Accumulation of Malignant Renal Stem Cells Is Associated with Epigenetic Changes in Normal Renal Progenitor Genes


  • Sally Metsuyanim,

    1. Department of Pediatrics and Pediatric Stem Cell Research Institute, Sheba Medical Center, Tel Hashomer, Israel
    Search for more papers by this author
  • Naomi Pode-Shakked,

    1. Department of Pediatrics and Pediatric Stem Cell Research Institute, Sheba Medical Center, Tel Hashomer, Israel
    2. Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
    Search for more papers by this author
  • Kai M. Schmidt-Ott,

    1. Department of Nephrology and Hypertension, Medical Faculty of the Charité and HELIOS Klinikum Berlin, Max-Delbrueck-Center for Molecular Medicine, Berlin-Buch, Germany
    Search for more papers by this author
  • Gilmor Keshet,

    1. Department of Pediatric Hemato-Oncology and Sheba Cancer Research Center, Sheba Medical Center, Tel Hashomer, Israel
    Search for more papers by this author
  • Gideon Rechavi,

    1. Department of Pediatric Hemato-Oncology and Sheba Cancer Research Center, Sheba Medical Center, Tel Hashomer, Israel
    2. Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
    Search for more papers by this author
  • Danith Blumental,

    1. Department of Pediatrics and Pediatric Stem Cell Research Institute, Sheba Medical Center, Tel Hashomer, Israel
    Search for more papers by this author
  • Benjamin Dekel M.D., Ph.D.M.

    Corresponding author
    1. Department of Pediatrics and Pediatric Stem Cell Research Institute, Sheba Medical Center, Tel Hashomer, Israel
    2. Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
    • Pediatric Stem Cell Research Institute, Safra Children's Hospital, Sheba Medical Center, Tel Hashomer 52621, Israel. Telephone: 03-5302517; Fax: 03-5305787
    Search for more papers by this author


Recent studies indicate a dual epigenetic role of the Polycomb group (PcG) proteins in self-renewal of stem cells and oncogenesis. Their elevation in our previous human kidney microarray screen led us examine whether they participate in processes involving normal and malignant renal progenitors. We therefore analyzed the expression of the PcG genes (EZH2, BMI-1, EED, SUZ12) in relation to that of the nephric-progenitor genes (WT1, PAX2, SALL1, SIX2, CITED1) using real-time polymerase chain reaction and methylation assays during renal development, regeneration, and tumorigenesis. Although all of the nephric-progenitor genes were shown to be developmentally regulated, analysis of polycomb gene expression during murine nephrogenesis and in an in vitro induction model of the nephrogenic mesenchyme indicated dynamic regulation only for EZH2 in the normal renal progenitor population. In contrast, induction of adult kidney regeneration by ischemia/reperfusion injury resulted primarily in rapid elevation of BMI-1, whereas EZH2 was silenced. Analysis of renal tumorigenesis in stem cell-like tumor xenografts established by serial passage of Wilms' tumor (WT) in immunodeficient mice showed cooperative upregulation of all PcG genes. This was accompanied by upregulation of WT1, PAX2, and SALL1 but downregulation of SIX2. Accordingly, methylation-specific quantitative polymerase chain reaction demonstrated promoter hypomethylation of WT1, PAX2, and SIX2 in primary WT and fetal kidneys, whereas progressive WT xenografts showed hypermethylation of SIX2, possibly leading to loss of renal differentiation. PcG genes vary in expression during renal development, regeneration, and tumorigenesis. We suggest a link between polycomb activation and epigenetic alterations of the renal progenitor population in initiation and progression of renal cancer.

Disclosure of potential conflicts of interest is found at the end of this article.


Author contributions: S.M., N.P.-S., and K.M.S.-O.: collection and/or assembly of data, data analysis and interpretation; G.K.: technologic advising; G.R.: financial support, consulting and advising; D.B.: collection and/or assembly of data; B.D.: conception and design, financial support, data analysis and interpretation, manuscript writing.

