Mechanotransduction-Induced Gene Expression Reveals Activation of TGFβ/SKIL/TAZ Axis and Supports Invasive Phenotype in Triple-Negative Breast Cancer

Abstract

Triple-negative breast cancer (TNBC) is an aggressive subtype with limited treatment options. Emerging evidence shows that mechanotransduction, driven by matrix stiffness and mechanical signaling, promotes TNBC invasion and metastasis. As breast cancer progresses, expansion of fibroblasts and tumor-reactive stroma increases extracellular matrix deposition, generating matrix tension and enhancing mechanotransduction, which promotes cell proliferation, invasion, and metastatic potential through altered gene expression patterns. To investigate the molecular mechanisms underlying these changes, human TNBC cells were subjected to constant or oscillatory strain, followed by comprehensive transcriptomic analysis. Results revealed pronounced differential expression of genes involved in cell migration, adhesion, and transforming growth factor-β (TGFβ) signaling, with RT-PCR validation confirming SKI Like Proto Oncogene (SKIL) as the most strongly upregulated gene. Analysis of The Cancer Genome Atlas (TCGA) datasets indicated that SKIL is highly expressed in multiple breast cancer subtypes. Cross-sectional comparison of oscillatory strain-induced genes with TCGA data revealed coordinated upregulation of TGFβ, SKIL, and other genes associated with invasive phenotypes, immune suppression, and drug resistance, highlighting the vital role of TGFβ signaling. Transcription factor enrichment analysis further identified regulators linked to oncogenic pathways, including TGFβ effectors and Hippo signaling, supporting a mechanotransduction-driven transcriptional program in breast cancer.

1. Introduction

Triple-negative breast cancer (TNBC) is among the most aggressive breast cancer subtypes and is associated with poor prognosis and reduced overall survival [1,2,3]. The absence of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) makes TNBC unresponsive to hormone-based therapies, limiting treatment options primarily to chemotherapy and immunotherapy [4]. However, these strategies often show limited long-term efficacy and may contribute to the emergence of more aggressive, treatment-resistant disease. Therefore, identifying key molecular pathways and therapeutic targets within tumor cells and the tumor microenvironment (TME) is critical for improving treatment outcomes.

TNBC progression is driven by complex molecular, biochemical, and biophysical mechanisms. As tumors grow, they alter the mechanical properties of surrounding tissue and generate a stressed microenvironment that supports disease progression [5]. Mechanical forces interact with biochemical signaling to regulate proliferation, dormancy, immune suppression, and metastasis, although these mechanisms remain incompletely understood [6,7]. While matrix stiffness has been well documented as a driver of tumor aggressiveness, the effects of dynamic mechanical forces—particularly tensile and compressive stresses during tumor development—are less clearly defined [8,9]. Breast tumors develop within a heterogeneous and evolving mechanical landscape. Expanding tumor masses generate compressive stress, whereas tensile forces accumulate at the tumor periphery [10]. Cyclic strains (1–25%), together with the elastic modulus of glandular breast tissue (~6.6–7.5 kPa), contribute to this dynamic environment [11]. Compressive forces enhance invasive behavior and proliferation in breast cancer cells [12,13], and tensile strain is elevated in cancer-associated fibroblasts (CAFs) compared with normal fibroblasts, suggesting a role in tumor progression [14,15]. However, the broader impact of mechanical loading on metastasis and therapeutic resistance remains incompletely understood [16]. Mechanical forces within the TME can also impair drug delivery and limit immune cell infiltration [12,17], promote macrophage polarization toward a pro-tumorigenic M2 phenotype, and restrict cytotoxic T-cell infiltration [18,19,20]. We have previously shown that mechanical strain enhances proliferation and migration of TNBC cells and promotes immunosuppression through the release of PD-L1-enriched exosomes [12,21].

