Mutation-agnostic base editing of the progerin farnesylation site rescues Hutchinson-Gilford progeria syndrome phenotypes in neuromuscular organoids

Dong-Woo Kim; Eun-Ji Kwon; Beom-Jin Jeon; Dong-Hyeok Kwon; Goo Jang; Youngyoon Yoon; Yuna Hwang; Hyukjin Lee; Hyuk-Jin Cha

doi:10.1038/s41467-026-77251-3

. 2026 Aug 31;17:10380. doi: 10.1038/s41467-026-77251-3

Mutation-agnostic base editing of the progerin farnesylation site rescues Hutchinson-Gilford progeria syndrome phenotypes in neuromuscular organoids

Dong-Woo Kim ^1,^#, Eun-Ji Kwon ^1,^#, Beom-Jin Jeon ², Dong-Hyeok Kwon ², Goo Jang ², Youngyoon Yoon ³, Yuna Hwang ³, Hyukjin Lee ^1,⁴, Hyuk-Jin Cha ^1,^4,^✉

PMCID: PMC13627717 PMID: 42816477

Abstract

Hutchinson Gilford progeria syndrome (HGPS) is a fatal premature aging disorder caused by pathogenic farnesylated lamin A variants that disrupt nuclear architecture and DNA repair. Current therapies, including farnesyltransferase inhibitors, provide only modest survival benefits and lack molecular specificity, while mutation-specific genome-editing strategies cannot address atypical laminopathies. Here, we develop Farnesylation Amino acid Targeted Editing (FATE), a mutation-agnostic base-editing platform that selectively disrupts the LMNA farnesylation motif without affecting other farnesylated proteins. Using isogenic human pluripotent stem cell derived neuromuscular organoids (NMOs), we reveal muscle-specific progerin accumulation that sequesters 53BP1 and abolishes DNA damage foci formation. FATE eliminates perinuclear progerin, restores 53BP1 mobility, reconstitutes DNA repair foci, and normalizes heterochromatin architecture. Importantly, transient delivery of FATE mRNA conjugated with lipid nanoparticles to HGPS NMOs achieves efficient base editing and phenotypic rescue. These findings establish FATE as a mutation-independent therapeutic strategy targeting a fundamental pathogenic mechanism in HGPS and provide a proof-of-concept for RNA-based in situ genome editing in progeroid disease.

Subject terms: Stem-cell biotechnology, Targeted gene repair, Disease model

A mutation-agnostic base-editing strategy targeting the LMNA farnesylation site restores DNA repair and nuclear architecture in human progeria neuromuscular organoids and supports transient mRNA-based therapeutic genome editing.

Introduction

Hutchinson–Gilford progeria syndrome (HGPS; OMIM #176670) is a rare genetic disorder that causes premature aging and leads to early mortality, typically within the first two decades of life¹. In over 90% of cases, a de novo LMNA mutation (c.1824C>T; p.G608G) activates a cryptic splice donor in exon 11, producing progerin—a truncated form of lamin A that retains its C-terminal cysteine-serine-isoleucine-methionine (CSIM) CAAX motif and hence remains permanently farnesylated^2,3. Because progerin lacks the cleavage site for the ZMPSTE24 metalloprotease, it escapes normal post-translational processing and accumulates at the inner nuclear membrane, causing toxic nuclear-envelope deformation⁴. Beyond the classical c.1824C>T mutation, other LMNA variants (e.g., c.1821G>A, c.1822G>A, c.1868C>G, c.1940T>G, c.1968G>A) are known to impair ZMPSTE24-mediated cleavage, producing farnesylated lamin A variants with progerin-like properties⁵. Likewise, ZMPSTE24 loss-of-function mutations result in accumulation of farnesylated prelamin A that causes related laminopathies such as restrictive dermopathy and mandibuloacral dysplasia^6,7. Collectively, despite their distinct genetic causes, all of these conditions share a common pathogenic mechanism: persistent lamin A farnesylation.

HGPS phenotypes are most pronounced in mesoderm-derived tissues such as vascular smooth muscle, bone, skin dermis, and adipose tissue^8,9. The persistent accumulation of progerin disrupts the nuclear lamina, causing mechanical fragility, aberrant mitosis, and accumulation of DNA damage¹⁰. Impaired recruitment of DNA repair proteins such as 53BP1 and RAD51 compromises double-strand break (DSB) repair and triggers checkpoint activation, ultimately promoting chronic genomic instability and senescence¹⁰. Simultaneously, heterochromatin erosion and detachment of lamina-associated domains alter gene expression programs^11,12. Together, these nuclear and epigenetic perturbations drive progressive tissue degeneration and functional decline. The cardiovascular system, particularly vascular smooth muscle cells, is especially vulnerable to these stresses, and its degeneration underlies the cardiovascular complications that constitute the leading cause of death in HGPS patients⁹.

The first and only FDA-approved therapy for HGPS, lonafarnib (Zokinvy), is a farnesyltransferase inhibitor (FTI) that blocks progerin farnesylation, improves nuclear morphology, and extends median survival by approximately 2.5 years^13,14. This therapeutic benefit underscores the pathogenicity of farnesylated progerin; however, the activity of lonafarnib is not specific to progerin alone. Long-term FTI use is therefore constrained by systemic toxicity, as farnesylation is essential for the proper function of multiple cellular proteins, including small GTPases (e.g., RAS isoforms and RHOE) and nuclear lamins (LMNB1 and LMNB2). Lonafarnib-induced hypo-farnesylation of these proteins impairs signal transduction, cytoskeletal dynamics, and vesicular trafficking, with consequent off-target effects such as gastrointestinal symptoms, anemia, and myelosuppression^13,15. Other therapeutic strategies have emerged, each of which also carries limitations. Rapamycin, which promotes progerin clearance via autophagy, has shown efficacy in preclinical models but also possesses broad immunosuppressive activity that hampers its clinical utility¹⁶. Mutation-specific approaches include adenine base editor (ABE) to directly revert the pathogenic c.1824C>T mutation¹⁷, which is applicable only to classical HGPS carrying this specific variant. Other strategies with potentially broader applicability include antisense oligonucleotides that block the aberrant splice donor in LMNA^18,19, which could in principle be adapted to different splice-activating mutations, and CRISPR–Cas9-mediated selective disruption of lamin A while sparing lamin C²⁰, which is likewise independent of the underlying mutation but carries risks of large genomic deletions or rearrangements²¹. However, none of these approaches directly addresses disorders such as ZMPSTE24 deficiency, which also result in the accumulation of farnesylated prelamin A.

Here, we introduce Farnesylation Amino-acid Targeted Editing (FATE), a base editing approach that targets the CSIM motif of LMNA rather than a specific pathogenic nucleotide. By installing a cysteine-to-arginine substitution in the CAAX box, FATE abolishes farnesylation of not only prelamin A but also progerin, thereby preventing the aberrant retention at the inner nuclear membrane regardless of upstream mutation. Using neuromuscular organoids (NMOs) derived from isogenic LMNA-mutant human pluripotent stem cells (hPSCs), we demonstrate that FATE eliminates perinuclear progerin aggregates and restores DNA damage repair foci. Because FATE disrupts a shared pathogenic post-translational modification rather than a patient-specific LMNA mutation, it provides a mutation-agnostic, potentially pan-laminopathy therapy, expanding the scope of genome editing beyond existing HGPS-targeted approaches.

Results

Establishment of two isogenic NMO models with mesoderm-specific progerin expression

To overcome the limited availability of primary fibroblasts from HGPS patients for in vitro studies^12,16, human induced pluripotent stem cells (iPSCs) derived from HGPS individuals are widely utilized to generate cellular models for mechanistic investigations^22,23. To minimize confounding effects from genetic background variability^24,25, we established two isogenic pairs of human pluripotent stem cells (hPSCs) differing only at the LMNA locus: (i) a pair of patient-derived iPSCs carrying the HGPS c.1824C>T mutation and its corrected counterpart (henceforth HGPS- and Edit-iPSCs), and (ii) a pair of hESCs harboring either wild-type or engineered c.1824C>T mutant alleles (henceforth WT- and Mutant-hESCs) (Fig. 1a). These isogenic hPSC lines were generated using prime editor as previously described^26,27 (Fig. 1b). Consistent with earlier reports²², undifferentiated hPSCs exhibited negligible LMNA expression and thus no detectable progerin production, even in HGPS-iPSCs (Fig. S1a).

Fig. 1 — a Graphical description showing the establishment of two sets of isogenic hPSC pairs. Created in BioRender. Kwon, E. (2026) https://BioRender.com/1dpr3dh. b Sanger sequencing results showing the genotypes of the c.1824 region for each hPSC line. Created in BioRender. Kwon, E. (2026) https://BioRender.com/y9wtbah. c Graphical overview of neuromuscular organoid (NMO) differentiation. Bright-field images of representative organoid at different stages (days 2, 5, 10, and 30). Created in BioRender. Kwon, E. (2026) https://BioRender.com/1xhgqb5. d Immunofluorescence image showing established day 70 (D70) NMOs derived from two sets of isogenic hPSCs. Magenta indicates fast myosin heavy chain (muscle marker), and green indicates Tuj1 (neuron marker). Scale bar, 500 μm. Representative of three independent experiments; four organoids per group. e Immunofluorescence images of day 40 HGPS and Edit NMOs stained with Progerin (green), Desmin (magenta), and DAPI (blue). Scale bar, 25 μm. Representative of three independent experiments; four organoids per group. f, g Immunoblot analysis of D70 NMOs (both skeletal muscle regions and neural regions) derived from iPSC and hESC pairs, showing Lamin A/C and Vinculin. Red asterisks indicate progerin-specific bands. Representative of two independent experiments; each blot was performed on lysates pooled from ten organoids per group. h Immunofluorescence images of the muscular region of HGPS-NMOs at day 40 (D40), day 74 (D74), and day 156 (D156), stained with Progerin (green), Desmin (magenta), and DAPI (blue) (left). Quantification showing the percentage of Progerin-positive cells at each timepoint (right). Scale bar, 20 μm. Data are presented as mean values ± SD; n = 4 (D40), 5 (D74), and 5 (D156) independent organoids. Statistical significance was determined by ordinary one-way ANOVA (F(2, 11) = 81.11, P < 0.0001) followed by Tukey’s two-sided multiple comparisons test with adjustment for multiple comparisons. Adjusted P values: D40 vs D74, P = 0.0022 (95% CI of the difference, 5.09–19.94); D40 vs D156, P = 2.3 × 10⁻⁷ (95% CI, 26.71–41.56); D74 vs D156, P = 1.2 × 10⁻⁵ (95% CI, 14.62–28.62).

Given that HGPS pathologies are prominent in mesoderm-derived tissues such as skeletal muscle^8,9, whereas ectodermal tissues such as neurons remain largely spared, we generated NMOs from the isogenic hPSC lines using established protocols²⁸ (Fig. 1c) to allow the simultaneous modeling of mesodermal (i.e., muscular) and ectodermal (neural) lineages within a shared genetic background. We established two NMO pairs, HGPS- vs. Edit-NMOs and Mutant- vs. WT NMOs, which exhibited comparable compartmentalization of neural and muscular tissues, as confirmed by expression of lineage-specific markers (Fig. 1d and S1b) and ultrastructural features such as M- and Z-bands, synaptic vesicles, and neural filaments (Fig. S1c). Despite their comparable tissue organization, immunofluorescence revealed progerin accumulation to occur exclusively in the muscular compartments of HGPS-NMOs (Fig. 1e) and Mutant-NMOs (Fig. S1d). Immunoblotting of whole NMOs lysates, containing both muscular and neural compartments, further validated robust progerin expression in the disease NMOs (HGPS: Fig. 1f; Mutant: Fig. 1g), whereas none was detected in their respective controls. Notably, progerin accumulation increased with NMO maturation: the proportion of progerin-positive cells within the muscular compartment rose from approximately 7% at Day 40 to 20% at Day 74 and 41% at Day 156 (Fig. 1h).

