregulin stimulation). Addition of heregulin further potentiates VEGF production in the erbB-2 transfectants, suggesting that the formation of heterodimers involving erbB-2 may play a role in VEGF regulation. This is supported by data obtained with T47D-5R, a cell line where the erbB-2 receptor surface expression was abolished. VEGF production in these T47D-5R cells was decreased compared to the T47D-puro control. The role of tyrosine kinases in the regulation of angiogenesis is supported by earlier reports showing that tyrosine kinase inhibitors, such as genistein, inhibit transcriptional and post-transcriptional hypoxia-induced VEGF expression (Mukhopadhyay et al., 1995; Levy et al., 1996). Since erbB-2 is (1) the preferred heterodimer partner of the three other members of erbB family; (2) essential for signal transmission by heregulin; and (3) enhances the binding anities of heregulin to its receptors, these mechanisms may all contribute to the enhanced VEGF secretion seen in our cell models. It is worth noting that loss of erbB-2 cell surface expression did not completely abolish VEGF secretion. This suggests that additional factors besides erbB-2 may mediate the response to heregulin. For example, interaction of heregulin with its cognate receptors (erbB-3 and erbB-4) with or without erbB-2 as a dimer partner or the relative levels of expression of the other members of the erbB family members could aect VEGF expression. Treatment with heregulin induced the secretion of VEGF in most cancer cell lines tested, while heregulininduced mitogenesis was cell-type dependent. For

example, heregulin has only a minor or no eect on SKBR3, H661, B2BN1 and B2BE6 cell proliferation, while the same dose resulted in a substantial induction of VEGF secretion. These observations suggest that heregulin-induced mitogenic activity and heregulininduced VEGF secretion act through distinct mechanisms; the increased VEGF secretion is not merely the consequence of cell proliferation. This hypothesis is supported by earlier studies showing that angiogenesis is independent of tumor cell proliferation (Holmgren et al., 1995) and that several cell cycle-related genes (p21, p16 and p27) were either unaected or decreased in VEGF-treated cells (Watanabe et al., 1997; Brown et al., 1997). In human normal non-immortalized primary cells, which express very low to undetectable levels of the erbB receptors (not shown), heregulin failed to stimulate cell proliferation and VEGF secretion, suggesting that heregulin-mediated VEGF secretion is associated with a neoplastic phenotype. Indeed, transforming oncogenes and tumor promoters such as v-ras and v-raf (Grugel et al., 1995), mutant ras (Rak et al., 1995), and phorbol myristate-12,13-acetate (PMA) (Enholm et al., 1997) have been shown to induce VEGF expression. Regulation of VEGF occurs partly at the transcriptional level and mostly at the post-transcriptional level (Shweiki et al., 1992; Stein et al., 1995; Levy et al., 1996). The half-life of VEGF was reported to be approximately 40 min under normoxic conditions, which increased to approximately 110 min under hypoxic conditions (Levy et al., 1996). Under our conditions, the increased levels of VEGF mRNA following stimulation by heregulin can not be explained by a change in mRNA stability. The halflife was not signi®cantly aected between unstimulated and heregulin-stimulated cells (not shown). Several erbB-coupled signaling pathways including the src/raf/MAPK pathway (reviewed in Dougall et al., 1994) mediate VEGF regulation by hypoxia (Rak et al., 2000; Arbiser et al., 1997; Milanini et al., 1998; Mukhopadhyay et al., 1995). We and others have shown that the ras pathway plays an important role in erbB signaling (Ben-Levy et al., 1994; Yen et al., 1997), and that heregulin-induced mitogenesis involves the activation of MAPK (Marte et al., 1995). VEGF mRNA expression has been shown to be upregulated by the ras oncoprotein (Feldkamp et al., 1999; Arbiser et al., 1997; Rak et al., 1995). Additionally, inhibitors of ras function such as dominant negative ras mutants (Feldkamp et al., 1999), farnesyl-transferase inhibitors (Feldkamp et al., 1999) and lovastatin (Feleszko et al., 1999) are able to suppress VEGF production in rasoverexpressing or ras-transformed cells. Furthermore, inhibition of PI3-K (which has been shown to play an important role in erbB signaling, particularly with respect to erbB-3/erbB-2 heterodimers) led to inhibition of VEGF expression and angiogenesis (Rak et al., 2000; Arbiser et al., 1997). Taken together, these observations suggest that VEGF and erbB receptors share similar signaling pathways, which may also account for the biological eect of heregulin on the regulation of VEGF. Further studies will determine the role of heregulin signaling through erbB receptors on the regulation of VEGF in cancer cells. In summary, this study supports cooperation between members of the erbB family of tyrosine kinase


