Microtubule forces drive nuclear damage in LMNA cardiomyopathy

Abstract

Nuclear homeostasis requires balanced forces between the cytoskeleton and the nucleus. Mutations in LMNA, which encodes lamin A/C, weaken the nuclear lamina, leading to nuclear damage and muscle disease. Disrupting the linker of nucleoskeleton and cytoskeleton (LINC) complex, which connects the cytoskeleton to the nucleus, may ameliorate LMNA-associated cardiomyopathy, yet the cardioprotective mechanism remains unclear. Here we developed an assay to quantify the coupling between cardiomyocyte contraction and nuclear deformation and interrogate its dependence on the nuclear lamina and LINC complex. The LINC complex was mostly dispensable for transferring contractile strain to the nucleus, and its disruption did not rescue elevated nuclear strain in lamin A/C-deficient cardiomyocytes. Instead, LINC complex disruption eliminated the microtubule cage encircling the nucleus. Microtubule disruption prevented nuclear damage and preserved cardiac function in lamin A/C deficiency. Computational modeling revealed that microtubule forces create local stress concentrations that damage lamin A/C-deficient nuclei. These findings identify microtubule-dependent force transmission as a pathological driver and therapeutic target for LMNA cardiomyopathy.

Fig. 1. Active sarcomere–nuclear strain coupling in beating cardiomyocytes.

a, Representative live isolated adult rat cardiomyocyte stained with SiR-actin and Hoechst 33342 to visualize sarcomeres and DNA, respectively (top). Cardiomyocytes were stimulated at 1 Hz, and the nucleus was imaged at 90 fps to follow nuclear deformation during the contraction cycle (bottom: representative time-lapse images; Supplementary Video 1). Nuc., nuclear. b, Sarcomere length, nuclear length and nuclear width recordings over time for a single contraction cycle. c, Sarcomere strain, nuclear length strain and nuclear width strain over time. d, Quantification of peak sarcomere and nuclear compression. e,f, Sarcomere–nuclear coupling represented with a plot of nuclear length versus sarcomere length (e) and respective nucleus strain versus sarcomere strain (f). The latter, dimensionless strain coupling map depicts the dampened strain on the nucleus during systolic compression and diastolic re-lengthening, as marked by the deviation from a linear correlation (dotted line). g, Sarcomere–nuclear strain dampening during systole quantified from area above linear correlation (left schematic). Diastolic dampening is quantified from area under end-systolic linear correlation (right schematic). Data are presented as mean ± s.e. for 20 cells from a single adult rat heart to illustrate the approach used throughout this study. Statistical significance was determined by two-tailed t-test.

Fig. 2. Distinct effects of LINC complex disruption and MT depolymerization on resting nuclear morphology and active sarcomere–nuclear strain coupling.

a, Schematic of cytoskeletal-to-nucleoskeletal connections and the experimental perturbations. b, Live three-dimensional super-resolution imaging of resting adult rat cardiomyocyte nuclei (mid-nuclear planes are displayed). c,d, Nuclear volume (c) and aspect ratio (d) measurements after AdV DN-KASH or colchicine treatment. For c: AdV empty and AdV DN-KASH (48 hours): N = 3, n = 30, DMSO and colchicine (24 hours): N = 3, n = 36. For d: AdV empty and AdV DN-KASH (48 hours): N = 3, n = 51, DMSO and colchicine (24 hours): N = 3, n = 51. e,f, Active two-dimensional imaging of electrically stimulated cardiomyocytes. Sarcomere strain (top) and nuclear strain (middle) over time and sarcomere–nuclear strain coupling (bottom) for AdV DN-KASH (e) and colchicine (f) compared with their respective controls. g, Snapshots of live, WT cardiomyocyte labeled with SPY-555 tubulin and Hoechst 33342 during diastole and peak systole demonstrating MT cage buckling (arrow) during contraction. h, Quantification of sarcomere–nuclear strain dampening during the systolic and diastolic phases, for the indicated perturbations (see schematics in Fig. 1g). i, Integrated nuclear strain over time during the contractile cycle for the indicated perturbations. AdV empty and AdV DN-KASH (48 hours): N = 3, n = 51, DMSO and colchicine (24 hours): N = 3, n = 51. Data are presented as mean ± s.e. on individual cells. For c, d, h and i, individual cells are indicated with open circles, and individual animal replicate means are indicated with closed triangles connected by lines between experimental groups. Statistical significance was determined by two-tailed t-test. Colch, colchicine. Panel a created with BioRender.com.

