IJC International Journal of Cancer

Transcription

1 IJC International Journal of Cancer Targeting inhibitor of apoptosis proteins by Smac mimetic elicits cell death in poor prognostic subgroups of chronic lymphocytic leukemia Daniela Opel 1, Andrea Schnaiter 1, Dagmar Dodier 1, Marjana Jovanovic 1, Andreas Gerhardinger 1, Irina Idler 1, Daniel Mertens 1, Lars Bullinger 1, Stephan Stilgenbauer 1 * and Simone Fulda 2,3,4 * 1 Department of Internal Medicine III, University Hospital Ulm, Ulm, Germany 2 Institute for Experimental Cancer Research in Pediatrics,, Goethe University, Frankfurt, Germany 3 German Cancer Consortium (DKTK), Heidelberg, Germany 4 German Cancer Research Center (DKFZ), Heidelberg, Germany Inhibitor of apoptosis (IAP) proteins are highly expressed in chronic lymphocytic leukemia (CLL) cells and contribute to evasion of cell death and poor therapeutic response. Here, we report that Smac mimetic BV6 dose-dependently induces cell death in 28 of 51 (54%) investigated CLL samples, while B-cells from healthy donors are largely unaffected. Importantly, BV6 is significantly more effective in prognostic unfavorable cases with, e.g., non-mutated VH status and TP53 mutation than samples with unknown or favorable prognosis. The majority of cases with 17p deletion (10/12) and Fludarabine refractory cases respond to BV6, indicating that BV6 acts independently of p53. BV6 also triggers cell death under survival conditions mimicking the microenvironment, e.g., by adding CD40 ligand or conditioned medium. Gene expression profiling identifies cell death, NF-jB and redox signaling among the top pathways regulated by BV6 not only in CLL but also in core-binding factor (CBF) acute myeloid leukemia (AML). Consistently, BV6 stimulates production of reactive oxygen species (ROS), which are contributing to BV6-induced cell death, since antioxidants reduce cell death. While BV6 causes degradation of cellular inhibitor of apoptosis (ciap)1 and ciap2 and nuclear factor-kappab (NF-jB) pathway activation in primary CLL samples, BV6 induces cell death independently of caspase activity, receptor-interacting protein (RIP)1 activity or tumor necrosis factor (TNF)a, as zvad.fmk, necrostatin-1 or TNFa-blocking antibody Enbrel fail to inhibit cell death. Together, these novel insights into BV6-regulated cell death in CLL have important implications for developing new therapeutic strategies to overcome cell death resistance especially in poor prognostic CLL subgroups. Key words: IAP proteins, Smac mimetic, CLL, cell death Abbreviations: ALL: acute lymphoblastic leukemia; AML: acute myeloid leukemia; ATP: adenosine 5 0 -triphosphate; CBF: corebinding factor; ciap: cellular inhibitor of apoptosis; CLL: chronic lymphocytic leukemia; GOEAST: Gene Ontology Enrichment Analysis Software Toolkit; IAP: inhibitor of apoptosis; IGVH: immunoglobulin heavy-chain variable region gene; NAC: N-acetylcysteine; Nec-1: necrostatin-1; NF-jB: nuclear factor-kappab; PARP: poly(adp-ribose) polymerase; RIP: receptor-interacting protein; ROS: reactive oxygen species; TNF: tumor necrosis factor; TRAIL: tumor necrosis factor-related apoptosis-inducing ligand; XIAP: X-linked inhibitor of apoptosis; zvad.fmk: N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone Additional Supporting Information may be found in the online version of this article. Conflict of Interest: Nothing to report *S.S. and S.F. shared senior authorship Grant sponsor: Deutsche Forschungsgemeinschaft, BMBF DOI: /ijc History: Received 19 Nov 2014; Accepted 29 May 2015; Online 19 June 2015 Correspondence to: Prof. Dr. Simone Fulda, Institute for Experimental Cancer Research in Pediatrics, Goethe-University, Komturstr. 3a, Frankfurt, Germany, Tel.: , Fax: , simone.fulda@kgu.de Chronic lymphocytic leukemia (CLL) is the most common type of adult leukemia in the Western world. 1 Key prognostic factors besides the Rai and Binet stage are chromosomal aberrations like 17p or 11q deletion, TP53 mutation and immunoglobulin heavy-chain variable region gene (IGVH) mutation status. Although individualized therapeutic strategies have been developed, the disease is still incurable. Especially patients with 17p deletion and TP53 mutation have a very poor prognosis. 2 Therefore, new therapeutic targets that act independently of functional p53 in CLL are needed. CLL is characterized by the accumulation of B- lymphocytes which has been largely attributed to defective cell death pathways rather than to aberrant proliferation. 3 5 Apoptosis represents one of the key forms of programmed cell death. 6 Necroptosis is a caspase-independent mode of programmed cell death that typically occurs when essential factors of apoptosis such as caspases are inhibited and is regulated by RIP1 and RIP3. 7 Inhibitor of apoptosis (IAP) proteins are major regulators of cell death and survival processes. X-linked inhibitor of apoptosis (XIAP) acts as direct caspase inhibitor, 8 while ciap1 and -2 were shown to protect cells from cell death by acting primarily as E3 ubiquitin ligases. 9 Smac binds to IAP proteins as an endogenous neutralizer. 10

2 2960 Smac mimetic elicits cell death in CLL What s new? Inhibitor of apoptosis (IAP) proteins are highly expressed in chronic lymphocytic leukemia (CLL) cells, contributing to cell death evasion and poor therapeutic response. Here, the Smac mimetic BV6 induces cell death in unfavorable prognostic subgroups of primary CLL samples and also under leukemic microenvironment conditions. Cell death, NF-jB, and redox signaling are all major BV6-regulated pathways, with BV6 activating NF-jB via degradation of cellular IAP proteins. BV6 also stimulates ROS production and ROS-dependent cell death in primary CLL cells. The findings may help to develop new therapeutic strategies to overcome cell death resistance, especially in CLL subgroups that have a poor prognosis. High expression levels of IAP proteins in CLL may contribute to cell death resistance. IAP proteins, in particular XIAP, are overexpressed in CLL and this overexpression confers chemoresistance and is associated with a progressive course To counter IAP proteins several small-molecule IAP inhibitors, including Smac mimetics, have been developed. 15 It has been shown that Smac mimetics abrogate XIAP-mediated caspase inhibition and induce autoubiquitination and rapid proteasomal degradation of ciap1 and-2, leading to activation of NF-jB and TNFa production that triggers cell death in an autocrine/paracrine fashion First clinical studies using Smac mimetics as single agents or in combination with chemotherapy have been initiated in patients with solid tumors and lymphoma. 15,19 Investigating the therapeutic potential of the bivalent Smac mimetic BV6 in CLL is of high interest, as our previous experiments showed that inhibition of IAP proteins increases death receptor-induced apoptosis in primary CLL cells, markedly in subgroups with bad prognosis including patients with 17p deletion. 