High Throughput Sequencing Following Cross-Linked Immune Precipitation (HITS-CLIP) of Argonaute (AGO) Identifies Mir-193a as a Regulator of Jagged1 In Marrow Stromal Cells.

Abstract 3847

MicroRNAs (miRNAs) are small non-coding RNAs with important roles in the post-transcriptional regulation of up to 30% of all vertebrate genes. Traditional methods to determine miRNA-mRNA interactions have included transcriptional profiling of miRNAs, bio-informatic prediction of miRNA-mRNA binding, analysis of 3` untranslated region (3`UTR) binding of miRNAs and over-expression of miRNAs in relevant cell types. These studies however fall short of demonstrating direct interaction between a miRNA and its target mRNAs. We applied a recently described biochemical technique of high throughput sequencing following cross-linked immune precipitation (HITS-CLIP) to dissecting the miRNA-mRNA interactions in two functionally distinct human marrow stromal cell lines. HITS-CLIP relies on the ability of ultraviolet (UV) radiation to cross-link RNA to proteins they are bound to, followed by immune-precipitation of the RNA-protein complex to isolate the cross-linked RNA and sequencing by high throughput techniques. As miRNA-mRNA interactions occur in close proximity to the argonaute proteins (AGO), an anti-argonaute monoclonal antibody was used to isolate the Ago-miRNA-mRNA complexes.

The two stromal cell lines analyzed by HITS-CLIP (designated HS5 and HS27a) were isolated from a normal marrow primary long term culture (LTC), immortalized and extensively characterized for both function and expression profiles (mRNA and miRNA). HS5 was found to secrete growth factors that stimulate proliferation and differentiation of hematopoietic progenitors (G-CSF, IL-6, IL-1α and IL1β), whereas HS27a expresses activities associated with the stem cell niche (CXCL12, Angiopoietin-1, Jag1 etc). In keeping with this, HS5 conditioned media stimulated proliferation and differentiation of isolated CD34+ cells whereas HS27a supported CD34+ cells in an undifferentiated state. Sequence reads from the HITS-CLIP analysis from each of the cell lines were aligned to the human genome using the UCSC genome browser to identify Ago-mRNA and Ago-miRNA binding sites in both the cell lines. Interestingly, corresponding datasets from HS5 and HS27a were similar for the majority of mRNAs and miRNAs, but distinct for those mRNAs (such as Jag1, CXCL12, IL6 and GCSF) and miRNAs (such as miR-886-3p, miR-221, miR-181a and miR-193a) known to be differentially expressed between the two cell lines. We then validated the use of the HITS-CLIP strategy in stromal cells by analyzing one such Ago-mRNA binding site for Jagged1 (Jag1). Jag1 is a ligand for Notch1 and is expressed in those cells that support the hematopoietic stem cell (HSC) niche. The Notch pathway is a highly conserved signaling system critical in regulating several tissue systems including hematopoietic cells. This binding site, 1749 bp downstream of the transcriptional start-site for Jag1 was significantly more enriched in HS5 compared to HS27a. The site was also a predicted binding site for miR-193a, a miRNA over-expressed in HS5 compared to HS27a cells. Over-expression of miR-193a in HS27a cells resulted in the down-regulation of Jag1 protein (as measured by Western blotting). To confirm the direct interaction between Jag1 and miR-193a, we cloned this purported binding site downstream of the luciferase gene and co-transfected the plasmid with miR-193a. Luciferase activity was down-regulated greater than 50% when compared to control transfections suggesting a direct effect of miR-193a on Jag1 transcript. In summary, our data suggest that HITS-CLIP methodology can be used to define in vivo spatial interactions between miRNA and mRNAs in the marrow microenvironment (ME). It can also be used to define miRNA-based regulation of specific genes such as Jag1, which are critical to defining functional niches in the ME.