ABSTRACT
Mutations in the human LMNA gene, which encodes the nuclear envelope (NE) proteins lamins A and C, cause autosomal dominant Emery-Dreifuss muscular dystrophy, congenital muscular dystrophy, limb-girdle muscular dystrophy, and other diseases collectively known as laminopathies. The molecular mechanisms responsible for these diseases remain incompletely understood, but the muscle-specific defects suggest that mutations may render nuclei more susceptible to mechanical stress. Using three mouse models of muscle laminopathies, we found that Lmna mutations caused extensive NE abnormalities, consisting of chromatin protrusions into the cytoplasm and transient rupture of the NE in skeletal muscle cells. NE damage was associated with DNA damage, activation of DNA damage response pathways, and reduced viability. Intriguingly, NE damage resulted from nuclear migration in maturing skeletal muscle cells, rather than actomyosin contractility. NE damage and DNA damage was reduced by either depletion of kinesin-1 or disruption of the Linker of Nucleoskeleton and Cytoskeleton (LINC) complex. LINC complex disruption rescued myofiber function and viability in Lmna mutant myofibers, indicating that the myofiber dysfunction is the result of mechanically induced NE damage. The extent of NE damage and DNA damage in Lmna mouse models correlated with the disease onset and severity in vivo. Moreover, inducing DNA damage in wild-type muscle cells was sufficient to phenocopy the reduced cell viability of lamin A/C-deficient muscle cells, suggesting a causative role of DNA damage in disease pathogenesis. Corroborating the mouse model data, muscle biopsies from patients with LMNA muscular dystrophy revealed significant DNA damage compared to age-matched controls, particularly in severe cases of the disease. Taken together, these findings point to a new and important role of DNA damage as a pathogenic contributor for LMNA skeletal muscle diseases.
INTRODUCTION
Lamins A and C, together with the B-type lamins, are the major components of the nuclear lamina, which line the inner nuclear membrane. Lamins A/C play important roles in providing structural support to the nucleus and connecting the nucleus to the cytoskeleton2. In addition, they participate in transcriptional regulation, genome organization, and DNA damage and repair2–5. The majority of the over 450 LMNA mutations identified to date are responsible for autosomal dominant Emery-Dreifuss muscular dystrophy (AD-EDMD)6, characterized by slowly progressive skeletal muscle wasting, contractures of the elbow, neck, and Achilles tendons, a rigid spine, abnormal heart rhythms, heart block, and cardiomyopathy7. Other LMNA mutations cause congenital muscular dystrophy (LMNA-CMD), a particularly severe form of muscular dystrophy with onset in early childhood8, and limb girdle muscular dystrophy, which affects proximal muscles of the hips and shoulders9. It remains unclear how LMNA mutations result in muscle-specific defects, and the incomplete understanding of the disease pathogenesis presents a major hurdle in the development of effective treatment approaches.
One potential explanation, the ‘mechanical stress’ hypothesis, states that LMNA mutations linked to muscular phenotypes result in structurally impaired nuclei that become damaged in mechanically active tissues, such as cardiac and skeletal muscle5. This hypothesis is supported by findings of decreased nuclear stiffness in fibroblasts expressing LMNA mutations linked to striated muscle laminopathies, impaired assembly of mutant lamins in vitro, and anecdotal reports of NE damage in skeletal and cardiac muscle cells of individuals with AD-EDMD and LMNA-related dilated cardiomyopathies2, 5, 10, 11. However, striated muscle laminopathies involve progressive muscle wasting, and analysis of fixed tissues and cells provides only a snapshot of the development of disease. Thus, the functional relevance of NE damage, particularly whether the NE damage is a cause or consequence of muscle dysfunction, remains unclear.
To better understand the mechanistic link between NE damage and muscle dysfunction, we employed three laminopathy mouse models with varying degrees of disease severity, along with a recently developed long-term in vitro muscle differentiation platform12. Using high resolution time lapse microscopy, we tracked myonuclear shape and integrity over time and recorded structural and functional changes during the differentiation of primary mouse myoblasts into multinucleated myotubes and subsequent formation of mature, contractile myofibers (Fig. 1a, b). Combining these models with novel, fluorescent nuclear damage reporters13, both in vitro and in vivo, we obtained unprecedented systematic and detailed temporal and mechanistic information of disease progression in mouse models of laminopathies, and corroborated key findings in skeletal muscle biopsies from humans with LMNA muscular dystrophy.
We found that myonuclei in Lmna mutant muscle cells exhibited progressive NE damage in vitro and in vivo, including extensive chromatin protrusions into the cytoplasm and transient NE rupture. Intriguingly, NE rupture was associated with progressive DNA damage and DNA damage response activation. Inducing DNA damage in wild-type muscle cells was sufficient to provoke defects in cell viability similar to those observed in the lamin A/C-deficient cells. In addition, preventing NE rupture by LINC complex disruption was sufficient to reduce DNA damage and rescue myofiber viability and contractility in lamin A/C-deficient cells. Together, our findings indicate a causative role of NE rupture and DNA damage in progressive muscle decline and provide a novel explanation for how lamin mutations lead to muscle weakness and wasting in AD-EDMD.
RESULTS
Lmna mutations cause progressive decline in myofiber health in vitro and in vivo
To examine the effect of Lmna mutations on nuclear mechanics and muscle function in vitro, we isolated myoblasts from three established mouse models of striated muscle laminopathies, representing a spectrum of muscle wasting and disease severity (Fig. 1a, Suppl. Fig. S1): Lamin A/C-deficient (Lmna–/–) mice14, subsequently referred to as lamin A/C knock-out mice (Lmna KO); knock-in mice carrying the LmnaN195K/N195K mutation (Lmna N195K)15; knock-in mice carrying the LmnaH222P/H222P mutation (Lmna H222P)16; and wild-type littermates. While the Lmna N195K mice were originally described as a model for dilated cardiomyopathy15, in the C57BL/6 background used in our studies, the mice developed pronounced skeletal muscular dystrophy in addition to cardiac defects (Suppl. Fig. S1). For in vitro studies, we utilized a recently developed, three-dimensional culture protocol to differentiate primary myoblasts into mature, contractile myofibers over the course of ten days (Fig. 1b)17, 18. The resulting myofibers display the highly organized sarcomeric structure and evenly spaced peripheral myonuclear positioning characteristic of mature skeletal muscle fibers in vivo (Suppl. Fig. S2).
All myoblasts, including the Lmna KO, Lmna N195K and Lmna H222P mutant cells, successfully completed myoblast fusion, differentiated into myotubes, and matured into myofibers (Fig. 1c), consistent with previous studies on differentiation of Lmna KO myoblasts19, 20. Wild-type myofibers remained healthy and highly contractile up to ten days of differentiation
(Suppl. Movie 1). In contrast, the Lmna KO myofibers showed a decline in cell contractility, viability, and number of myonuclei, starting at day five of differentiation (Fig. 1c, d; Suppl. Fig. S3, Suppl. Movie 2). The Lmna N195K myofibers showed a similar, albeit slightly delayed decline in cell viability, contractility, and number of myonuclei by day ten of differentiation (Fig. 1d, Suppl. Fig S3, Suppl. Movie 3). The reduction in cell viability at day ten in the Lmna KO and Lmna N195K myofibers was associated with an increase in activated caspase-3 (Fig. 1e,f), indicating that reduced viability was due at least in part to cell-intrinsic apoptosis. Unlike the Lmna KO and Lmna N195K models, the Lmna H222P myofibers exhibited no significant decrease in viability, contractility, or number of myonuclei (Fig. 1d, Suppl. Fig S3, Suppl. Movie 4) within ten days of differentiation. Taken together, the long-term differentiation assays revealed a striking correlation among the defects observed in vitro, including loss of muscle cell viability and presence of apoptotic markers, with the severity of the disease in the corresponding mouse models (Fig 1a, Suppl. Fig. S1), suggesting that defects in the in vitro model may serve as prognostic markers for disease progression.
