Non-coding RNAs in gene therapy: mechanisms, delivery strategies, and clinical translation

Non-coding RNAs (ncRNAs) have evolved from being regarded as “genomic noise” to a significant asset for gene therapy, enabling programmable control of transcription, RNA stability, and translation. This review integrates ncRNA biogenesis, therapeutic mechanisms, delivery engineering, and clinical translation across RNA subclasses (miRNAs, siRNAs, piRNAs, lncRNAs, and circRNAs), connecting molecular discoveries to design parameters and regulatory-grade development standards. Non-coding RNAs operate through sequence-specific silencing, decoy or sponging interactions, scaffolding of protein complexes, and epigenetic alteration across several modalities, creating multiple therapeutic opportunities while necessitating specialized requirements for cellular localization and target validation. Comparative synthesis indicates that siRNA therapeutics have progressed most swiftly due to dependable cytosolic target engagement and modular chemistry, with approved treatments such as patisiran demonstrating prolonged in vivo gene silencing. Conversely, early miRNA replacement endeavors (e.g., MRX34) demonstrate how innate immune activation, formulation-specific toxicity, and clinical variability might compromise favorable biological results. Nuclear accessibility, tissue-specific expression, and inadequate mechanistic characterization represent substantial translational barriers for lncRNA and circRNA treatments. Delivery emerges as the principal rate-limiting factor, as lipid nanoparticles, viral vectors, polymers, and extracellular vesicles each negotiate tissue tropism, endosomal escape, manufacturability, and immunogenicity. These insights collectively support a translational roadmap that prioritizes safer and more selective delivery methods, including ligand-guided and hybrid carriers; clearly defined endpoints that distinguish preclinical potential from clinical validation; and sophisticated design strategies incorporating RNA editing technologies and AI-driven optimization.