Exon skipping is a molecular biology technique that employs antisense oligonucleotides (AONs) to modulate pre-mRNA splicing by inducing the exclusion of specific exons from the mature mRNA, thereby altering gene expression or restoring the open reading frame in mutated genes.[1] This process mimics natural alternative splicing events but is artificially directed to target deleterious mutations, preventing the inclusion of faulty exons during the splicing of introns from pre-mRNA transcripts.[1] The mechanism relies on AONs binding to complementary sequences on the pre-mRNA, such as splice acceptor/donor sites or exonic splicing enhancer (ESE) motifs, to sterically block the recruitment of splicing factors like SR proteins and the spliceosome, resulting in the skipped exon being omitted from the final mRNA product.[1]In therapeutic applications, exon skipping has emerged as a promising treatment for genetic disorders caused by frameshift mutations, most notably Duchenne muscular dystrophy (DMD), where it aims to convert the severe phenotype to the milder Becker muscular dystrophy by producing a truncated but partially functional dystrophin protein.[2] This approach is applicable to approximately 90% of DMD cases involving out-of-frame deletions or other mutations that disrupt the dystrophin gene's reading frame, with specific exons like 51 (~13% of patients), 45 (~8%), and 53 (~8%) being prime targets.[2] As of 2025, the U.S. Food and Drug Administration (FDA) has approved four AON-based drugs for DMD, including eteplirsen (Exondys 51) in 2016 for exon 51 skipping, golodirsen (Vyondys 53) and viltolarsen (Viltepso) in 2019 and 2020 for exon 53 skipping, and casimersen (Amondys 45) in 2021 for exon 45 skipping, marking the first personalized medicines for this disease.[2]Beyond DMD, exon skipping holds potential for various cancers or hemoglobinopathies like β-thalassemia by correcting aberrant splicing.[1] Ongoing research explores multi-exon skipping strategies (e.g., exons 45-55, applicable to approximately 39% of patients with certain mutations) to broaden coverage, advanced delivery methods like peptide-conjugated AONs for better muscle targeting, and next-generation chemistries to improve efficacy and reduce dosing frequency.[2] These developments underscore exon skipping's versatility as both a therapeutic tool and a research method for dissecting splicing regulation, though challenges remain in achieving widespread applicability and long-term safety.[1]
Introduction
Definition and Principles
Exon skipping is a RNA splicing process in which one or more specific exons are excluded from the mature messenger RNA (mRNA), thereby altering the final mRNA sequence and the resulting protein product. This exclusion occurs during the splicing of precursor mRNA (pre-mRNA), where introns are removed and exons are typically joined in a linear fashion to form the coding sequence. By skipping targeted exons, the process can modify the protein's structure, length, or function, making it a key mechanism in gene expression regulation.[1]The core principles of exon skipping revolve around its potential to correct genetic defects, particularly in cases of frameshift mutations that disrupt the open reading frame (ORF) of a gene. In such scenarios, skipping an exon can restore the ORF, allowing translation of a truncated but partially functional protein instead of a non-functional one due to premature stop codons. For instance, in Duchenne muscular dystrophy (DMD) caused by dystrophingene mutations, exon skipping converts the severe out-of-frame mutation into an in-frame variant resembling the milder Becker muscular dystrophy (BMD), thereby ameliorating the disease phenotype. This approach leverages the modularity of exons to bypass deleterious mutations without altering the overall genetic code extensively.[1][3]While natural exon skipping is a physiological aspect of alternative splicing—enabling cells to generate diverse protein isoforms from a single gene through endogenous regulatory elements—therapeutic exon skipping is an engineered intervention designed to address specific pathogenic mutations. Natural processes occur without external input, driven by cellular factors like splicing enhancers and silencers, whereas therapeutic strategies artificially induce skipping to rescue protein function in genetic disorders. This distinction highlights exon skipping's evolution from a biological phenomenon to a targeted therapy.[4][1]The conceptualization of therapeutic exon skipping emerged in the early 1990s, building on observations of natural skipping events in disease contexts. In 1991, Matsuo et al. reported the first instance of exon 19 skipping in the dystrophin pre-mRNA due to an intra-exon deletion in a DMD patient, suggesting potential for manipulation. This laid the groundwork for therapeutic proposals, with Dunckley et al. demonstrating induced skipping using antisense oligonucleotides in 1998, and Matsuo formally proposing its application for DMD treatment in 1995. These developments marked the shift toward using exon skipping as a strategy for genetic diseases.[5][6][7]