13 Therefore, we aimed in our study to better understand the mechanisms of Smac mimetic-induced cell death in CLL and to identify parameters of responsiveness to Smac mimetics in CLL. Material and Methods Cell culture and reagents A cohort of 51 CLL patients from our institution were analyzed with patient consent and local ethical committee approval. Clinical and genetic characterization was performed as previously described and is provided in Supporting Information Table 1. Because of the limited amount of primary material that was available from one individual patient, not all different sets of experiments could be performed on the same samples. Individual samples were randomly selected for individual experiments to avoid a potential selection bias. Also, it is indicated in the figure legends which primary CLL samples were used for which experiments. Lymphocytes were isolated from blood samples by using Biocoll solution and aliquots were frozen viably at 21968C in RPMI 1640 enriched with 40% FCS and 10% DMSO. Primary B-cells were isolated from buffy coats of healthy donors by ficoll separation (Biochrom, Berlin, Germany) and magnetic bead isolation using human CD19 Micro Beads (Miltenyi Biotec, Berglisch Gladbach, Germany). Lymphocytes were cultured in RPMI 1640 medium supplemented with 10% FCS, 1% penicillin/streptomycin and 2% glutamine (all from Biochrom) at cells/ml. Both CLL cells and primary B-cells had been viably frozen before the setup of the experiments, the viability of the cells was always above 90% at the time of initiation of culture. Medium was conditioned by cultivation of HS-5 fibroblast cells obtained from ATCC in big cell culture flasks for 4 days. Supernatant was taken off and centrifuged at 4,000g for 10 min to remove any remaining cells. Human B-acute lymphoblastic leukemia (ALL) cell line Reh was obtained from ATCC and cultured in RPMI 1640 as described above. Smac mimetic BV6, a bivalent Smac mimetic that antagonizes XIAP, ciap1 and ciap2, 16 was kindly provided by Genentech, and Enbrel was kindly provided by Pfizer. Fludarabine was obtained from Medac. ABT263 and Obatoclax were purchased from Selleck Chemicals (Houston, TX). The caspase inhibitor zvad.fmk was purchased from Bachem, Necrostatin-1 (Nec-1) from Biomol. All chemicals were purchased from Sigma (Darmstadt, Germany) unless indicated otherwise. Western blot analysis Western blot analysis of a total of cells of the CLL samples in 5 ml medium was performed as described previously 23 using the following antibodies: goat anti-ciap1 (R&D Systems, Wiesbaden, Germany), rabbit anti-ciap2 (Epitomics, Burlingname, CA, USA), mouse anti-parp (Clontech, Mountain View, CA, USA), rabbit anti-caspase-3, rabbit anti-ijba, mouse antiphospho-ijba and rabbit anti-p100/p52 (all from Cell Signaling, Beverly, MA, USA). Mouse anti-b-actin (Sigma) was used as loading control followed by goat-anti-mouse IgG or goat-antirabbit IgG or donkey anti-goat IgG conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Heidelberg, Germany). Enhanced chemiluminescence was used for detection (Amersham Bioscience, Freiburg, Germany). All Western blots shown are representative of at least two independent experiments. Densitometry analysis was done using ImageJ. Determination of cell death A total of cells were treated with BV6 alone or in combination with the other drugs simultaneously. Cell death was assessed by flow cytometry and Annexin-V-PE and 7-AAD staining (BD Pharmingen, Heidelberg, Germany) according to the manufacturer s instructions from a total of 10,000 events per sample. The untreated controls were exposed to the same volume of DMSO in which BV6 was diluted and percentage of specific cell death was calculated as follows: [experimental cell death (%) 2 spontaneous cell death by DMSO (%)]/ [100% 2 spontaneous cell death by DMSO (%)].

3 Opel et al Table 1. Functional annotation clustering of genes modulated by BV6 in primary CLL cells (A) or primary CLL and CBF AML (B) using DAVID Database ( Annotation cluster no. Representative annotation term Enrichment score p-value (A) 1 Regulation of programmed cell death E213 2 Regulation of cell death E213 3 Leukocyte activation E210 4 Regulation of I-kappaB kinase/nf-kappab cascade E205 5 Positive regulation of immune system process E209 8 Positive regulation of NF-kappaB transcription factor activity E Regulation of kinase activity E Positive regulation of cellular biosynthetic process E Regulation of nitric oxide biosynthetic process E Positive regulation of nitric oxide biosynthetic process E204 (B) 1 Intracellular organelle lumen E208 2 Chromosome organization E207 3 Protein catabolic process E206 4 Protein ubiquitination E206 5 Transcription regulation E206 6 Regulation of programmed cell death E204 7 Regulation of programmed cell death E Positive regulation of NF-kappaB transcription factor activity Determination of ROS A total of cells were treated with 5 mm BV6 or 10 mm Antimycin as positive control for 20 hr. Cells were incubated for 30 min with CM-H2DCFDA (Life technologies, Darmstadt, Germany) for ROS analysis by flow cytometry. Viability assay An adenosine 5 0 -triphosphate (ATP)-content measurement using the CellTiterGlo Luminescent Cell Viability Assay (Promega, Mannheim, Germany), which reflects the amount of viable cells, was performed in a total of cells per sample treated with BV6 alone or in combination with N- acetylcysteine (NAC) and Trolox simultaneously according to the manufacturer s instructions. Caspase activity assay Enzymatic activity of caspase-3 in living non-fixed, non-lysed primary CLL cells was determined by FACS analysis using rhodamine-conjugated caspase-3 substrate (zdevd-r110) from Molecular Probes. 24 Gene expression profiling A total of cells from B CLL patient samples were treated for 4 or 20 hr with 1 mm BV6 or DMSO as control. Total RNA was available from four CLL patient samples (#23, 44, 50, 51 as shown in Supporting Information Table 2). Preparation of DNA single-stranded sense target, hybridization to GeneChip Human Genome U133 Plus 2.0 Arrays (Affymetrix, Santa Clara, CA, USA) and scanning of the arrays (7G Scanner, Affymetrix) were in accordance with the manufacturer s protocol as previously reported. 25 Cell files were normalized and filtered using BRB Array Tools Version (available at html) by applying the JustRMA (robust multi-array average) algorithm and previously reported filtering criteria (data are available at GSE62533 in case of CLL and GSE46819 in case of CBF-AML). 25 BRB Array Tools Version was also used for data analysis using class comparison and pathway class comparison tools as previously described. 26 All tests were two-sided. An effect was considered significant if the p values were 0.05 or less. Results were visualized using TreeView. 27 For class comparison analysis all BV6-treated samples versus all DMSO-treated samples regardless of the incubation time were used, because at 4 hr the changes in gene expression levels were not very strong, which might be explained by the short incubation time (Supporting Information Table 2). In addition, we performed functional annotation analysis on the selected gene list using the database for Annotation, Visualization and Integrated Discovery 6.7 tool (DAVID, Gene annotation enrichment analysis and functional annotation clustering were selected as

4 2962 Smac mimetic elicits cell death in CLL annotation categories and high classification stringency was set for the analysis in the Functional Annotation option. In addition, GOEAST (Gene Ontology Enrichment Analysis Software Toolkit; was used to identify significantly enriched biological process GO terms. Figure 1.

