CAR T Cell Adds Scorpion Venom to Tackle Tumor Heterogeneity

Coauthor Christine Brown, the deputy director of the T Cell Therapeutics Research Laboratory at City of Hope who also provides oversight for clinical trials investigating CAR therapies for glioblastoma, says her goal with the new study was to “really expand the arsenal of viable CAR therapies for brain tumors,” adding that understanding the differences between various therapies will allow researchers to combine them into more effective treatments.

See “The Next Frontier of CAR T-Cell Therapy: Solid Tumors” 

CAR T cell therapy is approved for some forms of leukemia and lymphoma and has shown some antitumor activity in clinical trials for the treatment of glioblastoma, but patient response rates have remained frustratingly low. Glioblastoma poses a particular challenge to CAR therapies because of tumor heterogeneity: brain cancer cells have considerable phenotypic differences and not all of them carry the specific antigen that a given CAR T cell targets, which means some malignant cells can evade attack. “We wanted to develop an immunotherapy that recognized the greatest frequency of tumors and the greatest proportion of cells within those tumors,” Brown explains.

The idea to use scorpion venom as a targeting agent came from coauthor Michael Barish, a developmental and stem cell biologist at City of Hope. Chlorotoxin was known to bind to glioblastoma tumors, although the cell surface receptor it grabs hold of hasn’t yet been identified. Fluorescently labeled chlorotoxin can mark tumor cells during brain surgeries, and the peptide had also served as a coating for chemotherapeutics vehicles, but it had never been incorporated into a CAR. So they challenged graduate student and coauthor Dongrui Wang to generate a CAR that includes chlorotoxin and recognizes brain cancer cells. Wang succeeded, and his development led to both his dissertation and the team’s most recent publication.

“There’s a long history of [chlorotoxin] being used for other purposes besides what we know about [it] being toxic and poisonous to humans,” says Bryan Choi, a resident neurosurgeon and CAR T cell researcher at Massachusetts General Hospital who was not involved in the study. “The idea of it being used for glioblastoma or in CARs is . . . an exciting translation of that technology.”

In the first step of the experiment, Brown and colleagues screened the chlorotoxin peptide on a panel of cells from 23 tumor samples resected from 15 glioblastoma patients. Chlorotoxin clung strongly to almost all patient tumors, with more than 80 percent of the cells binding chlorotoxin. Only samples from two of the 15 patients revealed binding in less than 40 percent of total cells. Similarly, when examining chlorotoxin binding in patient-derived glioblastoma cell lines, the team found that 21 of 22 cultured cell lines showed greater than 70 percent binding. In contrast, three other targeting agents currently being evaluated for CAR T cell therapy bound less broadly and less uniformly to cells from the tumor samples and cultured lines. According to Brown, these results demonstrate chlorotoxin’s ability to address heterogeneity both within and between tumors. 

They then used Wang’s chlorotoxin CAR and showed that T cells could express the CAR and recognize malignant glioblastoma cells that are known to bind chlorotoxin. To test the CARs in vivo, the researchers engrafted patient-derived glioblastoma cells into the brains of immunocompromised mice and found that the CAR T cells had potent anti-tumor activity on par with another CAR that targets the IL13Rα2 receptor and is a potential candidate for glioblastoma treatment.

An immune synapse (yellow) forms at the interface between the large nucleus of a malignant glioblastoma cell to the left (blue) and the smaller nucleus of a CAR T cell to the right (also blue). The synapse indicates T cell activation, which leads to tumor killing.

CITY OF HOPE

One of the dangers of CAR T cell therapy is off-target binding, when the CARs cause adverse effects by killing the wrong cells. To assess the novel CAR’s safety in pre-clinical models, the researchers measured binding in a panel of nontumor human cells, revealing little to undetectable off-target recognition. In other words, the T cells almost exclusively attacked brain cancer cells and spared normal cells. A mouse model echoed the findings: the CAR T cells only targeted glioblastoma cells, and there were no adverse reactions in the mice when the therapeutic cells were delivered.

“I think [the murine model] is a very good surrogate as the chlorotoxin receptor is conserved between humans and mice,” says Wang. “So we’re pretty confident that what we found in mice regarding the safety and the efficacy of this CAR can be translated to humans.” Choi points out, however, that without knowing exactly how the target “interacts with cancer cells specifically, it’s really hard to know what the potential deleterious effects are on normal human tissues. . . . It’s really up to Phase one clinical trials to sort that out.”

Robbie Majzner, a CAR T cell researcher and pediatric oncologist at the Stanford University Medical Center who did not participate in the research, agrees that the newly engineered CAR “definitely is the beginning of overcoming tumor heterogeneity, and it’s clearly a good choice of antigen because it’s broadly expressed on individual glioblastoma cells at a high percentage.” He notes that chlorotoxin bound roughly 80 percent of cells, which means there still is a population of cells that must be overcome. “No single antigen or binder or protein is going to end up [being] the one solution,” he says. “We’re going to need multi-antigen targeting . . . chlorotoxin plus something else.” 

The study raises questions about why chlorotoxin binds to such a high proportion of glioblastoma tumor cells, the biology of which the researchers hope to investigate in future research. Barish says it likely comes down to evolution, because it was useful for a scorpion to “be able to specifically target different components of the nervous system of its prey.” The team is also interested in studying membrane-associated matrix metalloproteinase-2 (MMP2) after the experiment revealed that chlorotoxin CAR binding is correlated with the expression of MMP2 on target cells. This molecule appears to be a critical component of binding and activating the novel CAR, though more research is needed to figure out MMP2’s role in successful CAR binding. 

Brown’s team is currently screening patients for a new clinical trial to determine the safety of the chlorotoxin CAR developed in this study. Once they find their candidates, they’ll engineer patient-specific CAR T cells to be delivered locally via a catheter device implanted in the brain during tumor removal. Brown says the device will allow them to closely study how the patients respond to the treatment, further illuminating chlorotoxin’s potential role in the fight against glioblastoma.

D. Wang et al., “Chlorotoxin-directed CAR T cells for specific and effective targeting of glioblastoma,” Science Translational Medicine, doi:10.1126/scitranslmed.aaw2672, 2020. 

Amy Schleunes is an intern at The Scientist. Email her at aschleunes@the-scientist.com.

Correction (March 25): Christine Brown is the deputy director of the T Cell Therapeutics Research Laboratory at City of Hope, not the director. The Scientist regrets the error.