Cell death is not essential for caspase-1-mediated interleukin-1ß activation and secretion.

Caspase-1 cleaves and activates the pro-inflammatory cytokine interleukin-1 beta (IL-1ß), yet the mechanism of IL-1ß release and its dependence on cell death remains controversial. To address this issue, we generated a novel inflammasome independent system in which we directly activate caspase-1 by dimerization. In this system, caspase-1 dimerization induced the cleavage and secretion of IL-1ß, which did not require processing of caspase-1 into its p20 and p10 subunits. Moreover, direct caspase-1 dimerization allowed caspase-1 activation of IL-1ß to be separated from cell death. Specifically, we demonstrate at the single cell level that IL-1ß can be released from live, metabolically active, cells following caspase-1 activation. In addition, we show that dimerized or endogenous caspase-8 can also directly cleave IL-1ß into its biologically active form, in the absence of canonical inflammasome components. Therefore, cell death is not obligatory for the robust secretion of bioactive IL-1ß.

Evolutionary divergence of the necroptosis effector MLKL.

The pseudokinase, MLKL (mixed-lineage kinase domain-like), is the most terminal obligatory component of the necroptosis cell death pathway known. Phosphorylation of the MLKL pseudokinase domain by the protein kinase, receptor interacting protein kinase-3 (RIPK3), is known to be the key step in MLKL activation. This phosphorylation event is believed to trigger a molecular switch, leading to exposure of the N-terminal four-helix bundle (4HB) domain of MLKL, its oligomerization, membrane translocation and ultimately cell death. To examine how well this process is evolutionarily conserved, we analysed the function of MLKL orthologues. Surprisingly, and unlike their mouse, horse and frog counterparts, human, chicken and stickleback 4HB domains were unable to induce cell death when expressed in murine fibroblasts. Forced dimerization of the human MLKL 4HB domain overcame this defect and triggered cell death in human and mouse cell lines. Furthermore, recombinant proteins from mouse, frog, human and chicken MLKL, all of which contained a 4HB domain, permeabilized liposomes, and were most effective on those designed to mimic plasma membrane composition. These studies demonstrate that the membrane-permeabilization function of the 4HB domain is evolutionarily conserved, but reveal that execution of necroptotic death by it relies on additional factors that are poorly conserved even among closely related species.

RIPK1- and RIPK3-induced cell death mode is determined by target availability.

Both receptor-interacting protein kinase 1 (RIPK1) and RIPK3 can signal cell death following death receptor ligation. To study the requirements for RIPK-triggered cell death in the absence of death receptor signaling, we engineered inducible versions of RIPK1 and RIPK3 that can be activated by dimerization with the antibiotic coumermycin. In the absence of TNF or other death ligands, expression and dimerization of RIPK1 was sufficient to cause cell death by caspase- or RIPK3-dependent mechanisms. Dimerized RIPK3 induced cell death by an MLKL-dependent mechanism but, surprisingly, also induced death mediated by FADD, caspase 8 and RIPK1. Catalytically active RIPK3 kinase domains were essential for MLKL-dependent but not for caspase 8-dependent death. When RIPK1 or RIPK3 proteins were dimerized, the mode of cell death was determined by the availability of downstream molecules such as FADD, caspase 8 and MLKL. These observations imply that rather than a 'switch' operating between the two modes of cell death, the final mechanism depends on levels of the respective signaling and effector proteins.

Necroptosis induced by RIPK3 requires MLKL but not Drp1.

Necroptosis is a mechanism by which cells can kill themselves that does not require caspase activity or the presence of the pro-apoptotic Bcl-2 family members Bax or Bak. It has been reported that RIPK3 (receptor interacting protein kinase 3) activates MLKL (mixed lineage kinase domain-like) to cause cell death that requires dynamin-related protein 1 (Drp1), because survival was increased in cells depleted of Drp1 or treated with the Drp1 inhibitor mdivi-1. To analyze necroptosis in a system that does not require addition of tumor necrosis factor (TNF), we used a construct that allows RIPK3 to be induced in cells, and then dimerized via an E. coli gyrase domain fused to its carboxyl-terminus, using the dimeric gyrase binding antibiotic coumermycin. We have previously shown elsewhere that RIPK3 dimerized in this manner not only induces necroptosis but also apoptosis, which can be inhibited by the broad-spectrum caspase inhibitor Q-VD-OPh (QVD). In response to RIPK3 dimerization, wild-type mouse embryonic fibroblasts (MEFs) underwent cell death that was reduced but not completely blocked by QVD. In contrast, death upon dimerization of RIPK3 in Mlkl(-/-) MEFs was completely inhibited with QVD, confirming that MLKL is required for necroptosis. Similar to wild-type MEFs, most Drp1(-/-) MEFs died when RIPK3 was activated, even in the presence of QVD. Furthermore, overexpression of wild-type MLKL or dominant active mutants of MLKL (Q343A or S345E/S347E) caused death of wild-type and Drp1(-/-) MEFs that was not inhibited with QVD. These results indicate that necroptosis caused by RIPK3 requires MLKL but not Drp1.

