von Hippel–Lindau (VHL) disease is a dominantly inherited family cancer syndrome characterized by the development of retinal and central nervous system haemangioblastomas, renal cell carcinoma (RCC) and phaeochromocytoma. Specific germline VHL mutations may predispose to haemangioblastomas, RCC and phaeochromocytoma to a varying extent. Although dysregulation of the hypoxia-inducible transcription factor-2 and JunB have been linked to the development of RCC and phaeochromocytoma, respectively, the precise basis for genotype–phenotype correlations in VHL disease have not been defined. To gain insights into the pathogenesis of RCC in VHL disease we compared gene expression microarray profiles in a RCC cell line expressing a Type 1 or Type 2B mutant pVHL (RCC-associated) to those of a Type 2A or 2C mutant (not associated with RCC). We identified 19 differentially expressed novel VHL target genes linked to RCC development. Eight targets were studied in detail by quantitative real-time polymerase chain reaction (three downregulated and five upregulated by wild-type VHL) and for six genes the effect of VHL inactivation was mimicked by hypoxia (but hypoxic-induction of smooth muscle alpha-actin 2 was specific for a RCC cell line). The potential role of four RCC-associated VHL target genes was assessed in vitro. NB thymosin beta (TMSNB) and proteinase-activated receptor 2 (PAR2) (both downregulated by wt pVHL) increased cell growth and motility in a RCC cell line, but aldehyde dehydrogenase (ALDH)1 and ALDH7 had no effect. These findings implicate TMSNB and PAR2 candidate oncogenes in the pathogenesis of VHL-associated RCC.
Renal cell carcinomas (RCC) is the most common adult kidney tumour, accounting for approximately 2% of all adult malignancies. Although only 2% of RCC tumours are familial, the identification of the genetic basis for inherited RCC has provided important insights into the pathogenesis of sporadic RCC. Thus whereas germline mutations in the VHL tumour suppressor gene cause von Hippel–Lindau (VHL) disease, somatic VHL inactivation occurs in most sporadic clear cell RCC (which accounts for ∼80% of all sporadic RCC). VHL disease (OMIM 193300) affects ∼one in 36 000 individuals and is characterized by the development of retinal and central nervous system haemangioblastomas, RCC, phaeochromocytoma, pancreatic islet cell tumours and renal, pancreatic and epididymal cysts (Maher et al., 1991; Lonser et al., 2003; Maher, 2004). VHL disease displays both intra- and interfamilial variability related to allelic heterogeneity and genetic modifier effects (Crossey et al., 1994; Webster et al., 1998). Analysis of genotype–phenotype correlations in VHL disease led to a classification of VHL disease into four subtypes: Type 1 VHL is characterized by predisposition to haemangioblastomas and RCC but not phaeochromocytoma, Type 2A by haemangioblastomas and phaeochromocytoma but rarely RCC, Type 2B by haemangioblastomas, RCC and phaeochromocytoma and Type 2C by familial phaeochromocytoma only (Crossey et al., 1994; Neumann et al., 1995; Zbar et al., 1996; Woodward et al., 1997). Although most kindreds with Type 2 VHL disease harbour a missense mutation, deletions and truncating mutations account for most with a Type 1 VHL phenotype.
Although multiple functions have been ascribed to the VHL gene product, pVHL, undoubtedly a key function is the role of pVHL in regulating the degradation of the α subunits of the hypoxia-inducible transcription factors (HIF)-1 and-2 (Maxwell et al., 1999; Cockman et al., 2000; Kamura et al., 2000; Ohh et al., 2000). HIF-1 and -2 play a critical role in regulating cellular responses to hypoxia, including the regulation of a wide range of hypoxia-inducible genes involved in energy metabolism (e.g. glucose transporter 1), angiogenesis (e.g. vascular endothelial growth factor (VEGF)) and apoptosis (e.g. NIP3) (Carmeliet et al., 1998; Maxwell and Ratcliffe, 2002). In the absence of wild-type (wt) pVHL, HIF-1α and HIF-2α are stabilized and HIF-1 and -2 expression is upregulated under normoxic conditions. Accordingly RCC with VHL inactivation demonstrate diffuse expression of hypoxia-inducible genes such as VEGF and carbonic anhydrase 9 (Wykoff et al., 2000; Zatyka et al., 2002). However, pVHL also regulated target genes by HIF-independent mechanisms (Zatyka et al., 2002; Wykoff et al., 2004; Lee et al., 2005). There is considerable debate regarding the extent to which HIF-dependent and -independent mechanisms can account for pVHL RCC suppressor activity.
Previously, others and we have reported the identification of pVHL target genes using microarray analysis to document global messenger RNA (mRNA) expression patterns in VHL Wt and VHL-null RCC cell lines (Wykoff et al., 2000, 2004; Zatyka et al., 2002; Jiang et al., 2003; Staller et al., 2003; Maina et al., 2005). In order to identify pVHL targets that are linked to the development of RCC we analysed gene expression profiles in a VHL-null RCC cell line that expresses both HIF-1 and -2 and transfected with wt VHL or Type 2A, 2B and 2C mutants. Comparison of gene expression patterns allowed us to identify those genes that were dysregulated only by mutations that are associated with RCC susceptibility.
Materials and methods
VHL gene expression constructs containing empty vector (pcDNA3.1, Invitrogen, San Diego, CA, USA), wt pVHL (pcDNA 3.1-VHL (1-213). HA) and Type 2A, 2B and 2C mutant pVHL were created and established as stable transfectants into a VHL-defective RCC4 cell line as described previously (Clifford et al., 2001). Cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS), L-glutamine (2 μ M), penicillin (50 IU/ml), streptomycin sulphate (50 μg/ml) and G418 (1 mg/ml). Cells approaching confluence were harvested and the resulting RNA used for gene expression analysis by real-time polymerase chain reaction (PCR).
