Submicron traction-force mapping reveals mitochondrial regulation of nanoscale force coordination during rigidity sensing under oxidative stress Authors: Muhammad Anwar Ul Hassan, Seok Gi Kim, Dan Bi Im, Jung Seung Lee, Gwang Lee, Sungsu Park Published: 2026 ABSTRACT Oxidative stress is closely associated with aging, fibrosis, and cancer progression, yet how oxidative stress affects rigidity sensing remains poorly understood. Here, we employed submicron micropillar-based traction-force microscopy to investigate oxidative stress-associated mechanotransduction dysfunction in mouse embryonic fibroblasts (MEFs). Sublethal H2O2 exposure impaired stiffness-dependent cytoskeletal remodeling, focal adhesion maturation, and traction-force generation during cell spreading. Quantitative force mapping further revealed disrupted contraction-unit (CU)-associated nanoscale force coordination at the cell periphery, characterized by increased alignment of neighboring traction-force vectors and reduced opposing contractions. These defects were accompanied by fragmented mitochondrial networks and altered mitochondrial dynamics-associated gene expression. Importantly, mitofusin (MFN)-deficient MEFs phenocopied oxidative stress-induced impairments in stiffness-responsive spreading and traction-force generation, identifying mitochondrial network integrity as a critical regulator of rigidity sensing. Together, these findings establish a mechanistic link between oxidative stress, mitochondrial dynamics, and nanoscale force coordination during rigidity sensing, while highlighting the utility of submicron traction-force mechanophenotyping for quantitative analysis of oxidative stress-associated mechanobiological dysfunction.Graphical abstract Supplementary InformationThe online version contains supplementary material available at 10.1186/s40580-026-00570-0. Introduction Cells sense and respond to the mechanical properties of the extracellular matrix (ECM) through mechanotransduction, a process essential for cellular differentiation, tissue development, and wound healing [1–3]. This process arises from coordinated coupling between the actin cytoskeleton and the ECM via focal adhesions (FA), enabling cells to probe substrate stiffness through force generation and adhesion remodeling [4–6]. A key outcome of this coupling is rigidity sensing, in which cells adjust spreading, adhesion maturation, and traction-force according to matrix stiffness. Because rigidity sensing fundamentally depends on cellular force transmission, quantitative assessment of traction forces is essential for evaluating mechanobiological behavior. Submicron micropillar-based traction-force microscopy enables nanoscale mechanophenotyping of cellular rigidity sensing by resolving localized force transmission during mechanotransduction [7–11]. Our previous studies demonstrated that nanoparticle-induced oxidative stress impairs key aspects of rigidity sensing, including FA dynamics, traction-force generation, and stiffness-dependent mechanotransduction [12–14]. These impairments were associated with reactive oxygen species (ROS) generation [15, 16] and concomitant reductions in intracellular ATP levels. Excessive ROS disrupts cellular homeostasis and promotes senescence, in part through mitochondrial dysfunction characterized by fragmented mitochondrial networks and reduced ATP production [17–21]. Because actin remodeling, adhesion dynamics, and contractile force generation are energy-dependent processes, mitochondrial dysfunction provides a plausible mechanistic link between oxidative stress and impaired cellular mechanics [22, 23]. In addition to ATP supply, mitochondrial network organization can influence the spatial distribution of energy within cells, which is particularly relevant for peripheral actomyosin activity and FA reinforcement during cell spreading [24, 25]. However, how mitochondrial network integrity governs force-based rigidity sensing under oxidative stress has not been systematically investigated. Here, we investigated how oxidative stress alters rigidity sensing in MEFs using a submicron micropillar-based nanoscale force-mapping approach. To further examine the role of mitochondrial network integrity, we employed MFN-deficient MEFs as a complementary genetic model of disrupted mitochondrial fusion [26, 27]. We also included superoxide dismutase 1 (SOD1) co-treatment as an antioxidant condition to evaluate the ROS-dependent component of the oxidative stress response. Together, these approaches demonstrate that oxidative stress impairs stiffness-dependent mechanotransduction and nanoscale force coordination through disruption of mitochondrial network integrity, leading to defective cytoskeletal remodeling and reduced traction-force generation during cell spreading (Fig. 1).Fig. 1Overview of oxidative stress–induced disruption of rigidity-dependent mechanotransduction and micropillar-based force readouts. A Conceptual schematic comparing untreated and oxidatively stressed cells. In untreated cells, actin–focal adhesion (FA) coupling supports cytoskeletal tension and rigidity sensing, resulting in enhanced spreading and FA reinforcement on rigid substrates. An intact mitochondrial network provides ATP to sustain these force-generating processes. Under oxidative stress, increased ROS and reduced ATP are associated with mitochondrial fragmentation and weakened FA–cytoskeletal coupling, leading to reduced tension and loss of stiffness-dependent differences in cell spreading. B Schematic of submicron PDMS micropillar arrays used to quantify traction-forces based on pillar deflection (D, diameter; h, height; Δx, displacement; F, traction-force). C Definition of the directionality parameter (γ) used to quantify local edge contractions. γ ≈ 0 indicates opposing neighboring vectors (paired inward contractions), whereas γ ≈ 1 indicates aligned vectors, reflecting reduced opposing contractions and impaired contraction-unit (CU) activity Overview of oxidative stress–induced disruption of rigidity-dependent mechanotransduction and micropillar-based force readouts. A Conceptual schematic comparing untreated and oxidatively stressed cells. In untreated cells, actin–focal adhesion (FA) coupling supports cytoskeletal tension and rigidity sensing, resulting in enhanced spreading and FA reinforcement on rigid substrates. An intact mitochondrial network provides ATP to sustain these force-generating processes. Under oxidative stress, increased ROS and reduced ATP are associated with mitochondrial fragmentation and weakened FA–cytoskeletal coupling, leading to reduced tension and loss of stiffness-dependent differences in cell spreading. B Schematic of submicron PDMS micropillar arrays used to quantify traction-forces based on pillar deflection (D, diameter; h, height; Δx, displacement; F, traction-force). C Definition of the directionality parameter (γ) used to quantify local edge contractions. γ ≈ 0 indicates opposing neighboring vectors (paired inward contractions), whereas γ ≈ 1 indicates aligned