HHMI researchers have learned to program T cells as if they were "microscopic robots" -- to sense inputs and to respond.
Howard Hughes Medical Institute (HHMI) scientists have developed a flexible method of programming cells to find and respond to molecular signals of disease. The highly customizable system, known as synNotch, can be used to deliver therapeutic molecules to a disease site or modulate local immune activity. Ultimately, its developers say, synNotch could enable a wide range of cellular therapies for cancer, autoimmune diseases, and other ailments.
HHMI investigator Wendell Lim and his team at the University of California, San Francisco, reported on synNotch in three papers published this year in the journal Cell. In the most recent paper, published September 29, 2016, Lim and his colleagues describe how they have used synNotch to instruct immune cells to carry out specific activities in the presence of their targets, such as delivering therapeutic antibodies to tumors or triggering the release of signaling molecules that can dampen overactive immune responses.
Lim says synNotch could allow researchers to develop next-generation cancer immunotherapies and other true personalized therapies for patients. “A lot of precision medicine now is about gathering genomic information about the disease and determining statistically what drugs you should give that person,” he says. “The potential here is to take the information about someone’s tumor or disease and use it to design a therapy that really recognizes the signature of that disease, and also addresses that disease’s vulnerability.”
The synNotch system is an adaptation of a naturally occurring receptor molecule called Notch, which facilitates critical cell-to-cell communications in most organisms. Notch receptors are embedded in cells' outer membranes, with functional components protruding into both the cell’s interior and its external environment. When the exterior part of a Notch receptor connects with its molecular partner, its interior end is freed from the rest of the molecule and moves to the cell’s nucleus, where it activates specific genes.
Lim’s team created their synthetic receptors by swapping out the input and output domains of the natural Notch receptor. By selecting the right end pieces, synNotch users can pair the recognition of particular cell-surface proteins with a cellular response of their choosing. For this reason, Lim calls synNotch a universal sensor. “It is the first platform that allows us to easily and flexibly program both what a cell senses and how it responds,” he says.
In January, Lim’s team reported in Cell that they had used synNotch receptors to drive a variety of behaviors in lab-grown cells, such as producing a telltale fluorescent protein in the presence of molecular indicators of disease. Lim imagines a range of research applications for the system, including investigating how organs develop and how brain cells establish their connections.
For developing new clinical therapies, Lim says the real power comes when synNotch receptors are introduced into immune cells like T cells, which can travel throughout the body in search of their targets. In a second Cell publication also published in January, his team used synNotch receptors to improve T cells’ ability to discriminate between tumors and healthy cells. Their approach was similar to that of a cancer immunotherapy currently being tested in clinical trials called chimeric antigen receptor (CAR) T-cell therapy.
CAR T-cell therapy programs a patient’s own immune cells to recognize specific molecules on the surface of his or her cancer cells so that the immune system can launch a targeted attack.
The approach has effectively eliminated cancer cells for many patients with certain blood cancers. Expanding the strategy to treat other types of cancers presents a challenge, however, because few of the molecular signposts, or antigens, that flag cancer cells occur exclusively on tumors. So in programming immune cells to destroy any cell on which they encounter a cancer-related antigen, researchers are almost always directing the immune system to attack some healthy cells, too.
Lim and his colleagues used synNotch to make tumor-seeking T cells more discriminating. They engineered cells that launched a response only when they detected two different molecular signals, rather than relying on a single antigen to identify their targets. The team’s strategy was to create T cells with a synNotch receptor that recognizes a tumor-related antigen and responds by only triggering production of a CAR receptor specific to a second antigen. Activation of that second receptor triggers the cells to destroy their targets. In mice, T cells programmed in this way eliminated tumors whose cells had both antigens on their surfaces, while leaving healthy cells—as well as tumor cells with just one of the two target antigens—undisturbed.
In the experiments reported in their latest Cell paper, Lim and his colleagues demonstrated that in addition to directing immune cells to a specific target, synNotch receptors can also program T cells to respond to a signal in new ways. In their experiments, the researchers used synNotch receptors to drive T cells to release factors that amplify or dampen immune responses, deploy therapeutic antibodies, or develop into more specialized cells when their activating signal is present.
In mice, the researchers demonstrated that T cells with synNotch receptors could deliver therapeutic molecules specifically to tumors, altering the microenvironment without disrupting tissues elsewhere in the body. That’s important, Lim says, because many molecules that can be beneficial at a disease site can cause serious side effects elsewhere in the body.
Ultimately, Lim expects researchers will be able to use synNotch to develop customized cells that execute a set of therapeutic actions tailored to individual patients’ diseases. “Some of the most complex diseases that we face—things like cancer and autoimmunity and degeneration—are ones where there’s some sort of pathology in [the affected tissue’s] microenvironment. I think these are cases where a cell therapy that can go in and identify those pathologies and then modulate it in a systematic way could be very effective,” he says.