Myopathy-associated gene signatures in the muscle compartment of HGPS-NMOs

Given the selective accumulation of progerin in the muscle compartment of HGPS-NMOs (Fig. 1), we performed single-cell RNA sequencing (scRNA-seq) to further characterize cell-type-specific transcriptional alterations. Consistent with the cell populations previously defined in established NMO models²⁸, graph-based clustering of single-cell transcriptomes identified twelve distinct cell subtypes: skeletal muscle (CL3, CL7), neural cells (CL1, CL2, CL4, CL5), sclerotome (CL0), endothelial (CL9), and various epithelial populations (CL6, CL8, CL10, CL11) (Fig. 2a). The overall distribution of cell types was comparable between HGPS- and Edit-NMOs (Fig. 2b), and lineage-specific marker gene expression confirmed NMO-derived populations to have well-defined differentiation trajectories toward muscle and neural lineages (Fig. 2c). Within the skeletal muscle lineage, we identified a continuum of differentiation states—from myogenic progenitor/satellite-like cells (cluster 1) through intermediate myocytes (cluster 2) to mature contractile skeletal muscle fibers (cluster 3) (Fig. 2d, e). Consistent with the maturation-dependent increase in progerin-positive cells shown in Fig. 1h, LMNA expression was significantly higher in intermediate myocytes (cluster 2) and skeletal muscle fibers (cluster 3) than in myogenic progenitors (cluster 1) (Fig. S2a). Notably, HGPS-NMOs exhibited a reduced proportion of myogenic progenitors compared to Edit-NMOs, accompanied by increased fractions of differentiated myocytes and mature muscle fibers (Fig. 2d). This shift in cellular composition likely reflects an adaptive regenerative response to chronic muscle damage, and resembles patterns observed in scRNA-seq analyses of Duchenne muscular dystrophy²⁹. Consistent with this interpretation, gene set enrichment analysis revealed significant enrichment of myopathy-associated gene signatures specifically in the muscle lineage of HGPS-NMOs (Fig. 2f); in addition, HGPS muscle populations showed marked upregulation of genes involved in the Striated Muscle Contraction Pathway, including actinin family members (Fig. 2g). These findings align with transcriptomic hallmarks of muscle degeneration and compensatory remodeling.

a Integration, and clustering of single-cell transcriptomes from D70 HGPS- and Edit-NMOs. Cell clusters were annotated based on the enrichment of canonical cell type marker genes. b Comparison of cellular composition between HGPS- and Edit-NMOs. c Identification of skeletal muscle and neural cell populations using lineage-specific markers. d Sub-clustering of the skeletal muscle trajectories reveals 3 distinct subpopulations. Shades of green indicate skeletal muscle subclusters. Relative proportions of each subcluster in HGPS organoids are shown below. e Dot plot showing representative gene expression across three muscle subclusters. f Human Phenotype Ontology terms enriched in HGPS muscle cells compared to Edit controls. Enrichment was assessed by pre-ranked gene set enrichment analysis (fgsea) with genes ranked by average log₂ fold change. P values were two-sided and were adjusted for multiple comparisons across all tested gene sets by the Benjamini–Hochberg procedure; adjusted P < 0.05 was considered significant. g Pathway enrichment analysis of skeletal muscle compartments in HGPS- and Edit-NMOs. Created in BioRender. Kwon, E. (2026) https://BioRender.com/vneh5lf (Fig. 2a, g).

In contrast, the neural lineage—comprising trunk neural crest derivatives (cluster 7) that further differentiated into glial and Schwann cells (cluster 8) (Fig. S2b and c)—exhibited no enrichment of disease-related gene sets. Specifically, in HGPS neural populations, both gene enrichment and differential expression analysis failed to reveal transcriptomic signatures associated with pathological phenotypes (Fig. S2d and e). Overall, this muscle-specific, neuron-sparing transcriptomic pathology highlights the value of the HGPS-NMO model for dissecting lineage-restricted mechanisms in progeroid disease.

Impairment of DNA damage repair machinery in the muscle compartment of HGPS-NMOs

Following identification of robust progerin accumulation in the muscle compartment (Fig. 2f) and the muscle-specific HGPS transcriptomic signature (Fig. 3), we investigated the underlying mechanisms by analyzing gene sets with reduced enrichment specifically in HGPS-derived muscle cells compared to Edit controls. Gene sets exclusively altered in the muscle compartment of HGPS—either upregulated in Edit-NMOs or downregulated in HGPS-NMOs—likely represent HGPS-specific transcriptional responses. Notably, gene sets that were upregulated in the muscle compartment of Edit-NMOs compared to HGPS-NMOs, but not in the neural compartment, were enriched for DNA damage repair pathways (highlighted in red, Fig. 3a), consistent with previous reports linking genome instability to HGPS pathogenesis^10,30. Additionally, gene sets downregulated in HGPS-NMOs included pathways associated with metabolism and autophagy (highlighted in blue, Fig. S3a), both of which have been implicated in skeletal muscle dysfunction in HGPS mouse models^16,31,32.

Fig. 3 — a Heatmap of GSVA-based KEGG pathway enrichment in neural and muscle compartments of HGPS and Edit organoids; DNA repair pathways in red, metabolism/autophagy sets in blue. b Cells with LMNA in the top 25% were designated the LMNA-high group (top); UMAP shows LMNA-high and LMNA-low distribution across organoids (bottom). c Heatmap of DNA repair–related gene set enrichment in LMNA-high and LMNA-low groups within neural and muscle compartments of HGPS and Edit organoids. d Immunofluorescence of D70 HGPS and Edit NMOs stained with Progerin (green), 53BP1 (magenta) and DAPI (blue). Scale bar, 25 μm. e Immunoblot of D70 NMOs from iPSC pairs, showing α-tubulin, Lamin A/C, pHH3 (T3), PCNA and 53BP1. f Immunofluorescence of D70 HGPS and Edit NMOs stained as in d, under mock and 5 Gy X-ray irradiation for 1 h. Scale bar, 25 μm. Quantification of progerin-positive cells with nuclear-membrane 53BP1 (53BP1-NM) in the skeletal muscle (S.M.) and neural (N) regions (right). Data are mean values ± SD; n = 6 independent organoids per group. Two-tailed unpaired t test (S.M., mock vs IR): t(10) = 0.41, P = 0.69 (95% CI, −21.79 to 15.00), ns. The neural region was negative (0%) and excluded. g, h, Immunofluorescence of HGPS and Edit NMOs stained with Progerin (green), DAPI (blue), and DNA damage markers (magenta) under mock, 1 h, and 4 h after 5Gy irradiation: g γH2AX; h pATM (S1981). Scale bar, 25 μm. Foci failed to form in progerin-positive muscle cells of HGPS-NMOs (g: 322 of 323 cells, n = 4 organoids, D82; h: 58 of 58 cells, n = 5 organoids, D70). i TEM of the muscle region of D70 NMOs from two isogenic HGPS hPSC pairs, showing nuclear heterochromatin. j Immunofluorescence of the muscle region of D70 HGPS- and WT-NMOs stained with Progerin (green), H3K9me3 (magenta), and DAPI (blue); insets, boxed regions. Scale bar, 20 μm. k, Defective 53BP1, γH2AX and pATM (S1981) localization in progerin-expressing HGPS- and Mutant-NMOs. Created in BioRender. Kwon, E. (2026) https://BioRender.com/bm7b99k. Representative of three independent experiments with four organoids per group (d, j), or two with twelve organoids (e); i, two organoids.

Given that HGPS-iPSCs are heterozygous and thus express both wild-type and mutant LMNA alleles at equivalent levels (Fig. S3b), we hypothesized that for individual cells, progerin expression would scale proportionally with LMNA transcript abundance. We therefore stratified cells based on LMNA expression, defining the top 25% as LMNA-high and the remainder as LMNA-low (Fig. 3b). This designation represents a relative transcript-level classification rather than a direct measurement of progerin protein. Nevertheless, among the transcriptomic features examined in our single-cell dataset, LMNA abundance provides the most direct proxy for potential progerin burden, because the progerin-regulating splicing factors SRSF1 and SRSF2 were neither differentially expressed between the LMNA-high and LMNA-low groups nor preferentially enriched in the muscle compartment^33,34 (Fig. S3c). Strikingly, within the muscle compartment of HGPS-NMOs, DNA damage repair gene sets were more suppressed in the LMNA-high group (Fig. 3c), supporting an association between increased LMNA transcription, greater potential progerin burden, and impaired DNA repair capacity in HGPS muscle cells.

Lamin A and C are known to interact with p53-binding protein 1 (53BP1)³⁵, which is a key mediator of DSB repair through its recruitment to DNA damage foci³⁶; notably, that recruitment can be disrupted by the presence of lamin B³⁷ or prelamin A³⁸. To assess whether progerin impairs 53BP1 localization, we examined 53BP1 distribution in HGPS-NMOs. Intriguingly, in the muscle compartment of HGPS-NMOs, 53BP1 was selectively enriched at the nuclear membrane, where progerin was prominently accumulated, suggesting that 53BP1 is displaced from its normal nucleoplasmic distribution (Fig. 3d and S3d). Immunoblotting of the nuclear membrane fraction from whole NMOs further supported the increased membrane association of 53BP1 in HGPS-NMOs (Fig. 3e). Consistently, this nuclear membrane enrichment was not observed in WT-NMOs (Fig. S3d), suggesting that progerin-dependent sequestration prevents 53BP1 recruitment to DNA damage sites. Following ionizing radiation (IR), progerin-positive muscle cells failed to form 53BP1 foci, with 53BP1 remaining at the nuclear periphery, whereas neural cells lacking detectable progerin formed robust nuclear foci (Fig. 3f).

Similarly, γH2AX—a canonical marker of DSBs characterized by phosphorylation at serine 139³⁹—was absent from the nuclei of nearly all progerin-positive muscle cells of HGPS-NMOs even after IR (Fig. 3g), whereas normal γH2AX foci formation was preserved in the neural compartment of HGPS-NMOs (Fig. S3e). Likewise, phosphorylated ATM (pATM), the upstream kinase required for γH2AX activation, formed nuclear DNA damage foci in Edit-NMOs following IR but remained undetectable in progerin-positive muscle cells of HGPS-NMOs (Fig. 3h). This defective γH2AX and pATM response was associated with abnormal retention of ATM at the progerin-positive nuclear membrane: approximately half of the progerin-positive cells exhibited nuclear membrane–localized ATM (ATM-NM), which persisted even after IR (Fig. S3f). Mutant-NMOs, generated from c.1824C>T mutant hESCs, exhibited results consistent with those observed in HGPS-NMOs: γH2AX and pATM foci were absent in progerin-positive muscle cells (Fig. S3g and h) but preserved in the neural compartment (Fig. S3i).

Transmission electron microscopy (TEM) revealed that the loss of peripheral heterochromatin architecture—a hallmark of genomic instability in HGPS¹⁰—was specifically observed in the muscle compartments of HGPS- and Mutant-NMOs, but not in their respective isogenic controls, Edit- and WT-NMOs (Fig. 3i), whereas peripheral heterochromatin remained intact in the neural compartments (Fig. S3j). To complement the ultrastructural information obtained by TEM with a progerin-resolved molecular readout, we further assessed H3K9me3, a representative heterochromatin marker known to be reduced in HGPS fibroblasts²². Consistent with the TEM findings, H3K9me3 was markedly reduced in progerin-positive cells of HGPS-NMOs compared with WT-NMOs (Fig. 3j). Given the essential roles of ATM-mediated γH2AX formation and 53BP1 recruitment in DSB repair (Fig. 3k, left), persistent sequestration of these factors by progerin likely underlies the chronic DNA damage and genome instability characteristic of HGPS (Fig. 3k, right).

Specific farnesylation amino acid targeted editing (FATE) of LMNA

Previous gene-editing strategies for HGPS include: (i) Cas9-mediated disruption of lamin A, which can eliminate progerin expression but carries the risk of introducing large on-site deletions^40,41 or even chromosomal rearrangements⁴²; and (ii) ABE, which precisely corrects the pathogenic c.1824C>T; p.G608G mutation but is applicable only to HGPS patients carrying this classical HGPS variant. For patients harboring non-classical HGPS mutations that also result in aberrant progerin production, gene correction requires customized editing strategies—such as near-PAMless ABEs⁴³, prime editor (PE), or cytosine-to-guanine base editors (CGBE)—each of which presents distinct technical challenges, including the presence of bystander adenines (Fig. S4a). To overcome these limitations and establish a mutation-agnostic approach, we targeted progerin farnesylation—a well-established driver of HGPS pathology—by designing a base editing strategy to disrupt the conserved CSIM motif in LMNA. As proof of principle, we tested the effect of unfarnesylated progerin generated by substituting the cysteine residue with serine in a GFP–progerin reporter cell line. This substitution reduced nuclear membrane retention of progerin (Fig. S4b) and restored DNA damage repair foci (Fig. S4c), confirming that disrupting the farnesylation motif is a viable alternative to mutation-specific correction. Building on this rationale, we engineered a single-guide RNA (sgRNA) to introduce the arginine (CGC) substitution at position 661 (c.1981T>C; p.C661R) coding cysteine (TGC), thereby preventing farnesylation and abolishing progerin accumulation at the nuclear envelope—a mutation-agnostic strategy we term Farnesylation Amino acid Targeted Editing (FATE) (Fig. 4a, left). This approach circumvents both the genotoxic risks associated with Cas9-induced large deletions and the complexity of tailoring editing tools to individual HGPS variants, while abolishing post-translational farnesylation of progerin and preventing its aberrant retention at the inner nuclear membrane regardless of the upstream mutation (Fig. 4a, right). To identify optimal editing conditions, we evaluated multiple ABE-compatible guide RNAs that positioned the target TGC codon within the editing window of various protospacer adjacent motif (PAM) sequences and ABE versions (Fig. S4d). Among these, only the guide RNA sequence TGCTGCAGTTCTGGGGGCTC, which employs a TAA PAM, exhibited significant editing activity in HGPS-iPSCs (c.1824C>T; p.G608G) when paired with SpRY-ABE8e⁴⁴, and therefore was used for the generation of FATE-iPSCs (c.1824C>T; p.G608G and c.1981T>C; p.C661R).

Fig. 4 — a Graphical description showing the FATE strategy, which selectively abolishes farnesylation of progerin by introducing an additional c.1981T>C (p.C661R) mutation into the LMNA gene using base editing, without affecting other CAAX box–harboring proteins. Created in BioRender. Kwon, E. (2026) https://BioRender.com/43tng5t. b Genome-wide A-to-G conversion frequency at potential gRNA-targetable sites. c Dot plot showing genes with high A-to-G conversion rates in HGPS wild-type and HGPS plasmid-treated samples. d A-to-G conversion frequency in non-coding regions of three candidates by targeted deep sequencing. e DNA sequence logo representation of the CAAX motif in Small GTPase, Lamin, and CENP family genes.

