Erin M. Buckingham, Keith W. Jarosinski, Wallen Jackson, John E. Carpenter, Charles Grose

Varicella-zoster virus (VZV) is an extremely cell-associated herpesvirus, with limited egress of viral particles. The induction of autophagy in VZV-infected monolayers is easily detectable; inhibition of autophagy leads to decreased VZV glycoprotein biosynthesis and diminished viral titers. To explain how autophagic flux could exert a proviral effect on the VZV infectious cycle, we postulated that the VZV exocytosis pathway following secondary envelopment may converge with the autophagy pathway. This hypothesis depended on known similarities between VZV gE and autophagy-related (Atg) Atg9/Atg16L1 trafficking pathways. Investigations were carried out with highly purified fractions of VZ virions. When the virion fraction was tested for the presence of autophagy and endosomal proteins, microtubule-associated protein 1 light chain (MAP1LC3B) and Ras-like GTPase 11 (Rab11) were detected. By 2D and 3D imaging after immunolabeling, both proteins also colocalized with VZV gE in a proportion of cytoplasmic vesicles. When purified VZ virions were enumerated after immuno-electron microscopy, gold beads were detected on viruses following incubation with antibodies to VZV gE (∼100%), Rab11 (50%) and LC3B (30 %). Examination of numerous electron micrographs demonstrated that enveloped virions were housed in single-membraned vesicles; viral particles were not observed in autophagosomes. Taken together, our data suggested that some viral particles after secondary envelopment accumulated in a heterogeneous population of single-membraned vesicular compartments, which were decorated with components from both the endocytic pathway (Rab11) and the autophagy pathway (LC3B). The latter cytoplasmic viral vesicles resembled an amphisome. VZV infection leads to increased autophagic flux, while inhibition of autophagy leads to a marked reduction in virus spread. In this investigation of the proviral role of autophagy, we found evidence for an intersection of viral exocytosis and autophagy pathways. Specifically, both LC3-II and Rab11 proteins copurified with some infectious VZV particles. The results suggested that a subpopulation of VZV particles were carried to the cell surface in single-walled vesicles with attributes of an amphisome, an organelle formed from fusion of an endosome and an autophagosome. Our results also addressed interpretation of autophagy/xenophagy results with mutated herpes simplex virus lacking its ICP34.5 neurovirulence gene. The VZV genome lacks an ICP34.5 ortholog, yet we found no evidence of VZV particles housed in a double-membraned autophagosome. In other words, xenophagy, a degradative process documented after infection with HSV Δ34.5, was not observed in VZV infected cells.