vectors, reflecting reduced opposing contractions and impaired contraction-unit (CU) activity Methods/experimental Cell culture Wild type MEFs and MFN-deficient MEFs were used, including Mfn1-null MEFs (MFN1-/-; ATCC CRL-2992), Mfn2-null MEFs (MFN2-/-; ATCC CRL-2993) and Mfn1/Mfn2-null MEFs (DKO; ATCC CRL-2994). All cell lines were maintained in high glucose Dulbecco’s modified Eagle’s medium (DMEM) from HyClone Laboratories (Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS) (Corning, NY, USA), 100 units/ mL penicillin and 100 μg/ ml streptomycin (Gibco). Cells were incubated under humid conditions at 37 °C in 5% CO2. For cell passaging and collection, cultures were treated with 0.1% trypsin–EDTA (Gibco) at 37 °C and 5% CO2 for 3 min to detach cells. Following this, the cells were centrifuged at 1300 rpm for 3 min prior to downstream assays. H2O2 treatment and senescence analysis For oxidative stress induction, MEFs were seeded in 96-well plates (104 cells/well) and treated with H2O2 (Sigma-Aldrich, St. Louis, MO, USA) at 50, 100, 200 and 400 µM for 12 h following a published oxidative-stress induction protocol with minor modifications [28]. Based on our dose–response assessments of cell viability, senescence, and ATP levels, 200 µM H2O2 was selected for subsequent experiments. Cells were fixed with 2% paraformaldehyde for 10 min, rinsed with phosphate-buffered saline (PBS, pH 7.4) and stained with the CellEvent™ Senescence Green Detection Kit (Thermo Fisher Waltham, MA, USA) according to the manufacturer's instructions. Senescent cells were identified based on senescence-associated β-galactosidase (SA-β-gal) fluorescence using confocal laser scanning microscopy (CLSM; K1-Fluo) from Nanoscope systems (Daejeon, Korea). Senescence was quantified as the percentage of SA-β-gal–positive cells determined by cell counting. Western blotting MEFs were seeded in 6-well plates (5 × 105 cells/well) and cultured overnight. Cells were then treated with H2O2 as indicated and lysed in ice-cold radioimmunoprecipitation assay (RIPA) buffer (Thermo Fisher Scientific, Waltham, MA, USA). Lysates were vortexed and incubated at 4 °C for 1 h, followed by centrifugation at 14,000 × g for 15 min at 4 °C. The supernatants were collected, and total protein concentration was quantified using the BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein (40 µg) were separated by SDS–PAGE and transferred onto nitrocellulose membranes. Membranes were blocked with 3% (w/v) nonfat milk in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 1 h at RT and incubated with primary antibodies overnight at 4 °C. The following primary antibodies were used: total FAK (1:2000; Cell Signaling Technology), phospho-FAK (Y397) (1:1000; Abcam), and GAPDH (1:1000; Cell Signaling Technology). After washing with TBST, membranes were incubated with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG secondary antibody (1:8000; Cell Signaling Technology) for 1 h at RT. Protein bands were detected using enhanced chemiluminescence (ECL) reagent (Thermo Fisher Scientific). SOD1 co-treatment MEFs were co-treated with SOD1 (S5395, Sigma-Aldrich) at 1 µM and 200 µM H2O2 for 12 h at 37 °C in 5% CO2. H2O2 and SOD1 stock solutions were freshly prepared prior to each experiment, and treatments were performed with minimal light exposure. Following treatment, cells were subjected to downstream mechanotransduction and mitochondrial analyses as described in the respective sections. Immunocytochemistry MEFs were seeded on substrates with different stiffnesses (soft, 5 kPa; rigid, 3 GPa). Prior to cell seeding, substrates were functionalized with 10 µg/mL fibronectin (Sigma-Aldrich) at 37 °C in 5% CO2 for 60 min. H2O2-treated and SOD1 co-treated cells were cultured on the substrates for 1 or 6 h before fixation. Cells were fixed with 4% paraformaldehyde for 15 min at RT and permeabilized with PBS containing 0.2% Triton X-100 (Sigma-Aldrich) for 20 min. Cells were then blocked with 1% bovine serum albumin (BSA) for 1 h. F-actin was stained using Acti-stain™ 555 phalloidin (Cytoskeleton Inc., Denver, CO, USA) diluted 1:200 in PBS for 1 h at RT. Nuclei were stained with 0.2 µg/mL 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) in PBS for 30 min at RT. For FA analysis, cells were incubated with rabbit anti-paxillin antibody (Abcam) overnight at 4 °C, followed by incubation with Alexa Fluor 488-conjugated anti-rabbit IgG secondary antibody (A-11034; Invitrogen) diluted 1:1000 in 2% BSA for 45 min at RT. Fluorescence images were acquired using CLSM. Cell spreading area was quantified using ImageJ (NIH, Bethesda, MD, USA). Filopodia were analyzed using FiloQuant, a FIJI/ImageJ plugin. FA area was quantified using FIJI/ImageJ software. RT-qPCR Total RNA was extracted from wild type MEFs and MFN-deficient MEFs using an RNA isolation kit (Takara Bio Inc., Shiga, Japan). Isolated RNA was converted to complementary DNA (cDNA) by using a reverse transcription kit (Takara Bio Inc.) according to the manufacturer's instructions. RT-qPCR was performed using TaqMan Fast Universal PCR Master Mix (Applied Biosystems, Waltham, MA, USA) on a StepOnePlus system (Applied Biosystems). Relative quantification of gene expression was calculated using the 2−ΔΔCt method. TaqMan assay information used for RT-qPCR is provided in Table 1.Table 1Predesigned TaqMan gene expression assays used for RT-qPCRGene nameSymbolSpeciesTaqMan assay IDMitofusin 1Mitofusin 2OPA1Dynamin-1-like (Drp1)Glyceraldehyde-3-phosphate dehydrogenaseMfn1Mfn2Opa1Dnm1lGapdhMouseMouseMouseMouseMouseMm00612599_m1Mm00500120_m1Mm01349707_g1Mm01342903_m1Mm99999915_g1 Table 1Predesigned TaqMan gene expression assays used for RT-qPCRGene nameSymbolSpeciesTaqMan assay IDMitofusin 1Mitofusin 2OPA1Dynamin-1-like (Drp1)Glyceraldehyde-3-phosphate dehydrogenaseMfn1Mfn2Opa1Dnm1lGapdhMouseMouseMouseMouseMouseMm00612599_m1Mm00500120_m1Mm01349707_g1Mm01342903_m1Mm99999915_g1 Predesigned TaqMan gene expression assays used for RT-qPCR Mitofusin 1 Mitofusin 2 OPA1 Dynamin-1-like (Drp1) Glyceraldehyde-3-phosphate dehydrogenase Mfn1 Mfn2 Opa1 Dnm1l Gapdh Mouse Mouse Mouse Mouse Mouse Mm00612599_m1 Mm00500120_m1 Mm01349707_g1 Mm01342903_m1 Mm99999915_g1 Mitochondrial network staining and morphology analysis To visualize mitochondrial morphology, MEFs were stained with MitoTrackerTM Deep Red FM (M22426; Thermo Fisher Scientific) (excitation/emission: 644/665 nm). A working solution was prepared in pre-warmed serum-free medium, and live cells were incubated with the dye at 50 nM for 30 min at 37 °C in 5% CO2. Following incubation, cells were rinsed with warm PBS and imaged in fresh medium. Fluorescence images were acquired using CLSM. Mitochondrial network morphology was quantified using the MiNA plugin in FIJI/ImageJ [29]. Confocal images were analyzed on a per-cell basis by selecting individual cells as regions of interest to exclude background fluorescence. The MiNA workflow was applied to separate mitochondrial signals from background through thresholding and binarization, followed