Wilms' tumour 1 can suppress hTERT gene expression and telomerase activity in clear cell renal cell carcinoma via multiple pathways

Background: Wilms' tumour 1 (WT1) gene was discovered as a tumour suppressor gene. Later findings have suggested that WT1 also can be oncogenic. This complexity is partly explained by the fact that WT1 has a number of target genes. Method: WT1 and its target gene human telomerase reverse transcriptase (hTERT) were analysed in clear cell renal cell carcinoma (ccRCC). In vitro experiments were performed to examine the functional link between WT1 and hTERT by overexpression of WT1 isoforms in the ccRCC cell line, TK-10. Results: WT1 demonstrated lower RNA expression in ccRCC compared with renal cortical tissue, whereas hTERT was increased, showing a negative correlation between WT1 and hTERT (P=0.005). These findings were experimentally confirmed in vitro. The WT1 generated effect on hTERT promoter activity seemed complex, as several negative regulators of hTERT transcription, such as SMAD3, JUN (AP-1) and ETS1, were activated by WT1 overexpression. Downregulation of potential positive hTERT regulators, such as cMyc, AP-2α, AP-2γ, IRF1, NFX1 and GM-CSF, were also observed. Chromatin immunoprecipitation analysis verified WT1 binding to the hTERT, cMyc and SMAD3 promoters. Conclusion: The collected data strongly indicate multiple pathways for hTERT regulation by WT1 in ccRCC.

The Wilms' tumour 1 (WT1) gene is an important regulator of cell growth and development with expression in the developing embryo kidney, adult urogenital system, central nervous system and in the hematopoietic system (reviewed in Scharnhorst et al (2001)). In 1990, WT1 was discovered as a tumour suppressor gene in Wilm's tumour, a childhood kidney neoplasm (Committe TNWTS, 1991;Haber et al, 1990). This gene encodes a 49 -52 kDa protein with an N-terminal domain involved in RNA/protein interactions critical for its transcriptional regulatory function . The C-terminal domain is involved in RNA and protein interactions and harbours four zinc-fingers and two nuclear localisation signals, which permit binding to target DNA sequences (Bruening et al, 1996). By two main alternate splicing events including splicing of exon 5 (17 amino acids) and exon 9 (three amino acids: lysine, threonine and serine (KTS)), four WT1 protein isoforms are produced designated by the presence or absence of exon 5/KTS as A (À/À), B ( þ /À), C (À/ þ ) and D ( þ / þ ) (Haber et al, 1991). WT1A (À/À) was demonstrated to induce morphological changes and promote cell migration and invasion in ovarian cancer cells (TYK) (Jomgeow et al, 2006). In osteosarcoma cell lines, induction in programmed cell death was preferentially mediated by WT1B ( þ /À) through transcriptional repression of the EGFR gene (Haber et al, 1996). WT1D ( þ / þ ) can cause a morphological transition from an epithelial to a more mesenchymal phenotype (Burwell et al, 2007). As suggested by the presence of zinc-fingers, WT1 is a potent transcriptional regulator. A large number of genes coding for growth factors (e.g., TGF-b, CSF-1), growth factor receptors (e.g., insulin R, IGF-IR, EGFR), transcription factors (e.g., EGR, WT1, cMyc, Pax2, Dax-1 and Sry) and other proteins (e.g., ODC, MDR1, Hsp70, p21, Bcl-2) have been identified as WT1 target genes (reviewed in Ariyaratana and Loeb (2007); Yang et al (2007)). Also, the WT1 protein has interacting partners such as p53 and STAT3, which can modulate WT1 transcription functions (Maheswaran et al, 1993;Rong et al, 2006).
Telomerase activation is a key event during immortalisation and malignant transformation contributing to telomere stabilisation and extended replicative capacity. Human telomerase reverse transcriptase (hTERT) is a key component of the telomerase complex regulated on several levels. The hTERT promoter contains methylation-accessible CpG islands and its methylation status has been associated with hTERT transcriptional repression (Devereux et al, 1999;Chatagnon et al, 2009). The hTERT core promoter and sequences upstream contain numerous binding sites for positive and negative regulators of transcription suggesting a complex regulation (Takakura et al, 1999). A number of factors can directly or indirectly regulate the hTERT promoter, including activators (cMyc, Sp1, ER, HIF-1a, E6 protein, activating enhancer-binding protein (AP-2) and so on), as well as repressors (WT1, AP-1, p53, p73, MZF, pRb, E2F and so on) (Cukusic et al, 2008). We have recently shown that the PI3K pathway is one significant road for hTERT regulation in ccRCC and that cMyc binding to the hTERT promoter seemed important for this control (Sitaram et al, 2009). Oh et al (1999) identified WT1 as a transcriptional repressor of hTERT in virally transformed human embryonic kidney 293 cells, but the WT1 regulation seemed to be cell type specific. In this study, we could demonstrate negative associations between WT1 and hTERT and between WT1 and cMyc in clinical ccRCC samples, data that were verified by cell line transfection experiments. Forced expression of WT1 in the ccRCC TK-10 cell line reduced hTERT mRNA levels and telomerase activity by direct WT1 binding to the hTERT promoter, but also by affecting several genes known to regulate hTERT transcription. Our results suggest that WT1 can act as a tumour suppressor in ccRCC via multiple pathways leading to downregulation of hTERT.

