Researchers led by teams at NYU Langone Health and its Perlmutter Cancer Center have reported on the development of antibodies that can selectively target cancer-related mutations in the extracellular domain (ECD) of the HER2 (human epidermal growth factor receptor 2) protein without attacking its nearly identical normal counterpart on healthy cells. The newly reported studies demonstrated that the antibodies, in a T cell engager format, can effectively kill cancer cells expressing relevant HER2 mutations both in vitro, and in a mouse xenograft model.
While the team acknowledges that their developments are still in the early stages, they suggest that their approach could lead to new treatments for cancer patients with HER2 mutations, which exhibit minimal side effects. “We set out to make an antibody that can recognize a single change in the 600 amino acid building blocks that make up the exposed part of the HER2 protein, which conventional wisdom says is very difficult, said Shohei Koide, PhD, a professor in the department of biochemistry and molecular pharmacology at NYU Grossman School of Medicine and member of Perlmutter Cancer Center. “The fact that we were able to detect the difference of a single amino acid so cleanly was a surprise.”
Koide is lead author of the team’s published study in Nature Chemical Biology. In their report, titled “Selective targeting of oncogenic hotspot mutations of the HER2 extracellular domain,” the investigators concluded, “These results validate HER2 ECD mutations as actionable therapeutic targets and offer promising candidates toward clinical development.”
“Selective targeting of cancers remains an important challenge in cancer drug discovery,” the authors noted. And for some proteins a single change, or mutation, in its DNA instructions may be all that it takes to tip the balance between functioning normally and causing cancer. However, while they may be oncogenic, these slightly mutated proteins may still resemble their normal versions so closely that treatments designed to target the mutants could also harm healthy cells. Cancer genomics studies have identified numerous oncogenic mutations, which may represent tumor-specific antigens and—particularly those located in the extracellular region of a cell-surface protein—represent “attractive targets for antibody-based therapeutics,” However, the team continued, “there is a … rather surprising paucity of therapeutics selective to oncogenic mutations of cell-surface antigens. One reason for this under-exploration may be the challenge of selectively recognizing an oncogenic mutation while sparing the wild-type (WT) counterpart.”
Koide et al’s studies revolve around HER2, a protein that occurs on the surfaces of many cell types and that turns on signaling pathways that control cell growth. It can cause cancer when a single amino acid swap locks the protein into “always-active” mode, which in turn causes cells to divide and multiply uncontrollably. “Among the hyperactivating mutations identified, S310F and S310Y are the most common, ‘hotspot’ mutations found in the HER2 extracellular domain (ECD) found in multiple types of cancer,” the team explained.
Cancer can also result when cells accidentally make extra copies of the DNA instructions that code for the normal version of HER2 and express higher levels of the protein on their surfaces. There are a few FDA-approved therapies, including trastuzumab and pertuzumab, that can treat this kind of cancer, but these therapies all work at the level of HER2 on the cell surface, where only low levels of the mutated version of HER2 occur. “That means we cannot mark cancer cells just by looking at HER2 levels,” said Koide, who also serves as director of cancer biologics at NYU Langone. In addition, since some approved therapies cannot tell the difference between mutant and normal HER2, they are more likely to harm healthy cells expressing normal HER2.
The researchers harnessed a new protein-engineering technique to develop antibodies that recognize only the mutant HER2. “It is generally challenging to develop an antibody that recognizes a single point mutation with high selectivity,” they noted. “Here we demonstrate the feasibility of developing antibodies that are capable of recognizing a single point mutation in the context of a large antigen, through a carefully designed library sorting strategy and a structure-guided, iterative engineering approach.”
Using a process that mimics natural antibody development, the researchers subjected antibodies to multiple rounds of mutation and selection, looking for variants that recognized the mutant HER2 but not the normal version. By taking atomic images with a cryo-electron microscope, the team saw how their new antibodies interacted with HER2 spatially (kept two HER2 molecules from interacting to signal), which allowed them to continually refine their antibody designs.
The ability to selectively recognize mutant HER2 represented one aspect to developing an effective cancer treatment. Antibodies also need to work with the immune system to kill cancer cells. A particular challenge is when cancer cells have only small numbers of mutant HER2 on their surfaces to which an antibody can attach.
To address this challenge the researchers converted their antibody into a bispecific T cell engager, a molecule in which the antibody targeting the mutant protein is fused to another antibody that binds to and activates immune cells called T cells. One end of the antibody sticks to the mutant HER2 on a cancer cell, while the other triggers T cells to kill the cancer cell. Further testing showed this method killed mutant HER2 cancer cells in dishes but spared normal ones. “Altogether, our results demonstrate that our antibodies in the bispecific T cell engager format achieved extremely selective and potent cytotoxicity toward HER2 S310F/Y,” the investigators reported.
When the researchers tested their T-cell engagers in mice with mutant HER2 tumors, they found that treatment—particularly using one construct, sTL18—significantly reduced tumor growth. It did so without causing weight loss or visible sickness in treated animals, which suggested the treatment had few side effects in the animals. Koide noted that because there are differences between mouse and human proteins, it is possible the lack of obvious side effects stemmed from the antibody binding even less to mouse wild-type HER2 than to the human version. Future studies will be required to investigate further.
Moving forward, Koide said the researchers will continue fine-tuning their antibody with the goal of developing a therapeutic. While the T cell engager molecule was the most potent construct they tried, he said, there could be better options they have not tested yet. In addition, the team plans to apply the antibody engineering technique to develop highly specific antibodies that may treat cancers caused by other mutant proteins.
In their paper, the investigators commented, “Overall, the developed antibodies show great promise as candidates for therapeutics as well as tools for advancing mechanistic studies of the HER family receptors … Although it is challenging to develop antibodies specific to a single point mutation within a large molecular-weight protein, the remarkable selectivity of the antibodies developed in this work offers an optimistic outlook on the strategy of developing antibodies highly selective to other disease-driver mutations in cell-surface proteins.”