SEARCH

SEARCH BY CITATION

Keywords:

  • Wilms tumor;
  • nephroblastoma;
  • retinoic acid;
  • microarray;
  • E2F

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Wilms tumor is the most frequent renal neoplasm in children, but our understanding of its genetic basis is still limited. We performed cDNA microarray experiments using 63 primary Wilms tumors with the aim of detecting new candidate genes associated with malignancy grade and tumor progression. All tumors had received preoperative chemotherapy as mandated by the SIOP protocol, which sets this study apart from related approaches in the Unites States that are based on untreated samples. The stratification of expression data according to clinical criteria allowed a rather clear distinction between different subsets of Wilms tumors. Clear-cut differences in expression patterns were discovered between relapse-free as opposed to relapsed tumors and tumors with intermediate risk as opposed to high risk histology. Several differentially expressed genes, e.g.TRIM22, CENPF, MYCN, CTGF, RARRES3 and EZH2, were associated with Wilms tumor progression. For a subset of differentially expressed genes, microarray data were confirmed by real-time RT-PCR on the original set of tumors. Interestingly, we found the retinoic acid pathway to be deregulated at different levels in advanced tumors suggesting that treatment of these tumors with retinoic acid may represent a promising novel therapeutic approach. © 2005 Wiley-Liss, Inc.

Wilms tumor or nephroblastoma is one of the most common solid tumors in childhood, with an incidence of 1:10,000 live births. It arises from embryonic kidney cells and most frequently presents as a unilateral (95%) and sporadic (98%) tumor. According to the SIOP (International Society of Pediatric Oncology) protocol, the treatment of children over the age of 6 months is initiated with preoperative chemotherapy comprising actinomycin-D and vincristin for 4 weeks. Postoperative treatment after tumor nephrectomy depends on tumor histology, detection of metastases and local tumor stage. The overall 5-year-survival rate is approximately 90%.1

For most Wilms tumors, the molecular pathogenesis is still unclear. Mutations in WT1, a tumor suppressor gene at 11p132 and in β-catenin (CTNNB1),3 a component of the Wnt signaling pathway, are known to be involved in Wilms tumorigenesis. Both mutations are frequently associated with each other,4 but account for only 10–15% of Wilms tumors.5 However, they do not have a predictive value. Furthermore, p53 mutations frequently occur in anaplastic Wilms tumors, which belong to the group of tumors with high-risk histology.6

Allele loss studies and comparative genomic hybridization have identified several chromosomal regions that are frequently deleted. Alterations in some of these chromosomal regions are associated with poor outcome, e.g. chromosomes 1 and 16q.7 Nevertheless, it has not been possible up to now to delineate individual genes within these regions that initiate or promote Wilms tumor growth.

With the aim of detecting novel candidate genes important for malignancy grade and tumor progression, we performed cDNA microarray experiments based on 11,500 clones, using 63 primary Wilms tumors treated with preoperative chemotherapy, the largest cohort analyzed to date.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Wilms tumors

Primary tumor tissue was initially obtained from 77 Wilms tumors from the German SIOP/GPOH 93-01 Wilms tumor study. Detailed clinical data were collected for all samples. Sixty-seven of 77 tumors received preoperative chemotherapy as mandated by the SIOP protocol. To exclude expression differences due to the therapeutic approach (preoperative chemotherapy vs. primary operation), we included only those 67 tumors that were preoperatively treated in our final analysis. In most cases, histological diagnosis was available from both local and reference pathologists. Informed consent for biological studies had been obtained before surgery.

Isolation and amplification of RNA

Total RNA was isolated from frozen tumor tissue, using the RNA/DNA Midi Kit (Qiagen, Germany) or Trizol reagent (Life Technologies). RNA was treated with 30 U DNase, again purified (RNeasy Kit, Qiagen) and stored in water at −80°C. The reference sample for array hybridizations consists of an RNA pool from 14 Wilms tumors with different clinical data.

Since the amount of extracted RNA was not sufficient for array hybridization in most cases, we amplified mRNA from all total RNA samples using the MessageAmp aRNA Kit (Ambion, Huntingdon, UK). Briefly, the procedure consists of reverse transcription with an oligo(dT) primer bearing a T7 promoter and in vitro transcription of the resulting cDNA with T7 RNA polymerase to generate multiple copies of each mRNA. After one round of in vitro transcription starting with 1 μg of total RNA, the yield of amplified mRNA ranged between 10 and 50 μg.

cDNA labeling

Hybridization probes were generated by indirect labeling with Cy3 and Cy5 dyes, using the CyScribe cDNA Post Labeling Kit (Amersham Biosciences Europe, Freiburg, Germany). Three micrograms of amplified tumor and reference RNA, respectively, were reverse transcribed with nonamer primers, incorporating modified amino-allyl-dUTP. The RNA template was then degraded with 2 μl NaOH (2.5 N) at 37°C for 15 min, followed by neutralization with 10 μl HEPES (2 M). The cDNA was purified (PCR Purification Kit, Qiagen, Hilden, Germany) and labeled with Cy5 (tumor samples) and Cy3 (reference samples). Labeled tumor and reference samples were then combined and again purified to remove unbound Cy dyes.

Microarray experiments

Microarrays were generated as previously described8 using a GMS 417 arrayer (MWG Biotech, Ebersberg, Germany). The chips contained 11,500 clones from the human sequence-verified UniGene cDNA sets gf200, gf201 and gf202 (http://www.resgen.com). Each experiment was performed as sandwich hybridization, i.e. instead of a cover slip, a second microarray slide was used. This provides a replicated measurement for each hybridization, which can be used for quality control and for reduction of technical variability.