Metanephroi are the primordia of adult mammalian kidneys [1]. The early development of the metanephros is a complex process that involves highly regulated interactions between two derivatives of the intermediate mesoderm, the Wolffian duct and the metanephric mesenchyme. Reciprocal signaling between the metanephric mesenchyme and a derivative of the nephric duct known as the ureteric bud results in branching of the ureteric bud and condensation of metanephric mesenchyme at its tips. The condensed mesenchyme is thought to form a precursor cell population, which both maintains itself at the tips of the ureteric bud (via proliferation and/or addition from the surrounding noncondensed mesenchyme) and gives off cells that differentiate into nephrons, the functional filtration unit of the kidney [2]. Several genes have been identified that are expressed specifically in the undifferentiated metanephric mesenchyme and are required for proper differentiation of the metanephric kidney, including PAX2 [3], WT1 [4], EYA1 [5], SIX2 [6], the HOX11 paralogous group [7], CITED1 [8], and SALL1 [9]. These genes are considered early markers of kidney progenitor cells.

Wilms' tumor, a common pediatric kidney cancer, is classified as a primitive, multilineage malignancy of embryonic renal precursors [10], which fail to terminally differentiate into epithelium and continue to proliferate, thus forming blastemal elements in the tumor. The multipotency of these cells is supported by a characteristic histology of the tumor, which includes, apart from blastemal components, more mature epithelial and stromal cells and suggests that blastemal cells have differentiated at least in part [10]. At a molecular level, several studies have shown that Wilms' tumors express markers of early kidney development, and more recently, global gene profiling has provided additional evidence for the similarities that are suggested by morphology [11, [12]–13]. We could show that the overall gene expression profile of a Wilms' tumor (WT) specimen was most similar to that observed for a human 8-week-gestation kidney but not kidneys of later gestation [11]. Furthermore, more detailed microarray experiments have shown that genes that are overexpressed in Wilms' tumors tend to be expressed at the time of the first contact between the ureteric bud and the metanephric mesenchyme [12], the earliest stage of metanephric development.

It is becoming clear that many, if not most, malignancies arise from a population of cells that exclusively maintain the ability to self-renew and sustain the tumor via the expression of tumor-progenitor genes [14]. This progenitor population is often biologically distinct from the bulk of differentiated cancer cells that characterize the disease. Recently, we applied a strategy to identify normal and tumor-progenitor genes of the developing human kidney on the basis of the molecular global analysis of progressive WT xenografts, human fetal kidneys and their counterparts' human adult kidney, and renal cell carcinoma [13]. Upon serial passages in immunodeficient mice, the WT xenografts (originally derived from favorable histology nonsyndromic WT) displayed blastemal accumulation and dedifferentiation compared with original WT, which contained more mature epithelial and stromal cells and therefore positively selected the renal progenitor population in vivo. Interestingly, the WT xenografts highly overexpressed genes that have been shown to specify the normal renal progenitor cell population (PAX2, LIM1, EYA1, SIX1, SIX2, SALL1, CITED1, WT1) [3, [4], [5], [6], [7], [8]–9]. In addition, among the most significantly upregulated genes, we identified those of the Wnt/β-Catenin signaling pathway (FZD2, FZD7, SFRP1, CTNNBIP1), which have been thoroughly described in kidney development [15, 16], as well as genes of the Polycomb group (EZH2, BMI-1).

The Polycomb group (PcG) gene family, which functions in silencing gene expression through epigenetic modification, is highly conserved throughout evolution [17, 18]. Dysregulation of PcG genes, which were found to be upregulated in several kinds of human cancer, has been linked with the aberrant proliferation of cancer cells [17, 18]. Furthermore, several PcG genes appear to regulate self-renewal of specific stem cell types, suggesting a link between the maintenance of cellular homeostasis and tumorigenesis. Indeed, very recent results have suggested that dynamic repression of developmental pathways by Polycomb complexes is required for maintaining stem cell pluripotency and self-renewal of murine and human embryonic stem cells [19, 20]. The idea that the oncogenic function of the Polycomb genes might be associated with their role in stem cell maintenance led us to examine possible roles in processes involving normal and malignant renal progenitor cells, namely, development, regeneration, and oncogenesis.