receptors in the paracrine regulation of angiogenesis and emphasizes the contribution of erbB-2 heterodimers to this regulation. Our results suggest that this upregulation of angiogenesis may contribute to the poor prognosis and enhanced metastatic potential observed in the clinical setting in patients with erbBoverexpressing cancers.

Materials and methods Cell lines and cell culture The cell lines used in this study include the human breast adenocarcinomas: MCF7, SKBR3, T47D and BT20, and the Non-Small Cell Lung Cancer (NSCLC) cell lines H322 (bronchioalveolar carcinoma) and H661 (Large cell carcinoma). These cells were obtained from the American Type Culture Collection (ATCC). The MCF7 clones stably transfected with erbB-2 (MCF7-HER218) and its neomycin control (MCF7-neo22) (Benz et al., 1993) were a generous gift from Dr C Kent Osborne. The Simian virus 40 large Tantigen immortalized human bronchial cells transfected with the neomycin gene (B2BN1) or the wild type human erbB-2 gene (B2BE6) were kindly provided by Dr Brenda I Gerwin (Noguchi et al., 1993). The T47D-5R subline stably expressing a gene encoding a single chain antibody that speci®cally abolishes erbB-2 expression on the cell surface and the control T47D-puromycin were described earlier (Jannot et al., 1996). Cancer cells were maintained in the appropriate media (Mediatech, Washington, DC, USA) described by the ATCC supplemented with 10% fetal bovine serum and penicillin-streptomycin antibiotics. Normal human primary mammary (NHMEC) and bronchial (NHBEC) epithelial cells, and human umbilical vein endothelial cells (HUVEC) were obtained from Clonetics Corp. NHMEC and NHBEC cells were maintained in medium supplied by the manufacturer (BEGM CC-3170 and MEGM CC-3150 BulletKits, Clonetics, Walkersville, MD, USA). HUVEC cells were maintained in M199 medium supplemented with 2 mM L-glutamine, 10% fetal bovine serum (HyClone Laboratories, Logan, UT, USA), 90 mg/ml heparin sulfate (Sigma, St. Louis, MO, USA), and 20 mg/ml endothelial cell growth supplement (Collaborative Biomedical Products, Bedford, MA, USA), 100 mg/ml penicillin and 100 mg/ml streptomycin. All cells were maintained in an atmosphere of 5% CO2. Stable expression of human erbB-2 BT20 cells were seeded at a density of 36105 cells per well in 6-well culture plates and left to grow until 75% con¯uence. Cells were then transfected with 8 mg LipofectAMINE (GIBCO-BRL, Rockville, MD, USA) and 2 mg of either the erbB-2 expression construct, pEV7-HER2, or the pEV7 neomycin control (Zhang et al., 1996) according to manufacturer's instructions. Forty-eight hours after transfection, cells were split into 150 mm dishes and selected with 75 mg/ml neomycin (G418-sulfate, Mediatech). Following 3 weeks of selection, individual clones were collected using cloning rings. Cell proliferation assay Exponentially growing cells (1 ± 36103 cells per 100 ml) were seeded in 96-well plates and incubated for 16 h. Next, cells were starved for 24 h in serum-free medium, then treated continuously with heregulin b1 (HRG) (Neomarkers, Freemont, CA, USA). Cell proliferation was evaluated 96 h later, using the 3-(4,5-dimethylthiazo-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay as previously described (You et al.,

1998). Data are expressed as per cent cell growth in comparison to untreated cells.