Oligonucleotide array expression analysis
Poly (A) mRNA purification, biotin labelling of complementary RNA and hybridization were performed according to the manufacturers' protocols. Double-stranded complementary DNA (cDNA) was synthesized from 5 μg poly (A) mRNA using reverse transcriptase Avian Myeloblastosis Virus Superscript Choice system (Gibco BRL, Cheshire, UK) with a T7-poly dT primer. 1 μg of cDNA was in vitro transcribed (Ambion, Austin, TX, USA) in the presence of biotinylated uridine triphosphate and cytidine triphosphate (Enzo Diagnostics, Farmingdale, NY, USA). Target for hybridization was prepared and analysed as described previously (Maina et al., 2005).
Analysis of RNA by real-time quantitative RT–PCR
Total RNA from each sample used on the GeneChip was treated with DNase 1 (Qiagen, West Sussex, UK) according to manufacturer's protocol. Additional clones were also used to assess consistency of results across different RCC4 cell line clones. Real-time quantitative reverse transcriptase (RT)–PCR was carried out as described previously (Maina et al., 2005). Primer sequence for novel RCC specific VHL target genes are shown in (Table 1).
VHL target gene regulation by HIF-1 or -2
Details of methods and cells used for hypoxic response studies have been reported previously (Zatyka et al., 2002). Details of the HIF-1 and -2 small interfering RNA (siRNA)s used have been published in full elsewhere (Raval et al., 2005). These siRNAs were used to downregulate HIF-1 and/or HIF-2 in RCC4 cell lines containing VHL gene expression constructs with either empty vector (pcDNA3.1, Invitrogen) or the wt pVHL (pcDNA 3.1-VHL (1-213). HA). Cells were grown under normoxic and hypoxic conditions. Real-time PCR was performed to determine whether the hypoxia-inducible target genes were regulated by either HIF-1 and/or HIF-2. All results were referenced to samples with both HIF-1 and -2 intact. The data were normalized to β-actin.
Plasmid constructs and colony formation assay
The proteinase-activated receptor 2 (PAR2) and NB thymosin beta (TMSNB) expression constructs were made by cloning the full-length human coding regions of these two genes, from the RCC4 cell line, into the EcoR1-BamHI sites of pcDNA3.1HA vector (Invitrogen). Plasmid constructs were verified by sequencing. Two micrograms of empty vector or expression vector were transfected, using FuGENE 6 (Roche, Basel, Switzerland) reagent according to manufacturer's instructions into Caki-1 cells. aldehyde dehydrogenase (ALDH)7 and ALDH1 expression constructs were made by cloning their full-length human coding regions into the EcoR1-BamHII sites of pcDNA3.1 vector (Invitrogen). 1 μg of empty vector or expression vector was transfected using FuGENE 6 (Roche) according to the manufacturer's instructions, into 5 × 104 SKRC18 cells. Forty-eight hours after transfection, cells were seeded in serial dilution and maintained in DMEM plus 10% fetal bovine serum supplemented with 1 mg/ml G418 (Invitrogen). Fourteen to 21 days after initial seeding, surviving colonies were stained with 0.4% crystal violet (Sigma, Dorset, UK) in 50% methanol, and counted. Each transfection was carried out in triplicate. Additionally, replicate experiments were carried out to obtain further clones for expression analysis and further experiments.
For cell cycle analysis, the full-length human coding regions of TMSNB and PAR2 from the RCC4 cell line were sub-cloned into the EcoR1-BamHI sites of pEGFP-C2 vector (Clontech, CA, USA).
Anchorage-independent growth assay
Caki-1 clones stably expressing TMSNB or the empty vector control were suspended in DMEM 10% FCS agar. Cells were maintained by addition of 200 μl of DMEM 10% FCS, supplemented with 1 mg/μl G418 weekly. After 8 weeks of growth, a final count of colonies was carried out.
Western blot analysis
Protein lysates from 38 RCC tumours/normal pairs were analysed for PAR2 expression. Three RCC cell lines: Caki-1, RCC4-WT (RCC4 cells reconstituted with pVHL) and RCC4-null (RCC4 cells deficient for pVHL) were used for TMSNB and PAR2 colony formation assay. Protein extraction was carried out as described previously (Wiesener et al., 2001; Morris et al., 2005). For PAR2 analysis, 50 μg of each extract was resolved on 12% polyacrylamide gel. Proteins were transferred onto Immobilon P (Millipore, Bedford, MA, USA) for 1 h and probed with anti-PAR2 mouse monoclonal antibody (sc-13504, Santa Cruz Biotechnology, CA, USA) diluted 1/1000. Signals were detected with horseradish peroxidase-conjugated anti-mouse antibody (DAKO, Ely, UK) diluted 1/4000 and visualized by enhanced chemiluminescence (Amersham, Little Chalfont, Buckinghamshire, UK).
TMSNB Western blot analysis was carried out as described previously (Cha et al. (2003). Briefly, cell lysates were concentrated with the use of Microcon YM-3 filters (Millipore, Bedford, MA, USA). Equal amounts of each sample (100 μg) were separated on a 18% sodiumdodecyl sulphate gel. Proteins were transferred onto 0.2 μm pore size Immobilon transfer membrane (Millipore, Bedford, MA, USA) by electrotransfer, and probed with rabbit polyclonal antibody to thymosin β4 (aa 1–14) (ALPCO Diagnostics, Windham, NH, USA) 1/1000 dilution. Signals were detected with horseradish peroxidase-conjugated anti-Rabbit antibody (DAKO, Ely, UK) diluted 1/4000 and visualized with enhanced chemiluminescence (Amersham, UK). After analysis, membranes were stained with India ink for standardization, and quantification was performed using a Bio-Rad imaging densitometer with Quantity One software.
