Research Project
10211103
ABSTRACT Influenza virus (flu) ranks highest in disease burden of all infectious diseases as measured in disability-adjusted life years. Seasonal epidemics cause 200,000-500,000 worldwide deaths annually. The total economic burden of seasonal flu is estimated to range from approximately $26B to $87B each year in the US in terms of direct medical expenses and lost work and productivity. Additionally, at least six known flu pandemics have become global human catastrophes, most notably the Spanish Flu pandemic of 1918, which killed 3-5% of the world?s population. Any reduction in the infection rate, transmission, and severity of flu infection would greatly reduce our healthcare expenditures and improve the quality of life for millions of people every year. The current vaccines are formulated annually based on predictions of which circulating flu strains may be prevalent in a given season. The effectiveness of these vaccines varies from year to year based on the circulation of unexpected antigenic variants and other factors. Vaccine design is complicated the by the multiplicity of flu strains, each with rapidly-evolving dominant antigen epitopes (?decoy? epitopes) that largely stimulate strain- restricted immunity. One strategy for rational antigen design, termed Immune Refocusing Technology (IRT), involves introducing mutations that reduce the immunogenicity of these decoy epitopes thus shifting the immune response to target more widely-conserved subdominant epitopes. BMI has previously applied this IRT approach with some notable successes to other viral antigens (e.g. HRV and the RSV F protein), and we now focus on the major flu surface antigen glycoprotein HA using H1, H3, and B vaccine strains as parental antigens. The anticipated effort to design a suitably modified antigen would ordinarily involve a protracted process of trial-and-error testing of many potential candidates. However, we have recently developed the ANATOPE automated B cell epitope prediction software package with algorithm parameters tuned using methods in artificial intelligence. Our algorithm identifies epitopes with a significantly higher success rate than previously available prediction programs. This breakthrough allows us to assign immunogenicity ?strength? scores to particular antigen surface patches and will further guide and accelerate the design of mutant antigens that refocus the immune response to cross-strain conserved epitopes. In this application, we propose to engineer and test the immunogenicity of rationally-designed HA antigens containing mutations that both 1) dampen the immunogenicity of dominant strain-restricted decoy epitopes and 2) enhance the immunogenicity of conserved subdominant epitopes associated with broadly neutralizing antibodies. Follow- up studies will assess the rationally-designed antigens in a ferret challenge study and prepare the approach for translation into humans as a universal vaccine that does not require annual reformulation.
