Perlman Laboratory

Research Highlights

PEDIATRIC RENAL TUMOR CLASSIFIERS AND POTENTIAL THERAPEUTIC TARGETS

Pediatric renal tumors fall into four primary categories: Wilms Tumors (WT, 85%), Clear Cell Sarcomas of the Kidney (CCSK, 5%), Congenital Mesoblastic Nephromas (CMN, 4%), and Rhabdoid Tumors of the Kidney (RTK, 3%). While the majority of WTs are easily recognized by their characteristic combination of epithelial, blastemal and stromal elements, approximately 5% of WT are composed exclusively of undifferentiated blastemal or stromal elements. Such tumors are easily mistaken for CCSK, CMN, and RTK. It has been the experience of the COG Renal Tumor Pathology Center that approximately 25% of such lesions are misdiagnosed by the local pathologist. Even following central pathology review, a small but significant number of tumors are provided a diagnosis that is less than fully certain. The importance of an accurate diagnosis is most evident when the differences in therapy and prognosis between these four different tumors are considered. Our goal has been to utilize gene expression data to accurately and reliably distinguish CCSK, CMN, RTK, and WT, and to define diagnostically useful gene expression signatures for each of these entities.

In collaboration with Dr. Spencer Huang a biostatistician and faculty member in the Department of Preventive Medicine at Northwestern University Feinberg School of Medicine, we have developed a classifier that accurately predicts the category of pediatric renal tumors [1]. Furthermore, we have published the following tumor type-specific works showing: (1) neural markers are upregulated and the sonic hedgehog and AKT pathways are activated in CCSK [2], (2) significant receptor tyrosine kinase activation in CMN [3], (3) significant repression of neural development-associated genes and bivalently histone modified/polycomb target genes in RTK [4], and (4) the classification of WT into distinct, clinically relevant subsets based on gene expression data [5], including low-risk [6] and WT that relapse [7].

THERAPEUTICALLY APPLICABLE RESEARCH TO GENERATE EFFECTIVE TREATMENTS (TARGET)

The NCI TARGET initiative seeks to identify therapeutic targets for high-risk pediatric tumors through genomic sequencing supported by copy number and gene expression analysis. The integration of data from high-throughput molecular analyses has uncovered great genetic and epigenetic variability, even within tumors that are histologically similar. These findings suggest that this variability can be used as prognostic indicators. Understanding the molecular processes that are dysregulated by these genetic and epigenetic changes is a powerful approach for personalizing existing care as well as developing new therapeutic agents that specifically target aberrant pathways. The latter will result in treatments that are more effective, less toxic, and have lower long-term morbidity than existing treatments.

The High-Risk WT TARGET project, currently in progress, includes the analysis of favorable histology (FH) tumors that relapsed and tumors with anaplasia (unfavorable histology). Approximately 80% of WT patients lack anaplasia and thus constitute favorable histology WT. Although the majority of patients with FHWT have an excellent prognosis, approximately 15% of patients will relapse and 50% of these patients will not survive. Patients with anaplastic WT tend to have a poor outcome, with survival rates of approximately 60%. We are seeking to identify therapeutic targets for these high-risk tumors by using an integrative approach of genomic sequencing [whole genome sequencing, performed by Complete Genomics Inc. (CGI); whole exome sequencing, performed by Baylor College of Medicine (BCM)], copy number analysis (Affymetrix SNP 6.0, performed at St. Jude Children’s Research Hospital), gene expression analysis (Affymetrix U133_plus2, performed at Stanley Manne Children’s Research Institute), and DNA methylation analysis (Illumina 450K, performed at Northwestern University Genomic Core Facility). These data are being analyzed in collaboration with Dr. Daoud Meerzaman’s group at The Center for Biomedical Informatics and Information Technology (CBIIT).

The CCSK TARGET project is currently underway in an effort to better understand these frequently aggressive kidney tumors that usually occur in children <3 years of age. Despite efforts to better characterize CCSK, the cell of origin and specific genetic changes that are associated with tumorigenesis remain unclear in these tumors. Therefore, in order to develop therapeutic targets for CCSK, it will first be necessary to identify recurrent genetic and/or epigenetic changes in these tumors. To this end, the integrative approach described above for WT will be used for CCSK.

The RTK TARGET project is currently underway in an effort to better understand these rare, highly malignant tumors of infancy that have a poor prognosis. RTK consistently show loss of INI1/SMARCB1, one of over 10 non-catalytic subunits of the evolutionarily conserved SWI/SNF chromatin-remodeling complex. We previously reported that RTK show striking repression of genes associated with bivalent histone modification in embryonic stem cells and our preliminary data in an RT cell line (G-401) suggest that this repression is at least partially mediated by increased H3K27me3. Based on these observations, we hypothesized that the loss of INI1/SMARCB1 in RTK results in an overall failure to release the repressive H3K27me3 histone marker in genes responsible for early differentiation, contributing to the aggressive phenotype of these tumors. To test this hypothesis, ChIP-Seq will be performed in RTK tumor samples as well as established cell lines using six different histone markers, including H3K27me3. Furthermore, these tumor and cell line samples will be subjected to miRNA-Seq, whole genome sequencing, DNA methyl-Seq, and RNA-Seq. This large-scale effort is being performed in collaboration with Dr. Marco Marra of The British Columbia Cancer Agency. The overall goal is to develop new therapeutic targets for these poor-outcome tumors by gaining a detailed understanding of the RTK genome and epigenome.

Principal Investigator

Elizabeth J. Perlman, MD