Arrays were prehybridized for 30 min at 55°C with a blocking solution containing 1% bovine serum albumin, 3× SSC and 0.1% SDS. After washing with water, arrays were dried by brief centrifugation. To reduce unspecific background signals, Cot1 DNA and polyA were added to the labeled cDNAs. The addition of 10 μl SSC (20×) and 4 μl SDS (2%) resulted in a final volume of 100 μl. Hybridization samples were boiled for 2 min immediately before sandwich hybridization. After incubation in a humid chamber at 55°C for 16 hr, arrays were separated again and 4 washing steps were carried out, twice with 0.1× SSC/0.1% SDS and twice with 0.1× SSC. Finally, the arrays were washed in water and dried by centrifugation.

Standardization and quality control

Arrays were scanned separately for both Cy dyes, using a GMS 418 microarray scanner (MWG Biotech, Ebersberg, Germany). Spot intensities were extracted from a scanned image with ImaGene 3.0 software (BioDiscovery Inc., Marina Del Rey, USA). Software parameters such as “signal range” or “spot detection threshold” were optimized for maximum reproducibility prior to image analysis (not shown). For each spot, median signal and background intensities were obtained for both channels. To account for spot differences, the background was corrected and the ratio of the two channels was calculated and log2-transformed.

In order to balance fluorescence intensities for the two dyes and to allow for comparison of expression levels across experiments, the raw data were standardized. The first step consisted of a spatial- and intensity-dependent standardization similar to that in Ref.9 to correct for inherent bias on each chip. As each gene was measured twice in the sandwich hybridization, mean log-ratios M were calculated from replicates. When gene replicates differed more than the maximum of 3-fold and 75% of the calculated average log-ratio or when background intensity was higher than signal intensity, the spot was excluded from that array. In total, 1.3% (∼150 spots per chip) of all spots had to be flagged and filtered out.

Sandwich hybridizations were highly reproducible. Only 4 hybridizations (5%) were excluded because of inconsistencies between replicated measurements. In 1 tumor sample, one of the sandwich slides showed strong spatial effects; therefore only 1 measurement from this tumor was used in the final data analysis. Ten tumors were excluded from further analysis because they were primarily operated and did not receive preoperative chemotherapy as the majority of tumors did. The final data matrix consisted of 11,552 standardized gene expression measurements (log2-ratios) from 63 individual tumors.

Statistical methods

To compare expression profiles between two independent groups, the 2-sample Welch t-statistic was used for ranking the genes. To account for multiple testing, we estimated the false discovery rate (FDR), a procedure introduced by Benjamini and Hochberg.10 The FDR is based on permutation p-values and it estimates the proportion of false discoveries within a given set of genes. As a second method, to correct for multiple testing, we calculated the familywise error rate (FWER). This method has already been applied to microarrays by Callow et al.11 It represents a step-down permutation algorithm12 that takes the correlation of the variables into account. The procedure does not rely upon a normality assumption. For example, evaluating all genes with an adjusted p-value of less than 5% means that the probability of having one or more false positive genes within the whole list is less than 5%. Because of the small sample size for most groups, in combination with the pronounced heterogeneity typically seen in Wilms tumors, the usual selection criteria (FWER or FDR < 0.05) were inappropriate to show differences in marker list quality between the different clinical questions. Therefore, higher adjusted p-values were used as cut-off criteria to generate lists of potentially interesting genes and the quality of the resulting lists was then further evaluated using FDR.

Both methods are implemented in the multtest package of the Bioconductor project (http://www.bioconductor.org).13 The open source project Bioconductor provides a collection of statistical tools developed for the analysis of microarray data.

To avoid losing interesting candidate genes with large fold change differences, we generated additional lists with candidate genes in an exploratory manner by selecting only genes with a fold change difference of at least 2 and an absolute value of the t-statistic of more than 1.96. Because of the ad hoc selection procedure, p-values or FDRs were not calculated for these lists.

Note that the primary goal of this study was to screen for differentially expressed genes for further analysis. FDRs and adjusted p-values were therefore only composed for descriptive purposes, not to prove statistical significance. The cut off used for comparison of the different clinical criteria was derived from estimated adjusted p-values (FWER).

Cluster analysis

Prior to cluster analysis, the expression profile of each gene was centered by subtracting the mean observed value. Average linkage hierarchical clustering was then performed for genes as well as for chips, using the Euclidean distance metric as implemented in the program Genesis.14

Analysis of over-represented genes and themes within gene lists by EASE

EASE (Expression Analysis Systematic Explorer) is a software application that automates the process of biological theme determination. EASE calculates over-representation of certain genes in a given list of deregulated genes with respect to the total number of genes assayed. Statistical measure of over-representation is the “EASE score,” which was shown to reliably identify most robust categories of differentially regulated groups of genes (EASE score < 0.05).15

Real time RT-PCR

Real-time PCR reactions were performed on all original Wilms tumor samples. About 0.5 μg of the amplified RNA was used for cDNA synthesis with the RevertAid First Strand cDNA Synthesis Kit, respectively (MBI Fermentas, St. Leon-Rot, Germany). For RNA degradation, each reaction mix was then incubated with 1.25 U of RNaseH at 37°C for 30 min. Finally, water was added to a final volume of 200 μl.

