Muscle-specific gene editing therapy via mammalian fusogen-directed virus-like particles

Zhou, Shi-Kun; Luo, Jing-Tong; Chen, Yi-Fang; Lu, Zi-Dong; Jian, Qiu-Hong; Jiang, Shui-Qing; Li, Jing; Zhang, Xue-Qin; Tan, Xin-Yu; Yang, Xian-Zhu; Xu, Cong-Fei; Wang, Jun

doi:10.1038/s41467-025-64200-9

Download PDF

Article
Open access
Published: 15 October 2025

Muscle-specific gene editing therapy via mammalian fusogen-directed virus-like particles

Nature Communications volume 16, Article number: 9145 (2025) Cite this article

Subjects

Abstract

Muscle genetic defects can lead to impaired movement, respiratory failure, and other severe symptoms. The development of curative therapies is challenging due to the need for the delivery of gene-editing tools into skeletal muscle cells throughout the body. Here, we use muscular fusogens (Myomaker and Myomerger) to engineer muscle-specific virus-like particles (MuVLPs) for the systemic delivery of gene-editing tools. We demonstrate that MuVLPs can be loaded with diverse payloads, including EGFP, Cre and Cas9/sgRNA ribonucleoproteins (Cas9 RNPs), and can be delivered into skeletal muscle cells via targeted membrane fusion. Systemic administration of MuVLPs carrying Cas9 RNPs enables skeletal muscle-specific gene editing, which excised the exon containing a premature terminator codon mutation in a mouse model for Duchenne muscular dystrophy (DMD). This treatment restores dystrophin expression in various skeletal muscle tissues, including the diaphragm, quadriceps, tibialis anterior, gastrocnemius, and triceps. As a result, the treated mice exhibit a significantly increased capacity for exercise and endurance. This study established a platform for precise gene editing in skeletal muscle tissues.

Introduction

Muscle tissues play critical roles in generating force, enabling movement, and maintaining body posture^1,2,3. Muscle defects can lead to devastating diseases such as muscular dystrophy, highlighting the need for effective gene editing therapies^4,5. However, accurately delivering gene-editing tools to muscle tissues is challenging owing to the extensive distribution of these tissues throughout the body^2,6.

The most effective methods for delivering gene-editing tools to muscle tissues in vivo involve the use of viruses, such as adeno-associated viruses (AAVs), to deliver CRISPR-Cas9-encoding DNA^7,8. Viral vectors, particularly lentiviruses (LVs), have been pseudotyped with various envelope proteins to enable targeting of specific cell types. For instance, VSV-G confers broad tropism⁹, while engineered glycoproteins (e.g., rabies glycoprotein binding to neurons)¹⁰ or tissue-specific ligands (e.g., CD4-targeting variants)¹¹ enhance specificity. However, muscle-targeted pseudotyping of current viral vectors remains understudied, and these methods can lead to prolonged expression and off-target effects^12,13. In contrast, direct delivery of CRISPR-Cas9 ribonucleoproteins (RNPs) using virus-like particles (VLPs) has shown promise in reducing these issues, demonstrating success in delivering gene editing tools to various tissues and organs^14,15,16,17. Nonetheless, the inability to target VLPs specifically and effectively to skeletal muscles remains the main obstacle for this class of delivery vehicle^18,19,20.

In this study, we develop skeletal muscle-specific virus-like particles (MuVLPs) that can effectively carry different genetic modifiers and undergo membrane fusion with skeletal muscle cells for targeted delivery using muscular fusogens (Myomaker and Myomerger). When administered intravenously, MuVLPs carrying Cas9/sgRNA RNPs (MuVLP_Cas9) can selectively edit skeletal muscle cells throughout the body to excise mutant exons and thus restore dystrophin expression and treat dystrophy in a mouse model for DMD. Our findings demonstrate the potential of this targeted gene editing approach to address genetic muscle disorders.

Results

Engineering the skeletal muscle-specific delivery platform MuVLPs

VLPs are collections of viral proteins that can infect cells but lack viral genetic materials and have become promising vectors for delivering gene-editing tools^15,21. In this study, we generated HIV-1-based VLPs primarily composed of the Gag polyprotein (p55), which forms the core structure of the particles. The Gag polyprotein is processed during particle maturation by the HIV-1 protease, yielding distinct subunits: p17 (matrix, MA) and p24 (capsid, CA). The MA domain anchors the Gag polyprotein to the cell membrane during assembly, while CA forms the conical core encapsulating the payload²². Critically, these VLPs lack lentiviral long terminal repeats (LTRs), ensuring no genomic integration or viral replication (Fig. 1a).

**Fig. 1: MuVLP-mediated protein delivery is highly skeletal muscle specific.**

The cellular tropism of VLPs can be modulated by changing their envelope glycoproteins⁹. Myomaker and Myomerger are muscular fusogens essential for the development and regeneration of skeletal muscles and can be used as pseudotype lentiviruses for muscle transduction²³. To investigate whether VLPs can be programmed to target muscle cells, we produced HIV-1 VLPs pseudotyped with Myomaker and Myomerger (denoted as MuVLPs) by mimicking the process of producing lentiviral vectors. As shown in Fig. 1a, plasmids expressing Myomaker and Myomerger were used to replace vesicular stomatitis virus glycoprotein (VSV-G) for modifying MuVLPs. Given that Myomaker and Myomerger are essential membrane proteins for myoblast fusion and can induce fusion in nonfusing cells^{24,25,26,27,28}, we reasoned that muscular fusogens could facilitate the fusion of VLPs with muscle cells. The production process involved co-transfecting HEK-293T cells with plasmids encoding (1) Gag-Pol for particle assembly, (2) Myomaker and Myomerger for membrane fusion, and (3) a payload protein fused to the C-terminus of Gag via a protease-cleavable linker. Upon maturation, the HIV-1 protease cleaves the linker, releasing the payload into the particle core while retaining structural components (p17, p24). Western blotting analysis further validated the presence of Gag-derived p17 and p24, as well as Myomaker/Myomerger on the particle surface (Fig. 1b). The morphology of the MuVLPs was spherical and relatively uniform, with a particle size of 100–150 nm (Fig. 1c). These results demonstrate the successful incorporation of muscular fusogens into VLPs while preserving their structural integrity.

To investigate the membrane fusion specificity of MuVLPs, we generated a C2C12 cell line that can inducibly express Myomaker and Myomerger upon doxycycline treatment (iC2C12), enabling the assessment of membrane fusion efficacy in Myomaker⁺Myomerger⁺ target cells and Myomaker^-Myomerger^- nontarget cells (C2C12, NIH/3T3, A9, DC2.4, and RAW264.7) (Fig. 1d and Supplementary Fig. 1). After incubating DiD-labeled VLPs or DiD-labeled MuVLPs (whose envelopes were stained with DiD dye) with six types of DiI-labeled cells, only the iC2C12 cell membrane (green) showed colocalization with the DiD-labeled MuVLPs (red), resulting in a merged yellow color indicative of membrane fusion (Fig. 1e and Supplementary Fig. 2). No DiD signals were observed in either the Myomaker⁺Myomerger⁺ or Myomaker⁻Myomerger⁻ cells incubated with DiD-labeled VLPs. The membrane fusion efficacy of MuVLPs was measurable, with up to 69% of target cells exhibiting DiD signals, while little to no signal was detected in nontarget cells (Fig. 1f). In contrast, cells co-incubated with DiD-labeled VSVG-VLPs exhibited widespread DiD signals (e.g., iC2C12 (98%), C2C12 (89%), NIH/3T3 (52%) in vitro) but lacked muscle specificity (Fig. 1f). Subsequently, for analysis of the muscle specificity of MuVLPs in vivo, DiD-labeled VLPs, DiD-labeled MuVLPs or DiD-labeled VSVG-VLPs were intravenously injected into B10-Dmd-KO mice harboring a mutation in exon 4 of the Dmd gene resulting in the loss of dystrophin protein expression. After 8 h, VLPs predominantly accumulated in the liver, kidneys and lungs and showed little accumulation in the heart, spleen and skeletal muscle. Notably, MuVLPs-treated mice showed 5.15-fold higher DiD fluorescence intensity in the diaphragm (Diaph.) compared to VLPs-treated mice, and 3.44-fold higher than VSVG-VLPs-treated mice. Additionally, MuVLPs-treated mice showed significantly higher DiD fluorescence intensity in various skeletal muscles (e.g., tibialis anterior (TA), gastrocnemius (Gas.), quadriceps (Quad.), triceps (Tri.) compared to VLPs-treated mice and VSVG-VLPs-treated mice. (Fig. 1g and Supplementary Fig. 3). These results demonstrated that MuVLPs exhibit specific tropism toward cells expressing muscular fusogens and skeletal muscle tissues.

