Introduction

New therapeutic approaches for colon cancer rely on biological response modifiers targeting both the tumor and its microenviroment. 1,25-Dihydroxyvitamin D₃ (1,25(OH)₂D₃) induces growth arrest and differentiation of colon cancer cells (Diaz et al., 2000; Palmer et al., 2001) and has antitumor, antimetastic (Evans et al., 2000; Koli and Keski-Oja, 2000) and antiangiogenic (Iseki et al., 1999) effects in vitro and in animal models of colorectal cancer. Several nonhypercalcemic 1,25(OH)₂D₃ derivatives have recently entered clinical trials in patients with colorectal carcinomas and other neoplasms (Gross et al., 1998; Gulliford et al., 1998; Smith et al., 1999; Dalhoff et al., 2003). Despite the strong evidence supporting 1,25(OH)₂D₃ antitumor activity, it is still unclear how these results could be translated into useful therapeutic regimens and which types and individual patients would benefit from 1,25(OH)₂D₃-based treatments. Understanding the antitumor mechanism of 1,25(OH)₂D₃ combined with clinical research will contribute to this end.

Studies of the past 10 years defined a central role of the Id proteins in the control of multiple cellular processes ranging from differentiation to cell proliferation, apoptosis, invasion and migration (Yokota, 2001; Zebedee and Hara, 2001). The mechanism of action of Id family members includes formation of heterodimers with basic helix–loop–helix (bHLH), Ets and Pax transcription factors and direct interaction with critical targets, such as Rb. Their broad regulatory function makes Id proteins vulnerable targets in the process of malignant transformation. Members of the Id family are generally upregulated during tumorigenesis (Israel et al., 1999; Lasorella et al., 2001) and are necessary for tumor angiogenesis (Lyden et al., 1999; Benezra et al., 2001; Sikder et al., 2003).

Id proteins control epithelial cell function. Moderate expression of Id1, Id2 and Id3 in the normal colonic epithelium dramatically increases in colorectal human tumors and human colon carcinoma cell lines (Wilson et al., 2001). Ids have been implicated in several pathways central to colon carcinogenesis. The great majority of colon cancers are initiated by mutation of adenomatous polyposis coli (APC) gene and/or -catenin gene, concomitant with upregulation of c-myc and cyclin D genes (Behrens, 2000). Induction of the -catenin/TCF-4 pathway in colorectal tumors leads to increased Id2 expression (Rockman et al., 2001). Remarkably, strong evidence supports a role for Id2 as a major cause of uncontrolled proliferation in neuroblastomas (Lasorella et al., 2000).

To study the effect of 1,25(OH)₂D₃ on Id expression and on the angiogenic phenotype, we have used a subline of the human SW480 cell line derived from a Dukes' stage B colon carcinoma (Leibovitz et al., 1976) that expresses 1,25(OH)₂D₃ receptor (VDR; SW480-ADH). 1,25(OH)₂D₃ treatment induces proliferation arrest and epithelial differentiation in these cells mainly through induction of E-cadherin expression in adherent junctions and inhibition of TCF-4/-catenin complexes (Palmer et al., 2001). 1,25(OH)₂D₃ has a biphasic effect on SW480-ADH cells. First it causes a rapid increase in association between VDR and -catenin, blocks -catenin interaction with TCF-4 and therefore modulates TCF target genes. These events are followed by nuclear export of -catenin and its localization to the plasma membrane concomitant with E-cadherin protein induction. 1,25(OH)₂D₃ regulates the expression levels of numerous genes involved in transcription, cell adhesion, DNA synthesis, apoptosis, redox status and intracellular signalling (Palmer et al., 2003). Now we demonstrate changes in the Id expression profile due to 1,25(OH)₂D₃ in the VDR-positive SW480-ADH cells. Id1 levels were rapidly increased in hormone-treated cells and remained high during acquisition of a more differentiated phenotype. Conversely, Id2 expression was downregulated and contributed to the 1,25(OH)₂D₃ antiproliferative effect. We also found that 1,25(OH)₂D₃ altered several factors which contribute to the angiogenic phenotype of colon carcinoma cells.

Results

Id1 and Id2 genes are differentially expressed in colon carcinoma cell lines

We have analysed the expression of Id genes in SW480 and SW620 cell lines derived from a primary colon carcinoma and a lymph node metastasis from the same patient (Hewitt et al., 2000), as well as in two SW480 sublines, SW480-ADH and SW480-R. SW480-ADH are adherent, with flat and polygonal morphology, and respond to 1,25(OH)₂D₃ by forming epithelial islets characterized by high levels of E-cadherin expression (Palmer et al., 2001). Conversely, SW480-R cells are rounded, refractile and unresponsive to 1,25(OH)₂D₃ (Palmer et al., 2001). Id1 mRNA and protein were expressed at similar, relatively high levels in SW620, parental SW480 and SW480-R. Interestingly, SW480-ADH cells expressed Id1 at extremely low levels (Figure 1a and b). Conversely, Id2 mRNA and protein were expressed at higher levels in SW480-ADH cells (Figure 1a and b). Id3 mRNA was uniformly low in all the cell lines tested (Figure 1a).

Figure 1.

Id expression in colon carcinoma cell lines. (a) Northern blot analysis of Id1 and Id3 mRNA (left panels) and Id2 mRNA (right panels) levels in human colon carcinoma cell lines. GAPDH and 18S rRNA were used as loading controls. (b) Western blot analysis of Id1 and Id2 protein levels in human colon carcinoma cell lines. -Tubulin was used as loading control

Full figure and legend (65K)

1,25(OH)₂D₃ oppositely regulates Id1 and Id2 genes in SW480-ADH cells

We have investigated the possible link between morphological changes induced by 1,25(OH)₂D₃ in SW480-ADH cells (Palmer et al., 2001; Figure 2a) and Id protein expression. As shown in Figure 2b, 1,25(OH)₂D₃ induced rapid increase in Id1 protein levels, while Id2 levels were reduced. Id1 was increased as early as at 2 h of 1,25(OH)₂D₃ treatment; this increase reached maximum at 16 h and persisted for 24 h. In contrast, Id2 was significantly decreased at 16 h of treatment and the downregulation persisted for 24 h (Figure 2b). 1,25(OH)₂D₃ caused the expected upregulation of E-cadherin (Figure 2b) and morphological changes (Figure 2a), thus indicating proper cellular response.

Figure 2.

