Transcriptome studies of cell-fate and aging

Abstract: Microarrays enable the researcher to capture the expression levels from most genes in a single experiment. The methodology has opened up a new field with possibilities and challenges. This thesis focuses on both methodological and bioinformatics improvements in the application of microarrays; and in particular on the use of such approaches in cell-fate and aging research. We have introduced a method to fabricate microarrays on unmodified glass that enables a cost effective approach to produce microarrays; and refined the SAM (significance analysis of microarrays) approach, highlighting important technical aspects of this, the most widely used method to identify differentially expressed genes. Cell-fate is a term to describe how a cell ends its normal life span. We have used microarrays to identify genes that are associated with cell-fate outcomes. Initially, we studied early changes of gene expression during induction of cellular senescence and separated them from changes that are associated with growth arrest. We thereby identified a set of genes that are induced during induction of cellular senescence but not during growth arrest. We also studied a model-system that recapitulates a characteristic of cancer cells, apoptosis resistance, mediated by overexpression of the translation initiation factor eIF4E. We identified a set of transcripts that are overtranslated when eIF4E is overexpressed and demonstrated that two of them act as mediators of apoptosis resistance. These studies have contributed to the understanding of how the transcriptome is regulated during induction cell-fates and how regulation of translation can influence cell-fate decisions. Senescence has been hypothesized to cause aging. When analyzing microarray studies of cellular senescence and aging we found a common mammalian aging transcriptome that had at least two components and established that cellular senescence resembles aging in mice but not humans. These findings are important as it indicates that aging in mice and humans could be substantially different and that cellular senescence in tissues, probably does not contribute to aging in humans.