Virus-induced unfolded protein response attenuates antiviral defenses via phosphorylation-dependent degradation of the type I interferon receptor - PubMed (original) (raw)

Virus-induced unfolded protein response attenuates antiviral defenses via phosphorylation-dependent degradation of the type I interferon receptor

Jianghuai Liu et al. Cell Host Microbe. 2009.

Abstract

Phosphorylation-dependent ubiquitination and degradation of the IFNAR1 chain of the type I interferon (IFN) receptor is regulated by two different pathways, one of which is ligand independent. We report that this ligand-independent pathway is activated by inducers of unfolded protein responses (UPR), including viral infection, and that such activation requires the endoplasmic reticulum-resident protein kinase PERK. Upon viral infection, activation of this pathway promotes phosphorylation-dependent ubiquitination and degradation of IFNAR1, specifically inhibiting type I IFN signaling and antiviral defenses. Knockin of an IFNAR1 mutant insensitive to virus-induced turnover or conditional knockout of PERK prevented IFNAR1 degradation, whether UPR-induced or virus-induced, and restored cellular responses to type I IFN and resistance to viruses. These data suggest that specific activation of the PERK component of UPR can favor viral replication. Interfering with PERK-dependent IFNAR1 degradation could therefore contribute to therapeutic strategies against viral infections.

PubMed Disclaimer

Figures

Figure 1

ER stress induces IFNAR1 Ser535 kinase activity and promotes phosphorylation of IFNAR1 within its destruction motif in a manner that does not require Tyk2 activity but relies on activity of PERK. A. Lysates from KR-2 cells (lacking catalytic activity of Tyk2) transfected with Flag-IFNAR1 plasmid (0 – 3.0μg) were subjected to immunoprecipitation (IP) using anti-Flag antibody followed by immunoblotting (IB) using the indicated antibodies. Relative intensity of bands in IB using either anti-phospho-S535 (black squares) or IFNAR1 (R1, gray squares) or Flag (black squares) was quantified and plotted in the right panel. B. Lysates from KR-2 cells transfected with Flag-IFNAR1 (1.5-3.0μg) or empty vector were used as a source of kinase activity in an in vitro kinase assay using GST-IFNAR1 as a substrate. The reactions were analyzed by IB using anti-phospho-S535 antibody (upper panel; both shorter and longer exposures shown) and by Ponceau staining to detect the substrate levels (middle panel). Levels of Flag-IFNAR1 in the cell lysates were analyzed by IB using anti-Flag antibody (lower panel). C. 293T cells were treated with thapsigargin (TG, 1μM) for 30 min. Endogenous IFNAR1 was immunoprecipitated and analyzed by IB using indicated antibodies. Aliquots of the whole cell lysates were analyzed for levels of phospho- and total eIF2α. D. Cells harboring wild type (WT-5) or the kinase dead Tyk2 (KR-2) were treated with IFNα (1000IU/ml) or TG at indicated concentrations for 30 min. Endogenous IFNAR1 was analyzed by IP-IB as in panel C. E. WT or PERK-/- MEFs were treated with TG (1μM) or murine IFNβ (1000 U/ml) for 30 min. Mouse endogenous IFNAR1 was analyzed for its phosphorylation and levels using the indicated antibodies. F. MEFs from PERKfl/fl mice that received an empty vector (Mock) or vector for expression of Cre recombinase (Cre) were treated and analyzed as in panel F. Whole cell lysates from these cells were also analyzed by IB using the indicated antibodies.

Figure 2

ER stress promotes IFNAR1 ubiquitination and degradation in a ligand/Jak-independent manner A. Levels of endogenous IFNAR1 in 293T cells pre-treated or not with methylamine HCl (MA, 20mM) for 1 h and then treated with TG (1μM) for indicated time were analyzed by IP-IB. Levels of β-actin in whole cell lysates are also shown. B. Cells harboring the WT Tyk2 (WT-5) or the kinase dead Tyk2 (KR-2) were treated with TG as indicated and ubiquitination and levels of endogenous IFNAR1 were analyzed by IP-IB. Aliquots of whole cell lysates were also analyzed by IB using anti-β-actin antibody (lower panel). C. 293T cells were pre-treated or not with MA for 1 h and then treated with cycloheximide (Chx, 50μg/ml) alone or together with TG (1μM) for indicated times. Levels of endogenous IFNAR1 were analyzed by IP-IB. Levels of c-Jun and β-actin in whole cell lysates were also determined by IB using indicated antibodies. D. Ubiquitination of Flag-tagged IFNAR1 co-expressed with the indicated shRNA constructs in 293T cells was analyzed by IP using anti-Flag antibody followed by IB using anti-ubiquitin and anti-Flag antibodies as indicated. Aliquots of whole cell lysates were also analyzed by IB using anti-β-actin antibody (lower panel). E. PERKfl/fl MEFs that either underwent acute deletion of PERK (Cre) or not (Mock) were treated with 1μM of TG (together with Chx, 10μg/ml) for 45 min as indicated. Endogenous mouse IFNAR1 was analyzed by IP-IB using the indicated antibodies. Ig: heavy chain immunoglobulins. Whole cell lysates were also subjected to IB analysis to determine levels of phosphorylated β-catenin and eIF2α as well as total levels of PERK and eIF2α using respective antibodies. NS: non-specific band. F. Mouse Flag-IFNAR1 expressed in PERKfl/fl MEFs that either underwent acute deletion of PERK (Cre) or not (Mock) were analyzed by IB using anti-Flag antibody. Levels of PERK are shown in lower panel. NS: non-specific band that serves as a loading control.

Figure 3

ER stress ER stress promotes IFNAR1 degradation in a manner depending on IFNAR1 phosphorylation within its phospho-degron A. Targeting strategy to generate an S526A allele in mouse ES cells (C57/BL6). Position of mutated Ser residue, resistance markers, loxP sites as well as restriction sites and the probe used for Southern analysis are also shown. B. Southern analysis of several selected ES clones that underwent homologous recombination (marked by an asterisk) was performed on genomic DNA digested with KpnI. Correct targeting yielded a 10.6 kb band in clones 106 and 252 (besides the 8.5 kb band indicative of the WT allele). C. Embryoid bodies (EB) derived from the WT (WT/WT) or the mutant (S526A/WT) ES were trypsinized and the EB-derived cells were propagated on gelatinized plates. Cells were pre-treated or not with MA (20mM for 1 h) and then treated with TG (1μM, 15min) as indicated. Endogenous mouse IFNAR1 was immunoprecipitated and analyzed by IB using the indicated antibodies. Phosphorylation of eIF2α and levels of PKR were also determined in aliquots of whole cell lysates by IB. D. EB-derived cells were treated with TG (1μM) for indicated times and analyzed for total levels of endogenous IFNAR1 (by IP-IB). Levels of eIF2α phosphorylation and total β-catenin were shown as stress and loading controls, respectively.

Figure 4

Viral infection promotes phosphorylation-dependent ubiquitination and downregulation of IFNAR1 in a Tyk2-independent and S535/526-dependent manner. A. Ubiquitination, phosphorylation and total levels of endogenous IFNAR1 from KR-2 cells infected with VSV (for 16, 18 and 20 h) were analyzed by IP using anti-IFNAR1 antibody followed by IB using the indicated antibodies. Viral protein accumulation is shown by the levels of VSV-M. B. Levels of endogenous IFNAR1 in the lysates from Huh7 cells (parental or harboring either a full length or subgenomic HCV) were analyzed by IP-IB. Levels of β-actin in the lysates aliquots are also shown. C. Endogenous IFNAR1 proteins were immunopurified from the indicated cells (as in panel C) and loaded onto the gel to yield comparable levels of total IFNAR1 (lower panel). Phosphorylation of IFNAR1 was then analyzed by IB using indicated antibody (upper panel). D. MEFs from IFNAR1-/- mice were stably reconstituted with murine Flag-IFNAR1 (either wild type or S526A mutant) and then infected with VSV (for 16-18 h.) Levels of IFNAR1, VSV-M and β-actin were analyzed by IB. E. EB-derived cells of WT (WT/WT) or mutant (S526A/WT) genotype were infected (or not) with VSV for 12 h and lysed. Under these conditions, levels of VSV-M become saturated at 10hr post-infection (data not shown). Levels of endogenous mouse IFNAR1 were determined by IP-IB. Levels of β-actin and VSV-M in the lysates were also determined.

Figure 5

Role of PERK in virus-induced degradation of IFNAR1. A. Phosphorylation and levels of endogenous IFNAR1 in 2fTGH cells that received indicated shRNA constructs and then were infected with VSV (for 16, 18 and 20 h) was analyzed by IP-IB using the indicated antibodies. Aliquots of IP supernatants were used for analysis of VSV-M, p-eIF2α and eIF2α levels by IB. B. Control or PERK-depleted 2fTGH cells (as in panel A) were infected with VSV (for 17 h) and then treated with Chx (1 or 10μg/ml for 1.5 h). Total levels of IFNAR1 were determined by IP-IB. C. Levels of cell surface IFNAR1 analyzed by FACS using monoclonal anti-mIFNAR1 antibody in MEFs from PERKfl/fl mice (transduced with either empty vector (Mock) or construct for expression of Cre) were either left untreated (black line) infected with VSV (for 17 h, red line) or treated with TG (1μM for 4h, green line). Blue line represents the isotype Ig control. D. Levels of IFNAR1 and actin in Huh7 cells harboring the full-length or subgenomic HCV that were co-transfected with Flag-IFNAR1 and indicated shRNA constructs were analyzed using indicated antibodies.