Tissue samples
We performed the study on a total of 73 ccRCC tumour specimens and 26 tumour-free renal cortical tissue samples. The tissue specimens were collected between February 1988 and December 2003 under a protocol approved by the Human Ethics Committee of the Medical Faculty, Umeå University. Each patient participated after providing informed consent and during the later years also with informed and signed consent. All pathology specimens were reviewed by pathologists according to (Skinner et al, 1971). Tumour stages were classified according to the TNM classification 2002 (Greene and Sobin, 2002). Follow-up medical records of the patients were retrospectively updated by surgical urologists and were used for survival analysis.

RNA extraction and reverse transcription
Total RNA was isolated from snap-frozen tumour specimens and tumour-free renal cortical tissue using the TRIzol method (Invitrogen, Stockholm, Sweden). cDNA was prepared by reverse transcription with the Superscript II Reverse transcriptase kit according to the manufacturers' protocol (Invitrogen).

Genome-wide gene expression array
Total RNA (200 ng) of each sample was used to produce cRNA according to the provided protocol of the Illumina Total Prep RNA Amplification Kit (Ambion Inc., Austin, TX, USA).
A total of 750 ng biotinylated cRNA was used for hybridisation to a human HT12 Illumina Beadchip gene expression array according to the manufacturers' protocol (Illumina, San Diego, CA, USA). The arrays were scanned using the Illumina Bead Array Reader. For data analysis and normalisation the Illumina BeadStudio 3.2 software was used. Cell signalling pathway and network analysis was carried out with the Metacore software (GeneGo Inc, St Joseph, MI, USA). Samples were normalised by the quantile algorithm, genes with signal below background levels were excluded, and differentially expressed genes were identified by fold change calculations and with the Illumina custom differential expression algorithm (described in the Illumina Gene Expression Module user guide) to identify X2-fold and statistically (Po0.01) differently expressed genes.

Real-time PCR
Quantitative real-time PCR (qRT -PCR) using TaqMan technology was performed for WT1 expression. Following primers and probe given, a 119-bp product was used to detect WT1 mRNA levels. Forward primer: 5 0 -GCTATTCGCAATCAGGGTTACAG-3 0 (located on exon 1/2), reverse primer: 5 0 -TGGGATCCTCATGCTTGAATG-3 0 (located on exon 2); and TaqMan probe: 5 0 -CACACGCCCTCGC ACCATGC-3 0 (located on exon 2). The b-actin gene was used for normalisation of cDNA templates, and sequences of the primes and probe were previously described (Inoue et al, 1997). The qRT -PCR reaction was initiated with 2 min of incubation at 501C and then for 10 min at 951C, followed by 45 cycles of denaturation at 951C for 15 s and annealing at 601C for 2 min. Standard curves were generated by 10-fold dilutions of plasmid DNA containing the insert WT1 or b-actin genes.
The expression of hTERT mRNA was measured using the Light Cycler TeloTAGGG hTERT quantification kit (Roche Diagnostics, GmbH, Mannheim, Germany). By using a reference standard curve provided from the qRT -PCR kit, the relative hTERT mRNA expression (with reference to housekeeping gene, porphobilinogen deaminase (PBGD)) was calculated as previously described (Sitaram et al, 2009). For c-Myc, mRNA quantification was performed by qRT -PCR using SYBR green I technology (Roche Diagnostics) as previously described (Sitaram et al, 2009). Values of target gene expression were calculated with template normalised to house keeping gene PBGD.
SMAD3, ETS1 and AP-2a were analysed by TaqMan assays according to manufacturer's protocol with the TaqMan universal PCR mastermix and run on the ABI Prism 7000 Sequence Detection System, SMAD3 (Hs00232222_m1), ETS1 (Hs_00901425_m1) and AP-2a (Hs_01029410_m1) (Applied Biosystems, Foster City, CA, USA). cDNA from the T-cell lymphoma cell line (CCRF) was used to generate the standard curves. Collected data were normalised to b-actin as described above.

