JunB promotes cell invasion and angiogenesis in VHL-defective renal cell carcinoma

Article metrics

Abstract

Inactivation of the von Hippel–Lindau (VHL) tumor-suppressor gene causes both hereditary and sporadic clear-cell renal-cell carcinoma (ccRCC). Although the best-characterized function of the VHL protein (pVHL) is regulation of hypoxia-inducible factor-α (HIFα), pVHL also controls the development of pheochromocytoma through HIF-independent pathways by regulating JunB. However, it is largely unknown how these pathways contribute to the development and progression of ccRCC. In the present study, we confirmed that JunB was upregulated in VHL-defective ccRCC specimens by immunostaining. Short-hairpin RNA (shRNA)-mediated knockdown of JunB in 786-O and A498 VHL null ccRCC cells suppressed their invasiveness. In addition, JunB knockdown significantly repressed tumor growth and microvessel density in xenograft tumor assays. Conversely, forced expression of wild-type, but not dimerization-defective, JunB in a VHL-restored 786-O subclone promoted invasion in vitro and tumor growth and vessel formation in vivo. Quantitative PCR array analysis revealed that JunB regulated multiple genes relating to tumor invasion and angiogenesis such as matrix metalloproteinase-2 (MMP-2), MMP-9 and chemokine (C-C motif) ligand-2 (CCL2) in 786-O cells. JunB knockdown in these cells reduced the proteolytic activity of both MMPs in gelatin zymography and the amount of CCL2 in the culture supernatant. Moreover, shRNA-mediated knockdown of MMP-2 or inhibition of CCL2 activity with a neutralizing antibody repressed xenograft tumor growth and angiogenesis. Collectively, these results suggest that JunB promotes tumor invasiveness and enhances angiogenesis in VHL-defective ccRCCs.

Introduction

Kidney cancer accounts for 2–3% of adult malignancies, and it causes about 102 000 deaths worldwide annually (Rini et al., 2009). Clear-cell renal-cell carcinoma (ccRCC) is the most common histological subtype of kidney cancer, representing approximately 75–80% of RCCs (Linehan et al., 2007). von Hippel–Lindau (VHL) disease is caused by inactivating germline mutations of the VHL tumor-suppressor gene and is associated with an increased risk of ccRCC and other tumors, including pheochromocytoma, and hemangioblastoma of the retina, cerebellum and spinal cord (Woodward and Maher, 2006). In addition, it has been reported that biallelic inactivation of the VHL gene also occurs in the majority of sporadic ccRCCs (Young et al., 2009). These observations have motivated studies aimed at determining how the VHL protein (pVHL) suppresses tumor growth.

The best-characterized function of pVHL is its ability to regulate the hypoxia-inducible factor (HIF) through poly-ubiquitination and proteasomal destruction (Maxwell et al., 1999). HIF, especially HIF2α, appears to have a particularly important role in ccRCCs based on genotype–phenotype correlations in VHL disease and nude mouse xenograft assays using human VHL-defective ccRCC cell lines (Linehan and Zbar, 2004; Kaelin, 2008, 2009).

HIF activation caused by pVHL inactivation leads to excessive transcription of HIF target genes, including the angiogenic factor vascular endothelial growth factor (VEGF), which is responsible for higher vascularization of VHL-defective tumors. Indeed, molecules that target the HIF/VEGF pathway (mTOR inhibitors, sunitinib and sorafenib) are in clinical use and improve progression-free survival of patients with ccRCC significantly (Brugarolas, 2007; Rini et al., 2009).

Besides regulation of HIF, evidence is accumulating that pVHL also regulates extracellular matrix (ECM) formation, the microtubule cytoskeleton, apoptosis, cell motility and invasion (Frew and Krek, 2008; Roberts and Ohh, 2008; Calzada, 2010). In fact, it has been revealed that HIF-independent VHL functions affect the development and behavior of ccRCC (Yang et al., 2007; Frew and Krek, 2008; Kaelin, 2009). In this respect, we previously reported that pVHL regulated the abundance of JunB through atypical protein kinase-C (aPKC) in an HIF-independent manner, and that inappropriate accumulation of JunB through he inactivation of pVHL was responsible for the development of pheochromocytoma in VHL disease (Lee et al., 2005b; Nakamura and Kaelin, 2006; Woodward and Maher, 2006). However, the roles of JunB in the carcinogenesis of ccRCC remain to be elucidated.

JunB is a member of the activated protein-1 (AP-1) family that forms dimeric protein complexes with family members (jun, fos and ATF) (Eferl and Wagner, 2003; Piechaczyk and Farras, 2008). There is a growing body of evidence that AP-1 proteins have an important role in tumor invasion and angiogenesis (Eferl and Wagner, 2003; Ozanne et al., 2007; Schmidt et al., 2007). Therefore the aim of this study was to clarify the roles of JunB in the development and progression of VHL-defective ccRCC.

Here, we show that JunB is upregulated in sporadic ccRCC specimens with VHL inactivation. Short-hairpin RNA (shRNA)-mediated knockdown of JunB in VHL-defective ccRCC cells inhibited the invasive ability, angiogenesis and tumor growth of xenograft tumors. Present study revealed that JunB regulated tumor invasion and angiogenesis of ccRCC through matrix metalloproteinase-2 (MMP-2), MMP-9 and chemokine (C-C motif) ligand-2 (CCL2). Our results suggest that the JunB pathway could be a promising therapeutic target in these tumors.

Results

The JunB protein is highly expressed in tumor specimens from pVHL-defective ccRCC

To clarify whether JunB protein is upregulated in VHL-defective ccRCC, we performed immunohistochemical analysis of JunB for a total of 75 ccRCC specimens. The degree of immunopositivity in clinical samples was assessed as the percentage of cells positive for JunB. The patients’ characteristics and the status of VHL are shown in Supplementary Table 1. Of these 75 cases, VHL inactivation was observed in 43 cases (42 cases with VHL mutations and one case with promoter methylation). Positive staining was predominantly observed in the nuclei of tumor cells. RCC tumors harboring inactive VHL showed significantly higher JunB staining than those with wild-type (wt) VHL (Figures 1a and b). This result indicates that JunB protein accumulates in VHL-defective RCC cells.