Lmna mutant muscle cells have reduced nuclear stability that corresponds to disease severity
We hypothesized that the progressive deterioration of Lmna mutant myofibers results from damage to mechanically weakened myonuclei exposed to cytoskeletal forces. To test this hypothesis, we measured the nuclear deformability in primary myoblasts from the three laminopathy models using a novel high throughput microfluidics-based micropipette aspiration assay (Fig. 2a, Suppl. Movie 5)21. Nuclei from Lmna KO and Lmna N195K myoblasts were substantially more deformable than nuclei from wild-type controls (Fig. 2b,c). Intriguingly, myoblasts from Lmna H222P mice, which have a later disease onset and less severe muscle defects than the other two Lmna mutant models (Fig. 1a; Suppl. Fig. S1), had only a modest increase in nuclear deformability relative to wild-type controls (Fig. 2b,c; Suppl. Fig. S4a). Ectopic expression of lamin A significantly reduced the nuclear deformability defect in primary Lmna KO myoblasts (Suppl. Fig. S4b-d), confirming that the impaired nuclear stability was a direct consequence of altering the nuclear lamina. In addition, primary myoblasts from Mdx mice, which develop mild muscular dystrophy due to loss of the cell membrane-associated protein dystrophin, had nuclear deformation indistinguishable from wild-type controls (Suppl. Fig. S5), indicating that the defects in nuclear stability are specific to Lmna mutations and not muscular dystrophy in general.
To assess whether the observed defects in nuclear stability also occur in more mature, multinucleated myofibers, we subjected Lmna KO and wild-type myofibers to a ‘microharpoon’ assay, in which precise strain is exerted on the perinuclear cytoskeleton, and the induced nuclear deformation and displacement are used to infer nuclear stability and nucleo-cytoskeletal coupling, respectively22, 23. Lmna KO myofibers had significantly more deformable nuclei than wild-type controls (Fig. 2d,e; Suppl. Movie 6), consistent with the micropipette aspiration results in the myoblasts. Furthermore, analysis of Lmna mutant and wild-type myofibers at day five of in vitro differentiation revealed that Lmna KO, Lmna N195K, and Lmna H222P myofibers had significantly elongated myonuclei compared to wild-type controls (Fig. 2f,g), consistent with decreased nuclear stability in the Lmna mutant cells and with previous reports of elongated nuclei in muscle biopsies from laminopathy patients24. Taken together, these findings suggest that myopathic Lmna mutations that cause severe muscle defects result in mechanically weaker myonuclei.
Lmna mutant myonuclei display chromatin protrusions into the cytoplasm
Physical stress on the nucleus during external compression or confined migration can induce chromatin protrusions across the nuclear lamina into the cytoplasm, particularly in lamin-deficient cells13, 25. To test whether the mechanically weaker Lmna mutant myonuclei are prone to mechanically induced damage in muscle cells, we analyzed nuclear structure and morphology over the ten-day time course of differentiation and maturation of primary myoblasts from laminopathy mouse models and healthy controls. Despite their mechanically weaker nuclei (Fig. 2), the Lmna mutant myoblasts show no nuclear abnormalities prior to differentiation (Fig. 3b). Following the onset of differentiation, however, Lmna KO, Lmna N195K, and Lmna H222P myofibers exhibited striking chromatin protrusions that were absent in wild-type fibers. These protrusions extended beyond the (B-type) nuclear lamina up to tens of microns into the cytoplasm (Fig. 3a,b). The protrusions were enclosed by nuclear membranes, as indicated by the frequent presence of the nuclear membrane protein emerin, and occasional presence of nesprin-1 (Suppl. Fig. S6a); however, these NE proteins were often concentrated in punctae inside the protrusions and myonuclei. Other NE proteins, such as nuclear pore complex proteins, were largely absent from the protrusions (Suppl. Fig. S6b), suggesting an altered membrane composition in the chromatin protrusion, similar to what has been reported in analogous structures in migrating cancer cells13, 26.
The frequency of chromatin protrusion was highest in Lmna KO myofibers, followed by Lmna N195K and then Lmna H222P myofibers (Fig. 3b), correlating with the increased nuclear deformability in vitro (Fig. 2) and the disease severity in vivo (Fig. 1a; Suppl. Fig. S1). Intriguingly, while Lmna KO and Lmna N195K myofibers had extensive chromatin protrusions at day five of differentiation, the frequency of chromatin protrusions in the Lmna H222P cells was initially low, but increased from five to ten days of differentiation (Fig. 3b), matching the delayed disease onset and progressive phenotype of the Lmna H222P model in vivo. Ectopic expression of lamin A in Lmna KO myoblasts significantly reduced the occurrence of chromatin protrusions at ten days of differentiation (Fig. 3b), confirming that the protrusions were caused by loss of lamin A/C expression.
To confirm the results of the in vitro studies in vivo, we isolated single muscle fibers from the hindlimbs of Lmna KO and wild-type mice. Chromatin protrusions were not detectable in muscle fibers from wild-type mice (Fig. 3c,d). In contrast, nuclei in muscle fibers from Lmna KO mice had similar chromatin protrusions as observed in the in vitro differentiated myofibers (Fig. 3c,d). Interestingly, the prevalence of chromatin protrusions in the Lmna KO myonuclei strongly depended on the location within the muscle. Myonuclei at the myotendinous junctions (MTJ) had a higher frequency of chromatin protrusions than nuclei in the muscle fiber body (Fig. 3d), consistent with a previous report of nuclear abnormalities at the MTJ in Lmna KO mice27, and possibly reflecting increased mechanical stress at the MTJ.
Nuclear damage is intrinsic to lamin A/C-deficient nuclei
To address whether the observed NE defects in Lmna mutant muscle cells are nucleus-intrinsic or arise from altered signaling pathways or other cytoplasmic changes in the mutant cells, we generated “hybrid” myofibers by combining wild-type and Lmna KO myoblasts prior to differentiation. Following myoblast fusion, these cells formed multinucleated myotubes and myofibers that contained both wild-type and Lmna KO nuclei with a shared cytoplasm (Fig. 3e). Importantly, in skeletal muscle cells, each nucleus is thought to provide mRNA transcripts for the nearby cytoplasm (referred to as myonuclear domain), so that the local RNA and protein content primarily stems from the nearest myonucleus28, and the genotype of each myonucleus can be determined by antibody staining against lamin A (Fig. 3e). We quantified the number of nuclei with chromatin protrusions and compared genetically identical nuclei (e.g., wild-type or Lmna KO) from hybrid and isogenic control myofibers after ten days of differentiation (Fig. 3f). Hybrid myofibers comprising ∼80% wild-type nuclei and ∼20% Lmna KO nuclei appeared healthy. Nonetheless, Lmna KO nuclei within the hybrid myofibers showed the same relative frequency of chromatin protrusions as nuclei from isogenic Lmna KO myofibers (Fig. 3f). Conversely, wild-type nuclei in hybrid fibers lacked chromatin protrusions and were thus not adversely affected by the presence of Lmna KO nuclei in the shared cytoplasm (Fig. 3f). These results indicate that the defects in nuclear structure are intrinsic to the Lmna mutant myonuclei and not due to impaired muscle fiber health, altered cytoplasmic signaling, or changes in the cytoplasmic architecture in Lmna mutant muscle fibers.
Lmna mutant myofibers exhibit extensive NE rupture in vitro
Physical compression by cytoskeletal forces can result in NE rupture, with depletion of lamins exacerbating the frequency of NE rupture13, 29–33. To examine whether the reduced nuclear stability seen in Lmna mutant muscle cells (Fig. 2) leads to NE rupture in Lmna mutant myofibers, we modified primary myoblasts to co-express a fluorescent NE rupture reporter, consisting of a green fluorescent protein with a nuclear localization signal (NLS-GFP)13 and fluorescently labeled histone (H2B-tdTomato). NLS-GFP is normally localized to the nucleus, but rapidly spills into the cytoplasm upon loss of nuclear membrane integrity and is then gradually reimported into the nucleus after the nuclear membrane has been repaired13. In vitro differentiated Lmna KO myotubes frequently exhibited NE ruptures (Fig. 4a, Suppl. Movie 7), which were absent in wild-type controls. To investigate NE rupture in more detail, we stably modified primary myoblasts with another fluorescent NE rupture reporter, cGAS-mCherry, which accumulates at the rupture site13 (Fig. 4b). Unlike the transient cytoplasmic NLS-GFP signal, however, the cGAS-mCherry accumulation persists even after the NE has been repaired13, 29. Wild-type myotubes had no detectable accumulation of cGAS-mCherry (Fig. 4c). In contrast,
Lmna KO myotubes displayed a progressive increase in the number of nuclear cGAS-mCherry foci during differentiation, starting around day two, which could be rescued by ectopic expression of wild-type lamin A (Fig. 4c). Lmna N195K showed intermediate levels of NE rupture (Suppl. Fig. S7A), whereas Lmna H222P myotubes had cGAS-mCherry accumulation comparable to wild-type controls (Suppl. Fig. S7b), consistent with the milder defects in nuclear stability in the Lmna H222P mutant cells (Fig. 2b,c; Supp. Fig. S4a).
