Muscular dystrophy–associated lamin variants disrupt cellular organization through a nucleolar-ribosomal axis

Abstract

Emery-Dreifuss muscular dystrophy (EDMD) arises from mutations in nuclear lamins or emerin. Current pathological models emphasize defective nuclear mechanics and transcriptional regulation, yet these mechanisms cannot explain how lamina defects propagate across the cell to produce the complex pathology of laminopathies. Here, we reveal an emerging pathway linking nuclear lamina dysfunction to cytoplasmic reorganization. Using Caenorhabditis elegans EDMD models, we show that disease-linked lamin variants reduce cytoplasmic mesoscale crowding, increase molecular diffusivity, and disrupt nuclear positioning and endoplasmic reticulum architecture, which mirror phenotypes caused by ribosome depletion. Lamin dysfunction also lowers nucleolar fibrillarin levels and ribosome abundance, revealing a nucleolar-ribosomal axis that transmits nuclear defects to the cytoplasm. Loss of the redundant LEM-domain proteins emr-1 and lem-2 phenocopied lamin mutants, indicating that cytoplasmic disorganization is a shared hallmark of EDMD. These findings connect nuclear architecture to whole-cell biophysics and suggest therapeutic strategies aimed at restoring ribosome function.

INTRODUCTION

Emery-Dreifuss muscular dystrophy (EDMD) is a devastating neuromuscular disorder characterized by progressive skeletal muscle weakness, cardiac conduction defects, and joint contractures (1). EDMD arises primarily from mutations in either LMNA, encoding nuclear A-type lamins (lamins A/C), or EMD, encoding the inner nuclear membrane protein emerin (1). A-type lamins confer mechanical stability to the nucleus, connect the nucleus to the cytoskeleton via linker of nucleoskeleton and cytoskeleton (LINC) complexes, and regulate chromatin organization to control gene expression (2). Emerin is an inner nuclear membrane protein and member of the LEM-domain family (named for LAP2, Emerin, and MAN1) that interacts with lamins and chromatin to organize the nuclear periphery (3, 4). Together, these observations highlight that EDMD arises from disruption of the lamina network rather than a single protein, underscoring the challenge of understanding how lamina dysfunction mechanistically leads to the tissue-specific EDMD pathologies.

A defining cellular hallmark of EDMD is aberrant nuclear positioning in skeletal muscle fibers (5, 6). In healthy muscle, nuclei distribute peripherally near the sarcolemma, but patients with EDMD show nuclei inappropriately clustered at the fiber center (5). This nuclear mispositioning indicates disrupted nuclear-cytoskeletal coupling, likely through LINC complexes (7, 8). However, nuclear positioning defects alone cannot fully explain EDMD’s complex pathophysiology, as other diseases with prominent nuclear mispositioning present distinct progression patterns (9), indicating that additional cellular mechanisms must be disrupted.

Current laminopathy models focus primarily on altered mechanotransduction and dysregulated gene expression (10). While these mechanisms contribute to disease, neither fully accounts for the broad cellular disorganization observed in EDMD tissues, including ruffled nuclear envelope and mispositioned muscle nuclei (10). Recent advances in understanding cytoplasmic biophysical properties suggest a third possibility: lamin mutations might disrupt fundamental physical properties of the cellular interior, leading to widespread organizational defects (11, 12).

The eukaryotic cytoplasm is a highly organized, crowded environment where macromolecular interactions influence all cellular functions (13). Macromolecular crowding affects protein folding, enzyme kinetics, liquid-liquid phase separation, and organelle positioning (13). Ribosome concentration is a major determinant of cytoplasmic macromolecular crowding and cellular organization in yeast and mammalian cells (11). Ribosome depletion reduces cytoplasmic macromolecular crowding, disrupts nuclear positioning, and causes endoplasmic reticulum (ER) network collapse in C. elegans (12). Ribosomes assemble in the nucleolus before cytoplasmic export, creating a direct link between nuclear function and cytoplasmic organization (11, 14).

C. elegans provides unique advantages for investigating potential interactions between nuclear lamins, nucleolar function, and cytoplasmic macromolecular crowding. The nematode has a single lamin gene, lmn-1, whose product functionally substitutes for both A- and B-type mammalian lamins (7, 15, 16). Like human lamins, LMN-1 localizes to the nuclear periphery, interacts with chromatin, and is required for proper nuclear shape and positioning (7, 16, 17). C. elegans EDMD models engineered using CRISPR-Cas9 faithfully recapitulate many disease phenotypes, including reduced fitness, movement disorders, and nuclear morphology defects (17). The lmn-1(R64P) mutation, corresponding to the severe human variant LMNA p.R50P, is particularly disruptive, causing both polymerization defects in vitro, and EDMD-like phenotypes in vivo (15, 17). C. elegans is also well suited for measuring macromolecular crowding through in vivo nanorheology using genetically encoded multimeric nanoparticles (GEMs) (11, 12), fluorescent protein assemblies that report local crowding conditions through their diffusive behavior in live animals. This approach revealed that cytoplasmic mesoscale macromolecular crowding in C. elegans tissues is regulated by the giant ER/outer nuclear membrane Klarsicht/ANC-1/SYNE homology (KASH) protein ANC-1 and ribosome concentration (12), both of which independently influence nuclear positioning and ER morphology.

Here, we test whether EDMD-associated lamin mutations disrupt cellular organization through effects on mesoscale macromolecular crowding, ribosome biogenesis, and nucleolar function. Our findings reveal a nucleolar-ribosomal axis through which nuclear lamina defects propagate to cause broad cellular dysfunction, providing a mechanistic basis for the cellular disorganization observed in EDMD tissues.

RESULTS

EDMD-associated lamin variants disrupt cytoplasmic biophysical properties in vivo

To investigate how lamin mutations affect cytoplasmic biophysical properties in living tissues, we used in vivo nanorheology in living C. elegans. This approach provides real-time quantification of cytoplasmic biophysical properties in intact animals using 40-nm GEMs as fluorescent tracers to explore the mesoscale intracellular environment (11, 12). We quantified cytoplasmic biophysical properties by analyzing mean squared displacement (MSD) curves to calculate effective particle diffusion coefficients (D_eff) through tracking trajectories of thousands of individual GEMs with high-speed spinning-disc confocal microscopy. Measurements were made in the hypodermis and intestine of homozygous lmn-1 mutant animals and compared to heterozygous controls maintained by hT2 balancer (movies S1 and S2) (17, 18). In heterozygous control animals, most GEMs were highly constrained within the crowded cytoplasm, while a smaller population of GEMs showed faster, more diffusive movements (Fig. 1, A and C). This bimodal pattern was consistent with previous observations in wild-type animals and reflects the spatially heterogeneous nature of the intracellular environment (12).

