A family of anti-apoptotic regulators known as IAP (inhibitor of apoptosis) proteins interact with multiple cellular partners and inhibit apoptosis induced by a variety of stimuli. c-IAP (cellular IAP) 1 and 2 are recruited to TNFR1 (tumour necrosis factor receptor 1)-associated signalling complexes, where they mediate receptor-induced NF-κB (nuclear factor κB) activation. Additionally, through their E3 ubiquitin ligase activities, c-IAP1 and c-IAP2 promote proteasomal degradation of NIK (NF-κB-inducing kinase) and regulate the non-canonical NF-κB pathway. In the present paper, we describe a novel ubiquitin-binding domain of IAPs. The UBA (ubiquitin-associated) domain of IAPs is located between the BIR (baculovirus IAP repeat) domains and the CARD (caspase activation and recruitment domain) or the RING (really interesting new gene) domain of c-IAP1 and c-IAP2 or XIAP (X-linked IAP) respectively. The c-IAP1 UBA domain binds mono-ubiquitin and Lys48- and Lys63-linked polyubiquitin chains with low-micromolar affinities as determined by surface plasmon resonance or isothermal titration calorimetry. NMR analysis of the c-IAP1 UBA domain–ubiquitin interaction reveals that this UBA domain binds the classical hydrophobic patch surrounding Ile44 of ubiquitin. Mutations of critical amino acid residues in the highly conserved MGF (Met-Gly-Phe) binding loop of the UBA domain completely abrogate ubiquitin binding. These mutations in the UBA domain do not overtly affect the ubiquitin ligase activity of c-IAP1 or the participation of c-IAP1 and c-IAP2 in the TNFR1 signalling complex. Treatment of cells with IAP antagonists leads to proteasomal degradation of c-IAP1 and c-IAP2. Deletion or mutation of the UBA domain decreases this degradation, probably by diminishing the interaction of the c-IAPs with the proteasome. These results suggest that ubiquitin binding may be an important mechanism for rapid turnover of auto-ubiquitinated c-IAP1 and c-IAP2.

INTRODUCTION

The IAP (inhibitor of apoptosis) proteins are a family of cell-death regulators found in viruses and metazoans. IAPs can interact directly with a variety of inducers and effectors of apoptosis, and can block apoptosis induced by diverse stimuli [13]. This places IAPs in a central position as inhibitors of death signals that proceed through a number of different pathways. The IAPs contain one to three zinc-binding BIR (baculovirus IAP repeat) domains that are required for anti-apoptotic activity [4,5]. Most of them also possess C-terminal RING (really interesting new gene) domains that function as ubiquitin ligases [5,6]. Some IAPs, such as c-IAP (cellular IAP) 1 and 2, possess a CARD (caspase activation and recruitment domain) as well [7]. c-IAP1 and c-IAP2 were originally identified through their ability to interact directly with TRAF2 [TNF (tumour necrosis factor)-associated factor 2] [8,9]. Through TRAF2 interactions, c-IAP1 and c-IAP2 are recruited to TNFR (TNF receptor) 1- and 2-associated complexes, where they regulate receptor-mediated signalling [10,11]. c-IAP1 and c-IAP2 are critical regulators of NF-κB (nuclear factor κB) signalling pathways [12]. In the TNFα-induced canonical NF-κB pathway, c-IAP1 and c-IAP2 are required for RIP1 (receptor-interacting protein 1) ubiquitination and NF-κB activation [1315]. In the non-canonical NF-κB pathway, c-IAP1 and c-IAP2 ubiquitinate NIK (NF-κB-inducing kinase), leading to its proteasomal degradation and abrogation of NF-κB signalling [16].

The regulated degradation and modification of cellular proteins by the ubiquitin–proteasome system affects a range of vital cellular processes in both normal and tumour cells [17]. An increasing body of work suggests that E3 ubiquitin ligase activity is an essential function of IAPs [6,18]. IAPs are capable of promoting ubiquitination and subsequent proteasomal degradation of themselves and several of their binding partners, including TRAF2 and NIK [16,19,20]. A Drosophila IAP, DIAP1, inhibits cell death, at least in part, by mediating ubiquitination and proteasomal degradation of the Drosophila caspases and IAP antagonistic proteins Reaper, Hid and Grim [21,22]. IAP antagonists promote auto-ubiquitination and proteasomal degradation of c-IAP1 and c-IAP2 in a process that is very efficient and extremely rapid [13,16,23]. Proteasome inhibitors can block the degradation induced by IAP antagonists, but questions still remain regarding the regulation of c-IAP1 and c-IAP2 proteasomal degradation.

In the present study, we have identified the existence of a UBA (ubiquitin-associated) domain in IAPs and characterized its biochemical properties. Our results show that the UBA domain of c-IAP1 binds both mono-ubiquitin and Lys48- and Lys63-linked polyubiquitin chains, with the highly conserved MGF (Met-Gly-Phe) binding loop of the UBA domain playing a critical role in this interaction. The UBA domains of c-IAP1 and c-IAP2 are not critical for their engagement in TNF signalling complexes or for their E3 ubiquitin ligase activities. We demonstrate that the UBA domain plays an important role in IAP antagonist-stimulated proteasomal degradation of c-IAP1 and c-IAP2 by facilitating their recruitment to the proteasome.

EXPERIMENTAL

Plasmids, antibodies and immunoprecipitations

Plasmids expressing FLAG–c-IAP1, FLAG–c-IAP2, mutant FLAG–c-IAP1 and FLAG–c-IAP2 constructs that lack the ability to bind TRAF2 (T2bm), FLAG–c-IAP1 RING mutant (H588A), Smac (second mitochondrial-derived activator of caspase)–Myc, TRAF2–Myc, HA (haemagglutinin)–c-IAP2/MALT1 (mucosa-associated lymphoid tissue protein 1) (case2), and HA–case2 delBIR1 have been described previously [16,24]. Deletions and point mutations of c-IAP1 and c-IAP2 were generated by PCR and subcloned into the p3XFLAG-CMV-14 vector (Sigma) or pEF6/V5-His vector (Invitrogen). Site-specific mutants in c-IAP1, c-IAP2 and case2 were generated using a QuikChange® site-directed mutagenesis kit (Stratagene). All constructs were verified by sequencing of the entire coding region. Human recombinant soluble TNFα was from Genentech. The primary antibodies against c-IAP1 were purchased from R&D Systems (affinity-purified goat antibody) or PTG; anti-ubiquitin antibody was from Cell Signaling Technology; anti-TRAF2 antibodies were from Santa Cruz Biotechnology; anti-TNFR1 antibodies were from R&D Systems; anti-TRADD (TNF-associated death domain) and anti-RIP1 antibodies were from BD Biosciences; anti-FLAG M2 antibody was from Sigma, anti-HA antibody was from Covance, anti-RPN13 (regulatory particle non-ATPase 13) antibody was from Biomol; anti-Myc antibody was from Roche; anti-β-tubulin antibody was from ICN Biomedicals. MG132 was purchased from Calbiochem, and IAP antagonist BV6 [16] was from Genentech. Western blot analyses and immunoprecipitations were performed as described previously [25].

Cell lines, transfections, NF-κB activity assay and viability assays

HEK (human embryonic kidney)-293T cells, A2058 melanoma cells and HT1080 human fibrosarcoma cells were obtained from A.T.C.C. (Manassas, VA, U.S.A.); KMS18 multiple myeloma cells were from the Japanese Collection of Research Bioresources (JCRB). All cell lines were grown in 50:50 Dulbecco's modified Eagle's and FK12 medium supplemented with 10% FBS (fetal bovine serum), penicillin and streptomycin. HEK-293T, A2058 and HT1080 cells were transfected with FuGENE6 reagent (Roche). KMS18 cells were transfected by electroporation using Amaxa Nucleofactor technology (Lonza). HEK-293T cells were transfected with the indicated constructs, and NF-κB luciferase activity was measured using the dual-luciferase reporter assay (Promega) according to the manufacturer's instructions. Viability of KMS18 cells was measured using the CellTiter-Glo Luminescent Cell Viability Assay (Promega).