Given that systemic hypo-farnesylation is a major limitation of long-term FTI therapy in HGPS, we first assessed the target selectivity of FATE by whole-exome sequencing (WES) and targeted deep sequencing of HGPS-iPSCs with or without FATE. FATE achieved 48% A-to-G conversion at the intended LMNA site with no guide RNA–dependent off-target edits detected above 0.001% (Fig. 4b and S4e). At sequencing depths >400 reads, WES identified eight and five A-to-G conversion sites in HGPS-iPSCs with or without FATE, respectively; all showed low homology to the guide RNA (maximum 9/20 matching nucleotides) and corresponded to known non-pathogenic SNPs (Fig. 4c and S4f). To capture potential off-target activity in non-coding regions not covered by WES, we performed targeted deep sequencing of three candidate loci. Only one site (Off-3), containing a single mismatch to the guide RNA, exhibited detectable editing (~18% A•T-to-G•C conversion), and this site resides in an intergenic region, suggesting minimal clinical relevance (Fig. 4d and S4g). Importantly, consistent with its sequence-specific design, FATE did not alter the CAAX motifs of other farnesylated proteins—including small GTPases, lamins, and CENP family members—thereby avoiding the broad suppression of protein farnesylation observed with FTI treatment (Fig. 4e).

Loss of progerin accumulation and restoration of DNA damage foci in FATE-NMOs

FATE-iPSCs generated from HGPS-iPSCs (Fig. 5a) were differentiated into FATE-NMOs, which exhibited well-organized skeletal muscle and neural compartments (Fig. S5a). We next assessed whether disruption of the LMNA farnesylation site could reverse hallmark HGPS phenotypes, including perinuclear progerin accumulation and defective DNA damage foci (Fig. 5b). As anticipated, the characteristic perinuclear progerin aggregation observed in HGPS-NMOs was absent in FATE-NMOs; instead, progerin exhibited a diffuse nucleoplasmic localization consistent with effective disruption of membrane tethering (Fig. 5c). This redistribution likely accounts for the apparently weaker progerin immunofluorescence signal in FATE-NMOs, because the protein is dispersed throughout the nucleoplasm rather than concentrated at the nuclear periphery⁴⁵. Notably, this relocalization occurred despite the continued progerin expression in FATE-NMOs, in contrast to Edit-NMOs, in which progerin production was eliminated through correction of the pathogenic LMNA mutation (Fig. 5c). Intriguingly, lamin A in FATE-NMOs exhibited an upward shift in apparent molecular weight relative to both HGPS and Edit-NMOs (Fig. 5d, top). This shift likely reflects impaired ZMPSTE24-mediated processing, which requires prior farnesylation for substrate recognition and cleavage⁴⁶ (Fig. 5d, bottom). Consistent with the loss of perinuclear progerin accumulation, FATE-NMOs restored IR-induced formation of γH2AX and pATM foci (Fig. 5e, f). The abnormal nuclear membrane retention of ATM observed in HGPS-NMOs was also abolished in FATE-NMOs, resulting in a distribution comparable to that in Edit-NMOs (Fig. S5b). Furthermore, co-staining of 53BP1 with Emerin, a nuclear membrane protein⁴⁷, confirmed the release of 53BP1 from the nuclear periphery and restoration of IR-induced 53BP1 foci formation in FATE-NMOs (Fig. S5c).

To further illustrate the marked phenotypic changes in FATE-NMOs, we applied three-dimensional confocal imaging to visualize the loss of perinuclear progerin accumulation and restoration of DNA damage foci before and after IR (Fig. 5g and Supplementary Movies 1–6). TEM further revealed restoration of peripheral heterochromatin architecture in the muscle compartment of FATE-NMOs (Fig. 5h and S5d), a feature severely compromised in HGPS-NMOs and indicative of genomic instability¹⁰. In line with this, the reduced H3K9me3 signal observed in progerin-positive cells of HGPS-NMOs was restored in FATE-NMOs to a level comparable to that in Edit-NMOs (Fig. 5i), confirming recovery of the disrupted heterochromatin state at the molecular level. Collectively, these results indicate that FATE not only restores DNA damage response foci formation but also promotes broader recovery of nuclear architecture in a lineage-specific manner.

To define the transcriptional basis of phenotypic rescue, we performed scRNA-seq on FATE-NMOs and compared them with HGPS- and Edit-NMOs. FATE-NMOs maintained five major cell types, confirming preservation of lineage identity (Fig. S5e), and exhibited continuous differentiation trajectories within both the skeletal muscle (Fig. S5f) and neural lineages (Fig. S5g). Notably, the skeletal muscle lineage of FATE-NMOs showed increased representation of myogenic progenitor populations, mirroring Edit-NMOs (Fig. 2d). PCA of KEGG pathway enrichment profiles indicated that the muscle transcriptome of FATE-NMOs clustered closely with Edit-NMOs and diverged from HGPS-NMOs (Fig. 5j). Considering that the FATE dataset was generated independently, the separation observed among neural-lineage samples was interpreted with caution, while the overall neural lineage composition and differentiation trajectories remained preserved. In addition, disease-associated transcriptional alterations observed in HGPS-NMOs—including reduced myogenic progenitors (Fig. 2d), enrichment of myopathy-associated gene signatures (Fig. 2f), and downregulation of DNA damage repair pathways (Fig. 3c)—were consistently reversed in FATE-NMOs (Fig. 5k, l). Based on prior evidence of age-associated molecular changes in human muscle⁴⁸, we classified transcriptomic signatures from bulk RNA-seq of human muscle stem cells⁴⁹ as either young (15–40 years) or old (≥ 60 years). Gene set enrichment analysis revealed that both FATE-NMOs and Edit-NMOs were significantly enriched for the young signature, supporting a shift of the FATE muscle transcriptome toward the corrected state (Fig. S5h).

Unlike mutation-specific base editing strategies that target individual pathogenic variants^17,50, the FATE approach offers a key advantage by extending therapeutic applicability to non-classical forms of HGPS. As an initial test, we examined patient-derived fibroblasts carrying the non-classical c.1968+5 G > A mutation, which produces a Δ90 progerin variant (Fig. S5i). FATE treatment improved the abnormal nuclear morphology of these cells (Fig. S5j and k). However, because aHGPS fibroblasts exhibit limited proliferative capacity and readily adopt a senescence-like state, we reprogrammed them into aHGPS-iPSCs to enable more extensive analysis. We then generated FATE-edited aHGPS-iPSCs and differentiated both lines into aHGPS-NMOs and aHGPS-FATE-NMOs (Fig. 5m). In aHGPS-NMOs, 53BP1 was abnormally retained at the nuclear membrane and failed to form normal DNA damage foci following IR, whereas aHGPS-FATE-NMOs showed release of 53BP1 from the nuclear periphery and restoration of IR-induced 53BP1 foci formation (Fig. 5n). These findings provide functional evidence that FATE rescues a pathogenic DNA damage response phenotype in a non-classical HGPS background.

Recovery of HGPS-NMOs by delivery of FATE mRNA

Given the known risks of adeno-associated virus (AAV)–mediated delivery—including potential immunotoxicity, particularly in pediatric patients⁵¹, and its limited cargo capacity—we opted to avoid AAV for delivering ABE components. Instead, we tested synthetic mRNA for transient SpRY-ABE8e expression, a platform recently optimized for in vivo gene editing^52,53, in the context of FATE in HGPS-iPSCs. For proof of concept, mRNA-based FATE (mRNA-FATE) achieved higher editing efficiency in HGPS-iPSCs compared to plasmid-based delivery (Fig. 6a). We next encapsulated green fluorescent protein (GFP) mRNA-FATE in lipid nanoparticles (LNPs)⁵⁴, and applied them directly to Day 25 HGPS-NMOs by simple incubation (Fig. 6b). Interestingly, in our neuromuscular organoid model, GFP mRNA-LNPs showed preferential uptake in the muscle compartment. As a result, editing efficiency with ABE8e mRNA-LNP, measured by next-generation sequencing of whole organoids appeared low, reflecting dilution by unedited neuronal and non-muscle cells (Fig. 6c). To address this limitation, we delivered mRNA-LNPs by microinjection directly into the muscle region of NMOs (Fig. 6d), mimicking a local intramuscular injection strategy. To determine an appropriate dosage, GFP mRNA-LNPs were microinjected at 257.58 μg/ml, delivering a total of 515.16 ng (2 μl) of encapsulated mRNA. This resulted in robust GFP expression, confirming the feasibility of localized delivery into organoids (Fig. 6e, left). Following microinjection of ABE8e mRNA-LNPs into Day 26 HGPS-NMOs and an additional 10 days of culture, next-generation sequencing confirmed successful base editing at the target site, with a maximum efficiency of 8.55% (Fig. 6e, right). Given the <10% editing efficiency achieved with FATE mRNA, the observed phenotypic rescue—characterized by loss of perinuclear progerin aggregation and restoration of DNA damage foci—likely reflects a mixed population of edited and unedited cells within the organoids (Fig. 6f). Consistent with genetically edited FATE-NMOs, transient FATE mRNA delivery abolished progerin accumulation at the nuclear membrane, restored its diffuse nuclear distribution, and reestablished DNA damage foci marked by γH2AX (Fig. 6g) and 53BP1 (Fig. 6h). These results demonstrate that even partial, transient mRNA-mediated base editing is sufficient to restore functional DNA repair. Collectively, our findings show that transient FATE mRNA delivery can phenotypically and molecularly rescue HGPS-NMOs, underscoring its translational promise as a mutation-agnostic therapeutic strategy.

Fig. 6 — a FATE delivery into HGPS iPSCs via plasmid and mRNA (left) and A-to-G conversion by Sanger sequencing (right). No editing was detected in mock. Mean values ± SD; n = 4 (mock), 5 (plasmid), and 4 (mRNA) independent experiments. Ordinary one-way ANOVA (F(2, 10) = 32.61, P < 0.0001) with Tukey’s two-sided test. Adjusted P: mock vs plasmid, 0.0164 (95% CI, −47.56 to −5.24); mock vs mRNA, <0.0001 ( − 87.55 to −42.95); plasmid vs mRNA, 0.0013 ( − 60.01 to −17.69). Created in BioRender. Kwon, E. (2026) https://BioRender.com/dvc4h4z. b Delivery of FATE mRNA-LNPs into HGPS-NMOs. Created in BioRender. Kwon, E. (2026) https://BioRender.com/5foo7gc. c, Confocal images of D15 HGPS-NMOs in mock and GFP mRNA-LNP groups (left), and A-to-G conversion by NGS after FATE mRNA-LNP delivery (right). Mean values ± SD; n = 5 independent organoids per group. Two-tailed unpaired t test with Welch’s correction (t(4) = 11.13, P = 0.0004; 95% CI, 1.218 to 2.026). d FATE mRNA-LNP microinjection into D25 HGPS-NMOs and bright-field image. Created in BioRender. Kwon, E. (2026) https://BioRender.com/8kuoky4. e Confocal images of D25 HGPS-NMOs in mock and GFP mRNA-LNP groups after microinjection (left), and A-to-G conversion by NGS 10 days later (right). Low, 285.29 μg/ml; Medium, 570.58 μg/ml; High, 1587.65 μg/ml. Mean values ± SD; n = 5 (low), 6 (medium), and 6 (high) independent organoids. Welch’s ANOVA (W(2.0, 8.84) = 9.17, P = 0.0070) with Dunnett’s T3 two-sided test. Adjusted P: low vs medium, 0.0545 (95% CI, −2.28 to 0.02); low vs high, 0.0205 ( − 5.96 to −0.63); medium vs high, 0.1103 ( − 4.86 to 0.53); ns, not significant. f Progerin, γH2AX, and 53BP1 localization in D25 HGPS-NMOs after microinjection (edited vs non-edited), observed at D70. Created in BioRender. Kwon, E. (2026) https://BioRender.com/fhiilbp. Immunofluorescence of HGPS-NMOs microinjected with FATE mRNA-LNPs at D25, imaged at D70, stained with Progerin (green), Lamin A (gray), DAPI (blue), and DNA damage markers (magenta) 1 h after 5 Gy irradiation: g γH2AX; h 53BP1. Scale bar, 25 μm. Representative of three independent experiments; three organoids per group. Created in BioRender. Kwon, E. (2026) https://BioRender.com/srcmndt.

Discussion

Patient-derived fibroblasts have established key cellular hallmarks of HGPS, including nuclear envelope abnormalities, persistent DNA damage, and epigenetic defects^55–59. However, fibroblast-based models provide limited insight into the lineage-selective pathology of HGPS, particularly the differential vulnerability of mesodermal and neural tissues. Human iPSC-derived lineage models have partially addressed this limitation^22,23,60, but generally examine individual cell types in isolation.

Here, we established two complementary isogenic hPSC-derived NMO systems that enabled direct comparison of mesoderm-derived muscle and ectoderm-derived neural lineages within a shared genetic background (Fig. 1). Progerin accumulation (Fig. 2), sequestration of 53BP1 and ATM at the nuclear membrane, defective formation of γH2AX, pATM, and 53BP1 foci, and heterochromatin disruption were selectively observed in the muscle compartment (Fig. 3). These findings recapitulate the lineage-selective vulnerability of HGPS and demonstrate the utility of NMOs for examining progerin-dependent pathology and therapeutic rescue. The relative preservation of the neural compartment is also consistent with miR-9-mediated post-transcriptional restriction of lamin A/progerin in neural cells⁶⁰.

