Phosphorylation of STAT2 on serine-734 negatively regulates the IFN-α-induced antiviral response

Abstract

Serine phosphorylation of STAT proteins is an important post-translational modification event that, in addition to tyrosine phosphorylation, is required for strong transcriptional activity. However, we recently showed that phosphorylation of STAT2 on S287 induced by type I interferons (IFN-α and IFN-β), evoked the opposite effect. S287-STAT2 phosphorylation inhibited the biological effects of IFN-α. We now report the identification and characterization of S734 on the C-terminal transactivation domain of STAT2 as a new phosphorylation site that can be induced by type I IFNs. IFN-α-induced S734-STAT2 phosphorylation displayed different kinetics to that of tyrosine phosphorylation. S734-STAT2 phosphorylation was dependent on STAT2 tyrosine phosphorylation and JAK1 kinase activity. Mutation of S734-STAT2 to alanine (S734A) enhanced IFN-α-driven antiviral responses compared to those driven by wild-type STAT2. Furthermore, DNA microarray analysis demonstrated that a small subset of type I IFN stimulated genes (ISGs) was induced more by IFNα in cells expressing S734A-STAT2 when compared to wild-type STAT2. Taken together, these studies identify phosphorylation of S734-STAT2 as a new regulatory mechanism that negatively controls the type I IFN-antiviral response by limiting the expression of a select subset of antiviral ISGs.

Keywords: Interferon; JAK1; STAT2; Serine phosphorylation; Vesicular stomatitis; Virus infection.

Fig. 1.

Mass spectrometry identifies S734 as a type I IFN-inducible phosphorylatable residue in human STAT2 that is conserved in primates. (A) Tandem mass spectrometry (MS/MS) spectrum of STAT2 after treatment with IFN-α. +P (lower panel) indicates S734 as being phosphorylated and the underlined peptide mass (lower panel) indicates the presence of the peptide in the spectrum (upper panel). P, phosphoryl group (79.97 Da). Not all peaks are annotated in the mass spectrum. Matches were made with a mass tolerance of ±2 Da. (B) S734 is mapped to the transactivation domain (TAD) of STAT2. (C) Cross-species conservation of this phosphorylation site is seen in primate STAT2.

Fig. 2.

STAT2 is phosphorylated at S734 in response to IFN-α treatment. (A) U6A cells stably expressing empty vector, WT-STAT2 or S734A-STAT2 were treated with human IFN-α (1000 U/ml) for the indicated times. Phosphorylation of S734-STAT2 was detected by western blot analysis using an affinity purified polyclonal antibody raised against this phosphorylated epitope. IFN-α-induced tyrosine phosphorylation of STAT2 (pY-STAT2) and STAT1 (pY-STAT1) together with total levels of STAT2 and STAT1 were also assessed by western blot analysis. (B) Immunoreactivity of the pS734-STAT2 antibody is confirmed by a lambda protein phosphatase (LPP) assay. (C–F) Immortalized hTERT cells (C) and human colorectal cancer cells F6-8 (D), T29 (E) and (F) Stat2^−/− MEFs reconstituted with human STAT2 were stimulated with either human IFN-α (1000 U/ml) or murine IFN-β (1000 U/ml), accordingly, and examined as in A. GAPDH was used as an internal loading control.

Fig. 3.

Phosphorylation of S734-STAT2 is dependent on STAT2 tyrosine phosphorylation and catalytically active JAK kinases. (A) U6A cells stably expressing WT-, R409A/K415A (RKAA)-, or Y690F-STAT2 were left untreated or treated for 1 and 4 h with IFN-α (1000 U/ml). Tyrosine and serine phosphorylation of STAT2, and GAPDH were assessed by western blot analysis. Data were quantified and shown as the ratio of pS734-STAT2 to STAT2 and pY690-STAT2 to STAT2. The arrow indicates a non-specific protein band (NS). (B) U6A cells expressing WT-STAT2 were pre-treated or not with pan-JAK inhibitor overnight (o/n) and (C) U4A cells and U6 cells expressing WT-STAT2 stimulated with IFN-α (1000 U/ml) for the specified times were evaluated for pS734-STAT2, pY-STAT2, pY-STAT1, STAT1, STAT2 and GAPDH as indicated. (D) STAT1 immune complexes obtained from U6A cells expressing WT-STAT2 that were left untreated or treated with IFN-α (1000 U/ml) were evaluated for interactions with pS734-STAT2 and total STAT2 by western blot analysis.