Figure 1
figure1

JunB accumulates in VHL-defective ccRCC specimens. (a) Representative micrographs of immunohistochemical staining of JunB in normal kidney tissue (1), tumors with wt (2) or mutant VHL (3). Negative control of (3) stained with a control antibody (mouse IgG) (4). (b) Relative immunopositivity to JunB in VHL wt tumors versus VHL mutant tumors. P-value was assessed by Mann–Whitney nonparametric U-test. (c) Oncomine database analysis of JunB expression in ccRCC versus other histological types of RCC (non-ccRCC).

As for upregulation of JunB in VHL-defective RCC cells in vitro, we previously reported that pVHL mainly regulates the abundance of JunB mRNA by regulating he kinase activity of aPKC (Lee et al., 2005b). An analysis of gene expression data in the Oncomine website accords with our results by indicating that human primary ccRCC tissues express significantly higher levels of JunB mRNA compared with those with other histological types of RCC (Higgins et al., 2003) (Figure 1c). Collectively, these results strongly indicate that the JunB protein accumulates in ccRCC specimens, especially those with mutant VHL, as a result of an increase in JunB mRNA.

JunB regulates the invasive ability of VHL-defective RCC cells in an HIF-independent manner in vitro

To confirm that expression of JunB is regulated by pVHL, we next examined whether restoration of wt VHL in VHL-defective 786-O and A498 ccRCC cell lines reduced the amount of JunB. As expected, in 786-O cells stably transfected with a plasmid encoding hemagglutinin (HA)-tagged wt pVHL (clone WT8) and in A498 cells stably expressing the same plasmid (clone WTD10), JunB expression, along with that of HIF2α and Glut1, decreased. However, JunB expression did not recover when a constitutive active mutant of HIF2α was introduced to WT8 cells (WT8/HIF2αP531A cells) (Figure 2a). Moreover, shRNA knockdown of HIF2α in 786-O cells did not grossly affect the expression of JunB (Supplementary Figure 1). These results are in keeping with previous results (Lee et al., 2005b) that suggest that JunB expression is regulated by pVHL, mainly independent of HIF.

Figure 2
figure2

JunB regulates the invasive ability of VHL-defective ccRCC cells in an HIF-independent manner. (a) Immunoblot analysis of 786-O and A498 subclones with different VHL status and HIF reactivity. The subclones shown are VHL-defective ccRCC 786-O cells stably expressing (WT8) or not expressing (pRC3) an exogenous, HA-tagged VHL gene; VHL-defective ccRCC A498 cells stably expressing (WTD10) or not expressing (pRCJ17) an exogenous, HA-tagged VHL gene; and WT8 clones also expressing (HIF2α P531A) or not expressing (mock) a constitutively active HIF2α construct. Nuclear or whole-cell extracts were immunoblotted using the indicated antibodies. (b) Analysis of the invasive ability of each subclone by Matrigel Invasion Chambers. The micrographs below indicate that the cells invaded through the Matrigel chamber. Each experiment was performed in triplicate. The error bars represent the s.e. (c) Immunoblot analysis of 786-O and A498 subclones infected with retroviruses encoding scramble or JunB shRNAs (#1 and #2). Nuclear or whole-cell extracts were immunoblotted using the indicated antibodies. (d) Analysis of the invasive ability of 786-O or A498 cells subclones infected with scramble or JunB shRNAs by Matrigel Invasion Chambers. Each experiment was performed in triplicate. The error bars represent the s.e.

To examine the effect of VHL and HIF in vivo, we used scanning electron microscopy to analyze the ultrastructure of xenograft tumors from 786-O subclones. Interestingly, the scanning electron microscopic images demonstrated clear differences in ECM structure. Tumors from pRC3 (empty vector clones) had a loosely assembled ECM with fibrillar meshwork, whereas tumors from WT8 and WT8/HIF2αP531A had a densely assembled ECM with no fibrillar meshwork, indicating that ECM remodeling was regulated by pVHL in an HIF-independent manner, consistent with a previous report (Kurban et al., 2006) (Supplementary Figure 2). These findings prompted us to test the invasive ability of 786-O and A498 subclones expressing or not expressing VHL and HIF, as ECM remodeling is closely associated with the invasive ability of tumor cells. In keeping with previous studies (Koochekpour et al., 1999; Kurban et al., 2006), restoration of wt VHL significantly suppressed the invasive ability of both highly invasive 786-O and poorly invasive A498 cells. However, introducing constitutively active HIF2α to WT8 cells (WT8/HIF2αP531A cells) failed to restore it, indicating that the invasive ability of these cells, like their expression of JunB, was regulated by VHL, but independent of HIF activity (Figure 2b).

We next examined whether pVHL controls invasive ability through JunB. For this purpose, we stably knocked down JunB expression by using two independent shRNAs (#1 and #2), both of which markedly reduced endogenous JunB in 786-O and A498 cells without altering the expression of c-Jun, HIF-2α or Glut1 (Figure 2c). Downregulation of JunB significantly suppressed the invasive ability of both cell lines in vitro (Figure 2d). Collectively, these results suggest that deregulation of JunB, in an HIF-independent manner, contributes to the increased invasiveness of pVHL-defective tumor cells.

shRNA knockdown of JunB in VHL-defective RCC cells suppresses tumor growth, invasion and angiogenesis in vivo

To further clarify the role of the JunB that is upregulated in VHL-defective ccRCC cells, the effect of downregulating this AP-1-family protein was examined by using nude mouse xenograft models. Remarkably, JunB knockdown in both 786-O and A498 cells suppressed the growth of tumors from these cells when they were ectopically (subcutaneously) implanted (Figure 3a). In addition, we also performed orthotopic xenograft assays using highly invasive 786-O cells to examine whether JunB regulates the invasive ability of VHL-defective RCC cells in vivo. Microscopic examination of the tissue revealed that control 786-O/scramble tumor cells invaded into the parenchyma of the normal kidney, whereas 786-O/JunB shRNA cells hardly infiltrated (Figure 3b), indicating that JunB regulates the invasive ability of those cells in vivo.

