Two modified tumour proteins, or neoantigens, that possessed traits of possible vaccine candidates were found by researchers at the University of Arizona in a recent study that created a modelfor cutaneous squamous cell carcinoma, a form of skin cancer. Simultaneously, they used artificial intelligence to create 3D models to better understand and predict which neoantigens would activate T cells, a type of white blood cell vital to the immune system, to attack the cancer.
Tumour neoantigens are unique mutant proteins found in cancer cells. They act as a warning system, alerting the immune system to the threat that cancer cells pose. By identifying and characterising neoantigens, researchers can develop tailored tumour vaccines to help the immune system recognise and fight cancer cells.
The results suggest that identifying which neoantigens could be used in cancer vaccines to prevent tumours may depend heavily on their structural and physical properties.
According to senior author Dr. Karen Taraszka Hastings, a member of the U of A Cancer Centre and professor and chair of the Department of Dermatology at the University of Arizona College of Medicine in Phoenix, one of the challenges in creating tumor-based cancer vaccines is figuring out the best combination of neoantigens to elicit a T cell response that can eliminate a tumour. “We’re looking for ways to make it easier to choose which neoantigens to include in cancer vaccines, especially for cancers like melanoma and cutaneous squamous cell carcinoma that have a lot of mutations.”
Tumour vaccines can employ dozens of different peptides, or fragments of altered tumour proteins. Currently, there are experimental vaccines that target tumour mutations in people with melanoma, pancreatic cancer, and non-small cell lung cancer. However, it is challenging to identify which tumour peptide mutations will be most beneficial in a vaccine because there are thousands of them for certain diseases, like cutaneous squamous cell carcinoma.
The scientists found that both human and mouse cancers share many mutations, including the same important changes that are essential for the growth of tumours. Using a mouse model, they also identified two neoantigens that induced T cells to halt tumour growth.
Both neoantigens produced a strong anti-tumor response from T cells, whereas the peptides’ normal form did not. Upon closer examination, they found that although the immune system could see both neoantigens equally, their functions differed.
Before T cells can combat tumour neoantigens, the immune system must recognise them. The neoantigens are displayed by a group of proteins called the major histocompatibility complex, or MHC.
The team discovered that the MHC displayed the mutant Picalm peptide but not the normal peptide. This is likely the cause of the mutant Picalm peptide’s ability to trigger an anti-tumor T cell response. However, the mutant Kars peptide and the normal peptide exhibited similar mechanisms of binding to the MHC.
Hastings claims that “there is a different reason that mutated Kars elicits a T cell response that destroyed the tumour.”
To find out why, the researchers employed 3D, AI-based modelling to look for differences in the peptides’ structures and how they interacted with the MHC.
“We discovered that the mutated Kars peptide displays a distinct surface chemical structure on its 3D form that interacts with the T cell receptor,” Hastings explained. “This difference in peptide structure between mutated and normal Kars is probably recognised by the T cell receptor, which leads to a response that inhibits tumour growth in mice.”
Their analysis of all known cancer neoantigens that have been investigated independently for their ability to control tumour growth revealed that increased exposure of the modified peptide to the T cell receptor was essential.
“To further improve the selection of neoantigens for cancer vaccines, we are proposing the use of 3D structural modelling, which should enhance their efficacy,” she said.
Hastings claims that 3D modelling could be especially important for locating advantageous neoantigens for developing specialised vaccines against skin cancer and melanoma. It might be useful for other types of cancer as well. The group then plans to use patient tumour samples to test their hypotheses.
““The application of artificial intelligence to create a new method for developing personalized cancer vaccines highlights the transformative potential of this technology in cancer therapeutics,” said David Ebert. “Dr. Hastings’ research exemplifies the University of Arizona’s potential influence on the future of medicine and patient care.”
