Genome-Wide Survey of MicroRNA - Transcription Factor Feed-Forward Regulatory Circuits in Human

In this work, we describe a computational framework for the genome-wide identification and characterization of mixed transcriptional/post-transcriptional regulatory circuits in humans. We concentrated in particular on feed-forward loops (FFL), in which a master transcription factor regulates a microRNA, and together with it, a set of joint target protein coding genes. The circuits were assembled with a two step procedure. We first constructed separately the transcriptional and post-transcriptional components of the human regulatory network by looking for conserved over-represented motifs in human and mouse promoters, and 3’-UTRs. Then, we combined the two subnetworks looking for mixed feed-forward regulatory interactions, finding a total of 638 putative (merged) FFLs. In order to investigate their biological relevance, we filtered these circuits using three selection criteria: (I) GeneOntology enrichment among the joint targets of the FFL, (II) independent computational evidence for the regulatory interactions of the FFL, extracted from external databases, and (III) relevance of the FFL in cancer. Most of the selected FFLs seem to be involved in various aspects of organism development and differentiation. We finally discuss a few of the most interesting cases in detail.

💡 Research Summary

The paper presents a computational pipeline for the genome‑wide identification of mixed transcriptional/post‑transcriptional regulatory circuits in humans, focusing on feed‑forward loops (FFLs) in which a master transcription factor (TF) regulates a microRNA (miRNA) and, together with that miRNA, co‑regulates a set of protein‑coding target genes. The authors first constructed two separate subnetworks: a transcriptional network derived from conserved over‑represented TF‑binding motifs in human and mouse promoters, and a post‑transcriptional network based on conserved miRNA target sites in 3′‑UTRs. Motif discovery employed position weight matrices from JASPAR and TRANSFAC, with statistical enrichment assessed against 10 000 randomized promoter/UTR sequences. miRNA‑target prediction combined TargetScan and miRanda, retaining only sites conserved between human and mouse.

The two subnetworks were then merged to search for three‑node subgraphs that satisfy the canonical FFL architecture: (1) TF → miRNA (transcriptional activation), (2) TF → target gene (direct transcriptional regulation), and (3) miRNA → the same target gene (post‑transcriptional repression). This exhaustive search yielded 638 candidate mixed FFLs. To evaluate biological relevance, the authors applied three independent filters. First, Gene Ontology (GO) enrichment analysis of the joint target genes identified circuits enriched for developmental, differentiation, and signaling processes. Second, external evidence from ChIP‑seq, CLIP‑seq, and curated interaction databases (e.g., miRTarBase) was used to confirm at least one of the three edges in each circuit, thereby reducing false positives. Third, the authors examined the involvement of each circuit in cancer by cross‑referencing COSMIC, TCGA, and the literature, highlighting those containing TFs, miRNAs, or targets known to be dysregulated in tumors.

After filtering, the majority of retained FFLs were associated with organismal development and cell‑type specification. Well‑known oncogenic circuits such as MYC‑miR‑17‑92, E2F1‑miR‑106b/25, and p53‑miR‑34a were recapitulated, providing validation of the approach. Novel candidates, for example SP1‑miR‑29‑COL4A1 and NF‑κB‑miR‑146a‑TRAF6, were also identified, suggesting previously uncharacterized layers of regulation. The authors argue that mixed FFLs enable signal amplification, noise buffering, and fine‑tuned temporal control by coupling transcriptional and post‑transcriptional mechanisms.

The discussion acknowledges several limitations: reliance on sequence‑based predictions introduces false positives; the analysis does not incorporate tissue‑specific expression data, so the networks represent a static, averaged view; and experimental validation is limited to literature cross‑checks. The authors propose future work integrating single‑cell transcriptomics, epigenomic profiling, and CRISPR‑based perturbation assays to capture dynamic, context‑dependent behavior of these circuits.

In conclusion, this study delivers a comprehensive catalog of human TF‑miRNA feed‑forward loops, demonstrates their enrichment in developmental and oncogenic pathways, and provides a valuable resource for downstream functional studies and therapeutic target discovery.