IKKβ phosphorylation regulates RPS3 nuclear translocation and NF-κB function during infection with Escherichia coli strain O157:H7

Abstract

NF-κB is a major gene regulator in immune responses, and ribosomal protein S3 (RPS3) is an NF-κB subunit that directs specific gene transcription. However, it is unknown how nuclear translocation of RPS3 is regulated. Here we report that phosphorylation of RPS3 Ser209 by the kinase IKKβ was crucial for nuclear localization of RPS3 in response to activating stimuli. Moreover, virulence protein NleH1 of the foodborne pathogen Escherichia coli strain O157:H7 specifically inhibited phosphorylation of RPS3 Ser209 and blocked RPS3 function, thereby promoting bacterial colonization and diarrhea but resulting in less mortality in a gnotobiotic piglet-infection model. Thus, the IKKβ-dependent modification of a specific amino acid in RPS3 promoted specific NF-κB functions that underlie the molecular pathogenetic mechanisms of E. coli O157:H7.

Figure 1. RPS3 is phosphorylated and associates with IKKβ in response to NF-κB activation

(a) ³²P-labeling assays were performed with HEK 293T cells stimulated with TNF (20 ng/ml) for the indicated times. Whole-cell lysates were subjected to immunoprecipitation (IP) with RPS3 antibody, followed by autoradiography or immunoblotting with RPS3 antibody. (b) Whole-cell lysates (Input) from Jurkat cells stimulated as indicated were directly immunoblotted for indicated proteins, or serine-phosphorylated (p-Ser), tyrosine-phosphorylated (p-Tyr), or threonine-phosphorylated (p-Thr) proteins, or RPS3 after immunoprecipitation with RPS3 antibody (IP: RPS3). (c) Interaction between HA-RPS3 and Flag-IKKβ in HEK 293T cells assessed by immunoprecipitation and immunoblot (IP: HA). (d) Whole-cell lysates (Input) from Jurkat cells stimulated as indicated were directly immunoblotted, or after immunoprecipitation (IP) with RPS3 antibody, for IKKα, IKKβ, or RPS3. Data are representative of at least two independent experiments.

Figure 2. IKKβ kinase activity is required for the nuclear translocation of RPS3

(a) Jurkat cells were transfected with siRNAs for IKKα, IKKβ, or scrambled nonspecific (NS) siRNA. 72 h later, cells were either left untreated (Unstim) or stimulated with 50 ng/ml of TNF or 50 ng/ml of PMA plus 1.5 mM ionomycin (PMA+I). Confocal images of RPS3 (red) are shown with nuclei counterstaining (blue). Percentages (mean and s.d.) of cells (n >200) with nuclear RPS3 were quantified, as shown in the right panel. (b) Representative immunoblots using indicated antibodies of whole cell lysates (WCL) or nuclear subcellular fractions (Nuc) from Jurkat cells transfected with indicated siRNAs and left unstimulated or stimulated with TNF (T) or PMA plus ionomycin (P) for 30 min. (c) NF-κB luciferase assay (mean and s.d., n = 3) using Jurkat cells transfected with empty vector (Vec), Flag-IKKβ with SSAA mutation (SSAA) or SSEE mutation (SSEE) plasmids together with a 5 × Ig κB sites-driven luciferase reporter gene (left). Representative immunoblotting of the cytosolic (C) and nuclear (N) subcellular fractions derived from Jurkat cells overexpressing the indicated Flag-IKKβ constructs is shown. Hsp90 and PARP served as cytosolic and nuclear markers, respectively (panels b and c, right). (d) Jurkat cells overexpressing indicated Flag-IKKβ constructs were analyzed by confocal microscopy following fixation and staining for RPS3, Flag, and nuclei. Percentage (mean and s.d.) of cells (n > 200) with nuclear RPS3 with or without Flag-IKKβ expression was quantified. Data are representative of at least 200 cells from three (a, d), two (b), and four (c) independent experiments, respectively.

