Background
Multiple myeloma (MM) is an accumulative disease of mature plasma cells that is incurable in the majority of cases [
1]. Despite the development of highly effective drugs including proteasome inhibitors (e.g., bortezomib, carfilzomib), immunomodulatory agents (e.g., lenalidomide, pomalidomide), and various therapeutic CD38 antibodies (daratumumab, isatuximab, and MOR202) [
2‐
4], therapeutic resistance and disease progression invariably supervene. One of the hallmarks of MM cells is NF-κB pathway dependence [
5]. The canonical NF-κB pathway is classically activated by TNFα, whereas the non-canonical (alternative) pathway is activated by CD40 ligation and BAFF/APRIL [
6]. NF-κB is frequently activated in MM [
7] and represents a critical survival factor for these cells [
5]. Mutations in diverse NF-κB pathway components are common in MM [
8], making NF-κB a high-priority target in this disease [
9]. Notably, proteasome inhibitor (e.g., bortezomib) activity has been attributed to NF-κB inhibition [
9] presumably secondary to sparing of IκBα, which binds RelA and prevents nuclear translocation and activation [
10].
BIRC family genes (1–8) encode inhibitor of apoptosis proteins (IAPs) including NAIP, cIAP1, cIAP2, XIAP, survivin, BRUCE/apollon, livin/ML-IAP, and IAP-like protein 2 (ILP-2), respectively. Their unifying structural motif is the baculoviral IAP repeat (BIR) domain, while other important domains (e.g., RING) have recently been identified [
11]. IAPs (e.g., cIAP1/2) containing RING domains with E3 ubiquitin ligase activity trigger proteasomal degradation of proteins, including IAPs themselves [
12]. Following MOMP (mitochondrial outer membrane permeabilization), mitochondrial Smac (second mitochondrial activator of caspases) undergoes cytoplasmic release where it neutralizes IAPs, promoting apoptosis [
13]. IAPs are prevalent in many types of cancer, including MM [
14], and increased expression is associated with chemoresistance, disease progression, and poor prognosis [
14]. Moreover, pre-clinical studies suggest a role for Smac mimetics (SMs) in MM [
15]. Currently, several SMs, including the monovalent IAP antagonist LCL161 [
16] and the bivalent IAP antagonist birinapant (TL32711) [
17] have been evaluated in humans (NCT01681368).
BCL-2 family protein deregulation in MM and the ability of IAP antagonists to activate caspases directly, thus bypassing BCL-2/MCL-1 [
18], make such compound attractive candidates in this disease. Recent studies have highlighted cross talk between IAPs, and specifically cIAP1/2, in both the canonical and non-canonical NF-κB pathways, as well as the extrinsic apoptotic cascade. For example, cIAP1/2 are components of complex I containing RIP1 and TRADD implicated in canonical NF-κB pathway activation through the NEMO/IKKα/β complex [
19]. Additionally, cIAP1/2 negatively regulates the extrinsic apoptotic pathway by preventing RIP1-mediated formation of pro-death complex II (the ripoptosome), consisting of RIP1, cFLIP, FADD, and caspase-8 [
20]. While cIAP1/2 may directly mediate ubiquitination and degradation of NIK (NF-κB-initiating kinase), a key activator of the non-canonical NF-κB pathway, their predominant action is to promote ubiquitination/degradation of TRAF3 (TNF receptor-associated factor 3), required for NIK ubiquitination/degradation [
21]. Significantly, the non-canonical NF-κB pathway plays a key role in interactions between MM cells and the microenvironment (e.g., via BAFF, APRIL, CD40), which contributes to MM drug resistance [
22]. Previous studies have shown that SMs increase the activity of various conventional or targeted agents (e.g., cisplatin, TRAIL etc.) in various tumor types [
15,
23]. In MM cells, cIAP2 has been implicated in bortezomib resistance [
24]. Results of a recent study indicated that the Smac mimetic BV6 and bortezomib triggered cell death in B cell lymphoma cells in vitro [
25]. However, mechanisms by which clinically relevant SMs interact with proteasome inhibitors are largely unknown, particularly in MM, a disease in which bortezomib represents a mainstay of therapy, nor are data available involving primary MM cells or in vivo models. Here, we report that the clinically relevant SM birinapant (TL) interacts synergistically with bortezomib (Btz) in MM cells, including Btz-resistant cells, and that this interaction involves cIAP downregulation, non-canonical pathway interruption, and extrinsic apoptotic cascade activation. Moreover, similar interactions occur in primary MM cells, including populations enriched for stem cell-like cells and in in vivo models.
Materials and methods
Cell lines and reagents
Human NCI-H929, U266, and RPMI8226 cells (ATCC) were maintained as before. U266 and RPMI8226 cells were authenticated (Basic STR Profiling Service, ATCC® 135-X) by ATCC immediately after this study was completed. Btz-resistant cells (PS-R) and Btz-resistant RPMI8226 (8226/V10R) sublines were maintained as described [
26]. Primary bone marrow mononuclear cells were obtained with informed consent from patients with MM and analyzed as described in Additional file
1: Supplementary Materials and Methods.
All experiments used logarithmically growing cells (3–5 × 105 cells/ml). MycoAlert (Lonza, Allendale, NJ) assays were performed, demonstrating that cells were free of mycoplasma contamination.
Aminoactinomycin D (7-AAD) was purchased from Sigma/Aldrich, St. Louis, MO. Btz (Velcade®) was provided by Millennium Pharmaceuticals (Cambridge, MA, USA). TL32711 (NSC 756502; birinapant; TL) was provided by Tetralogic Pharmaceuticals (Indianapolis, IN, USA) through the Cancer Treatment and Evaluation Program (CTEP), NCI. The irreversible pan-caspase inhibitor (Z-VAD-FMK) was purchased from Selleck Chemicals (Houston, TX, USA). Drugs were dissolved in dimethyl sulfoxide (DMSO), aliquoted, and stored at − 80 °C; final DMSO concentrations were ≤ 0.1%.
Nuclear proteins were isolated using a Nuclear Extract kit (Active Motif, Carlsbad, CA) following the suppliers’ instructions.
Chemiluminescent DNA-binding ELISA-based assay for activated p65 and p52
DNA binding capacity of NF-κB was determined in U266 cell nuclear extracts using the TransAM® (Active Motif) NF-κB p65 or p52 Chemi Act Assay according to the suppliers’ instructions.
RNA interference, immunoblotting analysis, and analysis of cell death (apoptosis) and co-culture of MM cells with stromal cells
See Additional file
1: Supplementary Methods.
Animal studies
A xenograft murine model was used. NOD/SCID-γ mice (Jackson Laboratories) were injected subcutaneously with 5 × 106 U266 or bortezomib-resistant cells (PS-R) stably transfected with a construct encoding luciferase. Tumor growth was monitored weekly using calipers, and mean tumor volume was calculated using the formula (1/2 × [length × width2]). When mean tumor volumes reached 150–200 mm3 (18 days post-injection), animals were randomized into treatment groups. TL (TL32711) was dissolved in 12.5% Captisol (Ligand Pharmaceuticals) in distilled water. Bortezomib (Btz) was dissolved in DMSO and was diluted in 0.9% saline. Both were administered via intra-peritoneal (i.p.) injection of TL and Btz on days 1, 4, 8, and 11 of a cycle. Animals were each treated with two 14-day cycles. Control animals were injected with equal volumes of vehicle.
Mice were monitored for tumor growth with an IVIS 200 imaging system (Xenogen Corporation, Alameda, CA) weekly. Measurement of animal body weight was performed twice/week to monitor toxicity.