ChIP analysis
Chromatin immunoprecipitations (ChIPs) were performed using the Chromatin Immunoprecipitation Kit (Upstate Millipore, Billerica, MA, USA). Approximately 2 -3 Â 10 6 WT1-transfected TK-10 cells were crosslinked with 1% formaldehyde, followed by glycine to quench unreacted formaldehyde. Chromatin was sonicated on ice to shear crosslinked DNA to about 200 -1000 bp in length using a Sonifier ultrasonic cell disruptor (Branson, Danbury, CT, USA) with 12 Â 10 s pulses. The sheared chromatin was resuspended in dilution buffer and 1% of the chromatin was removed as input, followed by immunoprecipitation using protein G magnetic beads with 2 mg of either anti-WT1 (C-19) antibody (Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA,) or normal rabbit IgG (Cell Signalling Technology Inc, Danvers, MA, USA) at 41C overnight with rotation. After the reversal of crosslinks by incubation in ChIP elution buffer containing proteinase K at 621C for 2 h, DNA was purified using spin columns.
The PCRs containing 2 ml of the immunoprecipitated DNA or input chromatin, primers and AmpliTaq Gold (Applied Biosystem) in a 50 ml volume were performed with initial denaturation at 951C for 10 min, followed by 35 cycles (951 for 30 s, 551C for 30 s and 721C for 45 s) and a final extension at 721C for 10 min. The primer sequences were as follows for hTERT promoter (NG_009265): The primer sequences for cMyc promoter were described previously (Han et al, 2004). PCR products were fractionated on 1% agarose gel, and ethidium bromide-stained DNA was visualised on Ultraviolet Transilluminator (Spectroline, Westbury, NY, USA).

Telomerase activity determination
Telomerase activity was evaluated using quantitative telomerase detection kit (QTD kit, Allied Biotech Inc, Ijamsville, CA, USA). A total reaction volume of 25 ml consisted of 12.5 ml of 2 Â QTD pre-mix (provided by the kit), 250 ng protein CHAPS extract (supplemented with 1 U ml -1 RNAsin and 1 mM DTT) and water. The qRT -PCR was performed on a 7000 sequence system (Applied Biosciences). Standard curve generated by TSR control template allows the calculation of the amount of template with telomeric repeat created by telomerase using 7000 SDS system software (Applied Biosystem).

Statistical analysis
Statistical analysis was performed using SPSS (version 15, SPSS Inc., Chicago, IL, USA) statistical software. Mann -Whitney U-test was used to compare differences in the gene expressions of two independent variables. Correlations between two variables were tested according to Spearman's correlation test (statistical significance Pp0.05).