To investigate how mechanical strain influences gene expression, we performed transcriptomic profiling of MDA-MB-231 cells exposed to constant or oscillatory strain followed by bulk RNA sequencing. Data were analyzed using the nf-core RNA-seq pipeline, and differentially expressed genes were assessed for enrichment using the STRING database. Oscillatory strain significantly altered genes involved in proliferation, epithelial-to-mesenchymal transition (EMT), and oncogenic signaling. Genes such as SKIL and TGFβ were strongly upregulated, with SKIL showing the highest induction. Analysis of The Cancer Genome Atlas (TCGA) dataset confirmed elevated SKIL expression in primary breast tumors, including TNBC. SKIL is a proto-oncogene that stabilizes TAZ within the Hippo pathway and functions as a transcriptional repressor of the TGFβ/Smad signaling axis [22]. Collectively, these findings suggest that mechanotransduction-induced upregulation of SKIL contributes to TNBC aggressiveness through modulation of TGFβ-associated oncogenic signaling.

2. Results

2.1. Oscillatory Strain Significantly Alters the Gene Expression Landscape of Human Breast Cancer Cells

Mechanical force emerges as a key mediator in breast cancer progression, metastasis, and immunosuppression [9,12,16]. To understand how mechanical force causes gene expression changes, human TNBC cells, MDA-MB-231, were subjected to constant strain and oscillatory strain for 48 h, after which total RNA was isolated, and RNA-sequencing was performed. At the outset, differential gene expression anagroups indicated a total of 112 genes between control group and constant strain-applied group, and a total of 1423 genes between control and oscillatory strain-applied group. There were 1378 genes that were differentially expressed between constant strain and oscillatory strain groups. After correcting for multiple testing, 101 genes remained significantly differentially expressed between oscillatory strain and control conditions (adjusted p < 0.05). However, there were no statistically significant differences in gene expression between either control and constant strain-applied groups or between constant strain and oscillatory strain-applied groups (Figure 1a). A heat map along with hierarchical clustering and grouping of genes based on expression similarity was generated to visualize changes in the top 100 differentially expressed genes, including both upregulated and downregulated genes (Figure 1b). As shown in the heat map, oscillatory strain induced pronounced changes in gene expression, with many genes exhibiting either strong upregulation or downregulation. Since no statistically significant changes in gene expression were seen in constant strain vs. control or oscillatory strain vs. constant strain, downstream functional analysis was focused on oscillatory strain induced gene expression changes. Prior to analyzing gene expression changes by functional ontology and pathway enrichment, the transcriptome data were verified by real-time PCR for the following transcripts: SKIL, TNFSF15, TGFB1, TCN1, ATOH8, and LGALS9B. Results confirmed the validity of RNA-Seq data, showing similar expression patterns to RNA-Seq analysis (Figure 1c).

