Researchers have improved the ability of cancer-killing T-cells to target pancreatic tumors rather than healthy tissue by engineering the cells to produce more receptors.
Researchers have improved the ability of cancer-killing T-cells to target pancreatic tumors rather than healthy tissue by engineering the cells to produce more receptors. As a result, the engineered cells are less likely to induce potentially dangerous or fatal side effects during cancer immunotherapy.
T-cells, a type of immune cell, naturally contain receptors that recognize foreign substances in the body. Depending on the type of T-cell, they either attack those substances or signal for a larger immune response.
Today, scientists use genetic engineering to direct these cells to attack a cancer patient’s own tumor—with impressive results. In clinical trials, children and adults with late stage leukemia that did not respond to other treatments had their cancer disappear for months to years after receiving engineered T-cells. To engineer the T-cells, clinicians first isolate T-cells from a cancer patient’s own blood. They then inject genetic instructions into the cells to produce synthetic receptors that bind to a protein (also called an antigen) found on the surface of the cancer cells. These receptors are called chimeric antigen receptors (CAR), and the cells that contain them are CAR T-cells.
Clinicians amplify the CAR T-cells in media. When there are enough of them, doctors inject the CAR T-cells into the patient. When CAR T-cells recognize an antigen on a tumor cell, the connection triggers the portion of the CAR dangling inside the cell. That section of the CAR then stimulates the T-cell to launch an immune attack. So far, the U.S. Food and Drug Administration has approved two CAR-T treatments for cancerous cells floating in the blood. It has not approved CAR-T treatments for solid tumors.
One reason is that many of the same surface receptors that populate cancer cells are also found (in lower abundance) on healthy cells. If CAR T-cells bind to healthy cells instead of tumor cells, this generates “on-target, off-tumor” side effects, which can include nausea, other types of illness, and, at least in one instance, death.
Researchers have been working on ways to improve the specificity of CAR T-cells for several years. One approach is to add a second CAR to the engineered T-cells. While this improves specificity, the effects are limited. Cancer cells mutate frequently and quickly. If one of the two proteins expressed by the tumor changes, the modified CAR-T cell loses its recognition specificity.
Juan Vera, an assistant professor at Baylor College of Medicine’s Center for Gene Therapy, and his collaborators had a different strategy. The researchers engineered T-cells to produce three receptors that recognize either a protein on pancreatic cancer cells or soluble molecules found floating around the tumor. Each engineered receptor transmits one part of the typical three-part signal that activates natural T-cells. These engineered T-cells only activate when all three components bind to their receptors, as the researchers explained in their recently published paper in Cancer Discovery.
The first engineered receptor is a CAR designed to recognize prostate stem cell antigen (PSCA) found on the surface of pancreatic cancer cells. When bound to its target, this receptor sends an internal signal that activates the T-cell. The second receptor recognizes a soluble molecule called transforming growth factor beta (TGF-β), which controls cell growth, proliferation, and death. It produces a co-stimulation signal that signals T-cells to multiply. The third receptor recognizes interleukin-4 (IL-4), a soluble molecule found in the environment surrounding tumor cells. It generates a signal for the CAR T-cell to produce interleukin-7, a chemical message that supports T-cell survival.
To mimic an “on-target, off-tumor” situation, the researchers injected cells that display the tumor surface protein, but did not produce the two soluble molecules recognized by the other receptors, into the left flank of a mouse. In the right flank, they injected tumor cells that produced all three components recognized by the engineered cells. The researchers then injected engineered T-cells, containing a fluorescent protein, into each site. After 18 days, they noticed the engineered T-cells growing in the right flank, but not the left where cells only expressed the tumor-specific protein. After 33 days, the lump in the right flank had completely disappeared.
“Regular CAR T-cells would have had the opposite effect,” Vera says.
In addition to specificity, the approach also delivers a second advantage when attacking solid tumors: the two soluble molecules the CAR-T cells recognize, IL-4 and TGF-β, ordinarily suppress immune response. Thus, the same molecules that would normally suppress T-cell activity become signals that activate a T-cell response.
The most interesting news of all, however, is that the use of three-part patterns to recognize tumor cells could be applied to other types of tumors, Vera says. Assuming clinical trials return positive outcomes, this could open the door to using CAR-T therapy to treat a much wider range of cancers.
Melissae Fellet is an independent technology writer.