Recently epigenetic modifications have drawn lots of attention in the oncology field. Cancer is characterized by aberrant patterns of gene expression of multiple genes. These major shifts in gene expression are believed to be due to genetic and epigenetic changes. Lots of high-throughput genome-wide sequencing data (e.g. ChIP-seq) and expression data (RNA-seq, microarray) are generated everyday to address this topic. To understand co-operation and co-regulation of hundreds of altered genes and many aberrant histone modifications, we need to have a collective view. Since multiple histone modifications exist, they are perceived to be interdependent and dynamically modulated in various biological processes, including disease. Histone modifications are controlled by histone modifying enzymes. So, we have interest on how these enzymes are expressed and co-regulated in cancer cells when compared to their normal counterpart.
This project started in 2010 with supervision of my mentor Dr. Nuria Lopez-Bigas and Dr. Elizaveta Benevolenskaya (University of Illinois, Chicago, USA). We performed gene expression analysis of histone demethylases (HDMs) and histone methyltransferases (HMTs), the enzymes that bind and enzymatically modify histone tails, in normal and tumor tissues. Using representative, internally consistent expression datasets from a variety of normal and tumor tissues and from cell lines, we have identified cell types and tumors where HDM and HMT genes are prominently expressed and underexpressed. Our study enabled identification of sets of co-regulated HDMs and HMTs, comprising HDM/HMT gene expression signature. Most of the analyses were done using our own software Gitools. Strikingly, the expression of HDM and HMT genes was highly correlated with overexpression or underexpression of their target genes and of genes characterized by specific histone modifications. Surprisingly, this analysis revealed multiple correlations in the expression levels of different HDMs and HTMs. In addition, we found that the HDM/HMT gene expression signature was correlated to the expression of HDM/HMT target genes and of genes with relevant histone modifications. The revealed correlations were strikingly different in cancer cells compared to normal cells. So, we may confer that coordinated regulation of enzymes regulating histone modifications may underlie global epigenetic changes occurring in cancer. The fact that multiple alterations in HDM and HMT genes can be discerned simultaneously suggests that this approach might be used to discover a combination cancer therapy that would be more effective and less toxic than single drugs.
Although the analysis was basically done in silico, however, after this initial assessment of expression patterns in various malignancies using cancer cell line data we surveyed transcript levels in hundreds of cancer tissue samples. For these purposes, we used RT-qPCR analysis, which is a better method for quantisation of gene expression than microarray studies. This wet lab experiments were performed by William and Laura, our colleagues in University of Illinois, USA.
All these results have just been published in PloS ONE journal: