Leukemia stem cells fight back against PI3K drugs by swapping one enzyme for another — but scientists found a way around it
Acute myeloid leukemia (AML) is notoriously hard to treat partly because leukemic stem cells — the root source of the disease — can develop resistance to therapies. This study investigated whether these stem cells depend on PI3 kinase (PI3K), a key cell-signaling enzyme. Researchers found that…
Leukemia stem cells fight back against PI3K drugs by swapping one enzyme for another — but scientists found a way around it
Acute myeloid leukemia (AML) is notoriously hard to treat partly because leukemic stem cells — the root source of the disease — can develop resistance to therapies. This study investigated whether these stem cells depend on PI3 kinase (PI3K), a key cell-signaling enzyme. Researchers found that targeting a specific version of PI3K, the p110-alpha isoform, was effective against leukemic stem cells.
However, when PI3K is blocked, the cancer cells adapt by downregulating EZH2, an epigenetic regulator, and compensating with a related enzyme, EZH1. This shift acts as a molecular escape hatch, allowing the cancer to survive treatment. The researchers showed that knocking down EZH1 — or simultaneously targeting both EZH1 and PI3K — re-sensitized the cancer cells, closing the escape route.
This work reveals a specific epigenetic resistance mechanism and points toward combination therapies that could prevent AML cells from evading PI3K inhibitors, potentially improving outcomes for patients who currently face high rates of relapse.
Key Findings
- Leukemic stem cells are selectively dependent on the PI3K p110-alpha isoform
- PI3K inhibition triggers compensatory upregulation of EZH1, an epigenetic enzyme
- EZH2 protein levels drop as a resistance mechanism in response to PI3K inhibition
- EZH1 knockdown resensitizes AML cells to PI3K inhibitors
- Combining PI3K and EZH1 inhibition may overcome this resistance mechanism
Implications
These findings identify a targetable epigenetic resistance pathway in AML and suggest that combination strategies — pairing PI3K inhibitors with EZH1 inhibitors — could prevent or overcome acquired resistance. This is particularly relevant for patients who relapse after initial therapy, which remains a major clinical problem in AML.
Caveats
This is a preprint (bioRxiv) and has not yet undergone peer review. The findings appear to be primarily from preclinical models (cell lines, likely mouse models); clinical validation in human AML patients would be needed before this translates to treatment changes. Summary based on abstract only.
Source: bioRxiv — 2026-04-11