A LC3-interacting motif in the influenza A virus M2 protein is required to subvert autophagy and maintain virion stability

Abstract

Autophagy recycles cellular components and defends cells against intracellular pathogens. While viruses must evade autophagocytic destruction, some viruses can also subvert autophagy for their own benefit. The ability of influenza A virus (IAV) to evade autophagy depends on the Matrix 2 (M2) ion-channel protein. We show that the cytoplasmic tail of IAV M2 interacts directly with the essential autophagy protein LC3 and promotes LC3 relocalization to the unexpected destination of the plasma membrane. LC3 binding is mediated by a highly conserved LC3-interacting region (LIR) in M2. The M2 LIR is required for LC3 redistribution to the plasma membrane in virus-infected cells. Mutations in M2 that abolish LC3 binding interfere with filamentous budding and reduce virion stability. IAV therefore subverts autophagy by mimicking a host short linear protein-protein interaction motif. This strategy may facilitate transmission of infection between organisms by enhancing the stability of viral progeny.

Graphical abstract

Figure 1

Relocalization of LC3 to the Plasma Membrane in Influenza-Infected Cells (A) A549 cells expressing GFP-LC3 were infected with PR8 strain IAV at an moi of 0.2, fixed at 16 hr p.i., and stained with anti-NP. Scale bar, 10 μm. (B) HCT116 cells expressing GFP-LC3 were infected as above and stained with WGA. Scale bar, 10 μm. (C) HCT116 cells expressing the indicated LC3 alleles fused to GFP were infected with IAV, fixed at 16 hr p.i., and stained with gold-labeled anti-GFP antibody. Black arrows, autophagosomes or autophagolysosomes; green arrows, labeling of plasma membrane. Scale bars, 500 nm. See also Figure S1.

Figure 2

Relocalization of LC3 to the Plasma Membrane Requires Influenza M2 (A) HCT116 cells expressing GFP-LC3 were infected with IAV, fixed at 16 hr p.i., and stained with anti-M2. Scale bar, 10 μm. (B) HCT116 cells expressing the indicated LC3 alleles fused to GFP were infected with WT or ΔM2 mutant IAV, fixed at 16 hr p.i., and stained with anti-NP. Scale bar, 10 μm. See also Figure S1.

Figure 3

A LIR Motif in Influenza M2 Is Required for LC3 Membrane Localization (A) M2 cytoplasmic tail (from amino acid 65 to C terminus) with predicted α helix and β strand. FVSI motif is underlined. (B) Comparison of M2 LIR motif with established LIRs. (C) Logo derived from 2,685 unique IAV M2 sequences. (D) LUMIER binding assay. Binding of luciferase-tagged cytoplasmic regions of the indicated IAV M2 variants expressed in 293ET cells to beads coated with purified GST or GST-LC3. (E) Fluorescence anisotropy of hydroxycoumarin-labeled M2 C terminus (DADDGHFVSIELE) against purified LC3. (F) Lysates of HCT116 cells expressing GFP-LC3 and infected with the indicated IAVs were immunoprecipitated with GFP-TRAP resin. Input and bound fractions were blotted with the indicated antibodies. Dotted line, irrelevant lanes removed. (G) HCT116 cells expressing GFP-LC3 infected with the indicated virus, fixed at 16 hr p.i., and stained with anti-HA. (H and I) Plasma membrane localization (H) or perinuclear accumulation (I) of GFP-LC3 in HCT116 cells expressing the indicated LC3 alleles and infected with the indicated viruses. Coverslips were assessed blindly in triplicate at 16 hr p.i. Mean and SD. ^∗∗p < 0.01 by ANOVA. (J) Lysates of HCT116 cells expressing GFP-LC3 were blotted for indicated antigens at 16 hr p.i. with indicated IAV. See also Figures S2 and S3.

Figure 4

M2 LIR Motif Required for Filamentous Budding and Virus Stability (A) MDCK cells were infected with the indicated PR8 MUd viruses at an moi of 5, fixed at 16 hr p.i., and visualized by scanning electron microscopy. Scale bar, 10 μm. (B) MDCK cells were infected with the indicated PR8 MUd viruses, fixed at 10 hr p.i., and stained without permeabilization with antiserum against PR8 virus. (Top) Maximum intensity z stack projection of confocal images. (Bottom) Side-on view of z stacks. Scale bar, 10 μm. (C) A549 cells infected with WT or mutant PR8 or PR8 MUd viruses at an moi of 1. Cell-free supernatants (16 hr p.i.) were harvested and frozen either immediately or after having been left at room temperature for 1 or 2 days. Experiments were performed in parallel. Titers were determined in triplicate after thawing by plaque assay; mean and standard deviation are shown. ^∗∗p < 0.01 by ANOVA. See also Figure S4.