Delivery of Gene Editing Components to Skeletal Muscle Using Lipid Nanoparticles: A Game Changer for Muscle Diseases

This is an original article, first published in CRISPR Medicine News.

Gene editing in its different forms holds unprecedented promises for the precise modulation or correction of the genetic variants that cause muscular dystrophies and myopathies. However, the simultaneous delivery of the gene editing components to various skeletal muscle groups remains a challenge for the application of gene editing to treat individuals affected with a muscle disease. In the neuromuscular field, there is already significant experience using viral vectors (particularly AAVs) to deliver therapeutic genes and nucleic acids in humans from clinical trials and increasingly real-world evidence from approved gene therapies for Spinal Muscular Atrophy (SMA) and Duchenne Muscular Dystrophy (DMD). Furthermore, evolved versions of AAV capsids with higher tropism for skeletal muscle and lower affinity for liver have been developed, as well as strategies to reduce the toxicity and immunogenicity of viral systems (Saha et al., 2025).

In parallel, there is a growing interest in the clinical application of Lipid-Nanoparticles (LNPs) to target different tissues. This interest has been fueled by the approval of Onpattro® (patisiran), a liver-directed siRNA-LNPs conjugate to treat hereditary transthyretin-mediated amyloidosis (hATTR‑PN), the mRNA vaccines, Comirnaty® (BNT162b2) and Spikevax® (mRNA‑1273), as well as of mRESVIA™ (mRNA-1345), and the mRNA vaccine against Respiratory Syncytial Virus (RSV) that was approved by both the FDA and EMA in 2024 for the treatment of lower respiratory tract disease (LRTD).

Delivery remains the key bottleneck

At present, there is no standard LNP formulation to deliver nucleic acids into the circulation to simultaneously reach various muscle groups, and studies describing the use of LNPs in vivo are still sparse. In fact, when LNPs have been used to deliver gene editing components to skeletal muscle for the purpose of treating a neuromuscular condition, the route of administration was almost exclusively local (intramuscular or via limb perfusion) rather than systemic.

To make things more difficult, and depending on the disease mechanism, multiple cell types would need to be genetically corrected for the therapies to work: the muscle fibres themselves, but also muscle stem cells (often termed satellite cells) and even interstitial cells such as fibroadipogenic precursors (FAPs).

However, AAV vectors used in gene therapy for muscle diseases (including Elevidys® approved for DMD) can transduce mature muscle fibers but they miss the muscle stem cell compartment. This is a significant bottleneck because in muscular dystrophies, there is continuous degeneration and regeneration, and when a genome-edited muscle fibre is damaged, it is replaced by a new fibre derived from a non-edited satellite cell, effectively "diluting" the effect of the editing over time (Arnett et al., 2014).

We will use the term Lipid Nanoparticle (LNP) to refer to the classic four-component formulation (ionizable cationic lipid ~50 mol %, cholesterol ~40 mol %, phospholipid or helper lipid ~10 mol %) and PEGylated lipid ~1.5 mol % (Vasileva et al., 2024).

LNPs serve to shield therapeutic RNAs and proteins from degradation, enhance their stability and circulation half-life, and minimise immunotoxic effects. Common administration routes for LNP-based therapeutics include subcutaneous (SC), intramuscular (IM), intravenous (IV), intradermal (ID), intraperitoneal (IP) and oral delivery. IM injection is widely employed for RNA vaccines, whereas IV administration is preferred for liver-targeted RNA therapies and for treating genetic disorders beyond the liver.

The route of administration and dose strongly influences which organs are exposed. High systemic doses can saturate pathways and change distribution. Pardi et al. (Pardi et al., 2015, J of control release) compared the expression level of luciferase after administering LNPs carrying luciferase mRNA in BALB/c mice via different routes: IV, IP, IM, SC, ID and intratracheal. The highest level of protein expression in the liver was achieved when the IV route was used, and the longest duration was with the IM and ID routes (up to 10 days).