Finally, to investigate whether MuVLPs can selectively deliver proteins to muscle cells, we generated a plasmid expressing EGFP with the C-terminus of Gag to package EGFP into VLPs (VLP_EGFP) and MuVLPs (MuVLP_EGFP) (Fig. 1h). As shown in Supplementary Fig. 4, we used Flow NanoAnalyzer to quantify particle concentration. Approximately 3.5 × 10¹⁰ MuVLPs are produced per 1 × 10⁷ HEK-293T cells. The EGFP fluorescence signal of the Bald-VLPs was consistent with the background noise level, with positive particles accounting for less than 1%. In contrast, VLP_EGFP, MuVLP_EGFP, and VSVG-VLP_EGFP showed positive particle percentages of 96%, 97%, and 96%, respectively (Supplementary Fig. 4b). After constructing and characterizing the VLP_EGFP, MuVLP_EGFP and VSVG-VLP_EGFP, we investigated the protein delivery specificity of MuVLP_EGFP by coincubation with six types of cells. After incubation with MuVLP_EGFP, EGFP was only significantly observed in iC2C12 cells (Fig. 1i and Supplementary Fig. 5), and the mean fluorescence intensity of EGFP in iC2C12 cells was also significantly greater than that in C2C12 cells, NIH/3T3 cells and other cells (Fig. 1j). In contrast, VSVG-VLP_EGFP efficiently transduced various cell types (e.g., C2C12, NIH/3T3) in vitro but lacked muscle specificity (Fig. 1j). Collectively, these results indicated the ability of MuVLPs to selectively fuse with skeletal muscle cells and deliver proteins specifically, laying the foundation for skeletal muscle-specific gene editing therapy.

MuVLP_Cre induces skeletal muscle cell-specific gene recombination

Subsequently, to investigate whether MuVLPs can induce gene recombination in muscle cells, a plasmid expressing Cre recombinase with the C-terminus of Gag was constructed to package Cre recombinase into VLPs (VLP_Cre) and MuVLPs (MuVLP_Cre) (Fig. 2a). As demonstrated in Supplementary Fig. 6, the presence of viral components and muscular fusogens on the MuVLPs and MuVLP_Cre was confirmed, as was the effective loading of Cre recombinase into VLP_Cre and MuVLP_Cre. To assess whether MuVLP_Cre could deliver Cre recombinase into skeletal muscle cells and exert its functionality, we generated B6-G/R;B10-Dmd-KO mice by mating female B10-Dmd-KO mice with male B6-G/R mice, which enabled us to trace the activity of Cre recombinase (Supplementary Fig. 7a). Primary skeletal muscle cells were isolated from the tibialis anterior muscle of B6-G/R;B10-Dmd-KO mice and cocultured with VLPs, MuVLPs, VLP_Cre, or MuVLP_Cre. After 72 h, red fluorescence was exclusively observed in skeletal muscle primary cells cocultured with MuVLP_Cre (Supplementary Fig. 7b), indicating the ability of the MuVLP_Cre to transduce skeletal muscle cells.

**Fig. 2: MuVLP_Cre-mediated targeted transduction of myogenic cells in vivo and in vitro.**

α7-Integrin⁺ myogenic cells regulate muscle cell functions, including proliferation, differentiation, and tissue repair, and play a vital role in muscle development and repair²⁹. Having observed the capacity of MuVLP_Cre to induce gene recombination in skeletal muscle cells, we postulated that MuVLP_Cre could be transduced into myogenic progenitor cells within muscle tissue. Subsequently, after intramuscular administration of MuVLP_Cre into the tibialis anterior muscle, tdTomato⁺ α7-integrin⁺ myogenic progenitors in B6-G/R;B10-Dmd-KO mice were analyzed seven days post-treatment (Fig. 2b). Standard flow cytometric gating strategies were utilized to identify α7-integrin⁺ myogenic cells³⁰ (Supplementary Fig. 8). As shown in Fig. 2c, ~18.1% of the α7-integrin⁺ cells were tdTomato positive after MuVLP_Cre treatment, significantly higher than the 6.3% in the VSVG-VLP_Cre-treated group (Fig. 2c). Importantly, a small number of tdTomato-positive non-myogenic cells were detected in the VSVG-VLP_Cre-treated group, indicating MuVLP_Cre’s specificity for myogenic cells (Fig. 2d).

Given the ability of intramuscularly injected MuVLP_Cre to transduce muscle cells in B6-G/R;B10-Dmd-KO mice, we investigated whether MuVLP_Cre could induce gene recombination in muscle cells through systemic administration. Three-week-old B6-G/R;B10-Dmd-KO mice were intravenously injected with MuVLP_Cre once a week for a total of four injections to observe its transduction effects on both nonmuscle and skeletal muscle tissues (Fig. 2e). TdTomato⁺ myofibers were exclusively detected in skeletal muscles treated with MuVLP_Cre, whereas no transduced myofibers were present in the groups receiving VLPs, MuVLPs, or VLP_Cre. (Fig. 2f, g and Supplementary Fig. 9a). Notably, approximately 40% of myofibers in the diaphragm were transduced, a pivotal observation given that respiratory failure is a primary cause of mortality in numerous myopathies³¹ (Fig. 2f, g). Additionally, tdTomato expression was also detected in the quadriceps (8.0%), tibialis anterior (8.8%), gastrocnemius (10%), and triceps (6.3%). Conversely, no evidence of transduction was observed in nonskeletal muscle tissues, including the heart, liver, spleen, lung, and kidney (Fig. 2h and Supplementary Fig. 9b). In summary, these findings suggest that MuVLP_Cre can transduce myogenic progenitor cells in muscle tissue and selectively target skeletal muscle tissue to induce gene recombination via systemic delivery, laying a solid foundation for precise gene editing in skeletal muscle tissue.

MuVLP_Cas9 enable selective excision of Dmd exon 4 in muscle cells

Frame-disrupting mutations in the Dmd gene, which encodes the protein responsible for muscular dystrophy, impair the integrity of myofibers and drive muscle degeneration^4,32. Removal of one or more exons from the mutated transcript can generate in-frame mRNA and truncated yet functional proteins^8,33,34. To enable MuVLPs to mediate exon deletion and restore dystrophin expression in the B10-Dmd-KO mice via CRISPR/Cas9 delivery, we screened sgRNA pairs capable of effectively excising Dmd exon 4 (Supplementary Fig. 10a). After transfection into C2C12 cells, the optimal sgRNA pairs were determined through genomic PCR and Sanger sequencing (Supplementary Fig. 10a). Concurrently, a plasmid containing the optimal sgRNA pair and Cas9, along with plasmids expressing muscular fusogen and Gag-Pol, was transfected into HEK-293T cells, resulting in the production of MuVLP_Cas9 (Fig. 3a, b and Supplementary Fig. 10b). Subsequently, we quantified Cas9 in VLP_Cas9 and MuVLP_Cas9 using an anti-Cas9 enzyme-linked immunosorbent assay (ELISA). The analysis indicated that both VLP_Cas9 and MuVLP_Cas9 encapsulate approximately 11 Cas9 molecules (Fig. 3c). To assess the specificity of MuVLP_Cas9 in targeting the Dmd gene within muscle cells, we incubated MuVLP_Cas9 with six distinct cell lines and measured Cas9 protein levels and gene editing efficacy (Fig. 3d). Immunofluorescence assays revealed that the Cas9 protein was delivered exclusively into iC2C12 cells and not into other cell lines (Fig. 3e). As shown in Supplementary Fig. 11, the Cas9-positive cell proportion in iC2C12 cells co-incubated with MuVLP_Cas9 reached 69.2%, significantly higher than that in iC2C12 cells co-incubated with VLP_Cas9 (10.7%). Additionally, iC2C12 cells co-incubated with MuVLP_Cas9 exhibited a significantly higher Cas9-positive proportion compared to C2C12 cells (7.5%), NIH/3T3 cells (3.9%), and other cell types (Supplementary Fig. 11). These results demonstrate that MuVLP_Cas9 selectively fuse with muscle cells and specifically deliver Cas9 protein. Furthermore, genomic PCR analysis revealed that MuVLP_Cas9 effectively transduced and edited iC2C12 cells but not C2C12 cells, NIH/3T3 cells, A9 cells, DC2.4 cells, or RAW264.7 cells (Fig. 3f and Supplementary Fig. 10c). Additionally, Sanger sequencing demonstrated that MuVLP_Cas9 successfully removed Dmd exon 4 in iC2C12 cells (Fig. 3g). These results confirm the ability of MuVLP_Cas9 to selectively deliver gene editing tools to cells expressing muscular fusogens, resulting in successful excision of the mutated Dmd gene exon.

