Introduction
MicroRNAs (miRNAs) are small, non-coding RNA molecules that function as post-transcriptional regulators of gene expression. While their role in silencing target messenger RNAs is well-documented, the mechanisms governing the expression of miRNAs themselves remain less understood. Most existing resources for miRNA regulation rely on manual literature curation, which is often incomplete and slow to update. To address this gap, researchers developed PuTmiR, a computational database designed to identify transcription factors (TFs) that potentially regulate human miRNAs by binding to nearby genomic regions.
Methodology
The PuTmiR database was constructed using a high-throughput computational pipeline that integrates genomic location data with protein-binding models:
- Genomic Mapping: Coordinates for 721 human miRNAs were retrieved from miRBase (v10).
- TFBS Prediction: Potential transcription factor binding sites (TFBS) for 398 TFs were identified using the UCSC Genome Browser and the TRANSFAC 7.0 database.
- Cross-Species Conservation: To minimize false positives, the authors filtered for binding sites conserved across human, mouse, and rat genomes, assuming that evolutionary conservation indicates functional importance.
- Statistical Scoring: Log-odds and z-scores were employed to rank the significance of predicted binding sites.
- Interface Development: A web-based tool was created to allow researchers to query specific miRNAs and explore predicted TF interactions within a customizable range of up to 10 kb.
Key Insights: The Importance of Downstream Regulation
A traditional focus in genomics is the upstream promoter region. However, PuTmiR highlights the significance of the Downstream Region (DSR). Analysis revealed that the number of putative TF-miRNA interactions in the 10 kb downstream of a miRNA often exceeds those in the 10 kb upstream. This suggests that enhancers and other regulatory elements located after the miRNA gene play a critical role in its expression.
Results and Network Analysis
The database identified a total of 11,485 regulatory relationships in the upstream region and 12,251 in the downstream region. Key findings include:
- Regulatory Complexity: Some miRNAs are predicted to be regulated by over 100 different TFs, indicating highly complex control networks.
- Network Topology: The TF-miRNA interaction network follows a power-law distribution, characteristic of scale-free networks. A few "hub" TFs regulate a disproportionately large number of miRNAs.
- Biological Validation: The authors demonstrated the utility of PuTmiR by identifying the protein YY1 as a significant regulator of "oncomiRNAs" in colon cancer, providing a starting point for experimental validation.
Limitations and Future Directions
While PuTmiR provides a powerful predictive tool, the authors emphasize that these interactions are "putative." Computational predictions require experimental validation through methods like ChIP-seq to confirm actual binding in specific cellular contexts. Furthermore, as human genome assemblies are updated (e.g., from hg18 to hg38), the database requires ongoing maintenance to ensure coordinate accuracy. Future iterations aim to integrate these computational predictions with experimentally validated data to create a more comprehensive regulatory map.
Summary updated