2.2. Oscillatory Strain Induces Upregulation of Genes Involved in Cell Migration

Upon confirming transcriptomic changes from RNA-Seq by RT-PCR analysis, gene ontology for biological processes was performed using the top 100 differentially expressed genes with a minimal ±1-fold change in expression. A bubble plot was generated to visualize the enrichment of biological processes based on Gene Ontology (GO) terms (Figure 2a), which revealed several significantly enriched processes, with signal strength values ranging from 0.54 to 0.64, where each bubble represented a biological process, bubble size indicated gene count, and color intensity reflected the False Discovery Rate (FDR). The top 10 enriched processes were shown in the bubble graph (Figure 2a). The FDR values for these enriched biological processes ranged from 5.3 × 10⁻⁶ to 3.3 × 10⁻⁴, indicating high statistical significance. To identify key modulators involved in the highly enriched biological processes following exposure to oscillatory strain, a protein–protein interaction (PPI) network analysis was performed for the top three enriched biological processes. As shown in Figure 2b, the PPI network for negative regulation of multicellular organismal processes is among the highly enriched biological processes, with the maximum number of nodes and edges. Each of the 21 nodes represents a protein, with edges denoting known or predicted interactions. This enrichment had a p-value of 2.39 × 10⁻⁶, and within this network, subgroups of proteins were associated with other biological processes, such as negative regulation of cell adhesion, regulation of multicellular organismal processes, regulation of neurogenesis, and regulation of nervous system development. The top enriched processes within this network were highlighted by unique colors, with respective color codes and associated processes shown in adjacent table. Further, PPI network analysis for the topmost enriched biological process, positive regulation of cell migration, included 14 nodes with a p-value of 4.87 × 10⁻⁶ (Figure 2c). Proteins in this PPI network were also part of other enriched biological processes, such as blood vessel development, regulation of chemotaxis, sympathetic neuron projection guidance, and regulation of lymphocyte proliferation, shown in inset, with unique colors and associated processes. To delineate molecular mediators of negative regulation of cell adhesion, a PPI network analysis was performed, revealing 11 nodes and 10 edges with a p-value of 4.87 × 10⁻⁶. Members of this network were also enriched for additional biological processes such as negative regulation of T-cell proliferation, negative regulation of cell activation, and regulation of lymphocyte proliferation. Upregulated genes were collectively associated with coordinated biological processes such as tumor progression, vascular development, and TGFβ signaling, highlighting the functional relevance of this gene set in regulating cell behavior, particularly in development, immune response, and cancer progression.

2.3. Oscillatory Strain Induces Expression of Genes Associated with TGFβ Signaling

Following gene enrichment analysis, reactome pathway analysis was performed for higher-order molecular network interpretation. This revealed enrichment of signaling pathways centered around TGFβ family members (Figure 3a). To include the maximum number of input genes, a PPI network was curated for signal transduction. As shown in Figure 3b, this network included 34 nodes and 28 edges, including members of the TGFβ family and providing comprehensive insight into signaling pathways involving members of the SMAD family and signaling by receptor tyrosine kinases (RTK). Most members in the signal transduction network (NRP2, FURIN, FSTL3, SERPINE1, INHBB, CGN, NOG, SKIL, NCOR2, and PMEPA1) were interconnected as part of the broad TGFβ signaling network.

Interestingly, four of the members, FURIN, SERPINE, SKIL, and NCOR2, were part of transcriptional activity involving the SMAD pathway whereas LAMB3, IRS1, COL1A1, VEGFC, NRP2, SPRY1, FURIN and PCSK6 were associated with RTK signaling (shown in the table at the bottom). In addition, three independent clusters were also observed: SPTB and SPTBN5; PLEKHG4 and AMIGO2; and PDE4B, PDE4D, and PDE7B. Functional aspects of these clusters revealed associations with platinum drug resistance and COPI-mediated anterograde transport (SPTB/SPTBN5) [23,24], cyclic AMP-mediated signaling pathways (PDE family clusters) [24,25], cytoskeleton dynamics at Golgi (PLEKHG4) [26], and tumorigenesis (AMIGO2) [27]. Collectively, this PPI network highlights a coordinated regulatory module centered on TGFβ signaling, with implications in cellular communication, EMT, immune modulation, and drug resistance in oncogenic signaling.

2.4. Oscillatory Strain Promotes Upregulation of Genes Associated with Invasive Breast Carcinoma

Gene ontology and pathway enrichment revealed that oscillatory strain-induced genes were associated with oncogenic signaling pathways centered on TGFβ signaling. To investigate clinical relevance, the TCGA dataset in UALCAN was used to compare expression of oscillatory strain-mediated genes between normal breast tissue (n = 114) and invasive breast carcinoma (n = 1097). The analysis revealed significant upregulation of many oscillatory strain-induced genes, including SKIL, SERPINE2, EPHB2, IRGQ, CGN, FURIN, LUPEXIN, COL1A1, ZNF385A, PGM2L1, SERPINE1, and TGFB1 in breast tumors with robust statistical significance (p ≤ 0.0001) (Figure 4a). This indicates a potential link between oscillatory strain-responsive genes and invasive breast carcinoma progression. Notably, COL1A1, TGFβ, and SERPINE1 were linked to YAP activation, TGFβ signaling, and EMT drivers [28]. COL1A1 plays key roles in matrix remodeling, exosome biogenesis, metastasis, and is recognized as an oncogenic driver in multiple malignancies, including TNBC [29].

Protein–protein interaction and pathway enrichment analysis of these genes showed strong similarities to oscillatory strain-induced pathways, including TGFβ signaling, SMAD2/SMAD3:SMAD4 transcriptional activity, ECM proteoglycans, EMT, and extracellular matrix organization (Figure 4b and inset). Strong PPI networking was observed among SERPINE1, COL1A1, FURIN, TGFβ, CGN, and SKIL, highlighting their central role in oscillatory strain-mediated oncogenic signaling. Consistent with these findings, previous work in ovarian cancer demonstrated that oscillatory strain activates EMT-associated genes, enhances invasive phenotypes, and increases EV production, suggesting a conserved mechanotransduction-driven signaling across tumor types [12,21,30].

2.5. Oscillatory Strain Mediates Upregulation of Proto-Oncogene SKIL

To identify molecular pathways specific to oscillatory strain, differential gene expression analysis was performed by comparing gene expression changes under oscillatory strain versus constant strain or control, as there was no difference between these two groups in adjusted p value. As shown in the volcano plot in Figure 5a, oscillatory strain caused upregulation of SKIL, NRP2, PMEPA1, and SPTB, and downregulation of MUC5AC, SYT9, ASB2, and RASL11A. Among these, SKIL was the most significantly upregulated gene under oscillatory strain. It is frequently amplified in multiple cancers and has been implicated in tumor growth and invasion by modulating transcriptional regulation of EMT-related transcription factors [31].

Hence, we focused further on investigating SKIL’s role in breast cancer through comparative gene expression analysis across various breast cancer types using the TCGA dataset in UALCAN. This analysis revealed that compared to normal breast epithelium, SKIL transcript levels were significantly elevated in luminal tumors and TNBC, whereas HER2+ tumors did not have significantly elevated expression (Figure 5b). To elucidate the SKIL-mediated signaling cascade, a PPI network of SKIL-interacting proteins was curated from the STRING database. As shown in Figure 5c, SKIL interacts with most SMAD family proteins, including SMAD2, SMAD3, SMAD4, SMAD5, SMAD6, and SMAD7, and other interactors such as NCOR1, SMURF2, SKI, and RNF11. Next, we subjected SKIL and its interacting partner proteins for biological enrichment. As shown in Figure 5d, this analysis revealed that SKIL and its interacting proteins were associated with multiple signaling pathways, including TGFβ signaling, SMAD protein signal transduction, ureteric bud development, BMP signaling. Among them, TGFβ/SMAD signaling was the highest enriched pathway regulated by SKIL and its interacting proteins.

2.6. SKIL Integrates Mechanical Cues to Activate YAP/TAZ Signaling and Cell Migration in TNBC

Our RNA-sequencing analysis identified SKIL as a key mediator of oscillatory strain-induced oncogenic signaling in the human TNBC cell line MDA-MB-231 (Figure 1c). To determine whether this response is conserved across additional TNBC cells, we examined SKIL transcript levels in Hs578T cells upon subjecting them to oscillatory strain. As shown in Figure 6a, SKIL expression was similarly upregulated in Hs578T cells, indicating that strain-responsive induction of SKIL is a shared feature of these TNBC lines.