5 Opel et al Statistical analysis Statistical significance was assessed by Student s t-test (twotailed distribution, two-sample unequal variance). Results BV6 induces cell death in primary CLL cells irrespective of poor prognostic factors In a first approach, we investigated the potential of the Smac mimetic BV6 to cause cell death in primary CLL cells. We defined response to BV6 when at least 30% specific cell death was induced by 10 mm BV6. Treatment with BV6 triggered cell death in a concentration-dependent manner; 55% (28/51) of patient samples responded to BV6, while primary B-cells from healthy donors remained largely unaffected (Fig. 1a). Importantly, most samples (10/12; 83%) derived from patients with 17p deletion who bear the poorest prognosis responded to BV6. Similarly, the majority (9/12; 75%) of cases with TP53 mutation were sensitive to BV6. Also, 59% (16/27) of cases with poor prognostic non-mutated IGVH status were susceptible to BV6. Interestingly, 75% (6/8) of Fludarabine refractory cases were sensitive to BV6 (Fig. 1b). In general, more samples derived from patients with poor prognosis features belonged to the BV6-responsive group rather than to the non-responsive group. Accordingly, among the BV6-responsive samples 57% (16/28) showed an unfavorable non-mutated VH status, 35% (10/28) harbored a 17p deletion and 32% (9/28) mutated TP53 (Supporting Information Table 1). Notably, samples derived from patients with unfavorable prognostic factors like non-mutated IGVH status, TP53 mutation or 17p deletion responded significantly better to BV6 than samples derived from patients with unknown or favorable prognosis, e.g., harbouring 13q deletion, 11q deletion or normal karyotype (Fig. 1c). Furthermore, BV6 efficiently triggered cell death in a patient cohort of 18 cases where Fludarabine was limited in its efficiency (Fig. 1d, left panel). Within this cohort of poor Fludarabine responders, BV6 induced cell death in both favorable subgroups (i.e., normal karyotype, 13q deletion) and unfavorable subgroups (i.e., 17p deletion, TP53 mutation, non-mutated IGVH gene) (Fig. 1d, right panel). In contrast, the ability of Fludarabine to induce cell death was significantly reduced in prognostic unfavorable samples in comparison to favorable samples (Fig. 1d, right panel). BV6 was not studied in ibrutinib/idelalisib-refractory cases. As survival signals provided by the bone marrow microenvironment have been implicated to confer resistance to CLL, 28 we next asked whether these microenvironmental cues protect against BV6-induced toxicity. To address this question, we assessed BV6-induced cell death in the presence and absence of CD40 ligand or conditioned medium. Of note, addition of CD40 ligand did not interfere with the ability of BV6 to induce cell death in the investigated cases (Fig. 1e), whereas we previously showed that it rescues CLL cells from spontaneous apoptosis in vitro. 29 Also, the response to BV6 was not significantly altered in the presence of conditioned medium compared to normal medium and BV6 at equimolar concentrations significantly exceeded Fludarabine-induced cell death under the same conditions (Fig. 1f). These data indicate that BV6 induces cell death also under survival conditions of the microenvironment. To investigate the question whether BV6 cooperates with other apoptosis-targeted drugs, we tested BV6 in combination with BH3 mimetics. BV6 cooperated with Obatoclax to induce cell death in CLL cells, Figure 1. BV6 induces cell death in prognostic unfavorable primary CLL cells and exceeds Fludarabine treatment. (a) Primary CLL patient samples and primary B-cells were treated for 48 or 24 hr with indicated concentrations of BV6. Cell death was determined by FACS analysis and percentage of specific cell death relative to solvent control DMSO is shown (mean cell death induced by DMSO alone: sensitive samples 42%; non-sensitive samples 43%, B-cells 21%). Mean and SEM of 28 BV6-sensitive CLL samples ( ), 23 BV6 non-sensitive CLL samples ( w) or six primary B-cell samples from healthy donors ( ) measured in duplicate or triplicate are plotted in dot-plot format. (b) Whole cohort of primary patient samples was divided into genomic subgroups plus Fludarabine refractory group. BV6-sensitive or non-sensitive cases in each subgroup were counted and percentage in order to all cases per subgroup in respect to the BV6 response is shown. (c) BV6- sensitive primary patient samples (all, n 5 28) treated with 10 mm BV6 were grouped based on their genetic background (X-axis) or Fludarabine refractory group and percentage of specific cell death analyzed by FACS, relative to solvent control DMSO is shown (mean cell death induced by DMSO: 42%; norm/13q-: 42%; IGVHu: 36%; IGVHm: 41%; tp53m: 45%; tp53u: 38%; 17p-: 44%; no 17p-: 40%; 11q-: 44%; no 11q-: 34%). Mean and SEM of individual CLL samples measured in duplicate or triplicate are shown; **p < 0.001; *p < 0.05; u: nonmutated; m: mutated; norm: normal; sens.: sensitive. (d) (left) BV6-sensitive primary CLL cells (n 5 18) were treated for 48 hr with 10 mm BV6 or 10 mm Fludarabine. Cell death was determined by FACS analysis and percentage of specific cell death, relative to solvent control DMSO is shown (cell death induced by DMSO: 36%). Mean and SEM of 18 individual CLL samples measured in duplicate or triplicate (samples no. 2, 6, 8, 15, 17, 18, 19, 21, 24, 28, 33, 36, 39, 40, 41, 43, 48, 50) are shown; **p < 0.001; *p < 0.05; (right) Graph illustrates specific cell death upon treatment with 10 mm Fludarabine or BV6 of genomic subgroups including norm/13q-, IGVH non-mutated, tp53 mutated and 17p deletion. (e and f) BV6-sensitive primary CLL cells were co-treated with 10 mm BV6 and 1,000 ng/ml CD40 ligand (e) or stimulated with BV6 in normal or conditioned medium at indicated concentrations (f) for 48 hr. Cell death was determined by FACS analysis and percentage of specific cell death, relative to solvent control DMSO is shown [cell death induced by DMSO: 45% (e) or 22% (f)]. Mean and SEM of five (e: samples no. 2, 12, 18, 19, 43) or six (f: samples no. 17, 24, 33, 43, 48, 50) individual CLL samples measured in duplicate or triplicate are shown; **p < (g) Primary CLL cells were treated for 48 hr with 5 mm BV6 alone or in combination with indicated concentrations of ABT263 (left) or Obatoclax (right). Cell death was determined by FACS analysis and percentage of specific cell death, relative to solvent control DMSO is shown (cell death induced by DMSO: 31% in case of Obatoclax; 40% in case of ABT263). Mean and SEM of four (Obatoclax: samples no: 2, 19, 43, 46) or nine (ABT-263: samples no: 2, 5, 12, 17, 19, 39, 43, 46, 48) individual CLL patients samples measured in triplicate are shown; *p < 0.05;**p <