RESULTS
WT1 expression was inversely correlated to hTERT and cMyc in ccRCC WT1 mRNA levels were analysed in 73 ccRCC specimens and 26 tumour-free renal cortical tissue samples using qRT -PCR. Significantly lower WT1 RNA levels were found in the tumour samples in comparison with renal cortical tissue (Po0.0001), as shown in Figure 1A, indicating a downregulation of WT1 in ccRCC. Immunoblotting for WT1 revealed lower protein levels in randomly selected tumour samples compared with tumour-free renal cortical tissues ( Figure 1B) We have previously demonstrated significantly higher mRNA levels of hTERT and cMyc in ccRCC compared with renal cortical tissue (Sitaram et al, 2009). In the present ccRCC samples, negative correlations were found between WT1 and hTERT (P ¼ 0.005, r ¼ À0.328, Figure 1C) and between WT1 and cMyc (P ¼ 0.05, r ¼ À0.246, n ¼ 64, not shown in figures). We could also demonstrate a tendency towards a negative association between WT1 and hTERT for a subset of samples with lower expression levels for both hTERT and WT1 (P ¼ 0.063, Figure 1C).

WT1 expression was not associated with clinical features
WT1 mRNA levels did not differ depending on patient age, gender, tumour grade or stage (P40.05, for each parameter, data not shown). No significant difference in survival time was observed when the ccRCC cases were subdivided into two subgroups with a cutoff at the median WT1 mRNA level (data not shown).

Forced expression of WT1 can suppress hTERT and cMyc mRNA levels
In order to answer whether the WT1 protein can function as a negative regulator of hTERT and/or cMyc transcription, we performed transfection experiments using TK-10 cells. High expression levels of WT1 isoforms A and D were demonstrated at 24 and 48 h after transfection (Figure 2A). It has been demonstrated that nuclear localisation of WT1 is a prerequisite for its transcriptional regulatory capacity (Ye et al, 1996). Therefore, immunostaining was performed 48 h after transfection verifying WT1 protein localisation in nuclei ( Figure 2B). WT1A and WT1D overexpression induced a decrease in hTERT and cMyc mRNA levels as shown in Figures 2C and D. The downregulation of cMyc was also demonstrated by immunoblotting (Figure 2A). The repressive effects on cMyc and hTERT varied between experiments. By plotting the mean expression values for hTERT and cMyc in a series of separate WT1 transfections, we found a strong correlation, indicating a similar degree of inhibition ( Figure 2E). This gives support for the idea that cMyc is involved in the hTERT regulation. Furthermore, both WT1A and WT1D transfections reduced telomerase activity after 24 and 48 h ( Figure 2F).

WT1 can regulate hTERT transcription via multiple pathways
Gene expression alterations induced by WT1 transfection in TK-10 cells were detected using whole genome expression array analysis and several genes reported to be WT1 targets were found to be affected (Figure 3). Further analysis was focused on hTERTregulating genes identifying an increased expression of several repressors. Fold changes, based on the array data, in expression of hTERT transcriptional regulators after WT1 transfection are summarised in Table 1. Two negative hTERT regulators, SMAD3 and JUN, were significantly upregulated. Moreover, the GM-CSF, ETS1 and IRF1 genes also with suggested negative effects on hTERT transcription demonstrated increased expression. In contrast, AP-2a, AP-2g and NFX1, all with potential positive effects on hTERT expression, were downregulated. The altered expression of SMAD3, ETS1 and AP-2a was validated by qRT -PCR, showing a good correlation with the array data ( Figure 4).

WT1 binds to the hTERT, c-Myc and SMAD3 promoters
To demonstrate that WT1 proteins bind to the promoters of the hTERT, cMyc and SMAD3 genes, we performed ChIP experiments. As shown in Figure 5, a single band was observed when chromatins from WT1-transfected TK-10 cells were immunoprecipitated with WT1 antibody. This was not observed when rabbit IgG antibody was used. By ChIP/PCR assay we could demonstrate direct binding of WT1 to the hTERT, cMyc and SMAD3 promoters.