3467

Western blot and immunoprecipitation analysis Cells at 60 ± 70% con¯uence were starved for 24 h in serumfree medium, and then treated with HRG for 5 ± 10 min when indicated. Cells were collected, washed twice in cold PBS, and then lysed directly using lysis buer (1% Triton X-100, 10 mM Tris-HCl pH 8.0, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 0.5% NP-40, 0.5 mM phenylmethylsulfonyl¯uoride, 10 mg/ml leupeptin, 10 mg/ml pepstatin, and 10 mg/ml aprotinin, 1 mM sodium orthovanadate). Cell lysates were incubated on ice for 30 min and the supernatant was collected after 13 000 g centrifugation at 48C for 10 min. Protein lysates were resolved by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose (Costar, Cambridge, MA, USA), blocked overnight in PBS containing 3% dried milk, then incubated with monoclonal antibody against erbB-2 (Ab-3, clone 3B5; Oncogene Research Products, Cambridge, MA, USA). Enhanced chemiluminescence detection was performed using ECL detection reagents (Amersham-Pharmacia, Piscataway, NJ, USA). Blots were subsequently stripped in 100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl pH 6.7, 508C for 30 min, and immunoblotted with monoclonal anti-GAPDH antibody (clone 6C5, Cedarlane Laboratories, Hornby, ON, Canada). For immunoprecipitation, 200 mg of protein were immunoprecipitated with a polyclonal anti-erbB-2 (clone C-18, Santa Cruz Biotechnology, Santa Cruz, CA, USA), as previously described (You et al., 1998). Samples were then loaded onto a 7.5% SDS-polyacrylamide gel, transferred to nitrocellulose (Costar), blocked, then incubated with antibody against erbB-2 (Ab-3, clone 3B5; Oncogene Research), or phosphotyrosine (4G10, Upstate Biotechnology Inc., Lake Placid, NY, USA). Detection was performed as described above for Western blotting. Measurement of VEGF by ELISA Cells were seeded at a concentration of 56103 per well using 96-well plates, and kept undisturbed until 70% con¯uence. Next, cells were starved for 24 h in serum-free medium, then treated with HRG in serum-free medium for 48 h. This condition was based on preliminary experiments indicating that VEGF expression was maximal at 48 h, compared to 6 and 24 h, and remained at that plateau thereafter. The supernates were collected and VEGF165 expression was determined using the Quantakine human VEGF Immunoassay kit (R&D Systems, Minneapolis, MN, USA). VEGF values were normalized to total protein concentration in each well. For calibration, we used human recombinant VEGF165 provided by the supplier to construct a standard curve and to obtain absolute values of VEGF protein content. Paracrine stimulation of endothelial HUVEC cell growth HUVEC cells were seeded at a density of 3000 per well, in 96-well plates, with regular complete medium. Twenty-four hours later, medium was replaced with a mixture of 30% conditioned medium and 70% regular serum-free medium. Conditioned medium was obtained from precon¯uent cultures of cancer cells, unstimulated or stimulated with 100 ng/ml HRG. Medium was centrifuged at 3000 g at 48C, and used immediately or stored at 7808C until use. HUVEC cells incubated with conditioned medium were kept undisturbed for an additional 96 h. When indicated, VEGF in the conditioned medium was neutralized using a polyclonal neutralizing antibody against human VEGF165 (MAb293A, R&D Systems) or normal goat IgG (10 mg/ml, R&D Systems) as a negative control. Cell proliferation was examined using the MTT assay as described above. Cell proliferation was Oncogene