Using our NMO model, we demonstrated that FATE disrupts the CSIM farnesylation motif of LMNA, releases progerin and DNA repair factors from the nuclear membrane, restores DNA damage foci, and recovers heterochromatin architecture (Fig. 5). Unlike strategies that correct the canonical c.1824C>T mutation^17,50, FATE targets a shared pathogenic post-translational modification and may therefore be applicable to non-classical LMNA variants and ZMPSTE24-deficient laminopathies that produce persistently farnesylated prelamin A. This mutation-agnostic potential was further supported in c.1968+5 G > A aHGPS-NMOs, in which FATE restored 53BP1 localization and IR-induced DNA damage foci formation (Fig. 5m, n). FATE also differs from lonafarnib by selectively modifying LMNA rather than broadly inhibiting protein farnesylation. Whole-exome and targeted sequencing detected no relevant editing in other genes encoding CAAX-containing proteins, although one intergenic candidate site showed measurable editing and will require further assessment (Fig. 4 and S4). Thus, FATE provides greater molecular specificity than systemic farnesyltransferase inhibition but will still require comprehensive off-target and in vivo safety evaluation. Complementing our NMO model of progerin-associated muscle pathology, a recently established blood vessel organoid model derived from HGPS patient iPSCs recapitulated vascular pathology and provides a promising platform for evaluating FATE in vascular tissues⁶¹.

Importantly, non-farnesylated progerin is not necessarily inert. Mice expressing non-farnesylated –SSIM progerin still developed progeria-associated phenotypes, although these were milder than those caused by farnesylated progerin and were accompanied by lower steady-state progerin levels⁶². In contrast, mice expressing a non-farnesylated –CSM progerin variant were largely free of progeria-like disease⁶³. These divergent outcomes indicate that the biological consequences of abolishing farnesylation depend strongly on the precise C-terminal sequence and protein context. The –RSIM species generated by FATE is distinct from both models, and its long-term safety, particularly in cardiac tissues, will require dedicated in vivo validation. Long-term assessment of neural-lineage integrity will likewise be necessary, although the current data do not indicate a coherent FATE-induced alteration of neural identity or function.

Finally, transient FATE mRNA–LNP delivery achieved partial editing but was sufficient to rescue disease-associated phenotypes in individual edited cells. The non-integrating nature and relatively low immunogenicity of mRNA–LNP delivery may permit repeated dosing, while advances in tissue-selective and muscle-directed LNP formulations offer potential routes to improve editing efficiency^64–67. In particular, tuning lipid composition or incorporating targeting ligands may enhance delivery to disease-relevant cell types and increase the proportion of edited cells. Future studies in HGPS animal models should therefore assess tissue-specific delivery, repeat-dosing efficacy, durability of correction, the capacity to reverse established progerin-associated pathology, and the long-term safety of the –RSIM lamin A/progerin species.

Methods

Ethics

All experiments involving human cells were conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board (IRB) of Seoul National University (SNU IRB protocol #2305/003-014). Human fibroblasts derived from patients with Hutchinson–Gilford progeria syndrome, including the non-classical c.1968+5 G > A mutation, were obtained through The Progeria Research Foundation. Written informed consent was obtained from all donors or their legal guardians prior to sample collection. Human fibroblasts were reprogrammed into induced pluripotent stem cells (iPSCs) following approved institutional protocols.

Sex of the human cell lines used in this study: the HGPS patient-derived iPSC line (Coriell AG28340) is from a male donor (46,XY); the human embryonic stem cell line WA09/H9 (WiCell) is female (46,XX); and the atypical HGPS patient-derived fibroblast line (Progeria Research Foundation, PSADFN576) is from a female donor. Sex was not considered in the study design, as the experimental focus was on isogenic comparison of LMNA genotypes within each cell line rather than on sex-based differences. Written informed consent for participation and for publication was obtained from all donors or their legal guardians prior to sample collection.

Culture of human pluripotent stem cells

Two human pluripotent stem cell lines were used in this study: HGPS patient-derived iPSCs (the following cell line was obtained from the NIGMS Human Genetic Cell Repository at the Coriell Institute for Medical Research: AG28340; referred to as HGPS-iPSCs) and the human embryonic stem cell line WA09 (H9; obtained from WiCell Research Institute, Madison, WI, USA; referred to as WT-hESCs). To generate isogenic controls, the c.1824C>T mutation in HGPS-iPSCs was corrected using the PE5stem prime editing system as described²⁶, resulting in mutation-corrected Edit-iPSCs. Conversely, the same c.1824C>T mutation was introduced into WT-hESCs using the same system to generate Mutant-hESCs. Two sets of human pluripotent stem cells (hPSCs) were maintained in mTeSR™1 medium (STEMCELL Technologies) on Matrigel-coated plates (Corning, 354277) at 37 °C in a humidified incubator with 5% CO₂⁶⁸. All hPSC lines used in this study were tested and confirmed negative for mycoplasma contamination. Cells were passaged approximately every four days using Versene solution (Thermo Fisher Scientific).

Generation of atypical HGPS iPSCs

Atypical HGPS patient dermal fibroblasts (PSADFN576, LMNA c.1968+5 G > A, heterozygous) were obtained from the Progeria Research Foundation Cell and Tissue Bank and maintained in DMEM (high glucose, without L-glutamine) supplemented with 15% fetal bovine serum, 1% penicillin–streptomycin, and 1% GlutaMAX at 37 °C under 5% CO₂. Reprogramming was performed using the CytoTune-2.0 Sendai Reprogramming Kit (Thermo Fisher Scientific), which delivers the reprogramming factors OCT4, SOX2, KLF4, and c-MYC through three non-integrating Sendai virus vectors (hKOS, hc-Myc, and hKlf4). Fibroblasts were transduced at the manufacturer-recommended multiplicity of infection (hKOS, 5; hc-Myc, 5; hKlf4, 3) according to the manufacturer’s instructions. Seven days after transduction, the cells were replated onto Matrigel-coated dishes (Corning, cat. no. 354277), and the medium was gradually transitioned to mTeSR1 (STEMCELL Technologies). Emerging iPSC colonies were manually picked between day 21 and day 28, clonally expanded, and maintained on Matrigel (Corning, cat. no. 354277) in mTeSR1. The heterozygous LMNA c.1968+5 G > A mutation was confirmed in the established aHGPS iPSCs by next-generation sequencing.

FATE plasmid construction

To generate FATE-iPSCs, the px330-mCherry plasmid (Addgene #98750, a gift from Dr. Jinsong Li) was modified by replacing the Cas9 coding sequence with the SpRY-ABE8e variant from pCMV-SpRY-ABE8e (Addgene #185671, a gift from Dr. David Liu) using Gibson assembly. Following this, a guide RNA sequence (5′-TGCTGCAGTTCTGGGGGCTC-3′) was inserted into the BbsI restriction site of the modified vector.

Generation of neuromuscular organoids (NMOs) from hPSCs via induction of human neuromesodermal progenitors

Human pluripotent stem cells (hPSCs) were dissociated into single cells using Accutase (STEMCELL Technologies) when cultures reached approximately 70–80% confluency. Cell counts were performed using the Countess III automated cell counter (Thermo Fisher), with the cell size threshold set to include only particles with a diameter ≥9 μm. Based on growth characteristics of each cell line, 1.3–1.5 × 10⁶ cells were seeded per well for WT-hESCs and Mutant-hESCs, whereas 1.0–1.2 × 10⁶ cells were used for HGPS-iPSCs and Edit-iPSCs. All cells were plated onto Matrigel-coated (Corning, 354277) 6-well plates. On day 0, to reduce stress from cell dissociation, 10 μM Y-27632 (a ROCK inhibitor) was added to the NMO basal medium, which also included 3 μM CHIR99021 and 40 ng/ml basic fibroblast growth factor (bFGF). The NMO basal medium was composed of a 1:1 blend of Advanced DMEM/F12 (supplemented with 1× N2) and Neurobasal medium (supplemented with 1× B27, 2 mM L-glutamine, 75 μg/ml BSA Fraction V, and 0.1 mM 2-mercaptoethanol). On day 1, fresh medium of the same composition—except without Y-27632—was applied, and daily medium changes were performed until day 3. This 2D induction protocol facilitated the differentiation of hPSCs into neuromesodermal progenitors (NMPs). On day 3, NMPs were enzymatically dissociated into single cells using Accutase and seeded at a density of 8000 cells per well into ultra-low attachment 96-well plates. Cell counts were obtained using a Countess III automated cell counter, and aggregation was initiated by centrifugation at 350 × g for 2 min. For organoid formation (day 0 of 3D culture), 100 μl of NMO basal medium supplemented with 50 μM Y-27632, 10 ng/ml bFGF, 2 ng/ml IGF1, and 2 ng/ml HGF was added to each well. On day 2, 50 μl of medium was carefully removed and replaced with 100 μl of fresh NMO basal medium containing 2 ng/ml IGF1 and 2 ng/ml HGF. On day 4, the medium was completely replaced with NMO basal medium without additional growth factors and subsequently refreshed every other day. On day 10, organoids were transferred into 60-mm ultra-low attachment dishes containing 5 ml of NMO basal medium, and on day 30, they were moved to 100-mm Petri dishes with 12 ml of the same medium. Medium changes continued every two days. To minimize mechanical stress, pipette tips were trimmed using autoclaved scissors during transfers. From day 10 onward, organoids were maintained on an orbital shaker set to 75 rpm to facilitate maturation. The experimental use of each hPSC-derived NMO model across immunoblotting, immunofluorescence, TEM, and scRNA-seq is summarized in Supplementary Table 1.

Immunohistochemistry of organoids

Organoids were collected at specific time points, rinsed once in DPBS, and fixed in 4% paraformaldehyde (PFA) at 4 °C for 1–2 h depending on the duration of culture. Post-fixation, the samples were washed three times in DPBS and placed in 30% sucrose at 4 °C overnight, allowing them to sink completely as an indication of full infiltration. For embedding, organoids were transferred to molds filled with Tissue-Tek O.C.T. compound. Cryopreservation was achieved by partially solidifying isopentane (2-methylbutane) in a metal container surrounded by liquid nitrogen in a styrofoam box. Once the isopentane appeared opaque, the molds were immediately submerged to freeze the samples. The frozen blocks were stored at –80 °C prior to cryosectioning. Tissue sections of 10 μm were sliced using a Leica CM1520 cryostat maintained at –19 °C and mounted onto glass slides. Before immunostaining, slides were equilibrated to room temperature and treated with 0.3% Triton X-100 in DPBS to eliminate any remaining O.C.T. compound. Non-specific binding was blocked for 1 h using either 5% normal goat serum or 4% BSA in DPBS. Sections were then incubated overnight (12–16 h) at 4 °C with primary antibodies diluted in the blocking buffer. The following day, sections were washed three times and treated with fluorescently labeled secondary antibodies for 1 to 1.5 h at room temperature. Nuclei were stained with DAPI (1 μg/ml) for 10 min and rinsed again. Final mounting was done using ProLong™ Gold Antifade Reagent. Confocal images were acquired using a Leica SP8 and a Zeiss LSM980 confocal microscope, and rotational movies (360° along the Y-axis at 250 fps) were generated using ZEN Blue software. Images were processed using LAS X (version 3.11.0) and ZEN lite (version 3.13). Antibodies used for immunostaining were listed in Supplementary Table 2.

Immunoblotting analysis

Organoids were homogenized in RIPA buffer (Biosesang) supplemented with 1 μM protease inhibitor and 10 μM sodium orthovanadate using a TissueLyser II (QIAGEN) set to a frequency of 30 Hz for 3 min. The homogenates were extracted on ice for 2 h with intermittent vortexing every 10 min. Lysates were then centrifuged at 13,000 rpm for 20 min at 4 °C, and the resulting supernatants were collected as total protein lysates. Protein concentrations were determined using the Pierce™ BCA Protein Assay Kit (#23225, Thermo Fisher Scientific). Equal amounts of protein were diluted with Bolt™ LDS Sample Buffer (4X) and Bolt™ Sample Reducing Agent (10X) to achieve uniform concentrations and then boiled at 95 °C for 10 min. Following protein extraction, immunoblotting was carried out using the following protocol. Approximately 20–30 μg of protein per sample was loaded onto 3–8% NuPAGE™ Tris-Acetate Mini Protein Gels (Invitrogen) and separated by electrophoresis at 150 V. The resolved proteins were transferred onto PVDF membranes using iBlot™ 3 Transfer Stacks (MIDI, PVDF format) via dry transfer. Membranes were blocked in 5% bovine serum albumin (BSA) in TBS-T for 1–2 h at room temperature and then incubated overnight at 4 °C with primary antibodies diluted in the same blocking solution. After three washes with TBS-T, membranes were incubated with HRP-conjugated secondary antibodies (1:10,000 dilution in TBS-T) for 1–1.5 h at room temperature. Signal detection was performed using either Miracle-Star (#16028, iNtRON Biotechnology) or Amersham™ ECL Prime (#GERPN2232, Cytiva) chemiluminescence reagents. Antibodies used for immunostaining were listed in Supplementary Table 2. Uncropped and unprocessed scans of all blots are provided in the Source Data file.

Real-time PCR analysis

Easy-BLUETM RNA isolation kit (iNtRON Biotechnology) was used for total RNA extraction. For reverse transcription, 5× PrimeScriptTM RT mix (TaKaRa) was used to generate cDNA. Quantitative real-time PCR was performed with SYBR Green PCR reagents (Life Technologies) using the QuantStudio3 (Applied Biosystems). Primer sequences are provided in Supplementary Table 4.

