Abstract
Abstract
Acute leukemias are classified based on their immunologic, cytogenetic and morphologic characteristics. However, in most instances, response to treatment and survival probability cannot be accurately predicted, suggesting that the disease is even more complex and heterogeneous than can be shown with current techniques. Although prognostic value has been shown for certain gene expression profiles, expression profile studies are limited by the fact that only a snapshot of mRNA content is obtained in a basal state, failing to represent how genes will respond to different stresses, also failing to detect the roles of genes expressed at lower levels for which major changes in expression levels are often lost in the “noise” of expression arrays. In order to overcome these issues and to provide a more accurate molecular phenotype of acute leukemias, we established an integrated epigenomic and genomic high-throughput platform using novel techniques and custom high-density oligonucleotide arrays. We combine studies of i) genome-wide cytosine methylation using a novel technique we developed that provides accurate quantitative determination of DNA methylation levels, using genome-wide custom oligo arrays, ii) chromatin structure by ChIP on chip for histone code settings associated with active or repressed genes using 24 K promoter tiling arrays, iii) gene copy number by array-based comparative genomic hybridization (array CGH) at 6 kb resolution genome-wide and iv) 36 K gene expression arrays. Results are validated by single locus quantitative PCR techniques. Cross platform integration is facilitated by use of NimbleGen oligo arrays for all studies and analysis using novel bioinformatics and statistical models. We used this integrative analysis platform to perform an in-depth analysis of the epigenomic basis of AML and ALL using primary patient samples and cell lines. The data allowed us to generate a “gene activity index” which identified the ability of genes to be expressed to be characterized genome-wide in AML and ALL cells. This data also allowed a much more comprehensive analysis of pathways active in these cells to be identified in comparison to expression arrays alone. Our current studies apply integrative platform and gene activity indexing to large series of patients enrolled in multicenter clinical trials in order to provide high-resolution analysis of the molecular basis of acute leukemia.