Mechanosensitive phosphorylation of NFATC4 at S213/S217 drives fibroblast-to-myofibroblast transition and fibrosis

Figure 1. Human lung fibroblasts activate mechanosensitive signaling pathways on stiff ECM substrates. (A) Immunofluorescence of YAP1 (green) in CCL-151 cells seeded for 2 h on fibronectin-coated substrates of indicated stiffness. Original magnification, ×40. (B) Experimental workflow for phosphoproteomic analysis of CCL-151 cells on varying substrate stiffnesses. FN, fibronectin. (C) Bar graph showing the number of quantified phosphosites per condition. (D) PCA of the stiffness-regulated phosphoproteome. (E) Pie charts showing the proportion of singly, doubly, and triply phosphorylated peptides (top) and phosphorylated S, T, and Y residues (bottom). (F) Pie chart showing the proportion of known, previously unreported, and regulatory phosphosites in PhosphoSitePlus. (G) Box-and-whisker plots of MYL9 T18/S19 phosphopeptide intensities across stiffness conditions. Immunofluorescence with phosphospecific MYL9 T18/S19 antibody (green) confirms stiffness-dependent induction; phalloidin (magenta) marks actin; DAPI (blue) marks nuclei. Original magnification, ×40. (H) Box-and-whisker plots of PXN S126 phosphopeptide intensities across stiffness conditions. Immunoblot with phosphospecific PXN S126…

Figure 2. Phosphoproteomics reveals mechanosensitive signaling pathways. (A) Heatmap of hierarchically clustered (Pearson) z scored MS intensities showing stiffness-dependent phosphorylation differences; 5 of 10 clusters exhibited clear stiffness-associated patterns (red numbers). (B) Dot plots showing scaled phosphorylation dynamics across soft to stiff substrates for selected phosphosites with known regulatory functions (PhosphoSitePlus), color coded by cluster. (C and D) IPA of phosphoproteomic data; enrichment significance shown as log P value shading for canonical pathways (C) and predicted upstream regulators (D).

Figure 3. A core mechanosensitive gene set downstream of differentially phosphorylated transcriptional regulators in vivo. (A) Analysis workflow. (B) Volcano plot of differential TF activity (healthy vs. pulmonary fibrosis fibroblasts), highlighting TFs with stiffness-dependent regulatory phosphosites (EP300, FOXO3, NFATC4, RB1, TP53, YBX1). (C) Bootstrap comparison of mean absolute TF activity differences for the stiffness-dependent TF subset versus random TF combinations. (D) Volcano plots of differential target gene expression (pulmonary fibrosis vs. controls) per TF regulon, with significantly upregulated genes highlighted (|log2FC| > 1 and FDR-adjusted P < 0.05; Mann-Whitney U test). (E) Heatmap of the 30 top upregulated NFATC4 target genes and the 5 top targets for each remaining TF in IPF versus healthy fibroblasts; color intensity reflects upregulation magnitude.

Figure 4. NFATC4, CTHRC1, and ASMA expression increases with IPF disease severity in human lung tissue. (A) Representative immunofluorescence images of lung tissue sections from control donors and IPF patients with mild (IPF1), moderate (IPF2), and severe (IPF3) disease. Left column: low-magnification overview (section diameter: 10 mm). Right columns: high-magnification images showing NFATC4 (magenta), CTHRC1 (red), ASMA (green), and DAPI (blue). The arrows indicate colocalization of NFATC4 in the nucleus. Scale bars: 50 μm. (B–D) Quantification of normalized expression levels for ASMA (B), NFATC4 (C), and CTHRC1 (D) across disease severity stages. Box-and-whisker plots show median, IQR, and whiskers to 1.5× IQR. Data represent mean ± SEM (n = 3). Statistics: *P < 0.05, **P < 0.001 versus control; Kruskal-Wallis test with pairwise Mann-Whitney U post hoc tests and Bonferroni correction for multiple comparisons.

Figure 5. NFATC4 knockdown attenuates fibroblast-to-myofibroblast transition, collagen deposition, and invasion. (A) qPCR of NFATC4, ASMA, and COL1A1 mRNA in CCL-151 transfected with siNC or siNFATC4 on stiff and soft substrates (n = 3). (B) Fluorescence intensity of intracellular ASMA, F-actin, and extracellular collagen type I, collagen type V, and Fibulin-1, normalized to DAPI (n = 6). (C) Collagen invasion assay. Confocal immunofluorescence images after 3D projection; ASMA staining (red); yellow arrow indicates an invasive cell. Scale bar: 500 μm. 3D visualization in Imaris (bottom panel). Quantification shown as relative mean ± SD from 2 replicates. (D) Collagen gel contraction in CCL-151 ± TGFB (1 ng/mL) over time (n = 3). Representative images at day 3 shown below; yellow circles outline gel pads. Statistics: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; unpaired 2-tailed Student’s t test with Bonferroni correction for multiple comparisons where applicable. Box-and-whisker plots show median, IQR, and whiskers to 1.5× IQR.

Figure 6. NFATC4 knockdown alters nuclear morphology and positioning on fibronectin micropatterns. (A) CCL-151 fibroblasts transfected with siNC or siNFATC4 were seeded on L-shaped fibronectin micropatterns and stained for pMYL9 (active myosin, green), paxillin (PXN; focal adhesions, red), and DAPI (nuclei, blue). siNC nuclei are positioned toward the adhesion-free corner and appear compact, whereas siNFATC4 nuclei are centrally located and more spheroid. Scale bar: 40 μm. (B) Schematic of the morphometric parameters used: cell area, semi-axes, orientation angle (α), circularity, and roundness. (C) Scatterplots of nucleus angle versus nucleus area for CCL-151 (siNC, n = 52; siNFATC4, n = 78) and MLT-018 (siNC, n = 108; siNFATC4, n = 100). (D) Quantification of nucleus area, circularity, and roundness in CCL-151 cells (siNC, n = 52; siNFATC4, n = 78). (E) Nucleus angle comparison between siNC and siNFATC4 in CCL-151 and MLT-018 cells. Statistics: *P < 0.05, ***P < 0.001; paired 2-tailed Student’s t test on per-run means (n = 3). Data are presented as mean ± SD. Box-and-whisker plots show median, IQR, and whiskers to 1.5× IQR. Pink lung icon, CCL-151; black lung icon, MLT-018.

Figure 7. NFATC4 phosphorylation regulates myofibroblast differentiation on soft and stiff substrates. (A) Representative immunofluorescence images of CCL-151 fibroblasts on soft (0.5 kPa) or stiff (32 kPa) substrates, transfected with control, WT NFATC4, phosphomimetic (S213D/S217D), or phospho-dead (S213A/S217A) constructs. Channels: NFATC4 (magenta), ASMA (green), DAPI (blue); arrows indicate nuclear NFATC4. Scale bars: 50 μm. (B and C) Normalized ASMA (B) and NFATC4 (C) intensity in CCL-151 cells on soft (gray) versus stiff (orange) substrates across conditions; data are presented as mean ± SD. Box-and-whisker plots show median, IQR, and whiskers to 1.5× IQR. Statistics: pairwise 2-sample z tests (n = 3 biological replicates); full results are provided in Supplemental Table 5.

Figure 8. CTHRC1 protein expression in CCL-151 and MLT-018 fibroblasts across substrate stiffnesses, NFATC4 constructs, and pharmacological treatments. (A and B) Western blots of CTHRC1 in CCL-151 and MLT-018 fibroblasts on soft (A) or stiff (B) substrates with NFATC4 constructs and inhibitors ± TGFB (5 ng/mL). (C) CTHRC1 band intensity quantification: CCL-151 soft (light gray), CCL-151 stiff (light orange), MLT-018 soft (dark gray), MLT-018 stiff (dark orange); normalized intensity values.

Figure 9. Model of the NFATC4 S213/S217 phospho-switch controlling fibroblast-to-myofibroblast transition. On soft ECM (<2 kPa), JNK signaling and TGFB/SMAD2/3 inputs are absent, NFATC4 remains unphosphorylated and cytoplasmic, and myofibroblast genes such as ACTA2 and CTHRC1 are not expressed. On stiff ECM (>8 kPa) with TGFB, mechanosensitive JNK activation phosphorylates NFATC4 at S213/S217, driving nuclear translocation and transcriptional induction of the myofibroblast program, including ACTA2 and CTHRC1, while parallel SMAD2/3 activation enhances the fibrotic gene program to firmly establish the CTHRC1+/ACTA2+ myofibroblast state. Mechanistic evidence is shown in the bottom panel: the phospho-dead mutant (S213A/S217A, blocked posttranslational modification [PTM]) fails to induce ACTA2 or CTHRC1 even on stiff ECM with TGFB; conversely, the phosphomimetic mutant (S213D/S217D, mimicked PTM) drives robust myofibroblast differentiation on soft ECM in the complete absence of TGFB. P, phosphorylation.