Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 292 (48), 19752-19766

Structural Analysis of the STAT1:STAT2 Heterodimer Revealed the Mechanism of Sendai Virus C Protein-Mediated Blockade of Type 1 Interferon Signaling

Affiliations

Structural Analysis of the STAT1:STAT2 Heterodimer Revealed the Mechanism of Sendai Virus C Protein-Mediated Blockade of Type 1 Interferon Signaling

Kosuke Oda et al. J Biol Chem.

Abstract

Sendai virus (SeV), which causes respiratory diseases in rodents, possesses the C protein that blocks the signal transduction of interferon (IFN), thereby escaping from host innate immunity. We previously demonstrated by using protein crystallography that two molecules of Y3 (the C-terminal half of the C protein) can bind to the homodimer of the N-terminal domain of STAT1 (STAT1ND), elucidating the mechanism of inhibition of IFN-γ signal transduction. SeV C protein also blocks the signal transduction of IFN-α/β by inhibiting the phosphorylation of STAT1 and STAT2, although the mechanism for the inhibition is unclear. Therefore, we sought to elucidate the mechanism of inhibition of the IFN signal transduction via STAT1 and STAT2. Small angle X-ray scattering analysis indicated that STAT1ND associates with the N-terminal domain of STAT2 (STAT2ND) with the help of a Gly-rich linker. We generated a linker-less recombinant protein possessing a STAT1ND:STAT2ND heterodimeric structure via an artificial disulfide bond. Analytical size-exclusion chromatography and surface plasmon resonance revealed that one molecule of Y3 can associate with a linker-less recombinant protein. We propose that one molecule of C protein associates with the STAT1:STAT2 heterodimer, inducing a conformational change to an antiparallel form, which is easily dephosphorylated. This suggests that association of C protein with the STAT1ND:STAT2ND heterodimer is an important factor to block the IFN-α/β signal transduction.

Keywords: innate immunity; interferon; negative-strand RNA virus; paramyxovirus; signal transduction; small-angle X-ray scattering (SAXS); surface plasmon resonance (SPR).

Conflict of interest statement

The authors declare that they have no conflicts of interest with the contents of this article

