Immunotherapies involving chimeric antigen receptors, or CARs, have shown some clinical success against blood cancers, but scientists have yet to figure out how to make them work against solid tumors. Now, a team of researchers has discovered that adding a molecular structure to CAR immune cells can make them more effective against a range of tumor types.
In a paper published Feb. 2 in Nature Biotechnology, scientists from St. Jude Children’s Research Hospital in Memphis described how they added an “anchor” molecule to human natural killer cells and T cells that had been engineered with CARs. The cells were implanted into mice with brain, lung and bone tumors, where they dramatically improved their suvival rates in some models.
To understand why the anchor molecule was effective, it helps to understand why the bond between CAR immune cells and some types of cancer cells can be too weak for the immune cells to work. CARs are engineered to recognize a protein called an antigen, found on the tumor’s surface. CARs bind to the cancer cells via a connection, known as an immune synapse, between the receptor and the antigen. When the synapse forms, the receptor sends signals down into the interior of the immune cell that it’s time to kill the cancer.
But this process doesn’t always work as well as it should with conventionally designed CAR immune cells. The signaling cascades that ultimately give the immune cell the greenlight to attack are very complex and can be difficult for the cell to understand. Scientists have theorized that this could be one of the reasons why CAR immunotherapies fail.
Looking to improve the quality of the signal, the researchers focused on making the signaling cascade components more organized. Much like keeping email responses to a single thread makes conversations easier to follow, organizing the signaling molecules could make it easier for the inside of the immune cell to understand messages from the CAR receptor.
To improve how the signaling molecules were arranged, the researchers looked to PDZ binding moieties, portions of molecules that recognize and bind to other molecules. PDZ binding moieties are found in around 400 proteins, many of which are found in cells with high polarity—like neurons, epithelial cells and endothelial cells—meaning among other things that the signaling molecules within them are structured in an organized way.
The researchers homed in on the proteins CRTAM and Scribble, the PDZ binding moieties of which are involved in immune synapse formation. The proteins’ interactions have been studied in both T cells and natural killer cells, and CRTAM is known to be involved in recognizing and destroying cancer cells, they noted in their paper.
Four amino acids from the PDZ binding moiety of CRTAM were taken to build the anchor protein, which was then added to a CAR on natural killer cells. This gave the CAR a way to effectively bind to the Scribble protein. Cell analyses showed that this approach “tuned” the immune synapse: There was greater signal transduction between the CAR and the inside of the cell. This translated to more efficient cancer cell elimination in every type of CAR the scientists tried, both in petri dishes and in mouse models of bone and lung cancer. In one model of localized bone cancer, around 80% of the mice with the newly designed CAR lived until the end of the four-month study. In contrast, half of the untreated mice had died by day 47.
The success of the anchor protein in CAR natural killer cells led the scientists to try it in CAR T cells, too. Here, they also saw improved cell signaling, along with a survival benefit in mouse models of glioma and aggressive bone cancer. Sixty percent of glioma models who received the PDZ CAR T cells cleared their tumors, while the median survival rate for the bone cancer models was about 30 percent higher in the PDZ CAR T group than in controls.
Scientists have studied different CAR designs in an effort to boost the therapy’s efficacy, but this is the first time anyone has tried using an anchor molecule to improve cell signaling, the St. Jude researchers said. Given its novelty, its ability to work across CAR types and in both natural killer and T cells—and the fact that it doesn’t require any new technology—the design should be tried in patients, senior author Stephen Gottschalk, M.D., said in a press release.
“Synapse tuning via anchor domains represents a fertile realm to explore for CAR-based immunotherapies,” the researchers wrote.