Dissecting cross‐lineage tumourigenesis under p53 inactivation through single‐cell multi‐omics and spatial transcriptomics

Abstract

BACKGROUND: Tumour suppressor genes, exemplified by TP53 (encoding the human p53), function as critical guardians against tumourigenesis. Germline TP53-inactivating mutations underlie Li-Fraumeni syndrome, a hereditary cancer predisposition disorder characterised by early-onset pan-tissue malignancies. However, the context-dependent tumour-suppressive mechanisms of p53 remain incompletely elucidated. This study aims to investigate the disruption of cellular homeostasis and tumourigenic mechanisms following p53 inactivation across distinct cell lineages. METHODS: Trp53 (encoding mouse p53) knockout mouse model was employed to study molecular alterations under p53-deficient conditions. Multi-omics analyses - including single-cell transcriptomics, single-cell ATAC-seq, spatial transcriptomics, whole genome sequencing, and CUT&Tag - were integrated to construct a Trp53 functional cell landscape. Deep learning-based gene network models were employed to reconstruct p53 regulatory networks and simulate in silico perturbations caused by p53 loss. RESULTS: Our analyses revealed transitional dynamics in immune, stromal, and epithelial cells from normal physiology to p53-deficient states and subsequent tumourigenesis. These transitions implicated critical pathways such as cell cycle regulation, stress response, metabolic reprogramming, and immune modulation, displaying both lineage-conserved and lineage-specific features. Tumour-prone cell populations exhibiting elevated differentiation plasticity were identified across lineages within tumourigenic trajectories. Spatial transcriptomic profiling confirmed the emergence of thymic tumour-initiating T-cell clusters characterised by deterministic chromatin architectural disruptions under p53-loss pressure. Notably, we uncovered a recurrent upregulation signature of ribosomal protein genes as an early pivotal molecular event preceding malignant transformation in p53-deficient oncogenesis. Finally, we decoded the p53 downstream regulatory network and computationally evaluated the perturbation effects of genetic inactivation at single-cell resolution. CONCLUSIONS: Our results elucidate the multiscale consequences of p53 inactivation while providing valuable resources for understanding tumour predisposition associated with p53-inactivating mutations and developing clinical interception strategies. KEY POINTS: Construction of a Trp53 functional cell landscape utilising single-cell multi-omics and spatial omics technologies. Reconstruction of p53 downstream regulatory relationships with lineage heterogeneity via machine learning-based gene network modelling. Dissection of shared and lineage-specific features during cross-lineage tumourigenesis under p53 deficiency.