Spatial ECM Stiffness Gradients Drive Tissue‑Specific Senescence Convergence via NELF/SPT6‑Mediated Transcriptional Noise

We hypothesize that progressive increases in extracellular matrix (ECM) stiffness create biomechanical gradients that amplify stochastic transcription elongation defects, leading to convergent downregulation of tissue‑specific genes across aged organs. This process is mediated by altered activity of the transcription elongation factors NELF and SPT6, which become dysregulated in stiff microenvironments and promote aberrant isoform production that erodes transcriptional identity. Consequently, tissues with higher baseline stiffness (e.g., lung, liver) show earlier and stronger convergence, while compliant tissues (brain, testis) resist the shift.

Predictions

• Spatial correlation: Regions of elevated ECM stiffness within a tissue will exhibit higher senescence marker expression and greater loss of tissue‑specific transcripts compared with softer niches.
• Causal link: Pharmacological softening of the matrix or genetic reduction of NELF/SPT6 activity in aged mice will attenuate convergence and preserve tissue‑specific gene profiles without globally suppressing senescence.
• Isoform shift: Stiffness‑induced NELF/SPT6 dysfunction will produce a detectable increase in intron‑retention or alternative‑splicing events in tissue‑specific genes, measurable by long‑read RNA‑seq.

Experimental design

1. Map stiffness and transcription: Perform AFM‑based elasticity mapping on cryosections of young (3 mo) and aged (24 mo) mouse lung, liver, spleen, and brain, followed by spatially resolved transcriptomics (Slide‑seqV2 or MERFISH). Overlay stiffness maps with expression of DiCo genes [1] and senescence markers (Ccl2, Il6, Cdkn2a) to quantify spatial correlations.
2. Mechanical intervention: Treat aged mice with LOX inhibitor (β‑aminopropionitrile) to reduce cross‑linking, or implant hydrogel scaffolds of defined modulus into liver parenchyma. Assess whether softened niches rescue tissue‑specific gene expression and reduce senescence burden relative to controls.
3. NELF/SPT6 perturbation: Use CRISPRi to knock down NELF or SPT6 specifically in stromal cells of aged lung and liver via AAV‑delivery. Measure changes in isoform ratios (PacBio) and DiCo gene convergence.
4. Readout: Compute convergence index (variance of tissue‑specific gene expression across organs) before and after interventions. A significant reduction in convergence index after stiffness reduction or NELF/SPT6 knockdown supports the hypothesis; no change falsifies it.

Mechanistic rationale ECM stiffness activates integrin‑FAK‑YAP signaling, which can phosphorylate NELF and alter its affinity for pausing complexes [3]. Simultaneously, mechanotransduction modulates chromatin accessibility, making SPT6‑dependent nucleosome remodeling less efficient. The combined effect increases transcriptional noise, disproportionately affecting low‑abundance, tissue‑specific transcripts that lack robust buffering mechanisms. As noise accumulates, expression profiles drift toward a common low‑variance state, producing the observed inter‑tissue convergence while preserving core senescence programs.

This hypothesis integrates the observed transcriptional convergence [1], spatially heterogeneous senescence niches [2], and the role of NELF/SPT6 in senescence‑associated RNA processing [3], directly addressing the lack of spatial senescence gradients and mechanistic drivers highlighted in the literature [4].