Regulation of Id expression by 1,25(OH)₂D₃ in SW480-ADH cells. (a) Phase-contrast micrographs showing differentiation induced by 48 h treatment with 1,25(OH)₂D₃. (b) Western blot analysis of regulation of Id1 and Id2 protein levels by 1,25(OH)₂D₃ treatment of SW480-ADH cells for the indicated time. Induction of E-cadherin expression was used as control for 1,25(OH)₂D₃ response in SW480-ADH cells. -Tubulin was used as loading control. Quantifications normalized against -tubulin are shown in graphs. (c) Immunofluorescence analysis of regulation of Id1 and Id2 expression by 1,25(OH)₂D₃ treatment of SW480-ADH cells for 48 h. Formation of epithelial islets was followed by E-cadherin staining. Images were taken by confocal microscopy

Full figure and legend (177K)

Id1 and Id2 regulation by 1,25(OH)₂D₃ was further confirmed by inmunofluorescence analysis (Figure 2c). In all adhesive, differentiated islets where E-cadherin was upregulated due to 1,25(OH)₂D₃ treatment, Id2 protein was downregulated and localized mainly in the nuclei. Increased Id1 protein was mainly detected in the cytoplasm of differentiated cells. Id3 protein levels and nuclear localization were not modified by 1,25(OH)₂D₃ treatment of SW480-ADH cells (Figure 2c). As was expected, the treatment of SW480-R or SW620 cells lacking VDR with 1,25(OH)₂D₃ yielded no modifications in expression or localization of Id proteins in agreement with the lack of VDR expression (data not shown).

In order to define the mechanism of Id regulation by 1,25(OH)₂D₃, we assessed Id mRNA by Northern blot analysis. Id1 mRNA levels were regulated by 1,25(OH)₂D₃ with kinetics similar to that found for protein levels (Figure 3a). Transcription and translation inhibitors were used to determine the type of regulatory mechanism by which 1,25(OH)₂D₃ is controlling Id levels; however, due to the short half-life of Id1 mRNA, this approach yielded no conclusive results (data not shown). To directly ascertain if Id1 regulation by 1,25(OH)₂D₃ in SW480-ADH cells occurred at transcription level, we used a reporter construct where human Id1 promoter was fused to a luciferase reporter gene. 1,25(OH)₂D₃ caused a consistent five fold induction of promoter activity (Figure 3b). In contrast, Id2 mRNA was rapidly downregulated by 1,25(OH)₂D₃ (Figure 4a). Run-on assays showed that 1,25(OH)₂D₃ caused a decrease in Id2 transcription rate (Figure 4b). These results show that Id1 and Id2 genes are regulated by 1,25(OH)₂D₃ at the transcription level.

Figure 3.

1,25(OH)₂D₃ regulates Id1 transcription in SW480-ADH cells. (a) Northern blot analysis of regulation of Id1 expression by treatment with 1,25(OH)₂D₃ during 4 or 24 h. The graph below shows the scintillation quantification by Instant Imager of Id1 normalized against cyclophilin. (b) Id1 promoter analysis in SW480-ADH treated with 1,25(OH)₂D₃ for 48 h

Full figure and legend (60K)

Figure 4.

1,25(OH)₂D₃ regulates Id2 transcription in SW480-ADH cells. (a) Northern blot analysis of regulation of Id2 expression by treatment with 1,25(OH)₂D₃. The graph below shows the scintillation quantification by Instant Imager of Id2 normalized against GAPDH. (b) Run-on analysis of regulation of transcriptional activity of Id2 by treatment with 1,25(OH)₂D₃ for 4 h. The graph below shows the scintillation quantification by Instant Imager of Id2 normalized against GAPDH. The empty plasmid pBSK is used as a control of unspecific binding

Full figure and legend (62K)

Id1 overexpression had no effect on E-cadherin levels in SW480-ADH cells

E-cadherin induction by 1,25(OH)₂D₃ is one of the major changes associated with the acquisition of a more differentiated phenotype in SW480-ADH cells (Palmer et al., 2001). Seeking for a role of Id1 induction in the regulatory networks that transduce 1,25(OH)₂D₃ effects in colon cancer cells, we have transiently overexpressed Id1 gene in SW480-ADH cells using both retroviral and adenoviral gene transfer systems under conditions which allowed transduction of 100% of the cells (Figure 5a and b). Neither retroviral (Figure 5a) nor adenoviral (Figure 5b) gene transfer of Id1 caused detectable variations in either basal or 1,25(OH)₂D₃-induced E-cadherin levels, despite the high ectopic levels of Id1 achieved using adenoviruses (Figure 5a and b). We have confirmed this result using reporter constructs for two different regions of the human E-cadherin gene promoter, a short 270-bp (-178 to +92) fragment containing the three E-boxes susceptible of Id1 regulation, and a long 1072-bp (-987 to +92) fragment regulated by 1,25(OH)₂D₃ in SW480-ADH cells (data not shown).

Figure 5.

Id1 overexpression does not modulate E-cadherin levels in SW480-ADH cells. Western blot analysis of E-cadherin expression in SW480-ADH cells: (a) infected with retrovirus-iresEGFP (6FP) or retrovirus-Id1-HA-iresEGFP (Id1). The overexpression of exogenous Id1 is detected by Western blot against HA. The graph below shows the quantification of E-cadherin signal normalized against -tubulin. Controls (grey bars), 1,25(OH)₂D3 treatments (black bars). Right images show transduction efficiency by detection of GFP expression and Hoeschst staining. (b) Infected with control adenovirus (mock) or adenovirus-Id1-HA (Id1). Id1 overexpression is detected by Western blot against HA. The graph below shows normalized quantification of E-cadherin signal against -tubulin. Controls (grey bars), 1,25(OH)₂D3 treatments (black bars). Empty bars correspond to not treated infected cells. Control and 1,25(OH)₂D3 treated non infected cells are also included. Right images show transduction efficiency by inmunofluorescent detection of Id1 expression using anti-HA and Hoeschst staining

Full figure and legend (169K)

Id2 overexpression interfered with the antiproliferative effect of 1,25(OH)₂D₃ in SW480-ADH cells