3468

expressed as per cent of cell growth in comparison with controls. Control experiments were conducted in parallel to examine the eect of HRG on HUVEC cell growth. In this case, HRG was added directly to a mixture of 30% control conditioned medium and 70% of serum-free HUVEC medium. Human recombinant VEGF165 dissolved in PBS, at a concentration of 5 ng/ml, was used as a positive control for HUVEC stimulation. Chick embryo chorioallantoic membrane (CAM)-angiogenesis assay The CAM angiogenesis assay was performed as essentially described with modi®cations (Al Moustafa et al., 1988; Brooks et al., 1995). Brie¯y, CAM of 3.5 day-old fertilized white Leghorn chicken embryos were used. Cells were seeded onto 13 mm round glass coverslips at a density of 16105 cells per well in 6-well plates. When the cells reached 80% con¯uence, the cells were incubated with fresh media for 2 h then the coverslips were washed twice with PBS and deposited on the CAM. For heregulin stimulation, 100 ng/ ml heregulin was added 1 h prior to washing and implantation. A round window was opened in the shell of the egg and coverslips were deposited on the CAM, with the cells in direct contact with the surface of the CAM. CAMs were examined daily for 7 days following implantation, when the angiogenic response peaked, then photographed in ovo with an Olympus SZH phase contrast microscope. Northern blot analysis Cells at 70% con¯uence were starved for 24 h in serum-free medium, then treated with 100 ng/ml HRG for the time points indicated. Cells were lysed directly and RNA was extracted using RNA-Plus reagent (Quantum Biotech, Montreal, QC, Canada). Twenty-®ve micrograms of total RNA were resolved by electrophoresis through a 1% denaturing formaldehyde gel, then transferred to Zeta-probe membrane (Bio-Rad, Hercules, CA, USA). Equal loading of

RNA in each lane was evaluated by ethidium bromide staining prior to transfer. Probes were prepared as follows: cDNA from H661 RNA was prepared by reverse transcription (Superscript Preampli®cation System for FirstStrand cDNA synthesis, Gibco-BRL), then subjected to PCR ampli®cation using paired forward and reverse primers for VEGF165 (5'-GCAGCTACTGCCATCCAATCG-3' and 5'TGTCACATCTGCAAGTACGTTCG-3'). The corresponding ampli®ed products were agarose gel-puri®ed, then 32Plabeled (Oligolabeling kit, Amersham-Pharmacia). Membranes were hybridized overnight with probe at 428C, then autoradiographed. Statistical analysis The two-tailed Student's t-test was used to compare VEGF secretion between transfectants and controls, in the absence and presence of HRG. Values with P40.05 were considered signi®cant.

Acknowledgments We sincerely thank Dr C Kent Osborne (UT Health Science Center, San Antonio, TX, USA) for providing the MCF7-neo22 and MCF7-HER218 clones, Dr Brenda I Gerwin (National Cancer Institute, Bethesda, MD, USA) for providing the B2BN1 and B2BE6 bronchial cell lines, and Dr Ke Zhang (Amgen Inc., Thousand Oaks, CA, USA) for the HER2 expression construct (pEV7-HER2) and its control. We also thank John Cho for his helpful technical assistance. This work was supported by grant No. 011320 from the Canadian Breast Cancer Research Initiative (CBCRI), and in part by grant MT 14357 from the Medical Research Council (MRC) of Canada. MA AlaouiJamali is supported by a scientist salary award from the Fonds de la Recherche en SanteÂ du Quebec (FRSQ). L Yen is supported by an FRSQ-FCAR-SanteÂ doctoral scholarship.

