Interferon-induced transmembrane protein 3 blocks fusion of sensitive but not resistant viruses by partitioning into virus-carrying endosomes.

Abraham L. Brass,Paul Meraner,Ruben M. Markosyan,Krishna C. Suddala,Sean P.J. Whelan,Fredric S. Cohen,Gregory B. Melikyan,Christine C. Lee,Tanay M. Desai,Mariana Marin

doi:10.1371/journal.ppat.1007532

Abstract

Late endosome-resident interferon-induced transmembrane protein 3 (IFITM3) inhibits fusion of diverse viruses, including Influenza A virus (IAV), by a poorly understood mechanism. Despite the broad antiviral activity of IFITM3, viruses like Lassa virus (LASV), are fully resistant to its inhibitory effects. It is currently unclear whether resistance arises from a highly efficient fusion machinery that is capable of overcoming IFITM3 restriction or the ability to enter from cellular sites devoid of this factor. Here, we constructed and validated a functional IFITM3 tagged with EGFP or other fluorescent proteins. This breakthrough allowed live cell imaging of virus co-trafficking and fusion with endosomal compartments in cells expressing fluorescent IFITM3. Three-color single virus and endosome tracking revealed that sensitive (IAV), but not resistant (LASV), viruses become trapped within IFITM3-positive endosomes where they underwent hemifusion but failed to release their content into the cytoplasm. IAV fusion with IFITM3-containing compartments could be rescued by amphotericin B treatment, which has been previously shown to antagonize the antiviral activity of this protein. By comparison, virtually all LASV particles trafficked and fused with endosomes lacking detectable levels of fluorescent IFITM3, implying that this virus escapes restriction by utilizing endocytic pathways that are distinct from the IAV entry pathways. The importance of virus uptake and transport pathways is further reinforced by the observation that LASV glycoprotein-mediated cell-cell fusion is inhibited by IFITM3 and other members of the IFITM family expressed in target cells. Together, our results strongly support a model according to which IFITM3 accumulation at the sites of virus fusion is a prerequisite for its antiviral activity and that this protein traps viral fusion at a hemifusion stage by preventing the formation of fusion pores. We conclude that the ability to utilize alternative endocytic pathways for entry confers IFITM3-resistance to otherwise sensitive viruses.

Highlights

Fusion of enveloped viruses with the host cell membrane is a key step leading to infection
The mechanism by which interferon-induced transmembrane proteins (IFITMs) interfere with the viral fusion step and the mechanism of virus escape from these restriction factors are poorly understood
Single virus and endosome tracking experiments demonstrate that the sensitive Influenza A virus is trapped within acidic interferon-induced transmembrane protein 3 (IFITM3)-positive endosomes that are not permissive for viral fusion

Summary

Introduction

Fusion of enveloped viruses with the host cell membrane is a key step leading to infection. The extensive conformational changes that ensue in the viral glycoproteins promote fusion between viral and cellular membranes [4,5,6]. Hemifusion is manifested as lipid mixing between viral and host membranes without viral content release, while full membrane fusion entails mixing of distinct aqueous contents delimited by the two membranes [4,5,6]. It has been demonstrated that sub-optimal conditions for membrane fusion including low density of viral glycoproteins, reduced temperature, and—where applicable—insufficiently acidic pH, favor dead-end hemifusion that does not progress to full fusion [9,10,11,12]. The progression to full viral fusion that culminates in the release of nucleocapsid into the cytoplasm is largely dependent on local conditions

Methods

Results

Discussion

Conclusion