Solid evidence in support of Id2 role in disruption of the G1 checkpoint in neuroblastomas (Lasorella et al., 1996, 2000, 2002; Liu et al., 2004) prompted us to investigate the possible contribution of the Id2 downregulation by 1,25(OH)₂D₃ to the antiproliferative effect of hormone treatment in SW480-ADH cells. For that purpose, we overexpressed Id2 in SW480-ADH cells using retroviral gene transfer. Figure 6a shows Id2 expression levels determined by Western blot and immunofluorescence in pools of SW480-ADH cells infected with Id2-expressing retrovirus. Sustained overexpression of Id2 by itself failed to stimulate S-phase entry of SW480-ADH cells; however, it abrogated the inhibitory effect of 1,25(OH)₂D₃ (Figure 6b). In contrast, Id1 overexpression had no consequence either for proliferation of SW480-ADH cells or for the antiproliferative effect of 1,25(OH)₂D₃ (data not shown). We further confirmed the contribution of Id2 to the antiproliferative action of 1,25(OH)₂D₃ by analysing the consequences of Id2 overexpression on the regulation of previously reported cell cycle targets of 1,25(OH)₂D₃ (Halline et al., 1994; Verlinden et al., 1998; Scaglione-Sewell et al., 2000; Jensen et al., 2001). 1,25(OH)₂D₃ treatment induced a prominent increase in p21 cyclin-dependent kinase inhibitor (CDKI) protein levels, that was abolished by Id2 overexpression (Figure 6c). Levels of cyclin A were significantly reduced by hormone treatment of SW480-ADH cells. Upon Id2 overexpression, cyclin A remained unchanged regardless of 1,25(OH)₂D₃ treatment (Figure 6c).

Figure 6.

Id2 overexpression interferes with 1,25(OH)₂D₃ antiproliferative effect in SW480-ADH cells. (a) Characterization of Id2-overexpressing SW480-ADH cells. SW480-ADH cells were infected with retrovirus-iresEGFP or retrovirus-Id2-iresEGFP. Id2 expression was detected by Western blot (upper panel) and immunofluorescence (lower panels) against Id2 protein. (b) Effect of Id2 overexpression on inhibition of thymidine incorporation by 1,25(OH)₂D₃. (c) Effect of Id2 overexpression on the regulatory effect of 1,25(OH)₂D₃ on p21 and cyclin A levels detected by Western blot

Full figure and legend (111K)

Despite consistent Id2 interference with cell cycle arrest by 1,25(OH)₂D₃, both morphological changes and E-cadherin regulation associated with hormone treatment still occurred in Id2-overexpressing SW480-ADH cells (data not shown).

125(OH)₂D₃ modulates the angiogenic phenotype of SW480-ADH cells

1,25(OH)₂D₃ has been previously shown to block angiogenesis (Iseki et al., 1999). This effect may stem from the modulation of the expression profile of key regulators of vascularization produced by the tumor cells. First, we have characterized the angiogenic profiles of SW480-ADH, SW480-R, SW480 and SW620 colon carcinoma cell lines by testing the media conditioned by these cell lines in the in vivo corneal neovascularization assay. Interestingly, the angiogenic potential of conditioned media (CM) from the more differentiated SW480-ADH cells was greatly reduced compared to the other colon carcinoma cell lines (Figure 7a). Moreover, SW480-ADH CM antagonized the activity of the conventional angiogenesis inducer, basic fibroblast growth factor (bFGF), in the corneal neovascularization assay, unveiling the presence of antiangiogenic factors (Figure 8a). SW480-ADH cells produced significantly lower levels of bFGF (Figure 7b). The expression of vascular endothelial growth factor (VEGF), a major inducer of angiogenesis in colon cancer, was reduced in SW480-ADH cells compared to SW480 parental line (Figure 7c), whereas PDGF-B, a multifunctional factor involved both in vessel growth and maturation, was highly expressed in SW480-ADH but not in SW480-R, SW480 parent or SW620 cells (Figure 7c). Paradoxically, thrombospondin-1 (TSP-1), a major inhibitor of angiogenesis, was markedly lower in SW480-ADH than in the other cell lines tested (Figure 7c).

Figure 7.

Angiogenic phenotype of colon carcinoma cell lines. (a) In vivo angiogenesis assay of CM from colon cancer cell lines. Photos of representative corneas are shown. The table shows positive corneas per total number of corneas. The right graph shows the percentage of positive corneas. (b) Western blot analysis of bFGF expression in CM from colon cancer cell lines. (c) Northern blot analysis of PDGF-B, TSP-1 and VEGF expression in colon cancer cell lines

Full figure and legend (120K)

Figure 8.

Regulation of the angiogenic phenotype of SW480-ADH cells by 1,25(OH)₂D₃. (a) In vivo angiogenesis assay of CM from SW480-ADH cells treated with 1,25(OH)₂D₃ for 48 h. Photos of representative corneas are shown. The table shows positive corneas per total number of corneas. The graph below shows the percentage of positive corneas. The effect of neutralizing antibodies against TSP-1 or VEGF on the angiogenic phenotype of SW480-ADH is shown. (b) Northern blot analysis using poly(A⁺)RNA of regulation of VEGF, TSP-1 and PDGF-B expression by 1,25(OH)₂D₃ treatment. The graphs in the right panels show the scintillation quantification by Instant Imager normalized against GAPDH. (c) Activation of human VEGF promoter by treatment with 1,25(OH)₂D₃ for 48 h. A positive control of the activation of this promoter was performed by treatment with 100 M deferoxamine (DFX) for 48 h. Controls (gray bars), 1,25(OH)₂ and DFX treatments (black bars). (d) Regulation of human TSP-1 promoter in SW480 cells by treatment with 1,25(OH)₂D₃ for 48 h. Controls (gray bars), 125(OH)₂ treatments (black bars)

Full figure and legend (187K)

We have then explored 1,25(OH)₂D₃ effect on the angiogenic phenotype of SW480-ADH cells. Conditioned medium from SW480-ADH cells treated with 1,25(OH)₂D₃ remained nonangiogenic, but hormone treatment abolished its ability to block bFGF-induced angiogenesis in the corneal neovascularization assay (Figure 8a). We have analysed the effect of 1,25(OH)₂D₃ on the levels of angiogenesis modulators in SW480-ADH cells and revealed a simultaneous upregulation of VEGF and TSP-1 by 1,25(OH)₂D₃ (Figure 8b). In contrast, the levels of PDGF-B mRNA were not altered by 1,25(OH)₂D₃ treatment (Figure 8b). To determine the mechanism of VEGF and TSP-1 regulation by 1,25(OH)₂D₃, we performed transactivation assays using reporter constructs encoding a luciferase reporter gene controlled by VEGF or TSP-1 promoters. 1,25(OH)₂D₃ increased both VEGF and TSP-1 promoter activity in the SW480-ADH cells (Figure 8c and d, respectively).

