Loss in MCL-1 function sensitizes non-Hodgkin's lymphoma cell lines to the BCL-2-selective inhibitor venetoclax (ABT-199)

As a population, non-Hodgkin's lymphoma (NHL) cell lines positive for the t(14;18) translocation and/or possessing elevated BCL2 copy number (CN; BCL2High) are exquisitely sensitive to navitoclax or the B-cell lymphoma protein-2 (BCL-2)-selective inhibitor venetoclax. Despite this, some BCL2High cell lines remain resistant to either agent. Here we show that the MCL-1-specific inhibitor A-1210477 sensitizes these cell lines to navitoclax. Chemical segregation of this synergy with the BCL-2-selective inhibitor venetoclax or BCL-XL-selective inhibitor A-1155463 indicated that MCL-1 and BCL-2 are the two key anti-apoptotic targets for sensitization. Similarly, the CDK inhibitor flavopiridol downregulated MCL-1 expression and synergized with venetoclax in BCL2High NHL cell lines to a similar extent as A-1210477. A-1210477 also synergized with navitoclax in the majority of BCL2Low NHL cell lines. However, chemical segregation with venetoclax or A-1155463 revealed that synergy was driven by BCL-XL inhibition in this population. Collectively these data emphasize that BCL2 status is predictive of venetoclax potency in NHL not only as a single agent, but also in the adjuvant setting with anti-tumorigenic agents that inhibit MCL-1 function. These studies also potentially identify a patient population (BCL2Low) that could benefit from BCL-XL (navitoclax)-driven combination therapy.


INTRODUCTION
Apoptosis or programmed cell death is an evolutionarily conserved cellular process that is required for normal embryonic development and maintenance of tissue homeostasis. The B-cell lymphoma protein-2 (BCL-2) family of proteins are essential regulators of apoptosis, functioning as either activators or inhibitors of cell death primarily at the mitochondrial level. This family of proteins consists of three groups that each contain at least one BCL-2 homology (BH) motif (BH1-4). The pro-apoptotic 'BH3-only' proteins, BID, BIM, PUMA, NOXA, BAD, BIK, BMF and HRK are activated or induced by cell death stimuli that, in turn, may activate the pro-apoptotic 'multidomain effector' proteins BAX and BAK. Once activated, these proteins homo-oligomerize to induce mitochondrial outer membrane permeabilization. Mitochondrial outer membrane permeabilization results in the release of pro-apoptotic factors such as cytochrome c from the mitochondria into the cytosol leading to apoptosome formation, caspase activation and DNA fragmentation. The anti-apoptotic members (BCL-2, BCL-X L , MCL-1, BCL-W and BFL-1) contain multiple BH motifs and function to inhibit apoptosis by direct interaction with the 'BH3-only' and multi-domain effectors via their BH3-binding grooves. Aberrant expression and/or function of BCL-2 family members results in deregulation of apoptosis that contributes to the development of a variety of human pathologies including cancer, neurodegeneration and autoimmunity. 1,2 Non-Hodgkin's lymphoma (NHL) represents a heterogeneous group of lymphoid-derived malignancies that include follicular lymphoma, diffuse large B-cell lymphoma and mantle cell lymphoma (MCL). The t(14;18) chromosomal translocation results in BCL2 hyperexpression by juxtaposing it to the immunoglobulin heavy chain gene enhancer, representing the primary tumorigenic event in most follicular lymphomas that is also found in~20% of diffuse large B-cell lymphomas. 3,4 Elevated expression of BCL-2 in diffuse large B-cell lymphoma is also associated with BCL2 gene amplification or transcriptional upregulation through constitutive activation of the NFκB pathway. 5,6 BCL-2 overexpression is associated with poor prognosis 5,7 in NHL by promoting cell survival and resistance to anti-tumorigenic agents. 1,2,8 Transgenic mouse models also reveal that MCL-1 and BCL-X L hyperexpression contribute to the onset and maintenance of hematological malignancies. [9][10][11][12] Navitoclax (ABT-263) is an orally bioavailable anti-tumorigenic agent that targets BCL-2, BCL-X L and BCL-W but not MCL-1 or BFL-1 (ref. 13) and is being evaluated in clinical trials as a single agent or in the adjuvant setting. However, BCL-X L -driven thrombocytopenia has been dose limiting in patients with hematological malignancies or small cell lung cancer. [14][15][16][17][18][19] Consequently, we developed the BCL-2-selective inhibitor venetoclax (ABT-199) that shows superior affinity to BCL-2 relative to navitoclax and circumvents BCL-X L -driven thrombocytopenia. 20 This attribute may permit attainment of higher plasma concentrations that translate into improved response rates in patients with BCL-2-dependent malignancies. Despite this, some cell lines of hematologic origin remain resistant to both venetoclax and navitoclax. 20 Although BCL2 is frequently mutated in NHL, 21,22 these mutations do not affect sensitivity to  and are unlikely to affect navitoclax or venetoclax efficacy. Mutations have been described in murine BCL2 following ABT-737/venetoclax acquired resistance, 23 however the analogous mutations in human BCL2 have not been reported in NHL patients. Therefore, potential inherent resistance factors may reside elsewhere in the apoptotic pathway. For example, MCL-1 has been identified by us and numerous other investigators as a factor that contributes to both intrinsic and acquired resistance to ABT-737, navitoclax and venetoclax. [24][25][26][27][28] Merino et al. 29 have suggested that navitoclax is not an efficient antagonist of BCL-X L in lymphoid cells, indicating that BCL-X L is in fact a resistance factor for ABT-737 (refs 29,30) and potentially navitoclax as well as, more obviously, venetoclax. Using highly potent and selective inhibitors of BCL-2, (ref. 20) BCL-X L (ref. 31) or MCL-1, (refs 27,28,32) and combinations thereof, we sought to further classify the survival dependency of NHL for anti-apoptotic BCL-2 family members. Consequently, these pre-clinical data inform on strategies to potentially improve on the clinical efficacy of venetoclax through co-inhibition of MCL-1 function.

