Circular RNA–Directed Antisense Therapies: From Biogenesis and BSJ-Specific Design to Next-Generation Delivery Systems
摘要
Circular RNAs (circRNAs) are a class of non-coding RNAs (ncRNAs), which is covalently closed structure and extreme resistance to exonuclease activity compared to linear isoforms. Understanding the role in pathologies, such as cancer and fibrosis, and functions, such as microRNA (miRNA) sponging and interaction with RNA-binding proteins (RBPs), they become ideal therapeutic targets. This review summarizes the circRNA biogenesis and discusses the antisense oligonucleotide (ASO) based treatment strategies, comprehensively. Particularly, this review discusses the gapmer and mixmer strategy, degrading circRNA via RNase H1 and functional inhibiton through steric blockade, respectively. The central challenge in circRNA-targeted therapy—achieving specificity for the circular isoform without degrading its linear mRNA counterpart—is addressed through thermodynamic discrimination principles. These principles, including Gibbs free energy minimization and mismatch penalty, are leveraged by back-splice junction (BSJ)-specific oligonucleotide design and locked nucleic acid (LNA) modifications. Furthermore, solution-oriented perspectives on next-generation delivery systems—such as ionizable lipid nanoparticles (LNPs), GalNAc (N-acetylglucosamine) conjugates, and exosomes—are presented to address the critical challenges of intracellular bioavailability and endosomal escape, which represent the most significant barriers to clinical translation.
Graphical AbstractThe precision and delivery triad of circRNA-targeted ASO therapy. A conceptual framework integrating target specificity, molecular mechanism, and clinical delivery for therapeutic targeting of circRNAs. (IA) circRNA biogenesis and the back-splice junction. Pre-mRNA is processed by two competing pathways. Canonical splicing joins exons in linear order to yield a 5′-capped, polyadenylated linear mRNA destined for translation. Back-splicing instead covalently joins the 3′ end of a downstream exon to the 5′ end of an upstream exon (here Exon 3 → Exon 2), generating a covalently closed circular RNA (circRNA). The resulting back-splice junction (BSJ) is a unique sequence not present in the cognate linear mRNA, making it the defining molecular signature and the sole circRNA-specific therapeutic target. (IB) Thermodynamic basis of target discrimination. An ASO complementary across the BSJ forms a perfectly matched, antiparallel duplex with the circRNA, yielding a large negative change in free energy (ΔG ↓↓) and a stable duplex. The same ASO aligned against the cognate linear mRNA encounters a central mismatch at the would-be junction, destabilising the duplex (ΔG ↑↑). This thermodynamic gap underlies the selective engagement of circRNA over its linear counterpart. (IIA) Gapmer mechanism — RNase H1-mediated cleavage. A gapmer ASO comprises chemically modified flanking wings (LNA) and a central DNA gap, hybridising across the BSJ in antiparallel orientation. RNase H1 recognises the resulting RNA:DNA heteroduplex and cleaves the circRNA strand opposite the central DNA gap. The circRNA is thereby linearised and subsequently degraded by cellular exonucleases, while the chemically protected wings leave the ASO intact for catalytic recycling and multiple-turnover activity. (IIB) Mixmer mechanism — steric blockade. A fully modified mixmer ASO (alternating LNA) cannot recruit RNase H1 and instead acts purely by steric hindrance. By hybridising across the BSJ and masking the adjacent miRNA response element (MRE), it blocks miRNA access and inhibits the circRNA’s sponging activity. The circRNA remains intact, signalling is silenced without cleavage, and the cognate linear mRNA — which lacks the BSJ — is preserved. (IIIA) Ionizable lipid nanoparticle (LNP) delivery and endosomal escape. At physiological pH 7.4 the ionizable lipids are near-neutral, minimising systemic toxicity. Following endocytosis, the acidic endosomal environment (≈ pH 5) protonates the lipid head groups; the resulting positive charge interacts with anionic endosomal lipids, disrupting the membrane. This triggers endosomal escape and releases the ASO cargo into the cytoplasm. (IIIB) Exosome-mediated (“Trojan horse”) delivery. Autologous or MSC-derived exosomes, bearing characteristic surface proteins and exhibiting low immunogenicity, encapsulate the ASO. These vesicles can cross the blood–brain barrier to target neuronal circRNAs, and deliver their cargo to the target cell by membrane fusion/endocytosis.