Real-time PCR was conducted in the iCycler (Bio-Rad, Munich, Germany). The standard PCR reactions (22 μl) contained 2 μl cDNA, 0.13 pmol FITC (Bio-Rad, Munich, Germany) and 0.75 μl SybrGreen (1:20,000 diluted, Eurogentec, Seraing, Belgium) for detection. PCR conditions were 40 cycles of denaturation at 95°C for 15 sec, followed by annealing and elongation at 60°C for 1 min. Finally, melt curves were established (50–93°C) in order to check the accuracy of the amplification. Real-time PCR reactions with the housekeeping gene HPRT were performed with the aim of equalizing concentration differences between the tumor samples. All primer sequences are available in the online supplementary Table 1.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Evaluation of expression data according to clinical criteria

A collection of Wilms tumors from the SIOP/GPOH 93-01 Wilms tumor study was used to isolate RNA and to hybridize cDNA microarrays with 2-color labeling as described. Microarray hybridizations of primarily operated tumors were excluded from further analysis. Expression data of 63 Wilms tumors that had received preoperative chemotherapy were finally evaluated for the following clinical criteria: metastasis, response to chemotherapy, histological risk grade (malignancy), relapse and survival. All these clinical criteria allowed stratification into two separate groups. The respective criteria, the number of tumors categorized in those groups and the number of genes selected as differentially expressed by FWER for every clinical criterion are summarized in Table I. Complete raw data are available as MIAMExpress submission A-MEXP-221 (http://www.ebi.ac.uk/miamexpress). As expected, the resulting differences in expression were not statistically significant for several of the clinical criteria tested, which may be due to heterogeneity within tumor subpopulations, small sample sizes for individual groups or the fact that some of the clinically observed differences just have no genetic basis. Strongest differences in expression were detected for different histological risk groups and for relapse. However, the gene lists generated contain a number of potentially interesting genes that can now be used for further analysis.

Table I. Stratification of Wilms Tumors and Number of Differentially Expressed Genes Selected Based on T-Statistic Ranking and False Discovery Estimates
Clinical criteriaNo. of Wilms tumorsNo. of differentially expressed genes with adjusted p-value <0.9 (FWER)False discovery rate (%)
Group 1Group 2
  • 1

    As nearly all relapses occur within three years after therapy, this timeframe was used for definition of relapse-free tumors.

  • 2

    The same group of relapse-free tumors was used as for comparison of relapsed tumors.

  • 3

    Tumor reduction of more than 50% after chemotherapy was defined as good response, tumor reduction of less than 50% or increase of tumor volume was defined as poor response.

Histological risk grade (1 = intermediate, 2 = high)5282027.5
Relapse1 (1 = no (>3 yr), 2 = yes)27102225.6
Survival2 (1 = relapse-free (>3 yr), 2 = death)2751231.5
Metastasis (1 = no, 2 = yes)5412750.5
Response to chemotherapy3 (1 = good, 2 = poor)3819667.5

Cluster analysis

On the basis of genes selected by lowest adjusted p-values for histological risk grade (intermediate/high risk) and for relapse (yes/no), we performed hierarchical clustering to visualize the discriminative potential of these genes. The Euclidean distance metric was used after centering the expression profile of each gene by subtracting the mean value.

In Figure 1a, intermediate risk histology is compared with high risk histology, using 74 differentially expressed genes with lowest adjusted p-values. There is a rather good separation between these two groups. Even within the intermediate risk group tumors with predominantly regressive changes are well separated. On the other hand, additional subclusters that are visually apparent do not correlate with either the indicated histological subtypes or with other clinical variables. Figure 1b shows the cluster analysis of relapsed tumors in comparison to all tumors with a relapse-free follow-up of at least 3 years (thus reducing the number of included tumors). In order to evaluate the potential predictive power of microarray analysis, we analyzed which tumors with high risk histology had relapsed later. High risk histology, i.e. blastemal predominant or anaplastic subtype, is the currently best criterion to predict outcome in affected children. Only 4 of the 10 relapsed Wilms tumors had been diagnosed with a high risk histotype. In contrast, the other 6 Wilms tumors that actually relapsed belong to the intermediate risk group, but they are recognizable by similar differential gene expression.

thumbnail image

Figure 1. Hierarchical cluster analysis of intermediate vs. high risk and relapse-free vs. relapsed Wilms tumors. Individual tumors are represented by vertical columns with additional histological or clinical information denoted by black squares. For each of the genes accession numbers and gene symbols (where available) are listed on the right. Green squares designate downregulation of a gene in a given tumor; red designates upregulation relative to the mean. Gray spots indicate missing or poorly reproducible data between replicated measurements of a given tumor. (a) Cluster analysis of 52 intermediate risk (predominantly regressive, stromal and epithelial as well as triphasic tumors) and 8 high-risk (anaplastic and blastemal) tumors using 74 differentially expressed genes with lowest adjusted p-values (FWER). Numbers in front of ws (Wilms tumor) numbers indicate intermediate-risk tumors (1) and high-risk tumors (2) Histological subtypes are given behind ws numbers (5 = triphasic, 4, 6 = epithelial, 7 = blastemal, 8 = stromal, 9 = regressive, 10 = anaplastic) and are additionally visualized as cluster below ws numbers. Additional subgroups of tumors become apparent especially among the intermediate risk group. (b) For the cluster of relapse-free (>3 years, group 1 as indicated before ws number) versus relapsed (group 2 as indicated before ws number) Wilms tumors 77 differentially expressed genes with lowest adjusted p-values (FWER) were used. The total number of tumors is reduced since tumors with a follow-up of less than 3 years had to be excluded. Note that only 4 of the 10 relapses occurred in Wilms tumors with histologically defined high risk (ws89, ws128, ws144, ws199).