6 2964 Smac mimetic elicits cell death in CLL whereas no interaction was observed for BV6 and ABT-263 (Fig. 1g). Together, this set of experiments indicates that (i) BV6 triggers cell death in more than half of primary CLL samples, including the majority of 17p-deleted cases, but spares normal primary B-cells; (ii) BV6 causes cell death in unfavorable CLL subgroups likely independently of the p53 pathway and (iii) BV6 is efficient in an environment enriched for survival factors. Figure 2. Heatmap of the top 200 differently expressed genes comparing BV6-treated versus solvent DMSO-treated samples of four CLL patient samples. BV6 expression-associated genes (rows) and CLL samples (columns) were hierarchically clustered (average linkage clustering; similarity metric: correlation, uncentered; selected gene names are depicted). Cases with low and high expression indicated by blue and red bars, respectively, grouped together as indicated (for annotated probe sets gene symbols are provided). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] BV6 modulates genes related to programmed cell death, NF-jB and redox signaling across tumor entities Based on our previous data showing that BV6-stimulated activation of transcription factors such as NF-jB and IRF1 is critically required for BV6-induced apoptosis, we performed gene expression profiling to identify genes that are specifically regulated in BV6-responsive CLL samples. A total of 1,068 genes were found to be significantly differentially expressed between the BV6-treated and DMSO control group (p < 0.05). Overall, 322 genes showed higher expression, while 746 genes were less expressed as summarized in Supporting Information Table 3. The heatmap of the top 100 upregulated or downregulated genes upon BV6 treatment is shown in Figure 2. Of note, the identified gene list includes several transcripts associated with regulation of cell death and NF-jB signaling (i.e., TRAF1, REL, TNFRSF4, NFKBIE, NFKB2, RIPK1, CFLAR, CASP8, CYLD, TNFAIP3) as well as genes that code for proteins involved in ROS modulation (i.e., SOD2, HSP90AB1) next to others associated with survival pathways, T-cell activation and inflammation (Supporting Information Table 3). To identify functional categories that were most significantly modulated by BV6 we used the DAVID Functional Annotation Clustering Tool. We identified 55 annotation clusters with enrichment scores >1.5 (Supporting Information Table 4). "Regulation of programmed cell death" and "NF-jB signaling" were represented by clusters 1, 2, 4 and 8, "regulation of nitric oxide biosynthetic process" by cluster 29 (Table 1A). In addition, using the Gene Ontology Enrichment Analysis Software Toolkit (GOEAST), we identified 432 enrichment sets (p < 0.05), including "regulation of programmed cell death," "response to stress," "cell death," "apoptotic process" and "regulation of I-jB kinase/nf-jb cascade" in line with the results obtained by using the DAVID tool (Supporting Information Table 4). To explore whether these BV6-stimulated changes in gene expression are specific for CLL, we compared these gene expression profiles in primary CLL samples with those that we previously described for BV6-sensitive primary CBF AML cases. 25 Pairwise analysis comparing BV6- and DMSOtreated samples irrespective of the tumor entity resulted in 868 distinctly regulated genes by BV6 (i.e., 644 downregulated and 224 upregulated genes, Supporting Information Table 5), including among others TRAF1, TNFAIP3, TNFRSF4, NFKBIA, BIRC3 and SOD3. Importantly, using

7 Opel et al DAVID functional annotation clustering tool, we identified among 70 annotation clusters with enrichment scores > 1.0 the two clusters regulation of programmed cell death and positive regulation of NF-jB transcription factor activity in both the CLL as well as the CBF AML cohort (Supporting Information Table 6, Table 1B), indicating that BV6 alters these gene clusters independently of the tumor entity. Furthermore, the comparison of the two GEP data sets (class BV6/class DMSO-CLL vs. class BV6/class DMSO- CLL1AML) revealed a 20% (214/1,068) overlapping rate. The gene list of these 214 overlapping genes was subsequently analyzed by DAVID functional annotation clustering tool. Interestingly, we identified regulation of programmed cell death and positive regulation of NF-jB transcription factor activity among the 39 clusters with enrichment scores > 1.0. In addition, 22 of the 214 overlapping genes showed an overlap with the top 100 enriched genes in the CLL gene list and 21 with the top 100 downregulated genes, including genes belonging to cell death and NF-jB signaling pathway (Supporting Information Table 7). Thus, we identified a general gene signature related to regulation of programmed cell death and NF-jB pathway signaling that is induced by BV6 across at least two distinct tumor entities. In addition, we found ROS-modulating genes to be differently expressed upon BV6 treatment in CLL and CBF AML pointing to modulation of redox signaling by BV6. BV6 triggers ciap degradation, NF-jB activation and TNFaindependent cell death in primary CLL cells Based on our GEP analysis that identified NF-jB signaling as one of the top-regulated pathways, we addressed the question whether BV6 leads to NF-jB pathway activation using Western blot analysis as previously reported. 33 To this end, we assessed the phosphorylation status and thereby the activation status of IjBa as well as proteolytic processing of p100 to p52, which are widely used as parameters to determine NF-jB pathway activation. BV6 stimulated an increase of phospho-ijba and p52 protein expression as well as a decrease of p100 levels (Fig. 3a, Supporting Information Fig. 1A), thus pointing to activation of both canonical and noncanonical NF-jB signaling. These findings are consistent with our data obtained by GEP analysis that showed activation of NF-jB signaling and underscore that BV6 stimulates NF-jB pathway activation in primary CLL cells. In parallel, BV6 caused rapid downregulation of ciap1 and ciap2 in a concentration- and time-dependent manner (Fig. 3a, Supporting Information Figs. 1B and 2), in line with the reported function of Smac mimetic to trigger NF-jB activation upon depletion of ciap proteins. 16,35 As Smac mimetics have been described to engage an NFjB-mediated, TNFa-driven autocrine/paracrine cell death loop, 9,16,35,36 we next investigated whether TNFa is required for BV6-induced cell death in primary CLL cells. However, addition of the TNFa-blocking antibody Enbrel failed to protect primary CLL cells against BV6-induced cell death (Fig. 3b), indicating that BV6-induced cell death occurs independently of TNFa. As a positive control, Enbrel significantly decreased BV6-induced cell death in the ALL cell line Reh (Fig. 3c). This set of experiments demonstrates that BV6 causes depletion of ciap proteins along with NF-jB activation, yet induces cell death independently of TNFa in primary CLL cells. BV6 triggers non-apoptotic cell death in primary CLL cells Based on our GEP analysis that discovered programmed cell death as one of the top-regulated pathways, we next monitored activation of caspases. Western blot analysis revealed that BV6 weakly induced cleavage of caspase-3 and poly(- ADP-ribose) polymerase (PARP) in a time-dependent manner, indicated by a minor accumulation of active caspase-3 fragments and PARP cleavage products as well as by a slight decrease of the pro-enzyme form of caspase-3 and full-length PARP (Fig. 4a). In line with these data, we found only a nearly twofold increase in caspase-3 activity in contrast to a tenfold increase of caspase activity after Staurosporine stimulation used as positive control (Fig. 4b). To examine whether caspase activity is required for cell death induction, we tested the effect of the broad-range caspase inhibitor zvad.fmk. Notably, addition of zvad.fmk failed to protect primary CLL cells against BV6-induced cell death (Fig. 4c). In fact, the addition of zvad.fmk significantly increased BV6- induced cell death in a cohort of 18 primary CLL samples (Fig. 4c), in line with recent reports that caspase-8 can limit non-apoptotic cell death. 37 In Reh ALL cells, which were used as positive control for caspase-dependent apoptosis, cell death was significantly reduced by zvad.fmk (Fig. 3c). These data indicate that BV6 induces caspase-independent cell death in primary CLL cells. The failure of caspase inhibition by zvad.fmk to prevent cell death points to alternative modes of cell death, for example, necroptosis that has recently been identified. 38 To test whether cells undergo apoptotic or necrotic cell death we simultaneously monitored phosphatidylserine exposure on the plasma membrane by Annexin-V staining as characteristic marker of apoptosis in parallel with 7-AAD staining to define the loss of plasma membrane integrity as an early marker of necrosis. Importantly, BV6 predominately increased the percentage of Annexin-V and 7-AAD doublepositive cells with only a minor increase in the percentage of single Annexin-V-positive cells (Fig. 4d). In addition, we found a significant increase in double-positive cells by treatment with BV6 plus zvad.fmk (Figs. 4d and 4e), suggesting that BV6 triggers caspase-independent, non-apoptotic cell death when caspase activation is blocked. To explore whether BV6 triggers necroptosis in CLL cells, we used the RIP1 kinase inhibitor Nec-1, as RIP1 has been reported to be a critical mediator of necroptosis. 38 Results showed that Nec-1 was unable to reduce BV6-induced cell death in primary CLL cells, demonstrating that BV6-

8 2966 Smac mimetic elicits cell death in CLL Figure 3. BV6 leads to ciap degradation, NF-jB activation and TNFa-independent cell death. (a) Representative Western blot showing the protein expression levels of ciap1, ciap2, XIAP, p-ijba, IjBa, p100 and p52 detected in primary CLL cells (samples no. 5, 22, 24, 39, 43) after 4 and 20 hr treatment with indicated concentrations of BV6. b-actin served as loading control. (b) BV6-sensitive primary CLL cells were treated for 48 hr with 40 mg/ml Enbrel, 10 mm BV6 alone or in combination. Cell death was determined by FACS analysis and percentage of specific cell death, relative to solvent control DMSO is shown (cell death induced by DMSO: 41%). Mean and SEM of 12 individual CLL samples measured in duplicate or triplicate (samples no. 2, 6, 8, 12, 18, 19, 28, 39, 41, 43, 48, 50) are shown. (c) Primary CLL cells and ALL cell line Reh were left untreated or were treated for 48 hr with DMSO, 20 mm zvad.fmk, 30 mm Nec-1, 40 mg/ml Enbrel, 10 mm BV6 or in case of Reh cells with 1 mm BV6 alone or in combination. Cell death was determined by FACS analysis and percentage of cell death relative to solvent control DMSO is shown (cell death induced by DMSO: 10%). Mean and SEM of two individual CLL samples measured in triplicate (samples no. 19 and 26) are shown; **p < mediated cell death does not depend on RIP1 activity (Fig. 1f). In this cohort of 17 primary CLL samples, we noticed even a slight increase in BV6-mediated cell death upon addition of Nec-1 (Fig. 1f), although this was not seen in two other samples (Fig. 3c). As a positive control, Nec-1 significantly decreased BV6-induced cell death in Reh ALL cells (Fig. 3c). Together, this set of experiments indicates that BV6 elicits non-apoptotic, caspase- and RIP1 kinase-independent cell death in primary CLL cells. BV6-induced cell death depends on ROS Based on our GEP data identifying regulation of ROSmodulating genes by BV6, we then examined ROS production by using the ROS-sensitive dye CM-H2DCFDA and Antimycin as positive control. BV6 significantly enhanced cellular ROS levels in primary CLL cells (Fig. 5a). To test whether increased ROS production is contributing to BV6- induced cell death we used NAC, a biosynthetic precursor of glutathione, as an antioxidant. Importantly, NAC significantly reduced BV6-mediated loss of viability in a concentrationdependent manner (Fig. 5b). In addition, NAC treatment substantially reduced BV6-induced cell death in primary CLL cells (Fig. 5c). To further test the requirement of ROS, we used the vitamin E derivative Trolox as additional antioxidant. Similarly, Trolox significantly reduced BV6-mediated cell death (Fig. 5c). Taken together, these experiments indicate that ROS contribute to BV6-induced cell death in primary CLL cells. Discussion Here, we provide evidence showing that the Smac mimetic BV6 represents a promising strategy to induce cell death in CLL cells. This conclusion is supported by data obtained in a large panel of primary CLL samples from different subgroups of patients, including cases with 17p deletion, the subgroup