In order to demonstrate the contribution of TSP-1 and VEGF to the changes in the angiogenic phenotype of SW480-ADH cells induced by 1,25(OH)₂D₃, we performed corneal neovascularization assay in the presence of neutralizing antibodies against each of these angiogenic factors. TSP-1 neutralizing antibodies shifted the angiogenic potential of SW480-ADH CM towards stimulatory and abrogated its ability to block bFGF-induced angiogenesis (Figure 8a). These results suggest that TSP-1 is critical for the reduced angiogenic potential of SW480-ADH cells.

The addition of neutralizing TSP-1 antibodies to CM from 1,25(OH)₂D₃-treated SW480-ADH cells caused the shift towards strong induction only in the absence of exogenous bFGF (Figure 8a). However, simultaneous neutralization of both VEGF and TSP-1 suppressed the robust angiogenesis displayed in the presence of TSP-1 antibody alone. Addition of neutralizing antibodies alone had no effect on the corneal neovascularization assay (data not shown). These results support the functional contribution of simultaneous regulation of TSP-1 and VEGF by 1,25(OH)₂D₃ to changes in the angiogenic phenotype of SW480-ADH cells by hormone treatment.

Discussion

1,25(OH)₂D₃ is a pleiotropic agent with a potential for prevention and treatment of multiple neoplasms. Considerable progress has been achieved in the development of synthetic 1,25(OH)₂D₃ analogs capable of interfering with tumor functions with minimal calcemic side effects (Guyton et al., 2001; Nishii and Okano, 2001). However, understanding of 1,25(OH)₂D₃ molecular mechanisms of action in particular tumor types would yield important insight into its target genes or processes and thus aid the development of more effective and better tolerated therapies combining 1,25(OH)₂D₃ analogs with other agents.

SW480-ADH cells, a subline of SW480 colon carcinoma cell line, expresses endogenous 1,25(OH)₂D₃ receptors and has been recently proven as a useful tool to evaluate the potential benefits of 1,25(OH)₂D₃ for the treatment and prevention of colon cancer (Palmer et al., 2003). SW480-ADH cells are profoundly affected by 1,25(OH)₂D₃ treatment, which halts their proliferation and induces phenotypic changes reminiscent of a more differentiated epithelial phenotype (Palmer et al., 2001). This work identifies Id proteins as novel targets of 1,25(OH)₂D₃ in colon carcinoma cells, rapidly modulated in response to hormone treatment. Id1 protein is a much earlier target of 1,25(OH)₂D₃ than E-cadherin, whose upregulation requires longer treatments (Palmer et al., 2001). However, Id2 protein levels are downregulated with a significant delay despite early regulation of transcriptional rate. The hormone-induced changes of Id genes are maintained up to 48 h, suggesting a role for these genes in the maintenance of the antiproliferative and differentiated state induced by 1,25(OH)₂D₃.

Tumorigenicity of SW480-ADH cells is significantly lower compared to SW480-R and SW620 (Palmer et al., 2001). The low expression level of Id1 in SW480-ADH cells is in agreement with the widespread correlation between high Id1 levels, aggressiveness and de-differentiated state of the tumors (Lasorella et al., 2001). 1,25(OH)₂D₃-dependent Id1 increase was associated with a more differentiated phenotype, which apparently contradicts the role ascribed to Id1 in differentiation processes (Yokota, 2001). On the other hand, our results are consistent with epithelial differentiation induced by glucocorticoids in mouse mammary epithelial cells, that is thought to be associated with Id1 increase and subsequent E-cadherin upregulation (Woo et al., 2000). However, in SW480-ADH, forced Id1 overexpression alone was insufficient to drive E-cadherin expression and associated epithelial differentiation. These differences may be due to the dissimilar regulatory networks controlling mouse and human E-cadherin promoter activity. Although transient overexpression of Id1 failed to alter either the kinetics of E-cadherin induction or the morphological changes associated with hormone treatment, it remains possible that Id1 regulation is necessary for the effects of 1,25(OH)₂D₃ on cell growth and differentiation. The regulation of Id1 by 1,25(OH)₂D₃ is apparently tissue specific, because in osteoblastic cells this hormone represses Id1 promoter activity, causing osteoblastic differentiation (Ezura et al., 1997). Additional research is needed to identify Id1 role among the plethora of processes and targets affected by 1,25(OH)₂D₃ in SW480-ADH cells.

Interestingly, in our system Id1 and Id2 were regulated by 1,25(OH)₂D₃ in an opposing manner. c-MYC (Lasorella et al., 2000) and TCF-4/-catenin (Rockman et al., 2001) are the main transcriptional activators of Id2 promoter, both negatively regulated by 1,25(OH)₂D₃ (Palmer et al., 2001). Initially ligand-activated VDR binds to -catenin, reducing the number of active TCF-4/-catenin complexes (Palmer et al., 2001). This mechanism may be responsible for early regulation of Id2 mRNA by 1,25(OH)₂D₃. At later times, additional mechanisms participate: 1,25(OH)₂D₃ induces E-cadherin expression, which sequesters -catenin to the plasma membrane, and also represses c-myc (Palmer et al., 2001); these are events which may contribute to late Id2 mRNA downregulation by 1,25(OH)₂D₃.

Enforced expression of Id2 prevented the antiproliferative effect of 1,25(OH)₂D₃ in SW480-ADH colon carcinoma cells. We showed that previously described targets of 1,25(OH)₂D₃ relevant for G1 progression or S-phase entry are affected in SW480-ADH cells in an Id2-dependent manner. Id2 factors impinge on the cell cycle machinery through at least two independent mechanisms that respectively affect pRb (Lasorella et al., 2000) and p21CDKI (Matsumura et al., 2002). Sustained overexpression of Id2 interferes with the modulation by 1,25(OH)₂D₃ of cyclin A, an S-phase cyclin, under direct E2F transcriptional control (Yam et al., 2002). Therefore, the downregulation of Id2 by hormone treatment most probably impinges on the Rb–E2F pathway and thus impaires S-phase entry. In agreement with our results, the antiproliferative effect of 1,25(OH)₂D₃ on human neuroblastoma also correlates with the regulation of myc-Id2–pRb pathway (Gumireddy et al., 2003). Id2 also interfered with regulation of p21CDKI by 1,25(OH)₂D₃, which most likely also contributed to exit from the cell cycle. Therefore, our results indicate that the effect of 1,25(OH)₂D₃ on proliferation of SW480-ADH cells is conducted by the downregulation of Id2 levels, pointing to a possible role of Id2 in hormone therapy of colon carcinoma.

An important trait defining tumor progression is the acquisition of an angiogenic phenotype, which endows tumors with the armory of factors needed to build a vascular network for the efficient support of local growth and the routes to invade distant new soil. The SW480 parent was strongly angiogenic, with high expression levels of VEGF and bFGF. Its less tumorigenic subline, SW480-ADH, produced bFGF and VEGF at significantly lower levels. Interestingly, the levels of antiangiogenic TSP-1 were also lower in SW480-ADH. However, the low underlying levels of angiogenic stimuli allowed TSP-1 to exert its inhibitory activity and to block the angiogenesis induced by exogenous bFGF in mouse corneal assay. Moreover, high levels of PDGF-B in SW480-ADH secretions on low VEGF background may enhance pericyte association with existing vasculature and thus interfere with capillary sprouting (Ramsauer and D'Amore, 2002). Our results show that 1,25(OH)₂D₃ has a complex regulatory action on the angiogenic phenotype of SW480-ADH cells. Unexpectedly, the hormone upregulates the expression of both VEGF and TSP-1, the two major opposing factors in control of tumor angiogenesis in opposite ways. VEGF induction by 1,25(OH)₂D₃ has been previously reported in osteoblasts (Schlaeppi et al., 1997). In our system, the levels of TSP-1, albeit increased by 1,25(OH)₂D₃, were apparently insufficient to counterbalance the combined proangiogenic activity of exogenous bFGF and higher amounts of endogenous VEGF resulting from hormone treatment in the corneal neovascularization assay.

An important role of Id genes in the control of tumor angiogenesis has been established over the last 5 years. Id factors impinge on the expression levels of key regulators of the angiogenic cascade including v, 6, 3, 4 integrins, hypoxia-inducible factor-1, matrix metalloprotease-2 and FGF receptor-1 (Ruzinova et al., 2003). In mouse embryonic fibroblasts from Id1 knockout mice, levels of antiangiogenic TSP-1 are increased and consequently the angiogenic balance is shifted towards an inhibitory phenotype (Volpert et al., 2002). However, the relevance of this regulatory loop in tumors has not been evaluated in depth. We have compared Id1 levels in a wide range of tumor cell lines and found that Id1 levels in SW480-ADH cells are strikingly low (data not shown). This enabled us to analyse the effect of Id1 overexpression on TSP-1 levels in colon carcinoma cells. Despite the high Id1 levels achieved by adenoviral gene transfer, we were unable to observe changes of the TSP-1 levels in SW480-ADH cells (data not shown). Consistent with our results, spontaneous tumor formation in Pten^+/-Id1^-/- or Pten^+/- Id1^-/-Id3^+/- mice has not been linked with modifications of TSP-1 production by the tumor cells (Ruzinova et al., 2003).

In animal models 1,25(OH)₂D₃ treatment causes significant decrease in tumor microvessel density (Iseki et al., 1999; Mantell et al., 2000; Bernardi et al., 2002). The effect of 1,25(OH)₂D₃ on tumor angiogenesis is a sum of hormone-induced changes in the angiogenic profile of tumor cells and direct 1,25(OH)₂D₃ effects on endothelial cells. Our results raise the possibility that the effects of 1,25(OH)₂D₃ on the angiogenic characteristics of tumor cells may be tumor-type dependent and, therefore, suggest that the potential of 1,25(OH)₂D₃ to halt tumor angiogenesis should be evaluated for each particular tumor.

Materials and methods

Cell culture

SW480, SW480-ADH, SW480-R and SW620 (donated by Dr F del Real) were grown in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FBS. 1,25(OH)₂D₃ (a gift from Dr Lise Binderup, Leo Pharmaceutical products, Copenhagen, Denmark) was dissolved in isopropanol. 1,25(OH)₂D₃ (10^-7 M) was added in fresh medium supplemented with 10% charcoal-treated FCS to remove liposoluble hormones. Treatments were sequentially added and simultaneously stopped, in order to minimize confluency effects.

Promoters and expression vectors

We used plasmids containing the -178/+92 and -987/+92 fragments of the human E-cadherin promoter (Palmer et al., 2001). The -2200/+1 human Id1 promoter fragment (pId1-SB-Luc) was a gift from Dr J Campisi (Nehlin et al., 1997). The construction containing a fragment -2033/+750 of TSP-1 promoter was a gift of Dr Volpert (Dameron et al., 1994). VEGF promoter construct -2274/+1 was donated by Dr Alfranca. We amplified by PCR the Id1 cDNA from pGEX-Id1A (containing the alternative splicing isoform Id1H, a gift from Dr E Hara) using 5'-GCGGATCCACCATGAAAGYCGCCAGTGGCAGC-3' (with a BamH1 site) and 5'-GCGATATCCGCGACACAAGATGCGATCGTC-3' (with an EcoRV site) as sense and antisense oligonucleotides. The PCR product was cut with BamHI and EcorRV and introduced into pcDNA3 (Invitrogene) containing the hemaglutinin (HA) epitope in 3' of EcoRV site to generate pcDNA3-Id1-HA. The bicistronic retroviral expression vector pLZR-Id1-HA-iresEGFP was generated by introducing the Id1-HA fragment from pcDNA3-Id1-HA digested with BamHI/XhoI into the pLZR-iresEGFP (donated by Dr A Bernad) digested with BamHI/XhoI. The adenoviral expression vector pDC315-Id1-HA under cytomegalovirus promoter control was constructed by introducing the previous Id1-HA fragment into pDC315 (AdeMax plasmid for adenovirus vector construction from Microbix Biosystem Inc.) digested with BamHI/SalI. Id2 cDNA obtained by BamHI/XhoI digestion from pcDNA3-Id2 (a gift from Dr Norton) was introduced in pLZR-iresEGFP to generate pLZR-Id2-iresEGFP.

Generation of retroviral particles and infections

The packaging cell line PA317 (ATCC) was used to generate a stable cell line to produce amphotropic retroviral particles containing pLZR-iresEGFP, pLZR-Id1-HA-ires-EGFP or pLZR-Id2-ires-EGFP. The sequence was integrated in PA317 genome by infecting with ecotropic particles obtained from transient transfection in 293T cells (ATCC) cotransfected with the pLZR-iresEGFP, pLZR-Id1-HA-iresEGFP or pLZR-Id2-ires-EGFP together with the expression vector of the retroviral genes gag-env-pol. The supernatants were added together with 8 g/ml polybrene (Sigma, St Louis, MO, USA) to subconfluent PA317 cultures in two cycles of 6 h. Stable cell lines PA317-iresEGFP, PA317-Id1-HA-iresGFP and PA317-Id2-iresGFP were selected by fluorescence-activated cell sorting (FACS) on a FAC-Star PLUS flow cytometer (Becton Dickinson, San Jose, CA, USA). Infection of SW480-ADH cells was performed by adding the supernatants of PA317-iresEGFP, PA317-Id1-HA-iresGFP or PA317-Id2-iresGFP cells together with polybrene to subconfluent cultures of SW480-ADH in two cycles of 12 h, replacing the medium with fresh medium overnight. Pools of cells expressing from moderate to high levels of EGFP were sorted and used for experiments.