**Fig. 3: MuVLP_Cas9 excises *Dmd* exon 4 in vitro.**

In vivo skeletal muscle-specific gene editing with MuVLP_Cas9 treats dystrophy in a mouse model for DMD

Given the robust transduction and gene editing capacity of MuVLP_Cas9 in iC2C12 cells in vitro, we investigated its potential for systemic gene editing in vivo. We administered MuVLP_Cas9 via weekly intravenous injections to 3-week-old B10-Dmd-KO mice over a course of four doses, aiming to evaluate its therapeutic efficacy in the context of DMD (Fig. 4a). No weight loss was observed during the treatment course, indicating that MuVLP_Cas9 demonstrated good tolerability and safety in the mice (Supplementary Fig. 12). One week after the last dose, we quantitatively evaluated the forepaw grip strength of the mice using a standardized grip strength meter. The B10-Dmd-KO mice receiving MuVLP_Cas9 treatment exhibited a significant increase in grip strength, in contrast to the lack of change in grip strength in the control groups treated with VLPs, MuVLPs, or VLP_Cas9 (Fig. 4b). Baseline data from WT mice show a grip strength of 10.4 g/g, while untreated B10-Dmd-KO mice exhibit weakness (6.1 g/g). MuVLP_Cas9-treated mice reached 7.5 g/g (72% recovery to WT levels). To further determine the functional benefits of MuVLP_Cas9 administration, we subjected the mice to a treadmill endurance test. As shown in Fig. 4c and Supplementary Movie 1, MuVLP_Cas9 treatment significantly extended the running time to exhaustion, indicative of increased muscular performance, an effect not observed in the control groups. WT mice ran for 35.5 min, compared to 16.9 min for untreated B10-Dmd-KO mice, and 24.3 min for MuVLP_Cas9-treated mice (68% recovery). Additionally, the serum creatine kinase (CK) level, a diagnostic marker of muscle tissue damage, was notably lower in the MuVLP_Cas9 group than in the PBS, VLPs, MuVLPs, and VLP_Cas9 control groups, implying a reduction in muscle injury (Fig. 4d). Additionally, WT serum CK levels were 31.9 U/L, compared to 1513.2 U/L for untreated B10-Dmd-KO mice; MuVLP_Cas9-treated mice showed a serum CK level of 309.7 U/L, representing a 79% reduction compared to the untreated group.

**Fig. 4: MuVLP_Cas9 specifically deletes *Dmd* exon 4 in skeletal muscle tissue and restores dystrophin protein expression.**

To investigate whether the therapeutic effects of MuVLP_Cas9 were attributed to gene editing in muscle cells, we assessed genomic modifications and protein production in both skeletal muscle and nonmuscle tissues at the end of the treatment. In various skeletal muscles of mice treated with MuVLP_Cas9, a deletion of Dmd exon 4 in genomic DNA was observed, whereas Dmd exon 4 remained unaltered in the mice receiving VLPs, MuVLPs, or VLP_Cas9 (Fig. 4e and Supplementary Fig. 13). Crucially, no gene editing was detected in nonmuscle tissues—including the heart, liver, spleen, lung, and kidney—across all treatment groups (Fig. 4e). These results underscore the skeletal muscle-specific gene editing precision of MuVLP_Cas9. Furthermore, Western blotting analysis (Fig. 4f) and immunofluorescence staining (Fig. 4g and Supplementary Fig. 14) revealed a marked absence of the dystrophin protein in the muscles of the B10-Dmd-KO mice treated with PBS, VLPs, MuVLPs, or VLP_Cas9, while those treated with MuVLP_Cas9 showed notable dystrophin protein restoration. Additionally, histopathological examination and biochemical analysis of liver injury markers in post-treatment mouse samples revealed no significant morphological differences or signs of hepatotoxicity following MuVLP_Cas9 administration (Supplementary Fig. 15, 16). Finally, we measured anti-MuVLP_Cas9 IgG antibody levels in the serum. As shown in Supplementary Fig. 17, no significant differences in anti-MuVLP_Cas9 IgG antibody levels were detected across all groups, suggesting that MuVLPs may elicit a limited humoral immune response under the tested conditions. Nonetheless, further studies are warranted to comprehensively assess the immunogenicity and long-term safety of repeated administrations. These findings highlight the efficacy and skeletal muscle-specific gene editing of MuVLP_Cas9, demonstrating its therapeutic potential without apparent adverse effects.

Finally, to explore the dose-dependence of MuVLP_Cas9 in muscle cells, we administered MuVLP_Cas9 intravenously at doses of 5.5 × 10¹⁰, 1.1 × 10¹¹, and 1.7 × 10¹¹ particles to 3-week-old B10-Dmd-KO mice once a week for 3 weeks, with four total doses, to evaluate how different doses of MuVLP_Cas9 affect editing efficiency (Fig. 4h). One month after the final dose, we assessed grip strength in the mice using a standardized grip meter. Mice treated with the 5.5 × 10¹⁰ dose of MuVLP_Cas9 showed a significant increase in grip strength, which further increased with higher doses (Fig. 4i). To further assess the functional benefits of MuVLP_Cas9 treatment, we placed the mice in a treadmill endurance test. As shown in Fig. 4j, with increasing MuVLP_Cas9 doses, the mice ran longer before exhaustion. Exon 4 deletion of the Dmd gene was observed in the skeletal muscles of MuVLP_Cas9-treated mice, and the number of exon 4-deleted DNA bands increased with higher doses (Supplementary Fig. 18a). Additionally, dystrophin expression in skeletal muscles also increased with higher doses (Supplementary Fig. 18b). To further quantify MuVLP_Cas9-mediated gene editing in muscle tissues, we performed qPCR using Taqman probes for the exon 19–20 junction to quantify total Dmd transcripts, and another probe targeting the exon 3–5 junction to quantify exon 4-deleted transcripts. As shown in Fig. 4k, after injecting 5.5 × 10¹⁰ MuVLP_Cas9, exon 4-deleted transcripts were found in multiple skeletal muscles, with targeting levels ranging from 2 to 8%. In contrast, after injecting 1.7 × 10¹¹ MuVLP_Cas9, exon 4-deleted transcripts were found in multiple skeletal muscles, with targeting levels ranging from 9 to 25%. These results demonstrate a dose-response relationship for MuVLP_Cas9 in muscle gene editing. Additionally, MuVLP_Cas9 can reduce muscle pathology indices in mice in a dose-dependent manner. High-dose MuVLP_Cas9 treatment significantly reduced the proportion of centrally-localized nuclei in diaphragm from 73.71% ± 0.53% (PBS) to 25.10% ± 1.91% (Fig. 4l). Dose-dependent reductions were also observed in quadriceps, tibialis anterior, gastrocnemius, and triceps muscles (Supplementary Fig. 19a). Fibrosis was evaluated through collagen deposition quantification using Masson’s trichrome staining. High-dose MuVLP_Cas9 (1.7 × 10¹¹) markedly decreased fibrotic areas in diaphragm from 20.78% ± 1.90% (PBS) to 3.29% ± 0.83% (Fig. 4m), with analogous dose-dependent reductions in fibrosis across the aforementioned muscle groups (Supplementary Fig. 19b). These data demonstrate that MuVLP_Cas9 can improve muscle pathology by reducing centrally-located nuclei, inhibiting fibrosis, and promoting muscle fiber structural recovery, further validating its therapeutic potential.

