engleski

Modeling epigenetic mechanisms of alternative splicing

The field of epigenetics is quickly growing as we realize its impact on the regulation of the gene expression, without a change in the DNA sequence. Three different systems are usually included within the epigenetics field: DNA methylation, Histone Modifications and RNA-associated Silencing. The outcome of these epigenetic changes can explain the differential expression between cells, tissues, organs and individuals, while they are genetically identical. In addition, epigenetics is important for X-chromosome inactivation in female mammals. Such is the importance of epigenetics that the disruption of any of its systems has been found to be related to diseases such as cancer or mental syndromes. On the other hand, alternative splicing is an expression edition process which allows eukaryotic cells to obtain several mature mRNA isoforms from the same gene. This means that different protein isoforms can be expressed specifically in each tissue or each development stage, implying great versatility for the organism. The alternative splicing regulation has been correlated with the conservation of the so called splice sites, recognized by the spliceosome, or specific splicing cis-regulatory elements, present in the primary transcript (pre-mRNA), where various proteins are bound. However, recent studies suggest that chromatin modifications have an effect on the alternative splicing end result, observing different exon inclusions when changing the local epigenetics. The interaction between elements on the chromatin structure, or the structure itself, and the RNA polymerase II may explain this phenomenon, as splicing is frequently co-transcriptional. Given the relationship between alternative splicing and epigenetic modifications, the present work intends to delve deeper into the matter by performing several logistic regression models. The statistical analysis presented in these pages focuses on the impact that histone modifications and DNA methylation have in the alternative splicing prediction of cassette exons in sixteen different tissues. The different models were developed combining histone modification and methylation data for the cassette exon and the surrounding introns. The results of these models highlight the importance of DNA methylation present in the cassette exon for determining the splicing outcome. However, methylation data alone doesn’t have predictive power, while this is true for histone modification data that have predictive values close to the model using all variables. This suggest a combining effect between epigenetic modifications. Additionally, a position-specific effect is shown for the DNA methylation while we do not see it in histone modifications.

epigenetics, histone modifications, DNA methylation, alternative splicing

nije evidentirano