Generation of adenoviral particles and infections

293 cells were cotransfected with pDC315-Id1-HA or pDC315 together with the expression vector containing adenovirus type 5 genome lacking the E1 and E3 genes. Viral particles in the supernanants were purified by cesium chloride ultracentrifugation. Adenovirus infection of the SW480-ADH cells was carried out using 30 multiplicity of infection (m.o.i.) in DMEM–0.5% FBS for 2 h, after which the medium was refreshed with DMEM–10% FBS, and 24 h later treatments were performed.

Antibodies

The following antibodies were used: rabbit polyclonal anti-Id1 (Santa Cruz Biotechnology, Inc.), rabbit polyclonal anti-Id2 (Santa Cruz Biotechnology, Inc.), rabbit polyclonal anti-Id-3 (Santa Cruz Biotechnology, Inc.), rat monoclonal anti-human E-cadherin (Zymed), mouse monoclonal anti--tubulin (Sigma), rabbit polyclonal anti-human bFGF (Santa Cruz Biotechnology, Inc.), anti-TSP-1 neutralizing antibodies (Ab-1, Novex), anti-VEGF neutralizing antibodies (R&D), anti-p21 (sc-3976, Santa Cruz Biotechnology, Inc.), anti-cyclin A (14531A, Pharmingen), mouse monoclonal anti-HA (Babko), goat anti-rabbit IgG (H+L) HRP conjugated (ICN Biomedicals), goat anti-mouse IgG (H+L) HRP conjugated (ICN Biomedicals), goat anti-rat IgG (H+L) HRP conjugated (Pierce Chemical and Co.), donkey anti-goat (H+L) HRP conjugated (Santa Cruz Biotechnology, Inc.), goat anti-rat IgG TRITC conjugated (Jackson Immunoreserach Laboratories), goat polyclonal anti-mouse IgG Alexa 488 (Molecular Probes), monkey polyclonal anti-mouse IgG AMCA conjugated Jackson Immunoreserach Laboratories) and goat polyclonal anti-rabbit IgG FITC conjugated (Jackson Immunoresearch Laboratories).

Immunofluorescence

Cells were washed with PBS and fixed in 4% paraformaldehyde in PBS for 15 min at room temperature (RT), permeabilized in 0,1% Triton X-100 in PBS for 15 min at RT, and blocked in 3% bovine serum albumine (BSA) in PBS for 30 min at RT. The slides were then incubated with the primary antibody diluted in 3% BSA in PBS for 1 h at RT. The dilutions used for the antibodies were: anti-Id1, anti-Id2 and anti-Id3 1 : 100; anti-E-cadherin 1 : 50; anti-HA 1 : 100. Then, the slides were incubated for 45 min with the corresponding secondary antibodies diluted in 3% BSA in PBS at RT. The dilutions used for secondary antibodies were: for anti-Id1, anti-Id2, anti-rabbit Alexa 488 conjugated antibodies, 1 : 500; for anti-E-cadherin, anti-rat TRITC-conjugated antibody, 1 : 500; for anti-HA, anti-mouse Alexa 488-conjugated antibody, 1 : 500. Then, the slides were incubated in Hoeschst 33258 (SIGMA) 1 g/ml for 5 min at RT, and mounted in Vectashield medium (Vector Laboratories). Confocal microscopy was performed in Leica SP2 microscope, using an argon laser for Hoeschst and Alexa-488, and a helium–neon laser for TRITC.

Western blotting

Cells were washed with PBS and lysed with RIPA-modified buffer with 1% SDS and phosphatase inhibitors (1 mM sodium orthovanadate, 25 mM -glycerophosphate, 100 mM sodium fluoride) and protease inhibitors (10 g/ml leupeptin, 10 g/ml aprotinin, 1 mM PMSF) for 15 min at 4°C. In all, 50 g of total protein was loaded per lane on SDS–PAGE gels and immunoblot was performed by protein transfer to PVDF membranes (Pall Corporation). Membranes were blocked by incubation with TBS-Tween 0.1–5% nonfat milk and incubated with the antibodies diluted in the same solution. Blots were developed using the ECL detection system (Amersham Pharmacia Biotech). Western quantification was performed using Scion Image software.

RNA preparation and Northern analysis

RNA extraction was performed by guanidinium–phenol–chloroform (Chomczynski and Sacchi, 1987) and 15 g of total RNA was loaded per lane. Purification of poly(A⁺) RNA was carried out as reported elsewhere (Vennstrom and Bishop, 1982). In all, 10 g of poly(A⁺) RNA were loaded per lane. Northern blots were performed on Biodyne A, 0.45 m (Pall Corporation) membranes. All probes were labelled by the random priming method using the kit Ready-to-Go (Amersham Pharmacia Biotech). Hybridizations were carried out overnight at 65°C in 7% SDS, 500 mM sodium phosphate buffer, pH 7.2, and 1 mM EDTA, as described by Church and Gilbert (1984). The following probes were used: complete cDNA for Id1, Id2 and Id3 (donated by Dr D Barettino); PDGF-B. TSP1, VEGF-A and PEDF (donated by Dr Noël Bouck); complete human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA and a fragment of the human cyclophilin cDNA. Quantifications were carried out detecting radioactive signals in the filters using an Instant Imager counter (Packard).