Figure 3. Importin-α-mediated nuclear translocation of RPS3 is IκBα degradation dependent

(a) Whole-cell lysates (Input) from Jurkat cells stimulated as indicated were directly immunoblotted, or after immunoprecipitation (IP) with RPS3 antibody, for importin-α imp-α, importin-β imp-β, or RPS3. (b) Immunoprecipitation (IP)/immunoblot of the association of the endogenous importin-α imp-α or importin-β imp-β to RPS3 in Jurkat cells overexpressing either wild-type or SSAA mutant HA-IκBα and stimulated with TNF as indicated. (c) Jurkat cells were transfected with nonspecific (NS) or IκBα siRNA. 72 h later, whole-cell lysates (Input) were immunoblotted directly or after immunoprecipitation (IP) with RPS3 antibody for indicated proteins. (d) Immunoblotting of cytosolic (Cyto) and nuclear (Nuc) subcellular fractions derived from Jurkat cells transfected with scrambled nonspecific (NS), or IκBα siRNAs. Hsp90 and PARP served as cytosolic and nuclear markers and loading controls, respectively. (e) Jurkat cells were pretreated with (+) or without (-) sodium pervanadate (Pv, 800 mM) for 2 h followed by a 30-min TNF stimulation. Whole-cell lysates (Input) were immunoblotted directly or after immunoprecipitation (IP) with anti-RPS3 antibody for indicated proteins. Data are representative of at least of two experiments.

Figure 4. IKKβ phosphorylates RPS3 at serine 209

(a) Autoradiograph (top) and Coomassie blue staining (bottom) of in vitro kinase assays performed with recombinant GST or GST-RPS3 protein using recombinant human IKKα (rIKKα) or IKKβ (rIKKβ) as kinases. The autophosphorylated IKKs (p-IKKα and p-IKKβ) and phosphorylated RPS3 (p-RPS3) are labeled, respectively. (b) In vitro kinase assays were performed using recombinant RPS3 protein with or without rIKKβ as kinases. After digestion, the phosphorylated peptides were enriched by TiO₂ and fragmented by mass spectrometer. MS/MS of the 1+ fragment ion displays indicative of phosphorylated KKPLPDHVpSIVEPKD based on a Mascot algorithm database search. The y6 ion (in red) shows the incorporation of the site of phosphorylation, which is further confirmed by the loss of H₃PO₄ from several ions (bottom). Schematic diagram of RPS3 with characterized domains (NLS, nuclear localization signal; KH, K homology) and the IKKβ phosphorylation site serine 209 highlighted in red (top). (c) Autoradiograph (top) and Coomassie blue staining (bottom) of in vitro kinase assays performed with recombinant wild-type or S209A mutant GST-RPS3 proteins using recombinant human IKKβ. The phosphorylated (p-RPS3), total GST-RPS3 proteins, and autophosphorylated IKKβ (p-IKKβ) are labeled, respectively. (d) Immunoprecipitation (IP)/immunoblot of phosphoserine for ectopically expressed RPS3 in HEK 293T cells, where wild-type or S209A mutant Flag-RPS3 was transfected with or without an IKKβ plasmid. (e) Immunoblot of phospho-RPS3 (p-RPS3) and RPS3 in whole cell lysates derived from Jurkat cells stimulated with TNF for the indicated periods (bottom). Densitometry of all bands was performed, and the intensity of each p-RPS3 band was normalized to corresponding RPS3 band. The fold change in p-RPS3/RPS3 ratio was further normalized to the unstimulated sample (top). Data are representative of four (a, d), and two (b, c, d) independent experiments, respectively.

Figure 5. Phosphorylation of RPS3 at serine 209 is critical for its nuclear translocation and NF-κB specifier function