by generation of skeletonized mitochondrial networks from the binary images. Mitochondrial branch length was quantified from the skeletonized network, whereas mitochondrial footprint was quantified from the thresholded mitochondrial area prior to skeletonization. The primary outcomes reported were mitochondrial branch length and mitochondrial footprint (total mitochondrial area per cell), reflecting mitochondrial network connectivity and coverage, respectively. Fabrication of soft and rigid substrates Soft PDMS substrates were fabricated using Sylgard Silicone Elastomer Kit (Dow Corning, Midland, MI, USA) as previously described [9, 12]. The silicone elastomer base was thoroughly mixed with its curing agent at a ratio of 75:1 (base: curing agent, w/w), which has been previously characterized to produce substrates with an elastic modulus of approximately 5 kPa [12]. The PDMS mixture was degassed for 30 min, spin-coated onto glass coverslips at 2000 rpm for 2 min to produce a PDMS layer with a thickness of 35 ± 5 µm, and subsequently cured at 70 °C for 4 h prior to use. Glass coverslips were used as rigid substrates with an elastic modulus of approximately 3 GPa. PDMS micropillar fabrication Micropillar arrays were fabricated from PDMS using a photolithography-generated silicon wafer mold containing an array of cylindrical holes (0.9 µm diameter, 2 µm depth, and 1.8 µm center-to-center spacing) as previously described [12]. PDMS base and curing agent were mixed at a ratio of 10:1 (base: curing agent, w/w) to produce PDMS with a Young’s modulus of approximately 2 MPa. The PDMS mixture was degassed for 15 min and spin-coated onto the silicon mold (10 s at 500 rpm followed by 30 s at 1800 rpm). The coated mold was subsequently degassed for an additional 30 min to remove trapped air bubbles and cured at 80 °C for 3 h. The pillars fabricated from this mold exhibited a bending stiffness (k) of 24.21 nN/μm, which was calculated using the Euler–Bernoulli beam theory (Eq. (1)) [30].1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ k = \frac{3}{64}\pi E\frac{{D^{4} }}{{L^{3} }} $$\end{document}where E, D and L represent the Young’s modulus, diameter and height of PDMS micropillars, respectively. Traction-force and rigidity sensing measurement Micropillar arrays were functionalized with fibronectin prior to cell seeding. Cells were observed within 1 h incubation (37 °C, 5% CO2) using an inverted microscope (Nikon ECLIPSE TE2000-U), controlled by SpinView software. The displacement of each pillar caused by the cells was quantified using the PillarTracker version 1.1.3 software from Mechanobiology Institute (MBI) at National University of Singapore, which employs a reconstruction method to create an accurate grid for assessing pillar deflections [12]. Pillars outside the cell boundary were used as reference points for the zero-force position. Each deflection (x) measurement was then multiplied by the pillar’s bending stiffness (k) to determine the traction-force based on Hooke’s law (Eq. (2)).2 F = k × x To measure cellular interactions with the substrate, the directional parameter (γ) was used. To calculate γ, the force vectors at the cell's leading edges were measured during initial contact period (30 min). The sum of the force vectors was then divided by their magnitudes (Eq. 3) [12].3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \gamma = \frac{{\sqrt {\left( {Ax + Bx} \right)^{2} + \left( {Ay + By} \right)^{2} } }}{{\sqrt {Ax^{2} + Ay^{2} } + \sqrt {Bx^{2} + By^{2} } }} $$\end{document}where A and B are the two neighboring force vectors. Statistical analysis and data correction Data are presented as mean ± standard deviation (SD). For comparisons between two experimental groups, an unpaired two-tailed Student’s t-test was used. For comparisons among multiple (≥ 3) experimental groups, one-way analysis of variance (ANOVA) was performed, followed by Tukey’s post hoc multiple-comparisons test. For quantitative morphology analyses, n = 30 denotes individual cells quantified per condition from three independent biological replicates. Data are presented as mean ± SD. For time-dependent morphology analyses, three-way ANOVA was performed with treatment (or genotype), substrate stiffness, and time as factors, including all interaction terms. For time-specific stiffness-response analyses, data obtained at 1 h and 6 h were analysed separately using two-way ANOVA with treatment (or genotype) and substrate stiffness as factors. Tukey-adjusted multiple-comparisons tests were applied following three-way ANOVA, whereas Šídák-adjusted multiple-comparisons tests were applied following two-way ANOVA. To reduce errors in micropillar experiments, corrections were made by reducing the mean deflection observed in the pillars located outside the cellular periphery (12). Results Oxidative stress induces senescence and impairs cytoskeletal and adhesion remodeling To establish an oxidative stress condition that induces senescence without compromising viability, MEFs were treated with increasing concentrations of H2O2 and evaluated for cell viability and senescence. Live/dead staining showed preserved viability at 100 and 200 µM H2O2, whereas 400 µM markedly increased cell death (Fig. 2A, B). In contrast, senescence-associated β-galactosidase staining revealed a pronounced increase in senescent cells at 200 µM, supporting this dose as an optimal condition for inducing senescence (Fig. 2A, C). Cellular ATP levels decreased in a dose-dependent manner with increasing H2O2 concentrations (Fig. 2D). In parallel, phosphorylation of focal adhesion kinase (p-FAK) was reduced, indicating impaired adhesion associated mechanotransduction signaling (Fig. 2E).Fig. 2Oxidative stress induces senescence and impairs cytoskeletal organization, focal adhesion (FA) formation, and ATP production in MEFs. A Representative Live/Dead staining with 4′,6-diamidino-2-phenylindole (DAPI) in MEFs treated with increasing H2O2 concentrations. Representative SA-β-gal staining with DAPI is also shown (scale bar: 200 μm). B, C Quantification of cell viability B MEFs treated with increasing and percentage of senescent cell C under the indicated H2O2 conditions (n > 300 cells). D Cellular ATP luminescence measured across increasing H2O2 concentrations. E Immunoblot analysis of phosphorylated FAK (p-FAK, Y397) and total FAK (t-FAK) with GAPDH as a loading control across increasing H2O2 concentrations. F Representative fluorescence images of cytoskeletal and adhesion features in untreated and 200 µM H2O2-treated MEFs, including F-actin staining, filopodia detection, skeletonized actin representation for branch analysis, and paxillin staining for FAs (scale bar: 30 μm). G–J Quantification of cell spreading area G, H filopodia number per cell, I average actin branch length, and J total FA area per cell (n = 30). Each dot represents a single cell; bars indicate mean ± SD. Statistical significance is indicated as ns (not significant), *P < 0.05, **P < 0.01, and ***P < 0.001. One-way ANOVA with Tukey’s multiple comparisons test was used for B-D, and unpaired t-tests were used for G-J Oxidative stress induces senescence and impairs cytoskeletal organization, focal adhesion (FA) formation, and ATP production in MEFs. A Representative Live/Dead staining with 4′,6-diamidino-2-phenylindole (DAPI) in MEFs treated with increasing H2O2 concentrations. Representative SA-β-gal staining with DAPI is also shown (scale bar: 200 μm). B, C Quantification of cell viability B MEFs treated with increasing and percentage of senescent cell C under the indicated H2O2 conditions (n > 300 cells). D Cellular ATP luminescence measured across increasing H2O2 concentrations. E Immunoblot analysis of phosphorylated FAK (p-FAK, Y397) and total FAK (t-FAK) with GAPDH as a loading control across increasing H2O2 concentrations. F Representative fluorescence images of cytoskeletal and adhesion features in untreated and 200 µM H2O2-treated MEFs, including F-actin staining, filopodia detection, skeletonized actin representation for branch analysis, and paxillin staining for FAs (scale bar: 30 μm). G–J Quantification of cell spreading area G, H filopodia number per cell, I average actin branch length, and J total FA area per cell (n = 30). Each dot represents a single cell; bars indicate mean ± SD. Statistical significance is indicated as ns (not significant), *P < 0.05, **P < 0.01, and ***P < 0.001. One-way ANOVA with Tukey’s multiple comparisons test was used for B-D, and unpaired t-tests were used for G-J We next investigated whether these changes were accompanied by alterations in cytoskeletal organization and adhesion structures. Compared with untreated cells, H2O2-treated MEFs exhibited markedly fewer filopodial protrusions (Fig. 2F). Although overall cell spreading area was not significantly altered (Fig. 2G), filopodia number per cell was significantly reduced (Fig. 2H). Consistent with impaired cytoskeletal remodeling, skeleton-based analysis revealed reduced actin branch length (Fig. 2I). Furthermore, paxillin staining demonstrated a significant decrease in total FA area per cell (Fig. 2J). Together, these results demonstrate that senescence-inducing oxidative stress reduces ATP production and FAK phosphorylation, while impairing cytoskeletal protrusion formation and FA structures, without significantly affecting overall cell spreading. Oxidative stress impairs stiffness-dependent spreading and FA remodeling To assess whether oxidative stress compromises rigidity sensing, untreated, H2O2-treated and SOD1 co-treated MEFs were seeded on soft (5 kPa) and rigid (3 GPa) substrates [9, 12] and analyzed during early (1 h) and late (6 h) spreading. In untreated cells, both F-actin organization and paxillin-positive FAs were enhanced on rigid substrates compared to soft substrates at both time points, consistent with stiffness-dependent cytoskeletal and adhesion remodeling (Fig. 3A, B).Fig. 3Oxidative stress impairs stiffness-dependent cytoskeletal and focal adhesion remodeling in MEFs. A, B Representative fluorescence images of F-actin (red) and paxillin in untreated, H2O2-treated and SOD1 + H2O2 co-treated MEFs plated on soft (5 kPa) and rigid (3 GPa) substrates and fixed at A 1 h and B 6 h after seeding. Untreated cells exhibited enhanced cytoskeletal organization and larger paxillin-positive FAs on rigid substrates, whereas H2O2-treated cells displayed similar actin/adhesion organization regardless of substrate stiffness. SOD1 co-treatment partially restored stiffness-associated cytoskeletal and adhesion remodeling (scale bar: 30 μm). C–E Quantification of time-dependent changes in cell spreading area C, cell aspect ratio D, and FA area E. F–H Quantification at 1 h of cell spreading area F, cell aspect ratio G and H FA area across conditions and substrate stiffness. I-K Quantification at 6 h of I spreading area, aspect ratio J and FA area K. For all quantitative morphology analyses, n = 30 denotes individual cells quantified per condition from three independent biological replicates. Each bar represents mean ± SD. Time-dependent data in panels C–E were analysed using three-way ANOVA with treatment, substrate stiffness, and time as factors, including interaction terms, followed by Tukey-adjusted multiple-comparisons tests. Data in panels F–H and I–K were analysed separately at 1 h and 6 h, respectively, using two-way ANOVA with treatment and substrate stiffness as factors, followed by Šídák-adjusted multiple-comparisons tests. Statistical significance is indicated as ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001 Oxidative stress impairs stiffness-dependent cytoskeletal and focal adhesion remodeling in MEFs. A, B Representative fluorescence images of F-actin (red) and paxillin in untreated, H2O2-treated and SOD1 + H2O2 co-treated MEFs plated on soft (5 kPa) and rigid (3 GPa) substrates and fixed at A 1 h and B 6 h after seeding. Untreated cells exhibited enhanced cytoskeletal organization and larger paxillin-positive FAs on rigid substrates, whereas H2O2-treated cells displayed similar actin/adhesion organization regardless of substrate stiffness. SOD1 co-treatment partially restored stiffness-associated cytoskeletal and adhesion remodeling (scale bar: 30 μm). C–E Quantification of time-dependent changes in cell spreading area C, cell aspect ratio D, and FA area E. F–H Quantification at 1 h of cell spreading area F, cell aspect ratio G and H FA area across conditions and substrate stiffness. I-K Quantification at 6 h of I spreading area, aspect ratio J and FA area K. For all quantitative morphology analyses, n = 30 denotes individual cells quantified per condition from three independent biological replicates. Each bar represents mean ± SD. Time-dependent data in panels C–E were analysed using three-way ANOVA with treatment, substrate stiffness, and time as factors, including interaction terms, followed by Tukey-adjusted multiple-comparisons tests. Data in panels F–H and I–K were analysed separately at 1 h and 6 h, respectively, using two-way ANOVA with treatment and substrate stiffness as factors, followed by Šídák-adjusted multiple-comparisons tests. Statistical significance is indicated as ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001 In contrast, H2O2-treated cells lost this stiffness-dependent response, exhibiting comparable F-actin organization and FA structures across soft and rigid substrates. This suggests that oxidative stress disrupts the ability of cells to respond to substrate stiffness through coordinated cytoskeletal and adhesion remodeling (Fig. 3A, B). In the SOD1 co-treatment group, actin organization and FA formation were again more evident on rigid substrates than on soft substrates, suggesting partial recovery of stiffness responsiveness under oxidative stress (Fig. 3A, B). Quantitative analysis supported these observations. We first examined time-dependent changes from 1 to 6 h under each stiffness condition. For cell spreading area, untreated cells showed time-dependent increases on both soft and rigid substrates, consistent