References Al Moustafa AE, Raes MB, Saule S and Dieterlen-Lievre F. (1988). Cell Dier. Dev., 25, 119 ± 134. Alroy I and Yarden Y. (1997). FEBS Lett., 410, 83 ± 86. Arbiser JL, Moses MA, Fernandez CA, Ghiso N, Cao Y, Klauber N, Frank D, Brownlee M, Flynn E, Parangi S, Byers HR and Folkman J. (1997). Proc. Natl. Acad. Sci. USA, 94, 861 ± 866. Balbay MD, Pettaway CA, Kuniyasu H, Inoue K, Ramirez E, Li E, Fidler IJ and Dinney CP. (1999). Clin. Cancer Res., 5, 783 ± 789. Beerli RR, Wels W and Hynes NE. (1994). J. Biol. Chem., 269, 23931 ± 23936. Ben-Levy R, Paterson HF, Marshall CJ and Yarden Y. (1994). EMBO J., 13, 3302 ± 3311. Benz CC, Scott GK, Sarup JC, Johnson RM, Tripathy D, Coronado E, Shepard HM and Osborne CK. (1993). Breast Cancer Res. Treat., 24, 85 ± 95. Brooks PC, Stromblad S, Klemke R, Visscher D, Sarkar FH and Cheresh DA. (1995). J. Clin. Invest., 96, 1815 ± 1822. Brown LF, Detmar M, Claey K, Nagy JA, Feng D, Dvorak AM and Dvorak HF. (1997). EXS., 79:233 ± 69, 233 ± 269. Carraway III KL, Weber JL, Unger MJ, Ledesma J, Yu N, Gassmann M and Lai C. (1997). Nature, 387, 512 ± 516. Chang H, Riese II DJ, Gilbert W, Stern DF and McMahan UJ. (1997). Nature, 387, 509 ± 512. Chen X, Levkowitz G, Tzahar E, Karunagaran D, Lavi S, Ben-Baruch N, Leitner O, Ratzkin BJ, Bacus SS and Yarden Y. (1996). J. Biol. Chem., 271, 7620 ± 7629.

Oncogene

Cheng SY, Huang HJ, Nagane M, Ji XD, Wang D, Shih CC, Arap W, Huang CM and Cavenee WK. (1996). Proc. Natl. Acad. Sci. USA, 93, 8502 ± 8507. Claey KP, Brown LF, del Aguila LF, Tognazzi K, Yeo KT, Manseau EJ and Dvorak HF. (1996). Cancer Res., 56, 172 ± 181. Claey KP and Robinson GS. (1996). Cancer Metastasis Rev., 15, 165 ± 176. Cross M and Dexter TM. (1991). Cell, 64, 271 ± 280. de Vries C, Escobedo JA, Ueno H, Houck K, Ferrara N and Williams LT. (1992). Science, 255, 989 ± 991. Dougall WC, Qian X, Peterson NC, Miller MJ, Samanta A and Greene MI. (1994). Oncogene, 9, 2109 ± 2123. Dvorak HF, Brown LF, Detmar M and Dvorak AM. (1995). Am. J. Pathol., 146, 1029 ± 1039. Enholm B, Paavonen K, Ristimaki A, Kumar V, Gunji Y, Klefstrom J, Kivinen L, Laiho M, Olofsson B, Joukov V, Eriksson U and Alitalo K. (1997). Oncogene, 14, 2475 ± 2483. Feldkamp MM, Lau N, Rak J, Kerbel RS and Guha A. (1999). Int. J. Cancer, 81, 118 ± 124. Feleszko W, Balkowiec EZ, Sieberth E, Marczak M, Dabrowska A, Giermasz A, Czajka A and Jakobisiak M. (1999). Int. J. Cancer, 81, 560 ± 567. Ferrara N, Houck KA, Jakeman LB, Winer J and Leung DW. (1991). J. Cell Biochem., 47, 211 ± 218. Ferrara N, Houck K, Jakeman L and Leung DW. (1992). Endocr. Rev., 13, 18 ± 32.


