Molecular Pathways: Targeting Inhibitor of Apoptosis Proteins in Cancer—From Molecular Mechanism to Therapeutic Application

No potential conflicts of interest were disclosed.

The members of the planning committee have no real or apparent conflict of interest to disclose.

Upon completion of the CME activity, the participant should have a better understanding of the molecular pathways that are regulated by IAP proteins. In addition, the participant should understand the rationale for the design of current therapeutics aimed at antagonizing IAP proteins and the preclinical or clinical development of IAP antagonists in cancer.

This activity does not receive commercial support.

Inhibitor of apoptosis (IAP) proteins are a family of eight human proteins, including neuronal AIP (NAIP), cellular IAP1 (cIAP1), cellular IAP2 (cIAP2), X chromosome-linked IAP (XIAP), survivin, baculovirus IAP repeat (BIR)–containing ubiquitin-conjugating enzyme (BRUCE/Apollon), melanoma IAP (ML-IAP), and IAP-like protein 2 [ILP-2; reviewed in ref. 1)]. As reflected by their name, they were initially characterized as endogenous inhibitors of caspases. However, besides the regulation of apoptosis, IAP proteins have also been implicated in the control of nonapoptotic processes, including differentiation, cell motility, migration, invasion, and metastasis (2).

All IAP proteins contain at least one of the signature BIR domains, a 70–80 amino acid segment that mediates protein–protein interactions between IAP proteins and caspases, thereby inhibiting caspase activation and activity. IAP proteins such as XIAP, cIAP1, cIAP2, and ML-IAP also contain the Really Interesting New Gene (RING) domain that exhibits E3 ubiquitin ligase activity (3). Depending on the chain type (e.g., K5-, K11-, K48-, or K63-linked chains), ubiquitination can lead to proteasomal degradation of substrates or can alter their signaling properties (3).

IAP proteins are key regulators of programmed cell death pathways. Apoptosis represents one of the best-characterized forms of programmed cell death and involves two major signaling pathways (4). In the extrinsic (death receptor) pathway, the binding of death receptor ligands such as TRAIL, to death receptors, such as TRAIL receptors, triggers recruitment of adaptor molecules [e.g., FAS-associated protein with death domain (FADD)] and caspase-8, which in turn drives activation of caspase-8 and, subsequently, of effector caspases such as caspase-3 (4). Furthermore, caspase-8–mediated cleavage of Bid into tBid provides a link between the extrinsic and intrinsic apoptosis pathways (4). Upon the engagement of the intrinsic (mitochondrial) pathway, mitochondrial proteins such as cytochrome c or second mitochondrial activator of caspases (Smac) are released into the cytosol and trigger apoptosis by promoting caspase activation (4). In the case of Smac, caspases are indirectly activated via Smac-mediated neutralization of IAP proteins, which in turn results in caspase activation. IAP proteins, in particular XIAP, block caspase activation by binding to caspases -3, -7, and -9 via the BIR domains (1) and negatively regulate both the intrinsic and extrinsic apoptosis pathways.

Furthermore, IAP proteins play an important role in the regulation of NF-κB signaling, in particular via ubiquitination events (1). NF-κB represents one of the key transcription factors that control various aspects of tumor biology, including cell death and survival signaling (5). Within the canonical NF-κB pathway, binding of TNF-α to TNF receptor 1 (TNFR1) triggers the recruitment of signaling molecules such as TNFR1-associated death domain protein (TRADD), receptor-interacting protein 1 (RIP1), TNF receptor-associated factor 2 (TRAF2), cIAP1, and cIAP2, which results in nondegradative ubiquitination of RIP1 via cIAP proteins and activation of inhibitor of NF-κB kinase (IKK; IκB kinase β) via a multiprotein complex containing TGF-β–activated kinase 1 (TAK1), TAK1-binding protein (TAB), and NF-κB essential modulator (NEMO; ref. 6). This leads to phosphorylation and proteasomal degradation of IκBα followed by nuclear translocation of NF-κB subunits to activate transcription of NF-κB target genes. NF-κB–responsive genes comprise proinflammatory genes, for example TNF-α and interleukin-8, antiapoptotic genes such as XIAP, Bcl-2, or Bcl-X_L, as well as proapoptotic genes such as CD95 and TRAIL receptors (5). cIAP proteins promote canonical NF-κB activation by ubiquitinating RIP1 (3).