Figure 3
figure3

shRNA-mediated JunB knockdown suppresses tumor growth, invasion and angiogenesis in VHL-defective ccRCC in vivo. (a) The sequential changes of subcutaneous xenograft tumors from 786-O or A498 subclones infected with scramble or JunB shRNAs. The numbers in parentheses indicate the number of animals developing tumors/total number of animals injected in each group. *P<0.001, P<0.001 versus 786-O/scramble; P=0.007, §P=0.02 versus A498/scramble. (b) Orthotopic implantation of 786-O cells infected with scramble or JunB shRNAs (#1, #2) for evaluation of invasive ability in vivo. Upper panel: Low-magnification view. Lower panel: High-magnification view. Margins of tumors are indicated by a dotted line. T, tumor; N, normal kidney tissue. (c) CD31, VEGF and Ki-67 staining of 786-O or A498 xenograft tumors developed from subclones infected with scramble or JunB shRNAs. The number of vessels and labeling index for Ki-67 in xenograft tumors developed from 786-O or A498 subclones are also shown. The error bars represent the s.e.

As shRNA-mediated knockdown of JunB did not grossly affect in vitro proliferation and cell-cycle progression (Supplementary Figures 3 and 4), we next explored the roles of JunB in regulating tumor growth in vivo. For this purpose, we first examined blood vessel formation by staining subcutaneous xenograft tumors from 786-O and A498 cells for CD31 (Figure 3c). Unexpectedly, tumors from subclones with JunB shRNAs showed markedly lower vessel formation. As no significant differences were observed in the Ki-67 index between tumors infected with scramble or JunB shRNAs (Figure 3c), it was suggested that poor vascularization might inhibit tumor growth. Samples were then stained with an anti-VEGF antibody. However, no apparent difference was observed in the intensity of VEGF staining among them (Figure 3c). Accordingly, 786-O/JunB shRNA cells secreted comparable amounts of VEGF in vitro (Supplementary Figure 5).

As a converse experiment, a 786-O subclone stably expressing VHL (WT8) was infected with a retrovirus encoding HA-tagged wt JunB or dimerization-defective JunB (ΔbZip JunB) (Figure 4a). Forced expression of wt JunB or ΔbZip JunB affected the expression of neither c-Jun nor HIF-2α (Figure 4a), nor cell proliferation in vitro (Supplementary Figure 3). Introduction of wt JunB did, however, increase the in vitro invasive ability of WT8 cells, and promoted tumor growth accompanied by augmented vessel formation in vivo (Figures 4b–d). Again, forcefully expressed JunB neither grossly affected the intensity of VEGF staining, the Ki-67 index nor in vitro VEGF secretion (Figure 4d and Supplementary Figure 5). As WT8 cells transfected with the ΔbZip JunB mutant failed to restore the phenotypes promoted by wt JunB, these results indicate that JunB transcripts control, at least in part, invasiveness and angiogenesis in VHL-defective ccRCCs.

Figure 4
figure4

Forced expression of JunB promotes cell invasion, tumor growth and angiogenesis in VHL-reconstituted 786-O cells (WT8). (a) WT8 cells were infected with a retrovirus encoding HA (mock), HA-tagged wt JunB or HA-tagged dimerization-defective JunB (ΔbZip JunB). Nuclear or whole-cell extracts were immunoblotted using the indicated antibodies. The arrow indicates ΔbZip JunB. (b) Analysis of the invasive ability of WT8 cells infected with mock, HA-tagged wt JunB or HA-tagged ΔbZip JunB by Matrigel Invasion Chambers. Each experiment was performed in triplicate. The error bars represent the s.e. (c) The sequential changes of xenograft tumors from WT8 cells infected with mock, HA-tagged wt JunB or HA-tagged ΔbZip JunB. The numbers in parentheses indicate the number of animals developing tumors/total number of animals injected in each group. *P=0.01 versus WT8/mock. (d) Left: CD31, VEGF and Ki-67 staining of xenograft from WT8 subclones. Right: Number of vessels and labeling index for Ki-67 in xenograft tumors developed from WT8 subclones. The error bars represent the s.e.

shRNA-mediated JunB suppression in VHL-defective RCC cells alters the expression of invasion- and angiogenesis-related genes

To identify genes under the control of JunB, the expression profile of 168 genes involved in angiogenesis and tumor invasion in 786-O/scramble and 786-O/JunB shRNA cells (#1 and #2) was compared by using a quantitative real-time reverse transcription–PCR array. As shown in Table 1, we identified 12 genes whose expression was changed at least 1.5-fold by both shRNAs.

Table 1 mRNA changes after shRNA-mediated knockdown of JunB in 786-O cells

Of genes relating to angiogenesis, CCL2 was downregulated by JunB shRNAs. This result was also confirmed by quantitative PCR using another set of primers (Figure 5a). In fact, the amount of CCL2 protein was significantly decreased in cell culture supernatant from 786-O/JunB shRNA cells (Figure 5b). Immunohistochemical studies also showed that the amount of CCL2 was reduced in xenograft tumors from 786-O/JunB shRNA cells (Figure 5c). As CCL2 has been demonstrated to act as a direct mediator of angiogenesis in in vivo models (Salcedo et al., 2000; Keeley et al., 2008), we next examined the effect of this chemokine in the tumorigenesis and angiogenesis of ccRCC. Inhibiting CCL2 activity with a neutralizing antibody significantly suppressed xenograft tumor growth in 786-O cells. Inactivation of CCL2 also significantly reduced microvessel density (MVD) 2 weeks after inoculation, although the effect was diminished at the time point of 6 weeks (Figure 5d).

Figure 5
figure5

JunB regulates CCL2 in 786-O cells. (a, b) Evaluation of the expression of CCL2 in 786-O/JunB shRNA cells by real-time PCR (a) and ELISA (b). (c) Evaluation of expression of CCL2 in xenograft tumors from 786-O/JunB shRNA cells by immunohistochemistry. (d) Effect of CCL2-neutralizing antibody on tumor growth (left) and MVD (right) in xenograft tumors from 786-O cells. The numbers in parentheses indicate the number of animals developing tumors/total number of animals injected in each group.

As for genes involved in tumor invasion, it was noted that expression of both MMP-2 and MMP-9 was reduced by JunB shRNAs (Table 1). Results of the PCR array analysis were confirmed by quantitative PCR (Figure 6a), and gelatin zymography revealed that shRNA-mediated knockdown of JunB reduced the proteolytic activity of both MMPs in vitro (Figure 6b). Moreover, in situ gelatin zymography applied to xenograft tumors showed that gelatinolytic activity was attenuated in tumors from 786-O/JunB shRNA cells (Supplementary Figure 6). To confirm the effects of endogenous or exogenous JunB on transcriptional activation of MMP-2, we next examined MMP-2 promoter activity using a luciferase reporter assay. As expected, JunB knockdown suppressed MMP-2 promoter activity in 786-O cells, and introduction of wt JunB, but not Δbzip JunB, promoted MMP-2 promoter activity in WT8 cells (Figures 6c and d). In addition, a site-specific mutagenesis of the AP-1-binding site in the MMP-2 promoter significantly attenuated its transcriptional activity in 786-O cells (Figure 6e), indicating that transcriptional activities were regulated, at least in part, by the AP-1-family protein.

