LLPS-Driven Membraneless Organelle Activity Landscape Reveals Molecular Subtypes, Immune States, and Prognostic Stratification in Lung Adenocarcinoma

Liquid–liquid phase separation (LLPS) drives the formation of membraneless organelles (MLOs) that regulate cellular compartmentalization and diverse biological processes. However, the landscape and clinical relevance of MLO-associated activity in lung adenocarcinoma (LUAD) remain unclear. We integrated curated MLO-associated gene sets with transcriptomic data to estimate MLO activity using single-sample gene set enrichment analysis. MLO-related molecular subtypes were identified via consensus clustering, and an integrative machine learning framework was used to develop and validate the Membraneless Organelle Risk Score (MLORS) across independent cohorts. We further examined associations of MLORS with immune features, genomic alterations, and predicted treatment sensitivity and evaluated the function and phase-separation properties of E2F7 in vitro. MLO-associated activity was broadly dysregulated in LUAD, with high-MLO and low-MLO subtypes exhibiting distinct immune profiles and survival outcomes. High MLORS was associated with poorer survival, a computationally inferred immune-suppressed microenvironmental profile, higher tumor mutational burden, and lower predicted sensitivity to selected EGFR-targeted agents and chemotherapy. E2F7 formed nuclear condensates with LLPS-like properties and promoted lung cancer cell proliferation and migration in vitro. These findings link transcriptionally inferred MLO activity to the heterogeneity and prognosis of LUAD, support MLORS as a candidate prognostic stratification tool, and identify E2F7 phase-separation behavior as a potential therapeutic vulnerability.