Within the noncanonical NF-κB pathway, cIAP proteins mediate under resting conditions the constitutive ubiquitination and proteasomal degradation of NF-κB–inducing kinase (NIK) together with TRAF2 and TRAF3, thereby negatively regulating noncanonical NF-κB signaling (7, 8). Activation of noncanonical NF-κB signaling, for example, in response to CD40 stimulation, terminates this cIAP-dependent NIK degradation and allows activation of IKKα via NIK, followed by processing of p100 to p52 and translocation of p52 to the nucleus to activate NF-κB target genes (9).

Moreover, IAP proteins have been implicated in the regulation of additional signaling cascades, for example, the mitogen-activated protein kinase (MAPK) pathway (10), TGF-β signaling (11, 12) and innate and adaptive immunity signaling pathways (13, 14). cIAP proteins have been shown to be required for activation of MAPK signaling by members of the TNFR superfamily (10). XIAP has been described to function as a cofactor in TGF-β signaling (11). The immunomodulatory functions of IAP proteins are mediated via their regulation of the NF-κB and MAPK signaling pathways or via the control of the activity of the inflammasome, a multiprotein complex that regulates an immune response to microbial triggers (13).

IAP proteins are highly expressed in multiple human malignancies and have been implicated in promoting tumor progression, treatment failure, and poor prognosis, indicating that IAP proteins represent relevant targets for therapeutic exploitation (1). Indeed, IAP proteins have been shown to play a critical role in the regulation of sensitivity versus resistance of cancers to current cytotoxic strategies. Therefore, many efforts have been made over the last decade to develop strategies to neutralize IAP proteins, including antisense oligonucleotides and small-molecule inhibitors (15, 16). Due to space limitations, the current review focuses on small-molecule inhibitors for therapeutic inhibition of IAP proteins.

In principle, IAP antagonists mimic the N-terminal part of the endogenous IAP antagonist Smac, which is required for binding to IAP proteins, and are composed of nonpeptidic elements (1). In addition to monovalent compounds that contain one Smac-mimicking motif, bivalent or dimeric IAP antagonists were developed, which consist of two monovalent Smac-mimicking units that are connected via a chemical linker. Bivalent IAP antagonists were reported to exhibit a several-fold higher antitumor activity than monovalent ones, due to their higher binding affinities and their higher potency to promote caspase activation and proteasomal degradation of cIAP proteins (17–20).

Similar to the Smac protein, small-molecule IAP antagonists bind to several IAP proteins, including XIAP, cIAP1, cIAP2, and ML-IAP. In principle, chemically distinct IAP antagonists can differentially neutralize IAP proteins. The binding of IAP antagonists to XIAP results in activation of caspases (21, 22). The interaction of IAP antagonists with cIAP proteins stimulates their dimerization and increases their E3 ligase activity (17–20, 23). This leads to increased autoubiquitination and degradation of cIAP proteins via the proteasome. Because cIAP proteins are responsible for the constitutive proteasomal degradation of NIK, a critical upstream component of the noncanonical NF-κB pathway (7, 8), IAP antagonist–mediated depletion of cIAP proteins leads to NIK accumulation and noncanonical NF-κB activation (Fig. 1). Subsequent upregulation of NF-κB target genes such as TNF-α can then engage TNFR1-mediated caspase-8 activation and apoptosis via a RIP1/FADD/caspase-8 cytosolic complex in an autocrine/paracrine manner (17–20) (Fig. 1). By depleting cIAP proteins, IAP antagonists suppress activation of the canonical NF-κB pathway. Initially, however, they may contribute to canonical NF-κB activation by stimulating the E3 ubiquitin ligase activity of cIAP proteins, thereby promoting RIP1 ubiquitination and NF-κB activation. This initial increase in E3 ligase activity of cIAP proteins is usually only transient because cIAP proteins are themselves autoubiquitinated and degraded. Five distinct IAP antagonists are currently under clinical development (Table 1).