Figure 6
figure6

JunB regulates MMP-2 and MMP-9 in 786-O cells. (a, b) Evaluation of the expression and the activity of MMP-2 and MMP-9 in 786-O/JunB shRNA cells by real-time PCR (a) and gelatin zymography (b). IDV: Relative integrated density value. (c, d) MMP-2 promoter activities were measured by using luciferase reporter assays in 786-O subclones using JunB shRNA (c) and in WT8 cells infected using wt JunB or Δbzip JunB (d). (e) A site-specific mutation was introduced in the AP-1-biding site of the MMP-2 promoter and MMP-2 promoter activities were measured in786-O cells.

In fact, shRNA-mediated knockdown of MMP-2 or MMP9 significantly inhibited in vitro cell invasion of 786-O cells by reducing gelatinolytic activity, which phenocopies the effect of JunB knockdown (Figures 7a and b). Similarly, a specific gelatinase (MMP-2 and MMP-9) inhibitor SB-3CT suppressed invasion of those cells (Supplementary Figure 7a). Notably, JunB-induced cell invasion was blocked by either knockdown of MMP-2 or treatment with SB-3CT in WT8 cells (Figure 7c and Supplementary Figure 7b). Previously, it was reported that pVHL regulated the expression of MMP-2 and MMP-9 (Koochekpour et al., 1999; Struckmann et al., 2008). In addition, Kurban et al. (2006) clearly showed that MMP-2 was regulated by pVHL in an HIF-independent manner in 786-O cells. These results, combined with our own, suggest that pVHL regulates the expression of MMP-2/9 and concomitant tumor invasion mostly through JunB in these cells. Importantly, shRNA knockdown of MMP-2 suppressed the tumor growth in xenograft model with decreased MVD (Figures 8a and b). No remarkable change was observed in either the intensity of VEGF staining or the Ki-67 index (Figure 8b). Collectively, these results highly suggest that MMP-2/9, particularly MMP-2, have significant roles on JunB-mediated tumor invasion and progression of pVHL-defective RCCs.

Figure 7
figure7

JunB promotes cell invasion of 786-O cells through MMP-2/9. (a) Gelatin zymography of 786-O subclones infected with lentiviruses encoding scramble or MMP-2 shRNAs or MMP-9 shRNAs. (b) Analysis of the invasive ability of 786-O subclones infected with scramble, MMP-2 shRNAs or MMP-9 shRNA by Matrigel Invasion Chambers. Each experiment was performed in triplicate. The error bars represent the s.e. (c) Left: MMP-2 expression measured by real-time PCR in WT8 cells overexpressing JunB infected with lentiviruses encoding scramble or MMP-2 shRNAs. Right: Invasive ability of WT8 overexpressing JunB infected with lentiviruses encoding scramble or MMP-2 shRNAs by Matrigel Invasion Chambers. Each experiment was performed in triplicate. The error bars represent the s.e.

Figure 8
figure8

shRNA-mediated MMP-2 knockdown suppresses tumor growth and angiogenesis in 786-O cells in vivo. (a) The sequential changes of xenograft tumors from 786-O subclones infected with scramble or MMP-2 shRNAs. The numbers in parentheses indicate the number of animals developing tumors/total number of animals injected in each group. *P=0.003; P<0.001 versus 786-O/scramble (b) Left: CD31, VEGF and Ki-67 staining of 786-O xenograft tumors from subclones infected with scramble or MMP-2 shRNAs. Right: Number of vessels and labeling index for Ki-67 in xenograft tumors developed from 786-O cells infected with MMP-2 shRNAs. The error bars represent the s.e.

Discussion

It is well known that JunB has an anti-oncogenic function that it exerts by antagonizing c-Jun-dependent proliferation and transformation (Piechaczyk and Farras, 2008; Shaulian, 2010). On the other hand, it was reported that homodimers of JunB positively regulated cell proliferation, whereas c-Jun:JunB heterodimers negatively regulated it in a mouse model, using a knock-in strategy and a transgenic complementation approach (Passegue et al., 2002). Indeed, it has been reported that JunB has oncogenic roles in particular cancer cells such as Hodgkin's lymphomas and anaplastic large cell lymphomas (Watanabe et al., 2005; Staber et al., 2007; Shaulian, 2010). The results in the present study indicate that JunB upregulated in VHL-defective ccRCCs augments the angiogenesis and invasion of tumors in vivo. Consistent with our result, there is a growing body of evidence that AP-1 subunits have an important role in tumor angiogenesis and invasion (Eferl and Wagner, 2003; Ozanne et al., 2007; Schmidt et al., 2007).

In terms of angiogenesis in VHL-defective ccRCCs, we clarified that JunB knockdown suppressed CCL2 expression in 786-O cells. CCL2 has a critical role in the recruitment of macrophages and the induction of angiogenesis (Balkwill, 2004; Keeley et al., 2008). As such, it was revealed that this chemokine regulated the angiogenesis of xenografts from pVHL-defective RCCs, although the effect of treatment with a neutralizing antibody was limited to the early phase of tumor formation. The pro-angiogenic effects of CCL2 should be further examined in VHL-defective ccRCCs by using shRNA-mediated knockdown of this chemokine in a future study. Importantly, it is already known that pVHL controls CCL2 transcripts in the livers of mice through inhibitory phosphorylation of CARD9, the nuclear factor-κB agonist (Yang et al., 2007). Moreover, it has been reported that transcriptional activation of the CCL2 promoter is regulated by both nuclear factor-κB and AP-1 in human endothelial cells (Martin et al., 1997). These results, combined with our own, indicate that pVHL might regulate the expression of CCL2 through JunB and the nuclear factor-κB pathway.