Run-on assay

Cells were collected at 4°C and nuclei were isolated by incubation with a lysis buffer containing 20 mM Tris-HCl, pH 8, 0.3 M sacarose, 60 mM KCl, 15 mM NaCl, 2 mM EDTA, 0.5 mM EGTA, 0.5 mM spermidin, 0.15 mM spermin, 0.5 mM -mercaptoethanol and 0.1% Nonidet P-40. Then, the nuclei were frozen in liquid N₂ and stored at -80°C in a solution containing 50 mM Hepes-NaOH, pH 8, 5 mM MgCl₂, 0.5 mM dithiothreitol, 1 g/ml BSA and 25% glycerin. The cDNAs of the genes to analyse were transferred to a Biodyne A membrane, using a dot-blot transfer system. The in vitro transcription was performed using 2–10 10⁶ nuclei in 50 l and mixed with 50 l of buffer containing the nucleotides and 100 Ci (-³²P)UTP (Amersham). The reaction was run for 20 min at 37°C and after that DNA was digested, and RNA isolated according to the method described previously. Membranes were hybridized with the labelled RNAs for 2–3 days at 42°C in hybridization buffer (200 mM sodium phosphate, pH 7.2, 1 mM EDTA, 7% SDS, 45% formamide, 250 g/ml tRNA).

Promoter analysis

SW480-ADH cells were seeded in 24-well plates at a density of 40 000 cells per well. After 24 h, cells were transfected in triplicate using Jet Pei reagent (Genycell). In all 0.5 g of promoter plasmid and 0.05 g of Renilla reporter under Timidine-kinase promoter (Promega), used as an internal control of eficiency of transfection, were transfected per well. After 16 h the medium was changed and treatments were added, and 48 h later luciferase activity was quantified. The luciferase activity was measured using a Dual-Luciferase Reporter Assay (Promega).

Thymidine incorporation assays

For thymidine incorporation assays 3 10⁴ cells were seeded in 24-well plates and 24 h later, 10^-7 M 1,25(OH)₂D₃ or vehicle was added for the next 24 h. Thymidine (1 Ci/ml) was added for the last 4 h. Cells were extensively washed with PBS prior to fixation with 10% TCA, followed by extraction of incorporated thymidine with 0.2 N NaOH, 1% SDS. Thymidine incorporation was quantified by scintillation counting. Each condition was determined in triplicate and similar results were confirmed in three independent experiments.

CM preparation

Two flasks of 75 cm² were seeded for each cell line. When they reached 80% confluency, they were washed two times with PBS and once with basal medium for 6 h, in order to completely eliminate serum. Then, 15 ml of basal medium supplemented with antibiotics, glutamine, and with 1,25(OH)₂D₃ where indicated, were added, and 48 h later the conditioned medium was collected. Cellular debris were eliminated by centrifugation at 4°C and inhibitors of proteases (PMSF) were added. Then the media were concentrated 100 times using Amicon Ultra devices with 10 kDa cutoff in a refrigerated centrifuge. Protein concentration was measured using Bradford Reagent (Bio-Rad) and checked by silver staining of SDS–PAGE gels. In all, 50 g of protein was loaded per lane for Western blot analysis.

In vivo angiogenesis assay

Cornea vascularization assay was performed as previously described. Sulcralfate/hydron pellets were prepared including the indicated substances and implanted into the avascular cornea of anesthetized C57/B16 mice. Where indicated the pellets contained 0.1 ng/l bFGF and 30 g/ml CM. In all, 0.5 g per pellet of anti-TSP1 or anti-VEGF neutralizing antibodies were included in the pellets where indicated. Results are shown as the number of positive corneas per total number of corneas analysed. Photos were taken 7 days post-implantation.