Evaluation of IAP antagonists as single agents revealed that they effectively trigger cell death in a small subset of human malignancies (19), indicating that IAP antagonist–based combination therapies might be required in the majority of cancers to achieve sufficient antitumor activity. Therefore, a variety of rational targeted combinations have been developed over the years together with different types of cytotoxic stimuli in a large set of cancer entities in vitro and in vivo to exploit additive or synergistic drug interactions (reviewed in ref. 1). It is interesting to note that there is some evidence from preclinical studies pointing to a therapeutic window in IAP antagonist–based combination regimens to prime malignant cancer cells for cell death induction, while sparing nonmalignant normal cells, e.g., peripheral blood lymphocytes, hematopoietic progenitor cells, and neuronal cells (24–27). However, many more studies need to be performed before any definite conclusion can be drawn. IAP antagonist–based combinations range from conventional chemotherapeutic agents and irradiation, the two pillars of many current treatment protocols, to death receptor agonists and various kinds of small-molecule signal transduction inhibitors.

Chemotherapeutic drugs of different pharmacologic classes, including doxorubicin, etoposide, gemcitabine, paclitaxel, cisplatin, vinorelbine, SN38, 5-fluorouracil (5-FU), and cytarabine, were shown to act in concert with IAP antagonists to exert antitumor activity in preclinical models of cancers (26, 28–32). In clinical trials, paclitaxel, daunorubicin, cytarabine, or gemcitabine have been selected for combination protocols (Table 1). Mechanistically, chemosensitization by IAP antagonists has been linked in some studies to an autocrine/paracrine TNF-α–driven loop that is engaged upon depletion of cIAP proteins (31). However, TNF-α–independent signaling events, i.e., the formation of a cell death complex in the cytosol containing RIP1, FADD, and caspase-8, were also shown to be critical for activation of cell death pathways in response to cotreatment with chemotherapeutics and IAP antagonists (32).

In addition to enhancing the antitumor activity of DNA-damaging drugs, IAP antagonists have been reported to increase radiosensitivity in various cancers (27, 33–36). Interestingly, radiosensitization by IAP antagonists was not restricted to the bulk population of the tumor, but was also found in tumor-initiating cancer stem cells that have been described to be particularly radioresistant (27).

Furthermore, the combination of IAP antagonists together with death receptor agonists turned out to be very potent to induce cell death even in otherwise resistant forms of cancer (22, 24, 25, 37–42). Accordingly, simultaneous neutralization of XIAP and cIAP proteins circumvents not only the requirement for mitochondrial amplification of death receptor–mediated apoptosis in many cancers, but also potentiates death receptor–initiated activation of caspase-8 by promoting the aggregation of caspase-8 together with FADD and RIP1 in a multimeric cytosolic complex (complex II) following the release of these signaling molecules from the receptor-bound death-inducing signaling complex (DISC). Just to give two examples, IAP antagonists were found to be able to prime pancreatic carcinoma cells, which are known to be notoriously refractory to most treatment approaches and to cell death induction, when combined with TRAIL receptor agonists such as soluble TRAIL ligand or monoclonal antibodies directed against one of the agonistic TRAIL receptors, resulting in enhanced apoptosis in vitro and reduced tumor growth in vivo (25, 39). Also, chronic lymphocytic leukemia (CLL) cells derived from patients belonging to the poor prognostic subgroups, e.g., those with TP53 mutation, 17p deletion, chemoresistance, or unmutated V(H) status, turned out to remain susceptible to cell death induction by IAP antagonists and TRAIL receptor agonists (38).

Moreover, IAP antagonists have successfully been combined with a range of signal transduction modulators depending on the cancer entity, including proteasome inhibitors, various kinds of kinase inhibitors, e.g., FMS-like tyrosine kinase (FLT)3 inhibitors, platelet-derived growth factor (PDGF) receptor inhibitors, insulin-like growth factor (IGF) receptor inhibitors, or EGF receptor inhibitors, as well as monoclonal antibodies targeting growth factor receptors (43–46). Also, IAP antagonists have been shown to increase the antitumor effects of immunotherapies both by priming cancer cells to immune cell–mediated cytotoxicity and/or by altering immune cell functions (47).

In addition to promoting apoptosis, there is accumulating evidence indicating that IAP antagonists can also potentiate nonapoptotic forms of cell death such as necroptosis, a recently identified programmed form of necrosis (48). Accordingly, IAP antagonists have been reported to promote the formation of the necrosome, a complex consisting of RIP1 and RIP3, under conditions in which caspase activation is inhibited (49, 50). The IAP antagonist–mediated amplification of necroptosis opens new perspectives to overcome treatment resistance, particularly in apoptosis-refractory forms of cancer.