Biological Context
In eukaryotic organisms, genes are organized into segments known as exons and introns, where exons represent the coding regions that are ultimately translated into proteins, while introns are non-coding sequences interspersed between them. The primary transcript, or pre-mRNA, is synthesized by RNA polymerase II as a direct copy of the gene, including both exons and introns, and undergoes processing in the nucleus before export to the cytoplasm as mature mRNA. This structure allows for the removal of introns during RNA splicing, ensuring that only the relevant exonic sequences are joined to form the functional mRNA.RNA splicing is a critical post-transcriptional modification that produces mature mRNA from pre-mRNA by excising introns and ligating exons, thereby enabling accurate protein synthesis and regulating gene expression. The process is mediated by the spliceosome, a large ribonucleoprotein complex composed of small nuclear ribonucleoproteins (snRNPs)—including U1, U2, U4, U5, and U6 snRNPs—along with numerous protein factors that recognize specific splice sites at exon-intron boundaries. These components assemble dynamically on the pre-mRNA to catalyze the two transesterification reactions that remove introns with high fidelity, preventing errors that could lead to dysfunctional proteins.Exon skipping is a naturally occurring form of alternative splicing, where one or more exons are omitted from the mature mRNA, resulting in protein isoforms with altered functionality, often in a tissue- or development-specific manner. For instance, in the human fibronectin gene, exon skipping generates isoforms that differ in their ability to bind extracellular matrix components, influencing cell adhesion in various tissues. Similarly, in the calcitonin gene, alternative splicing produces calcitonin in thyroid cells and calcitonin gene-related peptide (CGRP) in neuronal cells through exon inclusion or skipping, highlighting how this mechanism contributes to proteomic diversity without altering the genome. Such endogenous exon skipping events are regulated by splicing factors like SR proteins and hnRNPs that modulate splice site recognition.Mutations that disrupt splicing, such as those affecting splice sites, branch points, or regulatory elements, can aberrantly promote or inhibit exon skipping, leading to frameshifts, premature stop codons, or inclusion of intronic sequences in the mRNA, which often results in truncated or non-functional proteins underlying genetic diseases. For example, in certain inherited disorders, single nucleotide changes at splice junctions can cause exon skipping, producing aberrant isoforms that impair cellular function and contribute to pathology. This splicing vulnerability underscores the precision required in eukaryotic gene expression and the potential for therapeutic modulation to restore proper exon inclusion patterns.
Mechanism of Action
Pre-mRNA Splicing Basics
Pre-mRNA splicing is a critical eukaryotic process that removes non-coding introns from primary transcripts (pre-mRNAs) and joins coding exons to form mature messenger RNAs (mRNAs), enabling proper protein synthesis. This process occurs co-transcriptionally in the nucleus and is catalyzed by the spliceosome, a large ribonucleoprotein (RNP) complex composed of five small nuclear RNAs (snRNAs: U1, U2, U4, U5, U6) and over 100 associated proteins. The spliceosome assembles de novo on each intron in a stepwise manner, ensuring precise recognition and excision of introns while preserving exon integrity.[8]The splicing reaction proceeds via two sequential transesterification steps. First, the 5' splice site (5' SS), typically marked by a conserved GU dinucleotide, is recognized by base-pairing with U1 snRNA, while the 3' splice site (3' SS), ending in an AG dinucleotide, and the upstream branch point sequence (BPS)—often containing an adenine—are identified through interactions with U2 snRNA and auxiliary factors. In the initial step, the 2'-hydroxyl group of the BPS adenosine performs a nucleophilic attack on the 5' SS phosphate, cleaving the 5' exon and forming a lariat structure where the intron 5' end is linked via a 2'-5' phosphodiester bond to the BPS. The second step involves the 3'-hydroxyl of the freed 5' exon attacking the 3' SS, ligating the exons and releasing the intron lariat for degradation. These steps are highly conserved across eukaryotes, with the BPS-3' SS distance typically 18–40 nucleotides in mammals.