Figure 1.
Effects of CAPE on IbeA+ E. coli K1-induced NF-κB activation and pathogenicities in vitro and in vivo.
(A) IbeA+ E. coli K1 induced NF-κB activation in HBMECs was suppressed by CAPE. HBMECs were incubated with or without the NF-κB inhibitor CAPE (25 µM) for 30 min before stimulation with E44 or ZD1 (107/mL). IKK α/β phosphorylation (p-IKK α/β) in cytoplasmic fractions and NF-κB (p65) in nuclear fractions was examined after 2 h of stimulation with E. coli strains. The β-actin in both fractions was detected as internal loading controls. CON, control without E. coli stimulation. (B–C) Effects of CAPE (0–25 µM) on IbeA+ E. coli K1 penetration and PMN transmigration across HBMECs were examined. HBMECs were incubated with various concentrations of CAPE for 1 h before the invasion and PMN transmigration assays. (B). E. coli (107 CFU) were added to the HBMEC monolayers after CAPE treatment. Invasion assays were carried out as described in the Materials and Methods. (C) The CAPE-pretreated HBMECs were stimulated with E. coli (106 CFU) in the lower chamber for 2 h and incubated with PMN (106) in the upper chamber at 37°C for another 4 h. All assays were performed in triplicates. Results for invasion are expressed as relative invasion compared to the positive control without drug treatment (100%). Results for PMNT are expressed as the percentage of leukocyte transmigration of the total added. Both the invasion and PMNT assays were done with E44 (black column) and ZD1 (white column). E. coli meningitis was induced in neonatal mice with or without CAPE treatment (n = 5) as described in Methods and Materials. (D) Recruitment of PMN into the CSF; (E) Flux of albumin into the CNS; and (F) Levels of soluble NF-κB (p65) in CSF. The significant differences with regard to the controls without CAPE treatment were marked by asterisks (*P<0.05; **P<0.01).
Figure 2.
Inhibition of IbeA+ E. coli-induced IKK phosphorylation and NF-κB activation by MEK/ERK inhibitors.
(A) HBMECs were incubated with or without PD098059 (50 µM) for 60 min before stimulation with E44 or ZD1 (107/ml). (B) HBMECs were incubated with or without ERK89 (vimentin-binding domain, 25 µg/ml) and ERK312 (control peptide, 25 µg/ml) for 60 min before infection with E44 or ZD1 (107/ml). In both (A) and (B), ERK1/2 phosphorylation (p-Erk1/2), IKK α/β phosphorylation (p-IKK α/β) and IκBα degradation were examined in cytoplasmic fractions after 30 min of stimulation with E. coli K1 strains. NF-κB (p65) translocation to the nucleus was examined in nuclear fractions after 2 h of infection with E. coli K1 strains. β-actin in both fractions was detected as internal loading controls. CON, control without bacterial stimulation.
Figure 3.
Role of vimentin in IbeA+ E. coli K1-induced NF-κB activation.
(A) Immunofluorescence microscopy was used to examine the correlation between vimentin reorganization and NF-κB translocation to the nucleus after 2 h of stimulation with IbeA protein (0.1 µg/ml), E44 or ZD1 (25 MOI). HBMECs were triple-stained with the V9 antibody against vimentin conjugated to FITC (green), the rabbit antibody against NF-κB (p65) conjugated to rhodamine (red), and DAPI (blue). The merged images are shown in the right-hand panels (Merge). Arrows indicated cells with colocalization of vimentin and NF-κB (p65) Scale bar, 50 µm. (B) Blockage of IbeA+ E. coli K1-induced NF-κB activation in HBMECs by siRNA-mediated knockdown of vimentin. HBMECs were transfected with vimentin or control siRNA as described in Materials and Methods. After 24 h incubation, the cells were treated with E44 or ZD1 (107/ml) for 30 min or 2 h. Vimentin (VIM), α7 nAChR, ERK1/2 phosphorylation (p-Erk1/2), IKK α/β phosphorylation (p-IKK α/β), IκBα degradation, and PSF re-localization were examined in cytoplasmic fractions after 30 min of stimulation with E. coli K1 strains. NF-κB (p65) translocation to the nucleus was examined in nuclear fractions after 2 h of incubation with E. coli K1 strains. β-actin in both fractions was detected as internal loading controls. Control: HBMECs transfected with control siRNA; VIM KD: HBMECs transfected with vimentin siRNA; UNT: Untreated HBMECs.
Figure 4.
Time course analysis of IbeA-induced cytoplasmic activation and nuclear translocation of NF-κB.
HBMECs ware incubated with the IbeA protein (0.1 µg/ml) for 2, 6, and 24 h, respectively, and then the cytoplasmic and nuclear fractions were extracted. The cytoplasmic fractions were immunoprecipitated (IP) with the V9 anti-vimentin antibody and the rabbit anti-NF-κB (P65) antibody as described in Materials and Methods. The cytoplasmic and nuclear lysates (A), vimentin Co-IP complexes (B), and NF-κB (p65) Co-IP complexes (C) were subjected to western blot using the antibodies as described in Materials and Methods. CON: the IP control without primary antibodies incubation; 0 h: the control HBMECs without IbeA stimulation.
Figure 5.
Effects of vimentin head domain deletion on IbeA-induced NF-κB activation and interaction with β-tubulin.