As for molecules relating to tumor invasion, it was revealed that JunB regulated MMP-2 and MMP-9 in highly invasive 786-O cells. In fact, shRNA-mediated knockdown of these MMPs significantly suppressed in vitro cell invasion in 786-O cells. Notably, treatment with the specific gelatinase (MMP-2 and MMP-9) inhibitor SB-3CT, or shRNA knockdown of MMP-2, overrides the invasion conferred by forcefully expressed JunB in WT8 cells, indicating that JunB upregulates the invasion ability in these cells mainly through MMP-2/9, particularly MMP-2.

Importantly, shRNA knockdown of MMP-2 suppressed tumor growth in xenograft model with decreased MVD. This result might be explained by reference to the literature: it has been reported that MMP-2 and MMP-9 promote angiogenesis by releasing VEGF stored in the ECM, or by directly activating the bioavailability of VEGF (Lee et al., 2005a; Roy et al., 2009). Additionally, it has been shown that MMPs are able to degrade ECM components. In fact, pVHL-deficient cells fail to form ECM properly, which may facilitate neoangiogenesis and invasion of these cells, mostly through HIF-independent pathways (Ohh et al., 1998; Frew and Krek, 2008; Kurban et al., 2008; Calzada, 2010).

In terms of the possible regulation of ECM assembly by JunB, scanning electron microscopic images demonstrated that tumors from 786-O/JunB shRNAs had a densely assembled ECM with no fibrillar meshwork, which is similar to that in WT8 (pVHL-positive) cells. By contrast, tumors from 786-O/scramble cells had a loosely assembled ECM with fibrillar meshwork, which is similar to that in pRC3 (pVHL-negative) cells (Supplementary Figure 2). Recently, it was reported that the promoter region of COL14A1 was frequently methylated in RCC specimens (Morris et al., 2010). Interestingly, the expression profile analysis in the present study revealed that JunB represses the expression of COL14A1. Therefore, we suspect that this gene is down-regulated in RCC specimens. In fact, this suspicion is confirmed by data from the Oncomine website (Supplementary Figure 8). Similarly, expression of COL1A1, a target of upregulation by JunB, is overexpressed in RCC specimens (Supplementary Figure 8). These results indicate that JunB regulates several ECM components, linking the effects of this AP-1 protein to ECM remodeling and this link deserves examination to further our understanding of the roles of JunB in VHL-defective ccRCCs.

As for the potential clinical relevance of JunB target genes, it was reported that RCC patients with higher amounts of MMP-2 or MMP-9 in tumor tissues showed a trend of poor prognosis (Kallakury et al., 2001; Takahashi et al., 2002). It was also reported that the baseline serum level of MMP-9 was significantly higher among patients who did not respond to sunitinib treatment for metastatic ccRCC (Perez-Gracia et al., 2009). It is true that anti-angiogenic therapies, such as those targeting VEGF, demonstrate anti-tumor effects for patients with ccRCC (Brugarolas, 2007; Rini and Atkins, 2009). However, tumors receiving this therapy frequently acquire resistance or show initial resistance. Although the mechanisms of initial and acquired resistance remain to be elucidated, it is speculated that redundant angiogenic factors other than VEGF are related to the development of refractory disease (Rini and Atkins, 2009). Indeed, it is thought that VEGF-mediated signaling is the predominant stimulator of angiogenesis in VHL-defective ccRCC, but parallel angiogenesis-related pathways such as CCL2 can also drive tumor growth. In this sense, there is a possibility that expression of JunB-regulated genes affects the development of the above refractory disease.

Based on these results, we propose that the JunB pathway may be a promising therapeutic target for ccRCCs. Likewise, aPKC could be a useful target, as JunB is regulated through aPKC, although further investigation of JunB regulation by pVHL is required (Frew and Krek, 2008; Kaelin, 2008). Indeed, treatment of 786-O and A498 cells with the PKC inhibitor GF109203X efficiently reduced the invasive ability of those cells as well as markedly reducing JunB protein (Supplementary Figure 9).

In conclusion, we propose that JunB upregulation caused by inactivation of pVHL leads to tumor progression in VHL-defective ccRCCs. The aPKC/JunB pathway could be a target for suppressing the invasion and angiogenesis of these tumors.

Materials and methods

Patients and RCC samples

Tumor specimens from 75 patients who received radical or partial nephrectomy were obtained from a tissue bank in the Department of Urology at Kyoto University Hospital under the protocols approved by the University's institutional review board.

VHL genotyping and hypermethylation assays

Genotyping of VHL and determination of the methylation status of the VHL CpG island were performed as described previously (Nakamura et al., 2006).

Immunohistochemistry

After deparaffinization and antigen retrieval, endogenous peroxidase activity was blocked by 0.3% H2O2 in methyl alcohol. The glass slides were washed in phosphate-buffered saline and mounted with 1% horse normal serum in phosphate-buffered saline. Subsequently primary antibody was applied overnight at 4 °C. They were incubated with biotinylated horse anti-mouse serumin phosphate-buffered saline, followed by washes in phosphate-buffered saline. Avidin–biotin–peroxidase complex (ABC; ABC-Elite; Vector Laboratories, Burlingame, CA, USA) was applied together with the nuclear counterstaining with hematoxylin. JunB immunohistochemical analysis of 75 clinical ccRCC samples was performed on paraffin-embedded, formalin-fixed tissues. The degree of immunopositivity in clinical samples was assessed as the percentage of cells positive for JunB, and the assessment was performed by a pathologist (YM) with no prior knowledge of the VHL status of the samples. VEGF, Ki-67 and CCL2 staining of xenograft tumors was performed on paraffin sections and CD31 staining on frozen sections. Vessel counts were performed as described previously (Weidner, 1995). Numbers of total or Ki-67-positive nuclei were automatically quantitated by using the ImageJ software. Labeling index for Ki-67 was calculated as the percentage of positive tumor nuclei divided by the total number of tumor cells examined.

Antibodies

Antibodies were purchased commercially: JunB (for immunohistochemistry, Abcam, Cambridge, MA, USA; for western blotting, Santa Cruz Biotechnology, Santa Cruz, CA, USA), c-Jun (Cell Signaling, Beverly, MA, USA), HIF2α (Novus Biologicals, Littleton, CO, USA), Glut1 (Alpha Diagnostic Inc., San Antonio, TX, USA), HA (Covance, Berkeley, CA, USA), β-actin (Abcam), histone H2A (Cell Signaling), mouse CD31 (Dako, Carpinteria, CA, USA), VEGF (Santa Cruz Biotechnology.), Ki-67 (Novocastra, Newcastle Upon Tyne, UK) and CCL2 (R&D Systems, Minneapolis, MN, USA).