Silencing of PAR2 and TMSNB by RNA interference shRNAmir in SKRC18 cells
SKRC18 RCC cell line which showed high level of PAR2 and TMSNB expression was selected for siRNA transfection. A total of 2 × 105 cells were seeded into six-well tissue culture plate 24 h before transfection. On the day of transfection, cells were 50–60% confluence. Cells were transfected with the appropriate plasmids: TMSNB, PAR2, or nonspecific negative control (Open Biosystems, Catalog # RHS1703) short hairpin (sh)RNAmir plasmids, by using FuGENE 6 reagent (Roche) in accordance with the manufacturer's protocol. Two TMSNB shRNAmir plasmids were used to target the following mRNA sequences: 5′-IndexTermCTATGTTCCCTGGCTAAGA-3′, and 5′-IndexTermCATTGATGACCTTTGTGTA-3′. For PAR2 shRNAmir, 5′- IndexTermGCCTTATTGGTAAG GTTGA-3′ sequence was targeted. Forty-eight hours after transfection, cells were maintained in DMEM plus 10% fetal bovine serum supplemented with 2.5 μg/ml Puromycin (Life Technologies). Cells were harvested at different time points. Stable suppression of gene expression was confirmed using Western blot analysis or RT–PCR.
TMSNB RNA extraction and RT–PCR
The expression of TMSNB mRNA from SKRC18 cells was analysed in untreated controls, TMSNB shRNAmir treated and nonspecific control shRNAmir treated after 48, 72 h, and in stable clones. Total RNA was isolated from SKRC18 cells transfected with TMSNB shRNAmir and nonspecific shRNAmir using RNAzol Bee (Campro Scientific, Amersfoort, The Netherlands). Two micrograms of total RNA was used to generate cDNA using SuperScript III reverse transcriptase (Invitrogen). Primers and conditions used for TMSNB RT–PCR were: an initial denaturation of 10 min at 95°C followed by 35 cycles of 95°C for 30 s, 63.5°C for 30 s, 72°C for 30 s and a final extension of 10 min at 72°C. Primers used for TMSNB RT–PCR were sense primer, which spanned exons 2 and 3 (5′-IndexTermAGGAAACTATCCAGCAAGAGAAAGAGTG-3′) and antisense on exon 3 (5′- IndexTermGATGCAAGAACTACATACACAAAGGTC-3′) to give an expected product size of 307 bp. The PCR thermocycle consisted of an initial denaturation of 5 min at 95°C followed by 35 cycles of 95°C for 5 min, 59°C for 30 s, 72°C for 45 s and a final extension of 5 min at 72°C. Primers used for the glyceraldehyde-3-phosphatedehydrogenase (GAPDH) control were 5′-IndexTermTGAAGGTCG GAGTCAACGGATTTGGT-3′ and 5′-IndexTermCATGTGGGCCATGAGGTCCACCAC-3′. PCR products were visualized on 2% agarose gel with added ethidium bromide.
PAR2 protein extraction and Western blotting
SKRC18 cells trasnfected with PAR2 shRNAmir plasmid were harvested at different time points and lysed in mammalian cell lysis buffer, then Western blot analysis was performed with the use of conventional protocols as described previously.
Colony formation assay on stable SKRC18 clones with PAR2 or TMSNB suppression
For colony formation assay, stable clones of SKRC18 cells showing suppression of PAR2 or TMSNB expression, and SKRC18 cells stably transfected with nonspecific control shRNAmir were plated into a six-well plate at a density of 300 cells/well, and cells were maintained in DMEM plus 10% fetal bovine serum supplemented with 2.5 μg/ml puromycin (Life Technologies). Counting was performed 15–21 days after plating. Surviving colonies were stained with 0.4% crystal violet (Sigma) in 50% methanol, and counted. Each experiment was carried out in triplicate.
Student's t-test was carried out using SPSS 13.0 and statistical significance was taken at the 0.05 level.
Identification of pVHL target genes associated with RCC tumorigenesis
To identify pVHL target genes linked to the development of RCC, we compared gene expression patterns using the Affymetrix U95v2 GeneChip microarrays to assay 12 600 mRNA transcripts in stable RCC4 cell line transfectants expressing empty vector (VHL-null), full-length wt human pVHL (RCC4/wtVHL) and with pVHL bearing one of the missense germline point mutations (Types 2A, 2B or 2C VHL). Two independent experiments and three clones in total of each cell line were analysed. The results were interrogated to identify genes that were consistently dysregulated in the Type1 and 2B RCC4 cell lines versus the wt pVHL, RCC4 cell line, but that did not show changes in gene expression between Type 2A, 2C and wt pVHL comparisons.
We identified 19 genes (0.01 of total genes analysed) that demonstrated a ⩾twofold difference between RCC-associated pVHL mutants and non-RCC-associated pVHL in both experiments and in the three different clones (see Table 2). Fourteen of the 19 genes were upregulated by wt pVHL and five were downregulated by wt pVHL.
Real-time PCR analysis was undertaken in each of the cell lines for 8/19 pVHL targets. Expression was normalized to β-actin and the results of the microarray analysis were confirmed for all eight genes (see Table 3). Five of the novel pVHL targets were upregulated by wt pVHL (ITGB3BP (integrin beta-3 binding protein), ACTA2 (smooth muscle alpha-actin), ALDH1A1, ALDH7, UGT2B7 (3,4 catechol estrogen uridine diphosphate-glucuronosyltransferase) and three were downregulated TMSNB, PAR2 and INSIG2 (insulin-induced protein 2) (Table 4).
Regulation of RCC-associated pVHL target genes by oxygen
To determine whether the novel pVHL target genes were likely to be regulated by HIF-dependent or -independent mechanisms, we investigated whether they demonstrated evidence of hypoxic-induction. Thus we analysed the effect of hypoxia on expression of the eight target genes in a pVHL RCC4 cell line such that target gene expression was compared in hypoxia (1%) and normoxia (21%) by real-time PCR to expression of the known hypoxia-inducible pVHL target gene, PAI-1. As expected, expression of PAI-1 was upregulated by hypoxia (mimicking the effect of VHL inactivation) and two of the three target genes downregulated by wt pVHL (INSIG2 and TMSNB) were also induced by hypoxia. Of the five novel RCC-associated pVHL target genes that were upregulated by wt pVHL, hypoxia mimicked the effect of VHL inactivation for ITGB3BP, ACTA2, ALDH1A1 and UGT2B7. However, expression of ALDH7 and PAR2 were not hypoxia-inducible.
