Inhibitor of Apoptosis (IAP) Proteins
In 1995 and 1996, the first IAPs were identified in insects and mammals, and based on the definition that BIR-containing proteins are designated as IAP proteins, eight IAPs are encoded in the human genome (Fig. 1). Most IAPs harbor several functional structural domains in addition to the BIR domains, and it is therefore not surprising that individual IAPs serve distinct biological functions. This entry will primarily go through what is currently known about three IAP proteins, cIAP1 (also termed BIRC2, MIHB, API1, and RNF48), cIAP2 (also termed BIRC3, MIHC, API2, and RNF49), and XIAP (also termed BIRC4, MIHA, API3, hILP, and XLP2), that are structurally related and share functions important for cellular signaling processes (Srinivasula and Ashwell 2008; Vaux and Silke 2005).
IAP Structural Domains
As mentioned above, the defining feature of an IAP protein is the presence of one or more BIR domains and mammalian IAPs contain either one or three copies. cIAP1, cIAP2, and XIAP all have three BIRs and a C-terminal RING domain (Fig. 1). In addition, they comprise an ubiquitin-associated (UBA) domain, and cIAP1 and cIAP2 also contain a caspase-associated recruitment domain (CARD). BIR domains are ca. 70 amino acids in size and mediate protein-protein interaction with partners that contain an IAP-binding motif (IBM) as well as other non-IBM-type protein interactions. IBMs are four amino acid motifs starting with an exposed N-terminal alanine. IBMs have been described in several proteins including the processed forms of the mitochondrial factors Smac/DIABLO and Omi/Htr2A and in cleavage-activated caspases. Through their RING domains, IAPs interact with ubiquitin-conjugating enzymes (E2s) and facilitate the transfer of activated ubiquitin from the bound E2 to lysine residues on target proteins (ubiquitylation). The UBA domain enables IAPs to directly interact with polyubiquitylated proteins (Gyrd-Hansen and Meier 2010). The CARD functions as a self-regulatory domain by keeping cIAP1 in a monomeric state, thus limiting its ubiquitin ligase activity (Lopez et al. 2011).
IAPs in Apoptosis
Apoptosis is executed by a family of proteases termed caspases and can be triggered by extrinsic signals such as ligands that activate death receptors on the cell surface (e.g., Fas Ligand (FasL; also known as CD95 ligand and APO-1 ligand) and tumor necrosis factor (TNF)) or by intrinsic insults including DNA damage, oncogene activation, oxidative stress, and growth factor deprivation (Meier and Vousden 2007). The extrinsic pathway leads to activation of caspase-8, which in turn activates downstream effector caspases (e.g., caspase-3 and caspase-7) either directly or via the mitochondrial apoptosis pathway. Activation of the mitochondrial pathway, which also is engaged by intrinsic insults, leads to the release of proapoptotic factors from the mitochondrial intermembrane space (e.g., cytochrome c and Smac/DIABLO). In the cytosol, cytochrome c associates with Apaf-1 and caspase-9 to form the caspase-9 activating platform, termed the apoptosome. In turn, the active caspase-9 activates effector caspases through proteolytic cleavage. The release of apoptogenic factors from the mitochondria is considered the “point of no return,” and this process is tightly regulated by a network of pro- and antiapoptotic Bcl-2 family proteins including Bid, which links extrinsic signals and caspase-8 activation to the intrinsic mitochondrial pathway.
The strongest evidence that IAPs are important for regulation of apoptosis in vivo comes from studies of Drosophila melanogaster (fruit fly) IAP1 (DIAP1) (Ganesan et al. 2011; Orme and Meier 2009; Steller 2008). Genetic loss of DIAP1 results in spontaneous caspase activation causing widespread apoptosis and death of the embryo. Among the mammalian IAPs, XIAP is the only direct inhibitor of caspases, and overexpression of XIAP can inhibit caspase activation and apoptosis triggered by several stimuli (Eckelman et al. 2006; Srinivasula and Ashwell 2008). Conversely, cells that lack XIAP or are depleted for XIAP may become sensitized to apoptosis. Nonetheless, XIAP-deficient mice or mice expressing a truncated form of XIAP without the C-terminal RING domain are healthy and display no apparent defects, and cells isolated from these mice are not overtly sensitized to apoptosis except when treated with TNF in combination with cycloheximide (a protein synthesis inhibitor) (Gyrd-Hansen and Meier 2010). XIAP is therefore not essential for maintaining cell survival in mammals as DIAP1 is in flies, and the evidence rather suggests that XIAP modulates the apoptotic threshold under certain conditions and/or in certain cell types. This notion is supported by in vivo experiments showing that XIAP, in the absence of Bid, determines whether hepatocytes undergo apoptosis after exposure to FasL (Kaufmann et al. 2011). In wild-type mice, intraperitoneal injection of FasL results in massive apoptosis of hepatocytes in a manner that requires amplification of the caspase cascade through Bid-mediated release of apoptosis-promoting factors from the mitochondria. Bid-deficient mice are refractory to FasL treatment, but mice that are deficient for both Bid and XIAP are sensitive to FasL treatment and die from the injection of FasL akin to wild-type mice.
IAPs may also regulate the activity of caspases through RING-dependent ubiquitylation (Broemer and Meier 2009; Vucic et al. 2011). The RING activity of DIAP1 is important for inhibition of caspase activity and proper regulation of cell death in the fruit fly. DIAP1 ubiquitylation of active effector caspases does not cause proteasomal degradation of the caspases but inhibits their activity. Similarly, XIAP can facilitate caspase-3 ubiquitylation in a RING-dependent manner to regulate its activity, although here the ubiquitylation is also reported to target caspase-3 for proteasomal degradation.
IAPs in Immune Signaling
The first evidence that IAP proteins regulate inflammatory signaling and activation NF-κB transcription factors came from genetic studies in fruit flies. Here, DIAP2 was found to be indispensable for NF-κB activation and innate immunity in response to Gram-negative bacteria (Ganesan et al. 2011; Lopez and Meier 2010). Infection with Gram-negative bacteria elicits an innate immune response by activating the immune deficiency (Imd) pathway, which is critically dependent on the ubiquitin ligase activity of DIAP2. DIAP2 conjugates lysine (K) 63-linked ubiquitin chains onto Imd, which facilitates activation of downstream of kinases that, in turn, activate the transcription factor Relish (a NF-κB-like transcription factor) (Ganesan et al. 2011; Lopez and Meier 2010). Accordingly, DIAP2-deficient flies fail to induce Relish-mediated expression of antimicrobial peptides and rapidly succumb to bacterial infections. Intriguingly, XIAP-deficient mice were recently found to be unable to efficiently clear bacterial infections and to ultimately die from the infection (Damgaard and Gyrd-Hansen 2011).
The domain organization of DIAP2 is very similar to that of the mammalian IAPs and akin to the role of DIAP2; cIAP1, cIAP2, and XIAP have emerged as regulators of the inflammatory response downstream of pattern recognition receptors (PRRs) and members of the TNF receptor superfamily (Lopez and Meier 2010). cIAPs regulate activation of NF-κB transcription factors downstream of TNFR1, Toll-like receptor 4 ( TLR4), and the cytosolic bacterial sensors NOD1 and NOD2. A unifying point in these pathways is the activation of TAK1 (TGFβ-activated kinase 1; also known as mitogen-activated protein kinase kinase kinase 7 (MAP 3 K7)) which in turn facilitates activation of MAP kinases and the IκB kinase (IKK) complex consisting of the catalytic subunits IKKα and IKKβ and the obligate regulatory subunit NF-κB essential modifier (NEMO; also known as Open image in new window ). Once activated, IKK phosphorylates IκBα, an inhibitory subunit of NF-κB, which leads to its ubiquitylation and degradation by the proteasome. This enables nuclear translocation of NF-κB where it drives transcription of target genes required for the inflammatory response (Fig. 3). Activation of TAK1 and IKK is dependent on recruitment of the kinases to ubiquitin-modified proteins via the associated regulatory proteins TAB2/3 and NEMO, respectively. TAB2/3 and NEMO harbor ubiquitin-binding domains that mediate the recruitment to ubiquitylated proteins within the receptor complexes. In response to TNFR1 activation, ubiquitylation of RIPK1 and other components by cIAPs promotes the association of a trimeric ubiquitin ligase complex, termed LUBAC (linear ubiquitin chain assembly complex) (Walczak 2011). LUBAC expands ubiquitylation at the TNFR1 complex to enable efficient signaling and activation of the downstream kinases (Fig. 3). In addition, cIAPs are essential for signaling downstream of NOD1/2 where they are reported to ubiquitylate RIPK2 – a required adaptor similar to RIPK1. XIAP is also implicated in NOD2 signaling and was found to associate with RIPK2, but how XIAP contributes to signaling is currently unknown (Damgaard and Gyrd-Hansen 2011).
As mentioned above, cIAP1 and cIAP2 are required to prevent spontaneous activation the non-canonical NF-κB signaling pathway. This pathway is activated by receptors such as the B cell activating factor receptor (BAFF-R) and CD40 that are important for B cell function and lymphoid organ development. BAFF-R and CD40 stimulation leads to accumulation and self-activation of the NF-κB-inducing kinase (NIK). Upon activation, NIK activates an IKK complex consisting of IKKα dimers, and in turn, IKK facilitates activation of the NF-κB transcription factors (Bonizzi and Karin 2004).
The pathway is kept in check by a complex consisting of cIAP1, cIAP2, TRAF2, and TRAF3. TRAF3 links the complex to NIK, and the cIAP proteins ubiquitylate NIK (Lopez and Meier 2010). In response to CD40 or BAFF-R stimulation, cIAPs, through an unknown mechanism, shift target and instead ubiquitylate TRAF3, which leads to its degradation. This stabilizes NIK and allows activation of IKK and NF-κB. Treatment of cells with SM compounds results in a similar activation of NF-κB since cIAPs are degraded by the compound and no longer can mark NIK for degradation. This pathway is also activated by stimulation of the cytokine receptor Fn14 by its cognate ligand TWEAK. However, in this setting, Fn14 activation causes lysosomal degradation of the TRAF-cIAP complex, which thereby causes stabilization of NIK and subsequently activation of NF-κB.
cIAP-mediated ubiquitylation and degradation of TRAF3 also seems to contribute to TLR4 and RIG-I signaling (Lopez and Meier 2010). In response to TLR4 activation by LPS (a component of the bacterial cell wall), cIAPs are reported to ubiquitylate TRAF3 and cause its degradation which contributes to activation of MAP kinases, but not activation of NF-κB. RIG-I is an intracellular sensor of double-stranded viral RNA, and RIG-I activation leads to production of type I interferons such as IFNβ. cIAP-mediated degradation of TRAF3 after RIG-I activation is reported to be essential for IFNβ production after Sendai virus infection.
Other IAP Functions
IAPs are reported to negatively regulate the Ras-Raf-ERK signaling pathway. The pathway is activated by multiple growth factors and is frequently hyperactivated in cancer. XIAP, cIAP1, and cIAP2 were shown to interact with C-Raf and reduce the cellular levels of C-Raf in a ubiquitin-dependent manner, which involves the Hsp90 cochaperone and ubiquitin ligase CHIP (Gyrd-Hansen and Meier 2010). Accordingly, knockdown of XIAP or cIAPs led to an increase in cell motility and the invasive potential of HeLa cells in response to growth factor stimulation. Contrary to this, however, XIAP has been suggested to promote cell migration and metastasis in vivo in cooperation with survivin. The mechanism for this is not fully uncovered but appears to involve NF-κB activity.
XIAP is involved in the regulation of intracellular copper levels through its ubiquitylation of COMMD1 (copper metabolism gene MURR1 domain 1; also known as MURR1) (Srinivasula and Ashwell 2008). Defects in regulation of copper levels may cause severe pathologies such as Wilson’s disease, an autosomal recessive condition that presents with neurological damage to the basal ganglia and cirrhosis due to accumulation of copper, particularly in brain and liver. XIAP binds COMMD1 via its BIR3 domain and ubiquitylates it in a RING-dependent manner, which targets COMMD1 for proteasomal degradation. The physiological relevance of this function of XIAP is supported by the finding that liver tissue and fibroblasts from XIAP-deficient mice have increased COMMD1 levels and decreased copper levels compared to wild-type controls (Srinivasula and Ashwell 2008).
IAPs in Human Diseases
Deregulation of several IAP family members has been observed in cancer. Although most links are correlative, there is evidence that cIAP1 and cIAP2, in particular, may contribute directly to cancer development. For example, the locus harboring BIRC2/cIAP1 and BIRC3/cIAP2 as well as Yap1 and several MMP genes (11q21–q22) is found to be amplified in multiple human cancers including hepatocellular carcinoma (HCC), lung cancers, oral squamous cell carcinomas, medulloblastomas, glioblastomas, and pancreatic cancers (Gyrd-Hansen and Meier 2010; Srinivasula and Ashwell 2008). This is further supported by tumor models in mice where recurrent amplification of the 9qA1 locus, which is syntenic to the human 11q22 locus, was found in c-Myc-driven HCC in p53−/− mice and in spontaneously occurring osteosarcomas in p53+/− mice (Gyrd-Hansen and Meier 2010). In a separate report, cIAP1 was shown to positively regulate c-Myc activity by ubiquitylating the Myc-antagonistic protein, Max-dimerization protein-1 (Mad1), and facilitate its degradation (Vucic et al. 2011). The cIAP1-dependent decrease in Mad1 level correlated with increased proliferation and cell transformation in cell culture studies. However, whether this fully explains the role of cIAP1 in tumorigenesis in vivo needs further investigation.
Intriguingly, biallelic deletion of the locus encoding cIAP1 and cIAP2 is frequently observed in multiple myeloma (MM) together with other genes involved in regulation of NF-κB activity. MM is a B cell malignancy characterized by accumulation of antibody producing B cells (clonal) in the bone marrow. Combined, genetic lesions that cause deregulation of NF-κB activity were identified in ca. 20% of cases. Among the most frequent genetic alterations found were deletion of the BIRC2/cIAP1 and BIRC3/cIAP2 locus, deletion of TRAF3, TRAF2, and CYLD, and enhanced expression of lymphotoxin beta receptor (LTβR), CD40, TNFRSF13B/TACI, NFKB1, NFKB2, and NIK (Gyrd-Hansen and Meier 2010).
The mechanistic understanding of how amplification or deletion of the cIAP genes contributes to cancer development is currently not fully understood. In the case of MM, however, the identification of multiple genetic and expression abnormalities in MM that all affect genes involved in NF-κB signaling strongly suggests that unrestrained NF-κB activity contributes to driving these tumors. Analogous to this, chromosomal translocations that favor unrestrained NF-κB activity are frequent in MALT lymphoma (Ye et al. 2005). MALT lymphoma is an extranodal B cell lymphoma that arises in mucosa, primarily in the gastric tract and lungs, and is closely linked to Helicobacter pylori infections. The most common genetic lesion in MALT lymphomas is the reciprocal chromosomal translocation t(11;18)(q21;q21) that generates a fusion gene whose product consists of the N-terminal part of cIAP2 and the C-terminal portion of MALT1. The resulting chimeric protein cIAP2-MALT1 (also known as API2-MALT1) drives constitutive canonical NF-κB signaling without requirement for activation of cell surface receptors (Ye et al. 2005). The cIAP2-MALT1 fusion bypasses the normal regulation of B cell receptor signaling by facilitating constitutive K63-linked ubiquitylation of NEMO by the ubiquitin ligase TRAF6. In turn, ubiquitylated NEMO is retained by the cIAP2 UBA domain present in cIAP2-MALT1 to enable activation of IKK and NF-κB (Gyrd-Hansen and Meier 2010).
Although XIAP is described to be overexpressed in many types of cancer, no genetic lesions affecting XIAP in human malignancies have been identified. In contrast, XIAP mutations have recently been reported in patients suffering from X-linked lymphoproliferative syndrome type 2 (XLP-2), a condition characterized by deregulation of the immune system and defined by hemophagocytic lymphohistiocytosis (Pachlopnik Schmid et al. 2011). Following viral infection, XLP-2 patients frequently develop cytopenia, fever, splenomegaly and hemorrhagic colitis, suggesting that mutations in XIAP predispose patients to the development of immunodeficiency (Pachlopnik Schmid et al. 2011). This is consistent with the experimental evidence which indicates that XIAP contributes to innate immunity to bacteria by facilitating NOD2 signaling, but whether XIAP also functions in other immune-related signaling paradigms is yet to be investigated (Damgaard and Gyrd-Hansen 2011).
The understanding of the cellular functions of individual IAP proteins has come a long way since the initial observation that expression of a viral IAP can maintain viability of cells after viral infection. In addition to their well-described role as gatekeepers of caspase activity and apoptosis, IAPs have now emerged as key regulators of signaling pathways important for inflammation and innate immunity. The molecular details of how IAPs impinge on these processes are still scarce, and in-depth understanding of these processes will undoubtedly be pursued in the years to come. Intriguingly, the realization that IAPs regulate immune-related processes may have unanticipated implications for the clinical applications of IAP-antagonistic compounds to also include treatment of immune-related syndromes as well as cancers driven by uncontrolled inflammatory signaling.
- Pachlopnik Schmid J, Canioni D, Moshous D, Touzot F, Mahlaoui N, Hauck F, Kanegane H, Lopez-Granados E, Mejstrikova E, Pellier I, et al. Clinical similarities and differences of patients with X-linked lymphoproliferative syndrome type 1 (XLP-1/SAP deficiency) versus type 2 (XLP-2/XIAP deficiency). Blood. 2011;117:1522–9.PubMedCrossRefGoogle Scholar