X-ray studies show that influenza hemagglutinin (HA) forms an elongated structure connecting the influenza virus at one end to cell-surface receptors at the other. At neutral pH, the 20 N-terminal residues of HA2 - referred to as the fusion peptide - are buried in a hydrophobic pocket, about 100 Å away from the receptor-binding site, and thus seem unlikely to affect HA binding to the receptor. To test this assumption, we mutated residues in the fusion peptide, heterologically expressed the mutated proteins in COS7 cells, and examined their ability to bind fluorescently labeled red blood cells (RBCs). Surprisingly, a significantly reduced binding was recorded for some of the mutants. Ample experimental data indicate that HA has at least two forms: the native structure at neutral pH (N) that is metastable and the fusogenic form (F), observed at low pH, which is stable. Thus, a simple interpretation of our data is that HA can bind to its receptors at the RBC surface in the N form much more effectively than in the F (or in any other stable) form and that the altered binding properties are due to destabilizing effects of the mutations on the N form. That is, some of the mutations involve reduction in the free energy barrier between the N and F forms. This, in turn, leads to reduction in the population of the N form, which is the only form capable of binding to the cell-surface receptors. To explore this possibility, we estimated the stability free energy difference between HA wild-type (wt) and mutants in the N form using an empirical surface tension coefficient. The calculated stability differences correlated well with the measured binding, supporting the above interpretation. Our results are examined taking into account the available experimental data on the affinity of different soluble and membrane-attached forms of HA to its receptors.
Bibliographical noteFunding Information:
T.D. was a joint graduate student of Y.I.H. and Judith M. White; the mutations were carried out in White's lab. We thank Debora Fass, Michael H. Hecht, Amnon Horowitz, Gabriel Kaufmann, Nathan Nelson and Ezra Yagil for helpful discussions. The research was supported by fellowships from the Wolfson and Alon Foundations to N.B-T. We are grateful to the Bioinformatics Unit of the George S. Wise Faculty of Life Sciences at Tel Aviv University for providing computation facilities.
- Cell-surface receptor
- Continuum solvent model
- Fluorescent microscopy
- Fusion peptide
- Protein binding and stability