[8][9][10]Spliceosome assembly occurs in an ordered, ATP-dependent pathway involving four main complexes: commitment (E), pre-spliceosome (A), pre-catalytic (B), and catalytic (C). The E complex forms when U1 snRNP binds the 5' SS and U2 auxiliary factor (U2AF) recognizes the polypyrimidine tract upstream of the 3' SS, with branch point binding protein (BBP/SF1) aiding BPS identification. The A complex follows as U2 snRNP displaces BBP and base-pairs with the BPS, forming a bulged adenosine essential for lariat formation. The B complex assembles upon recruitment of the U4/U6.U5 tri-snRNP, which includes U4, U5, and U6 snRNAs; at this stage, the complex is catalytically inactive. Activation to B* involves release of U1 snRNP, unwinding of U4/U6 base-pairing by the ATPase Brr2, and recruitment of nineteen complex (NTC) proteins, positioning the reactive groups for catalysis. The C complex catalyzes the splicing steps, followed by disassembly to recycle components. This dynamic remodeling, driven by DExD/H-box ATPases like Prp2, Prp16, and Prp22, ensures fidelity and directionality.[8][9][10]Exon inclusion is regulated by splicing factors such as serine/arginine-rich (SR) proteins and heterogeneous nuclear ribonucleoproteins (hnRNPs), which bind exonic splicing enhancers (ESEs) and silencers (ESSs), respectively. SR proteins, through their RNA recognition motifs and RS domains, promote exon definition by stabilizing interactions between U1/U2 snRNPs across exons, facilitating splice site pairing in genes with long introns. In contrast, hnRNPs often antagonize SR proteins by binding ESSs or intronic splicing silencers (ISSs), repressing weak splice sites. Splice site selection is influenced by the exon definition model, prevalent in vertebrates with large introns, where factors bridge 3' SS and downstream 5' SS across the exon; this contrasts with the intron definition model in organisms like yeast with short introns, where assembly spans intron boundaries directly. Other factors include splice site strength (consensus sequence match), secondary structure, intron length, and cellular concentrations of splicing regulators.[8][9][10]Errors in splicing, such as aberrant splice site recognition or skipping, can introduce premature termination codons (PTCs) in the mRNA, triggering nonsense-mediated decay (NMD)—a surveillance pathway that degrades such transcripts via recognition of PTCs more than 50–55 nucleotides upstream of an exon junction. Retained introns or exon truncations may also lead to frameshifts, producing aberrant truncated proteins with dominant-negative effects or loss of function, contributing to diseases like spinal muscular atrophy. Accurate splicing fidelity is maintained by kinetic proofreading during spliceosome rearrangements and discard pathways for suboptimal complexes.[8][9][10]
Induced Skipping Process
Induced exon skipping is achieved through the use of therapeutic agents, primarily antisense oligonucleotides (AONs), that interfere with the pre-mRNA splicing machinery to exclude specific exons from the mature mRNA transcript. These AONs are designed to hybridize to target sequences within the pre-mRNA, such as exonic splicing enhancer (ESE) motifs or splice donor and acceptor sites, thereby masking them from recognition by splicing factors like SR proteins (e.g., SF2/ASF or SC35). By blocking the binding of these regulatory proteins, AONs prevent the inclusion of the targeted exon during spliceosome assembly, leading to its exclusion from the final mRNA product. This process exploits the natural flexibility of alternative splicing to redirect the spliceosome toward exon-skipping pathways.[1]The primary outcome of induced exon skipping is the removal of the targeted exon, which can restore the open reading frame in transcripts affected by out-of-frame mutations, such as deletions or nonsense mutations common in diseases like Duchenne muscular dystrophy (DMD). For instance, skipping an exon that disrupts the reading frame allows the production of a truncated but partially functional protein, with restoration levels often reaching 1-3% of normal in preclinical models. In cases of complex mutations involving multiple exons, multi-exon skipping can be induced by administering cocktails of AONs targeting adjacent exons, enabling frame correction for a broader range of genetic lesions and potentially benefiting up to 14% of deletion patients. This approach maintains protein functionality while avoiding complete loss of expression.