Single-cell and spatial transcriptomics define a C3-centred multicellular ecology in colorectal liver metastasis

Read the full article

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Colorectal liver metastasis (CRLM) is shaped by malignant-cell plasticity and microenvironmental remodeling, but how malignant, stromal, vascular and myeloid states assemble spatially remains unclear. We sought to define recurrent multicellular programs associated with CRLM. We integrated a 136,995-cell single-cell atlas with copy-number inference, consensus non-negative matrix factorization, regulatory-network and patient-aligned trajectory analyses, intercellular-communication modeling, four Visium sections, bulk-cohort survival analyses, virtual knockout, connectivity-map screening and molecular docking. Three malignant epithelial programs were resolved along a C1-to-C2-to-C3 continuum. The rare C3 state displayed secretory, invasive and epithelial-to-mesenchymal transition features and was associated with shorter recurrence-free survival. Fib- COL10A1 fibroblasts and Endo- GJA5 endothelial cells formed a recurrent stromal–vascular scaffold, whereas CTSL -program macrophages occupied focal interfaces whose geometry varied among metastases. Spatial pathway and multiview analyses implicated MAPK- and VEGF-associated activity. An INHBACST6LBHFSTL3 score was associated with recurrence in TCGA and showed partial external validation. Virtual knockout indicated convergent CST6 / LBH network effects, and connectivity-map analysis nominated PD-0325901 as a candidate for reversing the associated state; docking provided exploratory structural support for CST6 but not LBH . These analyses define a spatially flexible C3-centred ecology and a MAPK-linked network vulnerability in CRLM. The framework is computational and requires experimental validation.

Article activity feed