Recombinant protein production

Polyubiquitin chains were purchased from Boston Biochem. Mono-ubiquitin was bacterially expressed in a pET15b vector using either TB (tryptone broth) medium [enriched LB (Luria–Broth) medium] or minimal medium containing 15NH4Cl and [13C]glucose to produce 13C/15N-isotopically labelled protein. The c-IAP1 UBA domain (residues 379–445), CARD (residues 454–542 or 454–550), BIR3–UBA domain (residues 241–445) and UBA domain–CARD (residues 379–550), as well as c-IAP2 UBA domain constructs (residues 365–431), were subcloned into either the pET15b or pET28a vectors and expressed in bacteria using LB, TB or 15N-isotopically enriched minimal medium. Induction of protein expression was either carried out using manual induction with 1 mM IPTG (isopropyl β-D-thiogalactoside) or auto-induction using Overnight Express™ medium (Novagen). Purification of all expressed proteins used either a two- or three-step procedure, either affinity chromatography on Ni-NTA (Ni2+-nitrilotriacetate) resin (Qiagen) followed by size-exclusion chromatography (Superdex 75 16/60 or Superdex 200 16/60 columns; GE Healthcare) performed on an ÄKTA Explorer FPLC machine, or affinity chromatography on Ni-NTA followed by thrombin cleavage of the His6 affinity tag, dialysis and size-exclusion chromatography. PBS (pH 6.7–7.4), TBS (Tris-buffered saline; pH 7.2–7.4) and 20 mM Mes/150 mM NaCl (pH 6.0) were used as buffers. Full-length c-IAP1, c-IAP1 (M402A/F404A), c-IAP1 RGmt (H588A) and c-IAP1 ΔC7 (del 612–618) proteins were produced as described previously [16].

Ubiquitination assays and MS analysis

Ubiquitination reactions were performed at 17 °C as described previously [15,16] and analysed by the ubiquitin–AQUA (absolute quantification) method as described previously [26], with minor variations. Destained and dehydrated gel pieces from SDS/PAGE were swollen on ice for 20 min in the presence of 20 ng/μl trypsin in 50 mM ammonium bicarbonate/5% acetonitrile. Gel pieces were pulverized using a Teflon pestle, then digested for 16 h at 37 °C. Ubiquitin–AQUA peptide mixture was prepared in 10% acetonitrile/5% methanoic (formic) acid with each peptide at 250 fmol/μl. From this stock, 1500 fmol of ubiquitin–AQUA peptide was added to each sample. Peptides were extracted once using 2 gel vol. of 50% acetonitrile/5% methanoic acid, and samples were dried completely. Dry peptides were resuspended in 10% acetonitrile/5% methanoic acid/0.01% H2O2 at least 30 min before analysis. Peptides were loaded on to a 2.1 mm×150 mm Aquasil (3 μm particle size) C18 column at 200 μl/min and separated with a multi-stage gradient of buffer B (98% acetonitrile/0.1% methanoic acid) in buffer A (2% acetonitrile/0.1% methanoic acid) as follows: 5–12.5% buffer B (0–3 min); 12.5–25% buffer B (3–18 min); 25–65% buffer B (18–22 min); 65–85% buffer B (22–23 min). Multiple reaction monitoring was performed on a QTrap4000 (Applied Biosystems) mass spectrometer with dwell times of 90–220 ms per transition in a segmented run with four to fourteen multiple reaction monitoring transitions per segment. The area of each analyte and internal standard peak was used to determine the abundance of each peptide in the sample. Total ubiquitin abundance was determined as the average of values quantified from the Lys48, Lys63 and Lys11 loci.

SPR (surface plasmon resonance) measurements

SPR experiments were performed on a Biacore 3000 SPR machine using PBS with 0.005% Tween 20 and 0.01% sodium azide as a running buffer. Ubiquitin chains were immobilized directly on to CM5 chips using standard amine coupling chemistry. Measurements of ubiquitin–UBA domain interactions by SPR used a chip with approx. 1500, 7500 and 5000 RU (resonance units) of mono-ubiquitin, Lys48- or Lys63-linked polyubiquitin respectively, with the c-IAP1 UBA domain injected in a series of concentrations from 100 nM to 100 μM, and other constructs injected over the concentration range 500 nM–25 μM. Data were analysed using BiaEvaluation (GE Healthcare) and Scrubber 2 (BioLogic Software). Binding to mono-ubiquitin generally reached saturation rapidly and was ideal for performing equilibrium binding analysis. Binding kinetics of the UBA domain constructs to both Lys48- and Lys63-linked polyubiquitin chains could be fitted using a 1:1 binding model with a drifting baseline.

ITC (isothermal titration calorimetry) and CD

ITC experiments were performed on a Microcal VP-ITC calorimeter. Ubiquitin (945 mM) was titrated into the cell containing the c-IAP1 UBA domain (47 μM) in PBS. A total of 25 sequential 8 μl injections were made at 30 °C, with a stirring speed of 300 rev./min and a dwell time of 300 s. Control experiments in which ubiquitin was titrated into buffer alone revealed no significant enthalpy of dilution. CD spectra were acquired on an Aviv AV-215 spectropolarimeter. Spectra were acquired between 198 and 250 nm at 25 °C, scanning every 1 nm, with a 1 s averaging time, and corrected for the buffer signal. Mean residue molar ellipticity [θ] was calculated as follows:

 
formula

where θobs is the observed ellipticity in millidegrees, MRM is the total molecular mass of the protein divided by the number of amino acid residues, l is the optical pathlength in cm, and c is the final protein concentration in mg/ml.

NMR spectroscopy

1H-, 13C- and 15N-NMR spectra were acquired on a Bruker NMR spectrometer with a 1H Larmor frequency of 800 MHz. Sample buffers consisted of 20 mM Mes (pH 6), 150 mM NaCl in 90% 1H2O/10% 2H2O, with DSS {3-(trimethylsilyl)-[1,1,2,2,3,3-2H6]propanesulfonic acid} as an internal reference. Data were processed using TopSpin (Bruker) and analysed using CcpNmr Analysis (http://www.ccpn.ac.uk/software/ccpnmr-analysis/introduction).

RESULTS

Analysis of the primary structural requirements for IAP antagonist-stimulated proteasomal degradation of c-IAP1

To examine the primary structural requirements for IAP antagonist-stimulated proteasomal degradation of c-IAP1, a series of constructs was generated in which the functional region of each domain was mutated or the entire domain deleted (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/417/bj4170149add.htm). The stability of these constructs was investigated in HEK-293T cells using the IAP antagonist BV6 as a stimulus for degradation [16]. Consistent with previous reports [16], the wild-type c-IAP1 protein is fully degraded following treatment with BV6 (5 μM for 1 h) (Figure 1A). In addition, mutations in the BIR1 domain that prevent c-IAP1 from interacting with TRAF2 (T2bm), or mutations in the peptide-binding groove of the BIR2 domain of c-IAP1 (B2E/A or B2ED/AA), had only minor stabilizing effects on BV6-stimulated degradation of c-IAP1 (Figure 1A); the latter results are consistent with the low affinity of BV6 for the BIR2 domain of c-IAP1 [16]. In contrast, mutations in the peptide/BV6-binding groove of the BIR3 domain of c-IAP1 (B3E/A, B3D/A and B3ED/AA) completely blocked BV6-stimulated degradation (Figure 1A), in agreement with previous results [16]. The RING domain mutant (RGmt; H588A) of c-IAP1 also was not degraded following BV6 treatment, confirming the importance of auto-ubiquitination for this degradative process (Figure 1B). Similarly, deletion of the seven C-terminal amino acids of c-IAP1 (ΔC7), comprising a motif that is proposed to be critical for IAP dimerization and E3 ubiquitin ligase activity [2729], resulted in resistance to BV6-stimulated degradation (Figure 1B). The inability of c-IAP1 RGmt and ΔC7 constructs to mediate auto-ubiquitination was confirmed in a reconstituted ubiquitination assay (see Supplementary Figure S2 at http://www.BiochemJ.org/bj/417/bj4170149add.htm).

Analysis of the primary structural requirements for BV6-stimulated proteasomal degradation of c-IAP1

Figure 1
Analysis of the primary structural requirements for BV6-stimulated proteasomal degradation of c-IAP1

(A) Mutational analysis of the BIR domains of c-IAP1. HEK-293T cells were transiently transfected with the indicated constructs. After 24 h, cells were treated with BV6 (5 μM) for 1 h, and cellular lysates were examined with anti-FLAG antibodies. (B) Mutational analysis of the RING domain of c-IAP1. HEK-293T cells were transiently transfected with the indicated constructs, and cells were treated and lysates examined as in (A). (C) Mutational analysis of the region between the BIR3 and RING domains of c-IAP1. HEK-293T cells were transiently transfected with the indicated constructs, and cells were treated and lysates examined as in (A). W, Western blot.

Figure 1
Analysis of the primary structural requirements for BV6-stimulated proteasomal degradation of c-IAP1

(A) Mutational analysis of the BIR domains of c-IAP1. HEK-293T cells were transiently transfected with the indicated constructs. After 24 h, cells were treated with BV6 (5 μM) for 1 h, and cellular lysates were examined with anti-FLAG antibodies. (B) Mutational analysis of the RING domain of c-IAP1. HEK-293T cells were transiently transfected with the indicated constructs, and cells were treated and lysates examined as in (A). (C) Mutational analysis of the region between the BIR3 and RING domains of c-IAP1. HEK-293T cells were transiently transfected with the indicated constructs, and cells were treated and lysates examined as in (A). W, Western blot.