[1][11]Specificity in induced exon skipping relies on the sequence complementarity between the AON and its target, typically requiring 17-25 nucleotides of perfect or near-perfect matching to ensure selective binding and minimize off-target effects. Off-target skipping is reduced by selecting AONs that avoid homology with non-target transcripts (e.g., less than 17 consecutive nucleotides) and by targeting ESEs with high SR protein affinity, often located within 70 nucleotides of the acceptor splice site, which enhances efficiency while limiting unintended splicing alterations. Computational tools like GGGenome are employed to predict and avoid potential off-target binding sites during AON design.[1][12]Validation of induced exon skipping occurs through in vitro and in vivo methods that confirm the exclusion of targeted exons and restoration of protein function. In vitro assays using cultured cells, such as myotubes from patient-derived samples, employ reverse transcription polymerase chain reaction (RT-PCR) to detect skipped transcripts via size-shifted amplicons, often followed by sequencing to verify precise exon junctions. In vivo studies in animal models, like the mdx mouse for DMD, combine RT-PCR with Western blotting to quantify dystrophin protein levels and immunofluorescence to assess localization, ensuring the intervention yields functional outcomes without aberrant splicing. These methods have demonstrated consistent exon exclusion rates of up to 70% in optimized systems.[1][11]
Therapeutic Strategies
Antisense Oligonucleotides
Antisense oligonucleotides (AONs) serve as the cornerstone therapeutic agents in exon skipping therapies, functioning as synthetic nucleic acid analogs that bind to specific sequences in pre-mRNA to sterically hinder splicing factors and induce the exclusion of targeted exons during mRNA maturation.[13] These molecules are designed to restore the reading frame in mutated transcripts without altering the underlying DNA, offering a targeted approach to mitigate genetic disorders caused by frameshift mutations.[14]The design of AONs for exon skipping follows established principles to ensure effective binding and splicing modulation. Typically, AONs range from 15 to 30 nucleotides in length, with an optimal length of around 20 nucleotides to balance specificity, bindingaffinity, and manufacturability.[15] Target sites are strategically selected at splice acceptor or donor sites, exonic splicing enhancer (ESE) motifs, or intronic splicing silencers, with sequences proximal to the 3' acceptor site often yielding higher skipping efficiency due to their accessibility and influence on spliceosome assembly.[1] Computational tools, such as those predicting ESE motifs like RESCUE-ESE hexamers, aid in identifying high-potency targets while minimizing off-target effects.[15]Advancements in AON chemistry have evolved across generations to overcome the limitations of early designs, particularly in stability and cellular uptake. First-generation AONs, based on DNA with phosphorothioate (PS) backbones, provided basic nuclease resistance by replacing phosphodiester linkages with sulfur-containing bonds but suffered from rapid degradation, poor specificity, and immune activation due to their polyanionic nature.[16] Second-generation modifications, such as 2'-O-methyl (2'-OMe) ribose sugars combined with PS backbones, enhanced enzymatic stability and reduced immunogenicity while enabling steric blocking without RNase H cleavage, making them suitable for splicing modulation.[16] Third-generation chemistries further improved performance; morpholino oligomers (PMOs) feature a neutral phosphorodiamidate backbone and morpholine rings instead of ribose, conferring exceptional resistance to nucleases and proteases, high aqueous solubility, and minimal protein binding, often achieving superior efficacy in nuclear pre-mRNA targeting.[14] Locked nucleic acids (LNAs), with a methylene bridge locking the ribose in a C3'-endo conformation, offer increased binding affinity (Tm up to 10°C higher than DNA) and metabolic stability, though careful designs, such as chimeric combinations with other modifications like 2'-O-methyl, are required to mitigate potential toxicity such as hepatotoxicity.[14]Specificity and potency of AONs are rigorously evaluated in preclinical cell models to confirm targeted exon skipping without unintended transcript alterations. Human-derived cell lines, such as fibroblasts or myoblasts from relevant tissues, are transfected or exposed to gymnotic uptake of AONs at concentrations of 100 nM to 10 μM, followed by RNA extraction and reverse transcription polymerase chain reaction (RT-PCR) to quantify skipping efficiency as a measure for potential therapeutic advancement.[17]Western blotting assesses downstream protein restoration, while sequence-specific controls and bioinformatics predict off-target binding to ensure minimal genome-wide splicing disruptions.[17] These in vitro assays, often using patient-derived or CRISPR-edited cells, have validated PMOs and LNAs as outperforming earlier generations in potency, with skipping efficiencies exceeding 50% in optimized models.[14]
