In order to determine whether

In order to determine whether NF-κB could interact with GSTP1-1 promoter, we performed EMSA. One NF-κB like (−98κB) binding site was previously described in the GSTP1-1 promoter as a regulator element. However, few data about this site have been published and its role in the GSTP1-1 gene regulation remains unclear. Indeed, the sequence −98/−89 was shown to act as a repressor binding site in the human mammary carcinoma cell line MCF7 whereas it was not effective in the VCREMS cells, a MCF7 multidrug-resistant derivative cell line [32]. Authors showed that the silencing effect of this NF-κB like site occurred by interacting with the AP-1 site in MCF7 cells but not in the human SynaptoRedTM C2 carcinoma EJ cells [32], [33]. In contrast, Zhang et al. [65] provided evidence that this NF-κB site is not involved in silencing the TRE/ARE function in keratinocytes. On the other hand, Xia et al. [66] showed that response of GSTP1 to oxidants is mediated by a NF-κB like site while response to anti-oxidants is mediated by the AP-1 site. Thus, regulation of GSTP1 can result from cooperation between NF-κB like and AP-1 sites and the function of the −98/−89 sequence might be cell type and/or drug dependent. However, neither the capacity of the −98κB site to bind a NF-κB dimer nor the involvement of this transcription factor in the GSTP1-1 gene regulation had been demonstrated until now, particularly in leukemic cell lines in which the enzyme is largely expressed [9], [67]. By DNA sequence analysis [52] we identified a distal DNA sequence −323/−314 downstream from the minimal promoter as a putative NF-κB binding site. We then studied binding activity of both putative κB sites. EMSA experiments using K562 nuclear extracts and probes containing NF-κB consensus from the immunoglobulin κ light chain (C-κB) or the −323/−314 sequence (−323κB) exhibited similar patterns. Indeed, two binding complexes were induced by TNFα resulting from a specific interaction within the −323/−314 sequence as assessed by competition experiments with wild type (−323κB, C-κB) and mutated (−323κBM) probes. In contrast, we observed a single binding activity on the NF-κB like site (−98/−89) unaffected by TNFα treatments in K562 cells. Moreover, no binding of human recombinant NF-κB p50 protein to this probe was observed suggesting fundamental differences in the potential roles of −98/−89 and −323/−314 sequences. Additional studies are needed to demonstrate the importance of the sequences flanking the two NF-κB sites. NF-κB activation is regulated via a well known signaling pathway [13], [15], [16]. In order to characterize C1 and C2 bindings, we used MG-132 and BAY11-7082 which inhibit TNFα-induced NF-κB by blocking proteasome activity and phosphorylation of IκBα, respectively [61], [68]. The −323κB site was shown to bind TNFα-activated NF-κB dimers since both specific complexes (C1 and C2) were inhibited by MG132 and BAY11-7082. C1 and C2 complexes were then identified by super-shift experiments as p65/p65 and p50/p65 NF-κB dimers, respectively. In contrast, super-shift experiments did not reveal any binding of NF-κB dimers to the −98κB probe. In order to study the −323κB binding site we performed transfection experiments and so we could establish that its transactivation potential is similar to the one of the consensus site. Previous results clearly demonstrate that anti-inflammatory therapeutic agents block TNF-induced NF-κB activation, IκBα phosphorylation and activation of IKK, JNK, and AP-1, and suppressed TNF-induced apoptosis. Our results obtained by co-transfection of expression constructs coding for TNFR1, NIK, TRAF2, IKKβ, IκBα and p65/NF-κB specifically show that TRAF2, NIK, IκBα and NF-κB/p65 are involved in the regulated activation of the GSTP1-1 promoter. We show here for the first time that the GSTP1-1 promoter is regulated by a TNFR1, TRAF2, NIK pathway which leads to activation of NF-κB. The phosphorylation of IκBα is regulated by a large number of kinases, including IKK-α, IKK-β, IKK-γ, NIK, TGF-β-activated kinase-1, AKT, and mitogen-activated protein/extracellular signal-related kinases. However, only IKKβ mediates TNF-induced phosphorylation of IκBα at positions 32 and 36 which explains the strong negative effect of a dominant negative mutant of IκBα in our assays. On the other hand, a mutated form of the −396pGSTP promoter shows a strongly reduced luciferase activity if co-transfected with p65/NF-κB confirming the importance of NF-κB in the expression of GSTP1-1 gene. Nevertheless, this mutated construct maintains luciferase reporter gene activity which is potentially due to the presence of a consensus AP-1 site at −73. Our results confirm our previous reports describing the role of the −73 AP-1 site in constitutive and TPA-induced reporter gene activity [30].