Previously, we and others have demonstrated that certain pVHL hypoxia responsive target genes (e.g. CCND1 and GPR56) may be hypoxia-responsive in RCC cell lines but not in other types of cancer. To investigate this effect for hypoxia-responsive RCC-associated pVHL target genes we analysed the effect of hypoxia on expression of ITGB3BP, ACTA2, ALDH1A1, UGT2B7, INSIG2 and TMSNB of in breast (HBL-100) and bladder (EJ-29) cancer cell lines. Except for ACTA2, the results matched those seen in RCC4. However ACTA2, which was downregulated by hypoxia in a RCC4 cell line expressing wt VHL was not hypoxia-responsive in the two non-RCC cell lines.
To determine whether hypoxia-responsive RCC-associated pVHL target genes are regulated to an equivalent extent by the HIF-1 and -2 transcription factors, we analysed expression of these five genes in a normoxic VHL-null (expressing empty vector pcDNA3.1, Invitrogen) RCC4 cell line in which siRNA was used to differentially downregulate HIF-1α or -2α. SiRNA of HIF-1α and -2α had equivalent effects on the regulation of UGT2B7 and ACTA2, (Figure 1a and b) but TMSNB, ITGB3BP and ALDH1A1 were preferentially regulated by HIF-1 (Figure 1c, d and e). As expected, siRNA for HIF-1α and -2α has no effect on expression of the non-hypoxia-inducible target ALDH7 (Figure 1f).
VHL target gene expression in primary tumours
The relationship between pVHL regulation of target genes in an isogenic cell line and in primary RCC with VHL inactivation (somatic VHL mutation±3p25 allele loss) was investigated by analysing RNA or protein expression. Thus in nine primary RCC with VHL inactivation analysed by real-time PCR, the effect of VHL-inactivation in the RCC4 cell line was mimicked in 4/9 matched normal-tumour pairs for TMSNB, 3/9 for PAR2, 3/9 for ALDHIAI, 2/9 for ACTA2 and 2/9 for ITGB3BP, 1/9 for ALDH7 and 3/9 for UGT2B7 and 1/9 for INSIG2. In addition, PAR2 protein expression was assessed by Western blotting in another set of matched normal-tumour pairs of RCC. PAR2 protein expression was upregulated in 24% (9/38) of RCC (compared to the corresponding normal tissues) (see Figure 1g).
Effect of restoration of ALDH1 and ALDH7 expression on RCC cell line growth
pVHL inactivation is associated with downregulation of expression of ALDH1 and ALDH7. To define the effect of restoring ALDH1 and ALDH7expresion on a VHL-null RCC cell line, empty plasmid pcDNA3.1 or pcDNA3.l containing the full open-reading frame of ALDH1 and ALDH7 was transfected into the SKRC18 RCC cell line (a pVHL-null cell line with low expression of ALDH1 and ALDH7). Cells were selected with geneticin (G418), 48 h after transfection and resistant colonies developing 14–21 days later were stained with 0.4% crystal violet. There was no difference in the colony formation efficiency after transfection with ALDH1 or ALDH7 compared with transfection with empty vector control in three independent experiments.
PAR2 and TMSNB increase colony formation
To investigate whether the increased expression of PAR2 or TMSNB associated with VHL inactivation was likely to be implicated in renal tumorigenesis, we transfected wt PAR2 and TMSNB into the Caki-1 (wt endogenous VHL) RCC cell line. High transfection efficiencies were obtained for empty vector, PAR2 and TMSNB constructs (all transfections were spiked with empty green fluorescent protein plasmid). Forty-eight hours after transfection, expression of pcDNA3.1HA-PAR2 or pcDNA3.1HA-TMSNB was observed by Western blot Figure 2). Both PAR2 and TMSNB transfections significantly increased colony formation compared to empty vector control transfections in three independent experiments (mean 175%, P<0.0004 and mean 234%, P<0.003, respectively) (Figure 2).
TMSNB increases anchorage-independent growth
The effect of TMSNB on anchorage-independent growth in a soft agar colony formation assay was assessed in Caki-1 cells transfected with empty pcDNA3.1HA. Vector or pcDNA3.1HA-TMSNB. Following selection, clones expressing high levels of TMSNB were isolated and cells were seeded and incubated in soft agar for 8 weeks, each experiment was performed in triplicate with three independent clones. Cells transfected with pcDNA3.1HA-TMSNB showed robust colony growth, whereas there was little colony growth in cells transfected with empty pcDNA3.1HA vector alone. In a duplicate experiment using Caki-1 clones, both the number and size of colonies was increased. The number of large (⩾100 μm) colonies was increased 280% (P=0.0001) in clones expressing TMSNB when compared with the control clones (Figure 3).
Effects of PAR2 and TMSNB silencing in SKRC18 RCC cell line: decreased levels of PAR2 and TMSNB significantly alter the growth rate of SKRC18 cells and inhibit colony formation
As our studies of the effect of overexpression of PAR2 and TMSNB in a pVHL-wt RCC cell line had shown that both PAR2 and TMSNB increased cellular proliferation and cell growth, we next investigated the effects of silencing of PAR2 and TMSNB expression in a VHL-null cell line. A SKRC18 RCC cell line with high PAR2 and TMSNB expression was transfected with PAR2 or TMSNB shRNAmir plasmid. Stable clones of SKRC18 cells showing suppression of PAR2 or TMSNB were prepared. Colony formation assay was performed on these stable clones and compared to SKRC18 cells transfected with nonspecific control shRNAmir. The number of SKRC18 cell clones was then counted 14–21 days after transfection. Silencing of PAR2 or TMSNB significantly decreased the growth rate and colony formation of SKRC18 cells. The reduction in number and size of colonies in the silenced group than in the control group was statistically significant (mean 70%, P<0.0005 for PAR2 and mean 60%, P<0.0004 for TMSNB) (Figures 4 and 5).
