A Universal Flu Vaccine Based on Conformationally Locked Soluble Headless HA

Period of Performance: 02/15/2015 - 01/31/2016


Phase 1 SBIR

Recipient Firm

Avatar Medical, LLC
Brooklyn, NY 11220
Principal Investigator


DESCRIPTION (provided by applicant): An influenza pandemic remains an acute threat to world health;and stockpiling a universally protective influenza vaccine provides a strong defense against this potential catastrophe. The hemagglutinin (HA) protein is the primary target of humoral Ab responses to Influenza. Vaccines based on full-length HA give rise predominantly to strain-specific Ab responses to the immunodominant, and highly variable head domain, and lack breadth of protection. However, the majority of the broadly neutralizing antibodies (bnAbs) that have been isolated to date target HA's highly conserved stalk domain, where they recognize trimer-specific quaternary neutralizing epitopes (QNEs). An HA immunogen from which the head domain has been removed is expected to focus immune responses on these conserved stalk QNEs, resulting in broadly protective Ab responses. But current "Headless HA" constructs do not fold correctly or display only part of the stalk's key QNE(s), and thus have only shown limited heterologous protection. Avatar has developed a method of locking protein immunogens in their native, quaternary conformation, in order to better present broadly protective QNEs. Targeted dityrosine (DT) crosslinks are engineered into fully folded, native proteins by (i) introducing conservative Tyr substitutions, and then (ii) catalyzing DT bond formation, in order to conformationally lock the proteins. DT bonds are catalyzed enzymatically, and only form between Tyr side-chains in very close structural proximity and - unlike engineered disulfide bonds - are introduced after the protein has fully folded. DT bonds are safe and irreversible, and because the bonds are zero-length, this approach fully preserves protein structure, and furthermore avoids aggregation because DT bonds do not form spontaneously. We will engineer DT bonds in the stalk of the intact HA trimer to lock it in its native conformatio. This will allow us to subsequently remove the head by proteolysis, while maintaining the stalk's native, trimeric structure, generating a conformationally locked "DT-Headless HA". To this end, we will introduce both Tyr substitutions at targeted positions into the stalk, and proteolytic cleavage sites into the head domain of the same HA molecule. Following crosslinking of the stalk, and proteolytic removal of the head, we will analyze the DT-locked stalk trimer antigenically, by comparison to uncrosslinked and wild-type controls (AIM 1). Then we will perform lethal viral challenge studies in mice, to confirm that conformationally locked DT-Headless HA elicits robust homologous protection, and improved heterospecific neutralization responses, compared to full- length WT HA (AIM 2) By focusing Ab responses on the conserved stalk QNEs, and away from the immunodominant head of HA, the DT-Headless HA immunogen is expected to give rise to broadly neutralizing Abs that protect from homologous and drift variants, as well as group 1 heterologous, and perhaps even group 2 heterologous challenges.