Fractionation of organoids and immunoblotting analysis

Nuclear membrane fractionation was performed based on the protocol described⁶⁹. Briefly, 4–6 organoids were dissociated into single cells using Liberase™ TL Research Grade (Roche) following the manufacturer’s instructions. The resulting cell suspension was passed through a Falcon® 40 μm strainer to remove aggregates and debris. Cells were pelleted by centrifugation at 300 × g for 3 min, washed once with 1 ml pre-chilled PBS, and pelleted again under the same conditions. After gently removing the supernatant, the cell pellet was resuspended in lysis buffer (at a 1:5 to 1:8 v/v ratio), followed by pipetting and brief vortexing. The suspension was incubated on ice for 5 min to allow swelling. Cells were then lysed mechanically using a 26 G 1/2” KOVAX-SYRINGE. The lysate was carefully underlaid with 200 μl of 20% sucrose in 8% PVP buffer and centrifuged at 3500–4000 × g for 10 min at 4 °C. The supernatant, including the sucrose phase, was collected as the cytoplasmic fraction. The pelleted nuclei were resuspended in DNase buffer (1 ml per 1 × 10⁷ cells), vortexed, and incubated at room temperature for 30 min. The suspension was then underlaid with 100 μl of 30% sucrose prepared in 20 mM HEPES (pH 8), 0.1 mM MgCl₂, and 1:100 Solution P. Nuclear envelopes were pelleted by centrifugation at 18,000 × g for 10 min at 4 °C. The resulting supernatant was collected as the nucleoplasmic fraction, while the pellet represented the nuclear membrane fraction. For downstream SDS-PAGE and Western blot analysis, each fraction was processed by methanol precipitation. Specifically, 900 μl of methanol was added to 100 μl of sample, followed by vortexing and overnight incubation at –20 °C. Samples were centrifuged at 20,000 × g for 20 min at 4 °C, and pellets were briefly sonicated (2 s) in 500 μl of methanol, incubated again at –20 °C for 20 min, and re-pelleted. The final pellet was resuspended in a 1× dilution of Bolt™ LDS Sample Buffer (4X) and Bolt™ Sample Reducing Agent (10X) for protein denaturation.

8% PVP buffer

This buffer was composed of 8% polyvinylpyrrolidone (PVP-40), 20 mM potassium phosphate, and 7.5 μM magnesium chloride (MgCl₂). The pH was adjusted to 6.53 using concentrated phosphoric acid (H₃PO₄).

Lysis buffer

The lysis solution contained 6% PVP-40, 0.015% Digitonin, 0.015% Triton X-100, and 4 μM Cytochalasin B. Protease inhibitors were added at a final dilution of 1:100 from Solution P. Dithiothreitol (DTT) was included at 1 mM final concentration

DNase buffer

DNase treatment was performed using a buffer consisting of 10% sucrose in 20 mM HEPES (pH 8.0), supplemented with 0.1 mM MgCl₂, 10 μM CaCl₂, 1 mM DTT, 100 μg/mL heparin, and 0.1 μg/mL RNase A. DNase I (Roche) was added at 1:1,000 dilution from a 5 mg/mL stock prepared in 50% glycerol containing 10 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM DTT, and 2 mM MgCl₂. Protease inhibitors were also added at 1:100 dilution from Solution P.

Electron microscopy of organoids

Organoids were fixed overnight at 4 °C in Karnovsky’s fixative, consisting of 2% (w/v) paraformaldehyde and 2% (v/v) glutaraldehyde in 0.05 M sodium cacodylate buffer(pH 7.4). Following fixation, samples were washed three times for 5 min each at room temperature in 0.05 M sodium cacodylate buffer. Post-fixation was performed using 1% osmium tetroxide diluted in 0.1 M sodium cacodylate buffer. The samples were then rinsed three times with distilled water at room temperature, each for 5 min, and stained en bloc overnight at 4 °C in 0.5% uranyl acetate diluted in distilled water. After En Bloc staining, samples were again washed three times with distilled water for 5 min each and dehydrated through a graded ethanol series (30%, 50%, 70%, 80%, 90%, and 100%). Dehydrated samples were treated twice with 100% propylene oxide at room temperature for 15 min per wash. Resin infiltration was performed sequentially using mixtures of propylene oxide and Spurr’s resin at 1:1 and 1:2 ratios, each for 1.5 h at room temperature, followed by transition into 100% Spurr’s resin overnight. Finally, samples were embedded in fresh 100% Spurr’s resin in embedding molds and polymerized at 70 °C overnight. Ultrathin sections (~70 nm) were stained with 2% uranyl acetate followed by 3% lead citrate. Imaging was performed using a JEM-1010 transmission electron microscope (JEOL Ltd., Tokyo, Japan) operated at 80 kV. Images were acquired using a CCD camera equipped with an image capture system and processed with Radius 2.1 software (EMSIS GmbH, Münster, Germany).

Single cell preparation from organoids for scRNAseq

For single-cell RNA sequencing (scRNA-seq), 6 of day 70 organoids (HGPS-NMOs, Edit-NMOs, and FATE-NMOs) were washed with DPBS and transferred to 1.5 mL microcentrifuge tubes. Organoids were then incubated with 1 mL of Liberase™ TL Research Grade (Roche) at 37 °C for 30 min, following the manufacturer’s protocol. After enzymatic digestion, mechanical dissociation was performed by pipetting until a single-cell suspension was achieved. The suspension was passed through a 40 μm cell strainer (Falcon®) to eliminate cell aggregates and debris. The resulting cell suspension was washed with DPBS and resuspended in HypoThermosol® FRS (STEMCELL Technologies), followed by quality assessment using the Celinius single-cell QC platform.

Sample preparation and quality control

Single-cell RNA sequencing (scRNA-seq) was performed using the 10x Genomics Chromium platform. High-quality Dissociated neuromuscular organoid cells were prepared to ensure optimal assay performance. All suspensions were washed and resuspended in 1 x PBS containing 0.04% BSA buffer, using wide-bore pipette tips to prevent shearing. The quality of the final suspension was assessed by evaluating cell concentration, viability, and the degree of aggregation to ensure suitability for microfluidic partitioning.

scRNA-seq library preparation and sequencing

Single cells were partitioned into Gel Beads-in-emulsion (GEMs) using a Chromium controller Within each GEM, mRNA was captured and barcoded. scRNA-seq libraries were subsequently generated following the manufacturer’s protocol (10x genomics, CG000315), which includes reverse transcription, cDNA amplification, and library construction Chromium Next GEM Single Cell 3’ Kit v3.1 (10x Genomics, PN1000268) The quality and size distribution of the resulting cDNA and final libraries were assessed, and concentrations were quantified. Libraries were sequenced on a compatible platform using NovaSeq 6000 platform(Illumina) and a paired-end configuration, with read lengths structured to capture the cell barcode, Unique Molecular Identifier (UMI), and the transcript sequence.

Primary data processing

Raw sequencing base call (BCL) files were demultiplexed into FASTQ files using the cellranger mkfastq pipeline. The resulting FASTQ files were processed with cellranger (10x Genomics, v-8.0.1)count to align reads to a GRCh38 reference genome and generate a feature-barcode count matrix. This matrix, containing UMI counts per gene for each cell, served as the input for all subsequent analyses. All processing was performed using a consistent version of the Cell Ranger software to ensure reproducibility.

Single-cell RNA-seq data analysis

Single-cell RNA-seq datasets from HGPS, EDIT, and FATE samples were processed using Seurat v5.4.0. Cells with fewer than 200 detected genes or with more than 15% mitochondrial transcripts were removed. Normalization and variance stabilization were performed using an SCTransform-based approach, and the datasets were integrated after selecting 3000 highly variable features. Batch effects were corrected during the integration step. The integrated data were dimensionally reduced with principal component analysis (PCA), and the first 15 principal components were used for constructing a shared nearest neighbor graph and subsequent clustering. Low-dimensional visualization was performed with UMAP to explore cellular heterogeneity. Cell clusters were annotated into five major lineages (skeletal muscle, neuron, sclerotome, endothelial, and epithelial) based on canonical marker gene expression and module score analysis of curated gene sets. For cross-sample comparison, reference-based label transfer was applied to map the FATE dataset onto the integrated HGPS reference, and the query cells were projected onto the existing UMAP embedding.

Differential expression and pathway enrichment analysis

For lineage-specific comparisons, skeletal muscle and neuronal clusters were separately subsetted from the integrated HGPS and EDIT dataset to generate SM (skeletal muscle) and NR (neuron) objects. SCTransform normalization was applied independently to each subset. Differentially expressed genes (DEGs) between HGPS and EDIT were identified using a log2 fold change threshold > 1 and adjusted p-value < 0.05. Over-representation analysis of DEGs was conducted using EnrichR with the WikiPathways_2024_Human database to identify lineage-specific functional pathways. Additionally, gene set enrichment analysis (GSEA) was performed using the fgsea package with cell type signatures from the MSigDB C5:HPO collection (via msigdbr). Ranked gene lists based on log2 fold change were tested for enrichment, and significant pathways were defined by normalized enrichment scores (NES) with adjusted p values < 0.05. Directionality of enrichment was interpreted based on the sign of NES, with positive and negative values indicating enrichment in HGPS and EDIT, respectively. Enrichment results were visualized by plotting NES against –log10 adjusted p-values.

Pathway activity estimation and PCA analysis based on pseudobulk GSVA

To assess pathway activity at the group level, pseudobulk expression matrices were generated by averaging SCTransform-normalized expression values per condition and cell type (skeletal muscle and neuron) from both the integrated (HGPS and EDIT) and mapped (FATE) datasets. Gene Set Variation Analysis (GSVA) was conducted using the GSVA R package with curated gene sets from the KEGG Legacy and Human Phenotype Ontology collections, with emphasis on DNA damage repair pathways. GSVA scores representing relative pathway activity were computed for each pseudobulk sample. These scores were used for downstream comparative analysis, including hierarchical clustering and principal component analysis (PCA). PCA was performed on the GSVA enrichment matrix derived from KEGG Legacy pathways to evaluate global pathway activity patterns across HGPS, EDIT, and FATE samples.

Derivation and application of young muscle signature

To derive a transcriptional signature representing youthful skeletal muscle, we utilized publicly available human muscle single-cell RNA-seq data from Veronika R. K. et al.⁴⁸. Donor samples were stratified into young (ages 15–40) and old (ages 50–60 and 50–75) cohorts. Differential expression analysis was performed between the two groups, and genes with adjusted p value < 0.05 and average log2 fold change >1 were selected as significantly upregulated in the young group. These genes were defined as the young muscle signature.

To evaluate this signature in experimental samples, pseudobulk profiles from myogenic progenitor cells in the HGPS, EDIT, and FATE datasets were generated. GSVA was then applied to quantify the enrichment of the young muscle signature, enabling assessment of youthful transcriptional programs across conditions.

Whole exome sequencing (WES) and data processing

Whole exome capture libraries were prepared using the Agilent SureSelect Target Enrichment protocol for Illumina paired-end sequencing libraries (Version C2, December 2018) with 0.5 μg of input genomic DNA. The SureSelect Human All Exon V8 + UTR probe set was used in all cases. DNA quality and quantity were assessed with PicoGreen (Invitrogen) and Agilent TapeStation gDNA screentape. One microgram of genomic DNA was diluted in EB buffer and sheared to a target peak size of 150–200 bp using a Covaris LE220 focused ultrasonicator (Covaris, Woburn, MA) according to the manufacturer’s recommendations. Fragmented DNA underwent end repair, A-tailing, and adapter ligation, followed by PCR amplification. For exome capture, 250 ng of DNA library was hybridized with the capture probes at 65 °C for 24 h using the heated lid option (105 °C) on a thermal cycler, washed, and PCR amplified. Final libraries were quantified using qPCR (KAPA Library Quantification Kit for Illumina platforms) and qualified on an Agilent TapeStation D1000 screentape. Sequencing was performed on the Illumina NovaSeqX platform (Illumina, San Diego, USA) with a target coverage depth of approximately ×400, yielding an average achieved coverage of ~250×. Raw sequencing reads in FASTQ format were subjected to initial quality control to assess read quality, adapter content, and duplication rates. Reads were aligned to the human reference genome (GRCh38) using BWA-MEM, and PCR duplicates were removed with Picard MarkDuplicates. Base quality score recalibration (BQSR) was performed using GATK BaseRecalibrator, and the resulting BAM files were processed following GATK Best Practices. Variant calling was carried out using GATK HaplotypeCaller in GVCF mode, followed by joint genotyping with GenotypeGVCFs. Variants were filtered based on quality metrics with GATK VariantFiltration and functionally annotated using SnpEff and dbNSFP. Final results were compiled into both VCF and tabular formats for downstream analyses.

Targeted deep sequencing and data processing

Genomic DNA from induced pluripotent stem cells (iPSCs) and nuclear membrane organoids (NMOs) was extracted using the Wizard Genomic DNA Purification Kit (#A1120, Promega) to evaluate target editing efficiency. The target locus was amplified using Solg™ Pfu-X DNA Polymerase (#SPX16-R25h, Solgent) to generate sequencing libraries. Libraries were sequenced on an Illumina MiSeq platform (Binonics, South Korea) with 150-bp paired-end reads at a depth of ~30,000 reads. FASTQ files were aligned to the human reference genome (hg38) using BWA-MEM (BWA v0.7.19-r1273), and the resulting alignments were sorted and indexed using SAMtools (v1.23) for downstream analysis. Primer sequences used for targeted deep sequencing are provided in Supplementary Table 3.