Cell culture

VHL-deficient 786-O and A498 RCC cells were purchased from the American Type Culture Collection (Rockville, MD, USA). 786-O and A498 subclones stably transfected with pRc-CMV (clone pRC3 and pRCJ17, respectively) or pRc-CMV-HA-VHL (clone WT8 and WTD10, respectively), and WT8 subclones further stably transfected with either pIRES-puro-HA (WT8/mock) or pIRES-puro-HA-HIF2α P531A (WT8/HIF2α P531A), were described previously (Kondo et al., 2002, 2003; Nakamura et al., 2006). 786-O and A498 cells stably infected with shRNA retroviruses were selected in the presence of 1 mg/ml G418. WT8 cells stably infected with pBabe-puro retroviruses and 786-O cells stably infected with shRNA lentiviruses were selected in the presence of 1.5 μg/ml puromycin.

Immunoblot analysis

JunB and c-Jun were detected in nuclear extracts, and HIF2α and Glut-1 were detected in whole-cell extracts. Immunoblotting was performed as described previously (Lee et al., 2005b).

Invasion assay

In vitro tumor cell invasion was measured using BD Biocoat Matrigel Invasion Chambers (Becton Dickinson, Franklin Lakes, NJ, USA). Cells (786-O and A498 subclones, 4 × 104) in serum-free Dulbecco's modified Eagle's medium (500 μl) were plated in the upper chamber and incubated with Dulbecco's modified Eagle's medium containing 10% fetal calf serum in the bottom of the chamber at 37 °C in 5% CO2, 786-O for 24 h and A498 for 48 h.

Retroviruses

The oligonucleotide sequences used in the construction of the shRNA vector were as follows: JunB#1 (5′-IndexTermGATCCAATGGAACAGCCCTTCTACCATAGTGCTCCTGGTTGTGGTAGAAGGGCTGTTCCATTTTTTTTAT-3′) and (5′-IndexTermCGATAAAAAAAATGGAACAGCCCTTCTACCACAACCAGGAGCACTATGGTAGAAGGGCTGTTCCATTG-3′) (Watanabe et al., 2005); JunB#2 (5′-IndexTermGATCCGGTGAAGACACTCAAGGCTTAGTGCTCCTGGTTGAGCCTTGAGTGTCTTCACCTTTTTTAT-3′) and (5′-IndexTermCGATAAAAAAGGTGAAGACACTCAAGGCTCAACCAGGAGCACTAAGCCTTGAGTGTCTTCACCG-3′) (Staber et al., 2007); and control scramble (5′-IndexTermGATCCGTACAGCGGTCCAATCATAGTAGTGCTCCTGGTTGCTATGATTGGACCGCTGTACTTTTTTAT-3′) and (5′-IndexTermCGATAAAAAGTACAGCGGTCCAATCATAGCAACCAGGAGCACTACTATGATTGGACCGCTGTACG-3′). The oligonucleotides were annealed and then ligated into the BamHI/ClaI sites of the pSINsi-hU6 vector (Takara Bio, Shiga, Japan).

Previously, we made pDEST 47-JunB and pDEST-ΔbZip JunB (dimerization-defective JunB) (Lee et al., 2005b). To make pBABE-puro-HA-JunB and pBABE-puro-HA-ΔbZip JunB, the former two were PCR-amplified using the primers 5′-IndexTermCGGGATCCATGTGCACTAAAATGGAACAGCCCT-3′ and 5′-IndexTermCGGAATTCTGGGACAACTCCAGTGAAA-3′, digested using BamHI and EcoRI, and ligated into pBABE-puro-HA cut with these two enzymes.

G3T-hi packaging cells were infected with retroviral plasmids by using a Retrovirus Packaging Kit (Ampho; Takara Bio), according to the manufacturer's instructions.

Lentiviruses

Lentivirus-based plasmids containing shRNAs to human MMP-2 and MMP-9 were purchased from Open Biosystems (Huntsville, AL, USA). Non-silencing shRNAs (Open Biosystems) were used as negative control. The experimental procedures for shRNA transfection were according to the Open Biosystems technical manual.

Nude mouse xenograft assays

All experiments involving laboratory animals were performed in accordance with the Guideline for Animal Experiments of Kyoto University. For subcutaneous xenograft assays, 107 cells were injected subcutaneously into both flanks of 6-week-old female BALB/cAJcl nude (nu/nu) mice (CLEA Inc., Tokyo, Japan). Orthotopic xenograft was performed as described previously (Yang et al., 2007). Presence of tumor mass was finally confirmed by hematoxylin and eosin staining.

VEGF and CCL2 ELISA

The VEGF and CCL2 proteins in the supernatant were quantified using VEGF and CCL2 ELISA kits (R&D Systems). Concentration was normalized to total cell protein.

Real-time PCR and PCR arrays

cDNA synthesis and real-time PCR were performed as described previously (Terada et al., 2010). The primer sequences were as follows: CCL2, 5′-IndexTermCCCCAGTCACCTGCTGTTAT-3′ (sense) and 5′-IndexTermAGATCTCCTTGGCCACAATG-3′ (antisense); MMP-2, 5′-IndexTermATAACCTGGATGCCGTCGT-3′ (sense) and 5′-IndexTermAGGCACCCTTGAAGAAGTAGC-3′ (antisense); MMP-9, 5′-IndexTermGAACCAATCTCACCGACAGG-3′ (sense) and 5′-IndexTermGCCACCCGAGTGTAACCATA-3′ (antisense); and GAPDH, 5′-IndexTermGAAGGTGAAGGTCGGAGTC-3′ (sense) and 5′-IndexTermGAAGATGGTGATGGGATTTC-3′ (antisense).

RT2 Profiler human angiogenesis and Extracellular Matrix and Adhesion Molecules PCR Arrays (SA Biosciences, Frederick, MD, USA), containing 84 × 2 angiogenesis- and invasion-related genes, plus housekeeping genes and controls, were used according to the manufacturer's protocols.

Blocking CCL2 with neutralizing antibody

Neutralizing antibodies against human CCL2 (MAB679; R&D Systems) or control mouse IgG were administered to mice with subcutaneous tumors from 786-O cells, by injecting intraperitoneally every 4 days from the date of tumor cell inoculation to the end of the mouse experiment at a dose of 10 μg per mouse (Lu and Kang, 2009).