Download figure to PowerPoint

“Top candidate” genes

With the intention of finding additional candidate genes with large fold change differences that may be more readily scorable in a diagnostic setting, we established further gene lists on the basis of combined selection criteria considering t-statistics and fold change differences. In this exploratory approach, genes with expression differences of at least 2-fold and an absolute t-statistics value of more than 1.96 were selected. Because of the ad hoc selection procedure, p-values and FDRs were not calculated for these lists. Since the comparison of relapsed and relapse-free tumors is of particular interest, the first 40 genes with highest t-statistic value are summarized in Table II. This table also contains the 10 “top candidates” of further lists for the histological risk, metastasis and survival. Complete lists are available as online supplementary Table 2.

Table II. “Top Candidate Genes” for Different Clinical Criteria with Expression Differences of at Least 2-Fold and t-Statistic Values of more than 1.96
t + fct-StatFold change2GenBank accessionGene symbol
  • a

    For relapsed tumors, the 40 genes with highest t-statistic values are shown. Genes whose deregulation was validated by real time qPCR on the original set of tumors (see Fig. 2) are given in boldface. For the clinical criteria high risk histology, survival and metastasis, the first 10 genes are given. Complete lists are available as online supplementary Table 2.

  • 2

    The regulation refers to the group in italic of tumors in column 1.

Relapse (yes/no)5.192.40AA488336DRG1
5.112.02H09614CTPS
5.062.34AA480995MTHFD2
4.792.23T59641PBX2
4.632.11AA449762ILF3
−4.47−2.15N70734TNNT2
4.462.14W96224NDUFV3
−3.84−2.22T66180THRA
−3.81−2.82AA047039EIF1AY
−3.81−2.33AA083478TRIM22
3.712.13AA598610MEST
−3.70−2.49N51278CX3CR1
3.652.42N53057CHEK1
3.582.02R00884DHFR
3.542.13H56918EIF4A1
−3.53−2.14AA099369CUL5
3.502.35H59204CDC6
−3.43−2.40AA046700FBXO32
−3.42−2.47AA046679 
3.402.22AA430744EZH2
−3.38−2.23H11453 
3.362.17N70010CDCA5
−3.27−2.03AA192166 
−3.27−2.66T70352OLFML2A
3.232.64AA446462BUB1
−3.22−2.05AA458965NK4
−3.22−2.34AA427947 
−3.20−2.68R34224MGC16121
3.172.40R06900RAMP
3.152.32AA968443 
−3.14−2.15R91916CXCR6
−3.11−2.17W84612YPEL3
3.102.21H25560DGAT2
−3.09−2.35AA291749ESR1
−3.02−2.00W47350RARRES3
3.012.34T66935DKFZp762E1312
2.962.70AA701455CENPF
−2.92−2.64T53298IGFBP7
−2.92−2.31H73914LDB2
−2.89−2.19H0911113CDNA73
Histological risk (high/standard)−6.19−2.21R01167FLJ23467
−6.15−2.26AA126862LOC92689
−5.89−2.75N30615 
−5.79−3.28AA046700FBXO32
−5.75−2.77AA464711C3AR1
−5.61−2.63AA099369CUL5
−5.45−2.52R42713IDI2
−5.38−2.68N71028MS4A6A
−5.21−2.22AA035455BACH
−5.15−2.29AA701545RNASE6
Survival (no/yes)7.962.79AA480995MTHFD2
7.482.88AA488336DRG1
6.622.62T59641PBX2
−6.22−3.45N30615 
5.842.04AA186605SLC5A6
5.602.71N53057CHEK1
−5.13−2.57N30436 
5.122.36AA278749NCBP1
5.112.14H09614CTPS
−5.03−2.16W47179CTSB
Metastasis (yes/no)4.682.32AA489246ST14
3.402.89AA401441BF
3.292.95AA664406C4A
3.262.16AA129777SLC16A3
3.122.05AA424584LTBP2
2.902.06AA430625DPYD
2.772.03N93686 
2.682.45N30986 
−2.61−2.22W68220KIAA0101
2.532.89AA485867MARCO
thumbnail image

Figure 2. Comparison of the magnitude of gene expression changes between relapsed as opposed to nonrelapsed Wilms tumors as measured by microarray analysis and qPCR on the same set of tumors. The genes selected were differentially expressed by microarray analysis and were then assessed by qPCR for the same stratified tumor groups. Bars represent median values in each case. Genes are shown in their order of decreasing t-statistic values (cf. Table 2 and online supplementary Table 2).

Download figure to PowerPoint

Analysis of over-represented genes in relapsed and high risk tumors

EASE was applied to calculate over-representation of certain genes in our two lists of deregulated genes in relapsed and high risk tumors with respect to the total number of genes assayed. Genes involved in cell cycle progression and mitosis are significantly over-represented in relapsed tumors. In contrast, we found a significant over-representation of defense and immune response genes in high risk tumors. Complete lists of over-represented genes of different categories (EASE score < 0.05) are provided as supplementary Table 3.

Validation of array data by qPCR

A subset of array data was validated by real-time or quantitative RT-PCR (qPCR) on the original set of tumors. We designed primers for 9 genes with diverse biological functions and strong fold changes that had been selected in the relapsed vs. relapse-free tumor comparison. Expression levels were standardized with the housekeeping gene HPRT. Median cycle differences were calculated for the same groups of relapsed and relapse-free tumors that had been used for array data stratification (Fig. 2). Genes are listed according to their t-statistic values, i.e. in the order of Table 2 and online supplementary Table 2. Three of the validated genes are also included in Figure 1b (CTPS, TRIM22, THRA).

