GENE THERAPY TAKES MANY FORMS, BUT SOME OF THE MOST PROMISING involve taking cells from the body and altering them so that they can perform new tasks. One example is chimeric antigen receptor T-cell therapy (CAR T), the first gene therapy to get approval from the Food and Drug Administration. CAR T reprograms immune cells called T cells and improves their ability to find and destroy cancer cells. The revolutionary treatment has achieved remission rates of 70% to 90% in patients who receive it.

But applications of the therapy are limited. So far, re-engineering T cells doesn’t work for most cancers, and its side effects—including high fever and neurological complications—can be as dangerous as the cancer itself, which is why CAR T is used only when traditional therapies have failed. One possibility for getting around those obstacles involves something called a synthetic gene circuit.

A circuit on a computer makes simple, formulaic decisions in response to inputs, and gene circuits work in much the same way. The first gene circuit was created in 2000, when researchers equipped an E. coli bacterium with a “toggle switch”—a string of DNA that allowed certain genes to be turned on and off by means of external chemical signals and temperature changes. Since then a variety of circuits have been built, including one that mimics the signals bacteria use to communicate with each other, and another that spits out fluorescent proteins at regular, clock-like intervals.

Scientists now are exploring how this approach might improve CAR T. “If you want to use cells as a platform for therapies, you should fully utilize their capabilities to act as living, sensing, actuating devices,” says Wendell Lim, a synthetic biologist at the University of California, San Francisco. Lim founded Cell Design Labs, acquired by Gilead at the end of 2017, to work on this next generation of CAR T cell therapy.

Lim wants to use gene circuits to extend the abilities of engineered immune cells. He points out that T cells naturally travel through the body and sense disease, detecting subtle differences between invading cells and those that belong. “The real power of CAR T therapies will be to add greater versatility in how these cells sense and respond to other cells,” he says.

One helpful class of features would be a “kill switch,” genetic alterations that would allow CAR T cells to be disabled by a physician if the cells trigger an unwanted immune response or other dangerous side effects. One of Lim’s projects is developing an “on switch,” a trigger for the cell to turn itself on only when it senses a small-molecule drug administered by the physician to precisely tune and control CAR T-cell treatment.

Another gene circuit would add sophistication to cancer detection. The current version of CAR T-cell therapy locks only onto cells that present a single antigen, CD 19, that is found on the surface of B cells—a type of blood cell that becomes malignant in some blood cancers. A logical next step would be to find antigens that match up to other types of cancer, but that has proved difficult. “Many are starting to believe that solid cancers present no one specific tumor antigen to latch onto,” says Lim.

Gene circuits may solve that impasse. “If T cells could be programmed to recognize tumors based on two or three antigens, that might help,” he says, “which is similar to how facial recognition incorporates information about multiple points on a face and the relationship between those points.” Lim is developing circuits that can detect more than one antigen before launching an attack, and is testing this approach on classes of solid tumors that include glioblastoma and pancreatic cancer.

MIT synthetic biologist and electrical engineer Tim Lu has developed another gene-circuit therapy featuring more sophisticated sensors. In a paper published in Cell last fall, he described a sensor that will activate an immune response only if it senses two proteins—rather than just one—in a human model of ovarian cancer.

“I’m excited about the idea of creating a new class of medicines based on programming living cells,” says Lu, who recently launched a company called Senti Biosciences. “The first generation of CAR T drugs didn’t have the ability to change their activity or adapt. Modifying cells in new ways gives us that capability.”

Gene circuits might also be able to incorporate other kinds of drugs, such as the checkpoint inhibitors used in immunotherapy, into modified immune cells. That would allow them not only to destroy individual cells, but also to dispense medication to disable a tumor’s defenses against the immune system.

“We have to make CAR T cell therapies more powerful, but also more precise,” says Lim. “Genetic circuits may really be the key.”