The Analysis of the MCF7 Cancer Model System and the Effects of 5-AZA-2'-Deoxycytidine Treatment on the Chromantin State Using a Novel Microarray-Based Technology for High Resolution Global Chromatin State Measurement

A microarray method to measure the global chromatin state of the human genome was developed in order to provide a novel view of gene regulation. The 'chromatin array' employs traditional methods of chromatin isolation, microarray technology, and advanced data analysis, and was applied to a cancer model system. Chromatin is first separated by its condensation state using chromatin fractionation. By probing with a comparative genomic hybridization-style microarray, the chromatin condensation state of thousands of individual loci in an MCF7 tumor model cell line was determined and correlated with transcriptional activity. The chromatin array showed a significant portion (>3,000) of the genes were in a condensation state that was neither condensed or relaxed as a result of heterogeneity in the condensation states in the population. The utility of the chromatin array in deciphering gene regulation was demonstrated in a MCF7 cell line treated with 5 Aza dC, which disrupts genome methylation, and as a result causes global relaxation of chromatin structure. 5 Aza dC treatment results in strong changes in expression, and a normalized global chromatin relaxation of two-fold. A significant subset of 378 genes was condensed by 5 Aza dC treatment, indicating that a mechanism of chromatin regulation exists that can resist the effects of 5 Aza dC treatment. The genes with the largest changes in response to 5 Aza dC treatment showed a strong correlation with CpG island-based regulation (p < 0.0001), and a restoration of transcription patterns associated with normal mammary tissue. Analysis using splice-form specific microarray probes demonstrated that the chromatin state was not uniform across a gene. These findings indicate that certain gene regions exhibit differential sensitivity to 5 Aza dC treatment, and therefore may be regulated independently. Using functional annotation, expression microarray, and comparative genomic hybridization data, this work should provide a framework through which the biological implications of the relationship between chromatin accessibility and expression may be deciphered.