%0 Generic
%A Klein, Steffen
%C Heidelberg
%D 2023
%F heidok:33896
%R 10.11588/heidok.00033896
%T In situ cryo-correlative light and electron tomography of influenza A virus entry and its inhibition by IFITM3
%U https://archiv.ub.uni-heidelberg.de/volltextserver/33896/
%X The innate immune system is the first wall of defense against many infectious pathogens, such as viruses or bacteria. The antiviral interferon-induced transmembrane protein 3 (IFITM3) is one of the key players against enveloped viruses like the influenza A virus. IFITM3 is localized in the endosomal-lysosomal system and is known to prevent viral cytoplasmic entry. Different hypotheses on the mode of action of IFITM3 were proposed, but the underlying molecular mechanism still needs to be fully understood. Here, I am using a combination of cryo-light microscopy and in situ cryo-electron tomography to study the antiviral function of IFITM3 within the natural cellular environment in the context of an influenza A virus infection. To visualize the antiviral actions of IFITM3, I established a novel cryo-correlative light and electron microscopy method. This novel approach allowed me to localize trapped influenza A virus particles in the endosomal-lysosomal system of an IFITM3-overexpressing human epithelial lung cell line A549, which allowed me to study them by cryo-electron tomography. Structural analysis of IFITM3-positive multivesicular bodies revealed that IFITM3 does not alter the ultrastructural morphology of the endosomal-lysosomal system and does not modulate the number of intraluminal vesicles (ILVs). These results contradict the ’fusion decoy hypothesis,’ which suggests that an increased number of ILVs in the late endosomal lumen could redirect viral membrane fusion from the limiting late endosomal membrane to fusion with ILVs. High-resolution in situ cryo-electron tomography of influenza A virus particles within late endosomes revealed that IFITM3 traps influenza A virus particles in a hemifusion state at the limiting late endosomal membrane and ILVs. These findings support the previously formulated ’hemifusion stabilization’ hypothesis as they are the first direct proof of IFITM3-mediated hemifusion stabilization within the natural cellular environment. Furthermore, ultrastructural characterization of the hemifusion sites revealed the post-fusion form of the viral fusion protein hemagglutinin (HA). Thus, IFITM3 does not inhibit low-pH triggered HA conformational changes, indicating that IFITM3 inhibits membrane fusion indirectly by modulating the membrane properties of the late endosomal-lysosomal system and thus stabilizing hemifusion.