The complex genotype–phenotype correlations in VHL disease provide an opportunity to correlate specific pVHL functions to organ-specific tumorigenesis. Previously we and others investigated the effect of Type 1, 2A, 2B and 2C VHL mutations on the ability of pVHL to regulate HIF-1 and -2 and bind fibronectin. Although VHL mutations that are associated with RCC (Type 1 and 2B) were associated with loss of HIF-regulation and fibronectin-binding, Type 2A mutations also demonstrated partially impaired ability to regulate HIF suggesting that HIF dysregulation is necessary but not sufficient for RCC tumorigenesis (Clifford et al., 2001; Hoffman et al., 2001). Direct evidence for a role of HIF overexpression in renal tumorigenesis in VHL disease was obtained from experiments in which mutant HIF-1α and -2α resistant to pVHL-mediated degradation were overexpressed in a RCC cell line with wt pVHL. Interestingly HIF-2, but apparently not HIF-1, overexpression was associated with increased tumour growth (Kondo et al., 2003; Yang et al., 2003). These findings suggested that inhibition of HIF-2 activity might provide a novel therapeutic approach for VHL-related RCC. However HIF-1 and -2 regulate transcription of a wide repertoire of genes, including targets implicated in angiogenesis, cell metabolism, cell cycle control and cell migration (Wykoff et al., 2000, 2004; Zatyka et al., 2002; Jiang et al., 2003; Staller et al., 2003; Maina et al., 2005). As all HIF targets are unlikely to have equivalent effects on renal tumorigenesis the identification of specific RCC-related targets might offer a better approach to developing novel therapeutic strategies.
High-density oligonucleotide expression arrays have been used to identify a wide range of pVHL target genes (Wykoff et al., 2000, 2004; Zatyka et al., 2002; Jiang et al., 2003; Staller et al., 2003; Maina et al., 2005). Previously we identified 30 genes differentially expressed between a pVHL-wt and pVHL-null RCC4 cell line (Clifford et al., 2001). In the current study, we sought to preferentially identify pVHL targets implicated in RCC tumorigenesis using pVHL mutations associated with high to low susceptibility to RCC (Neumann et al., 1995; Zbar et al., 1996; Woodward et al., 1997). Using a significance threshold of twofold differential expression, we identified 19 novel genes that showed differential gene expression with the RCC-associated mutations (Type 1 and Type 2B) compared to wt pVHL and mutations not associated with RCC (Type 2A and 2C). The microarray expression results were verified for all eight of the RCC-associated pVHL targets that were analysed by real-time PCR, suggesting that our experimental approach had a low false-positive rate. 14 of 19 RCC-associated VHL target genes were upregulated by wt pVHL and might therefore be considered as potential tumour suppressors. However in many cases little information on gene function was available. For example KIAA0740 (RHOBTB1) is a member of the Rho guanidine triphosphatasase superfamily and is highly expressed in the kidney but it's precise function has not been defined (Li et al., 2004). The ITGB3BP (integrin beta-3 binding protein) gene encodes three transcripts, one of which, NRIF3, induces apoptosis human breast cancer cell lines (Huff et al., 2001). Thus in many cases further work is required to investigate the relevance of dysregulation to renal tumorigenesis. To investigate whether reduced expression of the aldehyde dehydrogenases (which catalyse the oxidation of various aliphatic and aromatic aldehydes to the corresponding carboxylic acids), ALDH1 (ALDH1A1) and ALDH7 (ALDH3B1), was likely to be causally associated with RCC tumorigenesis we re-expressed these genes in a VHL-deficient cell line. However no effect on in vitro growth was detected.
Five RCC-associated VHL targets (SLC5A3, KIAA0836 (D-glucuronyl C5-epimerase), INSIG2, TMSNB and PAR2 were downregulated by wt pVHL and so were considered potential RCC oncogenes. Interestingly two of these genes, TMSNB and PAR2, had been linked to tumorigenesis. Thus TMSNB (NB thymosin beta or TMSL8) is a member of the β-thymosin family. These highly conserved polar 5 KDa peptides are widely expressed and play an important role in the organization of the cytoskeleton. Thus they bind to and sequester actin monomers (G-actin) and therefore inhibit actin polymerization (Grant et al., 1995). In addition they have been implicated in angiogenesis, wound healing and apoptosis (Hall, 1991, 1994; Frohm et al., 1996; Malinda et al., 1997; Iguchi et al., 1999; Niu and Nachmias, 2000). Thymosin beta 4 and beta 10 mRNA expression was reported to be upregulated in RCC and normal human embryonic kidney (compared to normal kidney) (Dery et al., 1998). We found that overexpression of TMSNB increased growth of RCC cell lines in vitro and induced cell cycle progression. As TMSNB was hypoxia-inducible, these functions might also be implicated in non-VHL-related cancers with tumour hypoxia.
Proteinase-activated receptors (PARs) are G-protein coupled receptors (Nguyen et al., 1999). PAR2 (F2RL1) is one of the four known PARs and has been shown to play a role in secretion, smooth muscle contractility, neuromodulation, inflammation, fibrosis, angiogenesis and cancer metastasis (Bertog et al., 1999; Cocks et al., 1999; Danahay et al., 2001; Fiorucci et al., 2001; Hoogerwerf et al., 2001; Richard et al., 2001; Vergnolle et al., 2001; Frungieri et al., 2002; Hollenberg, 2002; Milia et al., 2002). We found that increased PAR2 expression enhanced, and reduced expression decreased, growth of RCC cell lines in vitro, suggesting that upregulation of TMSNB and PAR2 following VHL inactivation contributes to RCC tumorigenesis, although, we note that TMSNB and PAR2 were only upregulated in a proportion of primary RCC with VHL inactivation. However it is unclear whether this may be related to the effect of specific VHL mutations or whether the loss of pVHL regulation of TMSNB and PAR2 expression is modified by the additional genetic and epigenetic effects that occur in renal tumorigenesis.