Figures

Figure 1.
Figure 1.
Generation of N-terminal domain-deleted mutants of STAT1 and STAT2 and functional analysis of C proteins. A, schematic diagram of constructs of C′, C, Y1, Y2, and Y3. B, linear representation of the domains in human STAT1, STAT2, and their N-terminal domain-deleted mutants (STAT1ΔN and STAT2ΔN). N, N-terminal domain; CC, coiled-coil domain; DNA, DNA-binding domain; LK, linker domain; SH2, SH2 domain; Y, phosphorylated tyrosine residue; TA, transcription activation domain. C, to estimate the strength of the response to IFN-α, subconfluent 293T cells were transfected with pISRE-EGFP and an expression plasmid for FL-C, FL-Y1, or FL-Y3, and IFN-α (20 units/ml) was added to the culture medium at 6 h after transfection. At 24 h post-transfection, photographs were taken under an immunofluorescent microscope. D, HeLa cells were transfected with an expression vector for HA-STAT1, HA-STAT2, or FL-C. A portion of the cell lysates prepared at 24 h post-transfection were mixed as indicated and immunoprecipitated (IP) with an anti-FLAG antibody (Anti-FL) together with protein G-Sepharose. The immunoprecipitates were separated by SDS-PAGE followed by Western blot analysis (WB) using an anti-HA antibody (Anti-HA) and anti-FL antibody. A part of the cell lysates was used to confirm protein expression using anti-HA and anti-FL antibodies. E, HeLa cells were transfected with an expression vector for HA-STAT1 or HA-STAT1ΔN together with an expression vector for FL-C or FL-Y3. At 24 h post-transfection, the cell lysate was immunoprecipitated with anti-FL antibody and analyzed by Western blotting using anti-HA and anti-FL antibodies. *, a light chain of IgG.
Figure 2.
Figure 2.
Inhibition of IFN-α-induced tyrosine phosphorylation of STAT1 in the presence of C protein. U3A cells were transfected with an expression vector for HA-STAT1 or HA-STAT1ΔN together with an expression vector for FL-C (A) or FL-Y3 (B). At 0.5 h after stimulation with IFN-α (1,000 units/ml), proteins in the cell extracts were separated by SDS-PAGE for Western blot analysis with an anti-STAT1 antibody (anti-STAT1), an anti-Tyr701-phosphorylated STAT1 antibody (anti-pSTAT1) and anti-FL. C, the rate of phosphorylation inhibition was determined on the basis of averaged signal intensities of HA-pSTAT1 in the presence or absence of FL-C (A) or FL-Y3 (B), which was calculated from three independent experiments. Intensities of bands were measured using ImageJ version 1.47, and signal intensity of HA-STAT1 was used as an internal standard. An error bar indicates standard deviation. p value was calculated on the basis of Welch's test.
Figure 3.
Figure 3.
Inhibition of IFN-α-induced phosphorylation of STAT2 in the presence of C protein. U6A cells in a 35-mm dish were transfected with an expression vector for FL-STAT2 (1 μg) or FL-STAT2ΔN (1 μg) together with an expression vector for FL-C (0.4 μg) (A) or FL-Y3 (1 μg) (B). At 24 h post-transfection, the cells were stimulated with IFN-α (1,000 units/ml) for 1 h. C, U3A cells were transfected with an expression vector for FL-STAT2 together with an expression vector for FL-C or FL-Y3. At 24 h post-transfection, the cells were stimulated with IFN-α (1,000 units/ml) for 1 h. Proteins in the cell extracts were separated by SDS-PAGE for Western blot analysis using an anti-Tyr690-phosphorylated STAT2 antibody (Anti-pSTAT2) and anti-FL. D, the rate of phosphorylation inhibition was determined on the basis of averaged signal intensity of FL-STAT2 in A–C, which was calculated from three independent experiments. Signal intensity of FL-STAT2 was used as an internal standard.
Figure 4.
Figure 4.
Interaction between STAT1ND and STAT2ND. A, schematic diagram of the STAT1ND:STAT2ND fusion protein. B, distance distribution function of the fusion protein obtained by an SAXS experiment. C, heterodimer model of STAT1ND (cyan) and STAT2ND (green), shown as ribbon representation, is imposed on a low-resolution bead model calculated using DAMMIF by fitting to the experimental scattering curve shown in B. The STAT1ND:STAT2ND heterodimer model was created by replacing one subunit in the STAT1ND homodimer (PDB code 3WWT) with a STAT2ND homology model generated using SWISS-MODEL server (EXPASY) and the “align” command in PyMOL (58). D, an ensemble model generated by EOM analysis using atomic coordinates of the STAT1ND:STAT2ND heterodimer model. Cα atoms in the linker peptide are shown in sphere representation. STAT1ND and STAT2ND are colored in blue and orange, respectively, and their N and C termini are marked. Structural parameters obtained by EOM analysis are shown in the table. E, experimental scattering curve of the fusion protein and theoretical scattering curve of the ensemble model are colored in blue and magenta, respectively. F, the STAT1ND:STAT2ND heterodimer model is shown in ribbon representation. Carbon atoms in STAT1ND and STAT2ND are colored in cyan and green, respectively. The region enclosed with a red circle is enlarged below. Glu16, Asp23, and His19 from STAT1ND and His85, Arg88, and Arg92 from STAT2ND are shown in a stick model. The dotted lines represent possible hydrogen bonds. G, analytical size-exclusion chromatograms of STAT1ND:STAT2ND (black line), STAT1ND:STAT2NDR88E (gray line), or STAT1ND:STAT2NDR88E-R92E fusion protein (dotted line) are shown. All of the structural drawings were made using PyMOL.
Figure 5.
Figure 5.