Fig. 4.

Subcellular localization of phosphorylated S734-STAT2. Nuclear and cytoplasmic fractions of human (A) 2fTGH and (B) hTERT cells treated with IFN-α (1000 U/ml) for the given times were evaluated for pS734-STAT2, pY-STAT2, pY-STAT1, STAT1 and STAT2 by western blot analysis as indicated. Presence of GAPDH in the cytoplasmic extract and presence of Lamin B1 in the nuclear extract confirmed purity of fractions. The arrow indicates the upper band corresponding to pS734-STAT2. Actin was used as an internal loading control.

Fig. 5.

Phosphorylation of S734-STAT2 does not affect the antiproliferative effects of IFN-α. (A) Growth rate of U6A cells stably expressing WT-, S734A- or S734D-STAT2 was measured at 48 h and 72 h by an MTS assay. The inset shows a western blot analysis of STAT2 expression in U6A cells expressing empty vector, WT-, S734A or S734D-STAT2. (B) The same panel of cells was treated with 100 U/ml or 1000 U/ml of IFN-α for 72 h and cell viability determined by MTS assay. Results are presented as the percentage of viable cells in the presence of IFN-α compared against untreated cells. Data shown are from three or four independent experiments and presented as mean±s.e.m.

Fig. 6.

S734A-STAT2 enhances IFN-α-induced protection against VSV infection. (A) The U6A cell panel was left untreated or pre-treated with IFN-α (1000 U/ml) for 16 h and then infected with GFP-tagged vesicular stomatitis virus (VSV) for 36 h. Viral titers after 36 h post-viral infection (h.p.i.) were determined by plaque assay. Viral titers are shown on a log scale. (B) Western blot analysis of STAT2 expression in Stat2^−/− MEFs reconstituted with either human WT- or S734A-STAT2. (C) Stat2^−/− MEFs expressing human WT- or S734A-STAT2 were pre-treated or not with murine IFN-β (1000 U/ml) for 16 h and VSV infection was quantified by flow cytometry analysis of GFP-positive cells. Data are shown as the mean±s.e.m. from three or four independent experiments. **P<0.01 (Student's t-test). NS, not statistically significant.

Fig. 7.

S734-STAT2 regulates the expression of a subset of ISGs. (A) Microarray analysis identified a small subset of ISGs for which mRNA levels were found to be enhanced in S734A-STAT2 U6A cells when compared against WT-STAT2 U6A cells after 4 h of IFN-α (1000 U/ml) treatment. Genes marked in A with an asterisk were subsequently validated by qPCR using (B) U6A or (C) Stat2^−/− MEFs reconstituted with either human WT- or S734-STAT2 treated with or without IFN-α or IFN-β (1000 U/ml) for 6, 18 or 24 h. The mRNA levels were normalized to actin expression and calculated by using ΔΔCt method. Data are presented as the mean±s.e.m. fold change from untreated cells from three or four independent experiments. *P<0.05, **P<0.01 (Student's t-test). (D) WT-STAT2 and S734A-STAT2 U6A cells were stimulated or not with IFN-α (1000 U/ml) for 24 or 60 h and levels of ISG15 and OAS1 as well as pY-STAT2, pY-STAT1, STAT1, STAT2 and GAPDH as indicated, were assessed by western blot analysis.

Fig. 8.

Proposed model of S734-STAT2 phosphorylation in regulating the antiviral activity of type I IFNs. Upon binding of IFN-α and IFN-β to its receptor, JAK1 and TYK2 become activated and tyrosine phosphorylate STAT1 (pY701) and STAT2 (pY690). Activated STAT1 and STAT2, together with IRF9, as the ISGF3 complex, trigger the induction of antiviral ISGs. Subsequently, tyrosine-phosphorylated STAT2 (pY690) is phosphorylated at S734 (pS734) in a JAK-dependent manner, and this in turn, restricts the induction of ISGs with antiviral activity.