6 ± 1.7). Interestingly, in cells infected with E22ΔfliC for 2 h, IκB-α levels (13.7 ± 1.8) were lower than during E22 WT infection. However, at 4 h of infection with E22ΔfliC, IκB-α levels were higher (16.7 ± 0.2) than in cells infected see more with E22 WT (Fig. 5A, B). This indicates that both EPEC strains (E2348/69 and E22) provoke a strong and prolonged activation of NF-κB. E22 flagellum appeared to be required

to sustain the degradation of IκB-α at later stages of infection. To corroborate NF-κB activation, we also performed WB analysis of total and phosphorylated IκB-α (Fig. 5C). In mock-infected cells, we detected a clear and marked band of IκB-α (normalized band intensity value of 0.306 ± 0.016), but only a faint band of phosphorylated IκB-α (0.135 ± 0.40). In cells treated with HB101, no significant changes in phosphorylation of IκB-α (0.136 ± 0.033 at 2 h, and 0.129 ± 0.021 at 4 h) or IκB-α total levels (0.312 ± 0.054 at 2 h, and 0.315 ± 0.076 at 4 h) were detected. However, EPEC E2348/69 infection produced an intense IκB-α phosphorylation at 4 h (1.577 ± 0.117). This effect was accompanied by almost complete IκB-α

degradation (0.080 ± 0.070), indicating that all the remaining IκB-α was phosphorylated and markedly detected by the polyclonal anti-phospho-IκB-α antibody. However, at 2 h post-infection, only the degradation of IκB-α (0.232 ± 0.036) was observed, but no phosphorylation. During E22 WT infection, the degradation of IκB-α was not significantly different at 2 h of infection (0.389 ± 0.137); however, at 4 h, IκB-α Z VAD FMK degradation was lower (0.235 ± 0.038). p-IκB-α was clearly present already at 2 h (1.370 ± 0.076) Ureohydrolase and remained at 4 h (0.618 ± 0.043). These results confirm that E2348/69 as well as E22 infection promotes IκB-α phosphorylation and degradation. Since IκB-α phosphorylation and degradation are coupled, we only analysed IκB-α degradation in cells infected with E22 Δeae, ΔescN, ΔespA and ΔfliC mutants for 4 h (Fig. 5D). Contrary to the effect caused by E22 WT (0.235 ± 0.038), infection

with the intimin mutant did not induce IκB-α degradation (0.589 ± 0.238), and this value was higher than in mock-infected cells (0.306 ± 0.016). However, E22ΔescN, E22ΔespA and E22ΔfliC mutants induced lower IκB-α degradation than E22 WT strain (E22ΔescN: 0.289 ± 0.008, E22ΔespA: 0.278 ± 0.010 and E22ΔfliC: 0.275 ± 0.011). These data indicate that whereas T3SS and flagellin were confirmed to be implicated in the full activation of NF-κB, intimin decreases the activation of NF-κB. To understand the relationship between NF-κB and the activation of ERK1/2 with synthesis and secretion of proinflammatory cytokines during EPEC infection (for 4 h), we determined il-1β, il-8 and tnf-α expression by RT-PCR. Mock-infected cells expressed il-1β (Fig. 6A) and il-8 (Fig. 6C) mRNA (normalized intensity of the products: 0.680 ± 0.181 for il-1β and 0.593 ± 0.111 for il-8), but tnf-α mRNA was not detected in mock-infected cells (Fig. 6E).