DISCUSSION
In this study, we describe a functional link between WT1 expression and the transcriptional activity of one of its target genes, hTERT, in human ccRCC. This link was identified through (1) an inverse relationship between WT1 and hTERT gene expression levels in clinical ccRCC samples; (2) direct binding of WT1A and WT1D to the hTERT promoter; and (3) Figure 3 Altered gene expression of Wilms' tumour 1 (WT1) target genes and human telomerase reverse transcriptase (hTERT) transcriptional regulators by WT1 overexpression in TK-10 cells detected using expression array analysis. Network analysis was performed based on array data using the Metacore GeneGo software. Increased gene expression is indicated by a filled red circle on the upper right corner of each network object, whereas a blue dot indicates downregulation (X2-fold, Po0.01 compared with control). Gene/protein objects are represented by various shapes and colours depending on their functional annotations. Colours of the lines indicate inhibition (red) and activation (green), and gray arrows represent unspecified interactions (e.g., promoter binding). of the hTERT gene after experimentally induced WT1 overexpression. The collected data indicated that WT1 can control hTERT expression via multiple pathways and thereby act as a tumour suppressor in ccRCC. WT1 is known to be a potent transcriptional regulator of many downstream targets, and can thus function as a tumour suppressor or an oncogene depending on cell type and tumour entity. Recent studies have demonstrated high expression of WT1 in bone and soft tissue sarcomas, breast cancer and lung cancer (Loeb et al, 2001;Oji et al, 2002;Ueda et al, 2003). Therefore, the WT1 gene has been proposed as an oncogene in these contexts. In contrast to these observations, we here found significantly lower WT1 RNA levels in ccRCC samples compared with tumour-free renal cortical tissue. Only few studies have previously investigated WT1 in human RCC. Using northern blot analysis, Campbell et al (1998) demonstrated aberrant WT1 expression in four out of five RCC samples and in several RCC-derived cell lines arguing against WT1 being a tumour suppressor in this tumour type. In contrast, later studies have generated data in accordance with our study. Niu et al (2005) showed low WT1 transcription levels in ccRCC samples in comparison with tumour-free kidney tissue. By immunohistochemistry Nakatsuka et al (2006) found WT1 protein in four out of twelve (33%) ccRCC samples to be compared with 480% positivity in endometrial and brain tumours. On the basis of these reports and this study the collected data suggest that WT1 can act as a tumour suppressor in ccRCC.
The transcriptional activity of the hTERT gene is the net result of many positively and negatively acting factors. WT1 has been identified as a cell type-specific hTERT transcriptional repressor, acting through WT1-binding site in the promoter (Oh et al, 1999). Our clinical data, showing a negative correlation between WT1 and hTERT RNA levels, support the view that WT1 can suppress hTERT. In addition, the experimental cell line data further demonstrated that WT1 overexpression caused hTERT and telomerase repression. We could also show a direct binding of WT1 to the hTERT promoter in line with previous observations (Oh et al, 1999). Bellon and Nicot (2008) recently demonstrated that IL-2-induced upregulation of hTERT and telomerase activity in HTLV1-infected cells were due to sequestration of WT1 in the cytoplasm. In this study, we demonstrated nuclear localisation of WT1 protein after transfection, indicating that one prerequisite for its suppressive function was fulfilled.
One known activator of hTERT gene transcription is cMyc (Wu et al, 1999). In a previous study, we demonstrated a positive correlation between hTERT and cMyc in ccRCC (Sitaram et al, 2009). Cell line experiments showed that hTERT could be activated by a novel PTEN regulator, DJ-1, and cMyc seemed necessary for the upregulation (Sitaram et al, 2009). Studies of WT1 effects on the cMyc promoter have given divergent results indicating both stimulatory (Han et al, 2004) and repressive functions (Hewitt et al, 1995;Han et al, 2004). We observed a strong trend towards a negative correlation between WT1 and cMyc RNA levels in the clinical ccRCC samples, and overexpression of WT1 did suppress cMyc at the transcriptional and protein level. Our observations   Figure 5 DNA binding of Wilms' tumour 1 (WT1) protein to the human telomerase reverse transcriptase (hTERT), cMyc and SMAD3 promoters. Chromatin immunoprecipitation (ChIP)/PCR assay after forced overexpression of WT1A or WT1D in TK-10 cells as indicated. The location and size for each PCR product are illustrated on the right. TSS, transcriptional start site.