Friesel RE and Maciag T. (1995). FASEB J., 9, 919 ± 925. Graus-Porta D, Beerli RR and Hynes NE. (1995). Mol. Cell Biol., 15, 1182 ± 1191. Graus-Porta D, Beerli RR, Daly JM and Hynes NE. (1997). EMBO J., 16, 1647 ± 1655. Grugel S, Finkenzeller G, Weindel K, Barleon B and Marme D. (1995). J. Biol. Chem., 270, 25915 ± 25919. Harari D, Tzahar E, Romano J, Shelly M, Pierce JH, Andrews GC and Yarden Y. (1999). Oncogene, 18, 2681 ± 2689. Holmes WE, Sliwkowski MX, Akita RW, Henzel WJ, Lee J, Park JW, Yansura D, Abadi N, Raab H and Lewis GD. (1992). Science, 256, 1205 ± 1210. Holmgren L, O'Reilly MS and Folkman J. (1995). Nat. Med., 1, 149 ± 153. Im SA, Gomez-Manzano C, Fueyo J, Liu TJ, Ke LD, Kim JS, Lee HY, Steck PA, Kyritsis AP and Yung WK. (1999). Cancer Res., 59, 895 ± 900. Jannot CB, Beerli RR, Mason S, Gullick WJ and Hynes NE. (1996). Oncogene, 13, 275 ± 282. Jingjing L, Xue Y, Agarwal N and Roque RS. (1999). Invest. Ophthalmol. Vis. Sci., 40, 752 ± 759. Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, Saksela O, Kalkkinen N and Alitalo K. (1996). EMBO J., 15, 290 ± 298. Karunagaran D, Tzahar E, Beerli RR, Chen X, Graus-Porta D, Ratzkin BJ, Seger R, Hynes NE and Yarden Y. (1996). EMBO J., 15, 254 ± 264. Kumar R and Mendelsohn J. (1991). Curr. Opin. Oncol., 3, 70 ± 74. Lee J, Gray A, Yuan J, Luoh SM, Avraham H and Wood WI. (1996). Proc. Natl. Acad. Sci. USA, 93, 1988 ± 1992. Levy AP, Levy NS and Goldberg MA. (1996). J. Biol. Chem., 271, 2746 ± 2753. Lin P, Sankar S, Shan S, Dewhirst MW, Polverini PJ, Quinn TQ and Peters KG. (1998). Cell Growth Dier., 9, 49 ± 58. Maglione D, Guerriero V, Viglietto G, Delli-Bovi P and Persico MG. (1991). Proc. Natl. Acad. Sci. USA, 88, 9267 ± 9271. Marte BM, Graus-Porta D, Jeschke M, Fabbro D, Hynes NE and Taverna D. (1995). Oncogene, 10, 167 ± 175. Matthews W, Jordan CT, Gavin M, Jenkins NA, Copeland NG and Lemischka IR. (1991). Proc. Natl. Acad. Sci. USA, 88, 9026 ± 9030. Melnyk O, Zimmerman M, Kim KJ and Shuman M. (1999). J. Urol., 161, 960 ± 963. Meyer M, Clauss M, Lepple-Wienhues A, Waltenberger J, Augustin HG, Ziche M, Lanz C, Buttner M, Rziha HJ and Dehio C. (1999). EMBO J., 18, 363 ± 374. Milanini J, Vinals F, Pouyssegur J and Pages G. (1998). J. Biol. Chem., 273, 18165 ± 18172. Mukhopadhyay D, Tsiokas L, Zhou XM, Foster D, Brugge JS and Sukhatme VP. (1995). Nature, 375, 577 ± 581. Noguchi M, Murakami M, Bennett W, Lupu R, Hui Jr F, Harris CC and Gerwin BI. (1993). Cancer Res., 53, 2035 ± 2043. Olofsson B, Pajusola K, Kaipainen A, von Euler G, Joukov V, Saksela O, Orpana A, Pettersson RF, Alitalo K and Eriksson U. (1996). Proc. Natl. Acad. Sci. USA, 93, 2576 ± 2581. Park JE, Keller GA and Ferrara N. (1993). Mol. Biol. Cell, 4, 1317 ± 1326. Perrotte P, Matsumoto T, Inoue K, Kuniyasu H, Eve BY, Hicklin DJ, Radinsky R and Dinney CP. (1999). Clin. Cancer Res., 5, 257 ± 265.