9 Opel et al Figure 4.

10 2968 Smac mimetic elicits cell death in CLL with the worst prognosis, 20 samples with TP53 mutation or with unmutated IGVH gene. Importantly, BV6 is more effective in CLL cases with poor outcome, as BV6-induced cell death is higher in prognostically unfavorable cases than in prognostically favorable cases with normal or 13q karyotype. Also, BV6 may represent an alternative treatment strategy for Fludarabine-resistant cases, because it triggers cell death in CLL samples that are refractory to Fludarabine. In addition, BV6 exceeds Fludarabine s antileukemic activity especially in resistant samples with TP53 mutation or 17p deletion, which is associated with loss of p53. While some other studies using different Smac mimetics reported no correlation between cell death induction by Smac mimetic and biological parameters in primary CLL, 39,40 two studies using IAP inhibitors plus tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) showed cell death induction in CLL cases with poor prognosis. 13,41 Using whole-genome gene expression profiling, we identify two functional categories that are most significantly modulated by BV6, i.e., programmed cell death and NF-jB pathway, across tumor entities not only in CLL but also in primary CBF AML samples, underlining the broader relevance of this finding. Concomitant experiments demonstrate that treatment with BV6 activates NF-jB signaling and cell death pathways in primary CLL cells. Furthermore, gene expression profiling in CLL cells revealed downregulation of SOD2, an antioxidant enzyme involved in the control of redox homeostasis. 42 Consistently, parallel studies in primary CLL cells showed that BV6 stimulates production of ROS, which are contributing to BV6- induced cell death, as antioxidants attenuate BV6-mediated cell death. It is interesting to note that an antioxidant function has been ascribed to p53, 43 implying that TP53 mutation or p53 loss due to 17p deletion may lead to redox imbalance and constitutively increased ROS stress. The finding showing that BV6 modulates genes that code for proteins with antioxidative function indicates that BV6 might shift the redox balance resulting in an overall increase in ROS levels which may induce cell death when above a cellular tolerability threshold. Despite depletion of ciap proteins and NF-jB activation, BV6-induced cell death in primary CLL cells occurred independently of a TNFa-driven cell death loop previously described to mediate Smac mimetic-induced cell death in an autocrine/paracrine manner. 35,36,44 This conclusion is supported by the following lines of evidence: First, the TNFacatching antibody Enbrel fails to prevent BV6-induced cell death in primary CLL cells. Similarly, we recently described that BV6/Dexamethasone-induced cell death in childhood ALL occurs in a TNFa-independent manner. 45 Furthermore, no elevated TNFa levels were found in our GEP analysis, while this has previously been observed by other investigators in CLL cells upon treatment with another Smac mimetic. 40 Second, BV6-induced cell death is caspase-independent, because zvad.fmk is unable to protect CLL cells from BV6- triggered cell death and even significantly increases cell death, pointing to a caspase-independent mode of cell death, for example, necroptosis. However, the RIP1 kinase inhibitor Nec-1 fails to inhibit BV6-induced cell death. Thus, BV6 triggers caspase-independent, RIP1 kinase-independent non-apoptotic cell death in CLL cells. For survival of B-cells, a balance of NF-jB activity is required. Too much activation of NF-jB drives B-cells into apoptosis, e.g., during negative selection via inappropriate BCR stimulation. Activation of NF-jB might therefore especially in CLL play a role in programmed cell death by BV6, as this pathway was the strongest in the gene expression profiling. We previously reported that Smac mimetics can act as chemosensitizers in combination with anticancer drugs, 30,46 demethylating agents or glucocorticoids 45,47 and as radiosensitizer. 31,48,49 Our findings showing that BV6 cooperates together with Obatoclax to induce cell death in CLL cells, but not together with ABT-263 may be related to the differential activity profiles of these BH3 mimetics, as Obatoclax antagonizes BCL-2, BCL-xL, BCL-W and MCL-1, whereas ABT-263 targets BCL-2, BCL-xL and BCL-W, but not MCL Single-agent cytotoxicity of Smac mimetic compounds has been reported for CLL, 40 multiple myeloma or Figure 4. BV6 induces caspase-independent cell death in primary CLL cells. (a) Representative Western blot showing the protein expression levels of PARP and caspase-3 and cleaved caspase-3 in primary CLL cells (samples no. 15 and 34) after treatment with 5 mm BV6 for indicated times. b-actin served as loading control. (b) BV6-sensitive primary CLL cells (samples no. 2, 12, 17, 39) were treated for indicated times with 10 mm BV6, 1 mm Staurosporine and 20 mm zvad.fmk alone or in combination. Caspase-3 activity was determined by FACS analysis using 100 mm rhodamine-conjugated zdevd substrate after 30 min of incubation at 378C. SEM and mean fold increase in caspase activity of one experiment performed in duplicate is shown. (c) BV6-sensitive primary CLL cells were treated for 48 hr with indicated concentration of BV6 or zvad.fmk alone or in combination. Cell death was determined by FACS analysis and percentage of specific cell death, relative to solvent control DMSO is shown (cell death induced by DMSO: 40%). Mean and SEM of 17 individual CLL samples measured in duplicate or triplicate (samples no. 2, 6, 8, 12, 15, 17, 18, 19, 24, 28, 33, 36, 39, 41, 43, 48, 50) are shown; *p < (d and e) BV6- sensitive primary CLL cells (samples no. 12, 15, 24, 39) were treated for 48 hr with 10 mm BV6 in the presence or absence of 20 mm zvad.fmk. Cell death was determined by FACS analysis. The percentage of Annexin-V-positive/7AAD double-positive cells is shown (e) with mean and standard deviation of one experiment performed in duplicate; *p < (d) The percentage of Annexin-V (A)- and/or 7AAD (7A)- positive and -negative cells is shown with mean and standard deviation of one experiment performed in duplicate. (f) BV6-sensitive primary CLL cells were treated for 48 hr with 30 mm Nec-1 or 10 mm BV6 alone or in combination. Cell death was determined by FACS analysis and percentage of specific cell death relative to solvent control DMSO is shown (cell death induced by DMSO: 41%). Mean and SEM of 17 individual CLL patients measured in duplicate or triplicate (samples no. 2, 6, 8, 12, 15, 17, 18, 19, 24, 28, 33, 36, 39, 41, 43, 48, 50) are shown; *p < 0.05.