Materials and Methods

Establishment and Maintenance of WT Xenografts

WT xenografts were established from WT stage I with favorable histology as previously described [13]. Briefly, the original surgical samples were placed on ice, minced into 3–5-mm pieces, and implanted s.c. into SCID mice. After an initial latency period of 2 months, tumor growth was noted in 90% of the mice. Thereafter, the tumor was serially passaged s.c. with Matrigel (Becton, Dickinson and Company, Bedford, MA, as minced tumor pieces or by direct injection of single-cell suspension. The xenografts were maintained by serial passages in SCID and nude mice by harvesting the tumors under sterile conditions and placing them immediately in cooled Hanks' balanced saline solution (HBSS) (Sigma-Aldrich, St. Louis, Tumor cells were dissociated under sterile conditions, first by mincing the tissue with scissors to small fragments and then by gentle mechanical homogenization through a stainless steel mesh. Viable cells were separated from debris by layering over Ficoll-Paque 400 (Amersham Pharmacia Biotech AB, Uppsala, Sweden, and centrifugation at 500g for 20 minutes. Viable cells at the interface were collected, counted, and resuspended in cooled HBSS at a concentration of 2 × 107 cells per milliliter. For orthotopic implantation, 100 μl of tumor cell suspension (2 × 106 cells) was directly injected into several spots of the mouse kidney using a 27-gauge needle. The musculature layer of the abdominal wall and the skin were separately closed with 4-0 absorbable Vycril sutures (NHS Logistics, Derbyshire, U.K.,

WT Cultures

Primary WT samples were cleaned in HBSS, minced with scissors, and digested by incubation for 24 hours at 37°C in Iscove's modified Dulbecco's medium (IMDM) supplemented with collagenase II (Life Technologies, Grand Island, NY, The cell suspension was forced through a 100-μm mesh to achieve a single-cell suspension. To obtain primary tumor cell cultures, the resulting cancer cells were washed in IMDM plus 10% fetal bovine serum (FBS) (Life Technologies) and plated onto collagen-coated dishes in IMDM plus 10% FBS, supplemented with 100 ng/ml EGF, 100 ng/ml fibroblast growth factor (FGF)-2, and 10 ng/ml SCF (R&D Systems Inc., Minneapolis, Cells were passaged upon reaching confluence.


Immunostaining was performed as previously described [21]. Briefly, sections of 5 μm were mounted on SuperFrost/Plus glass (Menzel-Gläser, Braunschweig, Germany, and labeled with strept-avidin-biotin using a Histostain-Plus kit (Zymed, San Francisco, Heat-induced antigen retrieval was performed by controlled microwave treatment using an H2800 model processor (Energy Bean Sciences, Inc., Agawan, MA, in 10 mM citrate buffer, pH 6.0, for 10 minutes at 97°C. The sections were treated with 3% H2O2 for 5 minutes. Consecutive sections were incubated for 1 hour with polyclonal rabbit anti-human EZH2 antibody (Zymed) diluted 1:500. Immunostaining with this commercially available EZH2 antibody has previously been validated [22]. Negative control incubations were performed by substituting nonimmune serum for the primary antibody. Biotinylated second antibody was applied for 10 minutes followed by incubation with horseradish peroxidase (HRP)-conjugated streptavidin for 10 minutes. Following each incubation, the slides were washed thoroughly with Optimax wash buffer (BioGenex, San Ramon, CA, The immunoreaction was visualized by an HRP-based chromogen/substrate system, including 3,3′ Diaminobenzidine (DAB) (brown) chromogen (liquid DAB substrate kit; Zymed). The sections were then counterstained with Mayer's hematoxylin, dehydrated, and mounted for microscopic examination.

Dissection of Rat Metanephric Mesenchymes

Metanephric mesenchymes were dissected using minutien pins from embryonic day (E) 13.5 rats. Cross-contamination between ureteric buds and metanephric mesenchymes was ruled out by visual inspection and staining with lectin Dolichos biflorus, which selectively labels the ureteric bud at this stage [23]. For organ culture, rat metanephric mesenchymes were placed on filters (Transwell; collagen-coated, 0.4-μm pore size; Corning Enterprises, Corning, NY, and grown in Dulbecco's modified Eagle's medium/Ham's F-12 medium with insulin (5 μg/ml), transferrin (5 μg/ml), selenium (5 ng/ml), dexamethasone (5 μg/ml), prostaglandin (5 μg/ml), T3 (5 ng/ml; Sigma-Aldrich), FGF-2 (3 nmol/l), transforming growth factor-α (3 nmol/l), and leukemia inhibitory factor (50 ng/ml; all cytokines from R&D Systems).

Ischemia/Reflow Experiments

For unilateral ischemia/reflow, a flank incision was made, and the left renal pedicle was clamped for 40 minutes using a vascular clamp (Fine Science Tools, Inc., Foster City, CA, [24]. The abdomen was covered with gauze moistened in phosphate-buffered saline, and the mice were maintained at 37°C using a warming pad. After 40 minutes, the clamp was removed, and reperfusion was confirmed visually. To determine the extent of acute injury, control mice were sacrificed 24 hours after ischemia/reflow, and kidneys were collected and processed for histology using H&E and Sirius red staining.

RNA Isolation

Total RNA from human and mouse samples was isolated from each sample using Trizol (Life Technologies; Invitrogen, Carlsbad, CA, Total RNA from rat metanephric mesenchymes was extracted using the RNeasy mini kit (Qiagen, Valencia, CA, with on-column DNase digestion according to the instructions of the manufacturer. An Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA, was used to confirm RNA integrity.

Quantitative Reverse Transcription-Polymerase Chain Reaction

cDNA was synthesized using Super Script First-strand Synthesis System for polymerase chain reaction (PCR)-reverse transcription (RT) (Invitrogen) on total RNA. Real-time PCR of human and mouse samples was done using an ABI7900HT sequence detection system (PerkinElmer Life and Analytical Sciences, Boston,; Applied Biosystems, Foster City, CA, in the presence of SYBR Green (SYBR Green PCR kit; Qiagen) [13]. This fluorochrome incorporates stoichiometrically into the amplification product, providing real-time quantification of double-stranded DNA PCR product. Primers were designed to amplify an 80–120-base pair fragment with an annealing temperature of 50°C to 65°C. Primer sequences are described in Table 1.

Table Table 1.. Primer sequences for quantitative reverse transcription-polymerase chain reaction
original image

Methylation Assay

DNA samples were purified from tissues using ArchivePure DNA Tissue Kit (5 Prime, Gaithersburg, MD, Bisulfite conversion of DNA samples was performed using EZ-DNA Methylation Kit (Zymo Research, Orange, CA, Primers were designed for amplification of methylated DNA copies of WT1, SIX2, and PAX2 genes and of the ALU sequences (the endogenous control). Primer sequences are summarized in Table 2.

Table Table 2.. Primer sequences for methylation assay
original image

WT1, SIX2, and PAX2 probes contained a black hole quencher (BHQ-1) at the 3′ end and a 6-carboxyfluorescein fluorophore at the 5′ terminus. The ALU probe contained a groove binder nonfluorescent quencher (MGBNFQ) at the 3′ end and a 6FAM fluorophore at the 5′ terminus. Absolute quantification the real-time PCR assay was performed using an ABI 7900HT sequence detection system (PerkinElmer; Applied Biosystems) in the presence of TaqMan 1000 Reaction Gold (Applied Biosystems). Each reaction contained 1–1.8 μg of bisulfite-treated DNA. A standard curve for each tested gene was established using four points dilutions (1, 1:10, 1:100, 1:1000) of M.SssI (New England Biolabs, Ipswich, MA, DNA. For each sample, methylated copy number (MCN) was calculated by the following formula: MCN = log 10 [Tested gene average copy number/(Alu mean copy number × 0.00001)].

Statistical Analysis

Comparisons of continuous variables were performed by analysis of variance. To compare differences in mean value and SD of variables during follow-up, a two-tailed paired t test was performed. Statistical significance was considered at p < .05. All analyses were performed using SPSS version 11.0 for Windows (SPSS, Chicago,


PcG Gene Expression in Normal Nephrogenesis

Our initial goal was to establish the expression of the PcG genes (PCR2 components: EZH2, SUZ12, EED; PCR1 component, BMI-1) in relation to the nephric-progenitor genes (WT1, PAX2, CITED1, SIX2) during normal nephrogenesis. Review of our kidney microarray data showed that among the PcG genes only EZH2 displayed a significant difference in the hybridization signal between pooled human fetal and adult kidneys [13]. Nevertheless, the microarray data do not include temporal expression. For this purpose, we analyzed mRNA expression by quantitative reverse transcription (qRT)-PCR during mouse nephrogenesis, which continues 2 weeks postnatal. As expected, temporal expression patterns of the nephric-progenitor genes (WT1, PAX2, CITED1, SIX2) show discrete peaks of expression in early and mid-gestation and rapid downregulation in the neonatal (2 weeks) and adult kidneys (Fig. 1A). We found a similar expression pattern for EZH2, but not for SUZ12, EED, or BMI-1 (Fig. 1B), suggesting that only EZH2 is strictly developmentally regulated in the kidney. To determine whether this finding was specific for the progenitor compartment of the embryonic kidney, we analyzed the expression of the PcG genes in the organ culture model of epithelial differentiation of the isolated metanephric mesenchyme [23]. For this purpose, we microdissected metanephric mesenchyme from rat E13.5 embryos and induced epithelial differentiation over a 7-day time course in organ culture by applying a combination of growth factors [25]. As detected by real-time reverse transcription-PCR, only EZH2 was significantly decreased during epithelial differentiation of the metanephric mesenchyme, similar to the nephric-progenitor gene CITED1, whereas EED remained unchanged and BMI-1 was upregulated (Fig. 1C). We were unable to analyze SUZ12 as there is no known rat homolog. Consistent with the observed downregulation of EZH2 in adult mouse kidneys in vivo, this transcript decreased substantially, coinciding with epithelial differentiation and depletion of the progenitor pool in the organ culture system. Finally, immunostaining of a human 10-week-gestation kidney localized EZH2 to the nephrogenic mesenchyme (Fig. 1D–1G). In sum, these in vivo and in vitro findings suggest a role for EZH2, but not other PcG genes we analyzed, in the renal progenitor compartment during kidney development in the embryo.

Figure Figure 1..

Gene expression in normal nephrogenesis. Temporal expression patterns of nephric-progenitor (WT1, PAX2, CITED1, SIX2) (A) and polycomb group (EZH2, EED, SUZ12, BMI-1) (B) transcripts in the developing mouse kidney. For quantitative reverse transcription (RT)-polymerase chain reaction (PCR) experiments, we used pooled RNA isolated from five different mouse embryonic kidneys (E12, E13), neonatal (1 day, 2 W) and adult kidneys. (C): Regulation of EZH2, EED, BMI-1, and CITED1 during differentiation of nephrogenic mesenchyme. Freshly dissected rat metanephric mesenchymes were grown and differentiated in culture medium supplemented with fibroblast growth factor-2, transforming growth factor-α, and leukemia inhibitory factor. Real-time RT-PCR was performed using primers specific for EZH2, EED, BMI-1, and CITED1 on metanephric mesenchymes at baseline (day 0) and after 2 and 7 days of culture. Data were calculated as average ± SEM (three independent experiments). EZH2 and CITED1 are exclusively downregulated with differentiation. ***, p < .001; **, p < .01; *, p < .05 versus baseline levels (day 0). (D–G): Immunostaining of EZH2 in a human 10-W-gestation kidney; shown are negative control (D) and stained kidney (E) at low magnification (×4). High magnification (×40) of two separate fields (F, G) demonstrates expression in the nephrogenic mesenchyme. Abbreviations: E, embryonic day; RQ, 2(−ΔΔCt); W, week.

PcG Gene Expression in Renal Regeneration

The adult kidney possesses a remarkable regenerative capacity after acute ischemic and/or toxic injury. This regenerative capacity manifests itself by proliferation and migration of poorly differentiated cells along the denuded basement membrane of injured tubular segments within a few days after ischemic insult [26]. This process has previously been shown to be accompanied by reactivation of at least some of the nephric-progenitor genes [27]. We therefore determined whether this proliferative response is also associated with PcG gene induction. Mice were subjected to ischemia/reperfusion renal injury, and total RNA was extracted from regenerating kidney tissues at times ranging from 24 hours to 4 weeks after ischemia. Transcript levels of EZH2, SUZ12, EED, and BMI-1 were measured, along with WT1, PAX2, CITED1, SIX2, by real-time PCR at consecutive time points (Fig. 2). The intact adult kidney was used as a reference sample. Of the nephric-progenitor genes, we detected predominantly significant early upregulation of SIX2 transcripts (24–48 hours postischemia). This was accompanied by rapid and sustained induction of BMI-1 mRNA (up to 2 weeks postischemia) and also of EED (24 and 48 hours) and SUZ12 (48 hours). Interestingly, EZH2 was not induced early after ischemia but rather silenced during renal regeneration.

Figure Figure 2..

Temporal expression patterns of nephric-progenitor (WT1, PAX2, CITED1, SIX2) and polycomb group (EZH2, EED, SUZ12, BMI-1) transcripts following IR injury to murine kidneys. Real-time polymerase chain reaction of kidneys at different time points after ischemia (24 hours, 48 hours, 1 W, 2 W, 1 M) was performed (n = 3 for each time point after ischemic insult). Normalization was performed against control β-actin expression, and RQ was calculated relative to intact adult kidney. Data were calculated as average ± SD. ***, p < .001; **, p < .01; *, p < .05 versus levels at 1 M. Abbreviations: h, hour; IR, ischemia/reperfusion; M, month; W, week.

PcG Gene Expression in Renal Tumorigenesis

Recent interrogation of microarray data of blastema-enriched progressive WT xenografts indicated EZH2 and BMI-1 to be among the most significantly upregulated genes compared with human fetal kidneys [13]. To more closely examine the PcG genes in WT tumorigenesis and progression, we analyzed the expression of EZH2, SUZ12, EED, and BMI-1 in fifth- to sixth-generation WT xenografts and independent primary WT (favorable histology) using real-time PCR (list of primer sets are summarized in Materials and Methods). qRT-PCR analysis in which we used pooled RNA from mid-gestational kidney as a reference sample showed differential expression of the nephric-progenitor genes: significant upregulation of PAX2, WT1, and SALL1; significant reduction in SIX2 transcript levels; and no significant change in CITED1. Of the PcG genes, EZH2, SUZ12, and BMI-1 transcripts were all significantly elevated in the WT xenografts, whereas EED, although elevated, did not achieve statistical significance (Fig. 3). The xenografts showed concomitant upregulation of the cancer stem cell marker CD133 [28]. Taking into account the role of the polycomb gene group in silencing of developmental pathways, we also examined transcripts of endothelial differentiation (VEGF, SCL/TAL1, αSMA, FLK1, TIE2) and found most of them to be significantly downregulated in progressive WT xenografts (Fig. 4). FLK1 showed a tendency to reduction in transcript levels; however, high variance of expression in the fresh WT samples precluded this difference from being significant. After observing that positive selection of the WT progenitor blastema in vivo is accompanied by upregulation of the PcG genes, we wanted to determine whether WT differentiation would produce opposing results. Culturing of WT cells obtained from three different Wilms' tumors (WT2, WT5, WT7) in conditions known to induce mesoderm differentiation (10% FBS supplemented with FGF/EGF) demonstrated that serial in vitro passages (passage [P] 1, P4, P7) of cells originating from the three tumors resulted in upregulation of α-smooth muscle actin, TIE2, and FLK1 mRNA, indicating maturation along the myofibroblastic/endothelial lineage, frequently observed in WT stroma (Fig. 4). This progressive loss of undifferentiated cells in the WT cultures was accompanied by downregulation of the polycomb genes analyzed, EZH2 and BMI-1, along with CD133 and the nephric-progenitor gene PAX2, further suggesting a role for the PcG genes in the malignant progenitor population.

Figure Figure 3..

Quantitative reverse transcription-polymerase chain reaction analysis of nephric-progenitor (WT1, PAX2, SALL1, CITED1, SIX2) (A), polycomb group (EZH2, EED, BMI-1, SUZ12) and AC133 (B), and endothelial-differentiation (SCL/TAL1, TIE2, VEGF, VEGFR2) (C) genes in five independent samples of WT Xn and favorable histology fresh WT. Normalization was performed against control β-actin expression, and RQ was calculated relative to pooled mid-gestation human kidneys. Data were calculated as average ± SD. ***, p < .001; **, p < .01; *, p < .05 versus WT. Abbreviations: WT, Wilms' tumors; Xn, xenografts.

Figure Figure 4..

Temporal expression patterns of renal (PAX2), polycomb (EZH2, BMI-1), AC133, and endothelial/ mesenchymal (TIE2, FLK1, αSMA) mRNA in differentiating WT cultures. Three experiments originating from WT cells obtained from three different fresh WTs (WT2, WT5, and WT7) are shown. Total RNA was isolated from the WT cells in P1, P4, and P7, and transcript levels of each gene were analyzed by quantitative reverse transcription-polymerase chain reaction. Abbreviations: P, passage; Ct, cycle threshold; WT, Wilms' tumors.

Methylation of the Nephric-Progenitor Genes in Renal Tumorigenesis

Having established cooperative induction of the polycomb gene group in WT tumorigenesis and taking into account their known role in epigenetic modification of developmental genes, including direct control of DNA methylation [29], we performed methylation analysis of the nephric-progenitor genes in progressive WT xenografts, primary WT, and normal fetal and adult kidneys. We used bisulfite modification of DNA and methylation-specific quantitative polymerase chain reaction. The investigated gene promoters were chosen from loci previously reported to be methylated in lung (SIX2 [30]), endometrial (PAX2 [31]), and ovarian (WT1 [32]) cancers. As shown in Figure 5A, PAX2, SIX2, and WT1 were found to be hypermethylated in the adult kidney (AK), coinciding with their silencing with the completion of nephrogenesis. In contrast to AK, all of the nephric-progenitor genes were all significantly hypomethylated in primary WT (Fig. 5A). Although the fetal kidney showed a tendency for hypomethylation of the nephric-progenitor genes compared with AK, methylation levels of WT1 were significantly reduced compared with AK (Fig. 5A), whereas PAX2 and SIX2 did not achieve statistical significance (p = .061, p = .099, respectively). Analysis of progressive WT xenografts (Xn) showed differential promoter methylation patterns; hypomethylation for WT1 and PAX2, with further relaxation of WT1; and relative hypermethylation of SIX2 in the xenografts (Fig. 5A), corresponding with their expression levels in the progressive WT compared with primary WT (described above).

Figure Figure 5..

Epigenetic changes in nephric-progenitor genes. (A): Real-time polymerase chain reaction analysis of promoter methylation of nephric-progenitor genes (WT1, PAX2, SIX2) in AK, FK, WT, and progressive WT Xn. Five independent samples were analyzed in each group. Data were calculated as average ± SD. Levels of methylation are shown for each group. For WT1, **, p < .01 versus FK, WT, and Xn; #, p < .05 versus WT; for PAX2, *, p < .05 versus WT and Xn; for SIX2, **, p < .01 versus WT; ##, p < .01 versus WT. (B): A hypothetical model of epigenetic changes in renal progenitor cells during renal development and tumorigenesis. Abbreviations: AK, adult kidneys; FK, fetal kidneys; MCN, methylated copy number; PcG, Polycomb group; WT, Wilms' tumors; Xn, xenografts.


In the present study, we analyzed the expression of the polycomb group genes in complimentary models involving renal progenitor cells, namely, renal development, regeneration, and tumorigenesis. The PcG genes were examined alongside a set of early genes that specifically mark the renal progenitor pool of the nephrogenic mesenchyme (WT1, PAX2, SALL1, CITED1, SIX2), collectively termed nephric-progenitor genes. The latter were all shown to be dynamically regulated during nephrogenesis and silenced in adulthood; reactivated, in part, during regeneration of the adult kidney (SIX2); and overexpressed in the embryonic kidney malignancy, WT, and progressive stem-like tumor xenografts derived from primary WT (WT1, PAX2, SALL1). Thus, alternations of the normal renal progenitor population, which contributes to nephron formation and regeneration, are indicated in WT oncogenesis.

PcG gene expression analysis revealed hierarchy of polycomb gene activation in controlled (development, repair) versus uncontrolled oncogenic versus processes. Accordingly, during nephrogenesis a role in the early progenitor population is suggested only for EZH2, whereas in renal regeneration a role is suggested primarily for BMI-1. In contrast, cooperative and sustained induction of the PcG genes (EZH2, SUZ12, BMI-1) was found in the transformed progenitor blastema and progression of WT. The strength of the in vivo progenitor selection model is exemplified by the demonstration that once WT cells undergo in vitro differentiation along the mesodermal/stromal lineage, both PcG genes, PAX2 and CD133, are decreased. Together, these data demonstrate that the polycomb group is likely to be involved in alterations of the renal progenitor population in WT oncogenesis.

Interestingly, mutual PcG gene activation and heightened expression of the cancer stem cell marker CD133 in the progressive stem-like tumors were associated with downregulation of regulatory genes characteristic of endothelial development and specifically of the nephric-progenitor gene SIX2. The Pax-Eya-Six pathway is a conserved regulatory network in organogenesis [33]. Recent data demonstrate that a Hox11-Eya1-Pax2 regulatory network is necessary for the expression of the downstream targets SIX2 and GDNF to promote mammalian kidney development [34]. Furthermore, Wellik et al. [7] have shown that following removal of all six functional alleles of the HOX11 paralogous group, the expression of the early patterning molecules PAX2 and WT1 is normal, whereas expression of SIX2 and GDNF was specifically affected. Thus, although multipotent renal progenitors initiate and accumulate in WT, loss of renal differentiation indicated by SIX2 downregulation in the xenografts is likely to occur with WT progression, diverting the progenitor phenotype away from the renal lineage into a more malignant one. This is accompanied by downregulation of angiogenesis-related genes, leading to possible uncoupling of WT stem cell expansion and angiogenesis. A similar phenomenon has recently been demonstrated in progressive xenografts of human brain tumors that adopt a highly infiltrative and stem-like phenotype and expand independently of angiogenesis [35]. From a practical point of view, cancer treatment strategies of WT recurrence aimed at angiogenic targets [36, 37] therefore might not suffice, and there is a need to pursue the invasive stem-like cancer cells.

WT has previously provided significant information regarding the epigenetic events, such as loss of imprinting, that lead to the development of cancer in general [38]. Alterations of the renal progenitor population in WT oncogenesis can be linked to polycomb activation through epigenetic modification of the nephric-progenitor genes [17]. We therefore determined the methylation status of these genes and show it to be a mirror image of their mRNA expression pattern. Accordingly, hypomethylation of WT1, PAX2, and especially SIX2 in WT coincides with initial accumulation of progenitors of the renal lineage, whereas hypermethylation of SIX2 (similar to adult kidney levels and opposed to WT1/PAX2) in the progressive WT xenografts corresponds to its silencing and indicates loss of renal differentiation. These results are in accordance with reports on pluripotent embryonic stem cells [19, 20] in which developmental pathways are silenced by PcG proteins, representing a private case for the renal lineage. Our observations can be explained by both direct control of DNA methylation by the PcG group [29] and indirect transcriptional repression of HOX genes, including HOX11 [39], the absence of which, as stated before, exclusively affects SIX2 expression in metanephric development [7].

Overall, our results fit into the model previously hypothesized by Feinberg et al. [40] on the epigenetic progenitor origin of human cancer. Accordingly, we propose that multipotent renal progenitor cells residing within the nephrogenic mesenchyme, which are under tight control during renal development via the nephric-progenitor genes, undergo sequential epigenetic alterations involving PcG activation and epigenetic modifications of the nephric-progenitor genes (especially SIX2) (Fig. 5B), leading to tumor initiation and progression.

Disclosure of Potential Conflicts of Interest

The authors indicate no potential conflicts of interest.


This work was supported by the “Talpiut” Sheba Career Development Award, the Morris Kahn Career Development Award, Schreiber Foundation Tel Aviv University, the Ministry of Health and Israel Science Foundation Bat-Sheva De Rothschild Physician-Scientist Grant Award, and the Israel Science Foundation Project Grant (to B.D.). G.R. holds the Djerassi Chair in Oncology at the Tel Aviv University. This work was conducted in partial fulfillment of the requirements for Ph.D. degree of Sally Metsuyanim, The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Israel. We thank Dr. Yoram Cohen for excellent advice in the DNA methylation studies.