Identification of gRNA-dependent off-target Effects

To identify potential off-target effects associated with the designed guide RNA (gRNA), we computationally predicted candidate off-target loci with high sequence similarity (0–2 mismatches) using Cas-OFFinder implemented in RGEN Tools. A total of 37 candidate sites were identified. Among these, 19 sites were located within protein-coding regions, including 5 exonic sites, while the remaining candidates were distributed across intronic and intergenic regions. To assess editing events at these predicted loci, we applied genomic-context–appropriate sequencing strategies. Exonic candidate sites were analyzed using high-depth whole-exome sequencing (WES), whereas intronic and intergenic sites were evaluated by targeted deep sequencing. In both cases, aligned BAM files were processed to extract base-level read counts at each predicted off-target position. Editing frequencies were calculated by dividing the number of reads supporting the expected base substitution by the total read depth at that locus. To ensure conservative detection of off-target events, only substitutions exceeding the empirically estimated sequencing error rate (~1%) were considered indicative of bona fide off-target editing. This analysis revealed no detectable off-target editing above background levels at any of the predicted sites, supporting the high specificity of the gRNA used in the FATE strategy.

Identification of gRNA-independent off-target effects

To identify sequence alterations not attributable to gRNA targeting, we analyzed VCF files derived from WES data. Positions with a read depth >200 were retained for further inspection. Sites showing conversion in both Mock and FATE samples were excluded to focus on treatment-specific effects. As a result, six candidate sites, including one in LMNA, were uniquely observed in FATE-treated samples, while nine sites were identified only in Mock-treated samples.

Analysis of CAAX motif sequences

To examine the conservation of the CAAX motif, we extracted 12-nucleotide sequences corresponding to CAAX regions from genes known to harbor this motif, including LMNA, small GTPases, lamins, and the CENP gene family. Genomic coordinates of CAAX motifs were used to retrieve base-level read data from WES BAM files. Nucleotide frequencies at each position were aggregated across all selected motifs, and a sequence logo was generated to visualize base preferences and positional conservation within the motif.

mRNA synthesis via In vitro transcription (IVT)

Linear DNA templates were generated from the plasmid pCMV-SpRY-ABE8e (Addgene #185671) by PCR with PrimeSTAR® HS DNA Polymerase (Takara) following the manufacturer’s protocol. A 120-nt poly(A) tail was encoded in the template using a poly(T)-containing primer during PCR. The PCR products were purified using AccuPrep® PCR/Gel Purification Kit (Bioneer) and verified by agarose gel electrophoresis (1% agarose in 1× TBE buffer). In vitro transcription was performed using MEGAscript™ T7 Transcription Kit (Thermo Fisher Scientific), with CleanCap® AG (TriLink) and substituting UTP with N1-methyl-pseudouridine-5′-triphosphate (N1-Me-ΨUTP). Transcription reactions were carried out at 37 °C for 2 h, followed by DNase treatment to digest any residual template DNA. The resulting mRNA was purified using HiGene™ DNA/RNA Purification Mini Column (Biofact). Transcript integrity was confirmed by denaturing agarose gel electrophoresis (1% agarose in 1× MOPS buffer).

Preparation and characterization of LNPs

LNPs were prepared using a microfluidic mixing system (Ignite, Precision NanoSystems). A lipid mixture containing an ionizable lipid (undisclosed structure), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), cholesterol, and PEG-lipid were dissolved in ethanol at the specified molar ratio. The ethanol phase (97.5% v/v) was mixed with 10 mM citrate buffer, pH 3.0 (2.5% v/v). ABE mRNA and sgRNA were diluted in an aqueous phase (1× PBS and 10 mM citrate buffer, v/v = 2:1). LNPs were formulated by microfluidic mixing at a total flow rate of 12 mL/min with an ethanol-to-aqueous phase ratio of 1:3. The resulting LNPs were diluted 40-fold in 1× PBS and concentrated using Amicon® Ultra-15 centrifugal filters, 10 kDa MWCO (Millipore). Particle size and polydispersity index (PDI) of LNPs were measured by dynamic light scattering (DLS) using a Zetasizer Lab Blue (Malvern Panalytical). RNA encapsulation efficiency was quantified with Quant-iT™ RiboGreen® RNA Assay Kit (Thermo Fisher Scientific) as described below. Relative fluorescence (%)=(F_triton / F_control)X100.

mRNA-LNP microinjection

For FATE mRNA–LNP delivery into NMOs, organoids were transferred into a 100 μL basal medium drop using a micropipette and moved on the micromanipulator stage. A microinjector (CellTram, Eppendorf) connected to a non-heparinized microhematocrit capillary tube (DWK Life Sciences) was used as a holding pipette to hold the neural region of the organoid in place. A second microinjector (FemtoJet 4i, Eppendorf) was then used to perform three injections at a 90° angle into the muscle region under the following conditions: injection pressure of 150–250 hPa (adjusted by confirmation of LNP entry into the organoid), and consistent pressure of 0 hPa. To prevent backflow of the injected LNP solution from organoids, the injection needle was kept inside the organoid for 3 s before withdrawal. Following injection, organoids were transferred to NMO basal medium using a 1-ml pipette tip with the end trimmed and subsequently cultured for 10 days in 60-mm ultra-low attachment plates prior to NGS analysis.

Calculation of editing efficiency via mRNA delivery

To evaluate genome editing outcomes following mRNA delivery, Targeted deep sequencing data were analyzed as described above. Base composition at the editing site was quantified from the aligned reads, and the editing frequency was calculated using the same threshold-based approach. This allowed direct estimation of editing efficiency at the nucleotide level.

Statistics and reproducibility

Statistical analyses were performed using GraphPad Prism 10. Quantitative data are presented as mean ± standard deviation (SD), and the n reported in each figure legend denotes the number of independent organoids or samples analyzed. Comparisons between two groups were performed using an unpaired two-tailed t-test, with Welch’s correction applied when the groups had unequal variances. Comparisons among three or more groups were performed using ordinary one-way ANOVA with Tukey’s multiple comparisons test when variances were homogeneous, or Welch’s ANOVA with Dunnett’s T3 multiple comparisons test when variances were unequal. Differences in gene expression between single-cell populations were assessed using a two-sided Wilcoxon rank-sum test. Statistical significance is denoted as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, and non-significant comparisons are labeled ns (P ≥ 0.05). The specific test applied to each experiment is indicated in the corresponding figure legend. Statistical procedures for single-cell RNA sequencing, differential gene expression, and pathway enrichment analyses are described in their respective Methods sections.

Reporting summary

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

Supplementary information

Supplementary Information^{(10MB, pdf)}

41467_2026_77251_MOESM2_ESM.pdf^{(5.6KB, pdf)}

Description of Additional Supplementary Files

Supplementary Movie 1^{(24.8MB, mp4)}

Supplementary Movie 2^{(24.7MB, mp4)}

Supplementary Movie 3^{(24.6MB, mp4)}

Supplementary Movie 4^{(24.8MB, mp4)}

Supplementary Movie 5^{(24.6MB, mp4)}

Supplementary Movie 6^{(24.8MB, mp4)}

Reporting Summary^{(3MB, pdf)}

Transparent Peer Review file^{(13.2MB, pdf)}

Source data

Source Data^{(27.2MB, zip)}

Acknowledgements

Fibroblasts derived from the patient harboring the non-classical c.1968+5 G > A HGPS mutation were obtained through The Progeria Research Foundation (https://www.progeriaresearch.org).

Author contributions

H.J.C. conceived the overall study design, led the experiments, and wrote the manuscript. D.W.K. conducted experiments, engaged in critical discussions of the results, and wrote the initial draft of the manuscript. E.J.K. performed scRNAseq analysis and in silico analysis and wrote the initial draft of the manuscript. B.J.J., D.H.K., and G.J. conducted the gene delivery and sample preparation. H.L., Y.Y., and Y.H. provided mRNA and LNP. All authors contributed to writing and revising the manuscript and approved the final version.

Peer review

Peer review information

Nature Communications thanks Kan Cao, who co-reviewed with Huijing Xue; Susana Gonzalo and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

This research was supported by a grant of the Korea-US Collaborative Research Fund (KUCRF) (RS-2024-00466703), ABC-based Regenerative BioTherapeutics (ABC project) grant (RS-2024-00432867) and National Research Foundation of Korea (NRF) grant (RS-2025-00560943), funded by the Ministry of Science and ICT and Ministry of Health & Welfare, Republic of Korea.

Data availability

The single-cell RNA sequencing data generated in this study have been deposited in the Gene Expression Omnibus (GEO) database under accession code GSE305237. The whole-exome sequencing and targeted deep sequencing data generated in this study have been deposited in the NCBI BioProject database under accession code PRJNA1306877. Publicly available human skeletal muscle single-cell RNA sequencing data used to derive the young muscle signature are available in the ArrayExpress database under accession code E-MTAB-13874. Source data are provided with this paper.

Code availability

The custom R code used for the single-cell RNA-seq and base-editing specificity analyses is available at Zenodo. The specific version of the code associated with this publication is archived in ref. ⁷⁰ Zenodo and is accessible via [https://doi.org/10.5281/zenodo.21366828].

Competing interests

D.W.K. and H.J.C. have filed a patent application related to this work. The other authors declare no competing interests.

Footnotes

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

These authors contributed equally: Dong-Woo Kim, Eun-Ji Kwon.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77251-3.