Fig. 1. lmn-1 variants Y59C and R64P disrupt cytoplasmic biophysical properties.

(A and C) Representative inverted grayscale spinning disc confocal images of GEMs in the intestine (A) or hypodermis (C) of day 1 lmn-1(R64P)/hT2 (top) or lmn-1(R64P) (bottom) adults. Scale bar, 10 μm. Insets show magnified views of boxed regions. Right panels display representative GEM trajectories color-coded by frame number. Scale bar, 4 μm. Concentric circles indicate 2 and 4 μm radii from the trajectory origin. (B and D) Two-dimensional probability density histograms of D_eff versus the anomalous diffusion exponent (α) over the initial 100 ms for GEM diffusion in the intestine (B) or hypodermis (D) of day 1 lmn-1(R64P)/hT2 (top) or lmn-1(R64P) (bottom) adults. Marginal distributions shown as frequency histograms on top and right axes. Color scale indicates particle count per bin. (E and H) Gaussian mixture model (GMM) analysis of GEM D_eff in the intestine (E) or hypodermis (H) for the indicated genotypes. Distributions are plotted in log space; annotated values indicate the geometric mean of each identified peak. (F, G, I, and J) Bootstrap analysis (1000 iterations) of GMM-derived GEM diffusion parameters in the intestine (F and G) or hypodermis (I and J). Bar plots (F and I) show the percentage of the constrained GEM population; violin plots (G and J) show the D_eff of the unconstrained GEM population. All values represent mean ± 95% confidence interval (CI).

We next characterized GEM behavior in EDMD-associated lmn-1 homozygous mutants, focusing on two early-onset variants: lmn-1(R64P) and lmn-1(Y59C), respectively, corresponding to human LMNA p.R50P and p.Y45C (15, 17, 19). The lmn-1(R64P) mutation causes polymerization defects in vitro (15). Homozygous lmn-1(R64P) mutants showed substantial and reproducible increases in GEM mobility compared to balanced heterozygous controls in both the intestine and the hypodermis. In contrast, homozygous lmn-1(Y59C) mutants displayed marked increases in GEM mobility in the hypodermis but minimal effects in the intestine (Fig. 1, A and C, and fig. S1). To quantitatively analyze these distinct kinetic populations, we performed MSD analysis (fig. S2) to extract an effective diffusion coefficient (D_eff) and an anomalous exponent (α) for each GEM trajectory (Fig. 1, B and D), and then applied Gaussian mixture modeling to objectively separate trajectories into low and high mobility population using our previously developed analytical framework (Fig. 1, E and H) (12). In the diffusivity distributions, the left peak corresponds to GEMs whose motion is frequently hindered by cellular structures (“constrained” GEMs), while the right peak corresponds to GEMs diffusing more freely through the cytoplasm (“unconstrained” GEMs). We indicate the geometric mean of each peak to assist comparison of distribution shifts (Fig. 1, E and H). Homozygous lmn-1(R64P) and lmn-1(Y59C) mutants both showed a rightward shift in the unconstrained peak relative to their balanced controls, indicating enhanced diffusion of freely moving particles. To rigorously quantify this, we isolated the unconstrained population and used bootstrap analysis (1000 iterations) to estimate the arithmetic mean D_eff. This revealed that lmn-1(R64P) markedly increased unconstrained D_eff in both the intestine and hypodermis (Fig. 1, G and J), whereas lmn-1(Y59C) produced pronounced increases only in the hypodermis (Fig. 1, G and J, and fig. S1). Notably, the relative abundance and diffusivity of the constrained fraction remained largely unchanged across all genotypes (Fig. 1, F and I), indicating that lamin dysfunction primarily disrupts mesoscale macromolecular crowding rather than the structural constraints that physically confine GEM movement. Together, these results demonstrate that EDMD-associated lamin mutations alter cytoplasmic biophysical properties by reducing mesoscale macromolecular crowding, albeit with distinct tissue sensitivities.

EDMD-associated lamin variants recapitulate disease-relevant cellular defects

Disrupted nuclear positioning is consistently observed in the skeletal muscle of patients with EDMD (1, 5). We therefore examined whether our C. elegans lamin mutants recapitulate this phenotype by analyzing nuclear positioning in the hypodermis, a single-cell syncytial layer containing 139 nuclei that are evenly spaced apart and anchored in place by mechanisms involving the giant KASH protein ANC-1 (20, 21). We quantified nuclear anchorage defects by measuring the percentage of nuclei found in abnormal clusters (22, 23).

EDMD-associated lamin mutations caused significant nuclear clustering compared to controls (Fig. 2, A and B). The severe lmn-1(R64P) variant produced the strongest effect, with ∼35% of nuclei clustered, versus <5% in wild-type animals. lmn-1(Y59C) had an intermediate effect, consistent with its less severe impact on cytoplasmic crowding in the hypodermis. This nuclear mispositioning defect parallels those observed in EDMD patient tissues. The severity hierarchy, with R64P showing a stronger defect than Y59C, also matches known in vitro polymerization defects and organismal phenotypes (15, 17), demonstrating that EDMD-associated lamin mutations disrupt the mechanisms governing nuclear positioning.

Fig. 2. lmn-1 variants disrupt nuclear positioning and ER morphology in the hypodermis.

(A) Representative spinning disc confocal (left) or epifluorescence (right) images of GFP-labeled hypodermal nuclei (ycIs10, see Materials and Methods) in day 1 adult worms of the indicated strains. The GFP marker labels both hyp7 and seam cell nuclei where seam cell nuclei appear brighter. Only hyp7 nuclei were scored for clustering. Arrowheads indicate clustered hypodermal nuclei; dotted lines outline worm boundaries. Scale bar, 10 μm. (B) Nuclear clustering quantified as the percentage of clustered nuclei per worm in day 1 adults of the indicated strains. Each point represents one worm (n = 20 for each genotype). Welch’s ANOVA test: W (5, 50) = 44, P < 0.001. (C) Representative spinning disc confocal images of hypodermal ER organization in day 1 adults of the indicated strains. Left panels show ER morphology visualized by mKate2::TRAM-1 fluorescence; right panels show thresholded images with ER outline (cyan) and hypodermal boundary (red). Insets show magnified views of boxed regions. Scale bar, 10 μm. (D) ER occupancy (% of total hypodermal area) across the indicated genotypes (n = 10, 16, 31, and 18, left to right). Kruskal-Wallis test: for (D), H (3) = 1.491, P > 0.05 (ns). (E) ER perimeter complexity (see Methods) across the indicated genotypes (n = 10, 16, 31, and 18, left to right). Kruskal-Wallis test: for (E), H (3) = 30.458, P < 0.001. Each point represents one worm. Statistical significance is indicated as follows: ****P < 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, and ns: P > 0.05; exact P values are shown above each comparison in the figure.

To examine whether lamin mutations affect broader cellular architecture beyond nuclear positioning, we examined ER morphology. The ER forms extensive networks throughout hypodermal syncytia that depend on anc-1 pathways involved in cytoplasmic constraint and macromolecular crowding pathways reliant on ribosome concentration (12, 20). In balanced heterozygous controls, the hypodermal ER displays a complex branched network of interconnected tubules and sheets with small spaces between ER-rich regions (Fig. 2C). lmn-1(R64P) mutants, however, showed altered ER architecture, with larger void spaces suggesting network collapse (Fig. 2C). The mutant ER formed more sheet-like structures with reduced tubular branching, resembling simplified networks observed under cytoplasmic stress or crowding defect (24). We quantified these defects using described morphometric approaches (12, 25), measuring both ER occupancy (total area coverage) and network complexity (perimeter-to-area ratio). While lmn-1(R64P) mutants showed no significant change in total ER area, they exhibited marked reduction in network complexity (Fig. 2, D and E). Intriguingly, the lmn-1(Y59C) variant displayed a distinct pattern for this morphological phenotype; both lmn-1(Y59C)/hT2 heterozygotes and lmn-1(Y59C) homozygotes exhibited significantly reduced perimeter complexity (Fig. 2E). This finding suggests that the Y59C mutation exerts a dominant effect on ER network architecture that is independent of its effects on bulk cytoplasmic crowding, distinguishing it mechanistically from lmn-1(R64P). Notably, the nature of ER network defects in lmn-1(R64P) mutants closely matched those previously observed in ribosome-depleted animals (12), suggesting a shared underlying mechanism. These data demonstrate that lamin dysfunction produces cellular defects quantitatively and qualitatively similar to ribosome depletion effects, pointing toward a common pathway linking nuclear lamina integrity to cytoplasmic organization.

LMN-1 and ribosomes function in a common pathway controlling cellular organization

Since lmn-1(R64P) mutants phenocopy several cellular organization defects previously observed in ribosome-depleted animals, including reduced cytoplasmic crowding, nuclear mispositioning, and ER network collapse, we hypothesized that lamins and ribosomes function in the same pathway to regulate mesoscale macromolecular crowding and cellular organization. This model predicts that if lamins act upstream of ribosomes in the same pathway, combining lmn-1 mutations with partial ribosome depletion should produce epistatic rather than additive effects. To test this prediction, we performed genetic interaction studies using rps-18(RNAi) to partially deplete ribosomes in various lmn-1 mutant backgrounds and quantified three independent phenotypes: GEM diffusion, nuclear positioning, and ER architecture.

Ribosome depletion by rps-18(RNAi) robustly increased intestinal GEM diffusion in heterozygous lmn-1(R64P)/hT2, lmn-1(Y59C)/hT2, and homozygous lmn-1(Y59C) animals, as shown by GEM trajectories, two-dimensional probability density histograms, and quantitative analyses (Figs. 1, E and G; and 3, A to C and E), reproducing previous findings that ribosomes regulate cytoplasmic crowding (12). This response demonstrated that ribosome depletion retained its ability to disrupt cytoplasmic organization in genetic backgrounds with partial lamin function. However, rps-18(RNAi) failed to further increase cytoplasmic GEM diffusion in homozygous lmn-1(R64P) mutants, revealing a clear epistatic interaction (Figs. 1G and 3, B to E). This epistasis suggests that the severe lmn-1(R64P) mutation maximally disrupts the lamin-ribosome pathway. The differential responses between lmn-1(Y59C) (additive effects) and lmn-1(R64P) (epistatic effects) provide insight into the severity hierarchy of EDMD variants and their relative impacts on the lamin-ribosome regulatory axis.

Fig. 3. LMN-1 and ribosomes act in the same pathway to control cytoplasmic organization.

(A) GEM trajectories color-coded by frame number in intestinal cytoplasm of the indicated strains. Scale bar, 4 μm. Concentric circles indicate 2 and 4 μm radii from the trajectory origin. (B) Two-dimensional probability density histograms of GEM D_eff versus α. Marginal histograms are shown on top and right axes. Color scale indicates particle count per bin. (C) GMM analysis of intestinal GEM D_eff. (D and E) Bootstrap analysis (1000 iterations) of GMM-derived intestinal GEM diffusion parameters. Bar plots (D) show the percentage of the constrained GEM population; violin plots (E) show D_eff of the unconstrained population. All values represent mean ± 95% CI. (F) Representative spinning disc confocal images of GFP-labeled hypodermal nuclei. The marker labels both hyp7 and seam nuclei, with seam nuclei appearing brighter. Only hyp7 nuclei were scored. Arrowheads indicate clustered nuclei; dotted lines outline worm boundaries. Scale bar, 10 μm. (G) Nuclear clustering quantified as the percentage of clustered nuclei per worm (n = 20, 33, 20, 19, 20, and 13, left to right). Welch’s ANOVA: W (5, 43) = 115, P < 0.001. (H) Representative images of hypodermal ER organization. Left: mKate2::TRAM-1 fluorescence; right: thresholded images with ER outline (cyan) and hypodermal boundary (red). Insets show magnified boxed regions. Scale bar, 10 μm. (I and J) ER occupancy (I) and perimeter complexity (J; see Materials and Methods) across the indicated genotypes (n = 10, 16, 14, 12, and 12, left to right). Kruskal-Wallis: H (4) = 40 for (I) and 44 for (J), P < 0.001. Each point represents one worm. Significance: ****P < 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, and ns: P > 0.05; exact P values are shown above each comparison in the figure.

Nuclear positioning assays revealed similar epistatic patterns. While rps-18(RNAi) slightly enhanced nuclear clustering defects in lmn-1(Y59C) mutants, it produced no additional defect in lmn-1(R64P) backgrounds (Fig. 3, F and G). This consistent pattern across two independent cellular phenotypes strengthened the evidence for pathway convergence and indicated that genetic interactions were not specific to cytoplasmic crowding effects alone. As predicted by the pathway model, the ER morphology parameters in rps-18(RNAi); lmn-1(R64P) double-mutant animals remained comparable to rps-18(RNAi) heterozygous controls (Fig. 3, H to J), completing the epistatic pattern across all three cellular organization phenotypes we examined. Thus, across all three phenotypes, LMN-1 and ribosomes act in a common pathway controlling cellular organization.

LMN-1 regulates the nucleolar-ribosomal axis

Having demonstrated that lamins and ribosomes function in the shared pathway to regulate cellular organization, we sought to determine the mechanistic relationship between these factors. Ribosomes are synthesized and assembled within the nucleolus before export to the cytoplasm (14). Previous studies in mammalian tissue culture cells showed that lamins play important roles in nucleolar function (26, 27), but the effects of EDMD-associated lamin mutations on nucleolar organization and ribosome production in intact animal tissues remained unexplored. We hypothesized that lamin dysfunction disrupts cellular organization by interfering with ribosome biogenesis in the nucleolus.

To test this hypothesis, we first examined nucleolar organization in lmn-1 mutant animals using FIB-1::eGFP, a fluorescently tagged fusion of the C. elegans fibrillarin ortholog, an essential nucleolar protein (28). Fibrillarin is a well-established marker of nucleolar activity whose levels correlate with ribosome biogenesis capacity across multiple species (14, 29–31). Live imaging of nucleoli in the hypodermal syncytium allowed us to quantify nucleolar morphology and FIB-1::eGFP abundance (Fig. 4, A to H). In lmn-1(R64P) mutants, nucleolar size was comparable to lmn-1(R64P)/hT2 controls (Fig. 4B), yet mean FIB-1::eGFP intensity was significantly reduced (Fig. 4C), without a corresponding change in total integrated fluorescence intensity (Fig. 4D). This pattern, reduced local concentration but preserved total signal, suggests that R64P alters the internal organization or compaction of nucleolar contents rather than overall FIB-1 abundance. In contrast, lmn-1(Y59C) mutants showed smaller nucleoli with elevated mean FIB-1::eGFP intensity but reduced total integrated fluorescence relative to lmn-1(Y59C)/hT2 controls (Fig. 4, E to H), indicating an overall reduction in FIB-1 abundance concentrated into a smaller nucleolar volume. Thus, although both EDMD-associated lamin variants perturb nucleolar integrity, they do so in mechanistically distinct ways: R64P disrupts nucleolar organization without altering overall size or FIB-1 abundance, whereas Y59C produces smaller, denser nucleoli with reduced total FIB-1.

Fig. 4. EDMD-associated lmn-1 variants alter nucleolar organization and reduce ribosome abundance.

(A) Representative spinning-disk confocal images of the hypodermis in day 1 adult lmn-1(R64P)/hT2 and lmn-1(R64P) animals showing DIC (left) and FIB-1::eGFP fluorescence (right). Insets show magnified views of representative nucleoli. Fluorescence intensity is displayed using the viridis colormap. Scale bar, 10 μm. (B to D) Quantification of nucleolar area (B), mean FIB-1::eGFP fluorescence intensity (C), and integrated FIB-1::eGFP fluorescence intensity (D) in the indicated genotypes. Each dot represents one nucleolus (n = 126 and 111, left to right). (E) Representative spinning-disk confocal images of FIB-1::eGFP in the hypodermis of day 1 adult lmn-1(Y59C)/hT2 and lmn-1(Y59C) animals. Insets show magnified views of representative nucleoli. Fluorescence intensity is displayed using the viridis colormap. Scale bar, 10 μm. (F to H) Quantification of nucleolar area (F), mean FIB-1::eGFP fluorescence intensity (G), and integrated FIB-1::eGFP fluorescence intensity (H) in the indicated genotypes. Each dot represents one nucleolus (n = 101 and 94 nucleoli from 10 worms per genotype, left to right). (I and K) Representative spinning-disk confocal images of hypodermal GFP₁₁::RPS-18 fluorescence, reconstituted by split-GFP in day 1 adult lmn-1(R64P)/hT2 and lmn-1(R64P) animals (I), and lmn-1(Y59C)/hT2 and lmn-1(Y59C) animals (K). Scale bar, 10 μm. (J and L) Mean GFP₁₁::RPS-18 fluorescence intensity per animal in the indicated genotypes for R64P (J) and Y59C (L) backgrounds. Each dot represents one worm. Bars indicate mean ± 95% CI. Statistical significance was assessed by two-tailed Welch’s t test; exact p values are shown above each comparison. Asterisks denote significance as follows: ****P < 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, and ns, P > 0.05.

To further test whether lamin mutations perturb nucleolar organization, we analyzed a second independent nucleolar marker: NUCL-1::GFP, the endogenously tagged C. elegans ortholog of nucleolin (32). Consistent with the FIB-1::eGFP data, NUCL-1::GFP revealed nucleolar abnormalities in both lmn-1(R64P) and lmn-1(Y59C) mutant backgrounds (fig. S4), though the specific changes in nucleolar size and intensity differed between the two variants, consistent with the mechanistically distinct perturbations described above. Together, the concordant abnormalities detected with both FIB-1::eGFP and NUCL-1::GFP, markers with distinct nucleolar distributions, indicate that EDMD-associated lamin mutations broadly disrupt nucleolar organization rather than selectively affecting a single nucleolar component.

We next asked whether these nucleolar defects were accompanied by reduced ribosome abundance. Using endogenously tagged GFP11::RPS-18 visualized by split-GFP reconstitution in the hypodermis (33), we observed significant reductions in ribosomal fluorescence in both lmn-1(R64P) and lmn-1(Y59C) mutants relative to their respective balanced controls (Fig. 4, I to L). A second independent ribosomal marker, RPL-29::GFP (33), showed consistent reductions in hypodermal fluorescence intensity in both mutant backgrounds (fig. S5). Together, two independent nucleolar markers and two independent ribosomal markers point to the same conclusion: lamin dysfunction broadly perturbs nucleolar organization and reduces ribosome abundance.

Activation of mTOR signaling partially restores cytoplasmic crowding in EDMD-associated lamin mutants

The results above suggested that impaired ribosome biogenesis may underlie the altered cytoplasmic biophysical properties of lamin mutants. If so, stimulating ribosome production should partially restore cytoplasmic crowding. We therefore tested whether activation of mTOR signaling could suppress the crowding defects of lmn-1 mutants by depleting the components of the guanosine triphosphatase–activating protein activity toward Rags (GATOR) complex, specifically the GATOR1 components nprl-2 or nprl-3, which negatively regulate mTORC1 activity (34), and quantifying hypodermal GEM diffusion.

In lmn-1(R64P) mutants, depletion of either nprl-2 or nprl-3 shifted the GEM diffusivity distributions toward lower values relative to control(RNAi) animals (Fig. 5A), consistent with partial restoration of cytoplasmic crowding. Bootstrap analysis showed this effect was driven primarily by a reduction in the effective diffusion coefficient of the unconstrained GEM population, with little change in the relative abundance of constrained fraction (Fig. 5, B and D). A similar pattern was observed in lmn-1(Y59C) mutants, where depletion of either nprl-2 or nprl-3 likewise reduced unconstrained GEM diffusivity relative to controls (Fig. 5, C and D). Together, these data suggest that mTOR activation partially restores mesoscale crowding in lamin mutants without substantially altering structural constraints.

Fig. 5. Activation of mTOR signaling by depletion of GATOR1 components partially restores cytoplasmic crowding in EDMD-associated lmn-1 mutants.

(A and C) GMM analysis of GEM D_eff distributions in the hypodermis of day 1 adult lmn-1(R64P) (A) and lmn-1(Y59C) (C) animals subjected to control(RNAi), nprl-2(RNAi), or nprl-3(RNAi). Annotated values indicate the geometric mean of each identified peak. (B and D) Bootstrap analysis (1000 iterations) of GMM-derived GEM diffusion parameters in the hypodermis for the indicated conditions. Bar plots (B) show the percentage of the constrained GEM population; violin plots (D) show the mean D_eff values of the unconstrained GEM population. All values represent mean ± 95% CI.

Together, these findings support a model in which EDMD-associated lamin mutations impair a nucleolar-ribosomal axis that maintains cytoplasmic organization. By reducing nucleolar integrity and ribosome abundance, lamin dysfunction decreases cytoplasmic crowding and disrupts cellular organization, providing a molecular explanation for how nuclear lamina defects propagate to cause the broad cellular disorganization characteristic of EDMD.

Cytoplasmic crowding defects are specific to EDMD-associated nuclear envelope proteins

To determine whether cytoplasmic organization defects were specific to EDMD-associated proteins or represent a general consequence of nuclear envelope disruption, we examined how various nuclear envelope protein mutations affect cytoplasmic biophysical properties using GEM particle tracking. In humans, loss of emerin alone causes X-linked EDMD (1). However, in C. elegans, the emerin ortholog EMR-1 functions redundantly with another LEM-domain protein, LEM-2, during early embryogenesis (35, 36). Previous studies showed that codepletion of emr-1 and lem-2 recapitulate nuclear envelope defects and EDMD-like phenotypes observed in LMN-1–depleted animals (19, 35, 37). We found that single lem-2(tm1582) or emr-1(gk119) mutations produced minimal disruption to GEM diffusion patterns (Fig. 6, A to F, and fig. S3). However, emr-1(gk119); lem-2(RNAi) double mutants led to disruption of cytoplasmic crowding (Fig. 6, G to I). emr-1(gk119); lem-2(RNAi) animals had marked increased GEM diffusion coefficients, mirroring lamin mutants, demonstrating that combined loss of functionally redundant LEM-domain proteins resembles lamin dysfunction.

Fig. 6. Combined loss of LEM-domain proteins mimics lamin dysfunction and disrupts cytoplasmic crowding.

(A, B, and G) GMM analysis of GEM D_eff in the intestine (A and G) or hypodermis (B) of the indicated strains at day 1 of adulthood. Annotated values indicate the geometric mean of each identified peak. (C to F, H, and I) Bootstrap analysis (1000 iterations) of GMM-derived GEM diffusion parameters for the indicated strains. Bar plots (C, D, and H) show the percentage of the constrained GEM population in the intestine (C and H) or hypodermis (D); violin plots (E, F, and I) show the D_eff of the unconstrained GEM population in the intestine (E and I) or hypodermis (F). All values represent mean ± 95% CI.

We also tested canonical LINC complex components, including the inner nuclear membrane Sad1/UNC-84 (SUN) domain protein UNC-84 and the outer nuclear membrane partner, the KASH protein UNC-83 (38–40). These proteins are core constituents of the LINC complex that mediates nuclear-cytoskeletal coupling and interact with lamin (7). In mammals, LINC complex components such as SUN1, SUN2, nesprin-1, and nesprin-2 have also been implicated in EDMD (41, 42). However, animals carrying null mutations unc-84(n369) or unc-83(e1408) displayed only minimal alterations in cytoplasmic biophysical properties in both intestine and hypodermis (Fig. 6, A to F, and fig. S3).

These results indicate that disruption of individual nuclear envelope components does not substantially affect cytoplasmic organization. Across unc-84(n369), unc-83(e1408), lem-2(tm1582), and emr-1(gk119) single mutants, the percentage of constrained particles remained largely unchanged (Fig. 6, C and D), and diffusion coefficients of unconstrained particles showed only modest variations (Fig. 6, E and F). Thus, cytoplasmic crowding defects are not a general consequence of nuclear envelope dysfunction but instead represent a specific hallmark of EDMD-associated protein disruption. The pronounced effect of combined emerin/LEM-domain perturbation, contrasted with the minimal impact of other nuclear envelope components loss, highlights the distinctive requirement for EDMD-linked proteins in maintaining cytoplasmic biophysical properties.

DISCUSSION

This study uncovers a nuclear lamina-nucleolar-ribosomal axis through which defects in nuclear lamins propagate to the cytoplasm and disrupt cellular architecture. While previous studies identified connections between lamins and nucleolar function in cultured mammalian cells (26, 27), our work shows in intact tissues that this relationship directly influences cytoplasmic biophysical properties and cellular organization. Using C. elegans, we show that EDMD-associated lamin variants reduce mesoscale cytoplasmic crowding, misposition nuclei, and collapse ER network complexity through effects on nucleolar organization and ribosome abundance (Fig. 7).

Fig. 7. Nuclear lamina maintains cellular organization through a nucleolar-ribosome axis.

Schematic depicting nuclear envelope architecture and the proposed pathway linking lamina integrity to cytoplasmic organization. The nuclear lamina (LMN-1/lamin), inner nuclear membrane proteins EMR-1 (emerin; LEM-domain) and LEM-2 (LEM-domain), and LINC complex components UNC-84 (SUN protein), ANC-1 (giant KASH protein), and UNC-83 (KASH protein) are shown. Ribosomes (blue circles), actin filaments (dark green), microtubules (teal), and the ER network (gray) are depicted as indicated in the legend. Top (wild-type): An intact nuclear lamina preserves nucleolar integrity, supporting ribosome biogenesis. High ribosome abundance sustains cytoplasmic macromolecular crowding, stabilizing ER network morphology and cellular organization. Bottom [lmn-1(R64P)]: Lamina dysfunction disrupts nucleolar organization, reduces ribosome abundance, and decreases cytoplasmic crowding, leading to ER network collapse and broad cellular disorganization.

Integration with existing laminopathy models

Two prevailing models have historically shaped explanations for how lamin mutations drive pathology: (i) lamina defects weaken nuclear architecture and mechanotransduction, rendering muscle nuclei vulnerable to mechanical stress and aberrant signaling; and (ii) altered chromosome organization disrupts gene regulation, impairing tissue-specific gene activation (10). Our nuclear lamina-nucleolar-ribosome framework complements these models by revealing how nuclear defects translate into widespread cytoplasmic disorganization, a prominent but less well characterized feature of EDMD.

Rather than replacing existing models, our results suggest that the nucleolar-ribosomal axis may act downstream of both mechanical and transcriptional defects; nucleolar dysfunction could arise from impaired nuclear mechanics, altered chromatin organization, or disrupted rRNA transcription (27, 43–50). In support of this idea, we show that EDMD-linked lmn-1 mutations disrupt nucleolar integrity and reduce ribosome abundance, thereby affecting cytoplasmic biophysical properties and linking nuclear lamina defects mechanistically to cytoplasmic disorganization. Our model accounts for cytoplasmic phenotypes that earlier frameworks have not fully addressed, including altered macromolecular crowding, disrupted ER architecture and organelle mispositioning. More broadly, these findings are consistent with emerging evidence that the cytoplasm exhibits widespread mesoscale organization and regulated material properties (11, 51), and that nuclear envelope defects can perturb this organization through mechanisms extending beyond mechanics or transcription alone. Although our mTOR activation data support an upstream contribution of nucleolar-ribosomal dysfunction to the cytoplasmic phenotype, we cannot exclude the possibility that altered cytoplasmic organization feeds back on nucleolar state.

Evolutionary conservation and clinical relevance

The nucleolar-ribosome axis is highly conserved across species (52, 53), suggesting our findings in C. elegans are likely relevant to human disease. Mammalian lamins directly interact with nucleolar components and lamin mutations in patient-derived cells show nucleolar abnormalities (26, 27), resembling our observations. The conservation of ribosome-dependent cytoplasmic crowding from Escherichia coli to mammals (11, 12, 54) further supports the translational relevance of our findings.

This conservation has important clinical implications. If nucleolar dysfunction and ribosome depletion were central to EDMD pathogenesis, therapeutic strategies could target multiple points in this pathway: enhancing ribosome biogenesis, modulating mTOR signaling to restore ribosome production, or using crowding agents to compensate for reduced ribosome levels. Such approaches could potentially benefit patients with mutations in either LMNA or EMD, providing broader therapeutic options than mutation-specific strategies.

Tissue specificity and disease progression

The differential sensitivity between the intestine and hypodermis demonstrates the tissue-selective pathology characteristic of EDMD (19). Our finding that the lmn-1(Y59C) mutation affects GEM diffusion in the hypodermis but not the intestine illustrates how a ubiquitously expressed protein can produce tissue-restricted phenotypes. We propose that the hypodermal vulnerability reflects a combination of mechanical, biophysical, and epigenetic factors. First, the hypodermis occupies a structurally distinct mechanical environment: As a thin syncytium interposed between body wall muscle and the cuticle, it serves as the mechanical intermediary through which muscle contraction forces are transmitted to drive locomotion (55, 56). This places hypodermal nuclei under repetitive mechanical strain during movement, in contrast to the relatively protected intestinal cells. Consistent with this, the EDMD mutation L535P disrupts nuclear mechanical responses specifically in muscle and associated tissues, but not in other cell types (57). Second, tissues differ in their biophysical resilience and ribosome abundance (58, 59). The intestine maintains a ribosomal pool that appears to exceed its immediate translational demand. Even when ribosome biogenesis is partially compromised, the intestine may retain sufficient translational capacity to sustain normal cytoplasmic organization, whereas the hypodermis, which likely operates closer to its translational threshold, would be more vulnerable to nucleolar perturbation. Last, lmn-1(Y59C) specifically blocks the spatial relocation and activation of muscle gene promoters while leaving intestinal promoters unaffected (19), suggesting that epigenetic vulnerability is itself tissue-encoded (60). Together, tissues with high mechanical demands, limited translational reserve, and lamin-sensitive gene expression programs are predicted to be the earliest and most severely affected in EDMD (61), a prediction consistent with the clinical pattern of muscle and cardiac involvement in human disease.

Mechanistic insights from genetic interactions

The phenotypic parallels between EDMD-associated lamin mutants and ribosome depletion are notable: Both disrupt nuclear positioning alter ER morphology and reduce cytoplasmic crowding. The greater severity of lmn-1(R64P) relative to lmn-1(Y59C) is consistent with the larger polymerization defects caused by the R64P substitution (15, 19). Epistasis analyses further shows that lamin dysfunction and ribosome depletion act in a common pathway controlling cytoplasmic organization, whereas the giant KASH protein ANC-1 operates in a distinct pathway that organizes the cytoplasm by scaffolding the ER and imposing structural constraints (12). Together with the concordant reductions in nucleolar and ribosomal markers, these genetic relationships support a model in which EDMD-associated lamin mutations act upstream of ribosome biogenesis, coupling nuclear lamina dysfunction to cell-wide cytoplasmic disorganization.

Resolving the mTOR paradox

A paradox emerges when considering the hyperactivation of mTORC1 signaling in laminopathies (62). Since mTORC1 is a central driver of ribosome biogenesis (63), it is unclear how lamin dysfunction produces both elevated mTOR signaling and reduced ribosome abundance. We suggest this additional contradiction is resolved by considering that lamins contribute directly to nucleolar structural integrity (27, 44, 64), and that pathogenic lamin variants trigger ER stress responses that suppress rRNA transcription (65, 66), mechanisms that can limit ribosome production regardless of upstream mTOR activity. In this view, elevated mTOR signaling in laminopathies may represent a compensatory response to ribosomal insufficiency that is nonetheless unable to overcome nucleolar dysfunction. Consistent with this interpretation, our finding that further activating mTOR signaling through GATOR1 depletion partially restores cytoplasmic crowding indicates that the pathway retains functional capacity and that its output is limited at the nucleolar level rather than being fundamentally broken. This framework suggests that therapeutic strategies targeting nucleolar function or stress-induced suppression of rRNA transcription may be more effective complements to mTOR modulation than either approach alone.

Technical advances and broader applications

The development of GEM-based nanorheology for intact tissues represents a substantial methodological advance with applications beyond laminopathy research. Similar nuclear GEM approaches were recently applied in mammalian cells to probe nuclear mechanical properties (67). By applying tissue-specific GEM expression, we learned that C. elegans maintains distinct cytoplasmic biophysical properties: Both crowded and constrained. Our study showed that EDMD-associated lamin mutations selectively reduce the crowding of unconstrained cytoplasmic particles, suggesting a role in regulating the concentration or organization of crowding agents rather than structural barriers. Given that ribosome concentration is a primary determinant of cytoplasmic crowding in multiple species (12, 54, 68), these results strongly suggest that lamin mutations impair ribosome levels, distribution, or assembly. This approach could be applied to study cytoplasmic biophysics in other diseases involving protein aggregation, metabolic disorders affecting ribosome function, or aging-related changes in cellular organization. Direct measurement of pre-rRNA synthesis rates, for example, by single-molecule fluorescence in situ hybridization, would provide further mechanistic resolution of the nucleolar-ribosomal axis and remains an important direction for future work. The ability to quantify cytoplasmic crowding in living animals provides a new biomarker for disease severity and therapeutic response that was previously unavailable.

Therapeutic implications and future directions

Our finding that emr-1 and lem-2 function redundantly to regulate cytoplasmic crowding is relevant because these encode LEM domain proteins, including the C. elegans emerin ortholog (36), and mutations in human emerin cause X-linked EDMD (1). The emr-1; lem-2 double mutants phenocopy lamin mutations in C. elegans embryonic mitosis and viability, further supporting that they function in a common pathway (36). Thus, both major EDMD-associated proteins, lamin and emerin, contribute to regulating cytoplasmic biophysical properties.

The partial functional redundancy between emr-1 and lem-2 may explain why EDMD pathology is variable and tissue-specific, as functional compensation between nuclear envelope proteins could mask defects in most tissues. The identification of the lamina-nucleolar-ribosome axis as a common pathway affected in both lamin and emerin associated forms of EDMD suggests potential therapeutic opportunities. GEM-based nanorheology could serve as a useful tool for monitoring cellular organization in preclinical studies.

A limitation of this study is that all experiments were performed in C. elegans, which, while disease relevant, represents a simplified metazoan context. However, this system uniquely enables quantitative in vivo analysis of cellular biophysics with single-cell resolution, an approach not yet feasible in mammalian models. Future work will be needed to determine how the lamina-nucleolar-ribosomal axis is implemented in mammalian tissues and to identify additional molecular factors that modulate cytoplasmic organization. Such efforts will further clarify how nuclear architecture governs cellular organization across species and disease contexts.

In conclusion, this work reveals a mechanistic pathway through which nuclear lamina defects impair nucleolar function and ribosome biogenesis, propagating to disrupt cytoplasmic organization and whole-cell biophysics. The use of in vivo quantitative GEM-based nanorheology provides a powerful approach for measuring cellular material properties in intact tissues. Together, our findings show that both major EDMD-associated proteins, lamin and emerin, preserve cellular organization by sustaining cytoplasmic crowding, offering molecular insight into disease mechanisms and suggesting therapeutic strategies aimed at restoring ribosome function.

MATERIALS AND METHODS

C. elegans strains and maintenance

All C. elegans strains used in this study are listed in table S1. Worms were maintained on standard nematode growth medium (NGM) agar plates seeded with E. coli strain OP50 at 23°C (69). For all experiments, synchronized animals were cultured to young adult stage (24 hours post-L4) under nonstarved conditions. Experimental animals were maintained on fresh plates for a minimum of two generations before experimentation to ensure consistent physiological conditions.

RNA interference

Gene knockdown was performed using standard feeding RNAi methodology (70, 71). Bacterial strains harboring specific RNAi clones were obtained from the Ahringer RNAi library (Source Bioscience, Nottingham, UK) and verified by Sanger sequencing before use. Individual bacterial colonies were picked from fresh plates and grown overnight at 37°C in 1 ml of LB medium supplemented with carbenicillin (100 μg/ml) in 15-ml snap-cap tubes on a rotary shaker. Cultures were diluted 1:40 in 2 ml of fresh LB-carbenicillin medium and grown for 3 hours at 37°C. For dsRNA induction, 2 ml of prewarmed LB medium containing carbenicillin (100 μg/ml) and 1 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to achieve a final IPTG concentration of 0.5 mM, followed by incubation for 3 to 4 hours at 37°C. Induced bacterial cultures were spotted onto NGM agar plates containing carbenicillin (25 μg/ml) and 1 mM IPTG. Synchronized L2-L3 stage larvae were transferred to RNAi feeding plates and maintained at 23°C for 48 hours until reaching young adult stage. To achieve optimal GEM expression levels for single-particle tracking in hypodermal tissue, L3 stage worms expressing GEM-EGFP were fed bacteria expressing dsRNA targeting gfp for 48 hours until egg-laying commenced (12). L1 progenies were then transferred to standard NGM plates with OP50 bacteria and cultured until imaging age. This protocol reduced GEM concentration to appropriate levels for tracking analysis.

GEM detection and motion analysis

GEMs were imaged as previously reported (12). In brief, live imaging was performed using a Nikon Ti2 inverted microscope (Nikon Instruments, Melville, NY) equipped with a Yokogawa CSU-X1 spinning disk confocal scanner unit (Yokogawa Electric Corporation, Tokyo, Japan) and a Hamamatsu ORCA-Flash4.0 LT3 Digital sCMOS camera (Hamamatsu Photonics, Shizuoka, Japan). Images were acquired using a Nikon Plan Apo λ 100× oil immersion objective [numerical aperture (NA), 1.45] at 1060 × 1568 pixels resolution (0.065 μm/pixel). GEMs were visualized using 488-nm laser excitation at 100% power with 20 ms exposure time per frame in continuous acquisition mode (no interframe interval). Time series data were collected at 50 Hz acquisition rate for motion analysis. GEM particle detection and tracking were performed using published methods (12). In brief, trajectories shorter than 10 frames (0.2 s) were excluded from MSD analysis to ensure statistical reliability. For each valid trajectory, the effective diffusion coefficient (D_eff) was calculated by linear fitting of the first five data points (100 ms total duration) of the MSD curve. Two-component Gaussian mixture models were fitted to D_eff distributions to resolve subpopulations of constrained and unconstrained particles. Parameter distributions and confidence intervals were estimated using bootstrap resampling with 1000 iterations (12).

ER imaging and morphological analysis

ER networks were visualized using the mKate2::TRAM-1 fluorescent marker (strain UD756; table S1). Images were acquired using the Nikon Ti2 microscope system described above with 561-nm laser excitation at 50% power, 100 ms exposure time, and identical objective and camera configurations. ER occupancy and complexity analyses were carried out as previously described (12, 25). In brief, images underwent background subtraction and noise reduction before binary segmentation using Otsu’s thresholding method. Two morphometric parameters were quantified: (i) ER occupancy, defined as the ratio of ER-positive area to total cellular area, and (ii) ER network complexity, calculated as the perimeter-to-area ratio (P²/4πA), where P represents the total perimeter and A represents the total area of ER-positive regions.

Nuclear positioning measurements

Hypodermal nuclei were visualized using nuclear-localized GFP expressed from the integrated transgene ycIs10[p_col-10::nls::GFP::lacZ] (22). Young adult worms were cultured on either standard NGM plates with OP50 E. coli (controls) or RNAi feeding plates with HT115 E. coli expressing target dsRNA. Microscope slides were prepared as introduced previously (22). Nuclear positioning was assessed using an Andor BC43 microscope with 488-nm laser excitation and a 60×/1.42 oil immersion objective. To ensure systematic counting, the anus and pharynx of each worm were positioned in the same focal plane. Nuclear clustering was quantified by counting GFP-positive hypodermal nuclei in direct contact with neighboring nuclei (22, 23). Contacts along the perpendicular axis were excluded from analysis as the nuclear marker does not distinguish seam cell nuclei from hyp7 nuclear clusters. The percentage of clustered nuclei was calculated as the number of nuclei in contact divided by the total number of nuclei counted per animal.

Nucleolar measurements

Nucleoli in the hypodermal tissues were visualized using FIB-1::eGFP (72) or the endogenously tagged NUCL-1::GFP (32) in the indicated strains. Worms were synchronized to L4 stage and imaged 24 hours later as young adults. Animals were immobilized using 10 μM tetramisole in M9 buffer on 2% agarose pads using standard procedures (22). Confocal microscopy on BC43 was performed using an Andor BC43 microscope with 488-nm laser excitation (3.0% laser power, 100 ms exposure time) and a 60×/1.42 oil immersion objective. Bright-field imaging was performed concurrently for tissue localization. Nucleolar segmentation was performed using Otsu’s automatic thresholding method (73) in Fiji/ImageJ software, with size filters applied (0.5 μm² to infinity) to exclude artifacts. Regions of interest encompassing all identified nucleolar objects were generated and applied to original raw images for fluorescence intensity measurements. Mean fluorescence intensity was measured for each segmented nucleolus.

Ribosome visualization and analysis

Ribosomes were visualized using either GFP11::RPS-18, detected by split-GFP reconstitution using the endogenous allele yc113[GFP₁₁::rps-18] IV (12), or an RPL-29::GFP single copy insertion in the indicated strains (33). The complementary GFP_1-10 construct was expressed in hypodermal tissue using the transgene juEx5375[p_col-19::GFP _(1-10)+ p_ttx-3::RFP] (33), enabling reconstitution of functional GFP fluorescence specifically in the hypodermis. Young adult worms were immobilized using 10 μM tetramisole in M9 buffer on 2% agarose pads as described above. Images were acquired using the above-described Nikon Ti2 microscope system with a 100× oil immersion objective (NA 1.45) at 1060 × 1568 pixel resolution (0.065 μm/pixel). For GFP₁₁::RPS-18, fluorescence was captured using 488-nm laser excitation at 30% power with 50 ms exposure time. Because RPL-29::GFP is broadly expressed throughout the animal (33), the hypodermis was first identified by morphology before image acquisition. A single optical section through this tissue was then captured for fluorescence quantification (488-nm laser excitation, 5% laser power, 500 ms exposure). Hypodermal cell boundaries were manually delineated to define regions of interest for fluorescence intensity measurements. Average ribosome fluorescence intensity was quantified within the hypodermal syncytium.

Statistical analyses

All statistical analyses were performed using Prism GraphPad software (GraphPad Software Inc.). Sample sizes (n), degree of freedom and exact P values are reported in the figure legends. Nuclear positioning defects were analyzed using Welch’s analysis of variance (ANOVA) to account for unequal variances, followed by Dunnett’s T3 post hoc test for multiple comparisons. ER morphological defects were analyzed using Kruskal-Wallis followed by Dunnett’s T3 post hoc test for multiple comparisons. For nucleolus and ribosome quantification, two-tailed Welch’s t-test was performed. All statistical significance is indicated by asterisks (****P < 0.0001, ***P ≤ 0.001,**P ≤ 0.01, *P ≤ 0.05, and ns: P > 0.05) with exact P values shown above each comparison in each legend.

Supplementary Materials

The PDF file includes:

Table S1

Figs. S1 to S5

Legends for movies S1 and S2

Other Supplementary Material for this manuscript includes the following:

Movies S1 and S2