Figure 6

Viral infection inhibits Type I IFN signaling via accelerating Ser526 phosphorylation-dependent degradation of IFNAR1 A. 2fTGH cells infected or not with VSV (for 20 h) were treated with 50 IU/ml of IFNα or IFNγ for 30 min. Phosphorylation of Stat1 and total levels of Stat1, actin and VSV-M were analyzed by IB. B. Indicated Huh7 cell line derivatives were co-transfected with Flag-STAT1 and either empty vector (pcDNA3) or Flag-IFNAR1 (WT or S535A) as indicated. Lysates from these cells treated or not with IFNα (50 IU/ml) were immunoprecipitated using anti-Flag antibody and these reactions were analyzed by IB using the indicated antibodies. C. EB-derived cells of wild type (WT/WT) or mutant (S526A/WT) genotype were infected with VSV (for 12 h) and then were treated with murine IFNβ (100 IU/ml) or IFNγ (5ng/ml) for 30 min. Phosphorylation of Stat1 and total levels of Stat1 and actin were analyzed by IB. D. Titer of VSV produced in EB-derived cells 14 h after infection (an incubation of cells with VSV at MOI 1.0 for 1h) is depicted. The effect of IFNβ (20 IU/ml) added either 16 h prior to the infection (pre-treat) or immediately after infection (co-add) was determined. Data shown (the mean ± SD) are representative of two independent experiments (each in triplicate). Asterisk denotes p<0.01 in comparison with untreated cells.

Figure 7

Role of PERK in virus-induced suppression of Type I IFN signaling A. Control or PERK-depleted derivatives of 2fTGH cells were infected with VSV (for 20 h) and then treated with IFNα or IFNγ (50 IU/ml for 30 min). Phosphorylation and total levels of Stat1 and eIF2α were analyzed by IB. B. MEFs from PERKfl/fl mice transduced with either empty vector (Mock) or construct for expression of Cre were infected with VSV (for 20 h) and then treated with IFNβ (100 IU/ml) or IFNγ (5ng/ml) for 30 min. Phosphorylation and total levels of Stat1 were analyzed by IB. C. MEFs from PERKfl/fl mice transduced as indicated were infected with VSV at MOI 0.1 (+) or 0.5 (++) for 20 h and then treated with IFNβ for 30 min. IB analyses using indicated antibodies are shown. D. Indicated derivatives of Huh7 cells were co-transfected with indicated shRNA constructs and Flag-STAT1 and treated with IFNα (50 IU/ml) for 30 min. Stat1 proteins were immunoprecipitated using anti-Flag antibody and analyzed by IB using anti-phospho-Stat1 and anti-Stat1 antibody. E. Titer of VSV produced in control or PERK-depleted derivatives of 2fTGH cells 14 h after infection (an incubation of cells with VSV at MOI 1.0 for 1h) is depicted. The effect of IFNα (20 IU/ml) added either 16 h prior to the infection (pre-treat) or immediately after infection (co-add) was determined. Data shown (the mean ± SD) are representative of two independent experiments (each in triplicate). Asterisk denotes p<0.01 in comparison with untreated cells. F. MEFs from PERKfl/fl mice transduced as indicated were infected with VSV (MOI 1.0). 20 h after infection, viral titer in the culture supernatant was determined. Values represent the mean ± SD of three independent experiments each performed in triplicate. VSV-M protein levels analyzed by IB in cell lysates are also shown in the inset. G. 2fTGH and isogenic IFNAR2-deficient U5a cells were transduced with indicated shRNA constructs and then infected with VSV (MOI 1.0) for 18-20 h. Levels of VSV-M, ISG15 and β-actin were determined by IB. In a parallel experiment, these cells were treated with TG (1μM for 30 min) and analyzed for PERK levels by IP-IB (lower panel).

References

1. Baltzis D, Qu LK, Papadopoulou S, Blais JD, Bell JC, Sonenberg N, Koromilas AE. Resistance to vesicular stomatitis virus infection requires a functional cross talk between the eukaryotic translation initiation factor 2alpha kinases PERK and PKR. J Virol. 2004;78:12747–12761. - PMC - PubMed
1. Bergmann M, Garcia-Sastre A, Carnero E, Pehamberger H, Wolff K, Palese P, Muster T. Influenza virus NS1 protein counteracts PKR-mediated inhibition of replication. J Virol. 2000;74:6203–6206. - PMC - PubMed
1. Biron CA. Interferons alpha and beta as immune regulators--a new look. Immunity. 2001;14:661–664. - PubMed
1. Blalock JE, Baron S. Mechanisms of interferon induced transfer of viral resistance between animal cells. J Gen Virol. 1979;42:363–372. - PubMed
1. Brassard DL, Grace MJ, Bordens RW. Interferon-alpha as an immunotherapeutic protein. J Leukoc Biol. 2002;71:565–581. - PubMed

Virus-induced unfolded protein response attenuates antiviral defenses via phosphorylation-dependent degradation of the type I interferon receptor - PubMed (original) (raw)