MicroRNAs in Normal Human Terminal Granulocytic Differentiation.

Abstract 1463

Poster Board I-486

Normal human myelopoiesis is a complex biological process, where the balance between cell proliferation, differentiation and apoptosis is tightly regulated by a transcriptional program that results in the production of appropriate numbers of circulating mature myeloid cells. MicroRNAs (miRNAs) are small non-coding RNAs of 18∼25 nt that can affect cellular protein levels. Several studies show specific miRNA expression patterns in different subtypes of myeloid malignancies, however only limited data is available on miRNA expression patterns during normal myeloid differentiation of primary human cells. We set out to characterize miRNA expression patterns in the different stages of granulocytic differentiation in two models. First myeloblast, promyelocytes, metamyelocytes and granulocytes from normal human bone marrow were cell-sorted with flow cytometry using the markers CD10, CD11, CD34, CD36, CD45 and CD117. Second, CD34+ cells from primary human fetal livers were differentiated in vitro towards neutrophils. MiRNA expression levels were determined at different time points (day 0, 3 and 10), representing different stages of granulocytic differentiation. MiRNA expression was measured using the qPCR platform, containing 365 miRNAs, from Applied Biosystems. To identify potential miRNA target genes, we performed mRNA expression profiling in the latter in vitro differentiation. The negative correlations between miRNA and mRNA expression were identified and integrated with a target prediction database (Targetscan). The miRNA profiling showed that approximately 70% of the 365 miRNAs analyzed, were expressed during granulocytic differentiation and that the miRNA expression pattern during this process change significantly in both models. Principal component analysis showed clear separation of the different subsets of granulopoiesis based on the miRNA expression. We determined the differentially expressed miRNAs between the various subsets using ANOVA with a P value <0.05, after correction for multiple testing. We found 24 miRNAs to be differentially upregulated in the both models. The top 5 upregulated miRNA, with the highest fold change in granulocytes as compared to myeloblasts, were miR-223, miR-145, miR-148, miR-24 and miR-23a. We identified 27 miRNAs that were downregulated, the top 5 were of miR-10a, miR-196a, miR-130a, miR-135a and miR-125b. Concomitant miRNA and mRNA expression analysis of the in vitro model with the Targetscan database, demonstrates a potential regulatory role for these miRNAs in various processes, such as cell proliferation, apoptosis and cell cycle regulation. For example, miR-130a, miR-20b and miR-191, miR-301 expression levels were negatively correlated with E2F2 and SOX4 respectively. Furthermore, MAPK1 levels correlated inversely with miR-17-5p, miR-130a, miR-181b, miR-181d and miR-20b. We observed potential regulation of BCL2L11 by miR-10a, miR-10b and CDK6 by miR-148a, miR-148b, miR-191 and miR-21, as well as CHEK1 by the miR-15a and miR-16, LATS2 by miR-142-3p and CCND3 by miR-133a. In addition we also identified myeloid specific genes to be potentially regulated by miRNAs such as CEBPA by miR-181b, KIT by miR-148a, miR-148b and miR-301 and RUNX3 by miR-301. This is the first comprehensive study of miRNA expression in normal human granulocytic differentiation. We show in two models that the miRNA expression pattern changes during granulocytic differentiation. miRNA-mRNA analyses suggest involvement of miRNAs in regulation of important cellular processes during granulocytic differentiation. Experimental validations of several candidate targets as well as functional studies are currently ongoing.