Researchers at Zhejiang University have designed a synthetic genetic circuit that turns a subtle flaw in cancer cells into a death sentence, while leaving healthy tissue unharmed. The approach, reported in the journal Nature, shrank colorectal and pancreatic tumors in mice and worked especially well alongside existing immunotherapy, offering a genuinely new route for treating solid cancers.
The strategy targets an obscure cellular event called ribosomal frameshifting. When a cell makes a protein, a molecular machine called the ribosome reads genetic instructions in groups of three letters at a time. Occasionally it slips backward by one letter, shifting the entire reading frame and producing gibberish protein. In healthy mammalian cells the accident is extraordinarily rare, roughly once in every 100,000 translations, and the phenomenon is mostly known as a trick viruses use to pack more genes into small genomes.
Cancer cells are different. Through high-throughput screening of random genetic sequences, the team found that particular pairs of codons, including AGAACG and ACGAGA, make the ribosome slip frequently in cancer cells but rarely in normal ones. The effect appeared across colorectal, pancreatic, ovarian and melanoma cell lines. It arises because fast-growing tumors often run short of specific transfer RNAs and amino acids, causing ribosomes to pause and slide, while the cancer cells’ quality-control systems are weakened.
The researchers, led by Mao Yuanhui at Zhejiang’s Liangzhu Laboratory, engineered a circuit they call FAST, short for frameshifting-activated synthetic therapeutic circuit. Hidden inside it, one reading frame out of step, are the instructions for the active fragment of gasdermin D, a protein that punches holes in the cell membrane. In a normal cell the ribosome stays in its lane and the lethal fragment is never produced. In a cancer cell, the slip switches frames at exactly the right moment, gasdermin D is made, and the cell dies through pyroptosis, an inflammatory form of death that also summons the immune system.
Delivered systemically with an AAV9 viral vector in mice, FAST drove marked regression of MC38 colorectal tumors and of KPC pancreatic tumors, a notoriously stubborn model that arises spontaneously, and it cut spread to the liver and lungs. The dying tumors released alarm signals that drew interferon-producing CD4 and CD8 T cells, and combining FAST with checkpoint-blockade immunotherapy strengthened the effect. Mice cured of colorectal cancer rejected a second injection of the same tumor cells, indicating lasting immune memory, while major organs showed no detectable damage and blood tests stayed normal.
Because FAST keys into a state of translational chaos shared by many cancers rather than a single mutation, it could in principle reach tumors that lack obvious drug targets, the authors say. They are cautious about the road ahead. Delivery efficiency, immune reactions to the virus and introduced protein, long-term safety and differences in frameshifting among human tumors all need work before clinical trials. Still, as a proof that cancer’s own internal disorder can be wired into a programmable kill switch, the study opens a frontier distinct from both conventional chemotherapy and CAR-T cell therapy.

Dgcity