Lamin A/C-deficient myofibers experience extensive NE rupture in vivo
To test whether NE rupture occurs within Lmna KO muscle in vivo, we generated transgenic mice that express a fluorescent cGAS-tdTomato nuclear rupture reporter and crossed these mice into the Lmna mutant mouse models. Single hindlimb muscle fibers isolated from Lmna KO offspring expressing the cGAS-tdTomato reporter revealed a large fraction of myonuclei with cGAS-tdTomato foci, which were absent in both wild-type littermates expressing the cGAS-tdTomato reporter and in Lmna KO mice that were negative for the cGAS-tdTomato reporter (Fig. 4d-f). Within Lmna KO muscle fibers, the frequency of NE rupture was significantly higher at the MTJ than in the myofiber body nuclei (Fig. 4e; Suppl. Fig. S8), consistent with the increased frequency of chromatin protrusions in the MTJ myonuclei. The amount of cGAS-tdTomato accumulation scaled with disease severity: Lmna KO fibers had the highest amount of cGAS-tdTomato foci, Lmna N195K fibers had an intermediate amount, and Lmna H222P fibers had no nuclear cGAS-tdTomato accumulation, closely matching the in vitro data (Suppl. Fig. S7c). As an independent approach to detect loss of nuclear-cytoplasmic compartmentalization in muscle fibers, we analyzed the intracellular localization of endogenous heat shock protein 90 (Hsp90), which is typically excluded from the nucleus in healthy muscle fibers34. In our in vitro assays, Hsp90 was cytoplasmic in wild-type myofibers, whereas all three Lmna mutant models had increased nuclear Hsp90 levels during myoblast differentiation (Suppl. Fig. S9a,b). Similarly, muscle fibers isolated from Lmna KO mice, but not wild-type littermates, had a significant increase in nuclear Hsp90 (Suppl. Fig. S9c,d), confirming the occurrence of NE rupture in vivo. Taken together, these findings indicate widespread NE rupture in severe cases of laminopathic skeletal muscle.
Lmna KO myonuclei have increased levels of DNA damage in vitro and in vivo
Recent studies found that nuclear deformation and NE rupture can cause DNA damage in migrating cells13, 29, 35. To investigate whether chromatin protrusions and NE rupture can similarly lead to DNA damage in muscle cells, we quantified DNA damage in differentiating primary myoblasts by staining for γH2AX, a marker for double stranded DNA damage36. Both Lmna KO and wild-type myoblasts had elevated levels of DNA damage at the onset of differentiation (Fig. 5a,b), consistent with previous reports that show the transition from myoblasts to myotubes is associated with a transient increase in γH2AX levels37, 38. However, while γH2AX levels in wild-type myotubes subsequently decreased and then remained stable at low levels, the fraction of myonuclei with severe DNA damage in the Lmna KO cells continued to increase from five to ten days post-differentiation, with nearly 20% of Lmna KO myonuclei exhibiting severe DNA damage at day ten (Fig. 5b). Consistent with the increased DNA damage, Lmna KO myotubes exhibited significantly increased activity of the DNA-dependent protein kinase, DNA-PK (Fig. 5c), one of the major DNA damage response pathways in post-mitotic cells36. Single muscle fibers isolated from Lmna KO mice similarly contained many myonuclei with extensive γH2AX staining (Fig. 5d,e) and increased DNA-PK activity, especially at the MTJ, where over 10% of nuclei show very high intensity staining (Fig. 5f, Suppl. Fig S10), confirming the presence of extensive DNA damage in Lmna KO muscle fibers in vivo. The Lmna N195K single muscle fibers showed similar high levels of DNA damage compared to Lmna KO mice, whereas the Lmna H222P single muscle fibers showed an increase only in mid and low levels of DNA damage, consistent with the milder disease severity and other phenotypic markers (Fig. 5e). In contrast, muscle fibers isolated from wild-type mice contained only low levels of DNA damage (Fig. 5d,e). Of the myonuclei with extensive γH2AX staining, 82% also had chromatin protrusions, suggesting a link between physical damage to the nucleus and DNA damage (Suppl. Fig. S11).
To determine whether the accumulation of DNA damage during in vitro differentiation of Lmna KO myoblasts was caused by progressive new nuclear damage or defects in DNA damage repair, which had been reported in several progeroid laminopathies39–41, we subjected Lmna KO and wild-type myofibers to a pulse of gamma irradiation and monitored γH2AX levels at 3, 6, and 24 hours post-treatment. Consistent with previous studies42, irradiation resulted in a rapid increase in the number of γH2AX foci at 3 hours that then gradually resolved and returned to baseline by 24 hours post irradiation (Suppl. Fig. S12). Notably, following irradiation, Lmna KO myofibers displayed a DNA damage profile nearly identical to wild-type controls, suggesting that their ability to repair DNA damage is not significantly impaired, and the accumulation of DNA damage in the myotubes is more likely due to new incidents of damage.
Accumulation of DNA damage correlates with myofiber death
To test whether accumulation of DNA damage is sufficient to explain the progressive decline in myofiber viability in the Lmna KO cells, we subjected Lmna KO and wild-type myofibers to repeated treatments of phleomycin, a radiation mimetic agent, in conjuction with inhibition of DNA damage repair with NU7441, a DNA-PK-specific inhibitor, and/or KU55933, an ATM-specific inhibitor (Fig. 5g). Combined treatment with phleomycin and DNA damage repair inhibition (by one or both inhibitors in combination) resulted in an accumulation of DNA damage and loss of viability in wild-type myofibers comparable to that observed in untreated Lmna KO cells (Fig. 5h,i; Suppl. Fig.13). In contrast, phleomycin, alone or in combination with DNA damage repair inhibition, did not further reduce viability in Lmna KO myofibers (Suppl. Fig.13), suggesting that the preexisting DNA damage in these cells is already sufficient to drive myofiber decline. Since different types of DNA damage can elicit distinct repair responses in post-mitotic muscle cells43, the DNA damage and ensuing repair response associated with NE rupture may differ from that induced by either phleomycin or irradiation. Thus, it cannot be ruled out that the Lmna KO cells may exhibit some defects in DNA damage repair, or that additional mechanisms contribute to the pathogenesis in the Lmna mutant cells.
Nuclear damage in Lmna KO myofibers can be prevented by microtubule stabilization
We surmised that NE ruptures in Lmna mutant myofibers resulted from cytoskeletal forces acting on mechanically weak myonuclei, and that reducing mechanical stress on the nuclei would decrease nuclear damage. In striated muscle cells, the microtubule network remodels significantly upon differentiation to form a cage-like structure around the myonuclei44 (Suppl. Fig. S14), in large part due to the redistribution of centrosomal proteins at the NE45. To test if stabilizing this microtubule network and thereby reinforcing myonuclei can reduce chromatin protrusions and NE rupture, we treated in vitro differentiated myoblasts with low doses of the microtubule stabilizing drug, paclitaxel. Here, we focused on the Lmna KO model, which showed the most severe nuclear defects. The microharpoon assay confirmed that microtubule stabilization reinforced Lmna KO nuclei in differentiated myofibers and significantly reduced nuclear deformation in response to cytoplasmic force application (Fig. 6a,b, Suppl. Movie 8). Furthermore, paclitaxel treatment significantly reduced the percentage of nuclei with chromatin protrusions (Fig. 6c) and the incidence of NE rupture detected with the cGAS-mCherry reporter in the Lmna KO cells (Fig. 6d), suggesting that nuclear damage indeed arises from mechanical stress on the myonuclei and can be prevented by mechanically stabilizing the nuclei.