We investigated the relationship between expression of eight RCC-associated VHL target genes and hypoxia. Most targets (6/8) were hypoxia-responsive such that the effect of hypoxia mimicked that of VHL inactivation. Previously Type 2A mutations (that are not associated with RCC) affecting the pVHL HIF-binding domain were shown to be associated with HIF-dysregulation (Clifford et al., 2001). However the HIF-dysregulation associated with 2A mutations appears to be less severe than that with Type 1 and 2B mutations and Type 2A mutant pVHL retains significant ubiquitin ligase activity towards HIF-1α in vitro (Clifford et al., 2001; Knauth et al., 2006). These observations are consistent with our findings that both HIF-dependent and -independent target genes were associated with RCC susceptibility. We and others have observed that some pVHL target genes are hypoxia-inducible in RCC cell lines but not in other cancers such as breast or bladder. Such targets (e.g. ACTA2, CCND1 and GPR56) show tissue-specific responses to hypoxia (Zatyka et al., 2002; Bindra et al., 2002; Baba et al., 2003; Maina et al., 2005; Wang et al., 2005). This phenomenon may be implicated in the striking organ-specific pattern of tumour susceptibility observed in VHL disease. Among hypoxia-responsive genes we and others have observed that pVHL targets such as Cyclin D1 may respond differentially to HIF-1 and -2 (Lee et al., 2005; Maina et al., 2005; Raval et al., 2005). Furthermore, HIF-2 and -1 appear to have differential effects on renal oncogenesis with HIF-2 overexpression enhancing tumour growth in vivo (Kondo et al., 2002, 2003; Maranchie et al., 2002; Yang et al., 2003; Raval et al., 2005). Nevertheless, although the net effect of HIF-2 target genes may be more pro-oncogenic than that of HIF-1target genes, not all target genes will have equivalent effects. Thus we demonstrated that TMSNB overexpression stimulated RCC line growth in vitro, but was primarily regulated by HIF-1. This suggests that although, on balance, activation of HIF-1 target genes may not have a major effect on tumorigenesis, different HIF-1 target genes may have varying effects. TMSNB has not previously identified as a hypoxic response gene, but we found that it was hypoxia-inducible in all three types of cancer cell lines. This suggests that TMSNB may be implicated in non-VHL-related cancers in which tissue hypoxia leads to activation of the HIF pathway (Maxwell, 2005).
One of the key unanswered questions regarding pVHL function is the extent to which HIF-independent functions contribute to pVHL tumour suppressor function. It appears that pVHL regulation of JunB/EglN3 expression via a HIF-independent pathway has a critical role in phaeochromocytoma development (Lee et al., 2005) and HIF-independent pVHL functions have also been implicated in RCC (Clifford et al., 2001). We identified two non-hypoxia-responsive VHL target genes, PAR2 and ALDH7. Whereas we did not find evidence that ALDH7 had a significant effect on RCC growth, overexpression of PAR2 in a wt pVHL RCC cell line increased tumour cell growth and silencing of PAR2 in a VHL-null cell line reduced cell growth. Most Type 1 VHL mutations are deletions, truncating mutations before the elongin binding domain (Trp88, Asn90, Gln96, Tyr98 and Tyr112) or missense substitutions at residues buried within the that result in loss of structural integrity. In contrast Type 2 mutations are mainly missense and the prototypic 2B mutations at codon 167 interfere with elongin binding whereas this is retained with 2A and 2C mutations. Thus increased PAR2 expression associated with Type 1 and 2B mutations could be linked to failure to ubiquitinate pVHL targets that can still be bound by 2A and 2C mutant pVHL proteins. Identification of the mechanisms by which pVHL regulates PAR2 can provide important insights into pVHL HIF-independent tumour suppressor functions.
Baba M, Hirai S, Yamada-Okabe H, Hamada K, Tabuchi H, Kobayashi K et al. (2003). Loss of von Hippel–Lindau protein causes cell density dependent deregulation of cyclin D1 expression through hypoxia-inducible factor. Oncogene 22: 2728–2738.
Bertog M, Letz B, Kong W, Steinhoff M, Higgins MA, Bielfeld-Ackermann A et al. (1999). Basolateral proteinase-activated receptor (PAR-2) induces chloride secretion in M-1 mouse renal cortical collecting duct cells. J Physiol 521 (Part 1): 3–17.
Bindra RS, Vasselli JR, Stearman R, Linehan WM, Klausner RD . (2002). VHL-mediated hypoxia regulation of cyclin D1 in renal carcinoma cells. Cancer Res 62: 3014–3019.
Carmeliet P, Dor Y, Herbert JM, Fukumura D, Brusselmans K, Dewerchin M . (1998). Role of HIF-1alpha in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature 394: 485–490.
Cha HJ, Jeong MJ, Kleinman HK . (2003). Role of thymosin beta4 in tumor metastasis and angiogenesis. J Natl Cancer Inst 1995: 1674–1680.
Clifford SC, Cockman ME, Smallwood AC, Mole DR, Woodward ER, Maxwell PH et al. (2001). Contrasting effects on HIF-1alpha regulation by disease-causing pVHL mutations correlate with patterns of tumourigenesis in von Hippel–Lindau disease. Hum Mol Genet 10: 1029–1038.
Cockman ME, Masson N, Mole DR, Jaakkola P, Chang GW, Clifford SC et al. (2000). Hypoxia inducible factor-alpha binding and ubiquitylation by the von Hippel–Lindau tumor suppressor protein. J Biol Chem 275: 25733–25741.
Cocks TM, Sozzi V, Moffatt JD, Selemidis S . (1999). Protease-activated receptors mediate apamin-sensitive relaxation of mouse and guinea pig gastrointestinal smooth muscle. Gastroenterology 116: 586–592.