Gelatin zymography

Gels for gelatin zymography were purchased from Invitrogen (Carlsbad, CA, USA) and used according to the manufacturer's instructions. Cells were cultured in Opti-MEM for 24 h, and the medium was collected and concentrated using Amicon Ultra-4 (Millipore, Billerica, MA, USA). The samples were normalized for intracellular protein concentration.

Luciferase reporter assays

Full-length human MMP-2 promoter-luciferase constructs (pGL3-MMP2p-Luc) were made as described previously (Bian and Sun, 1997). A site-specific mutation (1270-TGACTTCT-1263→TGGATTCT) was introduced into the AP-1-binding site in the MMP-2 promoter as described previously (Song et al., 2006). Cells grown in 12-well tissue culture plates were transfected with 0.4 g of pGL3-MMP2p-Luc and 4 ng of pTK-RL using Lipofectamine 2000 (Invitrogen). After 24 h of incubation, the luciferase activity of the cell lysate was measured in triplicate by using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA), using a luminometer (MicroLumat Plus LB96V; Berthold Technologies, LLC, TN, USA).

Statistical analysis

Data were expressed as mean±s.e., except real-time PCR (mean±s.d.), and Student's t-test was used to analyze the differences between means. Mann–Whitney non-parametric U-test was performed to compare differences in JunB immunopositivity between VHL mutant groups and wt VHL groups. Inhibition of tumor growth with a CCL2 neutralizing antibody was analyzed by two-way repeated-measures analysis of variance. Statistical analyses were all performed using StatView ver. 5.0 (SAS Institute, Cary, NC, USA).

References

  1. Balkwill F . (2004). Cancer and the chemokine network. Nat Rev Cancer 4: 540–550.

  2. Bian J, Sun Y . (1997). Transcriptional activation by p53 of the human type IV collagenase (gelatinase A or matrix metalloproteinase 2) promoter. Mol Cell Biol 17: 6330–6338.

  3. Brugarolas J . (2007). Renal-cell carcinoma—molecular pathways and therapies. N Engl J Med 356: 185–187.

  4. Calzada MJ . (2010). von Hippel–Lindau syndrome: molecular mechanisms of the disease. Clin Transl Oncol 12: 160–165.

  5. Eferl R, Wagner EF . (2003). AP-1: a double-edged sword in tumorigenesis. Nat Rev Cancer 3: 859–868.

  6. Frew IJ, Krek W . (2008). pVHL: a multipurpose adaptor protein. Sci Signal 1: pe30.

  7. Higgins JP, Shinghal R, Gill H, Reese JH, Terris M, Cohen RJ et al. (2003). Gene expression patterns in renal cell carcinoma assessed by complementary DNA microarray. Am J Pathol 162: 925–932.

  8. Kaelin Jr WG . (2008). The von Hippel–Lindau tumour suppressor protein: O2 sensing and cancer. Nat Rev Cancer 8: 865–873.

  9. Kaelin Jr WG . (2009). Treatment of kidney cancer: insights provided by the VHL tumor-suppressor protein. Cancer 115: 2262–2272.

  10. Kallakury BV, Karikehalli S, Haholu A, Sheehan CE, Azumi N, Ross JS . (2001). Increased expression of matrix metalloproteinases 2 and 9 and tissue inhibitors of metalloproteinases 1 and 2 correlate with poor prognostic variables in renal cell carcinoma. Clin Cancer Res 7: 3113–3119.

  11. Keeley EC, Mehrad B, Strieter RM . (2008). Chemokines as mediators of neovascularization. Arterioscler Thromb Vasc Biol 28: 1928–1936.

  12. Kondo K, Kim WY, Lechpammer M, Kaelin Jr WG . (2003). Inhibition of HIF2alpha is sufficient to suppress pVHL-defective tumor growth. PLoS Biol 1: E83.

  13. Kondo K, Klco J, Nakamura E, Lechpammer M, Kaelin Jr WG . (2002). Inhibition of HIF is necessary for tumor suppression by the von Hippel–Lindau protein. Cancer Cell 1: 237–246.

  14. Koochekpour S, Jeffers M, Wang PH, Gong C, Taylor GA, Roessler LM et al. (1999). The von Hippel–Lindau tumor suppressor gene inhibits hepatocyte growth factor/scatter factor-induced invasion and branching morphogenesis in renal carcinoma cells. Mol Cell Biol 19: 5902–5912.

  15. Kurban G, Duplan E, Ramlal N, Hudon V, Sado Y, Ninomiya Y et al. (2008). Collagen matrix assembly is driven by the interaction of von Hippel–Lindau tumor suppressor protein with hydroxylated collagen IV alpha 2. Oncogene 27: 1004–1012.

  16. Kurban G, Hudon V, Duplan E, Ohh M, Pause A . (2006). Characterization of a von Hippel Lindau pathway involved in extracellular matrix remodeling, cell invasion, and angiogenesis. Cancer Res 66: 1313–1319.

  17. Lee S, Jilani SM, Nikolova GV, Carpizo D, Iruela-Arispe ML . (2005a). Processing of VEGF-A by matrix metalloproteinases regulates bioavailability and vascular patterning in tumors. J Cell Biol 169: 681–691.

  18. Lee S, Nakamura E, Yang H, Wei W, Linggi MS, Sajan MP et al. (2005b). Neuronal apoptosis linked to EglN3 prolyl hydroxylase and familial pheochromocytoma genes: developmental culling and cancer. Cancer Cell 8: 155–167.

  19. Linehan WM, Pinto PA, Srinivasan R, Merino M, Choyke P, Choyke L et al. (2007). Identification of the genes for kidney cancer: opportunity for disease-specific targeted therapeutics. Clin Cancer Res 13: 671s–679s.

  20. Linehan WM, Zbar B . (2004). Focus on kidney cancer. Cancer Cell 6: 223–228.

  21. Lu X, Kang Y . (2009). Chemokine (C-C motif) ligand 2 engages CCR2+ stromal cells of monocytic origin to promote breast cancer metastasis to lung and bone. J Biol Chem 284: 29087–29096.

  22. Martin T, Cardarelli PM, Parry GC, Felts KA, Cobb RR . (1997). Cytokine induction of monocyte chemoattractant protein-1 gene expression in human endothelial cells depends on the cooperative action of NF-kappa B and AP-1. Eur J Immunol 27: 1091–1097.

