Characterization of a trypsin-dependent avian influenza H5N1-pseudotyped HIV vector system for high throughput screening of inhibitory molecules

Abstract

In this study, we have generated and characterized an avian influenza H5N1 hemagglutinin (HA), neuraminidase (NA) and M2 ion channel pseudotyped HIV-based vector system (HaNaM-pseudotyped HIV vector). The cleavage site of the HA protein was modified to necessitate trypsin-dependent maturation of the glycoprotein. HA, NA and M2 were efficiently incorporated in HIV vector particles which could transduce different cell lines in a trypsin-dependent manner. Results also showed that the presence of avian influenza M2 and NA proteins maximized both vector production and transduction and that transduction was highly sensitive to the specific NA inhibitor oseltamivir (Tamiflu). H5N1 HaNaM-pseudotyped HIV vector system was also adapted for cell-based high throughput screening of drug candidates against influenza virus infection, and its high sensitivity to the specific oseltamivir validates its potential utility in the identification of new influenza inhibitors. Overall, the trypsin-dependent H5N1-pseudotyped HIV vector can mimic avian influenza virus infection processes with sufficient precision to allow for the identification of new antivirals and to study avian influenza virus biology in a lower biosafety level laboratory environment.

Fig. 1

Biochemical and functional analysis of H5N1-VLP harboring HA cleavage site modification. (A) Upper panel: modification of H5 HA cleavage. (B) 3 μg of H5N1 HA, NA and M2 protein expressors or VSV-G protein expressor were co-transfected with HIV vector containing GFP gene and HIV Gag-Pol expressing plasmid, as indicated in Section 2. At 48 h transfection, cells were labeled with [³⁵S]-methionine for 12 h. Purified H5N1-VLPs were treated (lane 2) or non-treated (lane1) with 4 μg of trypsin/ml for 1 h. Then, labeled VLPs were lysed with RIPA lysis buffer, immunoprecipitated by human anti-influenza serum (upper panel), rabbit anti-NA (middle panel) and mouse anti-P24 antibodies (lower panel) and analyzed by SDS-PAGE followed by autoradiography of gel. HA0: non-cleaved HA protein; HA1: HA surface subunit; HA2: HA transmembrane subunit; NA: neuraminidase; CA: HIV capsid p24 protein. (C) Dose-dependent trypsin activation of H5N1 HaNaM-pseudotyped VLP transduction. Equal amounts of purified HaNaM-pseudotyped VLPs were used to infect 293T cells in the absence or presence of various concentrations of trypsin, as indicated. The transduction efficiency (upper panel) was determined by measurement of GFP-positive cells with fluorescence-activated cell sorter (FACS) analysis at 48–72 h after infection. Intracellular GFP protein expression was detected by immunoprecipitation with anti-GFP antibody, followed by SDS-PAGE and anti-GFP western blot (lower panel). The results are representative for two independent experiments.

Fig. 2

Both NA and M2 proteins are required for a maximal HaNaM-pseudotyped VLP transduction. (A) The HIV vectors were co-transfected in 293T cells with H5N1 HA (3 μg) expressor alone, or co-transfected with different combinations of HA (3 μg), NA (3 μg) and M2 (3 μg) expressors, as indicated. At 48 h post-transfection, VLPs were collected from supernatants by ultracentrifugation and quantified by HIV-1 p24 detection ELISA. (B) Equal amounts of VLPs from panel A were used to transduce 293T cells, and at 48–72 h post-transduction, the percentage of GFP-positive cells was measured by FACS analysis. (C) The HA-, HA/NA- or HA/NA/M2-pseudotyped VLPs were produced in 293T cells that were co-transfected with HIV vectors, H5N1 HA(3 μg), NA (3 μg) and increasing amount of M2 expressing plasmids (0.5, 1, 3 μg). Then, equal amounts of VLPs were used to infect 293T cells and the expression of GFP protein in transduced cells was detected by immunoprecipitation with an anti-GFP antibody, followed SDS-PAGE and anti-GFP western blot. The results are representative for two independent experiments.

Fig. 3

The transduction efficiency of HaNaM-pseudotyped HIV VLPs in different cell types. (A) Equal amounts of HaNaM- or VSV-G-pseudotyped HIV VLPs (as adjusted by the level of HIV p24 antigen) were used to transduce human embryonic kidney 293T cells, canine kidney MDCK cells, A549 lung carcinoma cell and CD4+ C8166 T cells. The percentage of the GFP-positive cells in each cell population was measured by FACS analysis 48–72 h after transduction. (B) Fluorescence microscopic images of GFP-positive 293T (a) and A549 cells (b) 48 h post-transduction with HaNaM-pseudotyped VLPs. The results presented here were obtained from two independent experiments.

Fig. 4

NA inhibitor oseltamivir interferes with HaNaM-pseudotyped VLP production and its transduction. (A) HIV vectors were co-transfected with H5N1 HA, NA and M2 protein expressors or VSV-G protein expressor in 293T cells. After 12 h, the transfected cells were washed with serum-free medium and fresh DMEM added with or without 100 nM oseltamivir. At 48 h post-transfection, pseudotyped VLPs were pelleted by ultracentrifugation, and viral production was quantified by HIV-1 p24 antigen detection ELISA. (B) Equal amounts of HaNaM-pseudotyped VLPs from oseltamivir-treated or non-treated VLP producing cells were used to infect 293T cells and GFP-positive cells were quantified by FACS analysis. (C) Equal amounts of HaNaM- or VSV-G-pseudotyped VLPs produced in cells that were not treated with oseltamivir were incubated with 293T cells in the absence or presence of 100 nM oseltamivir for 2 h. The infection medium was replaced by fresh medium and incubated for 48 h. GFP-positive cells were quantified by FACS analysis. The results represent the mean ± standard deviations of three independent replicates. *, statistically significant (p < 0.05); **, very statistically significant (p < 0.01); ***, extremely statistically significant (p < 0.0001).

Fig. 5

Adaptation of HaNaM-pseudotyped vector transduction to a 96-well microplate assay. HIV vectors encoding for LacZ reporter gene were co-transfected with H5N1 HA/NA/M2 expressors or VSV-G expression plasmid in 293T cells. The supernatant containing HaHaM-pseudotyped VLPs were treated with trypsin (4 μg/ml) for 1 h at 37 °C before ultracentrifugation. (A) Two-fold serial dilutions of VLP stock were used to transduce 293T cells in 96-well plates. After 48 h, the β-Gal activity in cell lysates (upper panel) for each well was measured. Also, the β-Gal positive cells in the culture was stained by X-gal staining (lower panel). (B) Effect of oseltamivir on HaNaM-pseudotyped VLPs transduction. Equal amounts of HaNaM-pseudotyped VLPs from oseltamivir-treated or non-treated VLP producing cells were serially diluted and incubated with 293T cells cultured in 96-well plate. After 48 h, the cells were lysed and the β-Gal activity was quantified by with an enhanced β-galactosidase assay kit. The results are representative for three independent experiments.