Kinesin-mediated nuclear movements are responsible for nuclear damage in Lmna KO myonuclei
The findings that nuclear damage occurred only during myoblast differentiation, when the cytoskeleton significantly remodels, and that microtubule stabilization significantly reduced the amount of nuclear damage in Lmna KO myofibers, suggest that the nuclear defects result from cytoskeletal forces acting on the mechanically weaker Lmna mutant myonuclei. Physical stress may be imparted to myonuclei via (1) actomyosin-mediated contractile forces, and/or (2) forces due to nuclear movements at various stages of muscle development, including myoblast migration and fusion46, microtubule-driven spacing44, 47–54, shuttling to the periphery of myofibers55–57, and anchoring at the fiber periphery and neuromuscular junctions58. Recently, actomyosin contraction was shown to be a primary driver of NE rupture in chick cardiac tissue, and this NE rupture was rescued by treatment with blebbistatin59. To address whether NE damage in myotubes was caused by actomyosin contractility, we treated Lmna KO and wild-type myotubes with nifedipine, a muscle-specific calcium channel blocker. Nifedipine treatment effectively abrogated myotube contraction, but did not reduce the frequency of chromatin protrusions and NE ruptures in Lmna KO myonuclei (Suppl. Fig. S15), indicating that actomyosin contractility is not required to induce NE damage. The differences in our results could stem from differences in contraction strength between in vitro and in vivo systems as well as from the differences between cardiac and skeletal muscle cells. Therefore, we focused our attention to cytoskeletal forces exerted on the nucleus during nuclear migration in differentiating myotubes.
Based on the timing of the onset of NE rupture (Fig. 4c) and a progressive increase in the length of the chromatin tethers with differentiation (Suppl. Fig. S16), we examined the role of microtubule driven myonuclear spreading during myofiber maturation60, 61. Time-lapse sequences of Lmna KO myoblasts expressing the NLS-GFP and/or cGAS-mCherry reporters revealed that NE rupture frequently occurred when myonuclei were moved along the length of myotubes by microtubule-associated motors (Fig. 7a, Suppl. Movie 9). Thus, we reasoned that inhibiting nuclear movement should prevent nuclear damage in Lmna KO myofibers. Supporting this hypothesis, depletion of Kif5b (Suppl. Fig 17a), a subunit of kinesin-1 that is required for nuclear migration in myotubes 60–62, nearly abolished chromatin protrusions (Suppl. Fig. S17b), NE rupture (Fig. 7b,c), and high levels of DNA damage in the Lmna KO myotubes (Suppl. Fig. 17c,d). These findings indicate that nuclear movement by microtubule-mediated forces are sufficient to cause nuclear damage in Lmna mutant myofibers that correlates with increased DNA damage in these cells.
Reducing the mechanical forces acting on Lmna KO myonuclei improve myofiber function and viability
If mechanically induced NE rupture and DNA damage are causative for the decline in viability and contractility in the Lmna KO myofibers, then reducing the cytoskeletal forces on the myonuclei should improve myofiber health. Since depletion of Kif5b has detrimental long-term effects on myofiber function63, we utilized a system where the Linker of Nucleoskeleton and Cytoskeleton (LINC) is disrupted through expression of a dominate negative GFP-KASH2 (DN-KASH) protein64 (Suppl. Fig. 17a). As a control, we generated a similar construct that contained a double alanine extension (DN-KASHext). The DN-KASHext construct still targets to the NE, but cannot disrupt the LINC complex64, as confirmed by the retention of nesprin-1 at the NE (Suppl. Fig. 18a). The DN-KASH constructs were expressed under the control of an inducible promotor that allowed for controlling the onset of LINC complex disruption following the initial fusion events of myogenesis (Suppl. Fig. S18b). This experimental approach reduced force transmission from the cytoskeleton to the nucleus and limited nuclear spreading in the DN-KASH cells, but not the DN-KASHext cells, similar to the depletion of Kif5b (Suppl. Fig. 18b). Expression of the DN-KASH constructs had no effect on myofiber contractility or viability in Lmna WT myofibers (Suppl. Fig. S18c,d; Suppl. Movies 10-13). Expression of the DN-KASH construct, but not the DN-KASHext, starting on day three of differentiation significantly reduced the number of chromatin protrusions and incidence of NE rupture in Lmna KO cells (Fig. 7d,e, Suppl. Fig. S18e). Importantly, this reduction in nuclear damage was accompanied by a reduction in the fraction of cells with severe DNA damage (Suppl. Fig. 18f) and substantially improved myofiber contractility and viability (Fig. 7f-h; Suppl. Movies 14-17). These findings further support our model that mechanical forces induce NE rupture and DNA damage in Lmna mutant myofibers, leading to myofiber dysfunction and death.
Skeletal muscle biopsies from patients with LMNA-associated muscular dystrophy show DNA damage
Finally, to corroborate the findings from the in vitro and in vivo mouse models of striated muscle laminopathies in a clinically relevant context, we examined skeletal muscle biopsy samples from humans with LMNA-related muscular dystrophies and age-matched controls (Table 1). Muscle cryosections were immunofluorescently labeled for 53BP1, an established marker for DNA double strand breaks65, 66, and myofiber nuclei were identified based on labeling for DNA, actin, and dystrophin (Fig. 8a,b, Suppl. Fig. S19). Intriguingly, tissues from individuals with the most severe forms of the muscular dystrophy, i.e., those with early childhood and juvenile onsets, had significantly increased DNA damage compared to age-matched controls (Fig. 8c-e), thus closely mirroring the findings in the most severe laminopathy mouse models (Fig. 5e).
DISCUSSION
The mechanisms by which LMNA mutations result in skeletal muscle-specific diseases, such as AD-EDMD and LMNA-CMD, has long puzzled researchers and clinicians, presenting a major hurdle in the development of effective treatment approaches. Here, we present comprehensive new evidence in support of the ‘mechanical stress’ hypothesis. By using three mouse models of striated muscle laminopathies with varying onset and severity, we systematically assessed the frequency of nuclear defects, including chromatin protrusions, NE rupture, and DNA damage, in muscle fibers in vitro and in vivo, in a uniform genetic background. Our studies revealed a striking correlation between the nuclear defects observed during in vitro myoblast differentiation, myonuclear defects in isolated muscle fibers, and disease onset and severity in the mouse model. Notably, analysis of muscle biopsy tissue from individuals with LMNA muscular dystrophies showed a similar trend, with samples corresponding to the most severe forms of muscular dystrophy having the largest fraction of myonuclei with DNA damage. The paucity of available human muscle samples and effects of genetic background in these diseases, limit the conclusions that can be made.
Our findings of NE rupture are consistent with reports of NE damage and intrusion of cytoplasmic organelles into the nucleoplasm in skeletal muscle fibers of patients with EDMD67–70, cardiac myocytes in LMNA-dilated cardiomyopathy patients71, 72, lamin A/C-deficient mice14, 73, and muscle and tendons of lamin-deficient fruit flies74, 75. Unlike those previous reports, however, we now provide detailed information on the extent, timing, and cause of the NE damage, revealing a striking correlation of NE rupture and disease severity.
