Deep Mutational Scanning of H5 Influenza Hemagglutinin to Inform Influenza Virus Surveillance

University of Pennsylvania CEIRR (Penn-CEIRR) researchers Bernadeta Dadonaite, Jesse Bloom, Amy Baker, Scott E. Hensley, and Louise Moncla among others performed Deep Mutational Scanning of H5 Hemagglutinin to better assess how mutations might affect the virus’ risk to humans or effectiveness of current candidate vaccines. Learn more about their work here

Goal: This study uses pseudoviruses to measure how all mutations to the viral hemagglutinin (HA) protein of the H5 virus affect its ability to enter cells, use a2-6 sialic acids, increase HA stability, and reduce neutralization by polyclonal sera. 

Approach: A deep mutational scanning library was designed using the H5 HA from the A/American Wigeon/South Carolina/USDA-000345-001/2021 (H5N1) strain, which is one of the WHO’s clade candidate vaccine strains.  A lentivirus-based deep mutational scanning platform generated pseudoviruses encoding up to 10,773 mutants of HA linked to a nucleotide barcode. The effect of HA mutations on cell entry and preferred sialic acid receptor usage was quantified using 293T cells expressing a2-3 and a2-6 linked sialic acids. To identify mutations that increase HA stability, the pseudovirus libraries were incubated in progressively more acidic pH buffers to quantify their retention of infectivity. Polyclonal sera from mice or ferrets either vaccinated or previously infected with H5 clade was used to measure how all HA mutations are affected. 

Results: The effects of HA mutation on cell entry where highly variable, with some sites tolerating many mutations such as the head domain, and some using strong constraint including the stem helices and fusion machinery. Mutations that specifically enhanced entry into a2-6 versus a2-3 cells were at a handful of sites in the sialic-acid binding pocket including Q226L, which has previously been shown to enhance a2-6-linked sialic acid binding and occurred in prior H3 and H2 HA pandemic viruses. In addition, mutations at various sites were found to increase HA stability primarily concentrated in the stem’s alpha helices and the base of the head domain. The mutations that most strongly escaped serum neutralization were on the head of HA in previously defined antigenic regions. However, there was some variability in specific sites that caused the greatest mutations between the ferret and mouse sera.

Conclusion: This study provides the first comprehensive measurement of how HA mutations affect key viral replication factors contributing to the pandemic risk posed by H5. The research group also developed a website that makes it easier for investigators to look at the effects of all the H5 HA mutations on each property. These data could be very useful for assessing mutations to HA in newly observed strains that might affect risk to humans or effectiveness of current candidate vaccine strains. 


Reference: Dadonaite B. et al. (2024) Deep mutational scanning of H5 hemagglutinin to inform influenza virus surveillance. BioRxiv, DOI 10.1101/2024.05.23.595634. This article is a preprint and has not been certified by peer review. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.