Since YAP/TAZ are central regulators of mechanotransduction-driven pro-tumorigenic signaling, we next assessed whether oscillatory strain modulates their expression. We observed a significant increase in both YAP and TAZ transcript levels following application of oscillatory strain. SKIL has been reported to promote tumorigenesis in non-small cell lung cancer by upregulating TAZ [32], suggesting that strain-induced SKIL might function upstream of the YAP/TAZ axis in TNBC as well. To test this possibility, we silenced SKIL expression using siRNA and measured transcript levels of YAP and TAZ. As shown in Figure 6b, SKIL knockdown led to a significant reduction in both YAP and TAZ expression, in both MDA-MB-231 and Hs578T cells, supporting a regulatory role for SKIL in promoting YAP/TAZ activation under mechanical strain. We previously demonstrated that oscillatory strain enhances TNBC cell proliferation and migration [12]. Therefore, we further evaluated whether SKIL contributes to these phenotypes. Interestingly, SKIL silencing significantly decreased the migratory capacity of both MDA-MB-231 and Hs578T cells (Figure 6c), indicating that SKIL is required for strain-enhanced cell migration in TNBC.

2.7. Transcription Factor Enrichment and Interaction Network for Oscillatory Strain-Responsive Genes

To identify upstream regulators of oscillatory strain-responsive genes, transcription factor (TF) enrichment analysis was performed using the ChEA3 portal (ChIP-X Enrichment Analysis, version 3). The top 100 differentially expressed genes with a minimal ±1-fold change in expression were used for enrichment analysis. The top 10 transcription factors, based on MeanRank scores derived from multiple libraries, were identified. As shown in Figure 7a, these included ZNF469, SMAD3, TWIST1, ELK1, SNAI2, FOXD1, TWIST2, ATF3, GLI3, and PRRX1, all of which have established roles in EMT, extracellular matrix remodeling, and tumor progression, further validating our functional analysis.

The integrated ranking approach combines evidence from diverse sources, including ENCODE ChIP-Seq, GTEx co-expression, and TF perturbation signatures, improving prediction accuracy compared to single-library analyses [33]. The inset shows transcription factors and their associated genes, revealing that SMAD3 is one of the TFs with the maximal number of associated genes, including TGFB and SKIL. As shown in Figure 7b, TF enrichment analysis identified top-ranked nodes among TWIST1, SNAI2, and PRRX1, forming a regulatory network dominated by EMT-associated TFs, which are well-established drivers of EMT and metastatic progression. From this analysis, TWIST1 emerged as a central hub with connectivity to other EMT regulators such as TWIST2, ZNF469, and developmental factors FOXD1 and GLI3, suggesting crosstalk between EMT and stemness pathways. SMAD3, a key effector of TGFβ, reinforces TGFβ-mediated transcriptional programs. Additionally, stress-response TFs such as ATF3 and ELK3 reveal integration of stress-response co-regulatory mechanisms within this TF network. Taken together, TF enrichment and regulatory networks highlight oscillatory strain-mediated transcriptional architecture that promotes cellular plasticity, invasion, and therapy resistance, aligning with EMT-driven phenotypic transitions and aggressive behavior of breast cancer.

3. Discussion

Mechanical forces in the TME, including constant and oscillatory strain, significantly influence cancer cell behavior [9,13,16,30]. In breast cancer and TNBC models, oscillatory strain promotes proliferation, migration, invasion, and drug resistance [9,13,16,34,35]. Mechanical stimulation also enhances exosome production and activates pro-tumorigenic pathways such as YAP/TAZ signaling [12,31,36]. Although these effects are well documented, the underlying molecular mechanisms remain incompletely defined. We integrated transcriptomic profiling, bioinformatic analysis, and functional assays to characterize oncogenic programs induced by mechanical strain in TNBC cells. RNA sequencing revealed widespread gene expression changes under constant and oscillatory strain. Focusing on the top 100 genes ranked by fold change and adjusted p-value, oscillatory strain induced a strong transcriptomic shift. Enrichment analysis showed overrepresentation of pathways related to migration, invasion, loss of adhesion, suppression of T-cell proliferation, and drug resistance—processes previously linked to mechanical stimulation [13,21,34]. Pathway analysis highlighted activation of TGFβ/SMAD signaling and receptor tyrosine kinase networks. TGFβ and SMAD pathways are established regulators of EMT, immune suppression, and therapy resistance [12,21,35].

