New AI-Generated T-Cell Vaccine Showcases Long-Lasting Immunity Against COVID-19

Researchers have developed an AI-generated T-cell-based vaccine that demonstrates effectiveness against COVID-19 in mice, potentially offering long-lasting immunity against emerging variants. This approach, using the RAVEN platform, could also be used for developing improved vaccines for other seasonal viral diseases, such as influenza.

Artificial intelligence platform has the ability to efficiently design T-cell-based vaccines that offer broad protection. The platform could also be used to develop seasonal flu and other vaccines.

Researchers from Penn State have teamed up with Evaxion Biotech on a groundbreaking study that reveals the potential of an AI-generated vaccine in providing immunity against future COVID-19 variants. Unlike current COVID-19 vaccines, which target the spike protein of the SARS-CoV-2 virus and are prone to losing efficacy due to mutations, this vaccine focuses on triggering a T-cell response. This could result in a more durable immunity and serve as a model for other seasonal viral diseases, such as the flu. This study marks the first time an AI-generated vaccine has been tested and shown to be effective in a live viral challenge model.

The study conducted by Penn State and Evaxion Biotech researchers tested the effectiveness of a T-cell-based vaccine against SARS-CoV-2 in mice. The results showed a remarkable 87.5% survival rate among the vaccinated mice, compared to just 1 mouse from the control group. Furthermore, all the surviving mice that received the vaccine cleared the virus within two weeks after being challenged with a lethal dose of SARS-CoV-2.

The findings were recently published in the journal Frontiers in Immunology.

“To our knowledge, this study is the first to show in vivo [in a living organism] protection against severe COVID-19 by an AI-designed T-cell vaccine,” said Girish Kirimanjeswara, associate professor of veterinary and biomedical sciences, Penn State. “Our vaccine was extremely effective at preventing severe COVID-19 in mice, and it can be easily scaled up to start testing it in humans, as well. This research also paves the way for the potential rapid design of novel T-cell vaccines against emerging and seasonal viral diseases, like influenza.”

Why do we need a T-cell-based COVID-19 vaccine when the mRNA vaccines that are already in use are so effective?

Researchers used an artificial intelligence platform to create a T-cell-based COVID-19 vaccine that may last longer than the current vaccines. Credit: Girish Kirimanjeswara, Penn State

According to Kirimanjeswara, the spike protein of the SARS-CoV-2 virus is under heavy selection pressure, which can result in mutations that drive the emergence of new variants.

“This means that vaccine manufacturers will have to keep creating new vaccines that target new variants, and people have to keep getting these new vaccines,” he said.

Instead of targeting the constantly mutating spike protein, the team at Evaxion Biotech designed a vaccine that included 17 epitopes from various proteins of SARS-CoV-2 that are recognized by the immune system. These epitopes elicit an immune response from a broad selection of T cells, ensuring a sustained coverage of future variants.

“The virus would have to undergo too many mutations to be able to escape this T-cell-mediated immunity, so that is one advantage,” said Kirimanjeswara. “The second advantage is that T-cell-mediated immunity is usually long-lasting, so you don’t need repeated booster doses.”

If T cells are so great at remembering foreign agents, why were the first-generation COVID-19 vaccines designed to elicit responses from antibodies?

“It’s harder and takes longer to produce a T-cell-based vaccine than an antibody-based one,” said Kirimanjeswara. “Given the urgency with which we needed a vaccine to address the COVID-19 pandemic, it makes sense that vaccine manufacturers created an antibody-based vaccine. Now that the urgency has passed, a second-generation T-cell-based vaccine could be more effective and last longer.”

According to co-author Anders Bundgaard Sørensen, project director, Evaxion Biotech, other biotechnology companies are developing T-cell-based vaccines, but this team’s vaccine uses multiple types of artificial intelligence in a platform called RAVEN (Rapidly Adaptive Viral rEspoNse) to predict ideal targets for vaccines.

“RAVEN is really adaptable,” Sørensen said. “We don’t have to wait for a new strain of a virus to arrive to develop a vaccine. Instead, we can predict what will be needed in advance. That’s not something that others are doing right now.”

Sørensen noted, “It’s much easier to get broad coverage with a T-cell vaccine, as we can include multiple epitopes targeting different proteins.”

He added that, in addition to producing better COVID-19 vaccines, the RAVEN platform could be used to develop better influenza vaccines.

“Oftentimes, the influenza vaccines that are designed work only 30-40% of the time, so a lot of people end up getting sick,” he said. “As the world becomes increasingly integrated, that problem will become larger and larger. Our platform uses AI to better predict what will be needed.”

Sørensen noted that Evaxion benefitted from partnering with Kirimanjeswara and his Penn State colleagues because of their deep expertise in animal models of infectious disease and because the university houses a BSL-3 laboratory in which they could safely study the SARS-CoV-2 virus.

He said, “Our results are a testament to the power of industry-university partnerships.”

Reference: “DNA immunization with in silico predicted T-cell epitopes protects against lethal SARS-CoV-2 infection in K18-hACE2 mice” by Gry Persson, Katherine H. Restori, Julie Hincheli Emdrup, Sophie Schussek, Michael Schantz Klausen, McKayla J. Nicol, Bhuvana Katkere, Birgitte Rønø, Girish Kirimanjeswara and Anders Bundgaard Sørensen, 11 April 2023, Frontiers in Immunology.
DOI: 10.3389/fimmu.2023.1166546

Other Penn State authors on the paper include Katherine Restori, assistant research professor; McKayla Nicol, graduate student; and Bhuvana Katkere, assistant teaching professor. Other Evaxion Biotech authors on the paper include Gry Persson, project manager; Julie Hinchelli Emdrup, research associate; Sophie Schussek, competence manager; Michael Schantz Klausen, senior associate; and Birgitte Rønø, chief scientific officer.

Innovation Fund Denmark supported this research. The Huck Institutes of Life Sciences and the College of Agricultural Sciences provided support for studies performed at Penn State.

The Eva J. Pell Laboratory is an ABSL3 (animal biological safety level three) laboratory located on Penn State’s campus. It is a self-contained, standalone facility, which means that all materials, including waste products, are managed on site for increased safety. The lab is inspected by the National Institutes of Health and Centers for Disease Control and Prevention and is approved to conduct research on infectious agents.