Tumor metastasis is the leading cause of cancer-related mortality. During the metastatic process, disseminated tumor cells establish metastatic tumor microenvironments that function as an “ecosystem” supporting their survival. Accumulating evidence indicates that the features and states of the tumor microenvironment are closely associated with the efficacy of immunotherapy. Tumor cells can reshape specific immune microenvironments through intrinsic genetic alterations, thereby establishing immune barriers that limit therapeutic responses .
However, conventional tumor immunological screening approaches typically rely on tissue dissociation, resulting in the loss of critical spatial information within the microenvironment. Consequently, a systematic understanding of how tumor-intrinsic genetic alterations sculpt specific immune microenvironments and ultimately determine immunotherapy responses remains limited, and appropriate experimental strategies to address this question are still lacking.
In a study published in Cell, the team led by Prof. WANG Guangchuan from the Center for Excellence in Molecular Cell Science (Shanghai Institute of Biochemistry and Cell Biology) of the Chinese Academy of Sciences, in collaboration with the teams of Luonan Chen at Shanghai Jiao Tong University and Naihe Jing at the Guangzhou National Laboratory, developed a novel spatial perturbation screening platform termed CLIM-TIME (CRISPR–Laser-captured microdissection Integration Mapping of Tumor Immune Microenvironment).
This platform integrates CRISPR-based genetic screening with laser capture microdissection (LCM) of metastatic lesions, combined with transcriptomic profiling, deconvolution analysis, and immunofluorescence imaging, enabling systematic spatial dissection of metastatic tumor microenvironments. Using this approach, the study establishes causal links between tumor-intrinsic mutations, the spatial organization of lung metastatic microenvironments, and immunotherapy responses, and identifies key targets that remodel metastatic niches to enhance T cell infiltration and therapeutic efficacy.
Leveraging the CLIM-TIME platform, the researchers performed a systematic analysis of lung metastatic lesions driven by the loss of 391 commonly mutated tumor suppressor genes, profiling multiple dimensions including transcriptomic features, spatial immune cell distribution, immune evasion capacity, and responses to T cell–based therapies. The metastatic tumor microenvironments induced by distinct genetic perturbations were classified into seven major types, each characterized by distinct transcriptional programs.
Integrated analyses revealed that metastatic lesions resistant to T cell therapy predominantly exhibited two microenvironmental states: a myeloid-rich, T cell–excluded phenotype (T-IE/M-II) and an immune desert (ID) phenotype. Notably, the T-IE/M-II microenvironment was marked by significant enrichment of collagen-associated extracellular matrix (ECM) remodeling pathways and signaling pathways involving focal adhesion kinase (FAK).
Further analyses demonstrated that loss-of-function mutations in multiple tumor suppressor genes within the Hippo signaling pathway preferentially induced the formation of T-IE/M-II metastatic niches and conferred resistance to T cell–based therapies. Mechanistic studies using Nf2 loss as a representative example revealed markedly reduced T cell infiltration, increased myeloid cell accumulation, and enhanced collagen deposition within these metastatic lesions.
At the molecular level, Nf2-deficient tumor cells upregulated a suite of ECM-related genes, including the collagen crosslinking enzyme LOXL2, and simultaneously reprogrammed recruited myeloid cells to express ECM-associated genes and adopt an M2-like immunosuppressive phenotype. Together, these changes cooperatively promoted T cell exclusion and functional exhaustion.Clinical data analyses further showed that metastatic tumors with high LOXL2 expression exhibited significantly reduced T cell infiltration and poor responses to immunotherapy.
Genetic ablation of LOXL2 or mutation of its catalytic site led to decreased collagen deposition and myeloid cell infiltration, accompanied by increased T cell infiltration, effectively converting “T cell–resistant” metastatic lesions into therapy-responsive states. Across multiple mouse and human metastatic tumor models, both genetic targeting and small-molecule inhibition of LOXL2 markedly enhanced the efficacy of TCR-T and CAR-T cell therapies.
Collectively, this study establishes, for the first time at a high-throughput scale, a causal framework linking tumor-intrinsic genetic perturbations, microenvironmental architecture, and immunotherapy outcomes, providing a powerful new platform for dissecting mechanisms of immunotherapy resistance in metastatic disease. Building on this causal framework, the authors further applied machine learning approaches to identify key causal genes that determine immune states within metastatic tumors and developed more accurate predictive models for immunotherapy responses. These findings offer important theoretical and experimental foundations for overcoming immunotherapy resistance in metastatic cancer and for the development of novel therapeutic strategies and targets.
Reference: https://doi.org/10.1016/j.cell.2025.12.042
Appendix：
