Fig 1.
Baloxavir inhibits PA-X shutoff activity.
(A) PA and PA-X share the nuclease domain and any amino acid substitutions in the active site. Top: linear diagram of PA and PA-X proteins. Position of the F191 amino acid that marks the end of the shared sequence between PA and PA-X is indicated. Bottom: ribbon diagram of the nuclease domain structure with baloxavir acid (teal) bound in the active site (PDB ID: 6FS6) [12]. Side chains and positions of the conserved amino acids in the active site D108 (essential for nuclease activity) and I38 (site of baloxavir resistance mutations) are highlighted in red and brown, respectively. (B-D) 293A cells were co-transfected with the EGFP reporter plasmid and either the myc-tagged PA-X expression construct derived from the PR8 strain (PA-X) or the empty vector control (vect.). At 4 h post-transfection, cells were treated with the indicated concentrations of baloxavir acid (BXA) or vehicle control (-) and analysed at 24 h post-transfection. (B) Representative fluorescence microscopy images showing reporter expression. Scale bar = 500 µm. (C) The relative signal intensity in the GFP channel was quantified from images represented in panel B (N = 4, see materials and methods section for details). Multiple unpaired t-Tests were used to determine statistical significance for each BXA concentration and p values <0.05 presented above graphs (ns = non-significant). Error bars = standard deviation. (D) Whole cell lysates were collected from transfected cells treated as indicated and the expression of EGFP and PA-X was visualised by western blotting with antibodies to GFP and the myc tag, respectively. Actin was used as a loading control.
Fig 2.
Baloxavir resistance mutations impair PA-X shutoff activity.
(A-G) 293A cells were co-transfected with the EGFP reporter plasmid and the indicated myc-tagged wild-type (WT) or mutant PA-X expression construct derived from the PR8 strain and analysed at 24 h post-transfection. (A) Representative fluorescence microscopy images showing reporter expression. Scale bar = 500 µm. (B) The relative signal intensity in the GFP channel was quantified from images represented in panel A (N = 3, see materials and methods section for details). One-way ANOVA and Dunnett multiple comparisons tests were used to determine statistical significance (****P-value <0.0001; ***P-value < 0.001; **P-value < 0.01). Error bars = standard deviation. (C) Whole cell lysates were collected from cells transfected with the indicated PA-X expression constructs or with empty vector (vect.) and the expression of PA-X was visualised by western blotting with antibody to myc tag (myc). Actin was used as a loading control. (D) Representative fluorescence in situ hybridization (FISH) microscopy images showing levels and subcellular distribution of total cellular poly(A) RNA (magenta) and GAPDH transcript (white). White arrows indicate representative cells with high levels of GAPDH mRNA and poly(A) RNA in the cytoplasm, open arrows indicate representative cells with depleted GAPDH and poly(A) RNA. Open arrowheads indicate representative cells with depleted cytoplasmic GAPDH and increased nuclear poly(A) RNA signal. Scale bar = 100 µm. (E-G) The relative GAPDH signal (E), cytoplasmic poly(A) signal (F), and the cytoplasmic to nuclear poly(A) signal ratio (G) in GFP-positive cells was quantified from images represented in panel D as described in materials and methods (Cyt. = cytoplasmic; Nuc. = nuclear). Nonparametric Kruskal-Wallis tests with Dunn’s multiple comparisons were used to determine statistical significance (****P-value <0.0001; ***P-value < 0.001; **P-value < 0.01; *P-value < 0.05; ns: non-significant). Datapoints on bar graphs represent values from individual cells (23 cells per condition from 2 independent biological replicates). Error bars = standard deviation. (H) 293A cells were co-transfected with the firefly luciferase influenza replicon vector, the PR8 NP, PB2, PB1 expression constructs, and the indicated wild-type (WT) or mutant myc-tagged PR8 PA expression plasmids. At 4 h post-transfection, cells were treated with the indicated concentrations of baloxavir acid (BXA) and the luciferase activity was measured 24 h post-transfection. RLU = relative luminescence units.
Fig 3.
Baloxavir resistance mutations impair host shutoff by a laboratory-adapted influenza virus strain.