Delivery and Enhancement Techniques
Delivery of antisense oligonucleotides (AONs) for exon skipping primarily occurs through systemic or local routes, each presenting distinct advantages and challenges. Systemic delivery, typically via intravenous administration, enables broad tissue distribution but faces significant barriers, such as limited penetration across the blood-brain barrier (BBB) for central nervous system targeting and suboptimal uptake in cardiac muscle due to endothelial barriers and rapid clearance.[18][19] Local delivery, such as intramuscular injection, achieves higher concentrations in skeletal muscle but is less practical for widespread or multi-tissue applications in conditions like Duchenne muscular dystrophy (DMD).[20]To enhance AON delivery and efficacy, various strategies have been developed to improve cellular uptake, nuclear localization, and tissue specificity. Conjugation to cell-penetrating peptides (CPPs), such as cyclic peptides, facilitates endosomal escape and increases AON accumulation in target cells, including muscle tissues.[21] Lipid nanoparticles (LNPs) encapsulate AONs to protect them from degradation, enhance stability, and promote targeted delivery to skeletal and cardiac muscles via intravenous routes.[20] Viral vectors, particularly adeno-associated virus (AAV) serotypes like AAV9, offer long-term expression of AONs or splice-switching RNAs, though their immunogenicity and size limitations pose challenges for larger AON payloads.[22]Recent innovations from 2023 to 2025 have focused on optimizing nuclear uptake and multi-exon targeting. The DG9 peptide-conjugated phosphorodiamidate morpholino oligomer (DG9-PMO) enhances PMO nuclear import, leading to superior exon skipping and dystrophin restoration in both skeletal and cardiac muscles in preclinical DMD models.[23] Brogidirsen, a dual-targeting AON, simultaneously binds two sites on pre-mRNA to improve exon 44 skipping efficiency, demonstrating enhanced potency over single-target designs in early clinical evaluations.[24]Pharmacokinetics of AONs in exon skipping therapies are characterized by rapid plasma clearance followed by prolonged tissue retention, influencing dosing regimens. Plasma half-lives are short, typically 0.5–1.5 hours for initial distribution, with terminal half-lives in tissues ranging from 29–65 days, particularly longer in heart compared to skeletal muscle or liver.[25][26] Dosing often involves weekly or monthly intravenous infusions at 10–30 mg/kg to maintain therapeutic levels, with preferential accumulation in kidney, liver, and muscle, though cardiac distribution remains a focus for improvement.[27][28]
Applications in Disease Treatment
Duchenne Muscular Dystrophy
Duchenne muscular dystrophy (DMD) is caused by mutations in the DMD gene on the X chromosome, which encodes the dystrophin protein essential for muscle fiber integrity. Approximately 65-70% of these mutations are large deletions or duplications of one or more exons, with the remainder consisting of small mutations such as point mutations or insertions. These genetic alterations disrupt the open reading frame, leading to premature termination of translation and absence of functional dystrophin, resulting in progressive muscle degeneration.[29][29]A significant proportion of DMD mutations—particularly deletions and duplications affecting exons 45-53—are amenable to exon skipping therapy, potentially benefiting around 50% of patients by restoring the reading frame. The therapeutic rationale involves using antisense oligonucleotides to induce skipping of specific mutated or frame-disrupting exons during pre-mRNA splicing, thereby allowing production of a truncated but partially functional dystrophin protein analogous to that in the milder Becker muscular dystrophy (BMD). This shorter dystrophin retains key structural domains, enabling improved sarcolemmal stability and reduced muscle damage. Emerging applications include exon 44 skipping, with positive phase 1/2 trial data reported in 2024 and FDA Breakthrough Therapy Designation granted in 2025.[30][31][32][33]Mutation hotspots within the DMD gene make certain exons prime targets for skipping. Deletions amenable to exon 51 skipping occur in approximately 13% of patients, while those suitable for exon 45 or 53 skipping affect about 8% each, collectively addressing a substantial subset of cases concentrated in the distal hotspot region spanning exons 45-55. These percentages highlight the potential for personalized exon skipping strategies based on individual mutation profiles.[34][29]Preclinical studies in animal models, such as mdx mice carrying a dystrophinmutation analogous to human DMD, have demonstrated that exon skipping restores dystrophin expression and enhances muscle function. In mdx52 mice, which model exon 52 deletions amenable to exon 51 skipping, systemic administration of antisense oligonucleotides led to significant improvements in muscle histology, contractile force, and overall pathology, supporting the approach's efficacy in mitigating disease progression. Exon skipping therapies targeting exons 45, 51, and 53 have been approved for applicable DMD patients.[31]
Other Genetic Disorders
Exon skipping has shown promise in treating various monogenic disorders beyond Duchenne muscular dystrophy, particularly those where mutations disrupt protein function in genes encoding modular proteins amenable to truncation without complete loss of activity. This approach is viable when the targeted protein consists of distinct functional domains, allowing the exclusion of a mutant exon to produce a shortened isoform that retains partial functionality and avoids nonsense-mediated decay. Such criteria emphasize diseases with out-of-frame mutations or splice-site defects that can be corrected by restoring the reading frame through targeted skipping.[35]In USH2A-related retinitis pigmentosa, a form of Usher syndrome type 2, exon skipping targets mutations in exon 13 of the USH2A gene, which encodes usherin, a protein essential for retinal ciliary function. Preclinical studies using antisense oligonucleotides (AONs) have demonstrated efficient skipping of exon 13 in patient-derived retinal organoids and humanized mouse models, achieving up to 44% efficiency in vivo and restoring expression of related proteins like GPR98 and PDZD7 while improving ciliary structure without notable toxicity. These 2025 investigations highlight the potential for AON-based therapies to mitigate photoreceptor degeneration in affected individuals. As of November 2025, a phase 2b clinical trial for an exon 13 skipping AON was launched in December 2024.[36][37]Related splice-switching strategies, which inversely promote exon inclusion rather than skipping, have been explored in spinal muscular atrophy (SMA) and myotonic dystrophy type 1 (DM1). In SMA, AONs enhance inclusion of exon 7 in the SMN2 gene to boost production of full-length survival motor neuron (SMN) protein, addressing the natural skipping caused by a C-to-T transition; early studies reported up to 50% correction of splicing defects in cell models. Similarly, in DM1, AON-mediated splice-switching corrects aberrant inclusion patterns in the DMPK gene and other mis-spliced transcripts, inducing skipping of toxic expanded repeat-containing exons to alleviate RNA toxicity and restore normal protein isoforms in preclinical myoblast models.[38][39]Emerging candidates for exon skipping include beta-thalassemia, where mutations in the HBB gene lead to defective beta-globin production, and Huntington's disease (HD), involving expanded CAG repeats in the HTT gene. In beta-thalassemia, potential strategies aim to skip exons harboring splice-site mutations to normalize globin chain synthesis, though primarily conceptual at present with focus on frame-restoring approaches. For HD, partial exon exclusion via AONs targets exon 12 to disrupt caspase-6 cleavage sites in mutant huntingtin, reducing toxic N-terminal fragments and improving neuronal phenotypes in mouse models without affecting wild-type protein. These applications underscore the expanding scope of exon skipping for disorders with truncatable protein architectures.[40]
Clinical Development
Approved Therapies
The first FDA-approved exon skipping therapy was eteplirsen (Exondys 51), developed by Sarepta Therapeutics, which received accelerated approval on September 19, 2016, for the treatment of Duchenne muscular dystrophy (DMD) in patients with a confirmed mutation in the DMD gene amenable to exon 51 skipping.[41] This approval marked a milestone as the inaugural antisense oligonucleotide-based therapy for DMD, though it sparked significant controversy due to the limited clinical data available at the time, including a small open-label study showing only modest increases in dystrophin production without clear functional benefits.[42] The FDA's decision overrode internal reviewers' recommendations, relying on surrogate endpoints like dystrophin levels in skeletal muscle as a basis for accelerated approval, with the requirement for confirmatory trials to verify clinical benefit.