Statistical analysis
Values represent the means ± SD for at least three independent experiments performed in triplicate. The significance of differences between experimental variables was determined using two-tailed Student’s
t test. The significance of
P values are *
P < 0.05, **
P < 0.01, ***
P < 0.001, or ****
P < 0.0001. Synergism was determined by the method of Chou and Talalay [
27] using a commercially available software program (Calcusyn, Biosoft, Ferguson, MO). Kaplan-Meier analysis of mouse survival was performed with GraphPad Prism 6 software (La Jolla, CA), and a log-rank test (Mantel-Cox) was performed to compare survival curves.
Discussion
By bypassing inhibition of activated caspases by IAP family members (e.g., XIAP) [
36], SMs increase the activity of both conventional and novel cytotoxic agents [
37], presumably by lowering the apoptotic threshold. Recent attention has focused on the role of cIAP1/2 in governing the activity of pathways relevant to MM cell survival, including the non-canonical NF-κB pathway [
8] and the extrinsic apoptotic cascade [
38]. In this context, cIAP1 plays a complex role in regulating the activity of the non-canonical NF-κB pathway [
21], which has been implicated in microenvironmental forms of resistance in malignant B cells [
24,
39]. In addition, cIAP1 inhibits the ripoptosome (complex II), protecting cells from the lethal consequences of extrinsic apoptotic pathway activation [
20]. Such findings argue that perturbations in these pathways contribute functionally to IAPi/PI anti-myeloma activity and that this strategy may be effective in the setting of Btz or stromal factor-related forms of resistance.
The marked cIAP1 downregulation in TL/Btz-treated cells is counterintuitive given the mechanisms of the action of these agents. For example, the E3 ubiquitin ligase activity of the RING domain of cIAP1 triggers ubiquitination and subsequent proteasomal degradation of multiple proteins, including IAPs themselves [
12]. Nevertheless, the ability of SMs to downregulate cIAP1 is well described and has been attributed to E3 ligase activation requiring binding to TRAF2 and E2 [
40]. In addition, PIs might be expected to antagonize the proteasomal elimination of cIAP. However, in all MM cell types examined, combined TL/Btz exposure was associated with pronounced cIAP downregulation, arguing, albeit indirectly, that non-proteasomal mechanisms are responsible for cIAP downregulation in this setting.
Contrary to expectations, the TL/Btz regimen did not inactivate the canonical NF-κB pathway. While proteasome inhibitors such as Btz have classically been thought to inactivate NF-κB by preventing elimination of the NF-κB-inhibitory molecule IκBα [
41,
42], this has not been a universal phenomenon. For example, Btz has been shown to activate, rather than inhibit NF-κB signaling in MM and lymphoma cells [
29], possibly by triggering autophagic IκBα degradation. The results of the present studies revealed that combining Btz with TL increased phosphorylation and nuclear accumulation of p65, arguing that interruption of the canonical NF-κB does not underlie the anti-myeloma activity of the TL/Btz regimen.
In contrast, the TL/Btz regimen diminished non-canonical pathway activation, and this phenomenon contributed significantly to the regimen’s toxicity. In this regard, IAP antagonists have been referred to as “double-edged swords”/tumor suppressors [
43], as cIAP1 is involved in degradation of downstream non-canonical pathway components [
21]. However, cIAP inhibitors also upregulate the TRAF3-dependent E3 ubiquitin ligase, which opposes such effects [
44]. Notably, combined TL/Btz exposure induced marked TRAF3 accumulation, and TRAF3 knockdown significantly reduced TL/Btz toxicity. Such findings argue that the pro-apoptotic effects of TRAF3 upregulation by IAP antagonists predominate in this setting. Additionally, combined TL/Btz exposure downregulated TRAF2 and BCL-X
L, and both implicated in non-canonical NF-κB signaling [
30,
32]. The observation that both TRAF2 and BCL-X
L overexpression significantly protected cells from the regimen argues that their downregulation contributed to the regimen’s activity.