A549 cells were infected with the reverse-genetic derived PR8 wild-type (WT) or the indicated PA mutant influenza viruses at a multiplicity of infection (MOI) of 1.0 or mock infected and analyzed at 21 h post-infection. (A) Immunofluorescence microscopy images showing subcellular distribution of the cytoplasmic poly(A) binding protein (PABPC1, white). Infected cells are immunostained for NP expression (teal), nuclei are stained with Hoechst dye (purple). Scale bar = 100 µm. (B) Fraction of infected cells with nuclear PABPC1 accumulation was quantified from images presented in panel A (N = 4, see materials and methods section for details). (C) Accumulation of viral proteins in infected cells was visualized by western blotting using the indicated antibodies. Actin was used as a loading control. (D,E) Relative levels of host ACTB (D) and G6PD (E) transcripts in mock and virus-infected cells was determined using RT-qPCR (N = 3). In B, D, and E, one-way ANOVA and either Dunnett (B) or Tukey (D,E) multiple comparisons tests were used to determine statistical significance (****P-value <0.0001; ***P-value <0.001; **P-value < 0.01; ns: non-significant). Error bars = standard deviation.
Fig 4.
Baloxavir resistance mutations significantly attenuate host shutoff by the 2009 pH1N1 strain.
(A-B) 293A cells were co-transfected with the EGFP reporter plasmid and the indicated myc-tagged wild-type (WT) or mutant PA-X expression construct derived from the CA07 strain and analysed at 24 h post-transfection. (A) Whole cell lysates were collected from cells transfected with the indicated PA-X expression constructs or with empty vector (vect.) and the expression of PA-X was visualised by western blotting with antibody to myc tag (myc). Actin was used as a loading control. (B) The relative signal intensity in the GFP channel was quantified from fluorescence microscopy images (N = 3, see materials and methods section for details). (C-H) A549 cells were infected with the reverse-genetic derived CA07 wild-type (WT) or the indicated PA mutant influenza viruses at a multiplicity of infection (MOI) of 1.0 or mock infected and analyzed at 21 h post-infection. (C) Accumulation of viral proteins in infected cells was visualized by western blotting using the indicated antibodies. Actin was used as a loading control. (D) Immunofluorescence microscopy images showing subcellular distribution of the cytoplasmic poly(A) binding protein (PABPC1, white). Infected cells are immunostained for NP expression (teal), nuclei are stained with Hoechst dye (purple). Scale bar = 100 µm. (E) Fraction of infected cells with nuclear PABPC1 accumulation was quantified from images presented in panel D (N = 3, see materials and methods section for details). (F-H) Relative levels of host ACTB (F), G6PD (G), and IFNB1 (H) transcripts in mock and virus-infected cells was determined using RT-qPCR (N = 3). In B and E-H, one-way ANOVA and either Dunnett (E) or Tukey (B, F-H) multiple comparisons tests were used to determine statistical significance (****P-value <0.0001; ***P-value < 0.001; **P-value < 0.01; *P-value < 0.05; ns: non-significant). Error bars = standard deviation.
Fig 5.
PA-X proteins from H3N2 and H5N1 influenza virus strains have higher shutoff activity compared to H1N1-derived PA-X.