TCGA analysis demonstrated that SKIL, SERPINE1, SERPINE2, EPHB2, IRGQ, CGN, FURIN, LOX, COL1A1, ZNF385A, PGM2L1, and TGFB are associated with invasive breast carcinoma, supporting prior links between mechanical stress and invasiveness [13]. Among genes upregulated by oscillatory strain, SKIL, NRP2, PMEPA1, and SPTB clustered prominently, with SKIL showing the strongest induction. SKIL is upregulated across multiple breast cancer subtypes, including TNBC, and in other solid tumors [32,37,38]. Expanding tumors generate compressive stress and remodel the extracellular matrix (ECM), leading to collagen deposition, matrix stiffening, and vascular compression [30,39,40]. Impaired drainage increases interstitial pressure and produces both static compression and dynamic mechanical cues such as oscillatory strain near vessels [6,40]. Mechanistically, SKIL regulates contact inhibition and Hippo–YAP signaling. At low cell density, SKIL stabilizes TAZ by preventing its phosphorylation by LATS2, activating YAP/TAZ signaling; at high density, SKIL decreases and TAZ is degraded [38,41]. This balance is frequently disrupted in cancer [38]. TCGA data show increased SKIL expression in luminal tumors and TNBC but not in HER2-positive disease. Mechanical activation of TGFβ and Hippo pathways may further modulate SKIL, as TGFβ can regulate its stability and sustained mechanical strain promotes TGFβ activation [22].

Transcription factor enrichment identified ZNF469, SMAD3, TWIST1, ELK3, SNAI2, FOXD1, TWIST2, ATF3, GLI3, and PRRX1 as key regulators. These factors form a network centered on TGFβ and SKIL, with SMAD3 linking both pathways. Several promote EMT (TWIST1, SNAI2, and PRRX1), stemness (TWIST2, ZNF469, and FOXD1), or stress responses (ATF3, ELK3). Overall, oscillatory strain drives transcriptomic reprogramming through the TGFβ/SKIL/TAZ axis, promoting EMT, drug resistance, stemness, and immune suppression. SKIL induction was consistent in both MDA-MB-231 and Hs578T cells, suggesting a conserved mechanosensitive response. Increased YAP/TAZ activity and reduced signaling after SKIL knockdown support a model in which SKIL amplifies mechanotransduction upstream of Hippo signaling. Functionally, SKIL silencing reduced migration, confirming its role in strain-enhanced motility. These findings position SKIL as a central mediator linking mechanical strain to pro-tumorigenic signaling and suggest that targeting this axis may be therapeutically relevant.

Oscillatory strain also downregulated MUC5AC, SYT9, ASB2, RASL11A, and PDE7B. Reduced MUC5AC reflects loss of epithelial differentiation consistent with EMT and invasion [42]. Decreased SYT9 may alter Ca²⁺-dependent vesicle secretion under cytoskeletal remodeling [43]. ASB2 regulates degradation of actin-binding proteins and may affect motility when suppressed [44]. The role of RASL11A in TNBC remains unclear but its downregulation in malignancy suggests context-dependent tumor-suppressive functions [45]. PDE7B hydrolyzes cAMP; its suppression may increase cAMP/PKA signaling and has been linked to reduced proliferation in some cancer models [46,47]. Together, gene suppression under mechanical stress supports reprogramming toward increased plasticity and aggressiveness. The applied oscillatory strain parameters are physiologically relevant because they approximate the dynamic mechanical forces present within the breast TME, where tumor expansion, vascular pulsatility, extracellular matrix stiffening, and impaired lymphatic drainage generate fluctuating tensile and compressive stresses. These forces create both sustained and cyclic mechanical cues that activate key mechanotransduction pathways, including TGFβ and Hippo/YAP–TAZ signaling, which are known to regulate EMT, invasion, immune modulation, and therapy resistance. By modeling strain magnitudes and frequencies consistent with those reported in solid tumors, our system captures biomechanical conditions that are likely encountered by TNBC cells in vivo, thereby strengthening the translational relevance of the observed SKIL-dependent transcriptional reprogramming. Overall, the data support a model in which mechanical strain activates TGFβ signaling and induces SKIL expression, with SKIL enhancing YAP/TAZ activity by stabilizing TAZ upstream of Hippo signaling. However, because the current conclusions are largely based on transcriptomic analyses and SKIL knockdown studies, additional functional experiments such as proteomic, phosphoproteomic, and functional signaling analyses are required to determine whether TGFβ regulates YAP/TAZ directly, indirectly through SKIL, or via parallel mechanosensitive pathways.