Fig. 3. Cardiac-specific LINC complex disruption extends lifespan, improves cardiac function and protects against nuclear ruptures in Lmna N195K cardiomyopathy.

a, Mouse models and experimental timelines used in this study. b, Kaplan–Meier survival plot. N = 10 mice per genotype. c, Left ventricular ejection fraction measured by echocardiography at 10 weeks of age. Closed circles denote females; open circles denote males. WT: N = 9, csDN-KASH: N = 15, Lmna N195K: N = 9, Lmna N195K csDN-KASH: N = 13. Error bar represents mean ± s.d. Statistical significance was determined by one-way ANOVA with Tukeyʼs multiple correction. d, Representative Picrosirius red–stained heart sections and quantification of % fibrotic area. N = 3 animal replicates per group. Error bar represents mean ± s.d. Statistical significance was determined by one-way ANOVA with Tukeyʼs multiple correction. e, Representative images of cardiomyocyte nuclear morphology at 8–9 weeks of age. f, Quantification of nuclear length, width and aspect ratio. WT: N = 4, n = 58. csDN-KASH: N = 4, n = 69. Lmna N195K: N = 4, n = 55. Lmna N195K csDN-KASH: N = 4, n = 59. Data are presented as mean ± s.e. of individual nuclei (open circles); individual animal replicate means are indicated with closed triangles. g, Chromatin protrusion from the nucleus in a Lmna N195K cardiomyocyte. Representative mid-plane images of a normal Lmna N195K nucleus (top) and a nucleus with a chromatin protrusion (bottom) in Lmna N195K cardiomyocytes. h, Quantification of the percentage of nuclei with chromatin protrusions in each genotype by three independent, blinded scorers. Open circles represent the percentage of nuclei with chromatin protrusions quantified by each independent user for each animal; closed circles correspond to the mean for each animal. Mean ± s.e. is shown for each genotype. WT: N = 5, n = 223. csDN-KASH: N = 3, n = 75. Lmna N195K: N = 5, n = 287. Lmna N195K csDN-KASH: N = 3, n = 223. Statistical significance was determined by one-way ANOVA with Bonferroni correction. Panel a created with BioRender.com.

Fig. 4. Increased active nuclear strain in Lmna cardiomyopathy is not restored by cardiac-specific in vivo LINC complex disruption.

a, Representative snapshots of Lmna N195K cardiomyocyte nuclei during diastole and peak systole. b, Increased nuclear compression as evidenced by a downward shift in the laminopathy strain coupling curve. c,d, Representative snapshots of cardiac-specific LINC complex disruption in WT cardiomyocyte nuclei during diastole and peak systole (c) with no change in active strain coupling (d). e,f, Representative snapshots of cardiac-specific LINC complex disruption in Lmna N195K cardiomyocyte nuclei during diastole and peak systole (e) with no change in active strain coupling (f). Data are presented as mean ± s.e. g, Quantification of sarcomere–nuclear strain dampening during the systolic and diastolic phases, for the indicated groups (see schematics in Fig. 1g). h, Integrated nuclear strain over time during the contractile cycle for the indicated groups. WT: N = 4, n = 58. Lmna N195K: N = 4, n = 69. WT veh: N = 4, n = 81. csDN-KASH: N = 4, n = 69. Lmna N195K veh: N = 4, n = 55. Lmna N195K csDN-KASH: N = 4, n = 59. Data are presented as mean ± s.e. of individual cells (open circles); individual animal replicate means are indicated with closed triangles. Statistical significance was determined by two-tailed t-test. veh, vehicle.

