Similar to other positive-strand RNA viruses, rhinovirus, the causative agent of the common cold, replicates on a web of cytoplasmic membranes, orchestrated by host proteins and lipids. The host pathways that facilitate the formation and function of the replication membranes and complexes are poorly understood. We show that rhinovirus replication depends on host factors driving phosphatidylinositol 4-phosphate (PI4P)-cholesterol counter-currents at viral replication membranes. Depending on the virus type, replication required phosphatidylinositol 4-kinase class 3beta (PI4K3b), cholesteryl-esterase hormone-sensitive lipase (HSL) or oxysterol-binding protein (OSBP)-like 1, 2, 5, 9, or 11 associated with lipid droplets, endosomes, or Golgi. Replication invariably required OSBP1, which shuttles cholesterol and PI4P between ER and Golgi at membrane contact sites. Infection also required ER-associated PI4P phosphatase Sac1 and phosphatidylinositol (PI) transfer protein beta (PITPb) shunting PI between ER-Golgi. These data support a PI4P-cholesterol counter-flux model for rhinovirus replication.

Apoptosis and programmed necrosis (necroptosis) determine cell fate, and antagonize infection. Execution of these complementary death pathways involves the formation of receptor-interacting protein kinase 1 (RIPK1) containing complexes. RIPK1 binds to adaptor proteins, such as TRIF (Toll-IL-1 receptor-domain-containing-adaptor-inducing interferon-beta factor), FADD (Fas-associated-protein with death domain), NEMO (NF-κB regulatory subunit IKKγ), SQSTM1 (sequestosome 1/p62), or RIPK3 (receptor-interacting protein kinase 3), which are involved in RNA sensing, NF-κB signaling, autophagosome formation, apoptosis, and necroptosis. We report that a range of rhinoviruses impair apoptosis and necroptosis in epithelial cells late in infection. Unlike the double-strand (ds) RNA mimetic poly I:C (polyinosinic:polycytidylic acid), the exposure of dsRNA to toll-like receptor 3 (TLR3) in rhinovirus-infected cells did not lead to apoptosis execution. Accordingly, necroptosis and the production of ROS (reactive oxygen species) were not observed late in infection, when RIPK3 was absent. Instead, a virus-induced alternative necrotic cell death pathway proceeded, which led to membrane rupture, indicated by propidium iodide staining. The impairment of dsRNA-induced apoptosis late in infection was controlled by the viral 3C-protease (3Cpro), which disrupted RIPK1-TRIF/FADD /SQSTM1 immune-complexes. 3Cpro and 3C precursors were found to coimmuno-precipitate with RIPK1, cleaving the RIPK1 death-domain, and generating N-terminal RIPK1 fragments. The depletion of RIPK1 or chemical inhibition of its kinase at the N-terminus did not interfere with virus progeny formation or cell fate. The data show that rhinoviruses suppress apoptosis and necroptosis, and release progeny by an alternative cell death pathway, which is controlled by viral proteases modifying innate immune complexes.

BackgroundPicornaviruses are common human and animal pathogens, including polio and rhinoviruses of the enterovirus family, and hepatits A or food-and-mouth disease viruses. There are no effective countermeasures against the vast majority of picornaviruses, with the exception of polio and hepatitis A vaccines. Human rhinoviruses (HRV) are the most prevalent picornaviruses comprising more than one hundred serotypes. The existing and also emerging HRVs pose severe health risks for patients with asthma or chronic obstructive pulmonary disease. Here, we developed a serotype-independent infection assay using a commercially available mouse monoclonal antibody (mabJ2) detecting double-strand RNA.ResultsImmunocytochemical staining for RNA replication centers using mabJ2 identified cells that were infected with either HRV1A, 2, 14, 16, 37 or coxsackievirus (CV) B3, B4 or A21. MabJ2 labeled-cells were immunocytochemically positive for newly synthesized viral capsid proteins from HRV1A, 14, 16, 37 or CVB3, 4. We optimized the procedure for detection of virus replication in settings for high content screening with automated fluorescence microscopy and single cell analysis. Our data show that the infection signal was dependent on multiplicity, time and temperature of infection, and the mabJ2-positive cell numbers correlated with viral titres determined in single step growth curves. The mabJ2 infection assay was adapted to determine the efficacy of anti-viral compounds and small interfering RNAs (siRNAs) blocking enterovirus infections.ConclusionsWe report a broadly applicable, rapid protocol to measure infection of cultured cells with enteroviruses at single cell resolution. This assay can be applied to a wide range of plus-sense RNA viruses, and hence allows comparative studies of viral infection biology without dedicated reagents or procedures. This protocol also allows to directly compare results from small compound or siRNA infection screens for different serotypes without the risk of assay specific artifacts.