There was an overall good or even excellent correspondence between qPCR median values and array data. Even at the level of individual tumor samples, there was a good agreement between array and qPCR data (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We performed cDNA microarray experiments in 77 primary Wilms tumor samples and selected putative candidate genes according to different clinical stratifications. The final analysis contained 63 Wilms tumors all of which had undergone preoperative chemotherapy as mandated by the SIOP protocol.

Malignancy grade

Cluster analysis for histological risk grade clearly distinguishes between Wilms tumors of intermediate and high risk histology (Fig. 1a). The stratification was performed according to the revised Stockholm–Working-Classification.16 It should be stressed that therapeutic regimens in the European SIOP study mandate preoperative chemotherapy in most cases, which is different from the American NWTS study. It has been well documented that this leads to a quite different distribution of histological subtypes of Wilms tumor and concomitantly altered clinical course.17 Thus, it will remain open as to what extent the expression changes found in our patient cohort are different from those in the North American study.

The high risk tumor group comprises four predominantly blastemal and four anaplastic tumors. Despite histological differences between these high risk tumors they seem to have a common molecular basis that sets them apart from the intermediate risk group. The intermediate risk group comprises 52 Wilms tumors. Half of these were “classic” triphasic Wilms tumors presenting with a mixture of epithelial, stromal and blastemal cell types. The other half consists of predominantly stromal or epithelial and of regressive tumors. A partial clustering of histological subtypes can be seen, especially for regressive Wilms tumors in the left third of the cluster picture. Nevertheless, the separation of intermediate risk histologies into several subclusters lacks a clear clinicomorphologic correlate. This suggests that there may be additional underlying factors that are not covered by current diagnostic means. Part of this may also be due to the fact that the tumors generally display histological heterogeneity to different degrees, and thus, rather represent a continuum of histological subtypes with one often being predominant.

Relapse and survival

The cluster comparison between relapse-free tumors and those that later relapsed is based on 77 genes with lowest adjusted p-values. These genes clearly distinguished between the two groups (Fig. 1b). Only tumors with a relapse-free follow-up of at least 3 years were included in the first group. Expression data for survival overlap with those for relapse because most deaths in children with Wilms tumor occur as a consequence of relapse. More than half of the genes selected for survival were concurrently selected for relapsed tumors. In relapsed Wilms tumors, more genes with a lower FDR were selected due to the larger tumor cohort available in comparison to the stratification for survival.

An association between gene expression and prediction of outcome has recently been described in several other human cancers.18 Studies on breast cancer have delivered the most extensive and informative prognostic profiles.19 These studies suggest that the predictive power of microarray approaches may be even greater than that of other known predictive markers. However, the overall positive prediction of tumors rarely exceeded 60%. Here, we show that in Wilms tumors expression analysis may also help to predict tumor recurrence and outcome in future. Up to now, the risk for relapse in Wilms tumors is estimated from clinical and histological parameters and this stratifies the extent of postoperative therapy.16 Only 4 of the 10 relapsed tumors had high risk histology, whereas the majority of relapses occurred in tumors with intermediate risk histology. The stratification based on microarray data for relapsed vs. relapse-free Wilms tumors (Fig. 1b) points at a potential predictive value of expression data, in addition to histology. At this point, the number of analyzed Wilms tumors is definitely too small to calculate overall prediction rates. Prediction studies need both larger training and validation groups. However, expression profiling appears to deliver additional prognostic information, which may help to guide future research and to further optimize treatment of affected children.

Evaluation of other clinical features

Interestingly, cluster analysis for the stratification of tumors with and without metastasis was not unequivocal. The propensity of tumor cells to spread and colonize more distant sites is not easily identifiable in our tumor set. It may well be that the metastatic potential of individual subclones of tumor cells may get lost in a global analysis that is based on a comparatively large piece of tumor tissue. This is supported by the fact that in the literature metastasis-associated gene expression profiles are only available for comparisons between cell lines with different metastatic potential or genuine comparisons of primary tumor and metastases.

For the stratification of the clinical criterion response to chemotherapy, almost no expression difference could be detected. Response to preoperative chemotherapy was determined by measuring pre- and postchemotherapeutic tumor volumes. Tumor reduction of more than 50% was defined as good response (38 children), whereas tumor reduction of less than 50% or increase of tumor volume despite preoperative chemotherapy was defined as poor response (19 children). The lack of clear differentiating expression patterns was somewhat surprising in this case. Strongly hit tumors with massive apoptosis and large numbers of secondary cell types should be quite different from tumors that continue to grow almost unaffected by chemotherapy. However, it is possible that response to chemotherapy is controlled by relatively few genes that may not be represented on our microarrays.

Candidate genes for Wilms tumor development and progression

To date, 6 publications reported on microarray expression screening in Wilms tumors.20, 21, 22, 23, 24, 25 The aim of several of these studies, most of which are based on rather small numbers of Wilms tumor samples, was to identify genes differentially regulated in Wilms tumors compared to normal kidney tissue. Two recent studies sought to correlate expression profiles to clinical features. The study by Williams et al.22 compares 27 Wilms tumors with favorable histology, half of which later relapsed. The authors identified genes with potential biological interest, but state that their data are not useful for prediction. Probably due to the fact that they used prechemotherapy samples, there is only little overlap in genes associated with relapse compared to our study. The second recent study by Li et al.21 compares anaplastic tumors to tumors with favorable histology and additionally contains an overall comparison of Wilms tumors to normal kidney. Compared to our study, several genes were concordantly identified: CENPF, CCNA1, CDC2 (anaplastic vs. favorable histology) and EZH2 (Wilms tumor vs. fetal kidney). The first group of genes deregulated in advanced Wilms tumors is involved in cell cycle progression. The only difference between samples included in our study and that of Li et al. is the therapeutic approach: Wilms tumors analyzed by Li et al. were all primarily operated according to the American NWTS protocol, whereas tumor samples included in our study were derived from children who were treated with preoperative chemotherapy according to the European SIOP protocol. Since these genes were concordantly deregulated in both studies, they seem to be involved in tumor progression independent of the therapeutic approach.

