IFITM3 Clusters on Virus Containing Endosomes and Lysosomes Early in the Influenza A Infection of Human Airway Epithelial Cells

Kummer, Susann; Avinoam, Ori; Kräusslich, Hans-Georg

The analysis of uninfected and HK/1/68-infected A549 cells at different time points revealed a very low proportion of large clusters in uninfected cells or up to 2 h p.i. At 3 h p.i. and later, we observed a significant increase in cluster abundance ( Figure 3B Although A549 cells are an established model cell line for IAV research, adaptation to cell-culture conditions may have occurred; hence, we verified our results in primary human respiratory epithelial cells. To determine whether IAV infection also causes IFITM3 clustering in primary cells, we infected human small airway epithelial cells (HSAEpCs) with IAV PR/8/34 for different periods of time and performed an indirect immunofluorescence analysis for IFITM3 and NP (from incoming IAV particles) using confocal and STED microscopy ( Figure 4 , uninfected example see Figure S5 ). The IAV infected HSAEpCs revealed a co-localization of IFITM3 and IAV NP signals at apparently vesicular structures as early as 1 h p.i. (Figure 4A , white arrowheads). At later time points (3 h p.i.), this was more obvious, with a clear IFITM3 clustering on the NP-containing vesicular structures ( Figure 4B ). IFITM3 often exhibited a ring-like appearance, suggesting the coating of endosomal vesicles, and this phenotype became more obvious at later time points (6 h p.i; Figure 4C ). Some IFITM3-positive vesicles exhibited strong NP signals (e.g., Figure 4B ), suggesting IFITM3-coated vesicles carrying multiple IAV particles. A strong IFITM3 clustering with a ring-like appearance indicating vesicle coating was observed in both IAV-infected A549 cells ( Figure 5A ) and HSAEpCs at 10 h p.i. (Figure 5C ; white arrowheads indicating ring-like structures). In the case of the A549 cells, the IFITM3 signals in uninfected cells were weak and mostly diffuse. Upon IAV infection, > 90% of cells lacking a strong nuclear NP signal displayed IFITM3 clustering (e.g., the right cell in Figure 5A ), while this was only observed in < 10% of cells with a strong nuclear NP ( Figure 3B ). Given the high infection rate in this cell line and the fact that cytoplasmic NP was also observed in cells lacking the bright nuclear NP of replicating IAV ( Figure 1A) , we suggest that early IFITM3 clustering at vesicular structures correlates with an abortive infection in this cell type, and that only cells that do not induce the vesicular clustering of IFITM3 become infected. In general, the IFITM3 signals were stronger in HSAEpCs compared to A549 cells ( Figure 5A,C) , and IFITM3 clusters apparently coating cytoplasmic vesicles were also present in productively IAV-infected cells exhibiting a strong nuclear NP signal in this case ( Figure 5D ). No IFITM3 clustering at vesicular structures was observed in HSAEpCs in the absence of IAV infection, while the overall IFITM3 signal intensity was much higher in these primary cells compared to A549 cells without infection ( Figure S5 ). A strong IFITM3 clustering with a ring-like appearance indicating vesicle coating was observed in both IAV-infected A549 cells ( Figure 5A ) and HSAEpCs at 10 h p.i. (Figure 5C ; white arrowheads indicating ring-like structures). In the case of the A549 cells, the IFITM3 signals in uninfected cells were weak and mostly diffuse. Upon IAV infection, > 90% of cells lacking a strong nuclear NP signal displayed IFITM3 clustering (e.g., the right cell in Figure 5A ), while this was only observed in < 10% of cells with a strong nuclear NP ( Figure 3B ). Given the high inf