4. Materials and Methods

4.1. Cell Culture

Human TNBC cell lines, MDA-MB-231, were obtained from Dr. Danny Welch (University of Kansas), and the Hs578T cell line was a kind gift from Dr. Lalita Shevde-Samant (University of Alabama at Birmingham). Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin.

4.2. Exposure of MDA-MB-231 and Hs578T Cells to Strain

Then, 2.5 × 10⁶ cells were plated on collagen-coated 6-well UniFlex Culture plates (Flexcell International Corporation, Burlington, NC, USA) and allowed to adhere for 24 h. Fresh medium was then added, and plates were subjected to either 10% uniaxial oscillatory strain (OS) at 0.3 Hz for 48 h or 10% constant strain for 48 h using a FlexCell FX-5000 system (Flexcell International Corporation, Burlington). Strain parameters were selected to mimic deformations generated by breast cancer-derived CAFs and normal respiratory rates [14,15].

4.3. RNA-Sequencing

Total RNA was extracted using the PureLink™ RNA Mini Kit (Thermo Fisher, Waltham, MA, USA) according to the manufacturer’s protocol. RNA integrity was assessed with an Agilent Bioanalyzer 2100; samples with RNA integrity number (RIN) ≥ 8.0 were used for sequencing. RNA libraries were prepared using the Illumina TruSeq Stranded mRNA Library Prep Kit (Illumina Inc., San Diego, CA, USA) and sequenced on an Illumina NovaSeq 6000 platform to generate 150 bp paired-end reads (Illumina Inc., Novaseq600 sequencing system, San Diego, CA, USA). Raw reads were quality-checked using FastQC (v0.11.9), trimmed with Trimmomatic (v0.39), aligned to the human reference genome (GRCh38) using STAR aligner (v2.7.10a), and gene expression quantified with featureCounts (v2.0.3).

4.4. Transwell Migration Assay

Transwell cell migration assay was performed using Transwell chambers suitable for 24-well plates (8 µm, Corning, Glendale, AZ, USA). For assessment of migration capacity following SKIL silencing, cells were transfected with SKIL siRNA and control siRNA at 60–70% confluency, after 48 h of transfection following which the cells were trypsinized and resuspended in serum-free medium and seeded on the upper compartment of the Transwell insert and transferred in to a 24 well plate containing 500 µL DMEM medium in the lower chamber to act as a chemoattractant. After 24 h non-migrated cells remaining in the upper member surface were gently removed using moist cotton swab. Then, the membranes were fixed with 4% paraformaldehyde and stained with 1% crystal violet. The numbers of cells that had traversed to the underside of each membrane were imaged at 5× magnification using a Leica DMI4000 B inverted microscope (Leica Microsystem, Wetzlar, Germany) and number of migrated cells were quantified using captured images with Image J bundled with 64-bit Java 8 software.

4.5. Transfection

The siRNA against SKIL (AM16708, ID 107695) and negative control siRNA (AM4611) was purchased from Thermo Fischer (Thermo Fischer, USA). The siRNA transfection was done using Invitrogen lipofectamine RNAi Max reagent (Thermo Fischer, USA) as per the manufacturer’s protocol.