(A) The cytoplasmic fractions of the GFP–VRT, GFP-VH and GFP transductants were extracted and immunoprecipitated (IP) using the mouse anti-GFP antibody as described in Materials and Methods. The GFP-IP complexes were subjected to Western blotting using the rabbit polyclonal antibodies against GFP, NF-κB (P65), and β-tubulin. Band a, GFP–VRT (72 kDa); band b, GFP-VH (37 kDa); band c, GFP (27 kDa); band d, NF-κB (P65), (65 kDa); and band e, β-tubulin, (50 kDa). (B) Immunofluorescence images of the GFP–VRT and GFP transductants incubated with or without the IbeA protein (0.1 µg/ml) for 2 h. The cells were double-stained with the rabbit antibody against NF-κB (p65) conjugated to rhodamine (red), and DAPI (blue). Arrows indicate cells with NF-κB (P65) translocation to the nucleus, which was increased in the GFP transductants and reduced in GFP-VRT-transduced HBMECs upon stimulation with IbeA. Scale bar, 50 µm. (C) Western blot of the transduced HBMECs treated with the IbeA protein (0.1 µg/ml). ERK1/2 phosphorylation (p-Erk1/2), IKK α/β phosphorylation (p-IKK α/β), IκBα degradation, vimentin (VIM), GFP and PSF re-localization were examined in cytoplasmic fractions after 30 min of IbeA stimulation. NF-κB (p65) translocation to the nucleus was examined in nuclear fractions after 2 h of IbeA incubation. β-actin in both fractions was detected as internal loading controls.
Figure 6.
β-tublulin is required for IbeA+ E. coli K1-induced NF-κB activation.
(A) IbeA− and IbeA+ E. coli K1-induced β-tubulin/vimentin clustering and colocalization. Immunofluorescence microscopy was used to examine the clustering and reorganization of vimentin and β-tubulin after 2 h of incubation with the IbeA protein (0.1 µg/ml), E44 or ZD1 (25 MOI). HBMECs were triple-stained with the V9 antibody against vimentin conjugated to FITC (green), the rabbit antibody against β-tubulin conjugated to rhodamine (red), and DAPI (blue). The merged images are shown in the right-hand panels (Merge). Arrows indicated cells with colocalization between vimentin and β-tubulin around the perinuclear region. Scale bar, 50 µm. (B) Blockage of IbeA+ E. coli K1-induced cytoplasmic activation and nuclear translocation of NF-κB (p65) in HBMECs by the microtubule inhibitors. HBMECs were incubated with or without colchicines (Col, 2 µM), vincristine (Vin, 1 µM), nocodazole (Noc, 25 µg/ml) for 60 min before stimulation with E44 or ZD1 (107/ml). Phosphorylation of ERK1/2 (p-Erk1/2) and IKK α/β (p-IKK α/β) was examined in cytoplasmic fractions after 30 min of E. coli K1 treatment. NF-κB (p65) translocation to nucleus in nuclear fractions was examined after 2 h of E. coli K1 incubation. β-actin in both fractions was detected as internal loading controls. CON, control without bacterial stimulation.
Figure 7.
Inhibition of IbeA+ E. coli K1-induced pathogenicities, phosphorylation of ERK/IKK and nuclear translocation of NF-κB by knockdown of PSF.
HBMECs were transfected with PSF or control siRNA as described in Materials and Methods. IbeA+ E. coli K1 penetration (A) and PMN transmigration (B) across siRNA-transfected HBMECs were performed as described in the Materials and Methods. Both invasion and PMN transmigration assays were performed in triplicates. Results for invasion are expressed as a relative percentage compared to the penetration rate of E44 in the siRNA control (CON) (100%). Results for PMN transmigration are expressed as the percentage of PMN transmigration of total PMNs. The control siRNA-transfected HBMECs infected with E44 and ZD1 were used as the controls (panels A and B). The significant differences regarding to the control were marked by asterisks (*P<0.05; **P<0.01). (C) After transfection, the cells were stimulated with E44 or ZD1 (107/ml) for 30 min or 2 h. PSF re-localization, p-Erk1/2, p-IKK α/β and IκBα degradation were examined in cytoplasmic fractions after 30 min of stimulation with E. coli K1 strains. NF-κB (p65) and PSF in nuclear fractions were examined after 2 h of incubation with E. coli K1 strains. β-actin in both fractions was detected as internal loading controls. Control: HBMECs transfected with control siRNA; PSF KD, HBMECs transfected with PSF siRNA; UNT: Untreated HBMECs. (D) Time course analysis of IbeA-induced tyrosine phosphorylation of PSF. HBMECs were incubated with the IbeA protein (0.1 µg/ml) for 2, 6, and 24 hrs, respectively. The cytoplasmic fractions were extracted and immunoprecipitated (IP) using the anti-phosphotyrosine antibody as described in Materials and Methods. The Tyr-IP complexes were subjected to Western blot using the mouse monoclonal antibody against PSF. Total mouse IgG was detected as an internal loading control. CON: the IP control without primary antibody incubation; 0 h: the control HBMECs without IbeA incubation.
Figure 8.
Inhibition of IbeA+ E. coli K1-induced proteasomal degradation by knockdown of vimentin and VH domain deletion.
Immunoblotting analysis of polyubiquitinylated proteins (Ub-prs): (A) HBMECs with siRNA-mediated knockdown of vimentin; and (B) HBMECs transduced with the lentivirus constructs expressing GFP–VRT and GFP. In both (A) and (B), all the cells were incubated with or without E. coli K1 strains (E44 and ZD1, 107/ml) for 2 h. The Ub-prs were detected in cytoplasmic fractions to determine the proteasomal degradation as described in Materials and Methods. β-actin was used as an internal loading control. In all experiments, untreated HBMECs (UNT) were taken as controls.