Compared to previous expression studies on Wilms tumors, we analyzed a much larger set of 63 preoperatively treated tumors and considered several clinical criteria for the evaluation of expression data. The aim of our study was to assess relative expression differences between these tumors and to search for genes affecting tumor progression. To investigate whether certain groups of genes may be over-represented among differentially regulated genes, we performed EASE analysis of different gene lists. Among genes that were deregulated in relapsed tumors, those involved in cell cycle progression and mitosis were significantly over-represented. BUB1, CENPF, CENPE, CDC6, CKS2, CUL5, ESR1, UBE2C, MAD2L1, MAD2, CHEK1 and STK6 are examples of significantly (EASE score < 0.05) over-represented mitosis-associated genes in relapsed Wilms tumors. In tumors with high-risk histology that were compared to standard-risk tumors, we found a slightly different picture with significant over-representation of defense and immune response genes, e.g.C3AR1, TREM2, CSF1R, BF, IL16, HLA-C, IL1R1, and genes coding for extracellular region proteins, e.g.COL15A1, ELN, MMP2, NID2, SFRP4 and OSF2. Although high-risk histology is the best predictor for later relapse, it is intriguing that these two criteria appear to be characterized by rather different sets of biological processes, i.e. mitosis vs. defense.

Previous studies have proposed further genes with prognostic relevance in Wilms tumors, e.g. members of the INK4 family,26TERT,27VEGF and its receptor FLT1,28, 29DKK1,30TrkBfull,31TGFA and EGFR.32 In our study, however, no significant regulation of these genes could be found in any stratification of clinical features in Wilms tumors. The expression differences reported before may be due to the small number of tumors included in most of these studies. Furthermore, it may generally be rather difficult to predict outcome in Wilms tumors by the assessment of single genes. The range of single gene expression values measured by microarray and qPCR approaches in relapsed tumors also supports this assumption (data not shown). We therefore speculate that only a combination of expression data from a set of different genes may represent a reliable new approach for prediction of outcome in Wilms tumors.

The retinoic acid and E2F pathways as potential new targets

Interestingly, our analysis identified several retinol-related genes belonging to the retinoic acid receptor responder family. RARRES2 was found to be 2.2-fold downregulated in high malignant tumors, while RARRES3 was 2-fold downregulated in relapsed tumors and in tumors that led to death of the affected children.

Another retinoic acid induced gene, CTGF (connective tissue growth factor; also known as IGFBP8), has been identified in a screen for WT1-induced genes.33 Compared to normal kidney tissue, CTGF was overexpressed in a subset of Wilms tumors. We found CTGF to be downregulated in the evaluation of relapse and survival (2.9- and 3.8-fold downregulation, respectively), which represent important criteria for the prediction of prognosis. It is conceivable that CTGF is activated in nephrogenesis and early tumorigenesis, while its expression decreases with tumor progression.

Additionally, NK4 (natural killer cell transcript 4), RAMP (RA-regulated nuclear matrix-associated protein) and ENPP2 (Autotaxin) represent genes regulated by retinoic acid,34, 35 and they were also found to be strongly deregulated in relapsed Wilms tumors. Therefore, it appears that the retinoic acid pathway can be deregulated at different levels. We have since tested whether retinoic acid could be employed as a novel therapeutic agent in Wilms tumors by exposing cultured tumor cells.36 Our data suggest that retinoic acid may represent a novel therapeutic approach to treat tumors with evidence for impaired retinoic acid signaling.

Another pathway potentially involved in Wilms tumor progression is the RB-E2F pathway. Members of the E2F transcription factor family are important targets of the retinoblastoma (RB) pathway and they regulate transcription of a number of genes that control cell cycle progression.37EZH2 (enhancer of zeste homolog 2) is a direct target of E2F transcription factors.38 Overexpression of EZH2 and involvement in tumor progression was shown in different human cancers, e.g. breast cancer.38 Takahashi et al.20 and Li et al.21 found EZH2 overexpressed in Wilms tumors with favorable histology compared to normal kidney tissue. Our results confirm the involvement of EZH2 in Wilms tumorigenesis and additionally imply that EZH2 acts as a tumor progression factor since its expression is upregulated in recurrent Wilms tumors. In addition to EZH2, several other genes are known to be controlled by E2F39 and 4 of these were highly overexpressed in recurrent vs. nonrecurrent Wilms tumors (CDC6, CDC2, MYCN, DHFR). These results suggest an important role of the RB-E2F pathway in Wilms tumor formation and progression.