References

1. Behrens J. (2000) Ann. N.Y. Acad. Sci. 910: 21−33 discussion 33−35. | PubMed | ChemPort |
2. Benezra R, Rafii S & Lyden D. (2001) Oncogene 20: 8334−8341. | Article | PubMed | ChemPort |
3. Bernardi RJ, Johnson CS, Modzelewski RA & Trump DL. (2002) Endocrinology 143: 2508−2514. | Article | PubMed | ChemPort |
4. Chomczynski P & Sacchi N. (1987) Anal. Biochem. 162: 156−159. | Article | PubMed | ISI | ChemPort |
5. Church GM & Gilbert W. (1984) Proc. Natl. Acad. Sci. USA 81: 1991−1995. | PubMed | ChemPort |
6. Dalhoff K, Dancey J, Astrup L, Skovsgaard T, Hamberg KJ, Lofts FJ, Rosmorduc O, Erlinger S, Bach Hansen J, Steward WP, Skov T, Burcharth F & Evans TR. (2003) Br. J. Cancer 89: 252−257. | Article | PubMed | ChemPort |
7. Dameron KM, Volpert OV, Tainsky MA & Bouck N. (1994) Science 265: 1582−1584. | PubMed | ISI | ChemPort |
8. Diaz GD, Paraskeva C, Thomas MG, Binderup L & Hague A. (2000) Cancer Res. 60: 2304−2312. | PubMed | ISI | ChemPort |
9. Evans SR, Shchepotin EI, Young H, Rochon J, Uskokovic M & Shchepotin IB. (2000) Int. J. Oncol. 16: 1249−1254. | PubMed | ChemPort |
10. Ezura Y, Tournay O, Nifuji A & Noda M. (1997) J. Biol. Chem. 272: 29865−29872. | Article | PubMed | ChemPort |
11. Gross C, Stamey T, Hancock S & Feldman D. (1998) J. Urol. 159: 2035−2039 discussion 2039−2040. | Article | PubMed | ChemPort |
12. Gulliford T, English J, Colston KW, Menday P, Moller S & Coombes RC. (1998) Br. J. Cancer 78: 6−13. | PubMed | ISI | ChemPort |
13. Gumireddy K, Ikegaki N, Phillips PC, Sutton LN & Reddy CD. (2003) Biochem. Pharmacol. 65: 1943−1955. | Article | PubMed | ChemPort |
14. Guyton KZ, Kensler TW & Posner GH. (2001) Annu. Rev. Pharmacol. Toxicol. 41: 421−442. | Article | PubMed | ISI | ChemPort |
15. Halline AG, Davidson NO, Skarosi SF, Sitrin MD, Tietze C, Alpers DH & Brasitus TA. (1994) Endocrinology 134: 1710−1717. | Article | PubMed | ChemPort |
16. Hewitt RE, McMarlin A, Kleiner D, Wersto R, Martin P, Tsokos M, Stamp GW, Stetler-Stevenson WG & Tsoskas M. (2000) J. Pathol. 192: 446−454. | Article | PubMed | ChemPort |
17. Iseki K, Tatsuta M, Uehara H, Iishi H, Yano H, Sakai N & Ishiguro S. (1999) Int. J. Cancer 81: 730−733. | Article | PubMed | ChemPort |
18. Israel MA, Hernandez MC, Florio M, Andres-Barquin PJ, Mantani A, Carter JH & Julin CM. (1999) Cancer Res. 59: 1726s−1730s. | PubMed | ChemPort |
19. Jensen SS, Madsen MW, Lukas J, Binderup L & Bartek J. (2001) Mol. Endocrinol. 15: 1370−1380. | Article | PubMed | ChemPort |
20. Koli K & Keski-Oja J. (2000) Cell Growth Differ. 11: 221−229. | PubMed | ChemPort |
21. Lasorella A, Boldrini R, Dominici C, Donfrancesco A, Yokota Y, Inserra A & Iavarone A. (2002) Cancer Res. 62: 301−306. | PubMed | ISI | ChemPort |
22. Lasorella A, Iavarone A & Israel MA. (1996) Mol. Cell. Biol. 16: 2570−2578. | PubMed | ISI | ChemPort |
23. Lasorella A, Noseda M, Beyna M, Yokota Y & Iavarone A. (2000) Nature 407: 592−598. | Article | PubMed | ISI | ChemPort |
24. Lasorella A, Uo T & Iavarone A. (2001) Oncogene 20: 8326−8333. | Article | PubMed | ChemPort |
25. Leibovitz A, Stinson JC, McCombs WB, III, McCoy CE, Mazur KC & Mabry ND. (1976) Cancer Res. 36: 4562−4569. | PubMed | ISI | ChemPort |
26. Liu Y, Encinas M, Comella JX, Aldea M & Gallego C. (2004) Mol. Cell. Biol. 24: 2662−2672. | Article | PubMed | ChemPort |
27. Lyden D, Young AZ, Zagzag D, Yan W, Gerald W, O'Reilly R, Bader BL, Hynes RO, Zhuang Y, Manova K & Benezra R. (1999) Nature 401: 670−677. | Article | PubMed | ISI | ChemPort |
28. Mantell DJ, Owens PE, Bundred NJ, Mawer EB & Canfield AE. (2000) Circ. Res. 87: 214−220. | PubMed | ChemPort |
29. Matsumura ME, Lobe DR & McNamara CA. (2002) J. Biol. Chem. 277: 7293−7297. | Article | PubMed | ChemPort |
30. Nehlin JO, Hara E, Kuo WL, Collins C & Campisi J. (1997) Biochem. Biophys. Res. Commun. 231: 628−634. | Article | PubMed | ChemPort |
31. Nishii Y & Okano T. (2001) Steroids 66: 137−146. | Article | PubMed | ChemPort |
32. Palmer HG, Gonzalez-Sancho JM, Espada J, Berciano MT, Puig I, Baulida J, Quintanilla M, Cano A, de Herreros AG, Lafarga M & Munoz A. (2001) J. Cell Biol. 154: 369−387. | Article | PubMed | ISI | ChemPort |
33. Palmer HG, Sanchez-Carbayo M, Ordonez-Moran P, Larriba MJ, Cordon-Cardo C & Munoz A. (2003) Cancer Res. 63: 7799−7806. | PubMed | ISI | ChemPort |
34. Ramsauer M & D'Amore PA. (2002) J. Clin. Invest. 110: 1615−1617. | Article | PubMed | ChemPort |
35. Rockman SP, Currie SA, Ciavarella M, Vincan E, Dow C, Thomas RJ & Phillips WA. (2001) J. Biol. Chem. 276: 45113−45119. | Article | PubMed | ISI | ChemPort |
36. Ruzinova MB, Schoer RA, Gerald W, Egan JE, Pandolfi PP, Rafii S, Manova K, Mittal V & Benezra R. (2003) Cancer Cell 4: 277−289. | Article | PubMed | ISI | ChemPort |
37. Scaglione-Sewell BA, Bissonnette M, Skarosi S, Abraham C & Brasitus TA. (2000) Endocrinology 141: 3931−3939. | Article | PubMed | ISI | ChemPort |
38. Schlaeppi JM, Gutzwiller S, Finkenzeller G & Fournier B. (1997) Endocr. Res. 23: 213−229. | PubMed | ISI | ChemPort |
39. Sikder HA, Devlin MK, Dunlap S, Ryu B & Alani RM. (2003) Cancer Cell 3: 525−530. | Article | PubMed | ChemPort |
40. Smith DC, Johnson CS, Freeman CC, Muindi J, Wilson JW & Trump DL. (1999) Clin. Cancer Res. 5: 1339−1345. | PubMed | ChemPort |
41. Vennstrom B & Bishop JM. (1982) Cell 28: 135−143. | Article | PubMed | ISI | ChemPort |
42. Verlinden L, Verstuyf A, Convents R, Marcelis S, Van Camp M & Bouillon R. (1998) Mol. Cell. Endocrinol. 142: 57−65. | Article | PubMed | ChemPort |
43. Volpert OV, Pili R, Sikder HA, Nelius T, Zaichuk T, Morris C, Shiflett CB, Devlin MK, Conant K & Alani RM. (2002) Cancer Cell 2: 473−483. | Article | PubMed | ISI | ChemPort |
44. Wilson JW, Deed RW, Inoue T, Balzi M, Becciolini A, Faraoni P, Potten CS & Norton JD. (2001) Cancer Res. 61: 8803−8810. | PubMed | ChemPort |
45. Woo PL, Cercek A, Desprez PY & Firestone GL. (2000) J. Biol. Chem. 275: 28649−28658. | Article | PubMed | ChemPort |
46. Yam CH, Fung TK & Poon RY. (2002) Cell Mol. Life Sci. 59: 1317−1326. | Article | PubMed | ISI | ChemPort |
47. Yokota Y. (2001) Oncogene 20: 8290−8298. | Article | PubMed | ISI | ChemPort |
48. Zebedee Z & Hara E. (2001) Oncogene 20: 8317−8325. | Article | PubMed | ChemPort |

Acknowledgements

We acknowledge with gratitude all researchers who contributed the reagents: J Campisi for Id1 promoter, E Hara for pGEXId1A, JD Norton for pcDNA3-Id2, AG Herreros for E-cadherin promoter constructs, A Alfranca for VEGF promoter and A Bernad for pLZR-iresEGFP, I Palmero for anti-p21 antibody, C Cales for anti-cyclin A antibody. We specially acknowledge C Cales, M Campanero and I Palmero for their advice in cell cycle. This work was supported by grants SAF-2001-1349 and SAF-2004-04152 from Ministerio de Ciencia y Tecnología to B Jimenez, SAF2004-01015 to A Munoz and SAF-2004-07717 to M del Rio. NI Fernandez and M Garcia has been supported by a Comunidad Autonoma de Madrid fellowship.