Why does the immune system become less effective at fighting cancer as we age? A new study points to two molecular brakes on CD8 T cells, the immune cells that can directly kill cancer cells. Removing Dusp5 or Zfp219 improved tumor control in mouse experiments, but by different routes. Human data linked higher ZNF219 activity with age and poorer outcomes after immunotherapy. None of this shows that targeting either gene will improve cancer treatment in people.
The peer-reviewed study was published in Cell.
What T cells do against cancer
T cells are white blood cells that recognize specific molecular targets. CD8 T cells can find abnormal cells and destroy them by releasing toxic proteins, including granzymes, that trigger the target cell to die.
A successful anti-cancer response requires more than the ability to kill. Tumor-reactive T cells must multiply, enter the tumor, persist in a hostile environment and retain their effector function. With age, those steps often become less efficient. Both the T cells themselves and signals from the older tumor microenvironment can push them toward dysfunctional states.
This matters for immunotherapy. Immune checkpoint inhibitors such as anti-PD-1 drugs release inhibitory signals that restrain T cells. They can produce durable responses in some cancers, but they do not work for every patient. If older T cells cannot expand or maintain cytotoxic activity, removing one checkpoint may not be enough. Our coverage of a personalized cancer vaccine combined with checkpoint blockade shows how much modern cancer treatment depends on a sustained T-cell response.
A targeted CRISPR screen inside tumors
The researchers used CRISPR-Cas9 to switch off selected genes in naive mouse CD8 T cells. Their library targeted 60 transcriptional regulators and nuclear proteins, plus control genes, using 412 guide RNAs in total. This was a focused screen based on prior biology, not a genome-wide search.
The T cells were engineered to recognize ovalbumin, a model antigen placed in B16 melanoma cells. Researchers transferred the edited T cells into young or aged mice, implanted the tumors and recovered T cells from tumors and tumor-draining lymph nodes 15 days later.
Perturb-seq combined each cell’s CRISPR edit with single-cell RNA sequencing. That allowed the team to ask not only which edits helped T cells persist, but also which functional state each edited cell entered.
Several gene deletions increased T-cell abundance. Dusp5 and Zfp219 stood out because their loss pushed cells toward an effector-like state and improved tumor control in follow-up experiments.
DUSP5 affected the ability to multiply
DUSP5 is a phosphatase, an enzyme that removes phosphate groups from other proteins. It normally provides negative feedback to ERK1/2 signaling, a pathway involved in cell activation and proliferation.
Deleting Dusp5 increased activated ERK1/2 and raised the proportion of dividing tumor-reactive T cells. Those cells became more abundant, showed stronger effector-like programs and improved tumor control in both young and aged mice. The effect was therefore not specific to aging.
The result identifies DUSP5 as a regulator of T-cell expansion, but stronger ERK signaling is not automatically safe. A pathway that helps immune cells multiply can also produce unwanted activation or altered cell fate. The study used genetic deletion in engineered mouse T cells and did not test a DUSP5 drug.
ZNF219 affected the cells’ killing machinery
Zfp219, called ZNF219 in humans, is a DNA-binding transcriptional repressor. It can reduce access to gene programs rather than changing only one short signaling event.
Deleting Zfp219 changed chromatin, the DNA and protein structure that controls which genes are accessible. In aged tumors, the edited T cells increased activity around the Gzma and Gzmb genes and restored production of the cytotoxic proteins GZMA and GZMB. Tumor control improved in aged mice, but not young mice.
Transfer experiments suggested that the aged environment helped drive Zfp219 upward. Young T cells placed in aged hosts increased Zfp219, while aged T cells placed in young hosts did not show the same rise. That makes Zfp219 particularly interesting as a link between an older tumor environment and reduced T-cell killing capacity.
What the mouse immunotherapy experiment showed
Anti-PD-1 treatment alone was less effective in the aged mouse model. Combining anti-PD-1 with Zfp219-deficient T cells improved tumor control, increased tumor-infiltrating CD8 T cells and expanded cells producing granzymes. Three of eight aged mice experienced complete tumor clearance. The three cleared mice that were rechallenged resisted new tumor growth.
That is compelling preclinical evidence, but the model was highly engineered. The T cells recognized an artificial tumor antigen, the gene was deleted before transfer and the tumors were established mouse cell lines. Results in a small number of mice do not predict the response rate or safety of treatment in people.
What the human evidence actually showed
The study did not edit DUSP5 or ZNF219 in cancer patients. Instead, the researchers reanalyzed previously published single-cell datasets from tumor-infiltrating CD8 T cells across breast, lung, melanoma, colorectal and ovarian cancers.
DUSP5 expression was similar across age groups. ZNF219 expression increased with patient age. In separate pan-cancer cohorts treated with anti-PD-1 therapy, higher ZNF219 expression in CD8 T cells was associated with poorer overall survival and progression-free survival, both before treatment and in samples collected during or after treatment.
Those are observational associations. Older patients can differ in cancer type, stage, prior treatment, health status and many other factors. The analysis cannot prove that ZNF219 caused poorer immunotherapy response or that lowering it would improve survival.
The team also increased ZNF219 in primary human T cells from healthy donors. The cells persisted less well after transfer into immunodeficient mice carrying engineered human melanoma tumors, although their expansion in laboratory culture was not reduced. This is a human-cell experiment in mice, not a patient study.
No treatment exists yet
There is no approved cancer therapy that targets DUSP5 or ZNF219 for age-related T-cell dysfunction. The study does not establish a dose, delivery method or safe therapeutic window for either target.
Gene deletion might eventually be explored in T cells edited outside the body for adoptive cell therapy. A drug would face a different challenge because DUSP5 and ZNF219 may perform useful functions in other cells and tissues. Any treatment would need to improve tumor immunity without causing excessive immune activation, autoimmunity or harmful changes in T-cell differentiation.
Major limitations
- The CRISPR library covered only 60 selected targets rather than the full genome.
- The main experiments used engineered mouse T cells and artificial tumor antigens.
- Mouse group sizes were modest, including eight animals in the key combination-treatment group.
- Human survival findings were retrospective and observational.
- The study did not test gene inhibition in patients or evaluate long-term human safety.
- Different cancers and older patient populations may not share the same immune-aging mechanism.
What would need to happen next
Researchers first need to reproduce the effects in human tumor-reactive T cells from younger and older patients. Studies should test multiple cancers, naturally occurring tumor antigens and more realistic immune environments. They also need to compare partial inhibition with complete gene deletion and determine whether the effects last.
If those studies remain encouraging, an early clinical trial would begin with safety and feasibility, likely in a carefully selected cell-therapy setting. Only controlled patient studies could determine whether targeting either pathway improves response rates, progression-free survival or overall survival.
Bottom line: DUSP5 and ZNF219 help explain two parts of age-related T-cell dysfunction in experimental cancer models. One limits proliferation; the other restrains cytotoxic genes in aged tumors. The work offers plausible targets for further research, not evidence that immune aging has been reversed or that a new human cancer treatment is ready.