H2O2-induced NF-κB activation in T cells: the IKK-dependent pathway comes back into fashion

Read Also

Oxidant-induced signalling pathways have been intensely studied in T cell lines for many reasons. First, T cells are often submitted to ROS during inflammatory response, which can, in turn, influence a number of signalling pathways. For example, at a site of inflammation, H₂O₂ is produced by activated macrophages and neutrophils at an estimated rate of 2-6 x 10^-4 µM/h per cell and T cells may be exposed to 10-100 µM H₂O₂ in a physiological environment [15]. Secondly, it is now clear that the activation of T cells through their antigen receptors increases the level of intracellular ROS that, instead of being toxic, can actually play a positive role in controlling signalling pathways that lead to T cell proliferation [37]. Thirdly, T cell apoptosis is clearly regulated by ROS [38]. For example, a recent study in Jurkat leukemic cells has shown that NF-kB activation by H₂O₂ induces Bfl-1, which, in turn, attenuates Fas-mediated apoptosis [39]. Moreover, some compounds used in anti-leukemic chemotherapies induce cell death through ROS generation [40, 41]. For all of these reasons, understanding NF-κB activation mechanism by ROS in T cell was of importance. Until recently, all the works concerning NF-κB activation by ROS in T cells have highlighted an atypical mechanism of activation totally distinct from those triggered by pro-inflammatory cytokines. It involves phosphorylation of the inhibitor IkBa  on tyrosine 42 rather than the classical serines 32 and 36 by the IKK complex. This was true in murine T lymphocytes [42] and in human Jurkat T cells [43, 44]. Furthermore, the IκBα degradation mechanism appears to be proteasome-independent, but instead relies on a calpain-mediated digestion after phophorylation on S/T in the so-called PEST sequence of the inhibitor [42]. NF-κB activation induced by tyrosine phosphorylation of IκBα was also observed after pervanadate (a potent tyrosine phosphatase inhibitor) and hypoxia/reoxygenation treatment [44, 45]. This can occur in the absence of IκBα degradation; in this case, a dissociation mechanism from NF-kB has been described [46]. The discovery of the terminal tyrosine kinase that phosphorylates IκBα Y42 has been a challenge for many years. Livolsi et al. first demonstrated that the TCR-associated tyrosine kinases p56Lck and ZAP-70 were required for pervanadate-induced IκBα tyrosine phosphorylation, without showing that these kinases indeed phosphorylate IκBα directly [44]. Recently, Takada et al. reported that Syk tyrosine kinase was required for H₂O₂-induced IκBα tyrosine phosphorylation and NF-κB activation, and was capable of phosphorylating IκBα in vitro, suggesting that Syk may be the terminal tyrosine kinase responsible for IκBα tyrosine phosphorylation [43]. Our group has recently called this “Y42 paradigm” into question by studying the H₂O₂-induced NF-κB activation mechanism in T cells other than Jurkat cells, namely CEM and Jurkat JR (also termed Wurzburg). Unexpectedly, micromolar amounts of H₂O₂ were shown indeed capable of inducing IKK activation in these cell lines, leading to a classical IκBα phosphorylation on Ser32 and 36 [47]. No tyrosine phosphorylation was observed in this case. However, pervanadate treatment still induced a strong tyrosine phosphorylation of IκBα, suggesting that NF-κB activation mechanisms by H₂O₂ and pervanadate are different, at least in CEM and Jurkat JR cells [47]. In fact, the differences between Jurkat versus CEM and Jurkat JR cells in terms of oxidant-induced NF-κB activation mechanism relied on the expression of the SHIP-1 protein. SHIP-1, a lipid phosphatase, acts by dephosphorylating the membrane-bound PtdIns(3,4,5)P₃, generated by PI3Kinase, and has thus been described as a negative regulator of immune receptor, cytokine and growth factor receptor signalling [48]. Furthermore, SHIP-1 can interact with a large number of proteins via its SH2 and NPXY containing domains, thus influencing numerous signalling pathways [48]. It is now well known that Jurkat cells are deficient of SHIP-1 expression at the protein level, but that CEM cells express the protein normally [47, 49], which can, in turn, influence a number of signalling pathways [50]. The rescuing of Jurkat cells with SHIP-1 clearly made them shift to a classical mechanism dependent on IKK activation and phosphorylation of IκBα on serines 32 and 36 upon H₂O₂ stimulation. Furthermore, a less pronounced tyrosine phosphorylation of IκBα was observed in this case (Figure 3) [47]. As mentioned above, this observation was also made in Jurkat JR cells which is more sensitive to oxidant-induced NF-κB activation than the parental cell line Jurkat [19, 51], and expressed SHIP-1 normally. The analysis of the NF-κB activation pathway upon oxidative stress treatment in that cell-type also revealed an IKK-dependent mechanism [47]. This observation could explain why NF-κB activation might be more rapid and important in that subclone than in Jurkat cells, as observed by several authors [51, 52]. All this data clearly suggests that the atypical NF-κB activation pathway described in Jurkat cells treated by oxidative stress is only available in that cell type. NF-κB activation in other T cell lines is the classical IKK-dependent mechanism that rely on SHIP-1 (Figure 3). The tyrosine-phosphorylation mechanism is probably a rescue pathway adopted by SHIP-1 negative cells. The exact mechanism by which SHIP-1 acts to activate the IKK complex has still to be delineated. Both phosphatase and SH2 domains of SHIP-1 seem to be crucial in this process, but considerable work has yet to be carried out to find out the exact role of that protein in NF-κB redox regulation.