11 Opel et al Figure 5. BV6-induced cell death requires ROS. (a) BV6-sensitive primary CLL cells (samples no. 1, 12, 44, 51) were treated with 5 mm BV6 or 10 mm Antimycin for positive control for 20 hr. ROS production was determined by CM-H2DCFDA staining and flow cytometry; *p < (b) BV6-sensitive primary CLL cells (samples no. 2 and 24) were treated for 48 hr with 10 mm BV6, increased NAC concentrations ( mg/ml) alone or in combination. Cell viability was determined by GloMax viability assay and percentage of viability DMSO set 100% is shown. Mean and SEM of one experiment performed in triplicate are shown; **p < 0.001; *p < (c) BV6-sensitive primary CLL cells were treated for 48 hr with NAC, Trolox or BV6 alone or in combination at indicated concentrations. Cell death was determined by FACS analysis and percentage of specific cell death, relative to solvent control DMSO is shown (cell death induced by DMSO: 27%). Mean and SEM of six individual CLL samples measured in triplicate (samples no. 2, 6, 18, 24, 33, 50) are shown; **p < 0.001; *p < several solid tumors, 16,35,36,51 53 while another study showed that CLL cells remain resistant against Smac mimetics in the context of CD40 ligand stimulation. 39 Compared with CLL cells, BV6 showed little cytotoxicity against non-malignant B-cells, a finding also reported by other investigators in CLL. 40 As there is so far no evidence of a cell cycle dependency of Smac mimetic-induced cytotoxicity, Smac mimetics are of interest also for malignancies with a low proliferative capacity. In conclusion, we show that BV6 induces cell death especially in prognostically unfavorable subgroups of primary CLL samples. The correlation found between the responsiveness to BV6 and several poor outcome parameters is a new finding with important implications for the development of targeted therapies. Furthermore, our GEP analyses offer new insights into BV6-regulated signaling pathways in CLL. Together with the finding that BV6 efficiently induces cell death also under leukemic microenvironment conditions, e.g., by adding CD40 ligand or conditioned medium, BV6 may contribute to the development of future treatment strategies for CLL. As our study also shows that a substantial proportion of CLL samples are unresponsive to BV6, it will be important to further investigate mechanisms of resistance to Smac mimetics in future studies.

12 2970 Smac mimetic elicits cell death in CLL References 1. Hallek M, Cheson BD, Catovsky D, et al. International Workshop on Chronic Lymphocytic Leukemia Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: a report from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute- Working Group 1996 guidelines. Blood 2008;111: Zenz T, Mertens D, Kuppers R, et al. From pathogenesis to treatment of chronic lymphocytic leukaemia. Nat Rev Cancer 2010;10: Binet JL, Plunkett W, Robertson B, et al. What does apoptosis mean in CLL? Leuk Lymphoma 1996;22 (Suppl 2): Danilov AV, Danilova OV, Klein AK, et al. Molecular pathogenesis of chronic lymphocytic leukemia. Curr Mol Med 2006;6: Kolb JP, Kern C, Quiney C, et al. Re-establishment of a normal apoptotic process as a therapeutic approach in B-CLL. Curr Drug Targets Cardiovasc Haematol Disord 2003;3: Galluzzi L, Vitale I, Abrams JM, et al. Molecular definitions of cell death subroutines: recommendations of the Nomenclature Committee on Cell Death Cell Death Differ 2012;19: Vandenabeele P, Galluzzi L, Vanden Berghe T, et al. Molecular mechanisms of necroptosis: an ordered cellular explosion. Nat Rev Mol Cell Biol 2010;11: Eckelman BP, Salvesen GS, Scott FL. Human inhibitor of apoptosis proteins: why XIAP is the black sheep of the family. EMBO Rep 2006;7: Bertrand MJ, Milutinovic S, Dickson KM, et al. ciap1 and ciap2 facilitate cancer cell survival by functioning as E3 ligases that promote RIP1 ubiquitination. Mol Cell 2008;30: Liu Z, Sun C, Olejniczak ET, et al. Structural basis for binding of Smac/DIABLO to the XIAP BIR3 domain. Nature 2000;408: Chen R, Plunkett W. Strategy to induce apoptosis and circumvent resistance in chronic lymphocytic leukaemia. Best Pract Res Clin Haematol 2010;23: Grzybowska-Izydorczyk O, Cebula B, Robak T, et al. Expression and prognostic significance of the inhibitor of apoptosis protein (IAP) family and its antagonists in chronic lymphocytic leukaemia. Eur J Cancer 2010;46: Loeder S, Zenz T, Schnaiter A, et al. A novel paradigm to trigger apoptosis in chronic lymphocytic leukemia. Cancer Res 2009;69: Silva KL, Vasconcellos DV, Castro ED, et al. Apoptotic effect of fludarabine is independent of expression of IAPs in B-cell chronic lymphocytic leukemia. Apoptosis 2006;11: Fulda S, Vucic D. Targeting IAP proteins for therapeutic intervention in cancer. Nat Rev Drug Discov 2012;11: Varfolomeev E, Blankenship JW, Wayson SM, et al. IAP antagonists induce autoubiquitination of c-iaps, NF-kappaB activation, and TNFalphadependent apoptosis. Cell 2007;131: Varfolomeev E, Vucic D. (Un)expected roles of c- IAPs in apoptotic and NFkappaB signaling pathways. Cell Cycle 2008;7: Wu H, Tschopp J, Lin SC. Smac mimetics and TNFalpha: a dangerous liaison? Cell 2007;131: LaCasse EC, Mahoney DJ, Cheung HH, et al. IAP-targeted therapies for cancer. Oncogene 2008; 27: Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and survival in chronic lymphocytic leukemia. NEnglJMed2000;343: Krober A, Bloehdorn J, Hafner S, et al. Additional genetic high-risk features such as 11q deletion, 17p deletion, and V3-21 usage characterize discordance of ZAP-70 and VH mutation status in chronic lymphocytic leukemia. J Clin Oncol 2006;24: Zenz T, Krober A, Scherer K, et al. Monoallelic TP53 inactivation is associated with poor prognosis in chronic lymphocytic leukemia: results from a detailed genetic characterization with long-term follow-up. Blood 2008;112: Fulda S, Strauss G, Meyer E, et al. Functional CD95 ligand and CD95 death-inducing signaling complex in activation-induced cell death and doxorubicin-induced apoptosis in leukemic T cells. Blood 2000;95: Giagkousiklidis S, Vellanki SH, Debatin KM, et al. Sensitization of pancreatic carcinoma cells for gamma-irradiation-induced apoptosis by XIAP inhibition. Oncogene 2007;26: Luck SC, Russ AC, Botzenhardt U, et al. Deregulated apoptosis signaling in core-binding factor leukemia differentiates clinically relevant, molecular marker-independent subgroups. Leukemia 2011;25: Luck SC, Russ AC, Du J, et al. KIT mutations confer a distinct gene expression signature in core binding factor leukaemia. Br J Haematol 2010;148: Eisen MB, Spellman PT, Brown PO, et al. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 1998;95: Munk Pedersen I, Reed J. Microenvironmental interactions and survival of CLL B-cells. Leuk Lymphoma 2004;45: Bhattacharya N, Reichenzeller M, Caudron-Herger M, et al. Loss of cooperativity of secreted CD40L and increased dose-response to IL4 on CLL cell viability correlates with enhanced activation of NF-kB and STAT6. Int J Cancer 2015;136: Wagner L, Marschall V, Karl S, et al. Smac mimetic sensitizes glioblastoma cells to Temozolomideinduced apoptosis in a RIP1- and NF-kappaBdependent manner. Oncogene 2013;32: Berger R, Jennewein C, Marschall V, et al. NFkappaB is required for Smac mimetic-mediated sensitization of glioblastoma cells for gammairradiation-induced apoptosis. Mol Cancer Ther 2011;10: Abhari BA, Cristofanon S, Kappler R, et al. RIP1 is required for IAP inhibitor-mediated sensitization for TRAIL-induced apoptosis via a RIP1/ FADD/caspase-8 cell death complex. Oncogene 2013;32: Eckhardt I, Roesler S, Fulda S. Identification of DR5 as a critical, NF-kappaB-regulated mediator of Smac-induced apoptosis. Cell Death Dis 2013; 4:e Eckhardt I, Weigert A, Fulda S. Identification of IRF1 as critical dual regulator of Smac mimeticinduced apoptosis and inflammatory cytokine response. Cell Death Dis 2014;5:e Vince JE, Wong WW, Khan N, et al. IAP antagonists target ciap1 to induce TNFalpha-dependent apoptosis. Cell 2007;131: Petersen SL, Wang L, Yalcin-Chin A, et al. Autocrine TNFalpha signaling renders human cancer cells susceptible to Smac-mimetic-induced apoptosis. Cancer Cell 2007;12: Oberst A, Green DR. It cuts both ways: reconciling the dual roles of caspase 8 in cell death and survival. Nat Rev Mol Cell Biol 2011;12: Vandenabeele P, Declercq W, Vanden Berghe T. Necrotic cell death and necrostatins : now we can control cellular explosion. Trends Biochem Sci 2008; 33: Maas C, Tromp JM, van Laar J, et al. CLL cells are resistant to smac mimetics because of an inability to form a ripoptosome complex. Cell Death Dis 2013;4:e Scavullo C, Servida F, Lecis D, et al. Single-agent Smac-mimetic compounds induce apoptosis in B chronic lymphocytic leukaemia (B-CLL). Leuk Res 2013;37: Frenzel LP, Patz M, Pallasch CP, et al. Novel X- linked inhibitor of apoptosis inhibiting compound as sensitizer for TRAIL-mediated apoptosis in chronic lymphocytic leukaemia with poor prognosis. Br J Haematol 2011;152: Trachootham D, Alexandre J, Huang P. Targeting cancer cells by ROS-mediated mechanisms: a radical therapeutic approach? Nat Rev Drug Discov 2009;8: Sablina AA, Budanov AV, Ilyinskaya GV, et al. The antioxidant function of the p53 tumor suppressor. Nat Med 2005;11: Varfolomeev E, Goncharov T, Fedorova AV, et al. ciap1 and c-iap2 are critical mediators of tumor necrosis factor alpha (TNFalpha)-induced NF-kappaB activation. J Biol Chem 2008;283: Belz K, Schoeneberger H, Wehner S, et al. Smac mimetic and glucocorticoids synergize to induce apoptosis in childhood ALL by promoting ripoptosome assembly. Blood 2014;124: Stadel D, Cristofanon S, Abhari BA, et al. Requirement of nuclear factor kappab for Smac mimetic-mediated sensitization of pancreatic carcinoma cells for gemcitabine-induced apoptosis. Neoplasia 2011;13: Steinhart L, Belz K, Fulda S. Smac mimetic and demethylating agents synergistically trigger cell death in acute myeloid leukemia cells and overcome apoptosis resistance by inducing necroptosis. Cell Death Dis 2013;4:e Li W, Li B, Giacalone NJ, et al. BV6, an IAP antagonist, activates apoptosis and enhances radiosensitization of non-small cell lung carcinoma in vitro. J Thorac Oncol 2011;6: Vellanki SH, Grabrucker A, Liebau S, et al. Small-molecule XIAP inhibitors enhance gammairradiation-induced apoptosis in glioblastoma. Neoplasia 2009;11: Fulda S, Galluzzi L, Kroemer G. Targeting mitochondria for cancer therapy. Nat Rev Drug Discov 2010;9: Chauhan D, Neri P, Velankar M, et al. Targeting mitochondrial factor Smac/DIABLO as therapy for multiple myeloma (MM). Blood 2007;109: Gaither A, Porter D, Yao Y, et al. A Smac mimetic rescue screen reveals roles for inhibitor of apoptosis proteins in tumor necrosis factoralpha signaling. Cancer Res 2007;67: Krepler C, Chunduru SK, Halloran MB, et al. The novel SMAC mimetic birinapant exhibits potent activity against human melanoma cells. Clin Cancer Res 2013;19: