Targeted therapies specifically attach to and hinder cancer-causing proteins, but cancer cells can quickly evolve to thwart their action. A second drug class, immunotherapies, harnesses the immune system to attack cancer cells, but these agents often cannot detect the disease-causing changes happening inside cancer cells, which look normal from the outside.
Now, a new study led by researchers from Perlmutter Cancer Center at NYU Langone Health describes a strategy to overcome these limitations based on several insights. First, the research team recognized that drugs called “covalent inhibitors” form stable attachments with the disease-related proteins they target inside cancer cells. They also knew that proteins once inside cells are naturally broken down and presented as small pieces (peptides) on cell surfaces by major histocompatibility complex (MHC) molecules. Once bound to MHC, peptides are recognized as foreign by the immune surveillance system if they are sufficiently different from the body’s naturally occurring proteins.
Although tumor cells usually develop ways to escape immune surveillance, the study authors reasoned that a covalent inhibitor tightly bound to its cancer-related peptide target could act as an MHC-displayed “flag” recognized by immune proteins called antibodies. The team then engineered such antibodies and joined them to another antibody known to recruit T lymphocytes, the “killer cells” of the immune system, to form bi-specific antibodies that destroyed tumor cells.
“Even when genetic and other changes frustrate targeted therapies, they often still attach to their target proteins in cancer cells, and this attachment can be used to label those cells for immunotherapy attack,” says co-corresponding study author Shohei Koide, PhD, a professor in the Department of Biochemistry and Molecular Pharmacology at NYU Grossman School of Medicine and a member of Perlmutter Cancer Center. “Further, our system, conceptually, has the potential to increase the efficacy of any cancer drug when attached to the drug’s disease-related target where the combination can be displayed by MHCs.”
Published online October 17 in Cancer Discover, a journal of the American Association for Cancer Research, the new study tested the researchers’ approach on two U.S. Food and Drug Administration (FDA)–approved targeted drugs, sotorasib and osimertinib. Recently approved based on a study co-led by NYU Langone researchers, sotorasib works by attaching to an altered form of the protein KRAS called p.G12C, in which a glycine building block has been mistakenly replaced by a cysteine in its structure. This change causes the KRAS protein switch to become “stuck in the on mode” and signal for abnormal growth. Sotorasib effectively blocks this activated signal to start, but cancer cells rapidly become resistant.
In experiments with KRAS mutant cancer cells grown in a dish (cell cultures), the team’s HapImmune antibodies recognized them and recruited T cells, which led to the killing of treatment-resistant lung cancer cells, in which sotorasib attached to its target, KRAS p.G12C, and was displayed by MHCs. The team also developed bi-specific antibodies that bound to a peptide “flagged” with osimertinib, a drug that targets an altered form of epithelial growth factor receptor seen in other lung cancers, as well as prototypes that “saw” the drug ibrutinib when linked to its target, BTK, showing the technology’s broad potential, the researchers say.
The study revolved around the process where proteins inside human cells are broken down and replaced as a part of the normal lifecycle. Alongside this turnover runs an inspection system, in which protein fragments are delivered to a cell’s surface. T cells inspect these displayed complexes, and can notice, for instance, when a cell is displaying viral proteins, a sign that the cell is infected with a virus. The T cells then direct the killing of the virally infected cells.
The immune system can in some cases also recognize cells with cancerous changes underway inside by the proteins they display on their surfaces. However, because cancer-driving proteins arise from normal proteins, with differences between cancerous and normal fragments often minute, the system struggles to tell them apart. Even when patients develop T cells that can see these small differences, tumors respond with mechanisms designed to exhaust anti-tumor cells. In seeking to counter these mechanisms, the team’s central realization was that, among the proteins displayed by MHCs are fragments carrying drugs taken in by cells, which could be targeted by antibodies.
The current study also found that the team’s platform was effective against KRAS p.G12C mutant cells with different MHC types, also called human leukocyte antigen (HLA) supertypes. Usually, there is a strict pairing between MHC/HLA types and antibodies built to interact with certain T cells, which could potentially restrict the number of patients that could treated by this approach. The new study showed that the team’s antibodies recognize multiple MHC/HLA types, and so, in principle, could be deployed in 40 to 50 percent of the U.S. patient population with tumors bearing KRAS p.G12C.
“Our results further show that the antibodies attach to drug molecules only when presented by MHCs on cells, and so could be used in combination with a drug,” says study co-corresponding author Benjamin G. Neel, MD, PhD, the Laura and Isaac Perlmutter Director of Perlmutter Cancer Center. “When used in combination with such antibodies, a given drug would only need to flag cancer cells, not fully inhibit them. This creates the possibility of using drugs at lower doses, potentially, for reducing the toxicity sometimes seen with covalent inhibitors.”
Moving forward, the research team plans to study their platform in live animal models, and using more pairs of drugs and the disease-related protein fragments they target.
Along with Dr. S. Koide and Dr. Neel, the study was led by first authors Takamitsu Hattori and Lorenzo Maso of Perlmutter Cancer Center, as well as by Kiyomi Araki, Akiko Koide, James Hayman, Padma Akkapeddi, and Injin Bang. The work was supported by National Institutes of Health grants R21 CA246457, R21 CA267362, and R01 CA248896, as well as by Perlmutter Cancer Center Support grant P30CA016087.
Hattori, Maso, Dr. S. Koide, Dr. A. Koide, and Dr. Neel are listed as inventors of pending patents related to the study. NYU Langone has entered into a research and option agreement with ATP Research and Development to develop these inventions and potentially to form a startup company, with Dr. Neel and Dr. S. Koide as co-founders, to license and commercialize them. Dr. Neel holds equity in Northern Biologics, LTD; Navire Pharma; and Lighthorse Therapeutics, and holds equity and receives consulting fees from Arvinas, Inc.; Recursion Pharma; and GLG group. He also receives research funding from Repare Therapeutics. Dr. S. Koide is a co-founder and holds equity in Revalia Bio and receives research funding from Puretech Health, Argenx BVBA, and Black Diamond Therapeutics. These relationships are managed in keeping with the policies of NYU Langone.