AI-designed miniproteins target cancer cachexia; in one mouse model, they improved the response to immunotherapy
AI-designed miniproteins target cancer cachexia; in one mouse model, they improved the response to immunotherapy
Cachexia causes weight and muscle loss in people with cancer and can make treatment harder to tolerate. In a paper published on 7 July, researchers described short proteins that intercept GDF15 signaling. In mice bearing tumors that produced human GDF15, the treatment reduced wasting; in one model, combining it with anti-PD-1 controlled tumor growth more effectively.
Some tumors raise blood levels of GDF15, a stress-response protein. GDF15 binds to the GFRAL receptor on neurons in the brainstem and reduces appetite. This signaling pathway contributes to cachexia, which causes loss of fat and muscle, weakens patients, and can interfere with cancer treatment.
The authors provided protein-design algorithms with the surface of GDF15 that binds to the GFRAL receptor. The algorithms proposed sequences for short proteins. The team synthesized the candidates and measured their binding to GDF15. A miniprotein occupies the same surface as the receptor, preventing GDF15 from transmitting its signal to the brain.
To test this mechanism, the authors changed individual amino acids at the binding interface, which weakened binding. Matching charge substitutions in GDF15 and the miniprotein partly restored the miniprotein's ability to block the signal. The computational model predicted contact between electrically charged regions of the two proteins, and the laboratory experiment confirmed it.
The researchers then administered the miniproteins to mice bearing three types of tumors that produced human GDF15. In all three models, treatment reduced weight loss and prolonged survival; each survival analysis included ten animals. In one model, the mice ate more and retained adipose tissue and gastrocnemius muscle tissue.
In a separate experiment involving mice bearing tumors that produced human GDF15, the researchers combined the miniprotein with anti-PD-1, a drug that releases an inhibitory brake on immune T-cells. The combination suppressed tumor growth more effectively than either treatment alone. More T-cells capable of killing tumor cells entered the tumors. When the authors depleted these T-cells, the effect of the combination disappeared. This conclusion applies only to this mouse model.
An antibody against GDF15 has already been tested in humans. In a 2024 trial, 187 patients with advanced cancer, cachexia, and high GDF15 levels were assigned to one of three doses of the antibody ponsegromab or to placebo. At the 400 mg dose, participants gained more than 5% of their initial body weight over 12 weeks; participants in the placebo group lost about 0.5 kg. Ponsegromab is a different drug, but its results provide clinical support for targeting GDF15 itself.
For the new miniproteins, human trials will need to determine how long they remain in the blood, what dose is safe, and whether the approach works in tumors with different levels of GDF15.