[43]Subsequent approvals followed a similar accelerated pathway for additional exon skipping agents targeting different mutations in DMD patients. Golodirsen (Vyondys 53), also from Sarepta Therapeutics, was approved on December 12, 2019, for DMD patients amenable to exon 53 skipping; like eteplirsen, it is a phosphorodiamidate morpholino oligomer (PMO) that demonstrated increased dystrophin production in skeletal muscle biopsies from a small study of 12 patients.[44] Viltolarsen (Viltepso), developed by NS Pharma, received FDA approval on August 12, 2020, for the same exon 53 indication in DMD; phase 1/2 data from eight patients supported approval via surrogate dystrophin elevation, confirming its safety and tolerability profile.[45] Casimersen (Amondys 45), another Sarepta PMO, was approved on February 25, 2021, targeting exon 45 skipping in DMD patients, based on dystrophin increases observed in muscle biopsies from 12 participants in an open-label study.[46]These four therapies—eteplirsen, golodirsen, viltolarsen, and casimersen—represent all FDA-approved exon skipping drugs as of November 2025, each granted under the accelerated approval pathway that utilizes increases in dystrophin production as a surrogate endpoint reasonably likely to predict clinical benefit, amid the absence of curative options for DMD.[47] No additional exon skipping therapies achieved full or accelerated FDA approval between 2022 and 2025.[48] However, in September 2025, the FDA granted orphan drug designation to NS-051/NCNP-04, an investigational PMO from NS Pharma targeting exon 51 skipping in DMD, providing incentives for its further development following prior rare pediatric disease designation earlier that year.[49]
Recent phase 1/2 clinical trials have advanced exon skipping therapies for Duchenne muscular dystrophy (DMD), particularly targeting exon 44 mutations. Brogidirsen (NS-089/NCNP-02), a dual-targeting antisense oligonucleotide (ASO), is under investigation in the phase 1/2 PINPOINT trial for boys aged 4-15 amenable to exon 44 skipping. In 2025 interim data from six patients, brogidirsen induced a 32% increase in exon 44 skipping and restored dystrophin expression, with long-term 3.5-year follow-up showing sustained benefits and slowed disease progression without significant safety issues.[50]00672-4)[51]Avidity Biosciences' del-zota (delpacibart zotadirsen, AOC 1044), an antibody-oligonucleotide conjugate designed for enhanced muscle delivery, targets exon 44 skipping in the phase 1/2 EXPLORE44 trial. 2025 results demonstrated dose-dependent exon 44 skipping up to 74%, dystrophin production increases of 25% in skeletal muscle, and near-normalization of creatine kinase levels after multiple doses, alongside functional improvements in ambulation and cardiac efficacy markers such as reduced left ventricular strain. In July 2025, the FDA granted Breakthrough Therapy Designation to del-zota based on these promising early data. The therapy was well-tolerated, supporting ongoing open-label extension studies.[52][53][54][55]Expansion efforts in DMD exon skipping include trials exploring multi-exon combinations to address a broader patient population beyond single-exon targets. For instance, preclinical and early-phase strategies aim to combine ASOs for exons 44-46 or 51-53 skipping, potentially covering up to 60% of DMD cases by enabling frame restoration across diverse mutations, with 2024-2025 analyses confirming feasibility in patient-derived models.[2][56]In preclinical research, exon skipping for USH2A-related Usher syndrome type 2A has shown promise for restoring retinal function. 2025 studies using antisense oligonucleotides targeting exon 13 demonstrated efficient skipping in patient-derived retinal organoids, leading to production of functional usherin protein and partial recovery of ciliary motility essential for photoreceptor health. AAV-mediated approaches further enhanced exon excision in murine models, supporting progression toward clinical translation.[36][57][58]
Challenges and Future Directions
Limitations and Safety Concerns
Exon skipping therapies for Duchenne muscular dystrophy (DMD) face significant efficacy limitations, primarily due to the low levels of dystrophin restoration achieved in clinical settings, typically ranging from 0.9% to 7.8% of normal levels depending on the specific antisense oligonucleotide (AON) used.[59][59] Although preclinical and some clinical data suggest that even 5-10% dystrophin expression may confer functional benefits akin to milder Becker muscular dystrophy phenotypes, the partial restoration often fails to substantially halt disease progression in most patients.[60] Patient responses vary widely, influenced by factors such as mutation type, age at treatment initiation, and residual muscle integrity, leading to inconsistent therapeutic outcomes across individuals.[61]Mutation coverage remains incomplete, with current approved therapies applicable to only about 13-30% of DMD patients harboring amenable deletions, particularly those eligible for single-exon skipping like exon 51.[62] This restriction arises because exon skipping requires precise alignment of the mutation with target exons to restore the reading frame, excluding the majority of patients with duplications, nonsense mutations, or other variants.[2]Safety concerns include renal toxicity associated with phosphorothioate (PS)-modified AONs, as observed in trials of drisapersen, where elevated proteinuria and histopathological kidney changes necessitated dose limitations and contributed to its discontinuation.[63] Injection-site reactions, such as erythema, hyperpigmentation, and ulceration, are common with subcutaneous administration and can persist or worsen over time, affecting patient compliance.[64] Long-term immunogenicity poses a potential risk, particularly with repeated dosing, although phosphorodiamidate morpholino oligomers (PMOs) used in approved therapies like eteplirsen show minimal immune activation compared to earlier PS-AONs.[65]Off-target effects represent another hazard, where AONs may hybridize to unintended transcripts, causing aberrant exon skipping and production of novel, potentially dysfunctional protein isoforms that could exacerbate pathology.[66]Access barriers further limit widespread adoption, including exorbitant costs exceeding $1 million annually per patient for therapies like eteplirsen, compounded by the need for lifelong weekly intravenous infusions.[67][68] These factors, alongside delivery challenges such as poor muscle penetration, restrict equitable access primarily to patients with specific mutations in resource-rich settings.[69]
Emerging Advances
Recent advancements in exon skipping therapies are focusing on next-generation delivery systems to enhance tissue specificity and efficacy. Antibody-oligonucleotide conjugates (AOCs), which link antisense oligonucleotides (AONs) to monoclonal antibodies targeting muscle-specific proteins like transferrin receptor 1, have shown promise in improving delivery to skeletal and cardiac muscles in preclinical models of Duchenne muscular dystrophy (DMD). For instance, AOC 1044 achieved up to 50-fold higher phosphorodiamidate morpholino oligomer (PMO) accumulation in muscles compared to unconjugated PMOs, leading to enhanced exon 51 skipping and dystrophin restoration. As of October 2025, phase 1/2 trial data for del-zota (AOC 1044 targeting exon 44) demonstrated a statistically significant 25% increase in dystrophin production and functional gains in DMD patients.[70][71][72]CRISPR-based approaches are enabling permanent exon skipping by directly editing genomic DNA to delete mutated exons, offering a one-time treatment potential that overcomes the need for repeated AON dosing. Dual CRISPR-Cas3 systems have demonstrated efficient multi-exon deletions in DMD models, restoring the dystrophin reading frame with minimal off-target effects and sustained protein expression.[73] Base editors, such as those disrupting exonic splicing enhancers, have induced targeted exon skipping in up to 70% of cells, providing precise control for frame restoration in genetic disorders.[74] In 2024-2025, clinical trials such as HG302 using CRISPR/Cas12 for exon 51 skipping have initiated dosing in DMD patients.[75]To expand applicability, artificial intelligence (AI) tools are being developed for designing personalized AONs that optimize exon skipping efficiency based on patient-specific mutations. Platforms like eSkip-Finder and ASOptimizer use machine learning to predict optimal AON sequences, reducing design time and improving skipping rates by analyzing splicing patterns and binding affinities from large datasets.[76][77] Combining exon skipping with gene editing further broadens therapeutic scope; for example, prime editing strategies have corrected frameshift mutations in the DMD gene by inducing precise exon skips alongside base corrections.[78]Developments from 2024 to 2025 highlight improvements in PMO uptake, particularly for cardiac targeting. The DG9 peptide-PMO conjugate has boosted nuclear delivery in cardiomyocytes, achieving 41% dystrophin expression in heart tissue of DMD mouse models and improving cardiac function without toxicity.[23] Multi-disease platforms leveraging AON bioconjugation and AI optimization are emerging, enabling adaptable exon skipping strategies across genetic disorders like DMD and beyond, with preclinical data supporting broader tissue penetration and reduced dosing frequency.[79]Exon skipping holds potential for non-neuromuscular applications, such as oncology, where inducing skips in oncogenic transcripts could suppress tumor growth.