Although increased DNA damage and altered DNA damage repair have been reported previously in LMNA mutant cells, those cases were linked to progeroid diseases, including Hutchinson-Gilford progeria syndrome (HGPS) and atypical Werner syndrome (AWS)39–41, 76, 77. Our findings demonstrate for the first time that cytoskeletal forces acting on myonuclei result in nuclear damage and DNA damage in skeletal muscle fibers in vitro and in vivo. The precise mechanisms by which NE damage and NE rupture could cause DNA damage and cell death remains to be elucidated. The DNA damage could arise from exposure of genomic DNA to cytoplasmic nucleases following NE rupture, or nuclear exclusion and efflux of DNA repair factors, as previously discussed in the context of confined cell migration78–80. The correlation between DNA damage and NE defects in our studies (Suppl. Fig. S11) suggests that the DNA damage is linked to mechanically induced NE defects, which is further supported by the finding that depletion of kinesin-1 and disruption of the LINC complex abolished NE defects and the most severe DNA damage in Lmna KO cells (Fig. 7b-e, Suppl. Fig. 17b-d, S18e,f). Given the known association between lamin A/C and the DNA damage response protein 53BP181, loss of lamin A/C could also impair DNA damage repair efficiency, although this effect may be limited to proliferating cells and the associated replication stress82–84. In our experiments, post-mitotic Lmna KO myofibers exposed to DNA-damaging irradiation exhibited similar DNA damage repair dynamics as wild-type cells, indicating that the observed increase in DNA damage in Lmna KO myofibers is not caused by defective DNA damage repair. Nonetheless, it is possible that the efficiency of DNA damage repair decreases over the course of the differentiation, as the highest levels of DNA damage were found in myofibers at late stages of maturation. Such a decrease in DNA repair efficiency is consistent with work showing that satellite cells repair radiation-induced DNA double strand breaks more efficiently than their differentiated counterparts85, and reduced DNA damage repair could amplify the progressive DNA damage in differentiating myoblasts. In wild-type myofibers, repeated exposure to DNA damaging agents, when combined with DNA damage inhibition, was sufficient to induce cell death to the same extent as observed in untreated Lmna KO cells (Fig. 5g-i), demonstrating that accumulating DNA damage is sufficient to induce cell death even in post-mitotic cells such as myofibers.
The role of DNA damage response signaling in post-mitotic muscle function is an area of increasing interest. DNA damage results in rapid activation of DNA damage response pathways, including DNA-PK and ATM, which results in stabilization of p53, one of the primary DNA damage response pathway that can induce cell cycle arrest, senescence, and apoptosis.86, 87 The consequences of increased DNA damage response signaling in post-mitotic cells remain poorly characterized, but recent findings point to an intriguing role of DNA damage response pathways in skeletal and cardiac muscle. Increased activity of DNA-PK, one of the major DNA damage sensing pathways in interphase cells36, was recently linked to the age-related decline of metabolic, mitochondrial, and physical fitness of skeletal muscle cells34. Furthermore, cardiac-specific expression of Lmna D300N results in increased DNA damage and activation of p53 in a laminopathy mouse model, and cardiac specific deletion of the Trp53 gene encoding p53 significantly improved the cardiac defects, although only marginally improved overall survival77. Consistent with increased p53 signaling caused by accumulating DNA damage, we found evidence of caspase-3 activation and a progressive loss in viability in Lmna KO and N195K myofibers (Fig. 1e,f). While the mechanisms controlling apoptosis in a post-mitotic tissue such as skeletal muscle are not well understood and remain controversial88, this type of cell death has been observed in myofibers during other muscle wasting conditions89, in Lmna E82K mutant 90 and heterozygous Lmna+/– mouse hearts91, and recently in a cardiac-specific Lmna D300N mouse model77.
Although Lmna mutant myoblasts have mechanically weaker nuclei than wild-type myoblasts, we found that NE damage only arose during myoblast differentiation, when cytoskeletal forces acting on the myonuclei increase. Surprisingly, nuclear damage during in vitro myofiber differentiation was associated with kinesin-1 mediated nuclear movements, independent of actomyosin contractions (Fig. 7b,c, Suppl. Fig. 15, Suppl. Fig. 17b-d). Kinesin-1 applies localized point forces at the NE of skeletal myonuclei, either directly or through microtubules anchored at the NE through the LINC complex, to ensure correct nuclear positioning61. These forces acting on the weakened NE are likely sufficient to induce NE rupture in the Lmna mutant myonuclei, based on recent studies on lamin A/C-depleted cells subjected to precisely controlled tensile forces92. Furthermore, mutations in genes encoding lamins A/C and other NE proteins linked to muscular dystrophies could disrupt perinuclear cytoskeletal organization, including that of desmins73 and the perinuclear microtubule network, which helps to resist cytoplasmic strain and physically protect myonuclei, thereby further promoting nuclear damage in myofibers. At the same time, actomyosin contractility is likely to contribute to NE damage in muscle fibers in vivo, which can generate substantially higher forces than in vitro differentiated myofibers, and to NE rupture and DNA damage in cardiac myocytes59.
While we cannot exclude the possibility that altered cytoplasmic signaling and gene regulation pathways contribute to the increased NE rupture and DNA damage in Lmna mutant muscle cells77, 93, our data suggest that the damage is mechanically induced and nucleus-intrinsic, as depletion of Kif5b and disruption of the LINC complex substantially reduced DNA damage in Lmna KO myofibers (Figs. 3e,f, 4f,g, Suppl. Fig S18e,f). Our findings support a model in which cytoskeletal forces cause chromatin protrusions and NE ruptures in mechanically weakened Lmna mutant muscle cell nuclei, triggering DNA damage, which then leads to myofiber dysfunction and death in striated muscle laminopathies (Supp. Fig. 20). Given the diverse roles of lamins in cellular function, additional mechanisms may further contribute to the pathogenesis of laminopathies, and their specific contribution may depend on the particular mutation and cellular context.
Our experimental data provides novel mechanistic insights into the cellular processes that contribute to the development of striated muscle laminopathies, thereby informing future research efforts to target. Microtubule stabilization is one approach to correct the perturbed force balance and should be further explored as a therapy for striated muscle laminopathies. Paclitaxel was recently reported to improve cardiac conduction defects in Lmna H222P mice by restoring proper connexin 43 localization94. Here, we highlight an additional mechanism by which microtubule stabilization may mitigate damaging forces in striated muscle laminopathies. In addition, our work indicates that DNA damage is increased in mouse models of laminopathies, as well as human patients, and that targeting cell signaling pathways activated by DNA damage, such as DNA-PK34 and p5377 may provide a mechanism to ameliorate muscle wasting. Furthermore, our studies found that hybrid myofibers containing ∼20% Lmna KO and ∼80% wild-type nuclei were indistinguishable from their isogenic wild-type controls, suggesting that delivery of wild-type lamin A/C to a subset of myonuclei by gene delivery or stem cell therapy may be sufficient to rescue myofiber function. Further studies will need to identify the critical number of wild-type nuclei required to rescue cellular health in laminopathic tissue for therapeutic applications.
Beyond striated muscle laminopathies, insights gained from this work are highly relevant to other biological systems in which nuclei are exposed to physical stress from the cytoskeleton, such as in confined migration of cancer cells95, or intracellular nuclear positioning of polarized epithelial, or neuronal cells96. For example, in neuronal cells lacking functional lamin A97, defects in B-type lamins98 or other NE proteins (e.g., nesprin-1) may make these cells more susceptible to kinesin-mediated nuclear damage leading to neurodevelopmental defects (e.g., cerebellar ataxia). Taken together, these findings highlight a novel mechanism by which weakened myonuclei experience microtubule-mediated NE damage, leading to DNA damage and muscle dysfunction, potentially explaining the phenotypes seen in striated muscle laminopathies and a spectrum of other diseases caused by NE defects.
AUTHOR CONTRIBUTIONS
A.E., T.K., G.F., P.I., L.W., and J.L contributed to the conception and design of the work; A.E., T.K., G.F., P.I., J.P., S.I., S.M. contributed to data acquisition and analysis; A.E., T.K., G.F., L.W., G.B., J.L. contributed to the interpretation of data; A.E., T.K., G.F., J.L contributed to drafting the manuscript. All authors contributed to editing the manuscript.
METHODS
Animals
Lmna KO (Lmna–/–)14, Lmna H222P (LmnaH222P/H222P)16, and Lmna N195K (LmnaN195K/N195K)15 have been described previously. Lmna+/–, LmnaH222P/+, and LmnaN195K/+ mice were backcrossed at least seven generations into a C57-BL/6 line. For each mouse model, heterozygous mice were crossed to obtain homozygous mutants, heterozygous mice, and wild-type littermates. Lmna mutant mice were provided with gel diet (Nutri-Gel Diet, BioServe) supplement to improve hydration and metabolism upon onset of phenotypic decline. Dmdmdx mice have been described previously99; mice were obtained from the Jackson Laboratory in a C57BL background and hemi- or homozygous animals were bred to produce all hemi- and homozygous offspring. All mice were bred, maintained and euthanized according to IACUC approved protocols. Data from wild-type littermate controls for Lmna KO, Lmna N195K and Lmna H222P showed no difference in any of the experimental outcomes between the different wild-type littermates, so wild-type data was combined into a single group unless otherwise specified. For both in vivo and in vitro studies, cells and or tissues were isolated from a single mouse and counted as a single replicate. All data are based on at least two independently derived primary cell lines for each genotype.