The effects of mutations in the region of c-IAP1 between the BIR3 and RING domains were also explored (see Supplementary Figure S1). Deletion of the CARD (ΔCARD) had no significant effect on IAP antagonist-stimulated degradation of c-IAP1 (Figure 1C). However, deletion of the region between the BIR3 domain and the CARD (Δ358–454) resulted in stabilization of c-IAP1 in BV6-treated cells (Figure 1C). Therefore, in addition to the BV6-binding BIR3 domain and the E3 ubiquitin ligase RING domain, the region of c-IAP1 between the BIR3 domain and the CARD appears to modulate the stability of c-IAP1.

IAPs possess UBA domains

A PsiBLAST [30] search of the amino acid chain linking the BIR3 domain and the CARD of c-IAP1 shows a region of strong conservation between c-IAP1, c-IAP2, XIAP (X-linked IAP) and ILP2 (IAP-like protein 2), but absent from ML-IAP (melanoma IAP), NAIP (neuronal apoptosis inhibitory protein), survivin and BRUCE (BIR-containing ubiquitin-conjugating enzyme) (see Supplementary Figure S3). Secondary-structure prediction by PsiPRED [31] of this IAP segment shows a distinct core pattern of three α-helices marked by an invariant MGF motif in the turn between α1 and α2 (Figure 2A). Fold recognition alignments by MUSTER and HHPred [32,33] revealed a significant similarity of this region of IAPs to UBA domains, a class of ubiquitin-binding domains with a compact three α-helix fold (Figure 2A) [34]. Notably, most of the known UBA domains contain a highly conserved MGF/Y (Met-Gly-Phe/Tyr) ubiquitin-binding loop that also marks the predicted IAP UBA domains (Figures 2A and 2B). Comparative modelling of these latter UBA domains atop conventional UBA templates with MODELLER [35] shows that they parsimoniously adopt a typical three α-helix structure with the MGF loop, forming a surface hydrophobic patch (e.g. c-IAP1 in Figure 2B). Deletion of the UBA domain (ΔUBA: residues 384–444) rendered c-IAP1 largely resistant to IAP antagonist-stimulated proteasomal degradation (see Supplementary Figure S4 at http://www.BiochemJ.org/bj/417/bj4170149add.htm), thus confirming the results obtained with deletion of the extended region between the BIR3 domain and the CARD.

IAPs possess UBA domains

Figure 2
IAPs possess UBA domains

(A) Structure-guided alignment of the UBA domains from human c-IAP1, c-IAP2, XIAP and ILP2 (sequence boundaries are noted; MGF motifs are highlighted in white-on-black lettering) with several UBA domain folds from the PDB (http://www.rcsb.org), all determined by NMR methods: human STS1 (PDB code 2CPW), human USP (ubiquitin-specific protease) 5 (PDB code 2DAK), Arabidopsis USP14 (PDB code 1VEK), human TDRD3 (tudor domain-containing protein 3) (PDB code 1WJI), human ubiquilin 1 (PDB code 2JY5), human HHR23a (PDB code 1IFY) and human Cbl-b (PDB code 2JNH). (B) Homology model of human c-IAP1 UBA domain with the side chains of the MGF loop and the starting and the ending residues indicated. The top fold recognition hit of the c-IAP1 UBA-like sequence was to PDB code 2CPW (human STS1) with a significant P value of 4.6×10−5. (C) CD spectrum of human c-IAP1 UBA domain (residues 379–445) at a concentration of 100 μM in 20 mM Tris/HCl and 150 mM NaCl (pH 7.0).

Figure 2
IAPs possess UBA domains

(A) Structure-guided alignment of the UBA domains from human c-IAP1, c-IAP2, XIAP and ILP2 (sequence boundaries are noted; MGF motifs are highlighted in white-on-black lettering) with several UBA domain folds from the PDB (http://www.rcsb.org), all determined by NMR methods: human STS1 (PDB code 2CPW), human USP (ubiquitin-specific protease) 5 (PDB code 2DAK), Arabidopsis USP14 (PDB code 1VEK), human TDRD3 (tudor domain-containing protein 3) (PDB code 1WJI), human ubiquilin 1 (PDB code 2JY5), human HHR23a (PDB code 1IFY) and human Cbl-b (PDB code 2JNH). (B) Homology model of human c-IAP1 UBA domain with the side chains of the MGF loop and the starting and the ending residues indicated. The top fold recognition hit of the c-IAP1 UBA-like sequence was to PDB code 2CPW (human STS1) with a significant P value of 4.6×10−5. (C) CD spectrum of human c-IAP1 UBA domain (residues 379–445) at a concentration of 100 μM in 20 mM Tris/HCl and 150 mM NaCl (pH 7.0).

To investigate the biochemical properties of the c-IAP1 UBA domain, single- and double-domain c-IAP1 constructs [BIR3, UBA, UBA (M402A/F404A), CARD, BIR3–UBA and UBA–CARD] were subcloned into expression vectors, expressed in bacteria and purified to homogeneity. The individual UBA domain appeared to be well folded by CD spectroscopy (Figure 2C), and showed reasonable amide resonance dispersion in a 15N-HSQC (heteronuclear single-quantum coherence) NMR spectrum (see Supplementary Figure S5 at http://www.BiochemJ.org/bj/417/bj4170149add.htm). To verify the importance of the MGF loop for the function of the UBA domain, amino acid residues Met402 and Phe404 were mutated to alanine. The c-IAP1 UBA (M402A/F404A) double mutant (MF/AA) had NMR and CD spectra that were similar to those of the wild-type construct, suggesting that these two mutations did not affect the overall protein structure (results not shown).

c-IAP1 UBA domain binds mono-ubiquitin and polyubiquitin chains

Among ubiquitin-binding domains, UBA domains possess some of the highest ubiquitin-binding affinities [34], with some having selective affinity for polyubiquitin chains and others having selective affinity for single ubiquitin domains [36]. A hydrophobic patch on the UBA domains formed by the conserved MGF/Y sequence [37] frequently interacts with the conserved surface hydrophobic patch on ubiquitin formed by Leu8, Ile44 and Val70 [36]. Thus we sought to determine the affinity of the c-IAP1 UBA domain for multiple forms of ubiquitin through a combination of biophysical techniques and mutagenesis.

Ubiquitin–UBA domain interactions have been characterized extensively using NMR spectroscopy [3840–]. Thus NMR methods were used to investigate interactions between ubiquitin and the c-IAP1 UBA domain. Titration of 13C/15N-labelled ubiquitin with unlabelled UBA domain induced distinct ubiquitin chemical shift perturbations that saturated at approx. 1.2–1.4 molar equivalents, and which were consistent with slow exchange, on the NMR chemical shift time scale, between free and UBA-domain-bound states (Figures 3A and 3B). Ubiquitin amide resonances that experience significant chemical shift changes upon binding the UBA domain (Figure 3A) are associated with residues that are clustered around the hydrophobic surface patch (Leu8/Ile44/Val70) of ubiquitin, or with residues neighbouring them in the hydrophobic core. Examination of the methyl region of the 13C-HSQC spectra (Figure 3B) similarly revealed perturbations of resonances associated with the hydrophobic surface patch region (e.g. the δ-methyl groups of Leu8 and Ile44, and the γ-methyl groups of Val70), suggesting strongly that the hydrophobic patch of ubiquitin is the site of the interaction with the c-IAP1 UBA domain (see Supplementary Figure S6 at http://www.BiochemJ.org/bj/417/bj4170149add.htm).

c-IAP1 UBA domain interacts directly with mono-ubiquitin

Figure 3
c-IAP1 UBA domain interacts directly with mono-ubiquitin

(A) Superposition of 15N-HSQC NMR spectra of 13C/15N-labelled ubiquitin with either 0 (red) or 1.2 (blue) molar equivalents of unlabelled c-IAP1 UBA domain. Assignments for resonances that are significantly perturbed upon UBA domain binding are indicated; those shown in grey are located in the hydrophobic core of ubiquitin. (B) Superposition of methyl regions of 13C-HSQC NMR spectra of 13C/15N-labelled ubiquitin; colours and labels are as described for (A). (C) SPR spectra showing the interaction of the c-IAP1 UBA domain with mono-ubiquitin derivatized on to a Biacore CM5 chip (left), and the average response fitted to a 1:1 equilibrium binding model (right). (D) ITC data showing the titration of mono-ubiquitin (945 μM) into c-IAP1 UBA domain (47 μM) (upper) and the normalized enthalpy fitted to a 1:1 binding model (lower).

