Document resource
Background Personalized cancer immunotherapies have garnered much interest in recent years, owing to advancements in sequencing techniques required for accurate detection of individual cancer mutanomes. Presently, clinical successes have been achieved with the use of direct multi-neoepitope mRNA vaccines in combination with immune checkpoint inhibitors. However, direct mRNA vaccines are limited by their storage and physiological stability, as well as their lipid nanoparticle carriers, which may induce hypersensitivity depending on their formulation. DC vaccines offer a promising alternative due to their favorable safety profiles and potential to induce more robust and longer-lasting immunity. Furthermore, there are ongoing challenges in the selection and prioritization of therapeutic cancer neoepitopes, including the extended time periods and centralized equipment required to sequence and identify the mutanome, as well as suboptimal prediction algorithms that implement sequence-based methods with limited biochemical data for many diverse HLA alleles. Thus, we have sought to develop a faster tumor-to-prediction workflow to generate candidates for multiepitope RNA DC therapies.Methods We have evaluated the use of nanopore long-read cDNA-PCR sequencing of breast tumors for rapid neoepitope detection. We have also developed a custom prediction computational pipeline (NeoTarget) to prioritize the neoepitope repertoire across HLA class I alleles. This pipeline incorporates HLA-Inception, a deep convolutional neural network for neoepitope prediction based on molecular dynamic modeling of peptide-MHC electrostatic forces.Results Using both fresh and frozen resected breast tumors, we demonstrate detection of known chromosomal fusions and single nucleotide variants by nanopore cDNA-PCR sequencing, including unique mutations not detected by next-generation RNA-Seq (Illumina). We demonstrate successful generation of monocyte-derived DCs from healthy donors and ex vivo transfection with in vitro transcribed mRNA with greater than 95% efficiency.Conclusions Overall, this approach can be applied to rapidly identify and deliver neoepitopes into DCs for personalized neoantigen-based immune therapies in a scalable, GMP-compatible manner.Acknowledgements The author’s acknowledge resources and support from the Knowledge Enterprise Biosciences Genomics Core Facility and Research Computing Core Facility at Arizona State University. This work was partially supported by the Brenden Mann Foundation (FB). Supported in part by Mayo Clinic Breast Cancer SPORE grant P50 CA116201 from the National Institutes of Health.