LNP design governs where therapies end up

Starting from the formulation, structural changes in ionizable-lipid chemistry (head/linker/tail) influence organ uptake and endosomal escape, whilst changing the LNPs composition molar ratios can alter surface properties that control protein adsorption and circulation (Han et al., 2021). Size and surface charge alter clearance by the reticulocyte endothelial system (RES), mainly in the liver but also in spleen and lymph nodes, as well as tissue penetration. Particles larger than 100nm will be easily cleared by macrophages, and particles smaller than 10nm may be eliminated by the kidneys (Vasileva et al., 2024). Fine-tuned PEGylation can reduce liver uptake, prolong circulation and affect immunogenicity. Other properties, such as rigidity and shape, also affect circulation and cellular interactions.

The protein corona refers to the absorbed host proteins on the surface of the LNPs that direct them to specific cell receptors. It is estimated that roughly 80–90% of LNPs localise to the liver following iv administration. The reason for this is that ApoE and lipoproteins in serum cover LNPs and mediate their binding to low-density-lipoprotein receptor (LDLR) on hepatocytes (Sebastiani et al., 2021). Modulating this protein corona, for example, by blocking ApoE may redirect LNPs towards other cell types (Liu et al., 2023).

Adding targeting units such as peptides, antibodies, or glycans can increase uptake in intended tissues (but often at the cost of altered protein corona or clearance). Finally, an individual's physiological or disease state (inflammation, tumour vasculature, plasma protein composition) changes protein corona and vascular permeability (Liu et al., 2023).

Based on all of these factors, successful skeletal muscle targeting will require combined strategies including selecting ionizable lipids whose pK_a (the apparent pKa of a lipid nanoparticle is the pH value at which the ionizable lipid components within the LNP are 50% ionized or protonated, and 50% deionized or neutral) favour extra-hepatic uptake, modifying helper lipid charge as in the Selective Organ Targeting or SORT system (Cheng et al., 2020), controlling particle size (< 100nm may facilitate extra-hepatic distribution), accounting for the protein corona, using targeting units to engage specific cell types within skeletal muscle. Importantly, biodistribution should also be tested in relevant animal models to account for altered vasculature or tissue structure (e.g. muscle fibrosis, degeneration and regeneration, etc.) as well as considering proteins in plasma which can affect LNPs' tropism.

A search in the literature (PubMed) using a combination of MESH and free-text terms for nanoparticles, lipid nanoparticles, gene or genome editing, CRISPR/Cas systems and skeletal muscle (to filter for in vivo rather than studies in cell models solely) retrieved in December 2025 just over 10 studies. From those, only a handful describe the use of classical four-component LNPs where editing for a muscle disease is the primary focus.

Early studies show proof of concept in muscle

In a pivotal study, Kenjo and colleagues (Kenjo et al., 2023) tested different formulations of LNPs. They encapsulated luciferase (Luc) mRNA, and based on the transfection efficiency after IM injection, they chose the formulation which contained the ionizable lipid, TCL053. Then, they compared head to head the kinetics of the bioluminescence after IM injection of either the above LNP-Luc mRNA conjugates or of an AAV2 vector containing the luciferase sequence and they found that the signal in LNPs injected muscles was detected much earlier than the AAV injected muscles (4 hours versus 7 days) but lasted much less (undetectable by 2 days with LNPs versus 100 days for AAV transduced muscles). When they replaced Luc mRNA with Cas9 mRNA and a guide RNA targeting the Rosa26 locus and encapsulated them separately within the LNPs, they showed that editing had taken place in a dose-dependent manner despite transient expression of Cas9.

The scientist’s aim was to correct a deletion in the DMD gene that causes muscular dystrophy. After optimising a pair of guide RNAs flanking DMD exon 45 in patient-derived myoblasts, they analysed the efficiency of DMD exon 45 skipping in a humanised mouse model (hEx45KI-mdx44 mice). They injected the LNPs carrying Cas9 mRNA and one of the two gRNAs (packed separately in an LNP each), achieving around 10% of exon skipping when using the two gRNAs. Histological analysis of muscles injected with either LNP-CRISPR or PBS showed mild inflammatory infiltration near the site of injection in both cases. Also, liver damage markers and several cytokines were transiently elevated in both treatment groups but returned to basal levels after 7 days, indicating that LNPs IM injection was well tolerated.

Click here to download the Table.

To wrap up, work by Mochida and colleagues has unequivocally demonstrated that LNPs can reach skeletal muscle after IV injection. Furthermore, the therapeutic effect is sustained by the pool of edited cells that are the source of new myofibers after muscle injury, providing long-lasting genome editing. However, the authors acknowledge that efficient LNP delivery to skeletal muscle remains highly challenging and insufficient. They emphasise the need for further research to identify targeting moieties—such as peptides—that could enhance tropism toward specific muscle cell types. In the meantime, they propose leveraging the complementary strengths of LNP-CRISPR via IM administration and AAV-CRISPR via systemic administration, as these approaches are not mutually exclusive. A limitation of this study is the data on LNP-CRISPR distribution and editing in the liver and other organs after IV injection. Future studies should explore this in more detail.

Given their key role in LNP performance, the design of novel ionisable lipids is being pursued intensely by academics and industry. In a study published in 2023, researchers at the University of Toronto identified iso-A11B5C1 using combinatorial chemistry (Chen et al., 2023). This is a biodegradable ionizable lipid that showed high transfection efficiency in skeletal muscle following IM injection, whilst almost abolishing accumulation in liver and spleen.

To test the potential of this formulation for gene editing, the authors compared LNPs containing iso-A11B5C1 with the commercially available lipid SM-102 for delivering Cre mRNA to skeletal muscle in mTmG reporter mice whose cells express tdTomato (a red fluorescent protein). Successful delivery and functional expression of Cre mRNA results in genetic recombination and the replacement of tdTomato expression with membrane-bound EGFP (Enhanced Green Fluorescent Protein). Using this reporter system, researchers found comparable results with the two ionizable lipids after injection into the quadriceps; however, iso-A11B5C1 nanocarriers did not transfect the liver or spleen.

Although this is one of the few studies to report results from IV administration of LNPs into muscle, careful reading suggests that iso-A11B5C1 may not be the best candidate for systemic delivery of mRNA to skeletal muscle. When referring to IV administration of iso-A11B5C1 containing LNPs-Cre mRNA, the authors report little transfection to the liver and spleen, but the transfection efficiencies in skeletal muscles are not shown or mentioned specifically in the article. In fact, they argue that the higher pKa of the iso-A11B5C1 (pKa ≈ 7.2) may constrain it from travelling beyond the site of injection after IM delivery, which is consistent with previous findings that lipids with higher pKa showed low to no transfection after IV injection. Thus, it appears that iso-A11B5C1 is a good candidate for selective local targeting of skeletal muscle with negligible liver and spleen off-targeting, ideal for mRNA vaccines, but probably not for selective and broad targeting of skeletal muscle.

On the other hand and in addition to the low targeting for liver, iso-A11B5C1 was deemed less toxic than other ionizable lipids for two other reasons: i) LNPs containing biodegradable ionizable lipids are safer, enabling repeated dosing, as nonbiodegradable lipids can cause liver damage, ii) iso-A11B5C1 immunogenicity was low in cell studies (assessing the NF-kB pathway) and the animal studies (assessing the release of cytokines). Thus, iso-A11B5C1 containing LNPs could be a safe alternative for extra-hepatic delivery by IM injection of mRNAs, but their ability to reach various muscles after systemic administration is still unclear.

New strategies expand delivery beyond current limits

Many of the genetically determined muscle diseases are congenital, manifesting at birth, and progressive, albeit with variable rates of progression, so early intervention is key, implying that fetal medicine would be transformative for these diseases. In this scenario, LNPs may have a higher chance of success.

The fetal environment offers important advantages for LNP-mediated delivery. First, the high permeability of the vasculature during fetal life facilitates LNPs diffusion and entry into a broad range of tissues including the brain, secondly, the immune system at this stage is more tolerable in general and the RES system and clearance by macrophages is attenuated which means that LNPs spend more time circulating in the blood, thirdly, highly proliferative cells in embryonic tissues are more susceptible to transfection compared to slow-dividing cells and finally, the dose required to treat an embryo which weighs roughly 1% of a 1-year old child is much smaller which translates in larger safety margins and lower costs. Furthermore, several medical procedures are conducted nowadays in fetuses safely (Sagar and David 2024).

