TF Intrinsic Disorder‑Mediated Phase Separation Governs Promoter‑Proximal Pausing Stability and Explains the Accessibility‑Transcription Uncoupling in Aging

Hypothesis

Aging induces a tissue‑specific shift in the biophysical properties of master regulator transcription factors (TFs) that alters their ability to nucleate transcriptional condensates at promoters, thereby destabilizing RNA Polymerase II (Pol II) pausing without necessarily changing chromatin accessibility. This TF‑driven loss of pausing stability explains the observed uncoupling of increased promoter accessibility from transcriptional output in aged liver and predicts that restoring the native phase‑separation propensity of specific TFs will rescue pausing and gene expression independently of accessibility changes.

Mechanistic Basis

Recent work shows that "bridge" TFs such as E2F1 maintain stable expression yet exhibit altered target accessibility via 3D chromatin reorganization, linking differentially expressed genes between young and old states [2]. Concurrently, aged liver displays increased promoter accessibility at 63.4% of gained sites without corresponding transcriptional increases, mechanistically linked to reduced Pol II pausing stability [3]. We propose that the root cause is an age‑dependent modification (e.g., oxidative stress‑induced phosphorylation or acetylation) of intrinsically disordered regions (IDRs) within TFs that reduces their multivalent interaction capacity, impairing the formation of transcriptional condensates that stabilize paused Pol II. When TF IDRs lose optimal phase‑separation behavior, enhancer‑promoter looping may still occur (maintaining or increasing accessibility), but the microenvironment required for pausing fails, leading to productive transcription loss.

This mechanism also accounts for the modular decline in mitochondrial and ribosomal gene co‑expression observed with aging [5], as TFs controlling these modules (e.g., NRF1, MYC) rely heavily on IDR‑mediated condensates for high‑fidelity transcription.

Predictions

1. TFs showing unchanged expression but altered target accessibility in aged tissues will exhibit measurable changes in IDR‑driven phase separation (e.g., reduced droplet formation in vitro).
2. Artificially restoring the wild‑type IDR sequence or adding synthetic multivalent motifs to these TFs will increase Pol II pausing index at target promoters without further altering chromatin accessibility.
3. Rescue of pausing will correlate with restored transcriptional output of mitochondrial and ribosomal genes and with partial recovery of network information loss in aging muscle GRNs [1].
4. Conversely, disrupting IDR interactions in young cells will phenocopy the aging pausing defect and accessibility‑transcription uncoupling.

Experimental Approach

• In vitro: Purify candidate TFs (e.g., E2F1, EZH2) from young and old mouse liver; assess droplet formation using fluorescence microscopy and quantify partition coefficients of Pol II and Mediator.
• In vivo: Use CRISPR‑base editing to replace the endogenous TF IDR with a youthful sequence or to insert a synthetic low‑complexity domain in aged mice; measure chromatin accessibility (ATAC‑seq), Pol II pausing (PRO‑seq or NET‑seq), and nascent transcription (EU‑seq) in liver and muscle.
• Network readout: Apply single‑cell multi‑omics to compute GRN information loss before and after intervention, testing whether pausing rescue recovers up to 10% of lost information as predicted by single‑gene perturbation studies [1].
• Controls: Include TFs with stable IDRs (e.g., CTCF) to confirm specificity.

Potential Outcomes

• Support: IDR restoration normalizes pausing, increases transcription of mitochondrial/ribosomal genes, and recovers network information without global changes in accessibility, falsifying a pure stochastic damage model.
• Refutation: IDR manipulation fails to alter pausing or transcription despite confirmed changes in TF phase‑separation properties, indicating that accessibility‑transcription uncoupling arises from alternative mechanisms (e.g., promoter‑proximal nucleosome remodeling or TF‑cofactor stoichiometry shifts unrelated to condensates).

This hypothesis bridges TF network topology, chromatin dynamics, and phase‑separation biology, offering a concrete, falsifiable route to test whether rescuing TF material properties can reestablish transcriptional fidelity in aging.