DUSP4 protein keeps CAR-T cells killing colorectal tumours, NUS study finds

A protein called DUSP4 made CAR-T cells multiply faster and shrink colorectal tumours more effectively than standard CAR-T cells in laboratory models, according to a study published in Nature Communications on 15 August 2026 by researchers at the Yong Loo Lin School of Medicine, National University of Singapore (NUS Medicine), and collaborators.

The finding matters because CD8+ T cells — the immune system's front-line cancer killers — tend to burn out inside solid tumours. They stop multiplying. They stop attacking. That exhaustion is one of the main reasons immunotherapies have struggled against solid tumours, including colorectal cancer.

What did the researchers measure?

The team, led by Associate Professor Zhang Yongliang of the Department of Microbiology and Immunology and the Immunology Translational Research Programme at NUS Medicine, combined three lines of evidence: clinical data from colorectal cancer patients, experiments with human cells, and preclinical models.

In models lacking DUSP4, colorectal tumours grew more numerous and larger, and CD8+ T-cell responses weakened. Those cells produced fewer cancer-killing molecules and showed more markers of exhaustion. When the researchers analysed tumour tissue from 105 colorectal cancer patients, lower DUSP4 levels correlated with poorer overall survival. An independent dataset showed a similar trend.

Crucially, the effect appears immune-specific. Removing DUSP4 from cancer cells themselves had little effect on tumour growth. Removing it from blood-forming cells weakened CD8+ T-cell responses and increased tumour development in the preclinical models.

Assoc Prof Zhang said: "CD8+ T cells are among the immune system's main weapons against cancer, but prolonged exposure to tumours can gradually weaken their ability to function. Our findings suggest that DUSP4 helps keep these cells in a state where they can continue to multiply and attack cancer cells effectively."

How does the mechanism work?

The team mapped the pathway. DUSP4 regulates signalling through ERK2 and KLF2. Without sufficient DUSP4, ERK2 activity rises, which lifts KLF2 levels, which in turn suppresses the genetic programmes CD8+ T cells use to kill cancer. Keep DUSP4 active, and the T cells stay functional.

That mechanism gave the researchers a concrete engineering target for CAR-T therapy, where a patient's T cells are genetically modified to recognise cancer cells.

What happened with the engineered cells?

Using CRISPR activation (CRISPRa), the team boosted DUSP4 activity in human CAR-T cells. In laboratory experiments against colorectal cancer cells, the DUSP4-enhanced cells multiplied more readily and killed more cancer cells than standard CAR-T cells. In laboratory models, animals treated with the enhanced cells carried tumours that were smaller and lighter than those in animals receiving standard CAR-T cells.

Professor Nicholas R.J. Gascoigne, also of NUS Medicine's Department of Microbiology and Immunology and Immunology TRP, and a professor at the Immunology Center of Georgia and Georgia Cancer Center, Augusta University, said: "By increasing DUSP4 activity, we were able to improve both the growth and cancer-killing capacity of CAR-T cells in our experimental models. This gives us a potential strategy for engineering T cells that are better able to maintain their function in the difficult environment of solid tumours."

What are the study's limits?

The researchers flagged a significant caveat, and R&D managers evaluating the translational distance should weigh it carefully. The CAR-T results remain preclinical. The colorectal cancer cells used in the CAR-T model were engineered to carry CD19 — the target the CAR-T cells recognise — rather than a native colorectal cancer antigen. Further studies are needed before the approach can be assessed clinically.

The 105-patient survival correlation, while replicated in direction in an independent dataset, is observational rather than causal evidence in humans.

Next steps for the group are already defined: work out how DUSP4 itself is regulated within tumours, identify the other molecules through which it controls T-cell function, and explore practical ways of enhancing DUSP4 activity in engineered T cells — groundwork the team hopes will feed new approaches for T cell-based immunotherapies in colorectal cancer and other solid tumours.