Background: A new vulnerability in cancer cells adapting to inflammation

A preclinical study suggests that eliminating gene functions on which cancer cells depend to survive exposure to inflammatory cytokines may improve responses to immunotherapy. Across eight mouse tumor models, the researchers identified the GPI transamidase complex and FITM2 as vulnerabilities under interferon exposure. Removing these gene functions in tumor cells enhanced responses to immune checkpoint blockade in mice.

Antitumor immune responses impose inflammatory stress on cancer cells. Immune checkpoint blockade (ICB) intensifies this environment, but tumor cells also respond to cytokines and adapt. The possibility that resulting changes in transcription and cell state could create new vulnerabilities had not been fully explored. Identifying functions needed for adaptation to inflammation could also reveal targets for disrupting cancer-cell survival in the same environment.

The study by Collins K. Cheruiyot and colleagues was published in Nature Genetics on June 9, 2026.

Key findings: Loss-of-function screens in eight models and the mechanism of FITM2 loss

The researchers treated eight mouse tumor models separately with interferon gamma (IFNγ), interferon beta (IFNβ), and tumor necrosis factor (TNF), and performed in vitro genome-wide CRISPR loss-of-function screens. To identify inflammation-induced vulnerabilities, they compared how much cells with disrupted gene function were disadvantaged under cytokine exposure relative to unexposed conditions.

The screens identified components of the glycosylphosphatidylinositol (GPI) transamidase complex and the lipid phosphatase FITM2 as interferon-specific tumor dependencies. This means that cancer-cell survival in an interferon environment depends on these functions. In subsequent mouse experiments, removing GPI transamidase complex subunits or FITM2 from tumor cells also enhanced ICB responses.

The effects of FITM2 loss were investigated by combining functional genomics, metabolomics, and experiments using drugs to modulate downstream stress pathways. Cancer cells lacking FITM2 became vulnerable to IFNγ-induced endoplasmic reticulum stress and oxidative stress, showing a pattern leading to paraptosis-like cell death. The public descriptions of the extended data include increased intracellular vesicles, analyses of cell-death markers, and experiments using compounds that modulate oxidative stress. This mechanistic analysis concerns FITM2 loss; the same process cannot be assumed to apply to loss of the GPI transamidase complex.

Implications and outlook: Connecting genetic vulnerabilities to targets for combination therapy

These findings show that tumor-cell adaptation to inflammation can create new genetic dependencies. They provide a basis for follow-up research into targeting functions that help cancer cells withstand this environment, alongside treatments that strengthen immune responses.

However, gene loss in tumor cells does not guarantee the effectiveness of a drug administered to patients. The study also included analyses of genomic and transcriptomic data from existing cohorts of patients treated with ICB, but the available public data are insufficient to establish the clinical performance of a patient-selection test or prognostic prediction system. The therapeutic effects of drugs inhibiting these targets in patients also remain unproven.

Next steps include testing whether pharmacologically inhibiting target functions produces responses similar to gene loss, and whether tumor selectivity and safety for normal tissues can be achieved. The public abstract and descriptions of the extended data make it difficult to assess effect sizes in each model and the full scope of applicability. The findings should therefore be read as proposing preclinical targets for combination therapy rather than establishing a treatment for clinical use.

The study is described in the Nature Genetics paper abstract and public descriptions of the extended data.