  23. Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, Cockman ME et al. (1999). The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399: 271–275.

  24. Morris MR, Ricketts C, Gentle D, Abdulrahman M, Clarke N, Brown M et al. (2010). Identification of candidate tumour suppressor genes frequently methylated in renal cell carcinoma. Oncogene 29: 2104–2117.

  25. Nakamura E, Abreu-e-Lima P, Awakura Y, Inoue T, Kamoto T, Ogawa O et al. (2006). Clusterin is a secreted marker for a hypoxia-inducible factor-independent function of the von Hippel–Lindau tumor suppressor protein. Am J Pathol 168: 574–584.

  26. Nakamura E, Kaelin Jr WG . (2006). Recent insights into the molecular pathogenesis of pheochromocytoma and paraganglioma. Endocr Pathol 17: 97–106.

  27. Ohh M, Yauch RL, Lonergan KM, Whaley JM, Stemmer-Rachamimov AO, Louis DN et al. (1998). The von Hippel–Lindau tumor suppressor protein is required for proper assembly of an extracellular fibronectin matrix. Mol Cell 1: 959–968.

  28. Ozanne BW, Spence HJ, McGarry LC, Hennigan RF . (2007). Transcription factors control invasion: AP-1 the first among equals. Oncogene 26: 1–10.

  29. Passegue E, Jochum W, Behrens A, Ricci R, Wagner EF . (2002). JunB can substitute for Jun in mouse development and cell proliferation. Nat Genet 30: 158–166.

  30. Perez-Gracia JL, Prior C, Guillen-Grima F, Segura V, Gonzalez A, Panizo A et al. (2009). Identification of TNF-alpha and MMP-9 as potential baseline predictive serum markers of sunitinib activity in patients with renal cell carcinoma using a human cytokine array. Br J Cancer 101: 1876–1883.

  31. Piechaczyk M, Farras R . (2008). Regulation and function of JunB in cell proliferation. Biochem Soc Trans 36: 864–867.

  32. Rini BI, Atkins MB . (2009). Resistance to targeted therapy in renal-cell carcinoma. Lancet Oncol 10: 992–1000.

  33. Rini BI, Campbell SC, Escudier B . (2009). Renal cell carcinoma. Lancet 373: 1119–1132.

  34. Roberts AM, Ohh M . (2008). Beyond the hypoxia-inducible factor-centric tumour suppressor model of von Hippel–Lindau. Curr Opin Oncol 20: 83–89.

  35. Roy R, Yang J, Moses MA . (2009). Matrix metalloproteinases as novel biomarkers and potential therapeutic targets in human cancer. J Clin Oncol 27: 5287–5297.

  36. Salcedo R, Ponce ML, Young HA, Wasserman K, Ward JM, Kleinman HK et al. (2000). Human endothelial cells express CCR2 and respond to MCP-1: direct role of MCP-1 in angiogenesis and tumor progression. Blood 96: 34–40.

  37. Schmidt D, Textor B, Pein OT, Licht AH, Andrecht S, Sator-Schmitt M et al. (2007). Critical role for NF-kappaB-induced JunB in VEGF regulation and tumor angiogenesis. EMBO J 26: 710–719.

  38. Shaulian E . (2010). AP-1—the Jun proteins: oncogenes or tumor suppressors in disguise? Cell Signal 22: 894–899.

  39. Song H, Ki SH, Kim SG, Moon A . (2006). Activating transcription factor 2 mediates matrix metalloproteinase-2 transcriptional activation induced by p38 in breast epithelial cells. Cancer Res 66: 10487–10496.

  40. Staber PB, Vesely P, Haq N, Ott RG, Funato K, Bambach I et al. (2007). The oncoprotein NPM-ALK of anaplastic large-cell lymphoma induces JUNB transcription via ERK1/2 and JunB translation via mTOR signaling. Blood 110: 3374–3383.

  41. Struckmann K, Mertz K, Steu S, Storz M, Staller P, Krek W et al. (2008). pVHL coordinately regulates CXCR4/CXCL12 and MMP2/MMP9 expression in human clear-cell renal cell carcinoma. J Pathol 214: 464–471.

  42. Takahashi M, Oka N, Naroda T, Nishitani MA, Kanda K, Kanayama HO et al. (2002). Prognostic significance of matrix metalloproteinases-2 activation ratio in renal cell carcinoma. Int J Urol 9: 531–538.

  43. Terada N, Shimizu Y, Kamba T, Inoue T, Maeno A, Kobayashi T et al. (2010). Identification of EP4 as a potential target for the treatment of castration-resistant prostate cancer using a novel xenograft model. Cancer Res 70: 1606–1615.

  44. Watanabe M, Sasaki M, Itoh K, Higashihara M, Umezawa K, Kadin ME et al. (2005). JunB induced by constitutive CD30-extracellular signal-regulated kinase 1/2 mitogen-activated protein kinase signaling activates the CD30 promoter in anaplastic large cell lymphoma and Reed–Sternberg cells of Hodgkin lymphoma. Cancer Res 65: 7628–7634.

  45. Weidner N . (1995). Intratumor microvessel density as a prognostic factor in cancer. Am J Pathol 147: 9–19.

  46. Woodward ER, Maher ER . (2006). von Hippel–Lindau disease and endocrine tumour susceptibility. Endocr Relat Cancer 13: 415–425.

  47. Yang H, Minamishima YA, Yan Q, Schlisio S, Ebert BL, Zhang X et al. (2007). pVHL acts as an adaptor to promote the inhibitory phosphorylation of the NF-kappaB agonist Card9 by CK2. Mol Cell 28: 15–27.

  48. Young AC, Craven RA, Cohen D, Taylor C, Booth C, Harnden P et al. (2009). Analysis of VHL gene alterations and their relationship to clinical parameters in sporadic conventional renal cell carcinoma. Clin Cancer Res 15: 7582–7592.

Download references

Acknowledgements

We thank Dr William G Kaelin, Jr, for critical comments and for providing us with the different RCC cell lines. We appreciate all members of Cancer Research Courses for Integrated Research Training at Kyoto University Graduate School of Medicine for helpful advice and discussion. We also thank Tomoko Matsushita, Megumi Kuraguchi and Yuko Fujieda for technical assistance. Grant support: Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Author information

Correspondence to E Nakamura.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • renal-cell carcinoma
  • VHL gene
  • JunB
  • invasion
  • angiogenesis

Further reading