Crossey PA, Richards FM, Foster K, Green JS, Prowse A, Latif F, et al. (1994). Identification of intragenic mutations in the von Hippel–Lindau disease tumour suppressor gene and correlation with disease phenotype. Hum Mol Genet 3: 1303–1308.
Danahay H, Withey L, Poll CT, van de Graaf SF, Bridges RJ . (2001). Protease-activated receptor-2-mediated inhibition of ion transport in human bronchial epithelial cells. Am J Physiol Cell Physiol 280: C1455–C1464.
Dery O, Corvera CU, Steinhoff M, Bunnett NW . (1998). Proteinase-activated receptors: novel mechanisms of signaling by serine proteases. Am J Physiol 274: C1429–C1452.
Fiorucci S, Mencarelli A, Palazzetti B, Distrutti E, Vergnolle N, Hollenberg MD et al. (2001). Proteinase-activated receptor 2 is an anti-inflammatory signal for colonic lamina propria lymphocytes in a mouse model of colitis. Proc Natl Acad Sci USA 98: 13936–13941.
Frohm M, Gunne H, Bergman AC, Agerberth B, Bergman T, Boman A et al. (1996). Biochemical and antibacterial analysis of human wound and blister fluid. Eur J Biochem 237: 86–92.
Frungieri MB, Weidinger S, Meineke V, Kohn FM, Mayerhofer A . (2002). Proliferative action of mast-cell tryptase is mediated by PAR2, COX2, prostaglandins, and PPARgamma: Possible relevance to human fibrotic disorders. Proc Natl Acad Sci USA 99: 15072–15077.
Grant DS, Kinsella JL, Kibbey MC, LaFlamme S, Burbelo PD, Goldstein Al et al. (1995). Matrigel induces thymosin beta 4 gene in differentiating endothelial cells. J Cell Sci 108 (Part 12): 3685–3694.
Hall AK . (1991). Differential expression of thymosin genes in human tumors and in the developing human kidney. Int J Cancer 48: 672–677.
Hall AK . (1994). Molecular interactions between G-actin, DNase I and the beta-thymosins in apoptosis: a hypothesis. Med Hypotheses 43: 125–131.
Hoffman MA, Ohh M, Yang H, Klco JM, Ivan M, Kaelin Jr WG . (2001). von Hippel–Lindau protein mutants linked to type 2C VHL disease preserve the ability to downregulate HIF. Hum Mol Genet 10: 1019–1027.
Hollenberg MD . (2002). PARs in the stars: proteinase-activated receptors and astrocyte function. Focus on ‘thrombin (PAR-1)-induced proliferation in astrocytes via MAPK involves multiple signaling pathways’. Am J Physiol Cell Physiol 283: C1347–C1350.
Hoogerwerf WA, Zou L, Shenoy M, Sun D, Micci MA, Lee-Hellmich H et al. (2001). The proteinase-activated receptor 2 is involved in nociception. J Neurosci 21: 9036–9042.
Huff T, Muller CS, Otto AM, Netzker R, Hannappel E . (2001). beta-Thymosins, small acidic peptides with multiple functions. Int J Biochem Cell Biol 33: 205–220.
Iguchi K, Usami Y, Hirano K, Hamatake M, Shibata M, Ishida R . (1999). Decreased thymosin beta4 in apoptosis induced by a variety of antitumor drugs. Biochem Pharmacol 57: 1105–1111.
Jiang Y, Zhang W, Kondo K, Klco JM, St Martin TB, Dufault MR et al. (2003). Gene expression profiling in a renal cell carcinoma cell line: dissecting VHL and hypoxia-dependent pathways. Mol Cancer Res 1: 453–462.
Kamura T, Sato S, Iwai K, Czyzyk-Krzeska M, Conaway RC, Conaway JW . (2000). Activation of HIF1alpha ubiquitination by a reconstituted von Hippel–Lindau (VHL) tumor suppressor complex. Proc Natl Acad Sci USA 97: 10430–10435.
Knauth K, Bex C, Jemth P, Buchberger A . (2006). Renal cell carcinoma risk in type 2 von Hippel–Lindau disease correlates with defects in pVHL stability and HIF-1alpha interactions. Oncogene 25: 370–377.
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.
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.
Lee S, Nakamura E, Yang H, Wei W, Linggi MS, Sajan MP et al. (2005). Neuronal apoptosis linked to EglN3 prolyl hydroxylase and familial pheochromocytoma genes: developmental culling and cancer. Cancer Cell 8: 155–167.
Li D, Das S, Yamada T, Samuels HH . (2004). The NRIF3 family of transcriptional coregulators induces rapid and profound apoptosis in breast cancer cells. Mol Cell Biol 24: 3838–3848.
Lonser RR, Glenn GM, Walther M, Chew EY, Libutti SK, Linehan WM et al. (2003). von Hippel–Lindau disease. Lancet 361: 2059–2067.
Maher ER, Iselius L, Yates JR, Littler M, Benjamin C, Harris R et al. (1991). Von Hippel–Lindau disease: a genetic study. J Med Genet 28: 443–447.
Maher ER . (2004). Von Hippel–Lindau disease. Curr Mol Med 4: 833–842.
Maina EN, Morris MR, Zatyka M, Raval RR, Banks RE, Richards FM et al. (2005). Identification of novel VHL target genes and relationship to hypoxic response pathways. Oncogene 24: 4549–4558.
Malinda KM, Goldstein AL, Kleinman HK et al. (1997). Thymosin beta 4 stimulates directional migration of human umbilical vein endothelial cells. FASEB J 11: 474–481.
Maranchie JK, Vasselli JR, Riss J, Bonifacino JS, Linehan WM, Klausner RD . (2002). The contribution of VHL substrate binding and HIF1-alpha to the phenotype of VHL loss in renal cell carcinoma. Cancer Cell 1: 247–255.
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, 1999.