In summary, we identified several candidate genes involved in Wilms tumor development and progression. Since expression profiling harbors the potential to predict prognosis, which has already been shown in several other cancer types, the genes identified in relapsed Wilms tumors are of special interest. These differentially expressed candidate genes can now be screened in extended tumor cohorts to corroborate our findings. Our goal is to select a minimal set of genes whose expression can be used to identify patients at risk for relapse who may benefit from intensified therapeutic regimens or enhanced surveillance. Such screening procedures could be implemented in future study protocols as prospective prognostic studies. From a basic science perspective it will be very interesting to identify gene expression in individual cell types by mRNA in situ hybridization or antibody staining to better characterize their individual contribution to Wilms tumor development.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors would like to thank Angelika Filmer, Barbara Klamt and Sabine Erhardt for excellent technical assistance and Susan Nieschlag for language editing. This work is part of the PhD thesis of Birgit Zirn. This work was funded by the Graduiertenkolleg 639, the BMBF “Kompetenznetz für Pädiatrische Onkologie und Hämatologie”, the Sander Stiftung and parent donations. We gratefully acknowledge the efforts of all clinicians, nurses and patients who made it possible to collect the tumor specimens.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information
  • 1
    Graf N, Tournade MF, de Kraker J. The role of preoperative chemotherapy in the management of Wilms' tumor. The SIOP studies. International Society of Pediatric Oncology. Urol Clin North Am 2000; 27: 44354.
  • 2
    Gessler M, Poustka A, Cavenee W, Neve RL, Orkin SH, Bruns GA. Homozygous deletion in Wilms tumours of a zinc-finger gene identified by chromosome jumping. Nature 1990; 343: 7748.
  • 3
    Koesters R, Ridder R, Kopp-Schneider A, Betts D, Adams V, Niggli F, Briner J, von Knebel Doeberitz M. Mutational activation of the β-catenin proto-oncogene is a common event in the development of Wilms' tumors. Cancer Res 1999; 59: 38802.
  • 4
    Maiti S, Alam R, Amos CI, Huff V. Frequent association of β-catenin and WT1 mutations in Wilms tumors. Cancer Res 2000; 60: 628892.
  • 5
    Gessler M, Konig A, Arden K, Grundy P, Orkin S, Sallan S, Peters C, Ruyle S, Mandell J, Li F, et al. Infrequent mutation of the WT1 gene in 77 Wilms' Tumors. Hum Mutat 1994; 3: 21222.
  • 6
    Hill DA, Shear TD, Liu T, Billups CA, Singh PK, Dome JS. Clinical and biologic significance of nuclear unrest in Wilms tumor. Cancer 2003; 97: 231826.
  • 7
    Grundy RG, Pritchard J, Scambler P, Cowell JK. Loss of heterozygosity on chromosome 16 in sporadic Wilms' tumour. Br J Cancer 1998; 78: 11817.
  • 8
    Berwanger B, Hartmann O, Bergmann E, Bernard S, Nielsen D, Krause M, Kartal A, Flynn D, Wiedemeyer R, Schwab M, Schafer H, Christiansen H, et al. Loss of a FYN-regulated differentiation and growth arrest pathway in advanced stage neuroblastoma. Cancer Cell 2002; 2: 37786.
  • 9
    Yang YH, Dudoit S, Luu P. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 2002; 30: 15.
  • 10
    Benjamini Y, Hochberg YH. Controlling the false discovery rate: a practical and powerful approach for multiple testing. J R Statist Soc B 1995; 57: 289300.
  • 11
    Callow MJ, Dudoit S, Gong EL, Speed TP, Rubin EM. Microarray expression profiling identifies genes with altered expression in HDL-deficient mice. Genome Res 2000; 10: 20229.
  • 12
    Westfall PH, Young SS. Resampling-based multiple testing: examples and methods for p-value adjustment. New York: Wiley, 1993.
  • 13
    Ross I, Gentleman RR. A language for data analysis and graphics. J Comput Graphical Stat 1996; 5: 299314.
  • 14
    Sturn A, Quackenbush J, Trajanoski Z. Genesis: cluster analysis of microarray data. Bioinformatics 2002; 18: 2078.
  • 15
    Hosack DA, Dennis G,Jr., Sherman BT, Lane HC, Lempicki RA. Identifying biological themes within lists of genes with EASE. Genome Biol 2003; 4: R70.
  • 16
    Vujanic GM, Sandstedt B, Harms D, Kelsey A, Leuschner I, de Kraker J. Revised International Society of Paediatric Oncology (SIOP) working classification of renal tumors of childhood. Med Pediatr Oncol 2002; 38: 7982.
  • 17
    Weirich A, Leuschner I, Harms D, Vujanic GM, Troger J, Abel U, Graf N, Schmidt D, Ludwig R, Voute PA. Clinical impact of histologic subtypes in localized non-anaplastic nephroblastoma treated according to the trial and study SIOP-9/GPOH. Ann Oncol 2001; 12: 3119.
  • 18
    Russo G, Zegar C, Giordano A. Advantages and limitations of microarray technology in human cancer. Oncogene 2003; 22: 6497507.
  • 19
    van't Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, Mao M, Peterse HL, van der Kooy K, Marton MJ, Witteveen AT, Schreiber GJ, Kerkhoven RM, et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 2002; 415: 5306.
  • 20