Discussion

In the pursuit of precise and efficacious gene editing therapy, we designed a therapeutic approach for skeletal muscle-specific gene editing using mammalian fusogen-directed virus-like particles. By mimicking the natural fusion process of muscle cells^25,26, MuVLPs deliver gene editing tools selectively to skeletal muscle cells, enhancing specificity and efficiency while potentially reducing off-target effects, thereby augmenting safety and presenting a strategy for treating severe muscular genetic disorders, such as DMD.

By harnessing the surface expression of Myomaker and Myomerger on virus-like particles, MuVLPs achieve precise fusion with muscle progenitor cells harboring muscular fusogens, thereby effectively delivering gene editing tools to muscle cells. In vitro experimentation reveals exclusive transduction solely in muscular fusogen-positive cells, underscoring the remarkable specificity of this system. Remarkably, in vivo studies demonstrate that systemically administered MuVLPs selectively transduce skeletal muscles, notably exhibiting potent transduction efficacy in the diaphragm, a muscle essential for respiratory function³¹. Compared to conventional viral vectors like AAVs, which rely on muscle-specific promoters for tissue-restricted editing³⁵, MuVLPs exploit the inherent fusogenic properties of muscle cells, which may help confine gene editing to skeletal muscle tissue; however, a comprehensive assessment of off-target activity will be important in future work. While myotropic AAV capsids achieve high cardiac and skeletal muscle transduction with minimal hepatic uptake³⁶, MuVLPs, as a non-viral platform, circumvent immunogenicity and long-term persistence risks associated with viral vectors. Furthermore, although RNP delivery minimizes off-target editing by transient activity, chemical modifications to enhance RNP stability may inadvertently prolong its activity and associated risks³⁷. In contrast, MuVLPs combine muscle-targeting specificity with the transient nature of RNPs, thereby mitigating risks from prolonged editing. Moreover, the application of MuVLPs in a murine model of DMD not only reinstates dystrophin expression in various muscle tissues but also significantly enhances exercise capacity and endurance, unveiling the therapeutic potential of MuVLPs in muscle disease models.

Although MuVLPs have shown significant targeting specificity and editing efficiency in skeletal muscle, their gene editing capability in cardiac tissue is limited. MuVLPs demonstrate minimal accumulation in heart muscle, which is a significant limitation given that cardiomyopathy is a major contributor to morbidity and mortality in DMD patients. The low sensitivity of cardiomyocytes to MuVLPs may be related to the limited expression of Myomaker/Myomerger proteins in the heart or differences in the cell fusion mechanisms²⁶. Specifically, Myomaker is primarily expressed during the myoblast fusion stage, where it mediates the initial membrane contact and fusion signaling²⁶. Myomerger, on the other hand, plays an auxiliary role in the later stages of fusion to complete membrane fusion²⁵. The expression of both proteins is significantly reduced or even absent in mature myofibers²⁴. This observation suggests that, in its current design, the MuVLP platform may not be sufficient for correcting cardiac defects associated with muscular dystrophy. Addressing this limitation is critical for developing a comprehensive treatment strategy for DMD. Future work should focus on optimizing the MuVLP system—potentially by exploring alternative or additional fusogenic proteins, or by modifying the envelope composition—to enhance its transduction efficiency in cardiomyocytes. In parallel, a comprehensive assessment of long-term safety is essential before advancing MuVLPs to clinical applications. Future studies comparing the genomic integration profiles and long-term immunogenicity of MuVLP against viral vectors will further solidify its therapeutic utility.

In summary, this work harnesses the body’s inherent cell fusion mechanisms to mediate the delivery of gene editing tools, establishing a platform for gene editing therapies targeting muscle diseases and offering an approach for treating other conditions that require highly tissue-specific delivery. Moreover, by further optimizing the design and production of MuVLPs and exploring a wider range of cell membrane fusion proteins could expand the scope of MuVLPs applications to more types of cells and tissues, thereby opening up additional possibilities in the field of gene editing.

Methods

Ethical statement

This study complies with all relevant ethical regulations. All experiments conducted in this study were approved by the Animal Care and Use Committee of South China University of Technology (SCUT) (approval number: 2019012).

Cells and animals

The human cell line HEK-293T (CRL-3216) and the mouse cell lines C2C12 (CRL-1772), NIH/3T3 (CRL-1658), A9 (CRL-1811), and RAW264.7 (TIB-71) were obtained from the American Type Culture Collection (ATCC). The mouse cell line DC2.4 (CL-0545) was purchased from Procell Life Science & Technology Co., Ltd. C2C12 cells were infected with lentiviruses encoding Tet-on, Myomaker, and Myomerger to establish the Tet-on-Myomaker-Myomerger-C2C12 cell line (iC2C12). All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum. The cells were maintained in a humidified environment at 37 °C with 5% CO₂ and regularly tested for mycoplasma contamination.

B10-Dmd-KO mice (3–5 weeks old) and B6-G/R mice (3–5 weeks old) were purchased from GemPharmatech Co., Ltd (Jiangsu, China). Female B10-Dmd-KO mice were bred with male B6-G/R mice to generate male B6-G/R;B10-Dmd-KO mice. All mice were housed in specific pathogen-free (SPF) facilities at South China University of Technology and were cared for in accordance with the Guide for the Care and Use of Laboratory Animals. The mice were kept at a temperature of 20–25 °C, with humidity levels between 30 and 70% and a 12 h light/12 h dark cycle.

Plasmid construction

The Gag-Pol expression plasmid was constructed by amplifying the Gag-Pol sequence from the psPax2 plasmid (Addgene plasmid #12260) and cloning it into the pCMV plasmid using In-Fusion (TaKaRa Bio, Tokyo, Japan) cloning. The muscular fusogen expression plasmid was created by amplifying the Myomaker and Myomerger sequences from the cDNA of B10-Dmd-KO mice and cloning them into the pCMV plasmid using In-Fusion cloning. For lentivirus utilized in constructing iC2C12 cell lines, the Tet-on DNA fragment was synthesized by Sangon (Shanghai, China) and cloned and inserted into the transfer lentivirus plasmid pCDH-EF1 (Addgene plasmid #72266) along with the Myomaker and Myomerger sequences. For the generation of MuVLP_EGFP, Gag-EGFP was constructed by amplifying Gag from psPax2 and EGFP from pX458 (Addgene plasmid #73531). For the generation of MuVLP_Cre, the DNA fragments of the Cre recombinase were synthesized by Sangon. The Cre recombinase fragment was then amplified and cloned and inserted into the pCMV plasmid along with Gag using In-Fusion cloning to construct Gag-Cre. For the generation of MuVLP_Cas9, Gag-Cas9 was constructed by amplifying Gag from the psPax2 plasmid and Cas9 from the pX458 plasmid (Addgene plasmid #73531). The U6 DL-sgRNA and U6 DR-sgRNA sequences were also amplified from the pX458 plasmid. Gag-Cas9, U6 DL-sgRNA, and U6 DR-sgRNA were then cloned and inserted into the pCMV plasmid using In-Fusion cloning. The DNA sequences, primers, and sgRNAs used in this work can be found in Supplementary Tables 1–3.