4.6. Real-Time PCR

Real-time PCR (RT-PCR) was performed using SYBR™ Green on the QuantStudio 3 Real-Time PCR system (Thermo Fisher, USA), and data were analyzed using Design and Analysis software (DA2, Thermo Fisher, USA). The primers used were as follows:

• SKIL forward primer: 5′-ATGCAGGACAGTTGGCAGAA-3′;
• SKIL reverse primer: 5′-TCTGTCTTGCTTCCCGTTCC-3′;
• TNFSF15 forward primer: 5′-AAATCAGACAAGCAGGCCGA-3′;
• TNFSF15 reverse primer: 5′-AGACTTGGTCCCCATGAGGA-3′;
• TGFβ1 forward primer: 5′-TACCTGAACCCGTGTTGCTCTC-3′;
• TGFβ1 reverse primer: 5′-GTTGCTGAGGTATCGCCAGGAA-3′;
• TCN1 forward primer: 5′-TTTCACAATGGAGGAGCGCT-3′;
• TCN1 reverse primer: 5′-GCCTCCACTCAGAAGTTCCC-3′;
• ATOH8 forward primer: 5′-GTCCAAACTGGCCATCCTGA-3′;
• ATOH8 reverse primer: 5′-GGAGAAGCTGAGGTTGCTGT-3′;
• LGALS9B forward primer: 5′-GCAGGGTATGTGGTGTGCAA-3′;
• LGALS9B reverse primer: 5′-AGCAGAGGTCAAAGGGCATC-3′;
• TAZ forward primer: 5′-CACTGTGCTGATCGGGAAG-3′;
• TAZ reverse primer: 5′-TCCACAGCCGACTTGTTCTC-3′;
• YAP forward primer: 5′-TCCCAGATGAACGTCACAGC-3′;
• YAP reverse primer: 5′-GAGGACCTGAAGCCGAGTTC-3′.

4.7. Data Processing and Analysis

Differential gene expression analysis was performed using count-based statistical modeling implemented within the nf-core RNA-seq pipeline. Analysis of differentially expressed genes, heat maps, and volcano plots were developed using R (https://cran.r-project.org/, R version 4.2.2; URL accessed in January 2026). Differential gene expression was assessed using normalized RNA-seq count data, and nominal p-values were adjusted for multiple hypothesis testing using the Benjamini–Hochberg false discovery rate (FDR) method. Heat maps for the top 100 differentially expressed genes were generated using adjusted p-value < 0.05 and log₂ fold change ≥ ±3. Gene enrichment for biological processes and reactome pathway analysis were conducted using the STRING database (https://string-db.org), with significance determined by Benjamini–Hochberg adjusted p-values. SKIL expression in TNBC and invasive breast carcinoma was analyzed using UALCAN (https://ualcan.path.uab.edu) and MammoONC-DB (https://resource.path.uab.edu/MammOnc-Home.html), and transcription factor enrichment was performed using ChEA3 (https://maayanlab.cloud/chea3).

4.8. Statistical Analysis

Data were analyzed using GraphPad Prism v10. Student’s t-test was used to compare variables between two conditions. p values < 0.05 was considered significant.

5. Conclusions

Oscillatory strain in the TNBC drives transcriptional reprogramming that enhances expression of genes involved in tumor proliferation, invasion, EMT, and stemness via upregulation of key mediators, including SKIL, TGFβ, and YAP/TAZ, which are known to coordinate ECM remodeling and Hippo pathway activation. Transcription factor networks involving SMAD3, TWIST1/2, and SNAI2 further reinforce EMT and immunosuppressive programs, while downregulation of MUC5AC, SYT9, ASB2, RASL11A, and PDE7B indicate effects on cytoskeletal plasticity and survival signaling. Comparative analysis with TCGA datasets confirmed the clinical relevance of oscillatory strain-responsive genes in invasive TNBC. Together, these findings identify the TGFβ/SKIL/TAZ axis as a central driver of mechanically induced oncogenic signaling and highlight SKIL and its downstream effectors as potential therapeutic targets.