Maxwell PH, Ratcliffe PJ . (2002). Oxygen sensors and angiogenesis. Semin Cell Dev Biol 13: 29–37.
Maxwell PH . (2005). The HIF pathway in cancer. Semin Cell Dev Biol 16: 523–530.
Milia AF, Salis MB, Stacca T, Pinna A, Madeddu P, Trevisani M . (2002). Protease-activated receptor-2 stimulates angiogenesis and accelerates hemodynamic recovery in a mouse model of hindlimb ischemia. Circ Res 91: 346–352.
Morris MR, Gentle D, Abdulrahman M, Maina EN, Gupta K, Banks RE . (2005). Tumor suppressor activity and epigenetic inactivation of hepatocyte growth factor activator inhibitor type 2/SPINT2 in papillary and clear cell renal cell carcinoma. Cancer Res 165: 4598–4606.
Neumann HPH, Eng C, Mulligan L, Glavac D, Ponder BAJ, Crossey PA et al. (1995). Consequences of direct genetic testing for germline mutations in the clinical management of families with multiple endocrine neoplasia, type II. JAMA 274: 1149–1151.
Nguyen TD, Moody MW, Steinhoff M, Okolo C, Koh DS, Bunnett NW . (1999). Trypsin activates pancreatic duct epithelial cell ion channels through proteinase-activated receptor-2. J Clin Invest 103: 261–269.
Niu M, Nachmias VT . (2000). Increased resistance to apoptosis in cells overexpressing thymosin beta four: a role for focal adhesion kinase pp125FAK. Cell Adhes Commun 7: 311–320.
Ohh M, Park CW, Ivan M, Hoffman MA, Kim TY, Huang LE et al. (2000). Ubiquitination of hypoxia-inducible factor requires direct binding to the beta-domain of the von Hippel–Lindau protein. Nat Cell Biol 2: 423–427.
Raval RR, Lau KW, Tran MG, Sowter HM, Mandriota SJ, Li JL et al. (2005). Contrasting properties of hypoxia-inducible factor 1 (HIF-1) and HIF-2 in von Hippel-Lindau-associated renal cell carcinoma. Mol Cell Biol 25: 5675–5686.
Richard DE, Vouret-Craviari V, Pouyssegur J . (2001). Angiogenesis and G-protein-coupled receptors: signals that bridge the gap. Oncogene 20: 1556–1562.
Staller P, Sulitkova J, Lisztwan J, Moch H, Oakeley EJ, Krek W . (2003). Chemokine receptor CXCR4 downregulated by von Hippel–Lindau tumour suppressor pVHL. Nature 425: 307–311.
Vergnolle N, Wallace JL, Bunnett NW, Hollenberg MD . (2001). Protease-activated receptors in inflammation, neuronal signaling and pain. Trends Pharmacol Sci 22: 146–152.
Wang V, Davis DA, Haque M, Huang LE, Yarchoan R . (2005). Differential gene up-regulation by hypoxia-inducible factor-1alpha and hypoxia-inducible factor-2alpha in HEK293 T cells. Cancer Res 65: 3299–3306.
Webster AR, Richards FM, MacRonald FE, Moore AT, Maher ER . (1998). An analysis of phenotypic variation in the familial cancer syndrome von Hippel–Lindau disease: evidence for modifier effects. Am J Hum Genet 63: 1025–1035.
Wiesener MS, Munchenhagen PM, Berger I, Morgan NV, Roigas J, Schwiertz A et al. (2001). Constitutive activation of hypoxia-inducible genes related to overexpression of hypoxia-inducible factor-1alpha in clear cell renal carcinomas. Cancer Res 61: 5215–5222.
Woodward ER, Eng C, McMahon R, Voutilainen R, Affara NA, Ponder BAJ et al. (1997). Genetic predisposition to phaeochromocytoma: analysis of candidate genes GDNF, RET and VHL. Hum Mol Genet 6: 1051–1056.
Wykoff CC, Pugh CW, Maxwell PH, Harris AL, Ratcliffe PJ . (2000). Identification of novel hypoxia dependent and independent target genes of the von Hippel–Lindau (VHL) tumour suppressor by mRNA differential expression profiling. Oncogene 19: 6297–6305.
Wykoff CC, Sotiriou C, Cockman ME, Ratcliffe PJ, Maxwell P, Liu E et al. (2004). Gene array of VHL mutation and hypoxia shows novel hypoxia-induced genes and that cyclin D1 is a VHL target gene. Br J Cancer 90: 1235–1243.
Yang JC, Haworth L, Sherry RM, Hwu P, Schwartzentruber DJ, Topalian SL . (2003). A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer. N Engl J Med 349: 427–434.
Zatyka M, da Silva NF, Clifford SC, Morris MR, Wiesener MS, Eckardt KU et al. (2002). Identification of cyclin D1 and other novel targets for the von Hippel–Lindau tumor suppressor gene by expression array analysis and investigation of cyclin D1 genotype as a modifier in von Hippel–Lindau disease. Cancer Res 62: 3803–3811.
Zbar B, Kishida T, Chen F, Schmidt L, Maher ER, Richards FM et al. (1996). Germline mutations in the Von Hippel–Lindau disease (VHL) gene in families from North America, Europe, and Japan. Hum Mutat 8: 348–357.
About this article
Cite this article
Abdulrahman, M., Maina, E., Morris, M. et al. Identification of novel VHL targets that are associated with the development of renal cell carcinoma. Oncogene 26, 1661–1672 (2007). https://doi.org/10.1038/sj.onc.1209932
- renal cell carcinoma
- gene expression
Thymosin beta 15A (TMSB15A) is a predictor of chemotherapy response in triple-negative breast cancer
British Journal of Cancer (2012)
Proteinase-activated receptor 2 expression in breast cancer and its role in breast cancer cell migration
Hereditary leiomyomatosis and renal cell cancer: an unusual and aggressive form of hereditary renal carcinoma
Nature Clinical Practice Oncology (2007)