Fig. 5. In vivo LINC complex disruption eliminates perinuclear MT cage, leading to nuclear elongation.

a, Representative mid-plane immunofluorescence images of the MT network (α-tubulin), with a zoom-in on a merge with labeled nuclei (Hoechst) in the WT, csDN-KASH, Lmna N195K and Lmna N195K csDN-KASH adult mouse cardiomyocytes. b, Quantification of perinuclear MT enrichment defined as perinuclear (PN) to cytoplasmic (Cyt) α-tubulin ratio (illustrated on the image on the right). Data are presented as mean ± s.e. of individual nuclei (open circles); individual animal replicate means are indicated with closed triangles. Statistical significance was determined by one-way ANOVA with Bonferroni correction. c, Nuclear aspect ratio (AR) as a function of PN MT enrichment. Open circles represent individual nuclei from the respective groups. A fit to the pooled data is depicted with continuous black line. Gray area represents the error to the fit. Biphasic piecewise linear regression fit is indicated with a dashed black line (see Methods for details). d, Super-resolution Airyscan jDCV mid-plane image of a Lmna N195K cardiomyocyte nucleus (Hoechst) with the dense MT network (α-tubulin) penetrating the chromatin protrusion (white arrow). WT: N = 3, n = 85. csDN-KASH: N = 3, n = 71. Lmna N195K: N = 3, n = 70. Lmna N195K csDN-KASH: N = 3, n = 75.

Fig. 6. MT disruption protects from nuclear damage in Lmna N195K cardiomyocytes.

a, Quantification of number of cGAS foci per nucleus for WT (N = 4, n = 1,898) and Lmna N195K (N = 4, n = 2,915) groups. Bar graphs represent the mean ± 1 s.d. of the pooled nuclei. Superimposed black triangles represent the replicate means. Statistical significance was determined by two-tailed t-test. b, Immunofluorescent mid-plane image of the MT network (α-tubulin) on the tip of Lmna N195K nucleus, associated with a chromatin protrusion (Hoechst, arrow) and cGAS foci (cGAS–tdTomato) indicative of NE ruptures. c, Experimental design for live imaging of perinuclear cGAS–tdTomato foci in a Lmna N195K mouse model, after in vitro stimulation in the presence of isoproterenol or colchicine treatment. d,e, Representative MIP images of nuclei (Hoechst) and perinuclear cGAS–tdTomato foci, from freshly isolated (baseline) and contractility-induced (iso + stim) cardiomyocytes (d) and DMSO-treated or colchicine-treated cardiomyocytes (e). The 1-µm perinuclear rings used for foci identification are indicated with white lines; the detected cGAS foci are indicated with arrows. Quantification of the number and area of perinuclear cGAS foci, per nucleus (bottom). N = 6, n = 4,869 (baseline), n = 2,286 (1 hour iso + stim), n = 4,820 (24 hour DMSO), n = 4,263 (24 hour colchicine). Bar graphs represent the mean ± 1 s.d. of the pooled nuclei. Superimposed black triangles represent the replicate means, connected by lines to their respective treatment conditions. Statistical significance was determined by two-tailed t-test. Iso, isoproterenol; stim, stimulation. Panel c created with BioRender.com.

Fig. 7. MT disruption preserves cardiac function and protects from nuclear damage in cardiac-specific Lmna-depleted mice.