TNF can activate RIPK3 and cause programmed necrosis in the absence of RIPK1.

Ligation of tumor necrosis factor receptor 1 (TNFR1) can cause cell death by caspase 8 or receptor-interacting protein kinase 1 (RIPK1)- and RIPK3-dependent mechanisms. It has been assumed that because RIPK1 bears a death domain (DD), but RIPK3 does not, RIPK1 is necessary for recruitment of RIPK3 into signaling and death-inducing complexes. To test this assumption, we expressed elevated levels of RIPK3 in murine embryonic fibroblasts (MEFs) from wild-type (WT) and gene-deleted mice, and exposed them to TNF. Neither treatment with TNF nor overexpression of RIPK3 alone caused MEFs to die, but when levels of RIPK3 were increased, addition of TNF killed WT, Ripk1(-/-), caspase 8(-/-), and Bax(-/-)/Bak(-/-) MEFs, even in the presence of the broad-spectrum caspase inhibitor Q-VD-OPh. In contrast, Tnfr1(-/-) and Tradd(-/-) MEFs did not die. These results show for the first time that in the absence of RIPK1, TNF can activate RIPK3 to induce cell death both by a caspase 8-dependent mechanism and by a separate Bax/Bak- and caspase-independent mechanism. RIPK1 is therefore not essential for TNF to activate RIPK3 to induce necroptosis nor for the formation of a functional ripoptosome/necrosome.

In mouse embryonic fibroblasts, neither caspase-8 nor cellular FLICE-inhibitory protein (FLIP) is necessary for TNF to activate NF-κB, but caspase-8 is required for TNF to cause cell death, and induction of FLIP by NF-κB is required to prevent it.

Binding of TNF to TNF receptor-1 can give a pro-survival signal through activation of p65/RelA NF-κB, but also signals cell death. To determine the roles of FLICE-inhibitory protein (FLIP) and caspase-8 in TNF-induced activation of NF-κB and apoptosis, we used mouse embryonic fibroblasts derived from FLIP and caspase-8 gene-deleted mice, and treated them with TNF and a smac-mimetic compound that causes degradation of cellular inhibitor of apoptosis proteins (cIAPs). In cells treated with smac mimetic, TNF and Fas Ligand caused wild-type and FLIP(-/-) MEFs to die, whereas caspase-8(-/-) MEFs survived, indicating that caspase-8 is necessary for death of MEFs triggered by these ligands when IAPs are degraded. By contrast, neither caspase-8 nor FLIP was required for TNF to activate p65/RelA NF-κB, because IκB was degraded, p65 translocated to the nucleus, and an NF-κB reporter gene activated normally in caspase-8(-/-) or FLIP(-/-) MEFs. Reconstitution of FLIP(-/-) MEFs with the FLIP isoforms FLIP-L, FLIP-R, or FLIP-p43 protected these cells from dying when treated with TNF or FasL, whether or not cIAPs were depleted. These results show that in MEFs, caspase-8 is necessary for TNF- and FasL-induced death, and FLIP is needed to prevent it, but neither caspase-8 nor FLIP is required for TNF to activate NF-κB.

RIPK1 is not essential for TNFR1-induced activation of NF-kappaB.

On TNF binding, receptor-interacting protein kinase 1 (RIPK1) is recruited to the cytoplasmic domain of TNFR1, at which it becomes ubiquitylated and serves as a platform for recruitment and activation of NEMO/IKK1/IKK2 and TAK1/TAB2. RIPK1 is commonly thought to be required for the activation of canonical NF-kappaB and for inhibition TNFR1-induced apoptosis. RIPK1 has, however, also been reported to be essential for TNFR1-induced apoptosis when cIAPs are depleted. To determine the role of RIPK1 in TNF/IAP antagonist-induced death, we compared wild type (WT) and RIPK1(-/-) mouse embryonic fibroblasts (MEFs) treated with these compounds. On being treated with TNF plus IAP antagonist, RIPK1(-/-) MEFs survived, unlike WT MEFs, demonstrating a killing activity of RIPK1. Surprisingly, however, on being treated with TNF alone, RIPK1(-/-) MEFs activated canonical NF-kappaB and did not die. Furthermore, several cell types from E18 RIPK1(-/-) embryos seem to activate NF-kappaB in response to TNF. These data indicate that models proposing that RIPK1 is essential for TNFR1 to activate canonical NF-kappaB are incorrect.

Puma indirectly activates Bax to cause apoptosis in the absence of Bid or Bim.