with progressive spreading. H2O2-treated cells showed a time-dependent increase mainly on rigid substrates, whereas SOD1 + H2O2 co-treated cells showed increased spread over time, particularly on rigid substrates (Fig. 3C). Cell aspect ratio was quantified as a complementary readout of cell-shape adaptation during spreading. Untreated and H2O2-treated cells showed limited time-dependent changes in aspect ratio, whereas SOD1 + H2O2 co-treated cells showed time-dependent changes under both stiffness conditions (Fig. 3D). For FA area, untreated cells showed time-dependent increases, consistent with adhesion maturation during spreading. In contrast, H2O2-treated cells did not show a significant time-dependent increase in FA area under either stiffness condition, suggesting impaired adhesion maturation. SOD1 + H2O2 co-treatment restored time-dependent increases in FA area on both soft and rigid substrates (Fig. 3E). We next compared soft and rigid substrates at each time point. In untreated cells, spreading area was significantly higher on rigid substrates than on soft substrates at both 1 h and 6 h. In contrast, H2O2-treated cells did not show stiffness-dependent increases in spreading area, whereas SOD1 + H2O2 co-treated cells recovered stiffness-dependent spreading at both time points (Fig. 3F, I). For cell aspect ratio, untreated cells showed minimal stiffness-dependent differences. Interestingly, H2O2-treated cells exhibited increased elongation on rigid substrates despite impaired spreading and FA remodeling, suggesting abnormal cytoskeletal adaptation under oxidative stress rather than productive mechanosensitive remodeling. This stiffness-associated elongation was not observed in the SOD1 + H2O2 group (Fig. 3G, J). Consistent with the spreading results, untreated cells showed stiffness-dependent increases in FA area at both time points, whereas H2O2 -treated cells did not. SOD1 + H2O2 co-treatment partially restored stiffness-dependent FA remodeling at both 1 h and 6 h (Fig. 3H, K). Together, these results demonstrate that oxidative stress impairs stiffness-dependent spreading and FA remodeling, leading to defective rigidity sensing. Partial restoration by SOD1 co-treatment further supports a central role of ROS in mediating mechanotransduction dysfunction. MFN deficiency recapitulates oxidative stress–induced defects in stiffness-dependent remodeling To further connect oxidative stress with mitochondrial network regulation, we first examined the expression of mitochondrial dynamics-associated genes under H2O2 treatment and SOD1 co-treatment. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analysis showed that Mfn1 expression was not significantly altered, whereas Mfn2, Opa1 and Drp1 were significantly reduced following H2O2 treatment (Fig. 4A–D). SOD1 + H2O2 co-treatment partially restored Mfn2, Opa1 and Drp1 expression compared with H2O2 treatment alone, indicating that oxidative stress perturbs mitochondrial fusion/fission regulatory programs in a ROS-dependent manner (Fig. 4A–D).Fig. 4MFN deficiency recapitulates oxidative stress–induced defects in stiffness-dependent remodeling in MEFs. A–D RT-qPCR analysis of mitochondrial dynamics-related genes showing relative expression levels of A Mfn1, B Mfn2, C Opa1, and D Drp1 in untreated, H2O2-treated and SOD1 + H2O2 co-treated MEFs. E, F Representative fluorescence images of F-actin (red) and paxillin (green) in MFN-deficient MEFs (MFN1-/-, MFN2-/-, and double-knockout (DKO)) plated on soft (5 kPa) and rigid (3 GPa) substrates and fixed at E 1 h and F 6 h after seeding (scale bar: 30 μm). Quantification of time-dependent changes in G cell spreading area, H cell aspect ratio, and I FA area per cell. J-L Quantification at 1 h of J cell spreading area, K cell aspect ratio, and L FA area per cell across substrate stiffness. M–O Quantification at 6 h of cell spreading area M, N cell aspect ratio, and O FA area. For all quantitative morphology analyses, n = 30 denotes individual cells quantified per condition from three independent biological replicates. Each bar represents mean ± SD. RT-qPCR data in panels A–D were analysed using one-way ANOVA followed by Tukey’s multiple-comparisons test. Time-dependent morphology data in panels G–I were analysed using three-way ANOVA with genotype, substrate stiffness, and time as factors, including interaction terms, followed by Tukey-adjusted multiple-comparisons tests. Data in panels J-L and M–O were analysed separately at 1 h and 6 h, respectively, using two-way ANOVA with genotype and substrate stiffness as factors, followed by Šídák-adjusted multiple-comparisons tests. Statistical significance is indicated as ns, *P < 0.05, **P < 0.01, and ***P < 0.001 MFN deficiency recapitulates oxidative stress–induced defects in stiffness-dependent remodeling in MEFs. A–D RT-qPCR analysis of mitochondrial dynamics-related genes showing relative expression levels of A Mfn1, B Mfn2, C Opa1, and D Drp1 in untreated, H2O2-treated and SOD1 + H2O2 co-treated MEFs. E, F Representative fluorescence images of F-actin (red) and paxillin (green) in MFN-deficient MEFs (MFN1-/-, MFN2-/-, and double-knockout (DKO)) plated on soft (5 kPa) and rigid (3 GPa) substrates and fixed at E 1 h and F 6 h after seeding (scale bar: 30 μm). Quantification of time-dependent changes in G cell spreading area, H cell aspect ratio, and I FA area per cell. J-L Quantification at 1 h of J cell spreading area, K cell aspect ratio, and L FA area per cell across substrate stiffness. M–O Quantification at 6 h of cell spreading area M, N cell aspect ratio, and O FA area. For all quantitative morphology analyses, n = 30 denotes individual cells quantified per condition from three independent biological replicates. Each bar represents mean ± SD. RT-qPCR data in panels A–D were analysed using one-way ANOVA followed by Tukey’s multiple-comparisons test. Time-dependent morphology data in panels G–I were analysed using three-way ANOVA with genotype, substrate stiffness, and time as factors, including interaction terms, followed by Tukey-adjusted multiple-comparisons tests. Data in panels J-L and M–O were analysed separately at 1 h and 6 h, respectively, using two-way ANOVA with genotype and substrate stiffness as factors, followed by Šídák-adjusted multiple-comparisons tests. Statistical significance is indicated as ns, *P < 0.05, **P < 0.01, and ***P < 0.001 Having established that H2O2-treated cells lose stiffness-dependent spreading and FA remodeling (Fig. 3), we next investigated whether disruption of mitochondrial fusion produces a similar mechanotransduction defect. MFN-deficient MEFs (MFN1-/-, MFN2-/-, and DKO) were seeded on soft and rigid substrates and analyzed during early (1 h) and late (6 h) spreading. Loss of Mfn1 and Mfn2 expression in the respective knockout lines was confirmed by RT-qPCR (Fig. S1), validating these cells as genetic models of impaired mitochondrial fusion. Untreated MEFs were included as a control to compare the normal stiffness-responsive pattern with MFN-deficient cells (Fig. 4E, F). Compared with untreated cells, MFN-deficient cells exhibited impaired stiffness-dependent cytoskeletal organization and FA remodeling (Fig. 4E, F). We first examined time-dependent changes from 1 to 6 h under each stiffness condition. For cell spreading area, MFN-deficient cells showed weaker and more variable time-dependent changes compared with untreated cells, indicating altered spreading progression during early adhesion (Fig. 4G). For cell aspect ratio, MFN-deficient cells showed genotype-dependent changes over time, suggesting altered cell-shape adaptation during spreading (Fig. 4H). For FA area, time-dependent adhesion maturation was also altered in MFN-deficient cells, with partial preservation in some single-knockout conditions but a less consistent response in DKO cells (Fig. 4I). We next compared soft and rigid substrates at each time point. For cell spreading area, MFN-deficient cells failed to exhibit stiffness-dependent increases in spreading area observed in untreated cells at both 1 h and 6 h, indicating impaired stiffness-responsive spreading (Fig. 4J, M). For cell aspect ratio, MFN-deficient cells showed variable genotype-dependent responses without a consistent rigidity sensitive pattern (Fig. 4K, N). For FA area, MFN-deficient cells showed partial genotype- and time-dependent stiffness responses: MFN2-/- cells retained stiffness-dependent FA remodeling at 1 h, whereas MFN1-/- cells showed a stiffness-dependent increase at 6 h; however, DKO cells did not show a consistent stiffness-dependent FA response (Fig. 4L, O). This partial preservation of FA scaling in single-knockout cells suggests that MFN1 and MFN2 loss may affect adhesion remodeling differently, whereas combined loss of both MFNs produces a more consistent impairment. Consistent with these observations, skeleton-based analysis of the actin cytoskeleton revealed that the stiffness-dependent increase in actin branch length observed in untreated cells was attenuated in both H2O2-treated and MFN-deficient cells (Fig. S2). Together, these results show that MFN deficiency phenocopies oxidative stress–induced impairment of stiffness-dependent cytoskeletal and FA remodeling, supporting a critical role for mitochondrial fusion dynamics in rigidity sensing. Oxidative stress and MFN deficiency impair nanonewton-scale traction-force generation and nanoscale force coordination To directly evaluate whether oxidative stress and MFN deficiency affect force generation and rigidity sensing, we measured traction forces using submicron micropillar arrays and quantified local force coordination using the directionality parameter (γ) (Fig. 5A, B). γ was calculated from neighboring traction-force vectors within a 34.5 μm2 peripheral region, where higher γ values indicate increased alignment and reduced opposition between neighboring vectors, reflecting impaired local force coordination associated with rigidity sensing [31].Fig. 5Oxidative stress and MFN deficiency impair nanonewton-scale traction-force generation and CU-associated force coordination. A Schematic of submicron PDMS micropillar arrays used to quantify nanonewton-scale traction forces based on pillar deflection. B Schematic of the directionality parameter (γ), where higher γ values indicate increased alignment of neighboring traction-force vectors and reduced opposing force generation at the cell periphery [26]. C Representative nanonewton-scale traction-force maps of untreated, H₂O₂-treated, SOD1 + H₂O₂ co-treated, MFN1-/-, MFN2-/-, and DKO MEFs during early spreading. Yellow outlines indicate cell boundaries, and red arrows represent traction-force vectors (scale bar: 5 μm; force scale: 2 nN). D, E Quantification of D average traction-force per pillar and E total traction-force per cell (n = 30). F Enlarged peripheral regions showing local traction-force organization associated with contraction units (CUs). Blue ellipses highlight representative opposing traction-force pairs associated with CU-mediated rigidity sensing (scale bar: 1.8 μm; force scale: 2 nN). G Quantification of γ for each condition (n = 30). Each dot represents a single cell and bars indicate mean ± SD. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Statistical significance was determined using one-way ANOVA with Tukey’s multiple-comparisons test Oxidative stress and MFN deficiency impair nanonewton-scale traction-force generation and CU-associated force coordination. A Schematic of submicron PDMS micropillar arrays used to quantify nanonewton-scale traction forces based on pillar deflection. B Schematic of the directionality parameter (γ), where higher γ values indicate increased alignment of neighboring traction-force vectors and reduced opposing force generation at the cell periphery [26]. C Representative nanonewton-scale traction-force maps of untreated, H₂O₂-treated, SOD1 + H₂O₂ co-treated, MFN1-/-, MFN2-/-, and DKO MEFs during early spreading. Yellow outlines indicate cell boundaries, and red arrows represent traction-force vectors (scale bar: 5 μm; force scale: 2 nN). D, E Quantification of D average traction-force per pillar and E total traction-force per cell (n = 30). F Enlarged peripheral regions showing local traction-force organization associated with contraction units (CUs). Blue ellipses highlight representative opposing traction-force pairs associated with CU-mediated rigidity sensing (scale bar: 1.8 μm; force scale: 2 nN). G Quantification of γ for each condition (n = 30). Each dot represents a single cell and bars indicate mean ± SD. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Statistical significance was determined using one-way ANOVA with Tukey’s multiple-comparisons test During early spreading, untreated cells exhibited strong peripheral traction vectors and frequent opposing pillar displacements at the cell edge, consistent with coordinated local contractile activity during rigidity probing (Fig. 5C, F). In contrast, H2O2-treated and MFN-deficient cells displayed weaker edge-associated traction vectors and fewer opposing pillar displacements, indicating disrupted nanoscale force coordination at the cell periphery. SOD1 co-treatment partially restored peripheral traction patterns and opposing pillar displacements associated with local contractile coordination (Fig. 5C, F). Quantitative analysis supported these observations. The average traction force per pillar was significantly reduced in H2O2-treated and MFN-deficient cells compared with untreated controls, whereas SOD1 co-treatment partially restored force output relative to H2O2 alone (Fig. 5D). Similarly, total traction force per cell was reduced under oxidative stress and MFN-deficient conditions (Fig. 5E). Consistent with disrupted local force coordination, γ values were significantly elevated in H2O2-treated and MFN-deficient cells, indicating impaired formation of opposing traction-force pairs associated with rigidity sensing at the cell periphery (Fig. 5G). In contrast, SOD1 co-treatment shifted γ values toward those of untreated cells. Together, these