(A,B) 293A cells were co-transfected with the EGFP reporter plasmid and the indicated myc-tagged wild-type (WT) or mutant PA-X expression constructs derived from the SW13 (H3N2) or TX24 (H5N1) strains and analysed at 24 h post-transfection. (A) Representative fluorescence microscopy images showing reporter expression in cells co-transfected with the indicated WT or mutant SW13 PA-X constructs. Scale bar = 500 µm. (B) The relative signal intensity in the GFP channel was quantified from fluorescence microscopy images (N = 3, see materials and methods section for details). (C-G) 293A cells were co-transfected with the firefly luciferase reporter plasmid and the indicated myc-tagged wild-type (WT) or mutant PA-X expression constructs derived from the CA07 (H1N1), SW13 (H3N2), TX24 (H5N1), or BC24 (H5N1) strains and analysed at 24 h post-transfection. (C) The relative firefly luciferase activity was measured (N = 3). (D,E) Whole cell lysates were collected from cells transfected with the indicated PA-X expression constructs, empty vector (vect.), or catalytically inactive D108A mutant PR8 PA-X (108A) and the expression of PA-X was visualised by western blotting with antibody to myc tag (myc). Actin was used as a loading control. (F) Relative band intensities were quantified from western blot gels represented in D, E, and Fig 4A. (G) Reporter shutoff from panel C was normalized to PA-X expression levels from F (see materials and methods for details). In B,C, F, and G, one-way ANOVA and Tukey multiple comparisons tests were used to determine statistical significance (****P-value <0.0001; ***P-value < 0.001; **P-value < 0.01; *P-value < 0.05; ns: non-significant). Error bars = standard deviation. (H,I) Dynamic range and the limit of detection (L.O.D.) of the GFP-based (H) and firefly luciferase-based (I) reporter shutoff assays is represented based on raw signal values from reporter co-transfection with either vector control (vect.) or SW13 PA-X (SW), versus background transfection without the reporter (neg.). Error bars = standard deviation. L.O.D. (dotted line) is defined as 2 standard deviations over background.
Fig 6.
Baloxavir resistance mutations in PA of H5N1 influenza virus do not inhibit polymerase function and impair host shutoff.
(A,B) 293A cells were co-transfected with the firefly luciferase influenza replicon vector, the BC24 NP, PB2, PB1 expression constructs, and the indicated wild-type (WT) or mutant myc-tagged BC24 PA expression plasmids. At 4 h post-transfection, cells were treated with the indicated concentrations of baloxavir acid (BXA) and analyzed 24 h post-transfection. (A) The firefly luciferase activity was measured (N = 3). RLU = relative luminescence units. (B) Expression of PA in transfected cells was visualized by western blotting using anti-myc (myc) antibody. Actin was used as a loading control. (C-I). 293A cells were co-transfected with the EGFP influenza replicon vector, the BC24 NP, PB2, PB1 expression constructs, and the indicated wild-type (WT) or mutant myc-tagged BC24 PA expression plasmids and analyzed at 24 h post-transfection. (C) Immunofluorescence microscopy images showing subcellular distribution of the cytoplasmic poly(A) binding protein (PABPC1, white). Cells with active reconstituted replicon are visualized by GFP fluorescence (teal), nuclei are stained with Hoechst dye (magenta). Scale bar = 100 µm. (D) Fraction of GFP-positive cells with nuclear PABPC1 accumulation was quantified from images presented in panel C (N = 3, see materials and methods section for details). One-way ANOVA and Tukey multiple comparisons tests were used to determine statistical significance (****P-value <0.0001; ***P-value < 0.001; *P-value < 0.05). Error bars = standard deviation. (E) Immunofluorescence microscopy images showing subcellular distribution of the stress granule marker G3BP1 (teal) and the cytoplasmic poly(A) binding protein (PABPC1, magenta). (F) Immunofluorescence microscopy images showing subcellular distribution of the stress granule markers G3BP1 (teal) and TIAR (magenta). In E and F, insets highlight representative areas of the cytoplasm where stress granule markers display puncta. Scale bar refers to the inset dimensions (10 x 10 µm). (G-I) Number of PABPC1 (G) or G3BP1 (H) foci and the average foci size (I) were quantified from images presented in panel E (N = 3, violin plots represent distribution of individual values from at least 71 cells analyzed per condition, see materials and methods section for details). In panels G and I, nonparametric two-tailed Mann-Whitney test, and in panel H, unpaired t-Test with Welch’s correction were used to determine statistical significance (****P-value <0.0001; ns: non-significant).
Fig 7.
A working model diagram for the differential effects of baloxavir resistance mutations on PA vs. PA-X function.
Left: RNA polymerase cap snatching function involves coordinated action of the PB2 cap binding and the PA nuclease domains. Capture of the pre-mRNA 5’ end cap by the PB2 cap binding domain restricts movement of the RNA and orients it towards the PA endonuclease domain for optimal cleavage. Right: transient interaction of PA-X with RNA requires high nuclease activity for efficient cleavage. Structures for the trimeric RNA polymerase complex and the PA endonuclease domain are derived from PDB ID 7NHX [58]. Nuclease active site is represented by an orange triangle.