Signal Transducer and Activator of Transcription (STAT)-3-Dependent Regulation of Non-Coding RNA Genes in Chronic Lymphocytic Leukemia (CLL).

Abstract 2886

Non-coding RNAs regulate the expression of more than 30% of protein-coding genes both at a post-transcriptional and translational level. Although approximately 1000 microRNAs (miRs) have been identified in the human genome, little is known about the mechanisms that regulate miR expression. STAT3 regulates the transcription of miR-21, and miR181b-1, binds to their promoter and induce neoplastic cell transformation (Iliopoulos, Jaeger et al. 2010). Because STAT3 is constitutively activated in CLL cells (Hazan-Halevy, Harris et al. 2010) we sought to investigate how STAT3 affects non-coding RNA gene expression in CLL cells. We transfected peripheral blood CLL cells from 3 different patients with STAT3-shRNA and assessed non-coding RNA levels using a non-coding RNA array containing 2277 human miR probes, 960 from ultra-conserved genes and 3540 of long non-coding RNAs. When compared to transfection control, 152 probes from 78 non-coding RNA genes were differentially expressed (134 down-regulated and 18 up-regulated), suggesting that STAT3 affects the non-coding RNA network in CLL cells. Supervised clustering analysis was used to select genes for validation. By using quantitative RT-PCR we validated our gene array analysis. Similar to the data obtained by the non-coding RNA array, we found that transfection of CLL cells with STAT3- down-regulated the levels of miR-21, miR-155, and miR-320b. Binding site prediction programs and ChIP-seq data embedded in the UCSC genome browser determined that in 5 of 7 genes, down-regulated by STAT3-shRNA transfection, were either putative or experimentally confirmed STAT3-binding sites, indication that STAT3 directly regulates the transcription of those miRs. It has been shown that the interaction between miRs and single stranded RNA is dependent on base pairing in a seed region at positions 2 to 8. High levels of 4.8kb single stranded STAT3 RNA transcripts, present in CLL cells, provide a substrate for such paring. Therefore, we assumed that STAT3 functions as a “RNA sponge” soaking up miRs and altering their effective levels and function. To test this hypothesis we used the pattern-based RNA22 algorithm and identified potential miR targets. We than calculated the energy that would be released if the corresponding RNA/RNA complexes are saturated. We found that the energy released from binding of miRs to STAT3 sequences would be higher than energy released from binding to a random sequence with same length and base content suggesting that STAT3 “sponges out” miRs in a sequence specific manner. Thus, CLL cells are characterized by an ongoing interaction between STAT3-mediated transcriptional regulation of non-coding RNA and miR-mediated translational regulation of coding genes.