The BH3-only protein Bad is dispensable for TNF-mediated cell death

Tumor necrosis factor (TNF) is a pleiotropic cytokine critical for inflammation during innate and adaptive immune responses, orchestrating cytokine expression via activation of the canonical NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) signaling cascade. TNF can also trigger cell death by engaging TNF receptor I (TNFRI) along apoptotic or necroptotic pathways, which is believed to limit or exacerbate inflammatory responses, respectively.1 Excessive TNF release during fulminant bacterial infection can lead to a cytokine storm, vascular leakage driven by TNF-mediated endothelial damage, septic shock and death.2 The pro-death and inflammatory potential of TNF therefore need to be tightly regulated and cells have developed highly specialized means to control the biological outcome of TNFRI activation. Recent studies define a delicate network of linear and branched ubiquitylation/deubiquitylation-controlled signaling events downstream of TNFRI as critical for the decision between life and death upon TNF stimulation.3, 4 Its pro-death function is thereby constrained by different means but most prominently by NF-κB-dependent transcription of survival genes, such as Bcl-x, XIAP (X-linked inhibitor of apoptosis protein), cIAPs (inhibitor of apoptosis proteins) or c-FLIP (cellular FLICE inhibitory protein), a dimerization and activation partner of procaspase-8 that is rate limiting for TNF-driven apoptosis and also prevents RIPK1/RIPK3-dependent necroptosis that can ensue when caspase-8 is inhibited.5 To promote death, TNFRI signaling can also engage pro-apoptotic members of the Bcl-2 family, for example, the BH3-only proteins Bim, Bid, Puma and Bad (Bcl-2-antagonist of cell death). Bid is processed and thereby activated by caspase-8 in many cell types and systems. Bim activity, on the other hand, appears to be controlled by TNF-driven JNK-mediated phosphorylation and both, Bid and Bim, contribute to T-cell-driven TNF-dependent liver toxicity.6 TNF has also been shown to trigger NF-κB-dependent transcriptional activation of Puma and deficiency of this BH3-only protein can also provide protection from hepatocyte killing upon systemic TNF administration.7 Together these findings show that mitochondrial apoptosis, controlled by the Bcl-2 family, can be a critical event in pathological hepatocyte cell death in response to TNF. Bad phosphorylation on residues Ser-112, -136, -155, has previously been shown to prevent mitochondrial translocation of Bad and restrain its pro-death function (reviewed in Danial8). More recently, Yan et al.9 showed that the activity of the BH3 domain-containing protein Bad is repressed downstream of TNFRI activation by direct IKK-mediated phosphorylation on Ser-26. This study showed that mouse embryonic fibroblasts (MEFs), derived from Ikbkb−/− mice, that are deficient in IKKβ/IKK2, where Bad expression was also repressed by RNA interference, or MEFs from Bad−/− mice that were pre-treated with the IKK inhibitor PS-1145, showed reduced caspase-3/7 activity, delayed PARP cleavage and strongly reduced cell death upon TNF treatment. Similar observations were made in primary thymocytes and hepatocytes, pretreated with IKK-inhibitor prior to TNF stimulation. Most strikingly, Bad-deficient animals were protected from the lethal effects of hepatitis elicited by D-GalN priming and subsequent TNF injection. Together, these data provided a strong evidence for an NF-κB-independent anti-apoptotic function of the IKK complex, explaining, at least in part, the increased cell death susceptibility of IKKβ-deficient MEF over those lacking the NF-κB family members RelA plus cRel9 and supported a rate-limiting role of Bad in TNF-induced cell death signaling. Intrigued by an earlier report describing the generation and phenotype of Bad−/− mice that reported normal cell death responses upon TNF treatment,10 we decided to reinvestigate the role of Bad in TNF-driven cell death ex vivo and in fulminant hepatitis.