As an example of research into this field, Gao et al.in 2023 employed intrahepatic injection in utero to deliver Cre-mRNA/LNPs (containing the ionizable lipid D-Lin-MC3-DMA) complexes to Ai9 mice at 15.5. days of gestation. When analysing transfection efficiency in different tissues, observed as expression of tdTomato, they found high numbers of positive cells in the diaphragm and muscles of the back 48 hours after injection. Interestingly, some of the tdTomato cells also expressed markers of stem or progenitor muscle cells (Pax7 and myogenin).

Furthermore, they did not find any sign of toxicity in the organs of treated embryos or mothers, and the survival rates of both were almost 100% with no significant increase in the release of most cytokines tested. How permanent was the edition? Four weeks after birth, they could detect large numbers of tdTomato-positive differentiated myofibers expressing desmin and laminin in the diaphragm (almost 51% of laminin-expressing cells) and back muscles (around 24% of laminin-positive cells), which must have developed from the transfected muscle stem cells observed two days after treatment.

However, when instead of Cre mRNA they injected Cas9 mRNA and gRNA loaded together in the LNPs, they found that the rates of tdTtomato expression were much lower. For example, in the kidney, the total number of transfected cells fell from over 5% (percentage of tdTomato in total kidney cells) to less than 0.5%. Disappointingly, they did not report data in diaphragm and skeletal muscle when using Cas9 mRNA. The authors argued that even such low rates of editing may be sufficient to restore function in the case of, for example, metabolic diseases and that edited cells may expand due to their proliferative advantage over non-edited cells.

A comparable low level of editing was recently reported by scientists in Pennsylvania (Carpenter et al., 2025) when treating a severe mouse model for Spinal Muscular Atrophy (SMA) with LNPs carrying the adenine base editor ABE8.8 mRNA and a single guide RNA (sgRNA) to induce full-length SMN expression at the fetal (E14) and neonatal stages (P8) in the nervous system. A maximum of 5% and of 3% of edition were achieved in the cortex of fetal and neonatal treated mice, respectively, whereas editing in the spinal cord was negligible. Nonetheless, it was sufficient to increase the survival of mice treated at the fetal stage, leaving the door open for further studies.

Aside from LNPs, other non-viral delivery systems can be adapted to target skeletal muscle to treat muscle genetic diseases. A novel gene editing platform was recently described to overcome the difficulties of delivery to skeletal muscle. Viral-like particles were decorated with muscle-membrane fusion proteins (Myomaker and Myomaker) to promote fusion with muscle cells and editing in a DMD mouse model. By packaging Cas9 ribonucleoproteins into these particles, the researchers created MuVLPCas9, which was systemically administered (4 injections separated by 7 days and then the analysis at 7 days after the last injection) to DMD mouse models to excise the mutated exon in the Dmd gene with high specificity.

The therapy successfully achieved restoration of functional dystrophin protein across various skeletal muscles, including the diaphragm, and yielded a significant improvement in the mice's exercise capacity and endurance. Importantly, the MuVLP system demonstrated precision, exhibiting minimal off-target editing in non-muscle tissues like the heart, liver, and lung and allowed for repeated administration. This work establishes a promising, non-viral strategy for targeted gene editing in skeletal muscle diseases.

There are several other published studies using various types of synthetic nanoparticles (Zhu et al., 2023; Gee et al., 2020; Chen et al., 2019). In all these cases, the nanoparticles contained ribonucleoprotein (RNP) complexes of Cas9 protein and gRNAs. The advantage of delivering ribonucleoprotein complexes directly, where the Cas protein is bound to the guide RNAs, is that it circumvents translation of the mRNA into protein and reduces the off-target effects since editing is limited to the presence and activity of the nuclease.

In all the above studies, LNPs- RNPs conjugates were investigated exclusively in the context of IM injection, achieving a variable level of editing rates and recovery of the desired molecular or functional outcomes.

References

Arnett AL, Konieczny P, Ramos JN, Hall J, Odom G, Yablonka-Reuveni Z, et al. Adeno-associated viral vectors do not efficiently target muscle satellite cells. Mol Ther Methods Clin Dev. 2014;1:14038. doi:10.1038/mtm.2014.38.