Figure 3
c-IAP1 UBA domain interacts directly with mono-ubiquitin

(A) Superposition of 15N-HSQC NMR spectra of 13C/15N-labelled ubiquitin with either 0 (red) or 1.2 (blue) molar equivalents of unlabelled c-IAP1 UBA domain. Assignments for resonances that are significantly perturbed upon UBA domain binding are indicated; those shown in grey are located in the hydrophobic core of ubiquitin. (B) Superposition of methyl regions of 13C-HSQC NMR spectra of 13C/15N-labelled ubiquitin; colours and labels are as described for (A). (C) SPR spectra showing the interaction of the c-IAP1 UBA domain with mono-ubiquitin derivatized on to a Biacore CM5 chip (left), and the average response fitted to a 1:1 equilibrium binding model (right). (D) ITC data showing the titration of mono-ubiquitin (945 μM) into c-IAP1 UBA domain (47 μM) (upper) and the normalized enthalpy fitted to a 1:1 binding model (lower).

ITC and SPR measurements were used to determine the overall affinity and selectivity of the c-IAP1 UBA domain for different types of ubiquitin (Figures 3C and 3D and Tables 1 and 2). Isothermal titration of ubiquitin into the c-IAP1 UBA domain yielded a Kd of 5.8±0.3 μM, suggesting a strong affinity for mono-ubiquitin binding. SPR equilibrium affinity measurements of the c-IAP1 UBA domain binding to immobilized mono-ubiquitin gave a slightly weaker Kd estimate of 56±4 μM. No interaction could be observed between the c-IAP1 UBA MF/AA double mutant and mono-ubiquitin by NMR or SPR, suggesting conservation of the canonical ubiquitin-binding interface on the UBA domain [41]. Two-domain c-IAP1 constructs (BIR3–UBA and UBA–CARD) had slightly higher, but similar, equilibrium affinities for mono-ubiquitin (Table 2). SPR was also used to look at the overall binding kinetics of different UBA constructs with polyubiquitin chains linked to Lys48 and Lys63 (Table 1), although the overall accuracy was lower than that of the ITC and equilibrium SPR measurements, with the primary source of uncertainty being determination of the on-rate (ka). As with mono-ubiquitin, the UBA MF/AA double mutant showed no evidence of binding to either type of polyubiquitin chain by SPR measurements (Table 1) or in pull-down assay (see Supplementary Figure S7 at http://www.BiochemJ.org/bj/417/bj4170149add.htm). The binding of the UBA domain to either polyubiquitin linkage was marked by low on-rates (116–350 M−1·s−1) and moderate off-rates (1.6×10−3 s−1, 2.7×10−4 s−1); apparent affinities (1.1–12.7 μM) were similar to that estimated through ITC. Similar results were obtained for binding of BIR3–UBA and UBA–CARD constructs to either polyubiquitin linkage (Table 1). These results suggest that binding of the UBA domain to polyubiquitin is accomplished primarily through binding of one UBA domain to individual ubiquitin domains, with no particular selectivity for polyubiquitin over mono-ubiquitin.

Table 1
Binding kinetics to polyubiquitin chains

Fits are to a 1:1 binding model with a drifting baseline. There was no binding to c-IAP1 UBA (MF/AA).

 Lys48-linked polyubiquitin Lys63-linked polyubiquitin 
Protein ka (M−1 s−1kd (s−1Kd (μM) ka (M−1 s−1kd (s−1Kd (μM) 
c-IAP1 UBA 116±53 (1.57±0.54)×10−3 12.7±1.8 350±220 (2.7±1.5)×10−4 1.1±1.0 
c-IAP1 UBA–CARD 570±180 (8.2±0.1)×10−4 1.5±0.4 350±290 (1.87±0.23)×10−3 9.0±7.1 
c-IAP1 BIR3–UBA (3.2±1.7)×103 (8.3±0.3)×10−3 3.1±1.5 800±340 (4.0±0.2)×10−3 5.6±2.2 
 Lys48-linked polyubiquitin Lys63-linked polyubiquitin 
Protein ka (M−1 s−1kd (s−1Kd (μM) ka (M−1 s−1kd (s−1Kd (μM) 
c-IAP1 UBA 116±53 (1.57±0.54)×10−3 12.7±1.8 350±220 (2.7±1.5)×10−4 1.1±1.0 
c-IAP1 UBA–CARD 570±180 (8.2±0.1)×10−4 1.5±0.4 350±290 (1.87±0.23)×10−3 9.0±7.1 
c-IAP1 BIR3–UBA (3.2±1.7)×103 (8.3±0.3)×10−3 3.1±1.5 800±340 (4.0±0.2)×10−3 5.6±2.2 
Table 2
Equilibrium binding measurements to mono-ubiquitin
 Kd (μM) 
Protein Mono-ubiquitin (ITC) Mono-ubiquitin (SPR) 
c-IAP1 UBA 5.8±0.3 56±4 
c-IAP1 UBA (MF/AA) – No binding 
c-IAP1 UBA–CARD – 21±6 
c-IAP1 BIR3–UBA – 23±5 
 Kd (μM) 
Protein Mono-ubiquitin (ITC) Mono-ubiquitin (SPR) 
c-IAP1 UBA 5.8±0.3 56±4 
c-IAP1 UBA (MF/AA) – No binding 
c-IAP1 UBA–CARD – 21±6 
c-IAP1 BIR3–UBA – 23±5 

The UBA domain does not affect E3 ubiquitin ligase activity of c-IAP1

Ubiquitin ligase activity is critical for most of the signalling activities of c-IAP1 and c-IAP2 [6,12]. To assess the importance of ubiquitin binding for the E3 ubiquitin ligase activity of c-IAP1, recombinant full-length wild-type c-IAP1 and c-IAP1 MF/AA mutant were incubated in a reconstituted ubiquitination assay for 10 or 35 min. The forms of ubiquitin conjugated to c-IAP1 during auto-ubiquitination reactions were analysed using the ubiquitin–AQUA method [26]. Four gel regions from each sample were excised from the Coomassie Blue-stained SDS/PAGE gel, beginning immediately above the unmodified c-IAP1 bands (Figures 4A and 4B). Region D contained the first two visible ubiquitin–c-IAP1 bands, and region B contained the majority of multi- and polyubiquitinated c-IAP1 observed by Western blot against ubiquitin (Figures 4A and 4B). Regardless of the gel region, the composition of the ubiquitin linkages generated was very similar between the wild-type and MF/AA mutant versions of c-IAP1, with Lys11 being the most abundant linkage generated (Figures 4C–4F). Similar linkage profiles have been reported for other Ubc4/UbcH5a substrates, such as cyclin B1 and Murf1 [26,42]. The MF/AA mutant of c-IAP1 was somewhat more active than wild-type in generating high-molecular-mass ubiquitinated species in the longer-running reactions (Figures 4C–4E). Although minimal differences were observed after 10 min, comparison of total ubiquitin abundance in regions B and C for 35 min reactions showed 9.1 and 6.4 pmol respectively for MF/AA c-IAP1 compared with 5.1 and 4.5 pmol respectively for wild-type c-IAP1 (Figures 4D and 4E). As expected, ubiquitin in region D consists primarily of mono- and multi-mono-ubiquitinated forms of c-IAP1 (Figure 4F). On the basis of the low-molecular-mass smear in the Western blot, accumulation of polyubiquitin in Region D after 35 min is best accounted for not by ubiquitin attached to c-IAP1, but rather by free polyubiquitin chains or ubiquitin–UbcH5a (Figures 4B and 4F). Thus the UBA domain of c-IAP1 does not appear to have significant effect on the E3 ubiquitin ligase activity of c-IAP1.

In vitro ubiquitination activity of wild-type and MF/AA mutant c-IAP1

Figure 4
In vitro ubiquitination activity of wild-type and MF/AA mutant c-IAP1

(A) Coomassie Blue-stained gel containing equivalent amounts of in vitro c-IAP1 auto-ubiquitination reaction mixtures (10 and 35 min) performed using wild-type (WT) or MF/AA mutant (MF) proteins. Negative control reactions were performed for 35 min in the absence of ubiquitin (N). Arrowhead indicates unmodified c-IAP1 doublet band. Red broken lines mark the boundaries of four gel regions that were excised and subjected to in-gel trypsin digestion, followed by ubiquitin–AQUA analysis. (B) Western blot (W) using anti-ubiquitin antibody (P4D1) of in vitro auto-ubiquitination reactions. (CF) Stacked histograms display the amount of total ubiquitin in each corresponding gel region for each sample. The individual amounts of Lys48- (red), Lys63- (blue), Lys11- (green) and Lys6- (orange) linked chains are indicated by their respective colours. Mono-ubiquitin and chain end caps are indicated in yellow. MW, molecular mass; Ub, ubiquitin.