MuVLPs production and purification

MuVLPs production was achieved through transient transfection of HEK-293T cells. Cells were seeded in 100 mm cell culture dishes (Corning, CA, USA) at a density of 6 × 10⁶ cells per dish. After 24 h, transfection was performed using EZ Cell Transfection Reagent (Life-ilab, Shanghai, China) according to the manufacturer’s protocol. For VLPs production, 5 μg of plasmid expressing Gag-Pol was transfected into each dish of HEK-293T cells. For MuVLPs production, a mixture of 5 μg of plasmid expressing Gag-Pol and 10 μg of plasmid expressing muscular fusogens was transfected into each dish of HEK-293T cells. For VSVG-VLPs production, a mixture of 5 μg of plasmid expressing Gag-Pol and 10 μg of plasmid expressing VSVG (pMD2.G, Addgene plasmid #12259) was transfected into each dish of HEK-293T cells. For production of VLP_EGFP, VLP_Cre, or VLP_Cas9, a mixture of 5 μg of plasmid expressing Gag-Pol and 15 μg of plasmid expressing Gag-EGFP, Gag-Cre, or Gag-Cas9 was transfected into each dish of HEK-293T cells. For VSVG-VLP_EGFP or VSVG-VLP_Cre production, a mixture of 5 μg of plasmid expressing Gag-Pol, 10 μg of plasmid expressing VSVG (pMD2.G), and 15 μg of plasmid expressing Gag-EGFP or Gag-Cre was transfected into each dish of HEK-293T cells. For MuVLP_EGFP, MuVLP_Cre, or MuVLP_Cas9 production, a mixture of 5 μg of plasmid expressing Gag-Pol, 10 μg of plasmid expressing muscular fusogens, and 15 μg of plasmid expressing Gag-EGFP, Gag-Cre, or Gag-Cas9 was transfected into each dish of HEK-293T cells. After 72 h, the cell supernatants were collected and centrifuged at 1000 × g for 10 min to remove cellular debris. The supernatants were then filtered through a 0.45-μm PVDF membrane. The filtered supernatants were concentrated 1000-fold by ultracentrifugation on a 20% (w/v) sucrose cushion in PBS using an Optima XPN ultracentrifuge (Beckman Coulter, CA, USA) with an SW32Ti rotor at 120,000 × g for 2 h at 4 °C. The MuVLPs were resuspended in precooled PBS and centrifuged at 1000 × g for 10 min to remove residual debris. The total particle number was determined using a Flow NanoAnalyzer (NanoFCM Inc., Fujian, China), a device specifically designed to measure nanoparticles. The instrument was calibrated using quality control nanospheres following the manufacturer’s instructions. The concentration of VLPs was quantified by detecting scattering and fluorescence signals of individual particles. A particle size threshold (>50 nm) was applied to exclude cell debris interference, and the instrument was calibrated using standard silica beads of known concentration. EGFP fluorescence was excited using a 488 nm laser, and detected in the 525/40 nm channel. At least 10,000 particle events were collected per sample.

Western blotting analysis of MuVLPs protein content

Virus-like particles were lysed using RIPA lysis buffer (Biosharp, Anhui, China) at 4 °C with vortexing for 30 min. The total protein content of the particles was determined using the BCA Protein Assay Kit, and the samples were diluted with SDS loading buffer (Biosharp). Total protein (20 μg) was heated at 99 °C for 20 min. The proteins were then separated by 10-15% SDS‒polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The following antibodies were used for detection: Myomaker (1:1000, NBP2-94514, Novus Biologicals, CO, USA), Myomerger (1:1000, #AF4580, R&D, MN, USA), HIV-Gag (p55, p24, p17; 1:1000, #ab63917, Abcam, Cam, UK), EGFP (1:1000, MA1-952, Thermo Fisher, MA, USA), Cre recombinase (1:1000, #ab188568, Abcam), and Cas9 (1:2000, #ab189380, Abcam). The secondary antibodies used included anti-rabbit (1:5000, BL003A, Biosharp), anti-sheep (1:5000, #ab140470, Abcam), and anti-mouse (1:5000, BL001A, Biosharp) antibodies. Detection was performed using an enhanced chemiluminescence (ECL) kit (Thermo Scientific) on an ImageQuant LAS 4000 mini chemiluminescence imaging system.

The morphology of MuVLPs

MuVLPs were deposited onto copper grids and negatively stained with 2% phosphotungstic acid for 30 s, followed by drying on filter paper. The morphology of MuVLPs was examined using a field emission transmission electron microscope (Talos F200x, Thermo Fisher) operating at an accelerating voltage of 200 kV.

Fusion of MuVLPs with cells in vitro

VLPs, MuVLPs, and VSVG-VLPs were labeled with DiD dye (Beyotime, Shanghai, China), while iC2C12, C2C12, NIH/3T3, A9, DC2.4, and RAW264.7 cells were labeled with DiI dye (Beyotime, Shanghai, China). DiI-labeled cells were seeded in 24-well plates (5 × 10⁴ cells per well) and incubated with 1.8 × 10¹⁰ DiD-labeled VLPs, DiD-labeled MuVLPs or DiD-labeled VSVG-VLPs for 8 h. Fluorescence images were acquired using an LSM880 confocal microscope (Zeiss, Oberkochen, Germany), and confocal data were acquired with Zeiss ZEN2 (Black Edition) software. Analysis was performed using Zeiss ZEN2 (Blue Edition) software. The percentage of DiI-positive cells was determined using a FACSCelesta system (BD Biosciences, CA, USA). Data acquisition was carried out with BD FACS Diva software v8.0.1.1, and the data were analyzed using FlowJo software v10.0.7 (BD Biosciences).

Biodistribution of MuVLPs in vivo

All mouse experiments complied with all relevant ethical regulations. All the experiments in this research were approved by the Animal Care and Use Committee of South China University of Technology (SCUT) (official approval number: 2019012). A total of 1.1 × 10¹¹ DiD-labeled VLPs, DiD-labeled MuVLPs or DiD-labeled VSVG-VLPs were administered via intravenous injection into 3-week-old male B10-Dmd-KO mice (n = 3 mice per group). Eight hours post-injection, muscle tissues and normal organs were harvested and analyzed using an IVIS Lumina III in vivo imaging system (PerkinElmer, MA, USA). Data acquisition was performed using Living Image software v4.4 (PerkinElmer).

In vitro MuVLP_EGFP delivery experiments

iC2C12, C2C12, NIH/3T3, A9, DC2.4 and RAW264.7 cells were seeded in 24-well plates (5 × 10⁴ cells per well) and incubated with 1.8 × 10¹⁰ VLP_EGFP, MuVLP_EGFP or VSVG-VLP_EGFP for 12 h. For CLSM imaging, cells were fixed with 4% paraformaldehyde and stained for F-actin and nuclei using Alexa Fluor 568 Phalloidin (Invitrogen, MA, USA) and DAPI (Biosharp). Fluorescence images were captured using an LSM880 CLSM (Zeiss). The mean fluorescence intensity of EGFP-positive cells was analyzed using a BD FACSCelesta system.

MuVLPs transduce primary skeletal muscle cells in vitro

The tibialis anterior muscles of B6-G/R;B10-Dmd-KO mice were excised and minced into small pieces with curved scissors. Mononuclear cells were obtained by enzymatic digestion with 0.2% collagenase type II (Sangon) and 0.05% dispase in DMEM at 37 °C for 15 min, followed by an additional 10 min of digestion. The cells were centrifuged, filtered through a 70 µm strainer, and cultured in DMEM containing 20% donor horse serum, 1% penicillin, and 1% glutamine. The medium was refreshed every other day. After 7 days, primary skeletal muscle cells were harvested and counted. The cells were seeded in 24-well plates (1 × 10⁴ cells per well). The next day, the medium was replaced with DMEM containing 2% donor horse serum (Atlanta Biologics) and 1% penicillin‒streptomycin. On the fourth day, 1.8 × 10¹⁰ MuVLP_Cre or controls were added for transduction. After 72 h, the cells were fixed with 4% paraformaldehyde, and the nuclei were stained with DAPI (Biosharp). Fluorescence images were captured using an LSM880 CLSM (Zeiss).

