Identification of Long Non-Coding RNA Genes Important for Normal Hematopoiesis and Leukemogenesis Using Bioinformatics Tools

Only 1.5% of human genome encodes protein-coding genes, while more than 60% of human genome is transcribed for RNA molecules that have no-protein coding potentials. So far 6,000 long noncoding RNA genes (lncRNA) are documented (Evans et al., 2016). To investigate the functions of lncRNA genes poses huge opportunities as well as challenge to biomedical research. In this project, we interrogate important lncRNA genes that may be associated with normal and abnormal hematopoiesis using publically available databases containing RNA-seq data from different blood lineages and leukemia cells. The data mining and statistical analysis are performed by R program and oracle SQL code. Using RNA-seq data derived from purified blood lineages of human cord blood cells, we have identified long noncoding RNA genes transcribed exclusively in human hematopoietic stem cells as well as in different progenitor lineages. We have discovered two lncRNA genes: linc00028 and As-PP1 are regulated by RUNX1 in hematopoietic stem cells. Given that chromosome translocation t(8;21) generates a fusion protein RUNX1-ETO which often represses genes activated by RUNX1, we proceed to validate whether Linc00028 and As-PP1 may be regulated by transcription factor RUNX1 by chromatin immunoprecipitation assays. Further analysis implies that As-PP1 activated p53-mediated apoptosis via controlling p53 activity. Thus AML1-ETO may repress apoptosis via antagonizing the normal function of RUNX1 in repressing AsPP1 transcription. We have been studying how protein arginine methyltransferases 1 (PRMT1) promotes the progression of acute myeloid leukemia. In this vein, we found that inhibition of PRMT1 activity by small molecule PRMT1 inhibitor upregulated the expression level of the long noncoding RNA gene (As-PP1) and leads to leukemia cell apoptosis. Thus long noncoding RNA genes may serve as biomarkers for PRMT1 inhibitor treatment. By clustering genes of acute myeloid leukemia patients in TCGA database, we found unique lncRNA signature genes specifically associated to distinct groups of leukemia patients with PML-RAR alpha, FLT3 mutation and AML1-ETO (aka RUNX1-ETO) respectively. Whether these signature genes are useful to be applied as biomarkers will be further discussed in the presentation.