Figure 4
In vitro ubiquitination activity of wild-type and MF/AA mutant c-IAP1

(A) Coomassie Blue-stained gel containing equivalent amounts of in vitro c-IAP1 auto-ubiquitination reaction mixtures (10 and 35 min) performed using wild-type (WT) or MF/AA mutant (MF) proteins. Negative control reactions were performed for 35 min in the absence of ubiquitin (N). Arrowhead indicates unmodified c-IAP1 doublet band. Red broken lines mark the boundaries of four gel regions that were excised and subjected to in-gel trypsin digestion, followed by ubiquitin–AQUA analysis. (B) Western blot (W) using anti-ubiquitin antibody (P4D1) of in vitro auto-ubiquitination reactions. (CF) Stacked histograms display the amount of total ubiquitin in each corresponding gel region for each sample. The individual amounts of Lys48- (red), Lys63- (blue), Lys11- (green) and Lys6- (orange) linked chains are indicated by their respective colours. Mono-ubiquitin and chain end caps are indicated in yellow. MW, molecular mass; Ub, ubiquitin.

Importance of the UBA domains for c-IAP1- and c-IAP2-mediated signalling

To investigate whether the UBA domain of c-IAP2 is required for its recruitment to the TNFR1 signalling complex, HT1080 cells were transfected with constructs expressing vector, wild-type c-IAP2 or MF/AA mutant c-IAP2. Treatment of cells with TNFα triggered formation of the receptor-associated protein complex and recruitment of wild-type c-IAP2 and MF/AA mutant (Figure 5A). The importance of the UBA domain for binding to TRAF2 and Smac, two established binding partners for c-IAPs, was also tested. Mutations in the TRAF2-binding region of the BIR1 domain of c-IAP2 [24] abrogated interaction with TRAF2, whereas deletion or mutation in the UBA domain of c-IAP2 had no effect on binding (see Supplementary Figure S8A at http://www.BiochemJ.org/bj/417/bj4170149add.htm). In an analogous fashion, mutation of a residue critical for interaction with Smac [43] (Asp306 to alanine) prevented binding to Smac, whereas deletion or mutation in the UBA domain of c-IAP2 had no effect (see Supplementary Figure S8B). Similar results were obtained for c-IAP1 UBA domain mutants (results not shown). Thus the UBA domain is not involved in TNFR1 protein complex assembly and is not necessary for association of c-IAP1 and c-IAP2 with TRAF2 or Smac.

Importance of the UBA domain for c-IAP1- and c-IAP2-mediated signalling

Figure 5
Importance of the UBA domain for c-IAP1- and c-IAP2-mediated signalling

(A) The UBA domain of c-IAP2 is not required for recruitment into the TNFR1 signalling complex. HT1080 cells were transiently transfected with the indicated c-IAP2 constructs and treated with TNFα for 5 or 20 min. Cell lysates were immunoprecipitated with anti-TNFR1 antibodies and protein levels in cellular lysates and in the TNFR1-associated complex were determined by Western blotting with the indicated antibodies. (B) The UBA domain of c-IAP1 and c-IAP2 is not required for the stimulation of NIK degradation. HEK-293T cells were transiently transfected with the indicated constructs, and 24 h later, cellular lysates were examined with anti-Myc or anti-FLAG antibodies. (C) Expression of wild-type c-IAP1 and c-IAP2, but not of their MF/AA mutants, inhibits cell death induced by TNFα in the presence of cycloheximide. KMS18 cells were transfected with the indicated constructs and, 48 h later, were treated with TNFα and cycloheximide (1 μg/ml). Cell viability was determined as described in the Experimental section, and protein levels were determined by Western blotting as in (A). Results are means±S.D. for three independent experiments. IP, immunoprecipitation; W, Western blot.

Figure 5
Importance of the UBA domain for c-IAP1- and c-IAP2-mediated signalling

(A) The UBA domain of c-IAP2 is not required for recruitment into the TNFR1 signalling complex. HT1080 cells were transiently transfected with the indicated c-IAP2 constructs and treated with TNFα for 5 or 20 min. Cell lysates were immunoprecipitated with anti-TNFR1 antibodies and protein levels in cellular lysates and in the TNFR1-associated complex were determined by Western blotting with the indicated antibodies. (B) The UBA domain of c-IAP1 and c-IAP2 is not required for the stimulation of NIK degradation. HEK-293T cells were transiently transfected with the indicated constructs, and 24 h later, cellular lysates were examined with anti-Myc or anti-FLAG antibodies. (C) Expression of wild-type c-IAP1 and c-IAP2, but not of their MF/AA mutants, inhibits cell death induced by TNFα in the presence of cycloheximide. KMS18 cells were transfected with the indicated constructs and, 48 h later, were treated with TNFα and cycloheximide (1 μg/ml). Cell viability was determined as described in the Experimental section, and protein levels were determined by Western blotting as in (A). Results are means±S.D. for three independent experiments. IP, immunoprecipitation; W, Western blot.

Recent studies identified c-IAP1 and c-IAP2 as ubiquitin ligases responsible for the ubiquitination and proteasomal degradation of NIK [16]. MF/AA mutants of c-IAP1 and c-IAP2 were therefore examined for their ability to stimulate proteasomal degradation of NIK and found to be equally potent as their wild-type counterparts (Figure 5B). Since NIK degradation relies on the E3 ubiquitin ligase activity of the c-IAPs, these results are not surprising in the light of the findings presented in Figure 4 that mutations in the UBA domain do not overtly affect the ubiquitin ligase activity of c-IAP1.

c-IAP2 undergoes a genetic translocation t(11;18)(q21;q21) that fuses its BIR domains with paracaspase/MALT1 [44,45]. The c-IAP2–MALT1 fusion protein (case2) constitutively activates the NF-κB pathway, a potentially seminal activity for development of inflammation-associated tumours [46,47]. To investigate whether mutations in the UBA domain affect NF-κB activation by case2, wild-type and MF/AA mutant case2 were transfected in HEK-293T cells along with the luciferase reporter plasmid (see Supplementary Figure S9A at http://www.BiochemJ.org/bj/417/bj4170149add.htm). Wild-type and MF/AA mutant case2 stimulated NF-κB activation to comparable extents, whereas deletion of the BIR1 domain severely blunted potency, in agreement with previous results [24,48] (see Supplementary Figure S7A). c-IAP2–MALT1 fusion protein is postulated to mediate NF-κB activation by ubiquitinating NEMO (NF-κB essential modulator) [48]. Overexpression of wild-type or MF/AA mutant case2 resulted in similar enhancement of NEMO ubiquitination (see Supplementary Figure S9B). Therefore the ubiquitin-binding activity of the UBA domain of c-IAP2–MALT1 fusion protein does not appear to be critical for the activation of NF-κB or for NEMO ubiquitination.

To determine the relevance of ubiquitin binding for the anti-apoptotic activity of the c-IAPs, wild-type and MF/AA mutant c-IAP1 and c-IAP2 were transiently expressed in KMS18 multiple myeloma cells. These cells have a genetic ablation in the locus for c-IAPs [49,50] and thus allowed us to examine their anti-apoptotic role in the absence of endogenous c-IAPs. Expression of wild-type c-IAP1 or c-IAP2 efficiently reduced cell death caused by TNFα in the presence of cyclohexamide, while the MF/AA mutants did not provide significant protection against this apoptotic stimulus (Figure 5C). Thus mutations in the UBA domain and abrogation of ubiquitin binding diminish the anti-apoptotic activity of c-IAPs without affecting their recruitment to the TNFR1-associated signalling complex.

The UBA domains of c-IAP1 and c-IAP2 facilitate recruitment to the proteasome

Findings that deletion of the UBA domain affects IAP antagonist-stimulated proteasomal degradation of c-IAP1 and c-IAP2 (Figure 1 and Supplementary Figure S3) prompted us to investigate whether mutations in amino acid residues critical for binding ubiquitin (MF/AA) would have similar effects. To that end, wild-type and MF/AA mutant constructs of c-IAP1 and c-IAP2 were ectopically expressed in several cell lines and their stability following IAP antagonist treatment was investigated. In all cell lines, MF/AA mutants showed significantly decreased degradation compared with wild-type c-IAPs (Figure 6A). This difference appeared to be more pronounced in the case of c-IAP2, which is generally more stable than c-IAP1 [16] (Figure 6A).

Stability and association of c-IAPs and their MF/AA mutants with the proteasome

Figure 6
Stability and association of c-IAPs and their MF/AA mutants with the proteasome

(A) MF/AA mutants of c-IAP1 and c-IAP2 are more stable than the corresponding wild-type c-IAPs after exposure to IAP antagonists. A2058 and HT1080 cells were transiently transfected with the indicated c-IAP constructs and, 48 h later, were treated with BV6 (5 μM) for 30 min. Cellular lysates were examined with anti-FLAG antibodies. (B) MF/AA mutant c-IAP1 is recruited to the proteasome less efficiently than wild-type c-IAP1 after treatment with IAP antagonist BV6. HT1080 and 293T cells were transiently transfected with the indicated c-IAP1 constructs and, 48 h later, were treated with BV6 (5 μM) for 10 min in the presence of MG132. Whole-cell lysates (WCL) were immunoprecipitated with anti-FLAG antibodies, and protein levels in cellular lysates and in the immunoprecipitates were determined by Western blotting with the indicated antibodies. IP, immunoprecipitation; W, Western blot.

