Influenza A virus strains differ in sensitivity to the antiviral action of Mx-GTPase

Abstract

Interferon-mediated host responses are of great importance for controlling influenza A virus infections. It is well established that the interferon-induced Mx proteins possess powerful antiviral activities toward most influenza viruses. Here we analyzed a range of influenza A virus strains for their sensitivities to murine Mx1 and human MxA proteins and found remarkable differences. Virus strains of avian origin were highly sensitive to Mx1, whereas strains of human origin showed much weaker responses. Artificial reassortments of the viral components in a minireplicon system identified the viral nucleoprotein as the main target structure of Mx1. Interestingly, the recently reconstructed 1918 H1N1 "Spanish flu" virus was much less sensitive than the highly pathogenic avian H5N1 strain A/Vietnam/1203/04 when tested in a minireplicon system. Importantly, the human 1918 virus-based minireplicon system was almost insensitive to inhibition by human MxA, whereas the avian influenza A virus H5N1-derived system was well controlled by MxA. These findings suggest that Mx proteins provide a formidable hurdle that hinders influenza A viruses of avian origin from crossing the species barrier to humans. They further imply that the observed insensitivity of the 1918 virus-based replicon to the antiviral activity of human MxA is a hitherto unrecognized characteristic of the "Spanish flu" virus that may contribute to the high virulence of this unusual pandemic strain.

FIG. 1.

Effect of Mx1 on viral plaque formation. Vero-control cells and Vero-Mx1 cells were infected with 200 PFU of either influenza A virus strain FPV-B (upper wells) or A/Udorn/72 (lower wells) and were subsequently overlaid with medium containing agar. At 3 days p.i., cells were fixed and plaques were stained with crystal violet. Arrows mark plaques that have been formed on Mx1-expressing cells.

FIG. 2.

Effect of Mx1 on viral protein synthesis. Swiss 3T3 control cells and 3T3 Mx1-expressing cells were infected with 2 PFU per cell of different influenza A virus strains. At 6 h p.i., cells were lysed and proteins were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. (A) The viral NP was detected by Western blotting using a polyclonal rabbit serum. Detection of β-tubulin served as a loading control. (B) Expression of Mx1 in IFN-α-treated MEF and virus-infected 3T3-Mx1 cells. MEF were treated for 20 h with 1,000 U/ml of IFN-α B/D. 3T3-Mx1 cells were infected as described for panel A. Mx1 expression was monitored using a mouse monoclonal antibody. Detection of β-tubulin reveals loading of equal amounts of cell lysates. (C) Localization of Mx1 in infected cells. 3T3-Mx1 cells were either infected with 2 PFU per cell of A/Texas/36/91 or mock treated. At 8 h p.i., cells were fixed, permeabilized, and stained for Mx1 and viral NP using specific antibodies and Cy2 or Cy3 fluorophore-conjugated secondary antibodies, respectively.

FIG. 3.

Effect of Mx1 on viral polymerase activity. 293T cells were transfected with the following amounts of expression plasmids encoding PA, PB1, and PB2: 90 ng (A and B) or 5 ng (C) for A/WSN/33, 30 ng for A/Texas/36/91 (A and B), 5 ng for A/Turkey/England/91 (C), and 25 ng for SC35 (C). In addition, 100 ng of pRL-SV40-Rluc, 50 ng of pPOLI-Luc-RT, and 300 ng of the respective NP-encoding plasmid for strains A/WSN/33 and A/Texas/36/91 (A and B) or 500 ng of pPOLI-SP-Luc-RT and 500 ng of the respective NP-encoding plasmid for A/WSN/33, A/Turkey/England/91, and SC35 (C) were cotransfected. These amounts of expression plasmids had been determined before to lead to 50% of the maximum activity for each system. (A) To investigate the inhibitory potential of Mx1, increasing amounts of the Mx1-encoding plasmid were cotransfected. At 24 h after transfection, cells were lysed and luciferase activities were measured. Cotransfection of the inactive mutant Mx1(K49A) was used as a positive control, and activity in the presence of Mx1(K49A) was set to 100%. (B and C) Amount of Mx1 expression plasmid necessary for 50% inhibition of the different minireplicon systems, determined as described for panel A. Error bars, standard deviations calculated from three experiments.

FIG. 4.

Role of the viral NP in Mx1 sensitivity. 293T cells were transfected with the respective amounts of plasmids for minireplicon assays leading to 50% of the maximum activity of each system. The specific amount of Mx1-encoding plasmid causing 50% inhibition of luciferase activity or the same amount of a plasmid encoding the antivirally inactive mutant Mx1(K49A) was cotransfected: 37.5 ng for A/WSN/33 (A), 375 ng for A/Texas/36/91 (B), and 6 ng for SC35 (C). For each system, the NP-encoding plasmids of the three virus strains were cotransfected with the respective polymerase-encoding plasmids. Luciferase activity was determined 24 h after transfection. The values for the inactive mutant were set to 100%, and the values for Mx1 are expressed as percentages. Error bars, standard deviations calculated from three experiments. Abbreviations: wt, wild-type Mx1; K49A, mutant Mx1(K49A); WSN, A/WSN/33; Tx, A/Texas/36/91.

FIG. 5.

Inhibition of the polymerase activities of the highly virulent strains A/BM/1/18 and A/Vietnam/1203/04 by mouse Mx1 and human MxA. 293T cells were transfected with expression plasmids coding for the polymerase subunits (25 ng of each plasmid for A/BM/1/18 or 150 ng for A/Vietnam/1203/04), 300 ng of the plasmid encoding the NP of A/BM/1/18 or 500 ng of the NP-encoding plasmid for A/Vietnam/1203/04, 50 ng of pPOLI-Luc-RT, and 100 ng of pRL-SV40-Rluc. (A) Increasing amounts of pcDNA3-Mx1 were cotransfected. Twenty-four hours posttransfection, cells were lysed and luciferase activities were measured. The luciferase activities obtained by cotransfection of the inactive mutant Mx1(K49A) were set to 100%. (B and C) Titration of the amount of Mx1-encoding plasmid (B) or MxA-encoding plasmid (C) needed for 50% inhibition of the polymerase activities of strains A/BM/1/18 (BM) and A/Vietnam/1203/04 (VN). Error bars, standard deviations calculated from three experiments.