SURVEILLANCE AND PREDICTION OF SWINE INFLUENZA A VIRUS HEMAGGLUTININ GENE SEQUENCES FOR THE RAPID DEVELOPMENT OF SARNA-NANOPARTICLE VACCINES

<b>Project Methods</b><br> The overall objective of this multi-phase SBIR project is to design, develop, validate, and commercialize a system for more-effective swine IAV control that can be utilized by veterinarians/producers.Collection of nasal swab samples from Farrow to Wean Site 1 facilities and 4- month-old pigs. PI Lance Daharsh and co-investigator Hank Harris will coordinate sample collections with the two production facilities. During each collection, the production facility will collect 50 nasal swabs from Farrow to Wean Site 1 Facilities and 50 nasal swabs from 4-month- old pigs. Each facility will sample monthly for 6 months (July-Dec). All swabs will be labeled (location/date). Genvax will provide the BBL culture swabs and will arrange for shipping of the collected samples.Testing of collected nasal swab samples for swine IAV and sequencing positive samples. PI Daharsh will coordinate the shipment and storage of collected nasal swab samples. The samples will be submitted to the Iowa State Veterinary Diagnostic Lab (VDL). All samples will be screened for Influenza A by PCR. We assume there will be a positivity rate of 5% among samples. All PCR-positive samples will undergo full-length sequencing of the HA gene. The Iowa State VDL will generate and send results to Genvax via their online Client Portal. All newly generated HA sequencing data will be collected for sequential deposition into the Genvax Sequencing Database.Design and implementation of the Genvax Sequence Database. PI Daharsh, co- investigator Joel Harris, and Consultant Joe Webb will coordinate the design/implementation of the Genvax Sequence Database. Daharsh, Harris, and Webb will design a cloud computing database structure and implement it using Amazon AWS based tools. It will include associated services for database infrastructure (AWS DynamoDB), data storage (AWS S3), ML (AWS Sagemaker), and web app/website hosting (AWS API Gateway, Cloudfront, Route53).Aggregation and deposition of new sequence data and associated metadata into the Genvax Sequence Database. PI Daharsh and Lab Technician Naseer will design database protocols for adding sequences and metadata. Full-length swine IAV HA gene sequences and corresponding metadata from the past 10 years will be added to the database from online repositories, along with newly generated sequences from the Iowa State VDL Client Portal.Data analysis of gene sequences. PI Daharsh, Manager of Nanovaccines Dennis O'Neill, and Technician Naseer will perform data analysis on the sequences in the Genvax Sequence Database. We will summarize the prevalence/relatedness of swine IAV strains within Farrow to Wean Sites and 4-month-old pigs. This data will help to inform producers about the prevalence and possible spread of strains among sites. It will also provide clues about viral evolution and how often new strains are being introduced. Further, the Genvax team will conduct phylogenetic analyses on the collected sequences to see how they compare with historically collected data, including recently collected data from their surrounding region.Feature selection and ML predictions of viral evolution. The livestock sector needs a better way to manage sequencing data and potential genes of interest to aid in disease management. To provide added value to veterinarians/producers, the Genvax Sequence Database will be more than just a sequence repository. PI Daharsh and Consultant Webb will use the HA sequence and metadata information in the database to create predictive models of swine IAV evolution. The goal is to aid vaccine design by creating methods to help predict future circulating swine IAV strains within a specific producer facility or region. They will first evaluate the feature set from metadata using the Boruta algorithm to rank feature importance. Additional novel features may be created out of the existing metadata and evaluated for importance. At least 4 ML models will be tested using the highest-ranked features as input, including XGboost, random forest, neural networks, and rough set techniques. Additional ML models and ensemble approaches will also be tested. We will evaluate the feature set importance of each using SHAP values. The models will be implemented/evaluated using K-fold cross-validation and backtesting using historical data. Results will be presented as a confusion matrix and evaluated using traditional ML assessment methods, including F1 score and area under receiver-operating characteristic (AUROC).Production of the saRNA vaccine vector and insertion of the designed gene of interest. Manager O'Neill will coordinate production of the saRNA vaccine vector. Plasmid DNA is expanded in E. coli, these competent cells used for cloning are a DH5α derivative purchased from NEB. Antibiotic selection of the transformed cells with plasmid DNA is conducted with agar plates and LB broth containing carbenicillin. Plasmid DNA is extracted from the transformed cells using the endotoxin-free kit, ZymoPURE Plasmid Miniprep Kit. The restriction digest is purified with a column purification step to remove any residual E coli cells, so no living organism is present after this step in the saRNA production process. The carbenicillin selection gene is not part of the saRNA generated using the Promega RiboMAX Large Scale RNA Production System via the T7 promoter. The full-length saRNA is then 5'- Capped using the Vaccinia Capping Enzyme and mRNA Cap 2'-O-Methyltransferase with a One-Step capping Methyltransferase process.Design a 100% strain-matched gene of interest derived from a production facility. HA gene sequences from a production facility will be analyzed for predicted prevalence/ representation of circulating strains. PI Daharsh and co-investigator Hank Harris will choose the most prevalent and/or representative sequence for inclusion. Manager O'Neill will perform gene of interest integration with the vaccine vector by first conducting codon-optimization using the Integrated DNA Technologies (IDT) Codon Optimization Tool for expression within the Sus Scrofa pig genome. The optimized gene sequence will be ordered from IDT as a HiFidelity gBlock, which will be sequence-verified from the vendor. A seamless cloning strategy will be used for cloning all genes into the vector.In vivo vaccine trial in 3-week-old pigs. An in vivo study in pigs will be conducted at VRI-AMCV animal research facilities. The Genvax saRNA vaccine with the 100% strain- matched HA gene sequence and nanoparticle-delivery system will be injected intramuscularly into the ventral portion of the biceps femoris muscle. We will follow a prime boost vaccine strategy with a Day 1 prime injection followed by a boost injection 4 weeks later (Day 28). Each injection will use a 25-ug dose in 2 ml of volume. Animal well-being will be monitored throughout the study along with daily injection site monitoring. At minimum, a total of 5 animals will be injected with the vaccine along with a matched control animal group. Control animals will be injected with a similar volume of saline. Blood draws to collect serum samples will be performed on days 1, 14, 28, 42, and 56. All animals will be euthanized at the conclusion.Evaluation of vaccine efficacy. Serum samples collected during the in vivo vaccine trial will be submitted to the Iowa State VDL for HI testing. The assay will provide data on predicted protection provided by the vaccine against a homologous swine IAV strain. Titers >= 1:40 are considered to be protective. HI testing results from the VDL will be obtained via a client web portal by Genvax for further analysis and efficacy evaluation.Analysis of data from Objective 1 and 2, and Final Report. PI Daharsh will work with the other team members to assess Phase I results in terms of meeting or exceeding the feasibility metrics. A final report will be prepared that summarizes these results and provides recommendations for a follow-on Phase II project.