“SNAF” Computational Tool Identifies Splicing Neoantigen Targets for Immunotherapy

A team of multidisciplinary researchers at Cincinnati Children’s and the University of Virginia has developed a computational tool dubbed “SNAF” (spliced neo antigen finder) that they suggest may help scientists bring the emerging promise of cancer immunotherapy to a wider range of patients. Led by Guangyuan Li, PhD, and Nathan Salomonis, PhD, both with the Division of Biomedical Informatics at Cincinnati Children’s, the researchers say the tool has already helped to uncover shared immunogenic targets across various cancers, which could pave the way for a new wave of highly focused cancer treatments.

“The implications of this discovery are significant,” suggested H. Leighton Grimes, PhD, director of the Cancer Pathology Program at Cincinnati Children’s. “By identifying shared splicing neoantigens present in up to 90% of cancer patients, SNAF not only presents new targets for therapy but also challenges and expands our understanding of cancer biology.”

Grimes is co-author of the team’s published paper in Science Translational Medicine, titled “Splicing neoantigen discovery with SNAF reveals shared targets for cancer immunotherapy.” In their report the team concluded, “Applied broadly to new cancers and distinct forms of malignancy, we believe that SNAF could be used to identify splicing neoantigens that are unique and shared across malignancies and discover new sequence motif preferences that expand the repertoire of targets for precision cancer therapy.”

Immunotherapy has emerged as  “a crucial strategy” for combating cancer, by “reprogramming” a patient’s own immune system, the authors wrote… This therapeutic strategy effectively harnesses the patient’s immune system to fight cancer, often targeting neoantigens produced from genetic mutations. However, the approach typically benefits only those with a high mutational burden. “Focused targeting of patients with selective mutations has produced promising results in cancers with a high mutational burden, such as melanoma, non–small cell lung cancer, and microsatellite instability (MSI)–high (MSI-H) colorectal cancer,” the investigators noted. But while immune checkpoint blockade (ICB) has become the frontline treatment for patients with high mutational burden, “… such therapies are not used in many cancers with low mutational burden, such as glioma and leukemia.”

Though typically attributed to tumor-associated mutations, neoantigens can also be produced from diverse posttranscriptional regulatory mechanisms, the authors pointed out. SNAF aims to expand the universe of immunotherapy by identifying neoantigens arising from such post-transcriptional modifications—and particularly splicing errors—which until now have remained largely untapped.

Using new artificial intelligence approaches, SNAF predicts immunogenic peptides that T cells can recognize and novel proteins with altered extracellular components that B cells can target. This dual approach is vital in developing comprehensive immunotherapies that engage both arms of the adaptive immune system. “SNAF is an easy-to-use computational tool to identify, prioritize, and interpret distinct classes of splicing neoantigens,” the team stated. “The workflow incorporates advanced deep learning and probabilistic algorithms to discover immunogenic splicing neoantigens (SNAF-T workflow), full-length protein coding transmembrane tumor-specific isoforms (SNAF-B workflow), and regulators of altered splicing (RNA-SPRINT).”

While cataloging all possible neoantigens produced from alternative mRNA pathways, the team found that in patients with melanoma, the amount of splicing neoantigens correlated with patient survival and responses to immunotherapy. “Shared splicing neoantigens were found in up to 90% of patients with melanoma, correlated to overall survival in multiple cancer cohorts, induced T cell reactivity, and were characterized by distinct cells of origin and amino acid preference,” the investigators wrote. “Patients with high splice neoantigen burden skew toward poor outcomes and associate with genes important to block immune tumor recruitment.” One such splicing neoantigen prediction, SLC45A2, has emerged as a particularly promising target due to its high tumor specificity and immunogenicity.

In addition to T-cell neoantigens, the team discovered a novel class of tumor-specific extracellular neo-epitopes, termed ExNeoEpitopes, through their B-cell focused pipeline, SNAF-B. These ExNeoEpitopes show great promise for the development of monoclonal antibodies and CAR-T cell therapies, they suggested. “Such antibodies could represent powerful new molecular reagents for CAR-T or monoclonal antibody strategies for shared and patient-specific neoantigens,” they wrote. “… the validation of ExNeoEpitopes will rely on new proteogenomics approaches that leverage targeted long isoform sequencing and proteomics along with antibodies that target specific conformational epitopes.”

Tamara Tilburgs, PhD, co-corresponding author and a researcher in the Division of Immunobiology, said, “This is only the beginning. The SNAF workflow’s flexibility means it can be continuously adapted as we make further inroads into understanding and combating cancer.” And while noting limitations of their study, the team nevertheless concluded “Our study provides evidence that aberrant splicing in melanoma frequently results in shared MHC-presented neoantigens, which can be confirmed in different patient cohorts and used to predict survival and response to immunotherapy…our systematic identification of splicing neoantigens revealed potential shared targets for therapy in heterogeneous cancers.”

Salomonis and team are now applying the tools in the most difficult-to-treat cancers to find the best targets for therapy development and understand their single-cell origins.