a, Experimental scheme for concurrent MT depolymerization and cardiomyocyte-specific deletion of Lmna. b, Representative western blot measuring lamin A/C (left) and α-tubulin (right) protein levels 22 days after first tamoxifen (TMX) injection. c, Sequential % ejection fraction measured by echocardiography. Data are shown for pre-injection (day 0) and 11, 22 and 29 days after initial injection. Each data point represents one mouse. Red circles denote females; blue circles denote males. WT + veh: N = 8 (d29 N = 5), WT + colch: N = 9 (d29 N = 6), Lmna cKO + veh: N = 6 (Lmna cKO + vehicle-treated mice do not survive to day 29 after injection), Lmna cKO + colch: N = 9 (d29 N = 5). Error bars represent mean ± 1 s.d. Statistical significance was determined by two-way ANOVA with Tukeyʼs multiple correction. d, Representative cardiac tissue sections stained for WGA (yellow) and desmin (magenta). Images are MIPs of 4-µm-thick optical sections. Quantification of cardiomyocyte coverage area, inversely related to fibrosis, from WGA channel. Detected cardiomyocyte areas are highlighted on the images with semi-transparent white. Individual blue circles represent individual images; black triangles represent replicate means. N = 3 mice per group. WT veh: n = 35, WT colch: n = 37, Lmna cKO + veh: n = 39, Lmna cKO + colch: n = 38 images. Error bars represent mean ± 1 s.e. Statistical significance was determined by one-way ANOVA with Bonferroni multiple correction. e, Representative images of cardiomyocyte nuclei (Hoechst, gray) and surrounding MT network (α-tubulin, cyan). Images are MIPs of 4-µm-thick optical sections. Quantification of the percentage of nuclei with chromatin protrusions in each group by three independent, blinded scorers. Open circles represent the percentage of nuclei with chromatin protrusions quantified by each independent user for each animal; closed circles correspond to the mean for each animal. Mean ± s.e. is shown for each group. N = 3 mice per group. WT + veh: n = 114. WT + colch: n = 102. Lmna cKO + veh: n = 112. Lmna cKO + colch: n = 113 nuclei. Statistical significance was determined by one-way ANOVA with Bonferroni correction. f, Representative cardiac tissue sections stained for WGA (yellow), nuclei (Hoechst, gray), desmin (magenta) and α-tubulin (cyan). Images are MIPs of 4-µm optical sections. Cardiomyocyte nuclei are indicated with arrows. g, Quantification of cardiomyocyte nuclei AR, PN, Cyt and PN/Cyt. α-Tubulin intensity. N = 3 mice per group. WT veh: n = 113. WT colch: n = 112. Lmna cKO + veh: n = 111. Lmna cKO + colch: n = 112 nuclei. Statistical significance was determined by one-way ANOVA with Bonferroni correction. CM, cardiomyocyte; d, day; wo, weeks old. Panel a created with BioRender.com. Source data

Fig. 8. A computational model to explain nuclear damage in resting LMNA mutated cardiomyocytes and its rescue through LINC complex disruption.

a, Axisymmetric finite element (FE) model considering an imaginary stress-free configuration (top) for cardiomyocytes consisting of a round nucleus (which is further divided into nucleoplasm and NE and its underlying lamina), an ellipsoid MT cage and the surrounding randomly distributed unassembled myofibrils. Diastolic contractility, geometric constraints of the myocardium, restoring forces of titin proteins and compressive MT forces work in concert to deform this initial configuration to the physiological stressed configuration (bottom) where the nucleus is elongated and the myofibrils are aligned. b, Variations of the myofibril diastolic stress and MT compressive stress over the simulation time. c, Distribution of the longitudinal stress component in the cytoplasm and the MT cage in physiological conditions. d, Images of simulated nuclei (top) and comparison between the model predictions and the experimental data for nuclear aspect ratio for the four experimental groups (bottom). e, Simulation results for variation of the nuclear aspect ratio as a function of MT enrichment for WT and LMNA mutated (10 kPa) groups. Closed circles represent simulation data points; dashed lines represent piecewise linear regression fits for each group. f, Simulations show that, beyond a critical value for the MT compressive stresses, a form of instability emerges at the nuclear tips in which the location of maximum principal stress shifts from the middle of the nucleus to the tips (left, top and middle). Continuing the simulation after the instability by reducing MT compressive forces (simulating LINC complex disruption) shows that the nucleus becomes thinner and longer, with the maximum principal stress returning to the middle part (left, bottom). These observations are schematically shown in the right panel.

Extended Data Fig. 1. Validation of LINC complex and MT disruption in adult rat cardiomyocytes, and their effect on sarcomere–nuclear strain coupling.