Figure 6
Stability and association of c-IAPs and their MF/AA mutants with the proteasome

(A) MF/AA mutants of c-IAP1 and c-IAP2 are more stable than the corresponding wild-type c-IAPs after exposure to IAP antagonists. A2058 and HT1080 cells were transiently transfected with the indicated c-IAP constructs and, 48 h later, were treated with BV6 (5 μM) for 30 min. Cellular lysates were examined with anti-FLAG antibodies. (B) MF/AA mutant c-IAP1 is recruited to the proteasome less efficiently than wild-type c-IAP1 after treatment with IAP antagonist BV6. HT1080 and 293T cells were transiently transfected with the indicated c-IAP1 constructs and, 48 h later, were treated with BV6 (5 μM) for 10 min in the presence of MG132. Whole-cell lysates (WCL) were immunoprecipitated with anti-FLAG antibodies, and protein levels in cellular lysates and in the immunoprecipitates were determined by Western blotting with the indicated antibodies. IP, immunoprecipitation; W, Western blot.

Since ubiquitinated proteins destined for degradation are recruited to the proteasome, and the UBA domain mutants of the c-IAPs are not degraded as efficiently as their wild-type counterparts, we explored the recruitment of the UBA mutant proteins to the proteasome [51,52]. For this purpose, we ectopically expressed c-IAP1 or its MF/AA mutant and investigated their association with the proteasome receptor RPN13 [53,54] following IAP antagonist treatment in the presence of MG132 to prevent rapid degradation of the wild-type protein (Figure 6B). Treatment of cells with IAP antagonist BV6 promoted interaction of c-IAP1 with endogenous RPN13 (Figure 6B). In both cell lines examined, c-IAP1 association with RPN13 appeared to be more efficient than for the MF/AA mutant (Figure 6B). In a similar fashion, we observed much stronger binding to RPN10 with wild-type c-IAP1 compared with its MF/AA mutant (results not shown). Collectively, these results suggest that the UBA domains of c-IAPs play an important role in their stability and recruitment to the proteasome following exposure to IAP antagonists.

DISCUSSION

One of the most prominent features of IAP antagonists is their capacity to promote auto-ubiquitination and subsequent proteasomal degradation of c-IAP1 and c-IAP2. Our search for structural requirements within c-IAP1 that enable this rapid process identified a novel domain in the IAP family of anti-apoptotic proteins. The UBA domain gives IAPs the ability to bind ubiquitin and extends the list of their interaction partners. It also places IAPs in the ever-growing group of proteins whose binding to ubiquitin modulates their activity and their cellular fate [5557]. Interestingly, the UBA domain of c-IAP1 binds mono-ubiquitin and Lys48- and Lys63-linked polyubiquitin chains with comparable affinities. The selectivity of some UBA domains for relatively linear Lys63-linked polyubiquitin chains has recently been postulated to be a consequence of strong affinity for mono-ubiquitin [3840]. The ability to bind mono-ubiquitin as well as polyubiquitin chains suggests a wide variety of possible interactions that IAPs can engage through ubiquitin binding.

Investigation of various signalling activities for c-IAP1 and c-IAP2 showed that the UBA domain and its ability to bind ubiquitin are largely dispensable for the recruitment of c-IAPs to the TNFR1 signalling complex or for NF-κB activation. We have, however, observed diminished anti-apoptotic activity of c-IAP1 and c-IAP2 UBA domain mutants, suggesting the possibility that the UBA domain enables interaction of c-IAPs with novel binding partners. Future work involving proteomics and possibly screens of siRNA (small interfering RNA) libraries might identify such novel interactors and help to elucidate a mechanism for this observation. Equally tantalizing will be efforts to elucidate the effect of the XIAP UBA domain on caspase inhibition.

IAP antagonist-stimulated auto-ubiquitination of c-IAP1 and c-IAP2 leads to rapid proteasomal degradation of these proteins. This process requires, among other things, binding of IAP antagonists to the BIR3 domains of the c-IAPs, E3 ubiquitin ligase activity that will mediate auto-ubiquitination and efficient recruitment of ubiquitinated proteins to the proteasome machinery. The ability of the UBA domain to modulate this process implies that this domain should be able to alter one (or more) of these steps. We have shown that the UBA domain does not interfere with Smac binding, indicating that it does not affect binding of Smac mimetics/IAP antagonists to the c-IAPs. The ability of c-IAP1 and c-IAP2 wild-type and MF/AA mutants to promote NIK degradation to similar extents, together with detailed analysis of their E3 ubiquitin ligase activity in vitro, demonstrated that the UBA domains do not have significant roles in the ubiquitin ligase activity of c-IAPs. Nevertheless, MF/AA mutant c-IAPs were more stable following treatment with IAP antagonists than the wild-type proteins. These findings indicated that the mutant proteins were possibly not recruited to the proteasome very efficiently and that the UBA domains of c-IAPs might facilitate efficient degradation. Indeed, upon investigation of this possibility, we found that c-IAP1 with a mutated UBA domain was poorly recruited to the proteasome after exposure to the IAP antagonist BV6. In addition to testing the binding of ubiquitinated c-IAPs to proteasome receptors RPN10 and RPN13, we also examined the binding of ubiquitinated c-IAPs to shuttle receptors PLIC (protein linking IAP with cytoskeleton) 1 and 2 [51,52]. However, we could not detect their association with c-IAPs, possibly due to the low affinities of existing antibodies. Earlier studies have indicated that UBA domains can affect proteasomal degradation by regulating the access of the proteasome to the substrate [51,5860]. In a similar fashion, the UBA domains of the c-IAPs might allow fine-tuning of this critical process by modulating its timing and substrate-delivery priority. Proteasomal degradation of ubiquitinated c-IAP1 and c-IAP2 is an extremely rapid event, making it difficult to dissect various steps in the process. Nevertheless, our findings suggest that the UBA domains play an integral role in the recruitment of c-IAP1 and c-IAP2 to the proteasomal machinery and provide a foundation for future studies of IAP antagonist-mediated degradation of the c-IAPs.

We thank Ingrid Wertz, Ivan Bosanac, Borlan Pan, Pascal Meier, Kim Newton, Nobuhiko Kayagaki, Karen O'Rourke, Vishva Dixit, members of the Protein Engineering and Physiological Chemistry departments and the Oligo Synthesis and Sequencing facilities who provided help with insightful discussions, suggestions and reagents.

Abbreviations

     
  • AQUA

    absolute quantification

  •  
  • CARD

    caspase activation recruitment domain

  •  
  • HA

    haemagglutinin

  •  
  • HEK

    human embryonic kidney

  •  
  • HSQC

    heteronuclear single-quantum coherence

  •  
  • IAP

    inhibitor of apoptosis

  •  
  • BIR

    baculovirus IAP repeat

  •  
  • c-IAP

    cellular IAP

  •  
  • ILP2

    IAP-like protein 2

  •  
  • ITC

    isothermal titration calorimetry

  •  
  • LB

    Luria–Bertani

  •  
  • MALT1

    mucosa-associated lymphoid tissue protein 1

  •  
  • NF-κB

    nuclear factor κB

  •  
  • NEMO

    NF-κB essential modulator

  •  
  • NIK

    NF-κB-inducing kinase

  •  
  • Ni-NTA

    Ni2+-nitrilotriacetate

  •  
  • RING

    really interesting new gene

  •  
  • RIP1

    receptor-interacting protein 1

  •  
  • RPN

    regulatory particle non-ATPase

  •  
  • Smac

    second mitochondrial-derived activator of caspases

  •  
  • SPR

    surface plasmon resonance

  •  
  • TB

    tryptone broth

  •  
  • TNF

    tumour necrosis factor

  •  
  • TNFR

    TNF receptor

  •  
  • TRAF2

    TNFR-associated factor 2

  •  
  • UBA

    ubiquitin-associated domain

  •  
  • XIAP

    X-linked IAP

References

References
1
Deveraux
Q. L.
Reed
J. C.
IAP family proteins: suppressors of apoptosis
Genes Dev.
1999
, vol. 
13
 (pg. 
239
-
252
)
2
Salvesen
G. S.
Abrams
J. M.
Caspase activation: stepping on the gas or releasing the brakes?. Lessons from humans and flies
Oncogene
2004
, vol. 
23
 (pg. 
2774
-
2784
)
3
Vucic
D.
Targeting IAP (inhibitor of apoptosis) proteins for therapeutic intervention in tumors
Curr. Cancer Drug Targets
2008
, vol. 
8
 (pg. 
110
-
117
)
4
Liston
P.
Fong
W. G.
Korneluk
R. G.
The inhibitors of apoptosis: there is more to life than Bcl2
Oncogene
2003
, vol. 
22
 (pg. 
8568
-
8580
)
5
Salvesen
G. S.
Duckett
C. S.
IAP proteins: blocking the road to death's door
Nat. Rev. Mol. Cell Biol.
2002
, vol. 
3
 (pg. 
401
-
410
)
6
Vaux
D. L.
Silke
J.
IAPs, RINGs and ubiquitylation
Nat. Rev. Mol. Cell Biol.
2005
, vol. 
6
 (pg. 
287
-
297
)
7
Hofmann
K.
Bucher
P.
Tschopp
J.
The CARD domain: a new apoptotic signalling motif
Trends Biochem. Sci.
1997
, vol. 
22
 (pg. 
155
-
156
)
8
Rothe
M.
Pan
M. G.
Henzel
W. J.
Ayres
T. M.
Goeddel
D. V.
The TNFR2–TRAF signaling complex contains two novel proteins related to baculoviral inhibitor of apoptosis proteins
Cell
1995
, vol. 
83
 (pg. 
1243
-
1252
)
9
Rothe
M.
Wong
S. C.
Henzel
W. J.
Goeddel
D. V.
A novel family of putative signal transducers associated with the cytoplasmic domain of the 75 kDa tumor necrosis factor receptor
Cell
1994
, vol. 
78
 (pg. 
681
-
692
)
10
Shu
H. B.
Takeuchi
M.
Goeddel
D. V.
The tumor necrosis factor receptor 2 signal transducers TRAF2 and c-IAP1 are components of the tumor necrosis factor receptor 1 signaling complex
Proc. Natl. Acad. Sci. U.S.A.
1996
, vol. 
93
 (pg. 
13973
-
13978
)
11
Wang
C. Y.
Mayo
M. W.
Korneluk
R. G.
Goeddel
D. V.
Baldwin
A. S.
Jr
NF-κB antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c- IAP2 to suppress caspase-8 activation
Science
1998
, vol. 
281
 (pg. 
1680
-
1683
)
12
Varfolomeev
E.
Vucic
D.
(Un)expected roles of c-IAPs in apoptotic and NFκB signaling pathways
Cell Cycle
2008
, vol. 
7
 (pg. 
1511
-
1521
)
13
Bertrand
M. J.
Milutinovic
S.
Dickson
K. M.
Ho
W. C.
Boudreault
A.
Durkin
J.
Gillard
J. W.
Jaquith
J. B.
Morris
S. J.
Barker
P. A.
cIAP1 and cIAP2 facilitate cancer cell survival by functioning as E3 ligases that promote RIP1 ubiquitination
Mol. Cell
2008
, vol. 
30
 (pg. 
689
-
700
)
14
Mahoney
D. J.
Cheung
H. H.
Mrad
R. L.
Plenchette
S.
Simard
C.
Enwere
E.
Arora
V.
Mak
T. W.
Lacasse
E. C.
Waring
J.
Korneluk
R. G.
Both cIAP1 and cIAP2 regulate TNFα-mediated NF-κB activation
Proc. Natl. Acad. Sci. U.S.A.
2008
, vol. 
105
 (pg. 
11778
-
11783
)
15
Varfolomeev
E.
Goncharov
T.
Fedorova
A. V.
Dynek
J. N.
Zobel
K.
Deshayes
K.
Fairbrother
W. J.
Vucic
D.
c-IAP1 and c-IAP2 are critical mediators of TNFα-induced NF-κB activation
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
24295
-
24299
)
16
Varfolomeev
E.
Blankenship
J. W.
Wayson
S. M.
Fedorova
A. V.
Kayagaki
N.
Garg
P.
Zobel
K.
Dynek
J. N.
Elliott
L. O.
Wallweber
H. J.
, et al. 
IAP antagonists induce autoubiquitination of c-IAPs, NF-κB activation, and TNFα-dependent apoptosis
Cell
2007
, vol. 
131
 (pg. 
669
-
681
)
17
Hershko
A.
Ciechanover
A.
The ubiquitin system
Annu. Rev. Biochem.
1998
, vol. 
67
 (pg. 
425
-
479
)
18
Newton
K.
Vucic
D.
Ubiquitin ligases in cancer: ushers for degradation
Cancer Invest.
2007
, vol. 
25
 (pg. 
502
-
513
)
19
Conze
D. B.
Albert
L.
Ferrick
D. A.
Goeddel
D. V.
Yeh
W. C.
Mak
T.
Ashwell
J. D.
Posttranscriptional downregulation of c-IAP2 by the ubiquitin protein ligase c-IAP1 in vivo
Mol. Cell. Biol.
2005
, vol. 
25
 (pg. 
3348
-
3356
)
20
Li
X.
Yang
Y.
Ashwell
J. D.
TNF-RII and c-IAP1 mediate ubiquitination and degradation of TRAF2
Nature
2002
, vol. 
416
 (pg. 
345
-
347
)
21
Holley
C. L.
Olson
M. R.
Colon-Ramos
D. A.
Kornbluth
S.
Reaper eliminates IAP proteins through stimulated IAP degradation and generalized translational inhibition
Nat. Cell Biol.
2002
, vol. 
14
 pg. 
14
 
22
Wilson
R.
Goyal
L.
Ditzel
M.
Zachariou
A.
Baker
D. A.
Agapite
J.
Steller
H.
Meier
P.
The DIAP1 RING finger mediates ubiquitination of Dronc and is indispensable for regulating apoptosis
Nat. Cell Biol.
2002
, vol. 
14
 pg. 
14
 