NE rupture reporter mouse (cGAS/MB21D1-tdTom transgenic mouse)
To detect NE ruptures in vivo, we generated a transgenic mouse expressing FLAG tagged human cGASE225A/D227A fused to a tdTomato fluorescent protein (cGAS-tdTomato) under the control of the commonly used constitutive active CMV promoter. The cGAS mutations are in the magnesium-binding domain, abolishing the enzymatic activity and downstream production of interferon, while still remaining the ability to bind to genomic DNA. The mammalian expression cassette including promoter and terminator (CMV-3xFLAG-cGASE225A/D227A-tdTomato-SV40polyA) was released from the expression vector, removing the prokaryotic domains. The purified linear DNA was then injected into the pronucleus of fertilized embryos collected from super-ovulated C57BL/6 mice and transplanted into pseudo-pregnant recipients. The resulting transgenic mouse model was used to cross into the Lmna KO background to generate 3×FLAG-cGASE225A/D227A-tdTomato positive Lmna KO mice within two generations.
Myoblast isolation
Cells were harvested from Lmna KO, Lmna N195K, Lmna H222P, and wild-type littermates between 3-5 weeks for Lmna KO mice, 4-6 weeks for Lmna N195K, and 4-10 weeks for Lmna H222P mice using a protocol adapted from14. With the exception of the Lmna KO myoblasts, these time-points were prior to the onset of disease phenotypes. Myoblasts from wild-type littermates were harvested at the same time. Muscles of the lower hindlimb were isolated, cleaned of fat, nerve and excess fascia, and kept in HBSS on ice until all mice were harvested. The muscles were digested in 4 ml:1 g of tissue wet weight in a solution of 0.5% Collagenase II (Worthington Biochemicals), 1.2 U/ml Dispase (Worthington Biochemicals), 1.25 mM CaCl2 (Sigma) in HBSS/25 mM HEPES buffer. Digestion was carried out in a 37°C water bath for a total time of 60 minutes. At 20 minutes intervals, digestion cocktails were removed and triturated 40 times with a 5 ml pipet. In the case of difficult to digest tissues, an extra 25% of 1% Collagenase II was added to the digestion after 40 minutes.
When tissues were fully digested, the reaction was quenched using equal volumes of DMEM supplemented with 10% fetal bovine serum (FBS) and 1% P/S (D10 media, Gibco). The cell suspension was strained through 70 and 40 μm filters (Greiner Bioscience) sequentially to remove undigested myotube fragments and tendon. The cell suspension was centrifuged at 800 × g for 5 minutes and washed with 8 ml of D10 media for a total of four times. Cells were then resuspended in primary myoblast growth media (PMGM; Hams F-10 (Gibco) supplemented with 20% horse serum and 1% penicillin/streptomycin and 1 µl/ml basic fibroblast growth factor (GoldBio)) and plated onto a 2% gelatin coated T25 flask. Cells were allowed to sit undisturbed for 72 hours. Capitalizing on the fact that myoblasts adhere much more weakly than fibroblasts, cells were passaged using PBS (calcium- and magnesium-free) instead of trypsin to purify the myoblasts. Cells were washed for 2-3 minutes at room temperature using a volume of PBS sufficient to coat the bottom of the flask and dislodged using manual agitation. When necessary, a 0.000625% trypsin solution was used to aid in the myoblast removal. Myoblasts were re-suspended in PMGM and re-plated onto gelatin coated flasks. This process was continued 3-4 times until pure myoblast cultures were achieved100. Cells were maintained in culture on gelatin coated flasks with media changes every other day. All experiments were carried out prior to passage 12. Each independent experiment was done on a different set of lamin mutant and wild-type littermates such that each independent experiment was sourced from a different animal to account for heterogeneity in phenotype.
Myoblast differentiation
Myoblasts were differentiated according to a protocol modified from17. Coverslips for differentiation were prepared by first coating with CellTak (Corning) according to the manufacturer’s protocol and then coating with growth factor reduced Matrigel (Corning) diluted 1:100 with IMDM with Glutamax (Gibco). Pre-cooled pipette tips were used to avoid premature polymerization. Matrigel was allowed to polymerize at 37°C for 1 hour and the excess solution was aspirated. Primary myoblasts were seeded at a density of 55,000 cells/cm2 in PMGM. Cells were allowed to attach for 24 hours before being switched to primary myoblast differentiation media (PMDM) composed of IMDM with Glutamax and 2% horse serum without antibiotics. This timepoint was considered day 0. One day after the onset of differentiation, a top coat of 1:3 Matrigel:IMDM was added to the cells and allowed to incubate for 1 hour at 37°C. PMDM supplemented with 100 ng/ml agrin (R&D Systems) was added to the cells and henceforth replaced every second day. Cells were allowed to differentiate for a total of 0, 5, or 10 days.
Plasmids and generation of fluorescently labeled cell lines
Each of the mutant myoblast lines were stably modified with lentiviral vectors to express the nuclear rupture reporter NLS-GFP (pCDH-CMV-NLS-copGFP-EF1-blastiS) and cGAS-mCherry (pCDH-CMV-cGASE225A/D227A-mCherry2-EF1-blastiS). cGAS is a cytosolic DNA binding protein; we used a cGAS mutant (E225A/D227A) with abolished enzyme activity and interferon production, but that still binds DNA101 and serve as a NE rupture reporter13. For rescue experiments, Lmna KO cells were modified with human lamin A (pCDH-CMV-preLamin A-IRES-GFP-puro). To generate the DN-KASH and DN-KASHext constructs, GFP-KASH2 and GFP-KASH2ext were subcloned from previously published plasmids64 and inserted into an all-in-one doxycycline inducible backbone (pPB-tetO-GFP-KASH2-EIF1α-rtTA-IRES-Neo)102.
Viral and Piggybac modification
Pseudoviral particles were produced as described previously13. In brief, 293-TN cells (System Biosciences, SBI) were co-transfected with the lentiviral-containing, packaging and envelope plasmids using PureFection (SBI), following manufactures protocol. Lentivirus-containing supernatants were collected at 48 hours and 72 hours after transfection, and filtered through a 0.45 µm filter. Cells to be transduced were seeded into 6-well plates so that they reached 50-60% confluency on the day of infection and transduced at most 2 consecutive days with the viral supernatant using the TransDux Max system (SBI). The viral solution was replaced with fresh culture medium, and cells were cultured for 72 hours before selection with 1 μg/mL of puromycin or 2 μg/mL blasticidin S for 2-5 days. After selection, cells were subcultured and maintained in their recommended medium without the continued use of selection agents. For PiggyBac modifications, myoblasts were transfected with 1.75 μg of the PiggyBac plasmid and 0.75 μg of a Hyperactive Transposase using the Lipofectamine 3000 reagent according to the manufacture’s guidelines.
Extended imaging using incubator microscope
Long term imaging was performed using an Incucyte imaging system, which allows for incubator imaging to minimize the effects of humidity and CO2 changes. The differentiating cells expressing combinations of NLS-GFP and cGAS-mCherry were imaged using the Incucyte dual color filter module from day 0 to day 10, every 30-60 minutes with a 20× objective. Resulting images were analyzed using the Incucyte software, which performs fluorescence background subtraction using a top hat method and then subsequent thresholding. cGAS-mCherry cells were thresholded and then analyzed for increase in fluorescent foci over time to track the rate of increase in NE rupture or damage. NLS-GFP cells were used to investigate the frequency and presence of NE rupture. To verify the results obtained from the Incucyte, cells were fixed and stained with appropriate antibodies to evaluate DNA damage and NE rupture.
