60 / 2026-04-02 16:22:33

Single-cell multi-omics reveals disease-specific epithelial states and links genomic instability to epithelial reprogramming in endometrial polyps

Single-cell multi-omics,endometrial polyp,disease-specific epithelial states,genomic instability,organoid-based drug screening

Endometrial polyps (EPs) are common gynecological lesions characterized by focal hyperproliferation, yet their underlying cellular and genomic mechanisms remain poorly understood. In particular, how cell-type-specific states are linked to genome organization and contribute to disease progression is largely unexplored. Here, we performed an integrative multi-omics analysis combining single-cell RNA sequencing, spatial transcriptomics, and whole-genome sequencing (WGS) to dissect the cellular architecture and genomic features of EPs.

We first constructed a high-resolution atlas of normal endometrium and EP tissues, revealing a marked expansion of epithelial cells in EPs. Re-clustering of epithelial compartments identified two disease-enriched epithelial subpopulations with distinct transcriptional programs associated with proliferation, adhesion, and metabolic remodeling. Trajectory and spatial analyses further demonstrated that these epithelial states occupy specific positions along differentiation continua and exhibit consistent spatial localization within polyp tissues, suggesting the presence of stable disease-associated cellular states.

To investigate the genomic basis of these epithelial states, we integrated WGS data and identified widespread genomic instability in EPs, with structural variations (SVs) as a prominent feature. Notably, these genomic alterations were enriched in epithelial compartments and were associated with pathways involved in DNA repair, extracellular matrix remodeling, and cell proliferation. These findings suggest that structural variations may reshape local chromatin regulatory landscapes and contribute to the emergence and maintenance of disease-specific epithelial states.

To further explore translational potential, we established patient-derived epithelial organoids that recapitulated the proliferative and molecular characteristics of EP tissues. Building upon the identified disease-associated features, we implemented an AI-assisted drug prioritization framework integrating multi-omics signals and network-based scoring. This approach identified two candidate compounds, Nintedanib and Cabozantinib, both of which effectively suppressed the growth of polyp-derived organoids.

Collectively, our study links cell-state heterogeneity with genomic instability in EPs and suggests potential connection between structural genome alterations and epithelial cell reprogramming. By integrating multi-omics profiling with functional organoid models and computational drug screening, we provide a framework for understanding disease mechanisms and advancing precision therapeutic strategies in endometrial polyps.