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

Authors

Feifei Ma, Yige Yu

Published in

Functional & integrative genomics. Volume 26. Issue 1. Nov 09, 2026. Epub Nov 09, 2026.

Abstract

Colorectal liver metastasis (CRLM) develops through malignant-cell plasticity and microenvironmental remodeling, yet the spatial principles by which epithelial, stromal, vascular and myeloid states assemble remain poorly defined. We integrated a 136,995-cell single-cell atlas, four spatial transcriptomic sections and bulk-cohort survival analysis. cNMF resolved three recurrent malignant programs-C1 Proliferative-biosynthetic, C2 AP-1 stress-inflammatory and C3 Secretory-invasive immune-like-which recurred across donors (with a rare, patient-associated tuft/chemosensory component in C3) and followed a pseudotime-associated C1-to-C2-to-C3 ordering rather than a demonstrated temporal progression. C3 cells, rare overall and of donor-dependent abundance, showed epithelial-to-mesenchymal transition and invasive features, and in exploratory analyses their signature was associated with shorter recurrence-free survival. Continuous RGS1/CTSL programs captured macrophage heterogeneity more robustly than discrete subclusters. Spatial deconvolution revealed recurrent Fib-COL10A1-Endo-GJA5 proximity and variable C3 integration, with focal CTSL-program macrophage interfaces; MAPK and VEGF pathway activity was higher in C3-rich regions, although this association depended on how C3 was measured. An INHBA-CST6-LBH-FSTL3 score, which aggregates C3 and stromal expression, was associated with recurrence in TCGA, with partial external support in GSE39582 and no demonstrated gain over clinical variables. Virtual knockout of CST6 and LBH and LINCS screening prioritized MEK inhibition (PD-0325901) as a candidate for reversing the C3-associated transcriptional state; no target-engagement claim is made. Together, these findings define a spatially variable, C3-centred multicellular ecology in CRLM and provide a testable computational framework nominating recurrence-associated signatures and network-level candidates for experimental validation.

PMID:
43002319
Bibliographic data and abstract were imported from PubMed on 10 Nov 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.