Isolation of single muscle fibers
Single muscle fibers were harvested in a protocol adapted from Vogler et al.103. As previously described, fibers were isolated from the extensor digitorum longus (EDL) of male and female Lmna KO and wild-type litter mates at 5-6 weeks of age and Lmna H222P and wild-type litter mates were harvested at 6-8 weeks of age at 23-25 weeks of age in order to compare pre- and post-phenotype onset tissue14, 16, 104. Briefly, the EDL (extensor digitorus longus) and plantaris were isolated from the mouse and placed directly into a 1 ml solution of F10 media with 4,000 U/ml of Collagenase I (Worthington Biochemicals). The tissue was digested for 15-40 minutes depending on muscle size in a 37°C water bath with agitation by inversion every 10 minutes. The reaction was quenched by transferring the digestion mixture to 4 ml of PMGM. Single fibers were hand-picked from the digested tissue using fire polished glass Pasteur pipettes. When necessary, the tissue was further dissociated by manual pipetting with a large-bore glass pipet. Fibers were washed once in fresh media prior to fixation with 4% paraformaldehyde (PFA) for 15 minutes at room temperature and subsequent IF staining.
Pharmacological treatments
For preliminary experiments, myoblasts were differentiated using the standard protocol and treated with pharmacological treatments starting at day 5 of differentiation. For chromatin protrusion studies, paclitaxel was administered to differentiated myotubes in two 24 hours bursts at day 4 and day 6-post differentiation with a 24 hour recovery in between. Myotubes were then fixed in 4% PFA at day 7 and stained with anti-lamin B and DAPI in order to quantify the percentage of myonuclei with chromatin protrusions. For long term studies using the cGAS reporter, myotubes were treated with 10 nM of paclitaxel starting at day 5 and then media was refreshed every day. To inhibit myotube contraction, cells were treated with 5 µM nifedipine starting at day 5 and then media was refreshed every day. For DNA damage induction and inhibitor experiments, cells were treated with 20µg/ml of phleomycin for a two-hour pulse on Day 3,4, and 5 of differentiation. Concurrently, cells were treated with NU7441 (2 μM), KU55933 (2 μM) starting at day 2 of differentiation through day 10 of differentiation. For induction of DN-KASH and DN-KASHext, cells were treated with 1 μM doxycycline.
Biophysical assays
To evaluate nuclear deformability in high throughput, we designed and fabricated a microfluidic, micropipette aspiration device. The mask and wafers were produced in the Cornell NanoScale Science and Technology Facility (CNF) using standard lithography techniques. PDMS molds of the devices were cast using Sylgard 184 (Dow Corning) and mounted on coverslips using a plasma cleaner as described previously13. Three port entrances were made using a 1.2 mm biopsy punch. Pressures at the inlet and outlet ports were set to 1.0 and 0.2 psi (relative to atmospheric pressure, Patm), respectively, using compressed air regulated by a MCFS-EZ pressure controller (Fluigent) to drive single cells through the device. Myoblasts (∼5×106 cells/mL suspended in 2 % bovine serum albumin (BSA), 0.2 % FBS and 10 μg/mL Hoechst 33342 DNA stain in PBS) were captured within an array of 18 pockets, and then forced to deform into 3 µm wide × 5 µm tall micropipettes. The selected pressures resulted in detectable nuclear deformations without causing significant damage to the cells (tested using propidium iodide staining). The remaining port was set to Patm and outfitted with a handheld pipette to flush cells from the pockets at the start of each image acquisition sequence. Brightfield and fluorescence images were acquired every 5 seconds for a minimum of 60 seconds using an inverted microscope and 20×/NA 0.8 air objective. Nuclear protrusion length was calculated using a custom-written MATLAB program, made available upon request.
For the microharpoon studies, myoblasts were seeded in 35 mm glass bottom dishes and differentiated as previously described, except without the addition of a Matrigel top coat to allow microharpoon access. A Sutter P-97 micropipette puller was used to create microharpoons from borosilicate glass rods (Sutter; OD: 1.0 mm, ID: 0.78, 10 cm length) with tip diameters of ≈1 μm. Day 4 myotubes (Lmna KO and wild-type) were treated for 24 hours with either 50 nM Paclitaxel or the corresponding 0.1% DMSO. The following day, the microharpoon assay was performed as previously described by our laboratory22, with slight modifications to the pull parameters to accommodate myotubes. The microharpoon was inserted ≈5-7 μm from the edge of the nucleus and pulled 15 μm at a rate of 1 μm/s. Pull direction was always orthogonal to the long axis of the myofiber. Images were acquired at 40× (+1.6×) every 5 seconds. Nuclear strain and centroid displacement were calculated using a custom-written MATLAB program, made available upon request.
siRNA treatment
siRNAs used were as follows: Kif5b#3 (target sequence 5′-CAGCAAGAAGTAGACCGGATA-3′; Qiagen SI00176050), Kif5b#4 (target sequence 5′-CACGAGCTCACGGTTATGCAA-3′; Qiagen SI00176057), and non-target (NT) negative control (ON-TARGETplus non-targeting pool, Dharmachon, D-001810-10). Myoblasts were seeded at a density of ∼15,000 cells per well in a 96-well glass bottomed dish containing a matrigel coating. Once adhered, the myoblasts were transfected twice, 48 hours apart, with siRNA for NT or Kif5b using Lipofectamine RNAiMAX at a concentration of 150 nM in PMGM. After 12 hours, the myoblasts were switched to PMDM and differentiated for 5 days.
Immunofluorescence staining of mouse cells and tissues
Cells were fixed in pre-warmed 4% PFA at the appropriate time point(s) and washed with PBS. Cells were blocked and permeabilized with a solution of 3% BSA, 0.1% Triton-X 100 and 0.1% Tween (Sigma) for 1 hour at room temperature. Cells were stained with primary antibodies diluted in blocking solution according to Table 2 at 4°C overnight. Samples were washed with PBS and incubated for 1 hour at room temperature with 1:250 dilution of AlexaFluor antibodies (Invitrogen) and 1:1000 DAPI (Sigma). Single muscle fibers were stained using the same procedure in Eppendorf tube baskets with an increase in blocking solution Triton-X concentration to 0.25%.
Human patient biopsy staining
Following diagnostic testing, muscle biopsies were stored at - 80° C and subsequently utilized for research following protocols approved by the corresponding IRB, with informed consent from all participants. Cryopreserved human quadriceps muscle biopsy tissue from LMNA muscular dystrophy individuals and age-matched controls were used for immunostaining as described 105. An anti-rabbit polyclonal 53BP1 antibody (Novus) and anti-dystrophin mouse monocolonal antibody (Mab7A10, U of Iowa Hospitals and Clinics Pathology Core) were used at 1:1000 and 1:20 dilutions, respectively. Texas Red labeled phalloidin (Invitrogen) was used at 1:400 dilution. Secondary antibodies were a goat anti-rabbit Ig Alexa 488 conjugate (Invitrogen) and a goat anti-mouse IgG rhodamine Red-X conjugated (Molecular Probes), both used at 1:500 dilution. Slides were imaged on a Zeiss 710 confocal microscope (University of Iowa Central Microscopy Facility). The intensity of nuclear anti-53BP1 staining was quantified using ImageJ. Analysis of the human muscle samples was performed in a double-blinded manner. A pathologist generated the 10 µm thick cryo-sections and coded the samples, which were stained, imaged, and quantified prior to decoding by an independent individual.