(a) Nesprin-1 perinuclear enrichment in adult rat cardiomyocyte is lost following 48 h of adenoviral DN-KASH transduction, with no apparent changes in perinuclear sarcomeric organization (labeled with SiR-actin). (b) MT network (yellow) is enriched at axial nuclear tips depleted of myofibrils (magenta). Colchicine treatment (1 µM) for 24 h results in complete loss of the MT network (including the dense perinuclear MT cage), with no apparent change in perinuclear sarcomeric organization (labeled with SiR-actin). Representative nuclear mid-plane images are shown. (c) Sarcomere length, nuclear length, and nuclear width recordings over time during stimulated cardiomyocyte contraction. Left - cardiomyocytes treated for 48 h in culture with AdV empty (black) or AdV DN-KASH (blue). Right - cardiomyocytes treated for 24 h in culture with DMSO (black) or 1 µM colchicine (green). (d) Nuclear length versus sarcomere length coupling plots for AdV DN-KASH and colchicine treatments. AdV empty and AdV DN-KASH (48 h): N = 3, n = 51. DMSO and colch (24 h): N = 3, n = 51. Data presented as mean ± SE.

Extended Data Fig. 2. Validation of inducible, cardiac specific LINC complex disruption in WT and Lmna N195K mice.

(a) Nesprin-1 perinuclear enrichment in adult WT and Lmna N195K mouse cardiomyocytes is lost following cardiac specific in-vivo disruption of the LINC complex (csDN-KASH). Representative nuclear mid-plane images are shown. (b) Nesprin-2 perinuclear enrichment in cardiomyocytes is lost in csDN-KASH mice (left panel). Representative tissue sections are shown. Cardiomyocyte nuclei identified by myh7 and Hoechst counterstain (right panel) (c) Kaplan-Meier survival plot of the different experimental groups separated to males (solid line) and females (dashed line). (d) Left ventricular ejection fraction measured by echocardiography, for the Lmna N195K csDN-KASH mice at 10 and 12 weeks of age. 10 weeks: N = 13, 12 weeks: N = 4. Error bars represent mean ± SD. (e) Echocardiography measurements at 10 weeks of age for the indicated groups. Heart rate (HR), stroke volume (SV), % fractional shortening (FS), cardiac output (CO), corrected left-ventricular mass (LV mass), end-systolic and end-diastolic left ventricular diameters (Diameter, s and d), end-systolic and end-diastolic volume (ESV, EDV), end-systolic and end-diastolic left-ventricle anterior and posterior wall thickness (LVAW, s and d, LVPW s and d). Closed circles denote females, open circles denote males. WT: N = 9, csDN-KASH: N = 15, Lmna N195K: N = 9, Lmna N195K csDN-KASH N = 13. Error bar represents mean ± SD. Statistical significance determined by one-way ANOVA with Tukey multiple correction.

Extended Data Fig. 3. Nuclear and cardiomyocyte morphology in Lmna N195K mice.

(a) Quantification of nuclear aspect ratio in cardiomyocytes isolated from WT and Lmna N195K mice carrying cGAS-tdTomato reporter. Red lines represent mean of pooled nuclei, and black triangles correspond to individual replicate means. WT cGAS-tdTomato: N = 4, n = 1898. Lmna N195K cGAS-tdTomato: N = 4, n = 2915. (b) Isolated cardiomyocyte cell length, width and cell aspect ratio for the indicated mouse models. WT: N = 3, n = 133. csDN-KASH: N = 3, n = 105. Lmna N195K: N = 3, n = 113. Lmna N195K csDN-KASH: N = 3, n = 143. Data presented as mean ± SE. Statistical significance determined by 1-way ANOVA with Bonferroni correction. (c) NE rupture in Lmna N195K cardiomyocytes manifested by chromatin protrusion from the nucleus, with partial lamina coverage. Representative mid-plane confocal image of the nucleus is shown to illustrate ruptured lamina. Protrusions quantified from full data set in main Fig. 3h.

Extended Data Fig. 4. Active sarcomere–nuclear coupling for cardiac specific LINC complex disruption in WT and Lmna N195K mice.

Sarcomere length (SL), nuclear length (NL), and nuclear width (NW) recordings over time during stimulated contraction of freshly isolated cardiomyocytes 8-9 weeks of age. Sarcomere strain and nuclear strain over time are plotted below. (a) Lmna N195K (red) and WT littermate controls (black). (b) Cardiac specific LINC complex disruption (gray) and WT vehicle controls (black). (c) Lmna N195K with cardiac specific LINC complex disruption (blue) and littermate Lmna N195K controls (black). (d) NL versus SL coupling plots for the corresponding experimental groups. WT: N = 4, n = 58. Lmna N195K: N = 4, n = 69. WT veh: N = 4, n = 81. csDN-KASH: N = 4, n = 69. Lmna N195K veh: N = 4, n = 55. Lmna N195K csDN-KASH: N = 4, n = 59. Data presented as mean ± SE.