(a) Immunoblotting of the cytosolic (Cyto) and nuclear (Nuc) subcellular fractions derived from Jurkat cells overexpressing wild-type or S209A mutant Flag-RPS3 and stimulated with PMA plus ionomycin (PMA+I). Hsp90 and PARP were the cytoplasmic and nuclear markers, respectively (left). Densitometry of all bands was performed, and the intensity of each RPS3 band was normalized to loading hsp90 or PARP controls. The percentage of relative RPS3 was further normalized to the 0-min samples (set as 100%) in cells stimulated with or without PMA+I (right). (b) Immunoblotting of the cytosolic (Cyto) and nuclear (Nuc) subcellular fractions derived from Jurkat cells overexpressing IKKβ together with wild-type or S209A mutant Flag-RPS3 (left). Densitometry of all bands was performed and normalized as in (a), except further normalized to the control samples without overexpressing IKKβ (right). (c) Immunoblotting for endogenous and ectopically expressed RPS3 proteins in whole-cell lysates from Jurkat cells transfected with siRNA specifically targeting the 3′ untranslated region of RPS3 mRNA (RPS3-3′ UTR) or scrambled nonspecific siRNA (NS), plus wild-type (WT) or S209A mutant (S209A) Flag-RPS3 constructs. β-actin serves as a loading control. (d) NF-κB luciferase assay (mean and s.d., n = 3) of Jurkat cells transfected with siRNA specifically targeting the 3′ untranslated region of RPS3 mRNA (RPS3-3′ UTR) or scrambled nonspecific (NS) siRNA, together with either wild-type (WT) or S209A mutant (S209A) Flag-RPS3 constructs. A 5 × Ig κB site-driven luciferase construct was used as the reporter gene. Not statistically significant (NS), ** P < 0.01, calculated by Student's t-test. (e) Jurkat cells were left untreated (-) or stimulated (+) with PMA plus ionomycin (PMA+I) for 1 h, after transfection with siRNAs and Flag-RPS3 constructs as in (d). Cell extracts were analyzed by chromatin immunoprecipitation assays for the recruitment of Flag-RPS3 and endogenous p65 proteins to the promoters of NFKBIA, IL8, or ACTB, using Flag and p65 antibodies, respectively. Flag-RPS3- and p65-bound DNA was analyzed by quantitative real-time PCR in triplicate (primers, above diagrams) and normalized to input DNA, compared to cells transfected with a WT Flag-RPS3 construct and left untreated. Data are representative of at least of two independent experiments.

Figure 6. The bacterial effector protein NleH1 blocks RPS3 S209 phosphorylation

(a) 293T cells were transfected with control VN-HA or NleH1-HA plasmids and whole cell lysates were derived and immunoblotted for HA and β-actin as a loading control. (b) NF-κB luciferase assay (mean and s.d., n = 3) using 293T cells transfected with control VN-HA or NleH1-HA plasmids together with a 5 × Ig κB site-driven luciferase reporter gene. (c) 293T cells overexpressing either vehicle protein (HA) or NleH1-HA were stimulated with (+) or without (-) 50 ng/ml of TNF for 15 min. The derived whole cell lysates were immunoblotted for S209 phosphorylated RPS3 (p-RPS3) and indicated proteins. β-actin served as a loading control. (d) HeLa cells were left uninfected (Uninf) or infected for 3 h with wild-type (WT) E. coli O157:H7 or strains with isogenic deletions in the escN (ΔescN) or nleH1 (ΔnleH1) genes, followed by TNF treatment for the indicated periods. Whole cell lysates were extracted and immunoblotted with antibodies specific for normal RPS3 or S209 phosphorylated RPS3 (left). Densitometry of all bands was performed, and the intensity of each p-RPS3 band was normalized to corresponding RPS3 band. The fold change of p-RPS3/RPS3 was further normalized to the 0-min samples (set as 1.0) in cells infected with the indicated E. coli O157:H7 strains (right). (e) Transcript abundance relative to uninfected cells assessed by RT-PCR analysis of HeLa cells infected for 3 h with E. coli O157:H7 strains as in (d). The relative mRNA abundance of IL8, TNFAIP3, and NFKBIA were normalized to GAPDH expression (mean and s.d., n = 3). (f) Immunohistochemistry for S209 phosphorylated RPS3 in paraffin-embedded piglet colons derived from gnotobiotic piglets infected with E. coli O157:H7 EDL933 strains possessing (WT) or lacking NleH1 (ΔnleH1), using phospho-RPS3 antibody and 3,3′-diaminobenzidine as a substrate (brown). Nuclei were counterstained with hematoxylin (blue). Size bar represents 25 μm. Representative images from two piglets are shown. Data are representative of two (a, c), four (b), three (d, e), and six (f) independent experiments, respectively.