results demonstrate that oxidative stress reduces nanonewton-scale cellular force generation and disrupts nanoscale force coordination during rigidity sensing, while MFN deficiency phenocopies this mechanotransduction defect. Partial rescue by SOD1 further supports a ROS-dependent contribution to impaired force-based mechanosensing. Oxidative stress and MFN deficiency disrupt mitochondrial network connectivity associated with force-based mechanotransduction Given that mitochondrial function regulates cytoskeletal remodeling and force generation, we next examined whether the impaired traction-force generation observed under oxidative stress and MFN deficiency was accompanied by alterations in mitochondrial network architecture. Mitochondria were visualized using MitoTracker, and network morphology was quantified using Mitochondrial Network Analysis (MiNA) during early (1 h) and late (6 h) spreading. In untreated cells, mitochondria formed extended and interconnected networks that became more prominent over time (Fig. 6A, B). In contrast, H2O2-treated and MFN-deficient cells (MFN1-/-, MFN2-/-, and DKO) exhibited fragmented and punctate mitochondrial morphology at both time points. In the SOD1 co-treatment group, mitochondria remained largely fragmented despite partial recovery of stiffness-responsive remodeling and traction-force generation, suggesting that mitochondrial network morphology and mechanotransduction outputs did not recover in parallel. Nevertheless, mitochondrial footprint was partially preserved following SOD1 co-treatment, indicating maintenance of overall mitochondrial coverage despite limited recovery of elongated mitochondrial structures.Fig. 6Oxidative stress and MFN deficiency disrupt mitochondrial network connectivity and footprint during cell spreading. A, B Representative mitochondrial images of untreated, 200 µM H₂O₂-treated, SOD1 + H2O2 co-treated, and MFN-deficient MEFs stained with MitoTracker during A 1 h and 6 h spreading B. Top: brightfield images with cell outlines; bottom: MitoTracker fluorescence images (scale bar: 30 μm). C, D Quantification at 1 h of mitochondrial branch length C and mitochondrial footprint derived from MiNA/ImageJ analysis (n = 30) D. E, F Quantification at 6 h of mitochondrial branch length E and mitochondrial footprint (n = 30) F. Each dot represents a single cell; bars indicate mean ± SD. Statistical significance was determined using one-way ANOVA with Tukey’s multiple-comparisons test. Statistical significance is indicated as *P < 0.05, **P < 0.01, and ***P < 0.001. ns (not significant) Oxidative stress and MFN deficiency disrupt mitochondrial network connectivity and footprint during cell spreading. A, B Representative mitochondrial images of untreated, 200 µM H₂O₂-treated, SOD1 + H2O2 co-treated, and MFN-deficient MEFs stained with MitoTracker during A 1 h and 6 h spreading B. Top: brightfield images with cell outlines; bottom: MitoTracker fluorescence images (scale bar: 30 μm). C, D Quantification at 1 h of mitochondrial branch length C and mitochondrial footprint derived from MiNA/ImageJ analysis (n = 30) D. E, F Quantification at 6 h of mitochondrial branch length E and mitochondrial footprint (n = 30) F. Each dot represents a single cell; bars indicate mean ± SD. Statistical significance was determined using one-way ANOVA with Tukey’s multiple-comparisons test. Statistical significance is indicated as *P < 0.05, **P < 0.01, and ***P < 0.001. ns (not significant) Quantitative analysis confirmed these observations. At 1 h, mitochondrial branch length was significantly reduced in H2O2-treated cells and further decreased in MFN-deficient cells relative to untreated controls (Fig. 6C). Mitochondrial footprint was also significantly reduced across these conditions (Fig. 6D). While SOD1 co-treatment did not restore branch length, mitochondrial footprint remained partially preserved at levels comparable to untreated cells. Similar trends were observed at 6 h, with H2O2-treated and MFN-deficient cells displaying persistently reduced branch length and footprint compared with untreated cells, whereas SOD1 co-treatment showed only limited recovery of mitochondrial network metrics (Fig. 6E, F). Together, these results demonstrate that oxidative stress and impaired mitochondrial fusion converge on a shared mitochondrial phenotype characterized by disrupted network connectivity and reduced mitochondrial footprint. Notably, mitochondrial structural disruption persisted despite partial restoration of force-generating mechanotransduction, suggesting partial uncoupling between mitochondrial network architecture and nanoscale mechanotransduction outputs under oxidative stress conditions. Discussion Sublethal oxidative stress impaired rigidity sensing in MEFs by disrupting stiffness-dependent cytoskeletal and adhesion remodeling. H2O2-treated cells exhibited reduced filopodia formation, decreased FA area, diminished actin branch length, and impaired stiffness-responsive spreading behavior (Figs. 2 and 3). These findings indicate that oxidative stress compromises the ability of cells to reinforce cytoskeletal organization and adhesion maturation in response to substrate stiffness, processes that are essential for effective rigidity sensing and coordinated force transmission [32, 33]. Importantly, overall cell spreading was not completely abolished, suggesting that oxidative stress selectively impairs mechanosensitive remodeling rather than causing a general defect in cell adhesion. To further connect impaired rigidity sensing with mitochondrial regulation, we examined mitochondrial dynamics-associated gene expression and mitochondrial network architecture. H2O2 treatment reduced Mfn2, Opa1, and Drp1 expression, whereas Mfn1 expression remained relatively unchanged (Fig. 4A–D). Consistent with these molecular alterations, MitoTracker/MiNA analysis revealed fragmented mitochondrial networks with reduced branch length and mitochondrial footprint in H2O2-treated cells (Fig. 6). Because mitochondrial network organization influences not only ATP production but also intracellular bioenergetic distribution required for cytoskeletal remodeling and force transmission [34–37]. These findings are consistent with impaired mitochondrial network connectivity despite partial recovery of mechanotransduction outputs. Direct force measurements further demonstrated that oxidative stress impairs mechanotransduction at the functional level. Using submicron micropillar arrays, we observed significant reductions in both average and total traction-forces in H2O2-treated cells (Fig. 5D, E). In parallel, γ values were elevated, indicating disruption of opposing peripheral traction-force pairs associated with CU-dependent rigidity sensing (Fig. 5G) [7, 31, 38–40]. These findings indicate that oxidative stress disrupts not only overall force generation but also the nanoscale spatial coordination of local contractile activity required for rigidity probing. Consistent with this interpretation, ATP levels progressively