Allosteric Regulation of the M2 Protein from Influenza A by Cholesterol

Main Text

Protein structure and function are modulated by the interactions with their environment (1). For membrane proteins, the environment is represented by both the bulk water and lipid membranes that have an active role in shaping their structural topology. In the 1970s, Farías et al. (2) hypothesized that lipid membranes may influence enzyme cooperativity through allosteric changes. These authors showed that membrane composition and fluidity deeply affect ATPase activity. Hill coefficients for these enzymes were also correlated with the changes in the lipid uptake caused by the dietary regime of the animal sources. Indeed, different membrane compositions are able to activate or deactivate large membrane protein complexes such as Ca²⁺-ATPase and phospholamban (3). These effects have also been studied in GPCRs, which have an intimate relationship with lipids and cholesterol (4), acting as allosteric modulators (5).

In this issue of the Biophysical Journal, Ekanayake et al. (6) bring into focus the influence that cholesterol binding has on viral budding for the M2 protein from influenza A. These authors used solid-state nuclear magnetic resonance spectroscopy to characterize this protein in lipid bilayers without the constraints imposed by crystallization. They found that cholesterol interacts with and stabilizes M2’s amphipathic helices adsorbed in the membrane surface. The latter forces the protein to adopt the pyramidal shape that is essential for inducing the membrane curvature and pinching off the virus particle (see Fig. 1). While the channel activity of M2 influenza is mimicked by a truncated version of M2 containing the tetrameric assembly of the transmembrane domains, channel regulation and other essential functions of this protein can be understood only when its full-length structure is assembled. As the membrane components affect the biological functions of a number of membrane proteins, the modulation of the channel activity by the lipid compositions is an expected outcome. What is completely unexpected is the specific location of the cholesterol molecules, which had been thought to be located in a cholesterol-recognition sequence domain within the amphipathic helix. Instead, the cholesterol binding pocket is reminiscent of the cholesterol consensus motifs found in G-protein coupled receptors. Here, cholesterol interacts with the amphipathic helices of M2 at the membrane interface to a palmitoylation site. During viral budding, the amphipathic helices of M2 contribute to the pronounced membrane curvature, conferring an overall pyramidal shape upon the protein. Cholesterol molecules bind and stabilize these amphiphatic interactions, which are known to play a key role in the budding process (7).

Figure 1.

Cholesterol (blue) association with influenza M2 (yellow). Cholesterol interacts with the M2 amphipathic helices at the membrane-water interface. During viral budding, the surface helices confer an overall pyramidal shape upon the protein and contribute to membrane curvature. To see this figure in color, go online.

Additionally, the authors found that amantadine binding disrupts the interactions between these amphipathic helices and adjacent transmembrane helices, weakening interprotomer interactions. Interestingly, cholesterol prevents this process.

Finally, amantadine binding and protonation have mutually exclusive effects on the tetrameric assembly of the channel. While small but significant structural perturbations are found at the N-terminus of the transmembrane domain upon His³⁷ protonation, these changes were not detected upon drug binding. Most significantly, drug binding but not protonation perturbs the interactions between the amphipathic helices as well as the tetrameric structure of M2.

The ever-emerging role of lipid membranes and their components in modulating protein function is fascinating and relatively unexplored. The compelling evidence for a role of cholesterol in the stabilization of the amphipathic helices of M2 protein from Influenza A suggests possible allosteric effects that solid-state NMR is uniquely positioned to probe.

Editor: Francesca Marassi.