Extended Data Fig. 5. Cytoskeletal characterization of rat and mouse cardiomyocytes with LINC complex disruption.

(a) Quantification of α-tubulin enrichment at the nuclear tips and sides in the indicated mouse model cardiomyocytes. (b) Representative mid-plane immunofluorescence images of desmin intermediate filament network surrounding the nuclei (Hoechst) of the indicated groups. (c) Desmin western blot and quantification, normalized to H3, for the indicated groups. (d) Representative mid-plane immunofluorescence images of kinesin-1 and quantification of perinuclear kinesin-1 enrichment, defined as PN to Cyt kinesin-1 ratio. WT: N = 2, n = 46. csDN-KASH: N = 2, n = 51. Lmna N195K: N = 2, n = 50. Lmna N195K csDN-KASH: N = 2, n = 59. (e) Representative nuclear mid-plane images of adult rat cardiomyocytes labeled with α-tubulin and Hoechst (left), following 48 h of adenoviral LINC complex disruption. Quantification of perinuclear α-tubulin enrichment is shown on the right. AdV empty and AdV DN-KASH: N = 2, n = 29. (f) Representative nuclear mid-plane images of kinesin-1 in the entire adult rat cardiomyocyte following 48 h of adenoviral LINC complex disruption (left). Quantification of kinesin-1 intensity separately at the perinucleus and the cytoplasm is shown on the right. AdV empty: N = 2, n = 78, AdV DN-KASH: N = 2, n = 80. Data presented as mean ± SE. Statistical significance determined by two-tailed t-test. Source data

Extended Data Fig. 6. Full cardiomyocyte views and nuclear aspect ratio quantification for Lmna N195K cGAS-tdTomato experimental groups.

(a) Full cell views of representative maximum intensity projection images labeled for nuclei (Hoechst) and cGAS-tdTomato (red), from freshly isolated cGAS-tdTomato WT, and cGAS-tdTomato Lmna N195K mouse models. The middle panel shows a zoom in on the nucleus and the perinuclear region used for cGAS foci quantification. Relates to the full data set quantified in main Fig. 6a. (b) Full cell views of representative maximum intensity projection images labeled for nuclei (Hoechst) and cGAS-tdTomato (red) for the indicated experimental groups. (c) Nuclear aspect ratio (AR) for the experimental groups of cGAS-tdTomato Lmna N195K model. N = 6, n = 4869 (baseline), n = 2286 (1 h iso + stim), n = 4820 (24 h DMSO), n = 4263 (24 h colch). Red lines represent the puled nuclei mean for each experimental group. Superimposed black triangles represent the replicate means and connected by lines to their respective treatment conditions. Statistical significance determined by 1-way ANOVA with Bonferroni correction.

Extended Data Fig. 7. MT disruption protects from DNA damage in LMNA-deficient hiPSC-derived cardiomyocytes.

(a) Experimental design for the examination of DNA damage in hiPSC-derived cardiomyocytes upon transfection with non-targeted siRNA (siNT) or LMNA-targeted siRNA (siLMNA), and subsequent DMSO or colchicine (colch) treatment. Created in BioRender. Suay Corredera, C. (2025) https://BioRender.com/z6iym3d (b) Western blots, separate for replicate 1 and pooled lysates from replicates 2/3 showing the siRNA-mediated lamin A/C knock-down in hiPSC-derived cardiomyocytes. GAPDH was used as a loading control. (c) Representative mid-plane images of nuclei (Hoechst) and γH2A.X foci, from the four experimental groups of hiPSC-derived cardiomyocytes. Hoechst was used to outline the nuclei (white line), and the detected γH2AX foci are highlighted in yellow. (d) Quantification of mean nuclear γH2A.X intensity and γH2A.X foci fraction of nuclear volume coverage (top). Quantification of the normalized number and volume of γH2A.X foci (bottom). Each replicate is normalized to the mean value of the NT DMSO group. Individual nuclei presented as open circles with mean ± SE of the pooled nuclei. Closed triangles depict replicate means connected by lines to their respective experimental conditions. N = 3, n = 199 (NT DMSO), N = 3, n = 197 (NT colch), N = 3, n = 192 (siLMNA DMSO), N = 3, n = 203 (siLMNA colch). Statistical significance determined by one-way ANOVA with Bonferroni correction. Source data

