Given the critical importance of making the correct choices about cell life–death decisions in complex multicellular organisms, it is not surprising that the pathways governing caspase activation are under exquisite control by networks of proteins that directly or indirectly communicate with these proteases. A delicate balance between proapoptotic and antiapoptotic regulators of apoptosis pathways is at play on a continual basis, ensuring the survival of long-lived cells and the proper turnover of short-lived cells in a variety of tissues, including the marrow, thymus, and peripheral lymphoid tissues. The antiapoptotic proteins responsible for creating roadblocks to cell death have been mapped to specific caspase-activation pathways.
The Bcl-2 family represents a large group of proteins (number >26 in humans) that control mitochondria-dependent steps in cell-death pathways, including dictating whether Cyt-c is or is not released from these organelles (Fig. 15–2). Both proapoptotic and antiapoptotic Bcl-2 family proteins have been delineated.28 These proteins are best known for their roles in controlling the intrinsic (mitochondrial) cell-death pathway,20 although effects on the ER-pathway for cell death have also been documented.29 Even though the human genome encodes at least 26 Bcl-2 family proteins, only six of these are antiapoptotic (in humans, Bcl-2, Bcl-XL [BCL2L1], Bcl-W [BCL2L2], Mcl-1 [BCL2L3], Bfl-1 [BCL2L5], and Bcl-B [BCL2L10]). Several types of animal viruses also harbor Bcl-2 family genes within their genomes, including herpes viruses implicated in cancer such as the Epstein-Barr virus (EBV) and Kaposi sarcoma virus (KSV). The relative ratios of anti- and proapoptotic Bcl-2 family proteins dictate the ultimate sensitivity or resistance of cells to various apoptotic stimuli, including growth factor deprivation, hypoxia, radiation, anticancer drugs, oxidants, and Ca2+ overload.
Network of interactions among Bcl-2 family proteins. The functional and physical interactions among proapoptotic and antiapoptotic Bcl-2 family proteins are depicted. Illustrative members of the Bcl-2 family are shown.
Various Bcl-2 family members play important roles in controlling the life spans of hematopoietic cells, as evidenced by phenotypes generated in genetically engineered mice (gene knockouts and transgenics) and also (in some cases) by human clinical experiences with experimental therapeutics targeting some of these proteins. For example, antiapoptotic protein Bcl-2 is required for survival of mature T cells and B cells, with deficiency of Bcl-2 causing lymphopenia. Conversely, the proapoptotic protein Bim is necessary for limiting expansion of T and B lymphocytes, with deficiency of Bim causing lymphocytosis. Bim also plays important role in eradicating autoreactive T cells in the thymus (“negative selection”), having important implications for mechanisms of autoimmune diseases.30 Bcl-XL is required for platelet homeostasis, such that either genetic or pharmacologically induced Bcl-XL deficiency causes thrombocytopenia. Antiapoptotic protein Mcl-1 is particularly important for survival of the myeloid lineage in mice, as well as contributing to lymphocyte survival. Conversely, antiapoptotic protein Bcl-W is not required for hematopoiesis in mice, despite being widely expressed in myeloid lineage cells. It should be noted that direct comparisons of gene manipulations in mice with the human circumstance are not always possible because of genomic differences in the Bcl-2 family genes of mice versus humans (e.g., human Bfl-1 versus murine A1; human Bcl-B versus murine Boo/Diva).
Many members of the Bcl-2 family have a hydrophobic stretch of amino acids near their carboxyl-terminus that anchors them in the outer mitochondrial membrane.17 In contrast, other Bcl-2 family members such as Bid, Bim, and Bad, lack these membrane-anchoring domains, but dynamically target mitochondria in response to specific stimuli. Still others have the membrane-anchoring domain but keep it latched against the body of the protein until stimulated to expose it (e.g., Bax).31
Based on their predicted (or experimentally determined) three-dimensional structures, Bcl-2 family proteins can be broadly divided into two groups. One subset of these proteins is probably similar in structure to the pore-forming domains of bacterial toxins, such as the colicins and diphtheria toxin.32,33,34,35 These α-helical pore-like proteins include both antiapoptotic proteins (Bcl-2, Bcl-XL, Mcl-1, Bfl-1, Bcl-W, Bcl-B) and proapoptotic proteins (Bax, Bak, Bok, and Bid). Most of the proteins in this subcategory can be recognized by conserved stretches of amino acid sequence homology, including the presence of Bcl-2 homology (BH) domains, BH1, BH2, BH3, and sometimes BH4. However, this is not uniformly the case, as the Bid protein contains only a BH3 domain but has been determined to share the same overall protein-fold with Bcl-XL, Bcl-2, and Bax.33,34 Where tested to date, these proteins have all been shown to form ion-conducting channels in synthetic membranes in vitro, including Bcl-2, Bcl-XL, Bax, and Bid,36,37,38,39,40 but the significance of this pore activity remains unclear.
The other subset of Bcl-2 family proteins appears to have in common only the presence of the BH3 domain, including Bad, Bik, Bim, Hrk, Bcl-GS, p193, APR (Noxa), and PUMA. These “BH3-only” proteins are uniformly proapoptotic. Their cell-death–inducing activity depends, in most cases, on their ability to dimerize with antiapoptotic Bcl-2 family members, functioning as trans-dominant inhibitors of proteins such as Bcl-2 and Bcl-XL.41,42 However, some of these proteins (e.g., Bid, PUMA, Bim) can also interact with proapoptotic proteins (e.g., Bax, Bak), functioning as agonists of the killers, in addition to dimerizing with antiapoptotics (e.g., Bcl-2; Bcl-XL) to function as antagonists of these cell-survival proteins (see Fig. 15–2).28,43 Binding of Bid to Bax or Bak promotes insertion of these proteins into membranes where they oligomerize, apparently forming large pores through which molecules such as Cyt-c, SMAC, and Omi can escape from mitochondria or causing an increase in the permeability of the outer membrane of mitochondria through more complex mechanisms.44,45 Bax and Bak thus induce mitochondrial outer membrane permeabilization (MOMP), which is a critical event that not only causes release of death-inducing mitochondrial proteins but also secondarily causes necrosis by uncoupling of oxidative phosphorylation (when Cyt-c becomes limiting) and diversion of electrons from the respiratory chain into production of toxic free radicals.46,47
The BH3 domain mediates dimerization among Bcl-2 family proteins. This domain consists of an amphipathic α-helix of approximately 16 amino acids that inserts into a hydrophobic crevice on the surface of antiapoptotic proteins such as Bcl-2 and Bcl-XL.48 The BH3-only proteins link a wide variety of environmental stimuli to the mitochondrial pathway for apoptosis, with some examples outlined below.
In addition to mitochondria, mechanistic links for Bcl-2 family proteins to ER stress and autophagy have also been delineated. For example, Bax and Bak can bind the ER stress signaling protein, IRE-1, thereby stimulating its intrinsic autokinase activity and its endoribonuclease activity.49 Proapoptotic (Bak/Bax) and antiapoptotic (Bcl-2/Bcl-XL) family members also have opposing effects on basal ER Ca2+ levels, probably via effects on Ca2+ channel proteins in ER membranes (e.g., IP3Rs, BI-1, and TmBim3). Autophagy protein Beclin contains a BH3-like domain that mediates interactions with antiapoptotic Bcl-2 family proteins, which sequester Beclin and thereby reduce autophagic flux.50 Autophagy, which is a lysosome-dependent catabolic pathway, can either promote cell survival by providing access to nutrients during times of nutrient insufficiency and hypoxia, or it can cause cell death when stimulated to an extreme.51
The c-FLIP proteins are another type of apoptosis suppressor that operates via directly binding to certain caspases and their upstream activator, FADD. The c-FLIP gene of humans resides in a tandem gene cluster on chromosome 2, which contains the genes encoding procaspases 8 and 10, suggestive of gene duplication events. Two isoforms of c-FLIP are produced from a single gene, including the long form, which is highly similar in overall sequence to procaspases 8 and 10, containing tandem copies of DEDs, followed by a pseudocaspase domain that lacks enzymatic activity. The shorter isoform consists only of the DED domains, thus resembling analogous proteins encoded in the genomes of some mammalian viruses.52 FLIP-S is exclusively antoptotic whereas FLIP-L can be either pro- or antiapoptotic, depending on its levels of expression relative to procaspases 8 and 10.53 In general, FLIP proteins form complexes with procaspases 8 and 10, preventing their dimerization and activation, as well as competing for binding to adapter protein FADD, which is required for caspase recruitment to DR complexes.54,55 Thus, in most circumstances, FLIP proteins create blockades in the extrinsic pathway for apoptosis.
Additionally, a role in suppressing nonapoptotic cell death (necroptosis) has been described for FLIP in partnership with caspase-8.56 In this regard, TNFR1 signaling has been shown to stimulate caspase-independent cell death in some circumstances (commonly called “necroptosis”) via a mechanism that is suppressed by FLIP and caspase-8 but that is dependent on the protein kinase Rip3 (see section “Inhibitors of Apoptosis” below). It is thought that dimers consisting of the longer isoform of c-FLIP plus caspase-8 direct the proteolytic activity of caspase-8 to substrates that promote cell survival rather than cell death.57 Among the relevant substrates is the kinase Rip1, an upstream activator of Rip3. Thus, FLIP proteins play complex roles in cell death regulation mediated by various members of the TNF Receptor family.
The IAP proteins (n = 8 in humans) suppress apoptosis via a diversity of mechanisms, including directly binding to and inhibiting certain caspases.58,59 IAPs are characterized by the presence of protein interaction domains called BIRs (baculovirus internal repeats), numbering between 1 and 3 per protein. Most IAPs also carry RING domains that endow them with E3 ligase activity through interactions with ubiquitin conjugating enzymes (UBCs). Some of the apoptogenic proteins released from mitochondria, notably SMAC and HtrA2, bind certain BIRs and thereby compete for protein interactions on the surface of IAPs. Some examples of IAP mechanisms are provided here.
XIAP (so-called because its encoding gene resides on the X-chromosome) contains 3 BIR domains. BIR2 of XIAP binds downstream effector proteases, caspases-3 and -7, to suppress apoptosis at a distal point. BIR3 of XIAP binds upstream initiator protease, caspase-9, to suppress an apical step in the mitochondrial pathway for apoptosis.
The c-IAP1 (BIRC2) and c-IAP2 (BIRC3) proteins are also capable of binding to caspases 3, 7, and 9, although they are less potent by far as direct enzymatic inhibitors and may rely on their E3 ligase activity for controlling caspase degradation. However, these IAP family members also participate in other cell death-relevant mechanisms by impacting signal transduction by TNF family receptors. Binding of TNF to one of its principal cellular receptors expressed widely on cells, TNFR1, is capable of triggering at least three different signaling pathways, each involving overlapping but distinct protein complexes that are assembled at the receptor (Fig. 15–3). One of these TNFR1-initiated pathways causes caspase activation and apoptosis by DISC assembly (described in the section “Caspase Activation Pathways,” and Fig. 15-1 above). Another pathway causes activation of the kinase Rip3, usually via the upstream kinase Rip1, which associates with TNFR1 complexes.60 The Rip3-dependent cell death pathway is caspase-independent, leading to nonapoptotic cell death (“necroptosis”) through a process involving reactive oxygen species (ROS) generated by mitochondria. Another serine/threonine kinase, mixed-lineage kinase domain-like (MLKL) protein, appears to be a critical downstream mediator of Rip3-induced necroptosis. This Rip3-dependent pathway for necroptosis is suppressed by c-IAP1 and c-IAP2, probably via their roles as E3 ligases and possibly involving ubiquitin/proteasome-mediated reductions in Rip3 protein levels.61 Finally, TNFR1 stimulates a cell survival pathway in which c-IAP1 and c-IAP2 participate. In this TNFR1-mediated survival pathway, the kinase Rip1 comes together with the E3 ligases c-IAP1, c-IAP2, and tumor receptor-associated factor 2 (TRAF2) to stimulate noncanonical (lysine 63, rather than lysine 48) ubiquitination of Rip1, initiating a signal transduction pathway that causes activation of transcription factor nuclear factor (NF)-κB. NF-κB influences the expression of many target genes involved in host defenses and immune regulation, among which are several genes that suppress apoptosis. As a result, this NF-κB pathway nullifies the caspase pathway, negating apoptosis,62 in addition to accounting for the untoward inflammatory actions of this cytokine. Several antiapoptotic genes are among the direct transcriptional targets of NF-κB (REL)-family proteins, including the genes encoding c-FLIP, c-IAP2, Bcl-XL, and Bfl-1.
Opposing pathways for cell death and cell survival are induced by tumor necrosis factor receptor (TNFR). TNFR1 is the best studied of the death domain (DD)-containing TNF family receptors, which include in humans Fas (CD95), tumor necrosis factor–related apoptosis-inducing ligand (TRAIL) receptor-1 (TRAILR1, DR4), TRAILR2 (DR5), DR3, and DR6. DD-containing adapter protein TRADD (tumor necrosis factor receptor death domain) binds the DD in the cytosolic domain of TNFR1, which then connects to at least one of three different pathways that are outlined here. A cell survival pathway results in nuclear factor (NF)-κB activation, whereby TRADD recruits the DD-containing protein Rip1 and also binds the E3 ligase/adapter protein TRAF2 (tumor receptor-associated factor 2). Rip1 and TRAF2 bind c-IAP1 (inhibitor of apoptosis 1) and c-IAP2. The resulting complex promotes noncanonical ubiquitination of the kinase Rip1, triggering a signal transduction kinase pathway that results in activation of I-κB kinases (IKKs) that cause I-κB ubiquitination and proteasomal degradation, thereby releasing sequestered NF-κB to allow its translocation into the nucleus where it stimulates expression of multiple antiapoptotic genes. In the TNFR1-mediated apoptosis pathway, the DD of TRADD associates with the DD of FADD, which, in turn, binds caspases 8 and 10 via their death effector domains (DEDs), triggering protease activation and thereby stimulating apoptosis. The TNFR1-mediated pathway for necrosis (necroptosis) involves a cascade of events that include recruitment of Rip1, which, in turn, activates Rip3, which activates mixed-lineage kinase domain-like (MLKL) kinase and which causes mitochondrial and probably lysosomal changes that stimulate reactive oxygen species (ROS) generation and lead to necrosis.
In this regard, the c-IAP1 and c-IAP2 proteins were first identified because of their association with TNF receptor complexes. These IAPs bind the kinase Rip1 via their BIR3 domains, mediating noncanonical ubiquitination of Rip1 via interactions of atypical UBCs with their RING domains and possibly also indirectly via interactions of the E3 ligase TRAF2, which binds their BIR1 domains. The noncanonical ubiquitination of Rip1 is required for TNFR1-mediated NF-κB activation and suppression of cytokine-induced apoptosis. The c-IAP1 and c-IAP2 proteins also control the “alternative” pathway for NF-κB activation via yet another mechanism, which involves classical lysine 48 mediated polyubiquitination of the kinase Nik.
Additional members of the IAP family not described here (Survivin, Apollon/Bruce, ML-IAP, etc.) also have interesting mechanisms of interacting with components of cell-death pathways and they also can have other roles beyond cell-death regulation. For example, XIAP, c-IAP1, and c-IAP2 have other documented cellular activities, which include, for example, their interactions with kinases (e.g., Rip2) or kinase-binding adapter proteins (TAB/Tak) involved in processes such as innate immunity and morphogenesis. In these circumstances, the most relevant activity of IAPs appears to be their noncanonical E3 ligase activity, as well as a protein scaffold role where they serve as platforms for assembling multiprotein complexes. Additional roles for IAP family members include cell division, where, for example, the Survivin protein plays a fundamental role in chromosome segregation and cytokinesis.
Several of the IAPs are opposed by proteins released from mitochondria, SMAC and HtrA2. SMAC and Htra2 bind BIR domains on IAPs, thus displacing caspases and other associated proteins. In many cases, SMAC binding to IAPs induces their polyubiquitination and proteasomal degradation. Thus, factors that cause MOMP take the breaks off the caspases by eliminating various IAP family proteins.