Plate KH, Breier G, Weich HA and Risau W. (1992). Nature, 359, 845 ± 848. Pluda JM. (1997). Semin. Oncol., 24, 203 ± 218. Poltorak Z, Cohen T, Sivan R, Kandelis Y, Spira G, Vlodavsky I, Keshet E and Neufeld G. (1997). J. Biol. Chem., 272, 7151 ± 7158. Rak J, Mitsuhashi Y, Bayko L, Filmus J, Shirasawa S, Sasazuki T and Kerbel RS. (1995). Cancer Res., 55, 4575 ± 4580. Rak J, Mitsuhashi Y, Sheehan C, Tamir A, Viloria-Petit A, Filmus J, Mansour SJ, Ahn NG and Kerbel RS. (2000). Cancer Res., 60, 490 ± 498. Riese III DJ and Stern DF. (1998). Bioessays, 20, 41 ± 48. Rofstad EK and Danielsen T. (1999). Br. J. Cancer, 80, 1697 ± 1707. Rossler J, Breit S, Havers W and Schweigerer L. (1999). Int. J. Cancer, 81, 113 ± 117. Shweiki D, Itin A, Soer D and Keshet E. (1992). Nature, 359, 843 ± 845. Skobe M, Rockwell P, Goldstein N, Vosseler S and Fusenig NE. (1997). Nat. Med., 3, 1222 ± 1227. Sliwkowski MX, Schaefer G, Akita RW, Lofgren JA, Fitzpatrick VD, Nuijens A, Fendly BM, Cerione RA, Vandlen RL and Carraway III KL. (1994). J. Biol. Chem., 269, 14661 ± 14665. Stein I, Neeman M, Shweiki D, Itin A and Keshet E. (1995). Mol. Cell Biol., 15, 5363 ± 5368. Tan M, Yao J and Yu D. (1997). Cancer Res., 57, 1199 ± 1205. Terman BI, Dougher-Vermazen M, Carrion ME, Dimitrov D, Armellino DC, Gospodarowicz D and Bohlen P. (1992). Biochem. Biophys. Res. Commun., 187, 1579 ± 1586. Tischer E, Gospodarowicz D, Mitchell R, Silva M, Schilling J, Lau K, Crisp T, Fiddes JC and Abraham JA. (1989). Biochem. Biophys. Res. Commun., 165, 1198 ± 1206. Tsai JC, Goldman CK and Gillespie GY. (1995). J. Neurosurg., 82, 864 ± 873. Tzahar E, Levkowitz G, Karunagaran D, Yi L, Peles E, Lavi S, Chang D, Liu N, Yayon A and Wen D. (1994). J. Biol. Chem., 269, 25226 ± 25233. Viloria-Petit AM, Rak J, Hung MC, Rockwell P, Goldstein N, Fendly B and Kerbel RS. (1997). Am. J. Pathol., 151, 1523 ± 1530. Waltenberger J, Claesson-Welsh L, Siegbahn A, Shibuya M and Heldin CH. (1994). J. Biol. Chem., 269, 26988 ± 26995. Watanabe Y, Lee SW, Detmar M, Ajioka I and Dvorak HF. (1997). Oncogene, 14, 2025 ± 2032. Whittle C, Gillespie K, Harrison R, Mathieson PW and Harper SJ. (1999). Clin. Sci., 97, 303 ± 312. Yamada Y, Nezu J, Shimane M and Hirata Y. (1997). Genomics, 42, 483 ± 488. Yen L, Nie ZR, You XL, Richard S, Langton-Webster BC and Alaoui-Jamali MA. (1997). Oncogene, 14, 1827 ± 1835. You XL, Yen L, Zeng-Rong N, Al Moustafa AE and AlaouiJamali MA. (1998). Oncogene, 17, 3177 ± 3186. Yu D, Wang SS, Dulski KM, Tsai CM, Nicolson GL and Hung MC. (1994). Cancer Res., 54, 3260 ± 3266. Zhang D, Sliwkowski MX, Mark M, Frantz G, Akita R, Sun Y, Hillan K, Crowley C, Brush J and Godowski PJ. (1997). Proc. Natl. Acad. Sci. USA, 94, 9562 ± 9567. Zhang K, Sun J, Liu N, Wen D, Chang D, Thomason A and Yoshinaga SK. (1996). J. Biol. Chem., 271, 3884 ± 3890.

3469

Oncogene