Inhibition of the phosphorylation of STAT2 with decreased affinity to STAT1ND in the presence of Y3. A, HeLa cells were transfected with an expression vector for FL-STAT2, FL-STAT2R88E, or FL-STAT2R88E-R92E together with an expression vector for wild-type HA-STAT2. At 24 h post-transfection, the cell lysate was immunoprecipitated with anti-FL antibody and analyzed by Western blotting using anti-HA and anti-FL antibodies. B, U6A cells were transfected with an expression vector for FL-STAT2, FL-STAT2R88E, or FL-STAT2R88E-R92E together with an expression vector for FL-Y3. At 24 h post-transfection, the cells were stimulated with IFN-α (1,000 units/ml) for 1 h (B). Proteins in the cell extracts were separated by SDS-PAGE for Western blot analysis using anti-pSTAT2, anti-FL, anti-STAT1, and anti-phosphorylated STAT1 antibodies. C, the rate of phosphorylation inhibition was determined on the basis of averaged signal intensity of FL-pSTAT2 in B, which was calculated from three independent experiments. Signal intensity of FL-STAT2 was used as an internal standard. D, relative quantities of phosphorylated STAT1 were determined on the basis of the average signal intensity of pSTAT1 in B, which was calculated from three independent experiments. Signal intensity of STAT1 was used as an internal standard.
Figure 6.
Figure 6.
Response to IFN-α of U6A cells constitutively expressing STAT2 mutants. A, quantification analysis of constitutively expressed wild-type or a mutant of FL-STAT2 in U6A cells. The cell lysate was analyzed by Western blotting using an anti-STAT2 antibody (anti-STAT2) and anti-actin antibody (anti-actin). B, comparison of ISRE reporter activities at the basal level in U6A cells constitutively expressing wild-type or a mutant of FL-STAT2. To estimate the strength of ISRE reporter activities in the absence of IFN-α stimulation, subconfluent U6A cells constitutively expressing FL-STAT2 or a mutant were transfected with pISRE-EGFP. At 24 h after transfection, the cell lysate was analyzed by Western blotting using an anti-GFP antibody (anti-GFP). At the same time, U6A cells were transfected with pCAG-EGFP, and the cell lysate was used as an external standard. C, EGFP reporter assay for signal transduction of IFN-α using U6A cells. To estimate the strength of the response to IFN-α, subconfluent U6A cells and cells constitutively expressing FL-STAT2 or a mutant were transfected with pISRE-EGFP and an expression plasmid for FL-C or FL-Y3, and IFN-α (2,000 units/ml) was added at 16 h post-transfection. After 8 h, the cell lysate was analyzed by Western blotting using anti-GFP (anti-GFP), anti-actin, and anti-FL antibodies. Quantified intensity of EGFP calculated from three independent experiments is shown in the graph. An error bar indicates standard deviation. Signal intensity of actin was used as an internal standard in this figure.
Figure 7.
Figure 7.
Antiviral action of IFN-α in U6A cells expressing STAT2 mutants after SeV infection. A, U6A cells constitutively expressing FL-STAT2 or the mutant were infected with SeV. At 1 h post-infection, the cells were incubated with serum-free DMEM for 4 h at 37 °C, followed by the addition of IFN-α (2,000 units/ml) to the culture medium. After an additional incubation for 9 h, the cells were superinfected with rVSV-EGFP. After further incubation for 10 h, cell lysates were prepared to estimate the expression levels of rVSV-EGFP-derived EGFP and SeV C protein by Western blot analysis. B, ratio of EGFP expression in IFN-α-treated SeV-infected cells to that in mock-infected cells. C, to monitor the signal transduction of IFN-α in SeV-infected cells, subconfluent U6A cells were transfected with pISRE-EGFP. After 10 h, the cells were infected with SeV as mentioned above. The cells were incubated with serum-free DMEM for 4 h at 37 °C, followed by the addition of IFN-α (2,000 units/ml) to the culture medium. After an additional incubation for 9 h, the cell lysate was analyzed by Western blotting. Relative densities of EGFP in A and C are shown as a bar graph, in which density of EGFP in the IFN-α-untreated mock-infected cells was normalized to 1.
Figure 8.
Figure 8.
Binding between Y3 and STAT1ND:STAT2ND heterodimer. A, schematic diagram of the fusion protein to generate a disulfide-bonded STAT1ND:STAT2ND heterodimer. B, analytical size-exclusion chromatograms of the fusion protein (black line) and the disulfide-bonded STAT1ND:STAT2ND heterodimer in the presence (dotted line) or absence (gray line) of 1 mm dithiothreitol. C, experimental scattering curve of the disulfide-bonded STAT1ND:STAT2ND heterodimer in the presence of 0.5 mm cystine and theoretical scattering curve of STAT1ND:STAT2ND heterodimer are colored in blue and magenta, respectively. D, analytical size-exclusion chromatograms of Y3W125A (gray line) and disulfide-bonded STAT1ND:STAT2ND heterodimer in the presence (dotted line) or absence (black line) of Y3W125A. Proteins in fractions eluted from a mixture of the disulfide-bonded STAT1ND:STAT2ND heterodimer and 10-fold excess of Y3W125A were separated by SDS-PAGE, followed by silver staining and Western blotting using an anti-C antibody. E, binding of Y3W125A to the disulfide-bonded STAT1ND:STAT2ND heterodimer immobilized on the sensor chip. The maximal increase in RU indicates that one molecule of Y3W125A binds to the STAT1ND:STAT2ND heterodimer. The graph shows the binding ratio against the concentration of Y3W125A. Each point in the graph corresponds to experimental data, whereas the curve is theoretically drawn. The inset shows the change in sensorgrams after injecting Y3W125A. RU just before immobilization of the disulfide-bonded STAT1ND:STAT2ND heterodimer was set to 0.

Similar articles

See all similar articles

Cited by 1 PubMed Central articles

Publication types

Associated data

LinkOut - more resources

Feedback