FACS analysis of MuVLP_Cre-transduced muscle progenitor cells in vivo

Three-week-old male B6-G/R;B10-Dmd-KO mice were intramuscularly injected with 1.1 × 10¹¹ MuVLP_Cre (n = 5) or other controls (n = 3) once a week for a total of two injections. One week after the final injection, the tibialis anterior muscles were dissociated into single cells by enzymatic digestion in DMEM containing 0.2% collagenase type II (Sangon), 0.05% dispase (Sangon), 0.01% DNase I (Sangon), and 0.01% hyaluronidase (Sangon) at 37 °C for 30 min. Before labeling the markers on the cell surface, Fc receptors were blocked using anti-mouse CD16/32 antibodies. TdTomato-positive cells in the mouse tibialis anterior muscle were analyzed using APC-conjugated anti-mouse α7-integrin (Thermo Fisher), Alexa Fluor 700-conjugated anti-mouse CD45 (BioLegend, CA, USA), Alexa Fluor 700-conjugated anti-mouse CD31 (BioLegend), Alexa Fluor 700-conjugated anti-mouse Sca1 (BioLegend), and Alexa Fluor 700-conjugated anti-mouse Ter119 (BioLegend) antibodies. All samples were analyzed with a BD FACSCelesta system. The data were collected using BD FACS Diva software v8.0.1.1 and analyzed using FlowJo software v10.0.7.

Histology

Three-week-old male B6-G/R;B10-Dmd-KO mice were intravenously injection with 1.1 × 10¹¹ MuVLP_Cre (n = 5) or other controls (n = 3) once a week for a total of four injections. One week after the final injection, the skeletal muscles and major organs were harvested and fixed overnight in 4% paraformaldehyde, followed by an overnight incubation in 30% sucrose solution. The tissues were then embedded in optimal cutting temperature (OCT) compound and frozen. Sections (10 μm thick) were collected using a cryostat (CM1950, Leica, Wetzlar, Germany). After the sections were washed three times with PBS, they were mounted with antifade mounting medium containing DAPI (Beyotime). Fluorescence images were captured using an LSM880 CLSM (Zeiss).

Quantification of MuVLP_Cas9 protein content

MuVLP_Cas9 were lysed in RIPA lysis buffer to quantify the Cas9 protein. Quantification was performed using the FastScan™ Cas9 (S. pyogenes) ELISA Kit (29666 C, Cell Signaling Technology, MA, USA) according to the manufacturer’s protocol. A standard curve was generated using recombinant Cas9 (S. pyogenes) nuclease protein. The total particle number was determined using a Flow NanoAnalyzer to calculate the number of Cas9 proteins per MuVLP_Cas9.

Immunofluorescence

For coincubation experiments of VLP_Cas9 or MuVLP_Cas9 with cells, iC2C12, C2C12, NIH/3T3, A9, DC2.4, and RAW264.7 cells were seeded in 24-well plates (5 × 10⁴ cells per well) and incubated with 1.8 × 10¹⁰ VLP_Cas9 or MuVLP_Cas9 for 24 h. The cells were fixed with 4% paraformaldehyde for 20 min. After fixation, the cells were washed three times with PBS and permeabilized with 0.1% Triton X-100 for 10 min. The cells were then blocked with 10% goat serum at room temperature for 20 min and incubated with anti-Cas9 (1:100, #ab189380, Abcam) at 4 °C overnight. After being washed three times with PBS, the cells were incubated with a goat anti-rabbit Alexa Fluor 647-conjugated antibody (1:500, A0468, Beyotime) for 1 h, followed by three PBS washes. F-actin was stained with Alexa Fluor 488 Phalloidin (Invitrogen), and the cells were mounted with antifade mounting medium containing DAPI (Beyotime). Fluorescence images were captured using an LSM880 CLSM (Zeiss).

For analysis of dystrophin expression in muscle tissue, 3-week-old male B10-Dmd-KO mice were intravenously injection with 1.1 × 10¹¹ MuVLP_Cas9 (n = 5) or other controls (n = 3) once a week for a total of four injections. One week after the final injection, muscle tissues were collected and fixed overnight with 4% paraformaldehyde. The next day, the fixed muscles were transferred to 1 × PBS containing 10 mM glycine to quench free aldehydes for at least 24 h and then paraffin embedded. With a HistoCore AUTOCUT (RM2255, Leica), the paraffin blocks were cut into 5 μm sections. Deparaffinized sections underwent antigen retrieval via microwave-boiled citrate buffer (10 mM, pH 6.0) using a standardized protocol: slides were heated at 95 °C for 5 min, cooled to room temperature for 5 min, and reheated for an additional 5-min cycle to optimize epitope exposure. Following retrieval, the sections were blocked with 10% goat serum for 20 min and incubated with anti-dystrophin (1:100, D8168, Sigma, CA, USA) at 4 °C overnight. After three PBS washes, the sections were incubated with a goat anti-rabbit Alexa Fluor 647-conjugated antibody (1:500, A0468, Beyotime) for 1 h, washed three times with PBS, and mounted with antifade mounting medium containing DAPI (Beyotime). Fluorescence images were captured using an LSM880 CLSM (Zeiss).

Genomic DNA extraction and genomic PCR

Genomic DNA was extracted from tissues and cultured cells using a Genomic DNA Extraction Kit (D0063, Beyotime) according to the manufacturer’s instructions. The DNA concentration was measured using a NanoDrop system (Thermo Fisher), and 200 ng of DNA was subjected to PCR with PrimeSTAR Max DNA Polymerase (R045A, TaKaRa). The Dmd4 locus was amplified using the primers Dmd4_F and Dmd4_R (Supplementary Table 2). The PCR conditions were 98 °C for 3 min, followed by 35 cycles of 98 °C for 10 s, 58 °C for 15 s, and 72 °C for 10 s. A 2% agarose gel was used to extract the exon-excised bands, which were then cloned and inserted into the pUC19 plasmid and transformed into competent cells. Single colonies were analyzed by Sanger sequencing to confirm the correct deletion of sequences flanking the two guide RNAs.

Grip strength measurements

One week after the final injection in the treatment experiment, the forelimb grip strength of the mice was measured using a grip strength meter (KW-ZL, Calvin Bio, Nanjing, China). Each mouse was measured three times, and the average value was normalized to their body weight.

Treadmill fatigue tests

All mice were acclimated to the mouse treadmill (KW-PT/6, Calvin Bio, Nanjing, China) for 3 days before the start of the study and 1 week prior to sacrifice. On the fourth day, the mice were allowed to run to exhaustion, and the total running time and distance were recorded. Exhaustion or inability to run was defined as the point at which the mouse could no longer continue running after being stimulated with a 0.4 mA current. The treadmill settings were as follows: initial speed of 5 m/min, accelerated by 1 m/min every 5 min until exhaustion and stimulation current of 0.4 mA. The running performance of the mice is available in the Supplementary Movie 1.

Detection of dystrophin by Western blotting analysis

One week after the final dose in the treatment experiment, mouse muscle tissue was collected and placed in liquid nitrogen before being ground to a powder. After addition of RIPA lysis buffer, the mixture was vortexed at 4 °C for 30 min. The total protein content of the particles was determined using the BCA Protein Assay Kit, and the samples were diluted with SDS loading buffer (Biosharp). Total protein (30 μg) was heated at 99 °C for 20 min. The proteins were then separated by 4-10% SDS‒polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The following antibodies were used for detection: anti-dystrophin (1:1000, D8168, Sigma) and anti-GAPDH (1:2000, #ab8245, Abcam). The secondary antibodies used included anti-rabbit (1:5000, BL003A, Biosharp) and anti-mouse (1:5000, BL001A, Biosharp) antibodies. An ECL kit (Thermo Scientific) was used for detection with an ImageQuant LAS 4000 mini chemiluminescence imaging system.

Potential hepatotoxicity of MuVLP_Cas9 and muscle damage indicators

One week following the final dose in the treatment experiment, serum samples were collected. The serum levels of CK, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were measured using ELISA kits (Rito, Guangdong, China).