indicate that loss of WT1 function in ccRCC can result in increased cMyc expression, which may contribute to hTERT expression. The repressive effects of WT1 on hTERT and cMyc expression seemed to be parallel in repeated experiments, further indicating that cMyc acts as a regulator of hTERT in ccRCC.
Microarray analyses have previously been used to identify WT1 target genes (Kim et al, 2007). In this study, we found WT1 induced effects on a number of known targets for WT1, but also on hTERT transcriptional regulators previously not described as WT1 targets. Among negative hTERT regulators, SMAD3 and JUN were strongly upregulated by WT1 transfection. ChIP analysis revealed that WT1 can directly bind to the SMAD3 promoter. Direct repression of hTERT by SMAD3 via TGF-b signalling has been reported (Lacerte et al, 2008). JUN was originally thought to be identical to the transcription factor AP-1. However, it is now known that AP-1 constitutes a group of dimeric basic regionleucine zipper proteins that belong to the FOS, MAF and ATF subfamilies. JUN is the most potent transcriptional activator in this group of proteins (Hattori et al, 1988). Previous studies have shown that overexpression of AP-1 can downregulate hTERT transcription in cancer cells (Takakura et al, 2005). The combination of c-Fos/c-Jun or c-Fos/JunD strongly suppresses hTERT promoter activity in transient-expression experiments (Takakura et al, 2005).
IRF1 has been shown to be important in apoptosis and cell differentiation (Harada et al, 1993). A previous study demonstrated that IRF1 is a mediator for interferon-g-induced inhibition of hTERT expression and telomerase activity in cervical cancer cells (Lee et al, 2003). In this study we observed about two-fold increased expression of IRF1 by WT1D, indicating that it might have a negative effect on hTERT promoter activity. The clinical and experimental data presented here strongly suggest that WT1 contributes to hTERT inactivation, directly by acting on the promoter and indirectly via hTERT-negative regulators such as SMAD3 and JUN.
GM-CSF can regulate hTERT transcription both positively and negatively, but only in combination with other genes (Mano et al, 2000). GM-CSF was strongly upregulated by WT1 in our analysis, indicating that it might act as an hTERT repressor. Further, we also found that WT1 overexpression induced increased ETS1 expression. Transcriptional activation of ETS1 by WT1 has been reported in tumour vascularisation via regulation of endothelial cell proliferation and migration (Wagner et al, 2008). The role of ETS proteins, especially ETS1 and ETS2, in telomerase regulation seems to depend on ETS-binding sites in the hTERT promoter and protein -protein interactions (Dwyer et al, 2007). Xiao et al (2003) demonstrated an inhibitory as well as an activating effect on hTERT transcription mediated through different ETS-binding sites. In this study, increased ETS1 and reduced hTERT expression by WT1 may suggest a negative role of ETS1 in hTERT transcriptional regulation in ccRCC.
Two isoforms of NFX1 have been identified, the longer 1120-amino acid isoform as NFX1-123 and the shorter 833-amino acid isoform as NFX1-91 (Gewin et al, 2004). NFX1-123 was demonstrated to co-activate the hTERT promoter with cMyc, whereas NFX1-91 repressed the hTERT promoter (Gewin et al, 2004). Significant reduction of NFX1-123 expression by WT1 was found in our array analysis, suggesting that it might act as an hTERT activator in ccRCC. The protein AP-2b has been identified as a transcriptional activator of the hTERT promoter in human lung cancer cells (Deng et al, 2007). The family of AP-2 proteins consists of five different transcriptional factors (a, b, g, d and e), encoded by separate genes and with different biological functions. The AP-2-binding site at nucleotides -129 to -121 in the hTERT promoter is essential for transcriptional regulation during differentiation of human sarcoma cells (Ma et al, 2003), but the role of individual AP-2 family members has not been detailed. In this study AP-2a and AP-2g were downregulated by WT1, suggesting that WT1 thereby can affect hTERT regulation.
Interestingly, in T-cell cultures we observed that upregulation of telomerase was associated with decreased expression of GM-CSF, JUND, ETS1, SMAD3 and increased expression of AP-2a and cMyc (Degerman et al, 2010). The collected data strengthen the scenario of hTERT regulation via multiple pathways.
By gene expression array analysis, we found that forced expression of WT1 had a regulatory function on many known WT1 target genes. However, the effects on transcription of some target genes were different from previously published data in other cellular systems, further supporting the view of cell type specificity regarding WT1 function. In conclusion, this study strongly suggests that WT1 can act as a tumour suppressor in ccRCCregulating hTERT gene expression via multiple pathways.