Overall ten genes five of which are coding genes

Overall, ten genes, five of which are coding genes (FMOD, SULF1, COL12A1, STMN2 and MAEL), and five long-non-coding RNA (lncRNA) genes, were differentially expressed in hUC cells. lncRNAs are non-protein coding transcripts longer than 200 nucleotides. They have been implicated in the regulation of gene chemokine receptor antagonist at epigenetic, transcriptional and post transcriptional levels (Cao, 2014). Recently, their role in many biological processes, including pluripotency and differentiation, has been recognized. Further investigation of the lncRNAs found in our study may shed light as to their specific functions in hUC cells.
Disclosures
Acknowledgements The study was supported by OU-WB ISCRM, Oakland University and Michigan Head and Spine Institute. N. Beeravolu received Provost Graduate Research Award from Oakland University for this project. I. Khan was supported by Dr. Panjwani Center for Molecular Medicine and Drug Research, Pakistan. The authors acknowledge Dr. Shravan Chintala, Eye Research Institute for help confocal microscopy. We are also grateful for the Beaumont Biobank team, especially Judith Berry, RN. Barbara Preutz, Billie Ketelsen, and Evie Russell, Department of OB/GYN, Beaumont Hospitals as well as graduate students in our lab, especially Ali Alamri and Chris Lucier for assistance during the course of this study.
Introduction If each cell type is defined by the genes it expresses, then one would expect every cell type to show a distinct pattern of expression, characterizing that cell type. Such cell type-specific knowledge is important for advancing our basic understanding of biology and as a useful starting point for drug discovery. Such knowledge also sheds light on how one might reprogram one cell type in to another—a major hurdle in the process of direct reprogramming (Vierbuchen et al., 2010). However, elucidating a unique expression pattern for each cell type requires comparisons across a broad set of cell types. If one were to compare only fibroblasts to neurons, for example, one would find unique signatures distinguishing these cell types from each other, but not from other cells. Therefore, data-derived comparative signatures are context-dependent—subject to the diversity of cell types included in the comparison. Ignoring the context-dependency has led previous analyses astray—many genes that were identified as being expressed specifically in a particular cell type (i.e., markers for that cell type), were later found to be expressed in several different cell types (Juuri et al., 2012). A secondary goal was to find the unique regulatory core for each cell state—the elements which drive and maintain cell fate (Kim et al., 2008; Wang et al., 2006). Direct reprogramming of cells (e.g., from fibroblast to PSC), has shown that overexpression of a small number of transcription factors can drive a cell to become a completely different type of cell (Takahashi and Yamanaka, 2006). This direct reprogramming approach is quickly serving to identify robust transcriptional networks that drive particular cell fates, even when introduced into cells of a different germ layer. However, identifying these core expression factors has typically taken years of painstaking effort. Normally, the first step in identifying these small groups of cell fate drivers is to compare the gene expression of just two types of cells (or against several others in aggregate), and then to select the most upregulated transcription factors in the desired cell type. Next, through trial and error, cocktails of successful reprogramming factors (not necessarily unique) can sometimes be identified (Takahashi and Yamanaka, 2006; Vierbuchen et al., 2010). This overall approach has been hampered by the selection of factors based on expression differences between just two cell types (or based on comparing one cell type to several others in aggregate). Thus, our second goal was to streamline this type of reprogramming pre-selection process by obtaining a more refined comparison as a result of using a broader data set. We have named our overall approach, CEMA, for Core Expression Module Analysis.