Figure 7. NleH1 alters the substrate specificity of IKKβ to block IKKβ-mediated RPS3 phosphorylation

(a) Autoradiograph (left) and Coomassie blue staining (right) of in vitro kinase assays performed with recombinant His-NleH1 or His-NleH1 K159A. The autophosphorylated NleH1 (p-NleH1) and total NleH1 proteins are labeled, respectively. (b) HeLa cells overexpressing vehicle protein, wild-type (HA-NleH1) or kinase-dead [HA-NleH1 (K159A)] proteins were stimulated with (+) or without (-) 50 ng/ml of TNF for 30 min. The derived whole cell lysates were immunoblotted for S209 phosphorylated RPS3 (p-RPS3) and RPS3 (top). Densitometry of all bands was performed, and the intensity of each p-RPS3 band was normalized to corresponding RPS3 band. The fold change in p-RPS3/RPS3 ratio was further normalized to the unstimulated sample (bottom). (c) HeLa cells were left uninfected or infected for 3 h with either wild-type (WT) C. rodentium, a strain deleted for nleH (ΔnleH) genes, or ΔnleH strains complemented with wild type NleH1 (ΔnleH/nleH1) or kinase dead NleH1 [ΔnleH/nleH1 (K159A)], followed by TNF treatment for 30 min. Whole cell lysates were extracted and immunoblotted with antibodies specific for normal RPS3 or S209 phosphorylated RPS3 (p-RPS3) (top). Densitometry of all bands was performed, and the intensity of each p-RPS3 band was normalized to corresponding RPS3 band. The fold change of p-RPS3/RPS3 was further normalized to the untreated samples (set as 1.0) in cells without infection. (d) HeLa cells were transfected with indicated IKKβ constructs. In 48 h, cells were left uninfected (Mock) or infected for 3 h with wild type (WT) E. coli O157:H7, ΔescN, or ΔnleH1 strains. Nuclear proteins were extracted and immunoblotted for RPS3 and PARP (top panel). Densitometry of all bands was performed, and the intensity of each RPS3 band was normalized to corresponding PARP band. The relative fold change of nuclear RPS3 was further normalized to the mock infection samples, set as 1.0 (right). (e) Autoradiograph (left) and Coomassie blue staining (right) of in vitro kinase assays performed with recombinant RPS3 or GST-IκBα (1-54) proteins as substrates and recombinant NleH1 or human IKKβ as kinases. The phosphorylated and total proteins are labeled as indicated. Data are representative of at least of two experiments.

Figure 7. NleH1 alters the substrate specificity of IKKβ to block IKKβ-mediated RPS3 phosphorylation

(a) Autoradiograph (left) and Coomassie blue staining (right) of in vitro kinase assays performed with recombinant His-NleH1 or His-NleH1 K159A. The autophosphorylated NleH1 (p-NleH1) and total NleH1 proteins are labeled, respectively. (b) HeLa cells overexpressing vehicle protein, wild-type (HA-NleH1) or kinase-dead [HA-NleH1 (K159A)] proteins were stimulated with (+) or without (-) 50 ng/ml of TNF for 30 min. The derived whole cell lysates were immunoblotted for S209 phosphorylated RPS3 (p-RPS3) and RPS3 (top). Densitometry of all bands was performed, and the intensity of each p-RPS3 band was normalized to corresponding RPS3 band. The fold change in p-RPS3/RPS3 ratio was further normalized to the unstimulated sample (bottom). (c) HeLa cells were left uninfected or infected for 3 h with either wild-type (WT) C. rodentium, a strain deleted for nleH (ΔnleH) genes, or ΔnleH strains complemented with wild type NleH1 (ΔnleH/nleH1) or kinase dead NleH1 [ΔnleH/nleH1 (K159A)], followed by TNF treatment for 30 min. Whole cell lysates were extracted and immunoblotted with antibodies specific for normal RPS3 or S209 phosphorylated RPS3 (p-RPS3) (top). Densitometry of all bands was performed, and the intensity of each p-RPS3 band was normalized to corresponding RPS3 band. The fold change of p-RPS3/RPS3 was further normalized to the untreated samples (set as 1.0) in cells without infection. (d) HeLa cells were transfected with indicated IKKβ constructs. In 48 h, cells were left uninfected (Mock) or infected for 3 h with wild type (WT) E. coli O157:H7, ΔescN, or ΔnleH1 strains. Nuclear proteins were extracted and immunoblotted for RPS3 and PARP (top panel). Densitometry of all bands was performed, and the intensity of each RPS3 band was normalized to corresponding PARP band. The relative fold change of nuclear RPS3 was further normalized to the mock infection samples, set as 1.0 (right). (e) Autoradiograph (left) and Coomassie blue staining (right) of in vitro kinase assays performed with recombinant RPS3 or GST-IκBα (1-54) proteins as substrates and recombinant NleH1 or human IKKβ as kinases. The phosphorylated and total proteins are labeled as indicated. Data are representative of at least of two experiments.