Bcl-2 family members regulate apoptosis in response to cytokine withdrawal and a broad range of cytotoxic stimuli. Pro-apoptotic Bcl-2 family members Bax and Bak are essential for apoptosis triggered by interleukin-3 (IL-3) withdrawal in myeloid cells. The BH3-only protein Puma is critical for initiation of IL-3 withdrawal-induced apoptosis, because IL-3-deprived Puma(-/-) cells show increased capacity to form colonies when IL-3 is restored. To investigate the mechanisms of Puma-induced apoptosis and the interactions between Puma and other Bcl-2 family members, we expressed Puma under an inducible promoter in cells lacking one or more Bcl-2 family members. Puma rapidly induced apoptosis in cells lacking the BH3-only proteins, Bid and Bim. Puma expression resulted in activation of Bax, but Puma killing was not dependent on Bax or Bak alone as Puma readily induced apoptosis in cells lacking either of these proteins, but could not kill cells deficient for both. Puma co-immunoprecipitated with the anti-apoptotic Bcl-2 family members Bcl-x(L) and Mcl-1 but not with Bax or Bak. These data indicate that Puma functions, in the context of induced overexpression or IL-3 deprivation, primarily by binding and inactivating anti-apoptotic Bcl-2 family members.

Structure of the MDM2/MDMX RING domain heterodimer reveals dimerization is required for their ubiquitylation in trans.

MDM2, a ubiquitin E3-ligase of the RING family, has a key role in regulating p53 abundance. During normal non-stress conditions p53 is targeted for degradation by MDM2. MDM2 can also target itself and MDMX for degradation. MDMX is closely related to MDM2 but the RING domain of MDMX does not possess intrinsic E3-ligase activity. Instead, MDMX regulates p53 abundance by modulating the levels and activity of MDM2. Dimerization, mediated by the conserved C-terminal RING domains of both MDM2 and MDMX, is critical to this activity. Here we report the crystal structure of the MDM2/MDMX RING domain heterodimer and map residues required for functional interaction with the E2 (UbcH5b). In both MDM2 and MDMX residues C-terminal to the RING domain have a key role in dimer formation. In addition we show that these residues are part of an extended surface that is essential for ubiquitylation in trans. This study provides a molecular basis for understanding how heterodimer formation leads to stabilization of MDM2, yet degradation of p53, and suggests novel targets for therapeutic intervention.

Caspase inhibitors: viral, cellular and chemical.

Caspases, key mediators of apoptosis, are a structurally related family of cysteine proteases that cleave their substrates at aspartic acid residues either to cause cell death or to activate cytokines as part of an immune response. They can be controlled upstream by the regulation of signals that lead to zymogen activation, or downstream by inhibitors that prevent them from reaching their substrates. This review specifically looks at caspase inhibitors as distinct from caspase regulators: those produced by the cell itself; those whose genes are carried by viruses; and artificial caspase inhibitors used for research and potentially as therapeutics.

Identification of mammalian mitochondrial proteins that interact with IAPs via N-terminal IAP binding motifs.

Direct IAP binding protein with low pI/second mitochondrial activator of caspases, HtrA2/Omi and GstPT/eRF3 are mammalian proteins that bind via N-terminal inhibitor of apoptosis protein (IAP) binding motifs (IBMs) to the baculoviral IAP repeat (BIR) domains of IAPs. These interactions can prevent IAPs from inhibiting caspases, or displace active caspases, thereby promoting cell death. We have identified several additional potential IAP antagonists, including glutamate dehydrogenase (GdH), Nipsnap 3 and 4, CLPX, leucine-rich pentatricopeptide repeat motif-containing protein and 3-hydroxyisobutyrate dehydrogenase. All are mitochondrial proteins from which N-terminal import sequences are removed generating N-terminal IBMs. Whereas most of these proteins have alanine at the N-terminal position, as observed for previously described antagonists, GdH has an N-terminal serine residue that is essential for X-linked IAP (XIAP) interaction. These newly described IAP binding proteins interact with XIAP mainly via BIR2, with binding eliminated or significantly reduced by a single point mutation (D214S) within this domain. Through this interaction, many are able to antagonise XIAP inhibition of caspase 3 in vitro.

Apoptosis and toxicology--what relevance?

All cells are mortal-i.e. they can be killed if a vital metabolic process is blocked. All cells can engage in a variety of stress responses, such as the heat shock response, when vital processes are slowly, or only partially, inhibited. These stress responses involve detection of the damage, transduction of signals, and activation of a response, such as production of heat shock proteins, proteases, or chaperones. Many cells possess mechanisms whose purpose is to kill the cell. Such physiological cell death mechanisms are used to remove unwanted or damaged cells. Among metazoans, physiological cell death is implemented by a family of cysteine proteases, termed caspases, that exist in a latent state even in healthy cells. Cells killing themselves via activation of their caspases typically exhibit an appearance termed 'apoptosis'. Apoptosis is not only used to remove cells in physiological circumstances, such as during development, but is also a common response to cell stress. Thus many cells will detect damage to, or malfunctioning of, vital metabolic processes, and generate signals that lead to activation of the caspases, and apoptotic death of the cell. This has led to a great deal of confusion, because many drugs and toxins with known biochemical functions have been found to induce apoptosis, and rather than this being interpreted as a stress response, it has often wrongly been assumed that apoptosis is a direct effect of the drug or toxin.