Next-generation sequencing-based neutralization assays to forecast influenza virus clade growth.

University Of Washington, Seattle WA

Investigators

Abstract

Project Summary/Abstract Every year, seasonal influenza viruses emerge with hemagglutinin (HA) surface protein mutations that confer escape from previously-protective neutralizing antibodies. This process is known as antigenic drift, and it necessitates frequent vaccine updates. Historically, the viral strains with the most antigenically drifted HA variants tend to dominate a given influenza season. Previous work has taken advantage of this observation by training forecasting models on antigenic data (e.g., titers from neutralization or hemagglutination-inhibition assays) in addition to genomic and epidemiologic data. However, these assays are low in throughput and scope, and generally only completed for a limited number of strains. Nevertheless, incorporating antigenic measurements for even partial circulating variant diversity in forecasting models has improved their accuracy. Development of methods that could generate more comprehensive datasets – that is, measuring the neutralizing titers of all circulating HA diversity – is therefore an a ractive goal. Here, I propose the development and an application of a next-generation sequencing-based neutralization assay that would parallelize antigenic measurements of ~100 currently circulating pdmH1N1 and H3N2 influenza A viruses in high-throughput. In this approach, influenza viral variants are selected, barcoded and pooled to create a multiplexed virus library. This library will then be used to simultaneously measure neutralizing titers for all viruses against a given serum specimen. Current approaches frequently rely on ferret sera, which do not fully represent the complexities of epitope targeting exhibited by human sera. I propose using this method to profile serum from several different human cohorts. The proposed cohorts were selected to more adequately represent the heterogenous immune responses comprising population immunity as a whole. Using this dataset, I will then test the hypothesis that more comprehensive antigenic measurements will improve the ability to forecast influenza variant success. I will assess the predictive power of these measurements both on their own as well as incorporated into previously established fitness models. Overall, I will develop a method to parallelize neutralization assays with ~100 currently circulating influenza A viruses. I will then use computational methods to ask if these measurements improve our ability to predict influenza strain dominance for a given season. Importantly, a major goal of vaccine strain updates is that the vaccine strain matches the seasonally dominant variant(s). The approach set forth in this proposal therefore has the potential to improve our ability to select vaccine strains.