23
Vince
J. E.
Wong
W. W.
Khan
N.
Feltham
R.
Chau
D.
Ahmed
A. U.
Benetatos
C. A.
Chunduru
S. K.
Condon
S. M.
McKinlay
M.
, et al. 
IAP antagonists target cIAP1 to induce TNFα-dependent apoptosis
Cell
2007
, vol. 
131
 (pg. 
682
-
693
)
24
Varfolomeev
E.
Wayson
S. M.
Dixit
V. M.
Fairbrother
W. J.
Vucic
D.
The inhibitor of apoptosis protein fusion c-IAP2·MALT1 stimulates NF-κB activation independently of TRAF1 AND TRAF2
J. Biol. Chem.
2006
, vol. 
281
 (pg. 
29022
-
29029
)
25
Vucic
D.
Franklin
M. C.
Wallweber
H. J.
Das
K.
Eckelman
B. P.
Shin
H.
Elliott
L. O.
Kadkhodayan
S.
Deshayes
K.
Salvesen
G. S.
Fairbrother
W. J.
Engineering ML-IAP to produce an extraordinarily potent caspase 9 inhibitor: implications for Smac-dependent anti-apoptotic activity of ML-IAP
Biochem. J.
2005
, vol. 
385
 (pg. 
11
-
20
)
26
Kirkpatrick
D. S.
Hathaway
N. A.
Hanna
J.
Elsasser
S.
Rush
J.
Finley
D.
King
R. W.
Gygi
S. P.
Quantitative analysis of in vitro ubiquitinated cyclin B1 reveals complex chain topology
Nat. Cell Biol.
2006
, vol. 
8
 (pg. 
700
-
710
)
27
Linke
K.
Mace
P. D.
Smith
C. A.
Vaux
D. L.
Silke
J.
Day
C. L.
Structure of the MDM2/MDMX RING domain heterodimer reveals dimerization is required for their ubiquitylation in trans
Cell Death Differ.
2008
, vol. 
15
 (pg. 
841
-
848
)
28
Silke
J.
Kratina
T.
Chu
D.
Ekert
P. G.
Day
C. L.
Pakusch
M.
Huang
D. C.
Vaux
D. L.
Determination of cell survival by RING-mediated regulation of inhibitor of apoptosis (IAP) protein abundance
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
16182
-
16187
)
29
Mace
P. D.
Linke
K.
Feltham
R.
Schumacher
F. R.
Smith
C. A.
Vaux
D. L.
Silke
J.
Day
C. L.
Structures of the cIAP2 ring domain reveal conformational changes associated with E2 recruitment
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
31633
-
31640
)
30
Schaffer
A. A.
Aravind
L.
Madden
T. L.
Shavirin
S.
Spouge
J. L.
Wolf
Y. I.
Koonin
E. V.
Altschul
S. F.
Improving the accuracy of PSI-BLAST protein database searches with composition-based statistics and other refinements
Nucleic Acids Res.
2001
, vol. 
29
 (pg. 
2994
-
3005
)
31
Jones
D. T.
Protein secondary structure prediction based on position-specific scoring matrices
J. Mol. Biol.
1999
, vol. 
292
 (pg. 
195
-
202
)
32
Soding
J.
Protein homology detection by HMM-HMM comparison
Bioinformatics
2005
, vol. 
21
 (pg. 
951
-
960
)
33
Wu
S.
Zhang
Y.
MUSTER: improving protein sequence profile–profile alignments by using multiple sources of structure information
Proteins
2008
, vol. 
72
 (pg. 
547
-
556
)
34
Raasi
S.
Varadan
R.
Fushman
D.
Pickart
C. M.
Diverse polyubiquitin interaction properties of ubiquitin-associated domains
Nat. Struct. Mol. Biol.
2005
, vol. 
12
 (pg. 
708
-
714
)
35
Sali
A.
Blundell
T. L.
Comparative protein modelling by satisfaction of spatial restraints
J. Mol. Biol.
1993
, vol. 
234
 (pg. 
779
-
815
)
36
Hicke
L.
Schubert
H. L.
Hill
C. P.
Ubiquitin-binding domains
Nat. Rev. Mol. Cell Biol.
2005
, vol. 
6
 (pg. 
610
-
621
)
37
Mueller
T. D.
Feigon
J.
Solution Structures of UBA domains reveal a conserved hydrophobic surface for protein–protein interactions
J. Mol. Biol.
2002
, vol. 
319
 (pg. 
1243
-
1255
)
38
Kozlov
G.
Nguyen
L.
Lin
T.
De Crescenzo
G.
Park
M.
Gehring
K.
Structural basis of ubiquitin recognition by the ubiquitin-associated (UBA) domain of the ubiquitin ligase EDD
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
35787
-
35795
)
39
Long
J.
Gallagher
T. R.
Cavey
J. R.
Sheppard
P. W.
Ralston
S. H.
Layfield
R.
Searle
M. S.
Ubiquitin recognition by the ubiquitin-associated domain of p62 involves a novel conformational switch
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
5427
-
5440
)
40
Zhang
D.
Raasi
S.
Fushman
D.
Affinity makes the difference: nonselective interaction of the UBA domain of ubiquilin-1 with monomeric ubiquitin and polyubiquitin chains
J. Mol. Biol.
2008
, vol. 
377
 (pg. 
162
-
180
)
41
Swanson
K. A.
Hicke
L.
Radhakrishnan
I.
Structural basis for monoubiquitin recognition by the Ede1 UBA domain
J. Mol. Biol.
2006
, vol. 
358
 (pg. 
713
-
724
)
42
Newton
K.
Matsumoto
M. L.
Wertz
I. E.
Kirkpatrick
D. S.
Lill
J. R.
Tan
J.
Dugger
D.
Gordon
N.
Sidhu
S. S.
Fellouse
F. A.
, et al. 
Ubiquitin chain editing revealed by polyubiquitin linkage-specific antibodies
Cell
2008
, vol. 
134
 (pg. 
668
-
678
)
43
Liu
Z.
Sun
C.
Olejniczak
E. T.
Meadows
R. P.
Betz
S. F.
Oost
T.
Herrmann
J.
Wu
J. C.
Fesik
S. W.
Structural basis for binding of Smac/DIABLO to the XIAP BIR3 domain
Nature
2000
, vol. 
408
 (pg. 
1004
-
1008
)
44
Akagi
T.
Motegi
M.
Tamura
A.
Suzuki
R.
Hosokawa
Y.
Suzuki
H.
Ota
H.
Nakamura
S.
Morishima
Y.
Taniwaki
M.
Seto
M.
A novel gene, MALT1 at 18q21, is involved in t(11;18) (q21;q21) found in low-grade B-cell lymphoma of mucosa-associated lymphoid tissue
Oncogene
1999
, vol. 
18
 (pg. 
5785
-
5794
)
45
Dierlamm
J.
Baens
M.
Wlodarska
I.
Stefanova-Ouzounova
M.
Hernandez
J. M.
Hossfeld
D. K.
De Wolf-Peeters
C.
Hagemeijer
A.
Van den Berghe
H.
Marynen
P.
The apoptosis inhibitor gene API2 and a novel 18q gene, MLT, are recurrently rearranged in the t(11;18)(q21;q21) associated with mucosa-associated lymphoid tissue lymphomas
Blood
1999
, vol. 
93
 (pg. 
3601
-
3609
)
46
Lucas
P. C.
Yonezumi
M.
Inohara
N.
McAllister-Lucas
L. M.
Abazeed
M. E.
Chen
F. F.
Yamaoka
S.
Seto
M.
Nunez
G.
Bcl10 and MALT1, independent targets of chromosomal translocation in malt lymphoma, cooperate in a novel NF-κB signaling pathway
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
19012
-
19019
)
47
Uren
A. G.
O'Rourke
K.
Aravind
L. A.
Pisabarro
M. T.
Seshagiri
S.
Koonin
E. V.
Dixit
V. M.
Identification of paracaspases and metacaspases: two ancient families of caspase-like proteins, one of which plays a key role in MALT lymphoma
Mol. Cell
2000
, vol. 
6
 (pg. 
961
-
967
)
48
Zhou
H.
Du
M. Q.
Dixit
V. M.
Constitutive NF-κB activation by the t(11;18)(q21;q21) product in MALT lymphoma is linked to deregulated ubiquitin ligase activity
Cancer Cell
2005
, vol. 
7
 (pg. 
425
-
431
)
49
Annunziata
C. M.
Davis
R. E.
Demchenko
Y.
Bellamy
W.
Gabrea
A.
Zhan
F.
Lenz
G.
Hanamura
I.
Wright
G.
Xiao
W.
, et al. 
Frequent engagement of the classical and alternative NF-κB pathways by diverse genetic abnormalities in multiple myeloma
Cancer Cell
2007
, vol. 
12
 (pg. 
115
-
130
)
50
Keats
J. J.
Fonseca
R.
Chesi
M.
Schop
R.
Baker
A.
Chng
W. J.
Van Wier
S.
Tiedemann
R.
Shi
C. X.
Sebag
M.
, et al. 
Promiscuous mutations activate the noncanonical NF-κB pathway in multiple myeloma
Cancer Cell
2007
, vol. 
12
 (pg. 
131
-
144
)
51
Elsasser
S.
Finley
D.
Delivery of ubiquitinated substrates to protein-unfolding machines
Nat. Cell Biol.
2005
, vol. 
7
 (pg. 
742
-
749
)
52
Madura
K.
Rad23 and Rpn10: perennial wallflowers join the melee
Trends Biochem. Sci.
2004
, vol. 
29
 (pg. 
637
-
640
)
53
Schreiner
P.
Chen
X.
Husnjak
K.
Randles
L.
Zhang
N.
Elsasser
S.
Finley
D.
Dikic
I.
Walters
K. J.
Groll
M.
Ubiquitin docking at the proteasome through a novel pleckstrin-homology domain interaction
Nature
2008
, vol. 
453
 (pg. 
548
-
552
)
54
Husnjak
K.
Elsasser
S.
Zhang
N.
Chen
X.
Randles
L.
Shi
Y.
Hofmann
K.
Walters
K. J.
Finley
D.
Dikic
I.
Proteasome subunit Rpn13 is a novel ubiquitin receptor
Nature
2008
, vol. 
453
 (pg. 
481
-
488
)
55
Harper
J. W.
Schulman
B. A.
Structural complexity in ubiquitin recognition
Cell
2006
, vol. 
124
 (pg. 
1133
-
1136
)
56
Hoeller
D.
Hecker
C. M.
Dikic
I.
Ubiquitin and ubiquitin-like proteins in cancer pathogenesis
Nat. Rev. Cancer
2006
, vol. 
6
 (pg. 
776
-
788
)
57
Hurley
J. H.
Lee
S.
Prag
G.
Ubiquitin-binding domains
Biochem. J.
2006
, vol. 
399
 (pg. 
361
-
372
)
58
Deveraux
Q.
van Nocker
S.
Mahaffey
D.
Vierstra
R.
Rechsteiner
M.
Inhibition of ubiquitin-mediated proteolysis by the Arabidopsis 26 S protease subunit S5a
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
29660
-
29663
)
59
Raasi
S.
Pickart
C. M.
Rad23 ubiquitin-associated domains (UBA) inhibit 26 S proteasome-catalyzed proteolysis by sequestering lysine 48-linked polyubiquitin chains
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
8951
-
8959
)
60
Verma
R.
Oania
R.
Graumann
J.
Deshaies
R. J.
Multiubiquitin chain receptors define a layer of substrate selectivity in the ubiquitin–proteasome system
Cell
2004
, vol. 
118
 (pg. 
99
-
110
)

Author notes

1

All authors of this paper are or were employees and shareholders of Genetech, Inc.

2

Present address: Trubion Pharmaceuticals Inc., Seattle, WA 98121, U.S.A.

3

These authors contributed equally to this work.

4

Present address: Department of Molecular and Medical Pharmacology, University of California, Los Angeles, Los Angeles, CA 90095, U.S.A.

Supplementary data