Extended Data Fig. 8. Sequential echocardiography and survival in Lmna cKO mice.

(a) Sequential echocardiography measurements in cardiac specific Lmna depleted mice with in-vivo microtubule disruption. Data shown for pre-injection (day 0), 11, 22, and 29 days post initial injection. Herat rate (HR), stroke volume (SV), % fractional shortening (FS), cardiac output (CO), end-systolic and end-diastolic left ventricular diameters (Diameter, s and d), end-systolic and end-diastolic volume (ESV, EDV), corrected left-ventricular mass (LV mass), end-systolic and end-diastolic left-ventricle anterior and posterior wall thickness (LVAW, s and d, LVPW s and d). Red circles denote females, blue circles denote males. WT + veh: N = 8 (day 29: N = 5), WT + colch: N = 9 (day 29: N = 6), Lmna cKO + veh: N = 6 (Lmna cKO + vehicle treated mice do not survive to day 29 post injection), Lmna cKO + colch: N = 9 (day 29: N = 5). Error bars represent mean ± 1 SD. (b) Kaplan-Meier survival plots of the different experimental groups. Male: WT+veh: N = 2, WT+colch: N = 2, Lmna cKO+veh: N = 5, Lmna cKO+colch: N = 5. Female: WT+veh: N = 3, WT+colch: N = 2, Lmna cKO+veh: N = 5, Lmna cKO+colch: N = 6. Statistical significance determined by two-way ANOVA with Tukey multiple correction.

Extended Data Fig. 9. Cellular and immune marker quantification in Lmna cKO mice.

(a) Quantification of cardiomyocyte nuclear length, width, and area. (b) Quantification of perinuclear (PN), cytoplasmic (Cyt) and PN/Cyt. desmin intensity from maximum-intensity-projection images of cardiac tissue sections for the indicated groups. N = 3 mice per group. WT veh: n = 113. WT colch: n = 112. Lmna cKO + veh: n = 111. Lmna cKO + colch: n = 112 nuclei. Statistical significance determined by 1-way ANOVA with Bonferroni correction. (c) Representative cardiac tissue sections of the indicated groups, stained for CD45 (magenta), a common leukocyte marker, CD68 (yellow), which labels monocytes/macrophages and some other immune cells, and nuclei (Hoechst, gray). Images are maximum intensity projections of 3 μm optical sections. (d) Quantification of the percentage coverage area for CD45 and CD68. Blue open circles represent images/myocardial areas. Black triangles represent replicate means. N = 3 mice per group. WT + veh: n = 18, WT + colch: n = 18, Lmna cKO + veh: n = 18, Lmna cKO + colch: n = 17 images. Error bars represent mean ± SE. Statistical significance determined by one-way ANOVA with Bonferroni multiple correction.

Extended Data Fig. 10. Simulation parameters and results for the computational model describing perinuclear stress distribution in mature cardiomyocytes.

(a) Simulation results depicting stress distribution and average myosin motor density in the cardiomyocyte cytoplasm under physiological conditions. Radial stress component, tangential stress component and average motor density defined as $({\rho }_{11}+{\rho }_{22}+{\rho }_{33})/3$. (b) Model predictions for the maximum principal (maximum tensile) stress direction in the cytoplasm. (c) In vivo % volume changes relative to WT conditions. (d) Variations of the MT cage stiffness and active stress to simulate MT enrichment around the nucleus. (e) Simulation results for stress distribution in the microtubule cage in WT conditions. Radial stress component (Pa), Tangential stress component (Pa) and hydrostatic pressure (Pa).