Hypothesis

In SF3B1, U2AF1, SRSF2, or ZRSR2‑mutant tumors, splicing dysregulation remains adaptive as long as intracellular ATP keeps spliceosome condensates in a liquid‑like state. When ATP falls below a critical concentration (~1 mM in vivo), the same mutant spliceosomes undergo an ATP‑driven phase transition to a solid‑like gel, abolishing residual splicing fidelity and pushing transcriptomic noise past an entropy threshold that becomes deleterious to cell fitness. This transition converts splicing‑driven intratumor heterogeneity into a lethal splicing catastrophe, creating a synthetic‑lethal vulnerability that can be exploited by metabolic stressors or ATP‑lowering agents.

Mechanistic Rationale

1. Mutant spliceosomes alter condensate material properties – SF3B1‑K700E and similar changes increase multivalent RNA‑protein interactions, lowering the energy barrier for liquid‑to‑gel transition (see structural dynamics in 1).
2. ATP is a hydrotrope that prevents aberrant protein aggregation – ATP binds weakly to low‑complexity domains of splicing factors, maintaining solubility; depletion reduces this shielding effect (principle established for RBPs, applicable here by analogy).
3. Entropy threshold emerges from loss of splicing fidelity – Above the ATP threshold, mutant spliceosomes still produce a spectrum of isoforms that fuel plasticity and neoantigen load (2,3). Below it, gelation stalls the catalytic core, causing widespread intron retention and exon skipping that corrupt essential genes, raising isoform noise beyond the tolerated 15 000 cancer‑specific variants and triggering mitotic catastrophe or apoptosis.
4. Spatial prediction – Gelated spliceosomes should appear as immobile, hyper‑intense foci in ATP‑depleted tumor regions, detectable by ATP‑sensitive biosensors combined with long‑range RNA MAS‑seq.

Testable Predictions

• Prediction 1: In isogenic cell lines expressing SF3B1‑K700E, pharmacological ATP depletion (e.g., 2‑deoxy‑glucose + oligomycin) will increase the fraction of spliceosome condensates exhibiting FRAP recovery < 5 % (gel state) correlating with a > 2‑fold rise in intron retention measured by single‑cell long‑read sequencing.
• Prediction 2: Tumor xenografts treated with a glycolysis inhibitor will show spatially restricted spliceosome gel foci (visualized by spliceosome‑GFP) that overlap with regions of high ATP‑sensor loss and elevated γH2AX, indicating DNA damage from splicing catastrophe.
• Prediction 3: Combining a mild ATP‑lowering agent with the spliceosome inhibitor H3B‑8800 will produce synergistic cell death exclusively in SF3B1‑mutant lines (Bliss synergy score > 10), whereas wild‑type cells remain unaffected.
• Prediction 4: Patients with high tumor‑mutant spliceosome allele frequency and low tumor ATP signature (derived from PET‑FLT or MRI‑based metabolic imaging) will derive greater benefit from metabolic‑stress‑based therapies in retrospective cohorts.

Falsifiability

If ATP depletion fails to induce a detectable liquid‑to‑gel transition of mutant spliceosomes, or if gelation does not correspond with a sharp increase in deleterious isoform noise and loss of viability, the hypothesis is refuted. Conversely, observing the predicted ATP‑dependent phase change and its cytotoxic consequences would substantiate the model.

Therapeutic Implication

The entropy threshold frames a metabo‑synthetic‑lethal strategy: modest metabolic stressors that lower ATP just enough to push mutant spliceosomes over the phase‑transition line could selectively kill splicing‑addicted cancer cells while sparing normal tissue, offering a combinatorial avenue beyond direct spliceosome inhibition.