Carpenter MD, Kus N, Choi RB, Said H, Dumitru AM, Bauer JH, et al. Systemic mRNA-LNP administration in fetuses improves survival in a mouse model of spinal muscular atrophy. Mol Ther Methods Clin Dev. 2025;33(4):101599. doi:10.1016/j.omtm.2025.101599.

Chen J, Xu Y, Zhou M, Xu S, Varley AJ, Golubovic A, et al. Combinatorial design of ionizable lipid nanoparticles for muscle-selective mRNA delivery with minimized off-target effects. Proc Natl Acad Sci U S A. 2023;120(50):e2309472120. doi:10.1073/pnas.2309472120.

Cheng Q, Wei T, Farbiak L, Johnson LT, Dilliard SA, Siegwart DJ. Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR-Cas gene editing. Nat Nanotechnol. 2020;15(4):313-320. doi:10.1038/s41565-020-0669-6.

Gao K, Li J, Song H, Han H, Wang Y, Yin B, et al. In utero delivery of mRNA to the heart, diaphragm and muscle with lipid nanoparticles. Bioact Mater. 2023;25:387-398. doi:10.1016/j.bioactmat.2023.02.011.

Han X, Zhang H, Butowska K, Swingle KL, Alameh MG, Weissman D, et al. An ionizable lipid toolbox for RNA delivery. Nat Commun. 2021;12(1):7233. doi:10.1038/s41467-021-27493-0.

Kenjo E, Hozumi H, Makita Y, Iwabuchi KA, Fujimoto N, Matsumoto S, et al. Low immunogenicity of LNP allows repeated administrations of CRISPR-Cas9 mRNA into skeletal muscle in mice. Nat Commun. 2021;12(1):7101. doi:10.1038/s41467-021-26714-w.

Liu K, Nilsson R, Lazaro-Ibanez E, Duan H, Miliotis T, Strimfors M, et al. Multiomics analysis of naturally efficacious lipid nanoparticle coronas reveals high-density lipoprotein is necessary for their function. Nat Commun. 2023;14(1):4007. doi:10.1038/s41467-023-39768-9.

Mochida T, Fujimoto N, Asahina M, Asano S, Araki S, Inukai N, Hotta A. Muscle satellite cell editing by LNP-CRISPR-Cas9 to resist muscle injury. Cell Rep. 2025 Dec 16;45(1):116695. doi: 10.1016/j.celrep.2025.116695. Epub ahead of print. PMID: 41411128.

Pardi N, Tuyishime S, Muramatsu H, Kariko K, Mui BL, Tam YK, et al. Expression kinetics of nucleoside-modified mRNA delivered in lipid nanoparticles to mice by various routes. J Control Release. 2015;217:345-351. doi:10.1016/j.jconrel.2015.08.007.

Saha M, Bhake R, Meyers CA, Coleman K, Trivedi PD, Gallman M, et al. Enhancing AAV9-UFmuDys1 gene therapy efficacy through immunosuppression in mice with pre-existing immunity and enabling redosing strategies for Duchenne muscular dystrophy. Hum Gene Ther. 2025. doi:10.1177/10430342251385586.

Sagar R, David AL. Fetal therapies—stem cell transplantation; enzyme replacement therapy; in utero genetic therapies. Best Pract Res Clin Obstet Gynaecol. 2024;97:102542. doi:10.1016/j.bpobgyn.2024.102542.

Sebastiani F, Yanez Arteta M, Lerche M, Porcar L, Lang C, Bragg RA, et al. Apolipoprotein E binding drives structural and compositional rearrangement of mRNA-containing lipid nanoparticles. ACS Nano. 2021;15(4):6709-6722. doi:10.1021/acsnano.0c10064.

Vasileva O, Zaborova O, Shmykov B, Ivanov R, Reshetnikov V. Composition of lipid nanoparticles for targeted delivery: application to mRNA therapeutics. Front Pharmacol. 2024;15:1466337. doi:10.3389/fphar.2024.1466337.

Zhou SK, Luo JT, Chen YF, Lu ZD, Jian QH, Jiang SQ, et al. Muscle-specific gene editing therapy via mammalian fusogen-directed virus-like particles. Nat Commun. 2025;16:9145. doi:10.1038/s41467-025-64200-9.