References

1.Davies, B. S., Fong, L. G., Yang, S. H., Coffinier, C. & Young, S. G. The posttranslational processing of prelamin A and disease. Annu. Rev. Genomics Hum. Genet.10, 153–174 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Eriksson, M. et al. Recurrent de novo point mutations in lamin A cause Hutchinson–Gilford progeria syndrome. Nature423, 293–298 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Capell, B. C. et al. Inhibiting farnesylation of progerin prevents the characteristic nuclear blebbing of Hutchinson-Gilford progeria syndrome. Proc. Natl. Acad. Sci. USA 102, 12879–12884. 10.1073/pnas.0506001102 (2005). [DOI] [PMC free article] [PubMed]
4.Hamczyk, M. R., Del Campo, L. & Andrés, V. Aging in the cardiovascular system: Lessons from Hutchinson-Gilford progeria syndrome. Annu. Rev. Physiol.80, 27–48 (2018). [DOI] [PubMed] [Google Scholar]
5.Marcelot, A. & Worman, H. J. & Zinn-Justin, S. Protein structural and mechanistic basis of progeroid laminopathies. FEBS J.288, 2757–2772 (2021). [DOI] [PubMed] [Google Scholar]
6.Bergo, M. O. et al. Zmpste24 deficiency in mice causes spontaneous bone fractures, muscle weakness, and a prelamin A processing defect. Proc. Natl. Acad. Sci.99, 13049–13054 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Pendás, A. M. et al. Defective prelamin A processing and muscular and adipocyte alterations in Zmpste24 metalloproteinase–deficient mice. Nat. Genet.31, 94–99 (2002). [DOI] [PubMed] [Google Scholar]
8.Hennekam, R. C. Hutchinson–Gilford progeria syndrome: review of the phenotype. Am. J. Med. Genet. Part A140, 2603–2624 (2006). [DOI] [PubMed] [Google Scholar]
9.Merideth, M. A. et al. Phenotype and course of Hutchinson–Gilford progeria syndrome. N. Engl. J. Med.358, 592–604 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Liu, B. et al. Genomic instability in laminopathy-based premature aging. Nat. Med.11, 780–785 10.1038/nm1266 (2005). [DOI] [PubMed] [Google Scholar]
11.Chojnowski, A. et al. Heterochromatin loss as a determinant of progerin-induced DNA damage in Hutchinson–Gilford progeria. Aging Cell19, e13108 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Köhler, F. et al. Epigenetic deregulation of lamina-associated domains in Hutchinson-Gilford progeria syndrome. Genome Med.12, 1–16 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Gordon, L. B. et al. Clinical trial of a farnesyltransferase inhibitor in children with Hutchinson–Gilford progeria syndrome. Proc. Natl. Acad. Sci.109, 16666–16671 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Gordon, L. B. et al. Association of lonafarnib treatment vs no treatment with mortality rate in patients with Hutchinson-Gilford progeria syndrome. JAMA319, 1687–1695 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Suzuki, M. et al. FDA approval summary for lonafarnib (Zokinvy) for the treatment of Hutchinson-Gilford progeria syndrome and processing-deficient progeroid laminopathies. Genet. Med.25, 100335 (2023). [DOI] [PubMed] [Google Scholar]
16.Cao, K. et al. Rapamycin reverses cellular phenotypes and enhances mutant protein clearance in Hutchinson-Gilford progeria syndrome cells. Sci. Transl. Med.3, 89ra58–89ra58 (2011). [DOI] [PubMed] [Google Scholar]
17.Koblan, L. W. et al. In vivo base editing rescues Hutchinson–Gilford progeria syndrome in mice. Nature589, 608–614 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Puttaraju, M. et al. Systematic screening identifies therapeutic antisense oligonucleotides for Hutchinson–Gilford progeria syndrome. Nat. Med.27, 526–535 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Erdos, M. R. et al. A targeted antisense therapeutic approach for Hutchinson–Gilford progeria syndrome. Nat. Med.27, 536–545 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Beyret, E. et al. Single-dose CRISPR-Cas9 therapy extends lifespan of mice with Hutchinson-Gilford progeria syndrome. Nat. Med25, 419–422 10.1038/s41591-019-0343-4 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Hunt, J. M. T., Samson, C. A., Rand, A. D. & Sheppard, H. M. Unintended CRISPR-Cas9 editing outcomes: a review of the detection and prevalence of structural variants generated by gene-editing in human cells. Hum. Genet.142, 705–720 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Liu, G.-H. et al. Recapitulation of premature ageing with iPSCs from Hutchinson–Gilford progeria syndrome. Nature472, 221–225 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Zhang, J. et al. A human iPSC model of Hutchinson Gilford Progeria reveals vascular smooth muscle and mesenchymal stem cell defects. Cell Stem Cell8, 31–45 (2011). [DOI] [PubMed] [Google Scholar]
24.Musunuru, K. Genome editing of human pluripotent stem cells to generate human cellular disease models. Dis. Models Mech.6, 896–904 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Park, J.-C. et al. Gene editing with ‘pencil’rather than ‘scissors’ in human pluripotent stem cells. Stem Cell Res. Ther.14, 164 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Park, J.-C. et al. Enhancing genome editing in hPSCs through dual inhibition of DNA damage response and repair pathways. Nat. Commun.15, 4002 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Park, J.-C. et al. MutSα and MutSβ as size-dependent cellular determinants for prime editing in human embryonic stem cells. Mol. Ther. -Nucleic Acids32, 914–922 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Faustino Martins, J. M. et al. Self-organizing 3D Human Trunk Neuromuscular Organoids. Cell Stem Cell26, 172–186.e176 10.1016/j.stem.2019.12.007 (2020). [DOI] [PubMed] [Google Scholar]
29.Suarez-Calvet, X. et al. Decoding the transcriptome of Duchenne muscular dystrophy to the single nuclei level reveals clinical-genetic correlations. Cell Death Dis.14, 596. 10.1038/s41419-023-06103-5 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Gonzalo, S. & Kreienkamp, R. DNA repair defects and genome instability in Hutchinson-Gilford Progeria Syndrome. Curr. Opin. Cell Biol.34, 75–83 10.1016/j.ceb.2015.05.007 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Monnerat, G. et al. Metabolomic profiling suggests systemic signatures of premature aging induced by Hutchinson-Gilford progeria syndrome. Metabolomics15, 100. 10.1007/s11306-019-1558-6 (2019). [DOI] [PubMed] [Google Scholar]
32.Kreienkamp, R. et al. Doubled lifespan and patient-like pathologies in progeria mice fed high-fat diet. Aging Cell18, e12852 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Harhouri, K. et al. MG 132-induced progerin clearance is mediated by autophagy activation and splicing regulation. EMBO Mol. Med.9, 1294–1313 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Lee, J. M. et al. Modulation of LMNA splicing as a strategy to treat prelamin A diseases. J. Clin. Invest126, 1592–1602 10.1172/JCI85908 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Gibbs-Seymour, I., Markiewicz, E., Bekker-Jensen, S., Mailand, N. & Hutchison, C. J. Lamin A/C-dependent interaction with 53 BP 1 promotes cellular responses to DNA damage. Aging cell14, 162–169 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Panier, S. & Boulton, S. J. Double-strand break repair: 53BP1 comes into focus. Nat. Rev. Mol. Cell Biol.15, 7–18 (2014). [DOI] [PubMed] [Google Scholar]
37.Etourneaud, L. et al. Lamin B1 sequesters 53BP1 to control its recruitment to DNA damage. Sci. Adv.7, eabb3799 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Cobb, A. M. et al. Prelamin A impairs 53 BP 1 nuclear entry by mislocalizing NUP 153 and disrupting the Ran gradient. Aging cell15, 1039–1050 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Cha, H. et al. Wip1 directly dephosphorylates γ-H2AX and attenuates the DNA damage response. Cancer Res.70, 4112–4122 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Adikusuma, F. et al. Large deletions induced by Cas9 cleavage. Nature560, E8–E9 (2018). [DOI] [PubMed] [Google Scholar]
41.Shin, H. Y. et al. CRISPR/Cas9 targeting events cause complex deletions and insertions at 17 sites in the mouse genome. Nat. Commun.8, 15464 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Papathanasiou, S. et al. Whole chromosome loss and genomic instability in mouse embryos after CRISPR-Cas9 genome editing. Nat. Commun.12, 5855 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Walton, R. T., Christie, K. A., Whittaker, M. N. & Kleinstiver, B. P. Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants. Science368, 290–296 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Ren, Q. et al. PAM-less plant genome editing using a CRISPR–SpRY toolbox. Nat. Plants7, 25–33 (2021). [DOI] [PubMed] [Google Scholar]
45.Moiseeva, O., Lopes-Paciencia, S., Huot, G., Lessard, F. & Ferbeyre, G. Permanent farnesylation of lamin A mutants linked to progeria impairs its phosphorylation at serine 22 during interphase. Aging (Albany NY)8, 366 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Quigley, A. et al. The structural basis of ZMPSTE24-dependent laminopathies. Science339, 1604–1607 10.1126/science.1231513 (2013). [DOI] [PubMed] [Google Scholar]
47.Berk, J. M., Tifft, K. E. & Wilson, K. L. The nuclear envelope LEM-domain protein emerin. Nucleus4, 298–314 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Kedlian, V. R. et al. Human skeletal muscle aging atlas. Nat. aging4, 727–744 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Shen, X. et al. Nonlinear dynamics of multi-omics profiles during human aging. Nat. Aging4, 1619–1634 10.1038/s43587-024-00692-2 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Whisenant, D. et al. Transient expression of an adenine base editor corrects the Hutchinson-Gilford progeria syndrome mutation and improves the skin phenotype in mice. Nat. Commun.13, 3068 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Duan, D. Lethal immunotoxicity in high-dose systemic AAV therapy. Mol. Ther.31, 3123–3126 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Jiang, T. et al. Chemical modifications of adenine base editor mRNA and guide RNA expand its application scope. Nat. Commun.11, 1979 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Song, C.-Q. et al. Adenine base editing in an adult mouse model of tyrosinaemia. Nat. Biomed. Eng.4, 125–130 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Han, J. P. et al. In vivo delivery of CRISPR-Cas9 using lipid nanoparticles enables antithrombin gene editing for sustainable hemophilia A and B therapy. Sci. Adv.8, eabj6901 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Eriksson, M. et al. Recurrent de novo point mutations in lamin A cause Hutchinson-Gilford progeria syndrome. Nature423, 293–298 10.1038/nature01629 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Liu, Y., Rusinol, A., Sinensky, M., Wang, Y. & Zou, Y. DNA damage responses in progeroid syndromes arise from defective maturation of prelamin A. J. Cell Sci.119, 4644–4649 10.1242/jcs.03263 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Scaffidi, P. & Misteli, T. Lamin A-dependent nuclear defects in human aging. Science312, 1059–1063 10.1126/science.1127168 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Goldman, R. D. et al. Accumulation of mutant lamin A causes progressive changes in nuclear architecture in Hutchinson–Gilford progeria syndrome. Proc. Natl. Acad. Sci.101, 8963–8968 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Shumaker, D. K. et al. Mutant nuclear lamin A leads to progressive alterations of epigenetic control in premature aging. Proc. Natl. Acad. Sci. USA103, 8703–8708 10.1073/pnas.0602569103 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Nissan, X. et al. Unique preservation of neural cells in Hutchinson-Gilford progeria syndrome is due to the expression of the neural-specific miR-9 microRNA. Cell Rep.2, 1–9 (2012). [DOI] [PubMed] [Google Scholar]
61.Sun, X. et al. Vascular organoid model of Hutchinson-Gilford progeria syndrome uncovers repression of the SRF pathway in premature aging. Dev. Cell61, 505–517.e507 (2026). [DOI] [PubMed] [Google Scholar]
62.Yang, S. H., Andres, D. A., Spielmann, H. P., Young, S. G. & Fong, L. G. Progerin elicits disease phenotypes of progeria in mice whether or not it is farnesylated. J. Clin. Investig.118, 3291–3300 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Yang, S. H. et al. Absence of progeria-like disease phenotypes in knock-in mice expressing a non-farnesylated version of progerin. Hum. Mol. Genet.20, 436–444 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Kenjo, E. et al. Low immunogenicity of LNP allows repeated administrations of CRISPR-Cas9 mRNA into skeletal muscle in mice. Nat. Commun.12, 7101. 10.1038/s41467-021-26714-w (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Cheng, Q. et al. Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR–Cas gene editing. Nat. Nanotechnol.15, 313–320 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Tilstra, G. et al. Iterative design of ionizable lipids for intramuscular mRNA delivery. J. Am. Chem. Soc.145, 2294–2304 10.1021/jacs.2c10670 (2023). [DOI] [PubMed] [Google Scholar]
67.Ji, J. et al. Myomerger-derived peptide enhances skeletal muscle tropism and reduces liver transduction of lipid nanoparticles for gene delivery. Mol. Ther. Nucleic Acids37, 102785. 10.1016/j.omtn.2025.102785 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Kim, Y.-J. et al. Longitudinal analysis of genetic and epigenetic changes in human pluripotent stem cells in the landscape of culture-induced abnormality. Exp. Mol. Med.56, 2409–2422 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Udi, Y. et al. A general method for quantitative fractionation of mammalian cells. J. Cell Biol.222, e202209062 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Kwon, E.-J. Analysis code for “Mutation-agnostic base editing of the progerin farnesylation site rescues Hutchinson–Gilford progeria syndrome phenotypes in neuromuscular organoids”. Zenodo. 10.5281/zenodo.21366828 (2026). [DOI] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information^{(10MB, pdf)}

41467_2026_77251_MOESM2_ESM.pdf^{(5.6KB, pdf)}

Description of Additional Supplementary Files

Supplementary Movie 1^{(24.8MB, mp4)}

Supplementary Movie 2^{(24.7MB, mp4)}

Supplementary Movie 3^{(24.6MB, mp4)}

Supplementary Movie 4^{(24.8MB, mp4)}

Supplementary Movie 5^{(24.6MB, mp4)}

Supplementary Movie 6^{(24.8MB, mp4)}

Reporting Summary^{(3MB, pdf)}

Transparent Peer Review file^{(13.2MB, pdf)}

Source Data^{(27.2MB, zip)}

Data Availability Statement

[CR1] 1.Davies, B. S., Fong, L. G., Yang, S. H., Coffinier, C. & Young, S. G. The posttranslational processing of prelamin A and disease. Annu. Rev. Genomics Hum. Genet.10, 153–174 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Eriksson, M. et al. Recurrent de novo point mutations in lamin A cause Hutchinson–Gilford progeria syndrome. Nature423, 293–298 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Capell, B. C. et al. Inhibiting farnesylation of progerin prevents the characteristic nuclear blebbing of Hutchinson-Gilford progeria syndrome. Proc. Natl. Acad. Sci. USA 102, 12879–12884. 10.1073/pnas.0506001102 (2005). [DOI] [PMC free article] [PubMed]

[CR4] 4.Hamczyk, M. R., Del Campo, L. & Andrés, V. Aging in the cardiovascular system: Lessons from Hutchinson-Gilford progeria syndrome. Annu. Rev. Physiol.80, 27–48 (2018). [DOI] [PubMed] [Google Scholar]

[CR5] 5.Marcelot, A. & Worman, H. J. & Zinn-Justin, S. Protein structural and mechanistic basis of progeroid laminopathies. FEBS J.288, 2757–2772 (2021). [DOI] [PubMed] [Google Scholar]

[CR6] 6.Bergo, M. O. et al. Zmpste24 deficiency in mice causes spontaneous bone fractures, muscle weakness, and a prelamin A processing defect. Proc. Natl. Acad. Sci.99, 13049–13054 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Pendás, A. M. et al. Defective prelamin A processing and muscular and adipocyte alterations in Zmpste24 metalloproteinase–deficient mice. Nat. Genet.31, 94–99 (2002). [DOI] [PubMed] [Google Scholar]

[CR8] 8.Hennekam, R. C. Hutchinson–Gilford progeria syndrome: review of the phenotype. Am. J. Med. Genet. Part A140, 2603–2624 (2006). [DOI] [PubMed] [Google Scholar]

[CR9] 9.Merideth, M. A. et al. Phenotype and course of Hutchinson–Gilford progeria syndrome. N. Engl. J. Med.358, 592–604 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Liu, B. et al. Genomic instability in laminopathy-based premature aging. Nat. Med.11, 780–785 10.1038/nm1266 (2005). [DOI] [PubMed] [Google Scholar]

[CR11] 11.Chojnowski, A. et al. Heterochromatin loss as a determinant of progerin-induced DNA damage in Hutchinson–Gilford progeria. Aging Cell19, e13108 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Köhler, F. et al. Epigenetic deregulation of lamina-associated domains in Hutchinson-Gilford progeria syndrome. Genome Med.12, 1–16 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Gordon, L. B. et al. Clinical trial of a farnesyltransferase inhibitor in children with Hutchinson–Gilford progeria syndrome. Proc. Natl. Acad. Sci.109, 16666–16671 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Gordon, L. B. et al. Association of lonafarnib treatment vs no treatment with mortality rate in patients with Hutchinson-Gilford progeria syndrome. JAMA319, 1687–1695 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Suzuki, M. et al. FDA approval summary for lonafarnib (Zokinvy) for the treatment of Hutchinson-Gilford progeria syndrome and processing-deficient progeroid laminopathies. Genet. Med.25, 100335 (2023). [DOI] [PubMed] [Google Scholar]

[CR16] 16.Cao, K. et al. Rapamycin reverses cellular phenotypes and enhances mutant protein clearance in Hutchinson-Gilford progeria syndrome cells. Sci. Transl. Med.3, 89ra58–89ra58 (2011). [DOI] [PubMed] [Google Scholar]

[CR17] 17.Koblan, L. W. et al. In vivo base editing rescues Hutchinson–Gilford progeria syndrome in mice. Nature589, 608–614 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Puttaraju, M. et al. Systematic screening identifies therapeutic antisense oligonucleotides for Hutchinson–Gilford progeria syndrome. Nat. Med.27, 526–535 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Erdos, M. R. et al. A targeted antisense therapeutic approach for Hutchinson–Gilford progeria syndrome. Nat. Med.27, 536–545 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Beyret, E. et al. Single-dose CRISPR-Cas9 therapy extends lifespan of mice with Hutchinson-Gilford progeria syndrome. Nat. Med25, 419–422 10.1038/s41591-019-0343-4 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Hunt, J. M. T., Samson, C. A., Rand, A. D. & Sheppard, H. M. Unintended CRISPR-Cas9 editing outcomes: a review of the detection and prevalence of structural variants generated by gene-editing in human cells. Hum. Genet.142, 705–720 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Liu, G.-H. et al. Recapitulation of premature ageing with iPSCs from Hutchinson–Gilford progeria syndrome. Nature472, 221–225 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Zhang, J. et al. A human iPSC model of Hutchinson Gilford Progeria reveals vascular smooth muscle and mesenchymal stem cell defects. Cell Stem Cell8, 31–45 (2011). [DOI] [PubMed] [Google Scholar]

[CR24] 24.Musunuru, K. Genome editing of human pluripotent stem cells to generate human cellular disease models. Dis. Models Mech.6, 896–904 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Park, J.-C. et al. Gene editing with ‘pencil’rather than ‘scissors’ in human pluripotent stem cells. Stem Cell Res. Ther.14, 164 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Park, J.-C. et al. Enhancing genome editing in hPSCs through dual inhibition of DNA damage response and repair pathways. Nat. Commun.15, 4002 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Park, J.-C. et al. MutSα and MutSβ as size-dependent cellular determinants for prime editing in human embryonic stem cells. Mol. Ther. -Nucleic Acids32, 914–922 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Faustino Martins, J. M. et al. Self-organizing 3D Human Trunk Neuromuscular Organoids. Cell Stem Cell26, 172–186.e176 10.1016/j.stem.2019.12.007 (2020). [DOI] [PubMed] [Google Scholar]

[CR29] 29.Suarez-Calvet, X. et al. Decoding the transcriptome of Duchenne muscular dystrophy to the single nuclei level reveals clinical-genetic correlations. Cell Death Dis.14, 596. 10.1038/s41419-023-06103-5 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Gonzalo, S. & Kreienkamp, R. DNA repair defects and genome instability in Hutchinson-Gilford Progeria Syndrome. Curr. Opin. Cell Biol.34, 75–83 10.1016/j.ceb.2015.05.007 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Monnerat, G. et al. Metabolomic profiling suggests systemic signatures of premature aging induced by Hutchinson-Gilford progeria syndrome. Metabolomics15, 100. 10.1007/s11306-019-1558-6 (2019). [DOI] [PubMed] [Google Scholar]

[CR32] 32.Kreienkamp, R. et al. Doubled lifespan and patient-like pathologies in progeria mice fed high-fat diet. Aging Cell18, e12852 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Harhouri, K. et al. MG 132-induced progerin clearance is mediated by autophagy activation and splicing regulation. EMBO Mol. Med.9, 1294–1313 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Lee, J. M. et al. Modulation of LMNA splicing as a strategy to treat prelamin A diseases. J. Clin. Invest126, 1592–1602 10.1172/JCI85908 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Gibbs-Seymour, I., Markiewicz, E., Bekker-Jensen, S., Mailand, N. & Hutchison, C. J. Lamin A/C-dependent interaction with 53 BP 1 promotes cellular responses to DNA damage. Aging cell14, 162–169 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Panier, S. & Boulton, S. J. Double-strand break repair: 53BP1 comes into focus. Nat. Rev. Mol. Cell Biol.15, 7–18 (2014). [DOI] [PubMed] [Google Scholar]

[CR37] 37.Etourneaud, L. et al. Lamin B1 sequesters 53BP1 to control its recruitment to DNA damage. Sci. Adv.7, eabb3799 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Cobb, A. M. et al. Prelamin A impairs 53 BP 1 nuclear entry by mislocalizing NUP 153 and disrupting the Ran gradient. Aging cell15, 1039–1050 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Cha, H. et al. Wip1 directly dephosphorylates γ-H2AX and attenuates the DNA damage response. Cancer Res.70, 4112–4122 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Adikusuma, F. et al. Large deletions induced by Cas9 cleavage. Nature560, E8–E9 (2018). [DOI] [PubMed] [Google Scholar]

[CR41] 41.Shin, H. Y. et al. CRISPR/Cas9 targeting events cause complex deletions and insertions at 17 sites in the mouse genome. Nat. Commun.8, 15464 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Papathanasiou, S. et al. Whole chromosome loss and genomic instability in mouse embryos after CRISPR-Cas9 genome editing. Nat. Commun.12, 5855 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Walton, R. T., Christie, K. A., Whittaker, M. N. & Kleinstiver, B. P. Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants. Science368, 290–296 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Ren, Q. et al. PAM-less plant genome editing using a CRISPR–SpRY toolbox. Nat. Plants7, 25–33 (2021). [DOI] [PubMed] [Google Scholar]

[CR45] 45.Moiseeva, O., Lopes-Paciencia, S., Huot, G., Lessard, F. & Ferbeyre, G. Permanent farnesylation of lamin A mutants linked to progeria impairs its phosphorylation at serine 22 during interphase. Aging (Albany NY)8, 366 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Quigley, A. et al. The structural basis of ZMPSTE24-dependent laminopathies. Science339, 1604–1607 10.1126/science.1231513 (2013). [DOI] [PubMed] [Google Scholar]

[CR47] 47.Berk, J. M., Tifft, K. E. & Wilson, K. L. The nuclear envelope LEM-domain protein emerin. Nucleus4, 298–314 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Kedlian, V. R. et al. Human skeletal muscle aging atlas. Nat. aging4, 727–744 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Shen, X. et al. Nonlinear dynamics of multi-omics profiles during human aging. Nat. Aging4, 1619–1634 10.1038/s43587-024-00692-2 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Whisenant, D. et al. Transient expression of an adenine base editor corrects the Hutchinson-Gilford progeria syndrome mutation and improves the skin phenotype in mice. Nat. Commun.13, 3068 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Duan, D. Lethal immunotoxicity in high-dose systemic AAV therapy. Mol. Ther.31, 3123–3126 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Jiang, T. et al. Chemical modifications of adenine base editor mRNA and guide RNA expand its application scope. Nat. Commun.11, 1979 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Song, C.-Q. et al. Adenine base editing in an adult mouse model of tyrosinaemia. Nat. Biomed. Eng.4, 125–130 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Han, J. P. et al. In vivo delivery of CRISPR-Cas9 using lipid nanoparticles enables antithrombin gene editing for sustainable hemophilia A and B therapy. Sci. Adv.8, eabj6901 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Eriksson, M. et al. Recurrent de novo point mutations in lamin A cause Hutchinson-Gilford progeria syndrome. Nature423, 293–298 10.1038/nature01629 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Liu, Y., Rusinol, A., Sinensky, M., Wang, Y. & Zou, Y. DNA damage responses in progeroid syndromes arise from defective maturation of prelamin A. J. Cell Sci.119, 4644–4649 10.1242/jcs.03263 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Scaffidi, P. & Misteli, T. Lamin A-dependent nuclear defects in human aging. Science312, 1059–1063 10.1126/science.1127168 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Goldman, R. D. et al. Accumulation of mutant lamin A causes progressive changes in nuclear architecture in Hutchinson–Gilford progeria syndrome. Proc. Natl. Acad. Sci.101, 8963–8968 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Shumaker, D. K. et al. Mutant nuclear lamin A leads to progressive alterations of epigenetic control in premature aging. Proc. Natl. Acad. Sci. USA103, 8703–8708 10.1073/pnas.0602569103 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Nissan, X. et al. Unique preservation of neural cells in Hutchinson-Gilford progeria syndrome is due to the expression of the neural-specific miR-9 microRNA. Cell Rep.2, 1–9 (2012). [DOI] [PubMed] [Google Scholar]

[CR61] 61.Sun, X. et al. Vascular organoid model of Hutchinson-Gilford progeria syndrome uncovers repression of the SRF pathway in premature aging. Dev. Cell61, 505–517.e507 (2026). [DOI] [PubMed] [Google Scholar]

[CR62] 62.Yang, S. H., Andres, D. A., Spielmann, H. P., Young, S. G. & Fong, L. G. Progerin elicits disease phenotypes of progeria in mice whether or not it is farnesylated. J. Clin. Investig.118, 3291–3300 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Yang, S. H. et al. Absence of progeria-like disease phenotypes in knock-in mice expressing a non-farnesylated version of progerin. Hum. Mol. Genet.20, 436–444 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Kenjo, E. et al. Low immunogenicity of LNP allows repeated administrations of CRISPR-Cas9 mRNA into skeletal muscle in mice. Nat. Commun.12, 7101. 10.1038/s41467-021-26714-w (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Cheng, Q. et al. Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR–Cas gene editing. Nat. Nanotechnol.15, 313–320 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR66] 66.Tilstra, G. et al. Iterative design of ionizable lipids for intramuscular mRNA delivery. J. Am. Chem. Soc.145, 2294–2304 10.1021/jacs.2c10670 (2023). [DOI] [PubMed] [Google Scholar]

[CR67] 67.Ji, J. et al. Myomerger-derived peptide enhances skeletal muscle tropism and reduces liver transduction of lipid nanoparticles for gene delivery. Mol. Ther. Nucleic Acids37, 102785. 10.1016/j.omtn.2025.102785 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Kim, Y.-J. et al. Longitudinal analysis of genetic and epigenetic changes in human pluripotent stem cells in the landscape of culture-induced abnormality. Exp. Mol. Med.56, 2409–2422 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR69] 69.Udi, Y. et al. A general method for quantitative fractionation of mammalian cells. J. Cell Biol.222, e202209062 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Kwon, E.-J. Analysis code for “Mutation-agnostic base editing of the progerin farnesylation site rescues Hutchinson–Gilford progeria syndrome phenotypes in neuromuscular organoids”. Zenodo. 10.5281/zenodo.21366828 (2026). [DOI] [PubMed]

PERMALINK

Mutation-agnostic base editing of the progerin farnesylation site rescues Hutchinson-Gilford progeria syndrome phenotypes in neuromuscular organoids

Dong-Woo Kim

Eun-Ji Kwon

Beom-Jin Jeon

Dong-Hyeok Kwon

Goo Jang

Youngyoon Yoon

Yuna Hwang

Hyukjin Lee

Hyuk-Jin Cha

Abstract

Introduction

Results

Establishment of two isogenic NMO models with mesoderm-specific progerin expression

Fig. 1. Establishment of two sets of isogenic NMO pairs, with progerin expression restricted to the muscular compartment of HGPS- and Mutant-NMOs.

Myopathy-associated gene signatures in the muscle compartment of HGPS-NMOs

Fig. 2. Single-cell analysis of skeletal muscle cellular composition and developmental trajectories in HGPS and Edit NMOs.

Impairment of DNA damage repair machinery in the muscle compartment of HGPS-NMOs

Fig. 3. Impaired foci formation of DNA damage repair machinery and heterochromatin loss in the muscle compartment of HGPS-NMOs.

Specific farnesylation amino acid targeted editing (FATE) of LMNA

Fig. 4. Farnesylation Amino acid Targeted Editing (FATE) exclusively in LMNA.

Loss of progerin accumulation and restoration of DNA damage foci in FATE-NMOs

Fig. 5. Loss of progerin accumulation, restoration of DNA repair machinery foci formation, recovery of heterochromatin, and normalization of transcriptomic profiles in FATE-NMOs.

Recovery of HGPS-NMOs by delivery of FATE mRNA

Fig. 6. Recovery of disease phenotypes in HGPS-NMOs by delivery of FATE mRNA via direct microinjection.

Discussion

Methods

Ethics

Culture of human pluripotent stem cells

Generation of atypical HGPS iPSCs

FATE plasmid construction

Generation of neuromuscular organoids (NMOs) from hPSCs via induction of human neuromesodermal progenitors

Immunohistochemistry of organoids

Immunoblotting analysis

Real-time PCR analysis

Fractionation of organoids and immunoblotting analysis

8% PVP buffer

Lysis buffer

DNase buffer

Electron microscopy of organoids

Single cell preparation from organoids for scRNAseq

Sample preparation and quality control

scRNA-seq library preparation and sequencing

Primary data processing

Single-cell RNA-seq data analysis

Differential expression and pathway enrichment analysis

Pathway activity estimation and PCA analysis based on pseudobulk GSVA

Derivation and application of young muscle signature

Whole exome sequencing (WES) and data processing

Targeted deep sequencing and data processing

Identification of gRNA-dependent off-target Effects

Identification of gRNA-independent off-target effects

Analysis of CAAX motif sequences

mRNA synthesis via In vitro transcription (IVT)

Preparation and characterization of LNPs

mRNA-LNP microinjection

Calculation of editing efficiency via mRNA delivery

Statistics and reproducibility

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions

Peer review

Peer review information

Funding

Data availability

Code availability

Competing interests

Footnotes

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles