MeCP2-driven chromatin organization controls nuclear stiffness

Abstract

Cellular differentiation is driven by epigenetic modifiers and readers, including the methyl CpG binding protein 2 (MeCP2), whose level and mutations cause the neurological disorder Rett syndrome. During differentiation, most of the genome gets densely packed into heterochromatin, whose function has been simplistically viewed as gene silencing. However, gene expression changes reported in mutations leading to Rett syndrome have failed to be a predictor of disease severity. Here we show that MeCP2 increases nuclear stiffness in a concentration-dependent manner and dependent on its ability to cluster heterochromatin during differentiation. MeCP2-dependent stiffness increase could not be explained by changes in the expression of mechanobiology-related genes, but we found that it is disrupted by Rett syndrome mutations and correlated with disease severity. Our results highlight the impact of chromatin organization on the mechanical properties of the cell as an alternative or complementary mechanism to changes in cytoskeleton components.

Fig. 1. Nuclear stiffness does not depend on the cytosolic components.

a Scheme of the experiment. C2C12 myoblast nuclei were extracted to remove the cytosolic cytoskeleton. Atomic force microscopy force maps were acquired for the cells and the purified nuclei. The purified nuclei were seeded on top of a 0.5% agarose gel pad. To analyze the data, a mask was used to depict the cell, the nucleus area of the cell, or the purified nucleus. b Quantification of the nuclear volume in the different steps of the preparation. 3D confocal analysis of DAPI staining for cells seeded and fixed on glass coverslips (n = 35), isolated nuclei deposited on glass coverslips (n = 81), or seeded on 0.5% agarose pads (n = 35). Individual nuclei (n = 7) were selected from the atomic force microscopy images, and volume was calculated to confirm the results obtained in the 3D analysis. To assess the significance, a 2-sided t-test was performed to calculate a p value. Only significant differences are shown. *: p value < 0.05; **: p value < 0.001; ***: p value < 0.0001. c Histogram of elastic modulus values for cells grown on plates, showing measurements from the entire cell area (upper panel, n = 3) and from the nuclear region only (middle panel, n = 3). The lower panel (light gray histogram) shows data from isolated nuclei seeded on agarose (n = 37), with the agarose stiffness distribution shown in dark gray for comparison (n = 4). The median values for each condition are represented in dashed colored lines (blue: cell, nuclear area of the cell and purified nuclei, dark red: agarose) d Scheme of the probe–nucleus–agarose interface during a force spectroscopy measurement. Color gradients (red → yellow → green → blue) qualitatively represent the stress distribution across the interface, from high to low. Due to the forces (arrows) mainly from the agarose substrate, we performed a bottom effect cone correction in the elastic modulus calculation. Adapted from ref. .

Fig. 2. Heterochromatin clustering driven by MeCP2 increases nuclear stiffness.

a Visualization of the heterochromatin compartments by DAPI staining (high-density foci) in C2C12 myoblasts expressing different levels of MeCP2. The levels were defined by fluorescence-activated cell sorting (FACS), following the gate scheme depicted in the up. Scale: 5 µm. b Clustering analysis including the mean and 95% confidence interval of the number of heterochromatin compartments and the average volume of the heterochromatin compartments per cell calculated by 3D confocal microscopy in fixed cells on coverslips and stained with DAPI (n cells used: mb = 35; MeCP2 low = 13; MeCP2 high = 10). The violin plot representing the individual measurements is shown in Supplementary Fig. 1. c Elastic modulus distribution obtained from several nuclei per fraction, corresponding to different MeCP2 expression levels separated by FACS: mb (n = 35), low MeCP2 fraction (n = 48) and high MeCP2 fraction (n = 20). A Gaussian mixture model (GMM) with three components was used to highlight differences between conditions. The black line represents the overall model fit; cyan, blue, and red lines indicate the individual subpopulations. Dashed lines mark the mean of each population, and the colored numbers denote the relative weight of each subpopulation (in %), corresponding to the area under the curve.

Fig. 3. Deletion of MeCP2 severely impairs the increase of stiffness during neuronal differentiation.

a Differentiation model from embryonic stem cells (ESC) to neurons by leukemia inhibitory factor (LIF) deprivation, with an overview of the renewal abilities (green circle) and the MeCP2 levels after 7, 14, and 21 days (D7, D14, and D21, respectively). b Clustering analysis for wt and MeCP2 cells differentiated from ESC by LIF deprivation. 3D confocal analysis was performed in cells fixed and stained with DAPI (n cells used in each condition: wt D7 = 39; wt D14 = 16; wt D21 = 38; MeCP2 KO D7 = 28; MeCP2 KO D14 = 32, MeCP2 KO D21 = 19). The plot represents the mean and 95% confidence interval of each condition. The distribution of the data and the individual values can be found in Supplementary Fig. 3. c Comparison of the elastic modulus distribution in wt (blue) and MeCP2 KO (magenta) nuclei from cells differentiated from ESC by LIF deprivation. Nuclei used per condition: wt D7 = 44; wt D14 = 34; wt D21 = 48; MeCP2 KO D7 = 47; MeCP2 KO D14 = 40; MeCP2 KO D21 = 37. A Gaussian mixture model was applied for each dataset for a more quantitative view in Supplementary Fig. 3. d Differentiation from ESC to neurons by generating stable neural stem cells (NSCs). e Clustering analysis for wt and MeCP2 cells differentiated from stable NSCs. 3D confocal analysis was performed in cells fixed and stained with DAPI (n cells used in each condition: wt NSC = 85; wt neuron = 61; MeCP2 KO NSC = 87; MeCP2 KO neuron = 61). The plot represents the mean and 95% confidence interval of each condition. The distribution of the data and individual points can be found in Supplementary Fig. 3. f Comparison of the elastic modulus distributions in wt (blue) and MeCP2 KO (magenta) nuclei from cells differentiated from NSCs. Nuclei used in each condition: wt NSC = 18; wt neuron = 40; MeCP2 KO NSC = 29; MeCP2 KO neuron = 34. A Gaussian mixture model was applied for each dataset for a more quantitative view in Supplementary Fig. 3.

Fig. 4. Rett syndrome mutations of MeCP2 impair the increase in nuclear stiffness.

a Scheme of the main domains of the MeCP2 protein (MBD: methyl binding domain; TRD: transcription repression domain) and the location of the mutations studied. b Clustering analysis using 3D confocal microscopy on fixed C2C12 cells, untransfected or transfected with plasmids containing wt or mutant MeCP2 cDNA. Cells were fixed with formaldehyde and stained with DAPI (n number used for each condition: mb = 35, wt MeCP2 = 40; P101H = 15; R106W = 13; R133C = 20; A140V = 14; T158M = 14; R168X = 7; R255X = 20; R270X = 29; R294X = 19). Only transfected cells were analyzed; however, their expression levels were heterogeneous, which is reflected in the variability observed in the atomic force microscopy results. The mean and 95% confidence interval of the number of heterochromatin compartments and the average volume of the compartments per cell are represented. The violin plots containing the individual data are shown in Supplementary Fig. 5b, c. c Representative histogram of the elastic modulus with Gaussian mixture model (GMM) fits. Nuclei were purified from untransfected C2C12 cells (mb ut), seeded on an agarose pad, and analyzed by atomic force microscopy to determine their elastic modulus. A three-component GMM was applied, showing the overall model fit (black line) and the individual subpopulations (cyan, blue, and magenta lines). Histograms of the elastic modulus values for wild type MeCP2 and mutant nuclei are provided in Supplementary Fig. 5d. d Dendrogram showing the effect of Rett mutations on myoblasts based on elastic modulus values distribution. Elastic modulus histograms were fitted with a GMM as described above to obtain the mean modulus values of each subpopulation (Supplementary Fig. 5g). In addition, in order to make the populations comparable, all data were pulled together and fitted to 3 populations, followed by the assignment of the individual data to the three populations based on k-means to obtain the weight of the populations (see Supplementary Fig. 5f). All population means and weights were normalized using z-scores, and Euclidean distances were calculated and represented in a dendrogram. The mutants were classified into mild (gray) and severe (bold) based on previous publications. R270X is variably classified as mild or severe in the literature, likely due to differences in clinical scoring parameters and diagnostic criteria applied across cohorts. e Relationship between heterochromatin organization and stiffness contributions of nuclear subpopulations. Linear regressions showing the relationship between the heterochromatin organization index (HOI, log₁₀-transformed) and the weighted stiffness of the soft (left), mid (center), and stiff (right) nuclear fractions across MeCP2 mutants. Weighted stiffness values were calculated by multiplying the proportion of each population by its representative elastic modulus acquired from k-means clustering (1.7, 7.4, and 40.7 kPa for soft, mid, and stiff, respectively). The analysis revealed a negative trend for the soft fraction (β = –0.68, R² = 0.20), a weak positive trend for the mid fraction (β = 1.50, R² = 0.09), and a significant positive relationship for the stiff fraction (β = 7.91, R² = 0.40, p = 0.037), indicating that increased heterochromatin organization (higher HOI) is associated with greater mechanical stiffening of nuclei. f. Dendrogram of the Rett mutant rescue of NSC MeCP2 KO based on the elastic modulus distribution. The procedure was done as described for d, based on GMM populations (Supplementary Fig. 5g, h) and the k-means weights (Supplementary Fig. 5i).

Fig. 5. MeCP2 regulates the expression of the mechanotransduction-related genes Notch2 and Tgfbr1.

Volcano plots displaying the changes in overall expression of genes by the deletion of MeCP2 (a), or the mutation R106W (b), or T158M (c). Blue and red represent down- and up-regulated genes, respectively, with a false discovery rate (FDR) > 0.05. d Gene arrays for the main components of the nucleoskeleton and mechanotransduction pathways and their change due to MeCP2 deletion or mutation. Black outline: significant changes; black boxes: no reads available for these loci in the dataset. e Analysis of the expression of relevant genes in qRT-PCR in neural stem cell (NSC) differentiation. The bar plot shows the average and lines the standard deviation of three biological replicates, each of them done with three technical replicates. f Chromatin immunoprecipitation analysis of MeCP2 in 20 kb around the transcription start site (TSS) of the genes studied in d, as well as specifically for Tgfbr1 and Notch2.