Inactivation of the non-canonical pathway by TL/Btz was also associated with circumvention of microenvironmental factor pro-survival effects. Multiple studies have highlighted links between BAFF/APRIL-related activation of the non-canonical pathway and microenvironmental forms of drug resistance, particularly in malignant B cells [
45]. The importance of stromal cells in conferring drug resistance in MM is also well established [
46]. In accord with these findings, co-culture of MM cells with soluble factors (IL-6 or VEGF) or stromal cells failed to diminish TL/Btz toxicity, consistent with the regimen’s ability to disrupt alternative NF-κB pathway signaling.
Recent attention has focused on cross talk between components of the NF-κB pathway, particularly cIAP1/2, and activation of the extrinsic apoptotic cascade [
47]. For example, cIAP1/2 negatively regulates the extrinsic apoptotic pathway by preventing RIP1-mediated formation of pro-death complex II (the ripoptosome), consisting of RIP1, cFLIP, FADD, and caspase-8 [
20]. Consistent with these observations, the TL/Btz regimen robustly triggered caspase-8 activation. Significantly, regimen toxicity was markedly reduced in cells expressing dominant-negative FADD or caspase-8, indicating that extrinsic cascade activation played an important functional role in cell death. Such findings are consistent with previous reports indicating that concomitant activation of the extrinsic pathway dramatically amplifies the lethal effects of intrinsic pathway induction [
48] and that this phenomenon may be particularly relevant to MM cells [
47].
Importantly, the TL/Btz regimen was active against primary CD138
+ MM cells, but identical exposures were minimally toxic to normal hematopoietic progenitors. The basis for selectivity of this regimen remains to be determined, but may reflect preferential killing of neoplastic versus normal cells by proteasome inhibitors [
49], the increased dependence of the former on an intact NF-κB pathway for survival [
5], and selective targeting of neoplastic cells by IAP antagonists [
31]. Interestingly, more primitive MM cell progenitors (e.g., CD138
−,CD19
+, CD20
+,CD27
+) [
35] were at least, if not more, sensitive to this regimen than their more mature counterparts. This stands in contrast to standard cytotoxic agents, to which primitive, quiescent cells are generally more resistant [
35]. Notably, as with continuously cultured lines, stromal cells failed to protect primary CD138
+ MM cells from the TL/Btz regimen and were also fully active against Btz-resistant cells expressing upregulation of MCL-1 and downregulation of BIM [
26]. The latter capacity may reflect, at least in part, the ability of IAP antagonists to bypass apoptosis inhibition conferred by perturbations in BCL-2 family members through direct activation of caspases [
36].
The present findings demonstrate that TL and Btz co-administration was well tolerated in a murine xenograft MM model, and resulted in significantly greater tumor growth inhibition and animal survival than the agents alone in both Btz-naïve MM cells and Btz-resistant MM cells. The observation that tumor cells obtained from mice displayed several of the in vitro findings (e.g., inactivation of the non-canonical NF-κB pathway, downregulation of cIAP1/2) suggests that analogous mechanisms are operative in vivo. These results differ from those of a recent study in which combining Btz with another IAP antagonist (LCL-161) in a murine MM model did not result in survival benefit [
50]. This discrepancy may reflect multiple factors, including IAP antagonist-specific actions, the different model systems employed, or potentially immunologic actions of SMs [
51]. In our study, we speculate that birinapant activates the non-canonical NF-κB pathway as a compensatory cytoprotective action. However, bortezomib may block birinapant-activated non-canonical NF-κB signaling, thereby diminishing this protective effect and further enhancing cell death. Nevertheless, the present findings argue that a mechanism-based strategy involving concomitant cIAP1 and proteasome antagonism warrants attention in MM, particularly in the setting of Btz refractoriness or other forms of drug resistance (e.g., microenvironmental). Currently, given recent trials in MM combining IAP antagonists (e.g., LCL161) with cytotoxic agents (e.g., Cytoxan; NCT01955434), the concept of combining birinapant with proteasome inhibitors such as bortezomib warrants consideration. Efforts to develop this strategy further are currently underway.