Influenza Infects Lung Microvascular Endothelium Leading to Microvascular Leak: Role of Apoptosis and Claudin-5
(A) The reduction in claudin-5 levels is not due to decreased transcription. Levels of claudin-5 mRNA were assessed by quantitative PCR 24 hours after infection with replication-deficient influenza (MOI 4). Results are presented using the comparative CT method using 18S rRNA as a control. Results are from 3 experiments, p>0.05 for control versus infected. (B) Claudin-5 does not co-localize with lysosomes. Cells were infected with UV-irradiated influenza (MOI 8) for 16, 20, and 24 hours and were co-immunostained for LAMP1 and claudin-5. Images were from the 20-hour timepoint, but are reflective of all timepoints. Asterisks indicate cell nucleus, while arrows indicate cell membrane. Images are representative of 3 experiments. (C) Claudin-5 does not interact with poly-ubiquitin. The right-most blot shows the efficacy of the poly-ubiquitin antibody (FK1) since inhibition of the proteasome by MG-132 (MG) increases the expression of poly-ubiquitinated proteins in whole cell lysates. Cells were incubated with UV-influenza (MOI 8) for 24 hours. Whole cell lysates were immunoprecipitated for claudin-5 and blotted for poly-ubiquitin. IgH denotes the heavy chain of the IP antibody. Blots are representative of 3 experiments. (D) Inhibition of MMPs by marimastat blocks influenza-induced loss of claudin-5. Whole cell lysates were probed for claudin-5. Cells were treated with marimastat and infected with UV-flu (MOI 8) for 24 hours. Blot and quantitation are representative of 3 experiments. (E-F) Over-expression of claudin-5 blocks influenza-induced leak. Cells were transfected with either GFP or claudin-5 GFP. Arrow denotes membrane expression of claudin-5-GFP (A). Monolayers were then infected with UV-irradiated influenza (MOI 8) for 24 hours and the TEER was measured pre- and post-infection. Data are mean, SE for 3 experiments, *p<0.05.