groups including alternative splicing, protein phosphorylation and acetylation, gene transcription, ATP binding. The largest group, alternative splicing, is an expected result because DAZAP1 is a splicing factor and these target genes are originally selected based on their Author Manuscript Author Manuscript Author Manuscript Author Manuscript Nat Commun. Author manuscript; available in PMC 2014 August 27. Choudhury et al. Page 
7 significant change in splicing levels. Protein phosphorylation and acetylation are the next largest functional groups, suggesting that DAZAP1 is closely involved in regulating cellular signaling pathways. We then extracted all the DAZAP1 targets in the phosphoprotein and acetylation groups, and find that many of these targets were functionally connected into well-linked interaction networks as judged by a functional network analysis using STRING . Interestingly, the largest interaction network contained many genes in cell signaling pathways that control gene transcription, cell cycle and cell proliferation. This network also includes a small group of metabolic enzymes involved in ATP synthesis, suggesting that DAZAP1 may affect energy metabolism of cells. Subsequent analysis of splicing changes indicate that the splicing of ~60% of genes is positively regulated by DAZAP1, consistent with the role of DAZAP1 to promote exon inclusion when directly binding to pre-mRNA. We SB-590885 web further tested the results of mRNA-seq by randomly selecting several targets and assaying their splicing with RT-PCR in cells with overexpression or knockdown of DAZAP1. As shown in the example of WAC, a gene encoding WW domain-containing adapter protein with coiled-coil to regulating H2B monoubiquitination and gene transcription 49, DAZAP1 over-expression promotes inclusion of exon PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19845007 7 to produce more full-length canonical isoforms whereas RNAi of DAZAP1 has the opposite effect. We further validated additional positively regulated DAZAP1 targets involved in cell proliferation and RNA processing pathways. For example, the SRSF6 gene encodes a splicing factor that can enhance proliferation of lung epithelial cells 50, and MELK encodes a protein kinase that can regulate cell cycle progression by binding and phosphorylating CDC25B 51. We obtained reciprocal splicing changes for these candidate genes in cells with over-expression and knockdown of DAZAP1, suggesting that mRNA-seq indeed can reliably identify endogenous splicing events regulated by DAZAP1. This splicing regulation activity of DAZAP1 is comparable to a well-established splicing factor SRSF1. In addition, knock-down of DAZAP1 with shRNA in HeLa cells shows similar effects on splicing of DAZAP1 targets, indicating that the splicing regulation activity of DAZAP1 is not tissue or cell type specific. We also analyzed sequences around these DAZAP1 regulated exons to examine if they contain the RNA motifs recognized by DAZAP1. We extracted sequences within or at the downstream of the skipped exons that were positively or negatively regulated by DAZAP1, and compared the relative abundance of the DAZAP1 binding motifs with the control exon set that were not affected by RNAi of DAZAP1. Using an enrichment score of small motifs 52, we find, as expected, that DAZAP1 binding motifs are more enriched in exons positively regulated by DAZAP1 but are slightly depleted in exons negatively regulated by DAZAP1. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Nat Commun. Author manuscr