Five distinct IAP antagonists are currently undergoing evaluation in early clinical trials for the treatment of cancer (www.clinicaltrials.gov). Monovalent agents such as GDC-0917/CUDC-427 (Genentech, Inc./Curis), LCL161 (Novartis), and AT-406 (Ascenta Therapeutics) offer the advantage that they can be administered orally. By comparison, bivalent compounds such as HGS1029 (Aegera Therapeutics/Human Genome Sciences) and TL32711 (TetraLogic Pharmaceuticals) require intravenous administration but may turn out to be more efficacious, as indicated from preclinical studies. Nevertheless, the question remains open about whether mono- or bivalent IAP antagonists are the most promising candidates for further clinical development based on their pharmacodynamic and pharmacokinetic properties as well as their toxicity profiles. In addition, the question about which IAP proteins alone or in combination represent the most critical targets for therapeutic intervention in cancer has not yet been answered. Although neutralization of cIAP proteins has been shown to be instrumental for single-agent activity of IAP antagonists by engaging an NF-κB–dependent autocrine/paracrine TNF-α loop that triggers cell death upon concomitant depletion of cIAP proteins, the release of the XIAP-imposed block on caspases is considered to be critical for full activation of the effector phase of apoptosis.

Phase I trials examining the safety and pharmacologic properties of IAP antagonists trials have been completed for LCL161, HGS1029, and TL32711, demonstrating that IAP antagonists are in principle well tolerated (51–53). Furthermore, several combination protocols with IAP antagonists together with standard-of-care anticancer therapeutics have been initiated, including chemotherapeutic drugs such as paclitaxel, daunorubicin, cytarabine, and gemcitabine (Table 1). In addition, a trial testing TL32711 in combination with the demethylating agent 5-azacytidine has recently been launched (Table 1). Interestingly, there is also very recent evidence from preclinical studies showing synergistic antileukemic effects when an IAP antagonist was combined with 5-azacytidine (54).

Accompanying biomarker studies have demonstrated target inhibition by IAP antagonists by showing depletion of cIAP1 protein levels both in surrogate tissues such as peripheral blood mononuclear cells as well as in tumor tissue (51–53). In addition, increased levels of circulating cytokines and chemokines in plasma specimens were used as pharmacodynamic parameters. However, although the sensitivity of cancer cells to monotherapy with IAP antagonists has been linked to their ability to engage an autocrine/paracrine loop of TNF-α production, corresponding in vivo studies largely failed to detect a substantial increase in TNF-α levels in the circulation (19). This may be explained by local production of TNF-α rather than by its widespread release in the circulation. Also, additional cytokines or chemokines may be relevant for the antitumor activity of IAP antagonists. Although these markers may serve as indicators of target antagonism, they will likely not be suitable to predict treatment response.

Pharmacodynamic assays to determine treatment response include, for example, the detection of apoptotic markers in tumor biopsies such as cleavage products of caspases-3, -8, and PARP. However, the determination of markers of apoptotic cell death may not be sufficient to properly assess treatment response, because under certain conditions IAP antagonists can also trigger alternative forms of cell death besides apoptosis, for example necroptosis (49, 50).

In summary, IAP proteins represent promising targets for the development of small-molecule cancer therapeutics, as they are expressed at aberrantly high levels in multiple human malignancies and as they block cell death pathways while supporting cancer cell survival. The therapeutic potential of IAP antagonists may particularly reside in rational combinations together with other cytotoxic strategies to take advantage of additive or synergistic drug interactions. This has convincingly been demonstrated in various preclinical cancer models. In addition, clinical studies have recently been launched to test this concept. As many of these clinical trials are currently ongoing, it is too early to draw conclusions on the clinical efficacy of IAP antagonists.

The author thanks C. Hugenberg for his expert secretarial assistance.

This work has been partially supported by grants from the Deutsche Forschungsgemeinschaft, the Ministerium für Bildung und Forschung (01GM1104C), European Community, IUAP, Wilhelm Sander-Stiftung, and Jose Carreras-Stiftung.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.