decreased with increasing H2O2 exposure and were accompanied by reduced filopodia formation and diminished traction-force generation (Figs. 2 and 5). These findings establish that impaired cellular bioenergetics accompanies defective mechanotransduction under oxidative stress. The MFN-deficient models were used to determine whether disruption of mitochondrial fusion recapitulates the associated mechanical phenotype and demonstrated impaired stiffness-dependent remodeling and nanoscale force coordination. However, ATP levels and intracellular bioenergetic distribution were not assessed in the MFN-deficient MEFs. Therefore, the genetic findings support mitochondrial fusion as a regulator of rigidity sensing and force coordination, whereas the contribution of altered ATP availability to this phenotype remains to be determined. In addition, oxidative stress reduced FAK phosphorylation together with FA area (Fig. 2), suggesting impaired adhesion-associated signaling and weakened stiffness-dependent mechanical reinforcement [41–47]. Together, these observations support a model in which oxidative stress impairs rigidity sensing through combined mitochondrial bioenergetic dysfunction and defective adhesion-associated mechanotransduction. Importantly, MFN-deficient MEFs recapitulated key aspects of the oxidative stress phenotype, providing independent genetic evidence linking mitochondrial dynamics to force-based mechanotransduction. Across MFN1-/-, MFN2-/-, and DKO MEFs, stiffness-dependent spreading and traction-force generation were attenuated, while γ values were elevated, indicating impaired nanoscale coordination of rigidity-sensing contractions (Figs. 4 and 5). Mitochondrial morphology analysis further revealed reduced mitochondrial branch length and footprint in MFN-deficient cells, consistent with disrupted mitochondrial network connectivity (Fig. 6). Although FA remodeling exhibited partial genotype and time-dependent variability, the overall convergence between oxidative stress and MFN deficiency supports mitochondrial fusion dynamics as a critical regulator of cellular mechanotransduction and nanoscale force coordination [48–51]. Although the present study identifies mitochondrial fusion as a central regulator of rigidity sensing, fusion operates within an integrated mitochondrial quality-control network that also includes fission and mitophagy. Mitochondrial fission segregates dysfunctional mitochondrial segments, whereas mitophagy removes damaged mitochondria to maintain mitochondrial homeostasis [52, 53]. Although Drp1 expression was reduced following H2O2 treatment (Fig. 4D), this finding should be interpreted as a transcriptional change rather than a direct measure of DRP1-dependent fission activity. The fragmented mitochondrial phenotype remains consistent with the concurrent reduction in Mfn2 and Opa1 expression, indicating impaired fusion capacity. Moreover, DRP1 activity is regulated predominantly through post-translational mechanisms, including phosphorylation, mitochondrial recruitment, and oligomerization [54], whereas OPA1-dependent fusion is regulated through the processing of long and short OPA1 isoforms [55]. Because these protein-level regulatory events were not assessed, the present data do not define the functional balance between mitochondrial fission and fusion underlying the observed mitochondrial morphology. Accordingly, the convergence between oxidative stress and MFN deficiency should be interpreted within the broader context of coordinated mitochondrial quality-control processes that maintain mitochondrial network integrity under stress. An important finding of this study is that mitochondrial morphology and mechanotransduction outputs did not recover in parallel. Although SOD1 co-treatment partially restored stiffness-dependent spreading, FA remodeling, and traction-force generation (Figs. 3 and 5), mitochondrial branch length and network organization remained substantially impaired (Fig. 6). This dissociation indicates that partial recovery of mechanotransduction can occur without restoration of mitochondrial elongation during antioxidant co-treatment. Therefore, the present findings do not establish elongated mitochondrial network morphology as the sole upstream determinant of force generation. Rather, the genetic and antioxidant data together support a model in which mitochondrial fusion contributes to force-based rigidity sensing, whereas ROS-sensitive signaling regulates partially separable aspects of mechanotransduction. Because SOD1 was administered concurrently with H2O2 exposure and no post-stress washout and recovery experiment was performed, the present study demonstrates partial redox-responsive preservation of mechanotransduction outputs during antioxidant co-treatment but does not define the extent or kinetics of recovery after oxidative stress removal. These findings have broader implications for oxidative stress-related diseases, including fibrosis and cancer, where persistent oxidative stress and abnormal extracellular-matrix stiffening cooperatively promote pathological mechanotransduction [56, 57]. Our results suggest that oxidative stress impairs the ability of cells to appropriately sense and adapt to matrix stiffness through disruption of mitochondrial network integrity and nanoscale force coordination. Such defects may contribute to aberrant cell activation in fibrosis and altered tumor cell behavior, including invasion, progression, and therapeutic resistance, by compromising normal cell–matrix mechanical interactions. These findings also raise the possibility that preservation or restoration of mitochondrial network integrity could help normalize mechanotransduction under oxidative stress. Finally, submicron traction-force mapping provides a quantitative nanomechanical framework for investigating disease-associated alterations in rigidity sensing and evaluating therapeutic strategies targeting oxidative stress and mitochondrial dysfunction. Conclusions This study identifies mitochondrial network integrity as a critical regulator of force-based rigidity sensing under oxidative stress. By integrating submicron traction-force microscopy with mitochondrial and molecular analyses, we demonstrate that oxidative stress disrupts stiffness-responsive cytoskeletal remodeling and nanoscale contractile-force coordination through mitochondrial network dysfunction. The convergence between oxidative stress and MFN deficiency further establishes mitochondrial fusion dynamics as an essential determinant of cellular mechanotransduction during spreading. Beyond reductions in traction-force magnitude, oxidative stress disrupted nanoscale coordination of peripheral traction-force vectors associated with CU-dependent rigidity sensing. Together, these findings establish submicron traction-force mechanophenotyping as a sensitive framework for quantitative analysis of oxidative stress-associated mechanobiological dysfunction and mitochondrial mechanopathology. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1