Western analysis
Cells were lysed in RIPA buffer containing protease (cOmplete EDTA-Free, Roche) and phosphatase (PhosSTOP, Roche) inhibitors. Protein was quantified using Bio-Rad Protein Assay Dye and 25-30 µg of protein lysate was separated using a 4-12% Bis-Tris polyacrylamide gel using standard a standard SDS-Page protocol. Protein was transferred to a polyvinylidene fluoride (PVDF) membrane overnight at 4°C at a current of 40 mA. Membranes were blocked using 3% BSA in tris-buffered saline containing 0.1% Tween-20 and primary antibodies (Table 2) were diluted in the same blocking solution and incubated overnight at 4°C. Protein bands were detected using either IRDye 680LT or IRDye 800CW (LI-COR) secondary antibodies, imaged on an Odyssey® CLx imaging system (LI-COR) and analyzed in Image Studio Lite (LI-COR)
Imaging acquisition
Cells on coverslips and mounted single muscle fibers were imaged with an inverted Zeiss LSM700 confocal microscope. Z-stack were collected using 20× air (NA = 0.8), 40× water-immersion (NA = 1.2) and 63× oil-immersion (NA = 1.4) objectives. Airy units for all images were set between 1 and 1.5. Epi-fluorescence images were collected on a motorized inverted Zeiss Observer Z1 microscope equipped with CCD cameras (Photometrics CoolSNAP EZ or Photometrics CoolSNAP KINO) or a sCMOS camera (Hamamatsu Flash 4.0). H&E histology images were collected on an inverted Zeiss Observer Z1 microscope equipped with a color CCD camera (Edmund Optics, EO-0312C).
Image analysis
Image sequences were analyzed using ZEN (Zeiss), ImageJ, or MATLAB (Mathworks) using only linear adjustments uniformly applied to the entire image region. Region of interest intensities were extracted using ZEN or ImageJ. To quantify cleaved caspase-3 (i.e. active) area and myofiber health, maximum intensity protections were generated, which were then blinded to the observer. Cleaved caspase-3 area was calculated by thresholding of the caspase-3 and myosin heavy chain fluorescent signal and expressing the cleaved caspase-3 signal relative to the myosin heavy chain signal. Myofiber contractions were scored based on a minimum of 6 random fields of view per replicate using a blinded analysis according to the scales provided in Suppl. Fig. S3. To count the number of DNA protrusions, and DNA damage foci, confocal image stacks were three-dimensionally reconstructed and displayed as maximum intensity projections. Protrusions lengths were both counted and measured by the presence of DAPI signal beyond the lamin B rim of the nucleus. Aspect ratio was quantified based on a thresholded lamin B rim to avoid the confounding factor of the DNA protrusions outside the body of the nucleus. Nuclear rupture was detected by an increase of the cytoplasmic NLS-GFP signal, or the localization of cGAS-mCherry to the nucleus. For better visualization of NLS-GFP cells many of the fluorescent single color image sequences were inverted. Graphs were generated in Excel (Microsoft), and figures were assembled in Illustrator (Adobe). DNA damage was determined by counting H2AX foci and then binned based on foci number. If damage was so severe that individual foci could not be counted, these nuclei were placed in the >25 foci category. For Hsp90 quantification, average nuclear Hsp90 fluorescence intensity was determined from a single mid-nucleus z-plane image and normalized to the cytoplasmic intensity at two points immediately adjacent to the nucleus.
MTT assay
Myoblasts, seeded in a 96-well plate and differentiated as previously described for 0, 5, or 10 days, were assayed for cell viability according to the manufacturer’s instructions (Promega, CellTiter 96 Non-Radioactive Cell Proliferation Assay). Fresh differentiation media was added two hours prior to the addition of 15 μL MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide dye. After incubation for 3 hours in MTT dye, 100 uL of Stop Solution was added to solubilize the formazan product (appears purple). Following overnight incubation at 37°C and 5% CO2, the absorbance of each well (measured at 590 nm) was analyzed using a microplate reader.
Gamma irradiation
A pulse of gamma-irradiation (5 Gy) was administered to myotubes differentiating (5 days) in a 96 well plate. Non-irradiated controls, along with treated cells following 3, 6, or 24 hours of recovery, were PFA-fixed and stained with anti-γH2AX, anti-lamin B and DAPI. A custom macro was used to quantify the mean integrated density of nuclear γH2AX signal from maximum intensity projections of confocal z-stacks.
Statistical analysis
Unless otherwise noted, all experimental results were taken from at least three independent experiments and in vivo data were taken from at least three animals. For data with normal distribution, we used either student’s t-tests (comparing two groups) or one-way ANOVA (for experiments with more than two groups) with post-hoc tests. When multiple comparisons were made, we adjusted the significance level using Bonferroni corrections. All tests were performed using GraphPad Prism. Micropipette aspiration data were natural log-transformed (Suppl. Fig. S4a) and analyzed by linear regression of the log-log data. In addition, data was analyzed with a multilevel model, in which the log-transformed protrusion length was the dependent variable in the model and the log-transformed time, genotype, and their interaction were treated as independent fixed effects. Variance from individual experiments and other effects were considered in the model as random effects. Post-hoc multiple comparisons test with Dunnett correction were performed to determine differences between Lmna mutant cells (Lmna KO, Lmna N195K, and Lmna H222P) and control cells (pooled wild-type). Analyses were carried out using JMP software. For human tissue samples, nuclei were binned into eleven categories based on their intensity of 53BP1 staining as measured with ImageJ (arbitrary units). The relative percent intensities of each bin were plotted as a histogram for each genotype. Since the intensity values were not in a normal distribution, the Kruskal-Wallis one-way analysis of variance (ANOVA), followed by the Dunn post hoc test, were used to determine if there was a significant difference in staining among the genotypes. The Mann-Whitney non-parametric test was used for comparisons that involved one sample and one age-matched control. Unless otherwise noted, * denotes p ≤ 0.05, ** denotes p ≤ 0.01, and *** denotes p ≤ 0.001. Unless otherwise indicated, error bars represent the standard error of the mean (SEM). The data that support the findings of this study are available from the corresponding author upon reasonable request.
Data and code availability
The data that support the findings of this study are available from the corresponding authors upon reasonable request. MATLAB codes used for the microharpoon assay and micropipette aspiration analysis is available upon request.
ACKNOWLEDGEMENTS
The authors thank Colin Stewart for providing the Lmna KO and Lmna N195K mouse models, Edgar Gomes for help with the in vitro myocyte differentiation protocol, Alexandra Corbin for in vivo and in vitro protrusion analysis and in vivo Hsp90 localization analysis, Rebecca Mount for optimization of single fiber isolation and subsequent analysis, Daniel Huang for quantification of the Lmna N195K skeletal muscle fiber cross-sectional areas, Ern Hwei Hannah Fong for quantification of contraction scores, Francoise Vermeylen, Stephen Parry and Lynn Johnson from the Cornell Statistical Consulting Unit, and Katherine Strednak and the Cornell Center for Animal Resources and Education (CARE) for help in maintaining the Lmna mutant mice. Clinical data from LMNA individuals was provided by Katherine D. Mathews, M.D. (Vice Chair for Clinical Investigation, Director, Muscular Dystrophy Clinic, Director, Iowa Neuromuscular Program, Professor of Pediatrics – Neurology and Professor of Neurology). Technical assistance with human muscle immunohistochemistry was provided by Nicholas M. Shaw and Margaret R. Ketterer (U. Iowa). This work was supported by awards from the National Institutes of Health [R01 HL082792 and U54 CA210184], the Department of Defense Breast Cancer Research Program [Breakthrough Award BC150580], the National Science Foundation [CAREER Award CBET-1254846 and MCB-1715606], the Muscular Dystrophy Association [Development Award MDA603238 to T.J.K.] as well as generous gifts from the Mills family, a Fleming Postdoctoral Fellowship to T.J.K, and National Science Foundation Graduate Research Fellowships to A.J.E. [2013160437] and G.R.F. [2014163403]. Financial support from the Muscular Dystrophy Association (477283 to L.L.W); Burroughs Welcome Fund Collaborative Research Travel Grant (#1017502); University of Iowa Wellstone Muscular Dystrophy Cooperative Research Center (U54, NS053672 to S.A.M). This work was performed in part at the Cornell NanoScale Facility, a member of the National Nanotechnology Infrastructure Network, which is supported by the National Science Foundation [Grant ECCS-0335765].
Footnotes
The manuscript has been revised to include additional data and mechanistic insights. Several figures have been updated and revised to accommodate the new findings.
REFERENCES
- 1.
- 2.↵
- 3.
- 4.
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.
- 31.
- 32.
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.
- 49.
- 50.
- 51.
- 52.
- 53.
- 54.↵
- 55.↵
- 56.
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.
- 69.
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.
- 80.↵
- 81.↵
- 82.↵
- 83.
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