    Takahashi M, Yang XJ, Lavery TT, Furge KA, Williams BO, Tretiakova M, Montag A, Vogelzang NJ, Re GG, Garvin AJ, Soderhall S, Kagawa S, et al. Gene expression profiling of favorable histology Wilms tumors and its correlation with clinical features. Cancer Res 2002; 62: 6598605.
  • 21
    Li W, Kessler P, Williams BR. Transcript profiling of Wilms tumors reveals connections to kidney morphogenesis and expression patterns associated with anaplasia. Oncogene 2005; 24: 45768.
  • 22
    Williams RD, Hing SN, Greer BT, Whiteford CC, Wei JS, Natrajan R, Kelsey A, Rogers S, Campbell C, Pritchard-Jones K, Khan J. Prognostic classification of relapsing favorable histology Wilms tumor using cDNA microarray expression profiling and support vector machines. Genes Chromosomes Cancer 2004; 41: 6579.
  • 23
    Takahashi M, Yang XJ, Sugimura J, Backdahl J, Tretiakova M, Qian CN, Gray SG, Knapp R, Anema J, Kahnoski R, Nicol D, Vogelzang NJ, et al. Molecular subclassification of kidney tumors and the discovery of new diagnostic markers. Oncogene 2003; 22: 68108.
  • 24
    Rigolet M, Faussillon M, Baudry D, Junien C, Jeanpierre C. Profiling of differential gene expression in Wilms tumor by cDNA expression array. Pediatr Nephrol 2001; 16: 111321.
  • 25
    Li CM, Guo M, Borczuk A, Powell CA, Wei M, Thaker HM, Friedman R, Klein U, Tycko B. Gene expression in Wilms' tumor mimics the earliest committed stage in the metanephric mesenchymal-epithelial transition. Am J Pathol 2002; 160: 218190.
  • 26
    Arcellana-Panlilio MY, Egeler RM, Ujack E, Pinto A, Demetrick DJ, Robbins SM, Coppes MJ. Decreased expression of the INK4 family of cyclin-dependent kinase inhibitors in Wilms tumor. Genes Chromosomes Cancer 2000; 29: 639.
  • 27
    Dome JS, Chung S, Bergemann T, Umbricht CB, Saji M, Carey LA, Grundy PE, Perlman EJ, Breslow NE, Sukumar S. High telomerase reverse transcriptase (hTERT) messenger RNA level correlates with tumor recurrence in patients with favorable histology Wilms' tumor. Cancer Res 1999; 59: 43017.
  • 28
    Baudry D, Faussillon M, Cabanis MO, Rigolet M, Zucker JM, Patte C, Sarnacki S, Boccon-Gibod L, Junien C, Jeanpierre C. Changes in WT1 splicing are associated with a specific gene expression profile in Wilms' tumour. Oncogene 2002; 21: 556673.
  • 29
    Ghanem MA, van Steenbrugge GJ, Sudaryo MK, Mathoera RB, Nijman JM, van der Kwast TH. Expression and prognostic relevance of vascular endothelial growth factor (VEGF) and its receptor (FLT-1) in nephroblastoma. J Clin Pathol 2003; 56: 10713.
  • 30
    Wirths O, Waha A, Weggen S, Schirmacher P, Kuhne T, Goodyer CG, Albrecht S, Von Schweinitz D, Pietsch T. Overexpression of human Dickkopf-1, an antagonist of wingless/WNT signaling, in human hepatoblastomas and Wilms' tumors. Lab Invest 2003; 83: 42934.
  • 31
    Eggert A, Grotzer MA, Ikegaki N, Zhao H, Cnaan A, Brodeur GM, Evans AE. Expression of the neurotrophin receptor TrkB is associated with unfavorable outcome in Wilms' tumor. J Clin Oncol 2001; 19: 68996.
  • 32
    Grundy P, Telzerow P, Moksness J, Breslow NE. Clinicopathologic correlates of loss of heterozygosity in Wilm's tumor: a preliminary analysis. Med Pediatr Oncol 1996; 27: 42933.
  • 33
    Stanhope-Baker P, Williams BR. Identification of connective tissue growth factor as a target of WT1 transcriptional regulation. J Biol Chem 2000; 275: 3813950.
  • 34
    Park DJ, Vuong PT, de Vos S, Douer D, Koeffler HP. Comparative analysis of genes regulated by PML/RAR alpha and PLZF/RAR alpha in response to retinoic acid using oligonucleotide arrays. Blood 2003; 102: 372736.
  • 35
    Tice DA, Szeto W, Soloviev I, Rubinfeld B, Fong SE, Dugger DL, Winer J, Williams PM, Wieand D, Smith V, Schwall RH, Pennica D, et al. Synergistic induction of tumor antigens by Wnt-1 signaling and retinoic acid revealed by gene expression profiling. J Biol Chem 2002; 277: 1432935.
  • 36
    Zirn B, Samans B, Spangenberg C, Graf N, Eilers M, Gessler M. All-trans retinoic acid treatment of Wilms tumor cells reverses expression of genes associated with high risk and relapse in vivo. Oncogene 2005; 24: 524651.
  • 37
    Muller H, Bracken AP, Vernell R, Moroni MC, Christians F, Grassilli E, Prosperini E, Vigo E, Oliner JD, Helin K. E2Fs regulate the expression of genes involved in differentiation, development, proliferation, and apoptosis. Genes Dev 2001; 15: 26785.
  • 38
    Bracken AP, Pasini D, Capra M, Prosperini E, Colli E, Helin K. EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancer. EMBO J 2003; 22: 532335.
  • 39
    Trimarchi JM, Lees JA. Sibling rivalry in the E2F family. Nat Rev Mol Cell Biol 2002; 3: 1120.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This article contains supplementary material available via the Internet at http://www.interscience.wiley.com/jpages/0020-7136/suppmat .

FilenameFormatSizeDescription
jws-ijc.21564.tbl1.xls18KSupporting Information file jws-ijc.21564.tbl1.xls
jws-ijc.21564.tbl2.xls86KSupporting Information file jws-ijc.21564.tbl2.xls
jws-ijc.21564.tbl3.xls46KSupporting Information file jws-ijc.21564.tbl3.xls

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.