Histopathological examination

One week or 1 month after the final dose in the treatment experiment, the mice were euthanized, and their major organs and muscle tissues were collected and fixed overnight in 4% paraformaldehyde at 4 °C. The following day, the fixed organs and muscle tissues were transferred to 1× PBS containing 10 mM glycine to quench free aldehydes for at least 24 h, followed by paraffin embedding. With HistoCore AUTOCUT (RM2255, Leica), the paraffin blocks were sectioned into 5 μm slices and subjected to histopathological examination via hematoxylin and eosin staining.

For Masson’s trichrome staining, mouse muscle tissues were collected 1 month after the final treatment dose, following euthanasia. Tissue sections were prepared at a thickness of 10 μm. Fibrosis in the muscle tissues was assessed using a commercially available Masson’s trichrome staining kit, following the manufacturer’s instructions.

Detection of anti-MuVLP_Cas9 antibodies in plasma by ELISA

To evaluate the immune response against MuVLP_Cas9, plasma levels of anti-MuVLP_Cas9 antibodies were quantified using a sandwich ELISA. Briefly, blood samples were collected via retro-orbital puncture from DMD model mice one week after the final administration of MuVLP_Cas9. Whole blood was transferred into 1.5 mL microcentrifuge tubes and centrifuged at 2000 × g for 20 min at 4 °C to isolate plasma, which was stored at −80 °C until analysis. MuVLP_Cas9 (100 μL) were lysed in an equal volume of RIPA buffer (Biosharp, China) supplemented with 1 mM PMSF. The lysate was vortexed for 15 s, incubated on ice for 10 min, and centrifuged at 12,000 × g for 20 min at 4 °C. The supernatant was collected, and protein concentration was determined using a BCA Protein Assay Kit (Biosharp, China). For ELISA plate coating, 10 μg of MuVLP_Cas9 lysate per well was incubated in a 96-well plate at room temperature for 4 h. After washing three times with PBS, nonspecific binding sites were blocked with 5% BSA in PBS for 4 h at room temperature. Plasma samples were diluted 1:10 in PBS, and 100 μL of diluted plasma was added to each well and incubated at 37 °C for 30 min. Following four washes with PBS containing 0.05% Tween-20 (PBST), horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG secondary antibody (1:10,000, BL001A, Biosharp) was added and incubated at 37 °C for 30 min. After additional PBST washes, 100 μL of TMB substrate was added, and the reaction was terminated with 100 μL of stop solution after 10 min of incubation in the dark. Absorbance was measured at 450 nm using a microplate reader.

MuVLP_Cas9 dose-response study in B10-Dmd-KO mice

Three-week-old healthy male B10-Dmd-KO mice were randomly allocated into groups (n = 3 per group) and received weekly intravascular injections of MuVLP_Cas9 at doses of 5.5 × 10¹⁰, 1.1 × 10¹¹, or 1.7 × 10¹¹ particles per administration for four consecutive weeks. Assessments including forelimb grip strength measurement, treadmill exhaustion testing, Dmd exon4 deletion analysis, and dystrophin protein quantification in skeletal muscle were performed 4 weeks following the final injection (11 weeks of age).

Taqman-based determination of Dmd exon4 deletion by quantitative RT-PCR

Total RNA was extracted from tissues using RNAiso Plus reagent (TaKaRa) according to the manufacturer’s instructions. For tissues harvested from animals, 1 μg of RNA was used for cDNA synthesis in a 20 μL reaction using PrimeScript RT Master Mix (TaKaRa). Total Dmd transcripts were quantified using a Taqman probe targeting the exon 19–20 junction, and exon 4-deleted transcripts were quantified using a separate probe targeting the exon 3–5 junction (Supplementary Table 2). Taqman probes for 18S ribosomal RNA (Thermo Fisher) were used as an internal control. Each probe was run in triplicate with a 10 μL reaction volume and 20 ng of cDNA input. Taqman Fast Advanced Master Mix (LifeTech) was used, and the reactions were carried out according to the fast cycling conditions recommended in the manual. The percentage of exon 4-deleted transcripts relative to total Dmd transcripts was determined by calculating ΔCt values between the exon 3–5 and exon 19–20 junctions.

Statistical analysis

All values are expressed as the mean ± s.d. Comparisons between two groups were performed using a two-tailed t test, while multiple group comparisons were conducted using one-way ANOVA followed by Tukey’s multiple comparison test. Statistical significance was set at P < 0.05. Statistical analyses were carried out using GraphPad Prism v8.0 (GraphPad Software, Inc.).

Statistics and reproducibility

No statistical method was used to predetermine sample size. Sample sizes (n = 3–5) were selected based on common practice in the field and literature precedent. No data were excluded from the analyses. All replicates represent biological replicates (independent treatments in separate wells or animals). All samples, cells, and animals were allocated randomly into experimental groups. The investigators were not blinded to allocation during experiments and outcome assessment.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The data that support this study are available within the Article, Supplementary Information or Source data files. Source data are provided with this paper.

References

Mukund, K. & Subramaniam, S. Skeletal muscle: a review of molecular structure and function, in health and disease. Wiley Interdiscip. Rev. Syst. Biol. Med. 12, e1462 (2020).
Article PubMed Google Scholar
Bruusgaard, J. C., Liestøl, K., Ekmark, M., Kollstad, K. & Gundersen, K. Number and spatial distribution of nuclei in the muscle fibres of normal mice studied in vivo. J. Physiol. 551, 467–478 (2003).
Article CAS PubMed PubMed Central Google Scholar
Mercuri, E. & Muntoni, F. Muscular dystrophies. Lancet 381, 845–860 (2013).
Article CAS PubMed Google Scholar
Duan, D., Goemans, N., Takeda, S., Mercuri, E. & Aartsma-Rus, A. Duchenne muscular dystrophy. Nat. Rev. Dis. Prim. 7, 13 (2021).
Article PubMed Google Scholar
Mercuri, E., Bönnemann, C. G. & Muntoni, F. Muscular dystrophies. Lancet 394, 2025–2038 (2019).
Article PubMed Google Scholar
Wasala, N. B., Hakim, C. H., Chen, S. J., Yang, N. N. & Duan, D. Questions answered and unanswered by the first CRISPR editing study in a canine model of Duchenne muscular dystrophy. Hum. Gene Ther. 30, 535–543 (2019).
Article CAS PubMed PubMed Central Google Scholar
Levy, J. M. et al. Cytosine and adenine base editing of the brain, liver, retina, heart and skeletal muscle of mice via adeno-associated viruses. Nat. Biomed. Eng. 4, 97–110 (2020).
Article CAS PubMed PubMed Central Google Scholar
Tabebordbar, M. et al. In vivo gene editing in dystrophic mouse muscle and muscle stem cells. Science 351, 407–411 (2016).
Article ADS CAS PubMed Google Scholar
Strebinger, D. et al. Cell type-specific delivery by modular envelope design. Nat. Commun. 14, 5141 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Mazarakis, N. D. et al. Rabies virus glycoprotein pseudotyping of lentiviral vectors enables retrograde axonal transport and access to the nervous system after peripheral delivery. Hum. Mol. Genet. 10, 2109–2121 (2001).
Article CAS PubMed Google Scholar
Zhou, Q. et al. Exclusive transduction of human CD4⁺ T cells upon systemic delivery of CD4-targeted lentiviral vectors. J. Immunol. 195, 2493–2501 (2015).
Article CAS PubMed Google Scholar
Fu, Y. et al. High-frequency off-target mutagenesis induced by CRISPR-Cas nucleases in human cells. Nat. Biotechnol. 31, 822–826 (2013).
Article CAS PubMed PubMed Central Google Scholar
Nelson, C. E. et al. Long-term evaluation of AAV-CRISPR genome editing for Duchenne muscular dystrophy. Nat. Med. 25, 427–432 (2019).
Article CAS PubMed PubMed Central Google Scholar
Banskota, S. et al. Engineered virus-like particles for efficient in vivo delivery of therapeutic proteins. Cell 185, 250–265.e216 (2022).
Article CAS PubMed PubMed Central Google Scholar
Hamilton, J. R. et al. Targeted delivery of CRISPR-Cas9 and transgenes enables complex immune cell engineering. Cell Rep. 35, 109207 (2021).
Article CAS PubMed PubMed Central Google Scholar
Campbell, L. A. et al. Gesicle-mediated delivery of CRISPR/Cas9 ribonucleoprotein complex for inactivating the HIV provirus. Cell Rep. 27, 151–163 (2019).
CAS Google Scholar
Yao, X. et al. Engineered extracellular vesicles as versatile ribonucleoprotein delivery vehicles for efficient and safe CRISPR genome editing. J. Extracell. Vesicles 10, e12076 (2021).
Article CAS PubMed PubMed Central Google Scholar
Gee, P. et al. Extracellular nanovesicles for packaging of CRISPR-Cas9 protein and sgRNA to induce therapeutic exon skipping. Nat. Commun. 11, 1334 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Editas Medicine. Editas Medicine announces positive initial clinical data from ongoing phase 1/2 BRILLIANCE clinical trial of EDIT-101 for LCA10. Press release Available at: https://ir.editasmedicine.com/news-releases/news-release-details/editas-medicine-announces-positive-initial-clinical-data-ongoing (2021).
Gillmore, J. D., Maitland, M. L. & Lebwohl, D. CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis. N. Engl. J. Med. 385, 1722–1723 (2021).
Article PubMed Google Scholar
Hamilton, J. R. et al. In vivo human T cell engineering with enveloped delivery vehicles. Nat. Biotechnol. 42, 1684–1692 (2024).
Kaczmarczyk, S. J., Sitaraman, K., Young, H. A., Hughes, S. H. & Chatterjee, D. K. Protein delivery using engineered virus-like particles. Proc. Natl. Acad. Sci. USA 108, 16998–17003 (2011).
Article ADS CAS PubMed PubMed Central Google Scholar
Hindi, S. M. et al. Enveloped viruses pseudotyped with mammalian myogenic cell fusogens target skeletal muscle for gene delivery. Cell 186, 2062–2077.e2017 (2023).
Article CAS PubMed PubMed Central Google Scholar
Petrany, M. J. & Millay, D. P. Cell fusion: merging membranes and making muscle. Trends Cell Biol. 29, 964–973 (2019).
Article CAS PubMed PubMed Central Google Scholar
Bi, P. et al. Control of muscle formation by the fusogenic micropeptide myomixer. Science 356, 323–327 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Millay, D. P. et al. Myomaker is a membrane activator of myoblast fusion and muscle formation. Nature 499, 301–305 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Quinn, M. E. et al. Myomerger induces fusion of non-fusogenic cells and is required for skeletal muscle development. Nat. Commun. 8, 15665 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Zhang, Q. et al. The microprotein minion controls cell fusion and muscle formation. Nat. Commun. 8, 15664 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Chan, S. S. et al. Skeletal muscle stem cells from PSC-derived teratomas have functional regenerative capacity. Cell Stem Cell 23, 74–85.e76 (2018).
Article CAS PubMed PubMed Central Google Scholar
Maesner, C. C., Almada, A. E. & Wagers, A. J. Established cell surface markers efficiently isolate highly overlapping populations of skeletal muscle satellite cells by fluorescence-activated cell sorting. Skelet. Muscle 6, 35 (2016).
Article PubMed PubMed Central Google Scholar
Wahlgren, L., Kroksmark, A. K., Tulinius, M. & Sofou, K. One in five patients with Duchenne muscular dystrophy dies from other causes than cardiac or respiratory failure. Eur. J. Epidemiol. 37, 147–156 (2022).
Article PubMed Google Scholar
Koenig, M. et al. Complete cloning of the Duchenne muscular dystrophy (DMD) cDNA and preliminary genomic organization of the DMD gene in normal and affected individuals. Cell 50, 509–517 (1987).
Article CAS PubMed Google Scholar
Taglia, A. et al. Clinical features of patients with dystrophinopathy sharing the 45-55 exon deletion of DMD gene. Acta Myol. 34, 9–13 (2015).
PubMed PubMed Central Google Scholar
Echigoya, Y. et al. Long-term efficacy of systemic multiexon skipping targeting dystrophin exons 45-55 with a cocktail of vivo-morpholinos in mdx52 mice. Mol. Ther. Nucleic Acids 4, e225 (2015).
Article CAS PubMed PubMed Central Google Scholar
Bengtsson, N. E. et al. Muscle-specific CRISPR/Cas9 dystrophin gene editing ameliorates pathophysiology in a mouse model for Duchenne muscular dystrophy. Nat. Commun. 8, 14454 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Weinmann, J. et al. Identification of a myotropic AAV by massively parallel in vivo evaluation of barcoded capsid variants. Nat. Commun. 11, 5432 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Zhang, S., Shen, J., Li, D. & Cheng, Y. Strategies in the delivery of Cas9 ribonucleoprotein for CRISPR/Cas9 genome editing. Theranostics 11, 614–648 (2021).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

This work was supported by the National Key R&D Program of China (2022YFB3808100 (C.-F.X.)); the National Natural Science Foundation of China (52422305 (C.-F.X.), 32271442 (C.-F.X.), 32471434 (C.-F.X.), and 32430059 (J.W.)); Guangdong Basic and Applied Basic Research Foundation (2022B1515020025 (C.-F.X.)); the TCL Young Scholar (2024-2027 (C.-F.X.)).

Author information

These authors contributed equally: Shi-Kun Zhou, Jing-Tong Luo.

Authors and Affiliations

School of Biomedical Sciences and Engineering, South China University of Technology, Guangzhou International Campus, Guangzhou, P.R. China
Shi-Kun Zhou, Jing-Tong Luo, Yi-Fang Chen, Qiu-Hong Jian, Shui-Qing Jiang, Jing Li, Xue-Qin Zhang, Xin-Yu Tan, Xian-Zhu Yang, Cong-Fei Xu & Jun Wang
School of Medicine, South China University of Technology, Guangzhou, P.R. China
Zi-Dong Lu
National Engineering Research Center for Tissue Restoration and Reconstruction, Guangdong Province Key Laboratory of Biomedical Engineering, South China University of Technology, Guangzhou, Guangdong, P.R. China
Cong-Fei Xu & Jun Wang

Authors

Shi-Kun Zhou
View author publications
Search author on:PubMed Google Scholar
Jing-Tong Luo
View author publications
Search author on:PubMed Google Scholar
Yi-Fang Chen
View author publications
Search author on:PubMed Google Scholar
Zi-Dong Lu
View author publications
Search author on:PubMed Google Scholar
Qiu-Hong Jian
View author publications
Search author on:PubMed Google Scholar
Shui-Qing Jiang
View author publications
Search author on:PubMed Google Scholar
Jing Li
View author publications
Search author on:PubMed Google Scholar
Xue-Qin Zhang
View author publications
Search author on:PubMed Google Scholar
Xin-Yu Tan
View author publications
Search author on:PubMed Google Scholar
Xian-Zhu Yang
View author publications
Search author on:PubMed Google Scholar
Cong-Fei Xu
View author publications
Search author on:PubMed Google Scholar
Jun Wang
View author publications
Search author on:PubMed Google Scholar

Contributions

C.-F.X. and J.W. conceived the project, designed the experiments, and wrote and revised the manuscript. S.-K.Z. and J.-T.L. performed the experiments, analyzed the data and wrote the original draft. Y.-F.C., Z.-D.L., Q.-H.J., S.-Q.J., J.L., X.-Q.Z., X.-Y.T. and X.-Z.Y. assisted in the experimental design and data analysis. All the authors discussed the results and assisted in the preparation of the manuscript.

Corresponding authors

Correspondence to Cong-Fei Xu or Jun Wang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Baisong Lu and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Movie 1

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Zhou, SK., Luo, JT., Chen, YF. et al. Muscle-specific gene editing therapy via mammalian fusogen-directed virus-like particles. Nat Commun 16, 9145 (2025). https://doi.org/10.1038/s41467-025-64200-9

Download citation

Received: 06 August 2024
Accepted: 08 September 2025
Published: 15 October 2025
DOI: https://doi.org/10.1038/s41467-025-64200-9