MATERIALS AND METHODS
Reagents, cell culture and treatment NHL cell lines were obtained from the American Type Culture Collection or Deutsche Sammlung von Mikroorganismen und Zellkulturen and were cultured in Iscove's Modified Dulbecco's Media containing 10% human serum and 10 mM L-glutamine (all from Invitrogen Corporation, Carlsbad, CA, USA). All cell lines were tested for authenticity by short tandem repeat profiling and mycoplasm by the AbbVie Core Cell Line Facility. Cells were plated at a density of 0.25 × 10 6 cells/well in six-well plates for apoptosis assays, at 0.1 × 10 6 /ml for cell viability assays, and at 3 × 10 6 per 10 cm 2 petri dish for western blots. Navitoclax, venetoclax, A-1210477 and A-1155463 were dissolved in anhydrous dimethyl sulfoxide to a stock solution of 10 mM. Flavopiridol was dissolved in dimethyl sulfoxide at 1 mM. After overnight attachment, cells were treated for up to 48 h with vehicle alone, navitoclax, venetoclax, A-1155463, flavopiridol or A-1210477, or in the described combinations. Where indicated, cells were pre-treated for 60 min with z-VAD-fmk (50 μM; MP Biomedicals, Santa Ana, CA, USA). Navitoclax, venetoclax, A-1155463 and A-1210477 were synthesized as described. 20,[31][32][33] Unless otherwise indicated, all chemical reagents were obtained from Sigma Aldrich (St. Louis, MO, USA).

Cell viability
Cells (0.1 × 10 6 /ml) were treated in 96-well plates for 72 h and cell viability determined by CellTiter-Glo as described by the manufacturer's instructions (Promega Corporation, Madison, WI, USA). Responses were determined as a percentage of the control treated cells and EC 50 s determined from sigmoidal dose-response curves using GraphPad Prism (GraphPad Software, La Jolla, CA, USA).
Determination of BCL2, BCL2L1 and MCL1 CN DNA was isolated from NHL cell lines using DNeasy blood and tissue kit (Qiagen, Venlo, Netherlands; #69506) per manufacturer's protocol, except eluted in reduced EDTA TE buffer (Teknova, Hollister, CA, USA; T0223) and quantitated with PicoGreen assay (Molecular Probes, Thermo-Fisher, Waltham, MA, USA). Copy number was determined by SNP 6.0 assay (500 ng DNA input) per manufacture's protocol (Affymetrix cytogenetics copy number assay rev. 2) followed by data smoothing and quantitation of CEL files in Partek software (Partek Inc., St Louis, MO, USA).
Protein expression BCL-2, BCL-X L and MCL-1 protein expression were measured using an assay developed based on the Luminex technology (Austin, TX, USA). In brief, MCL-1, BCL-2 and BCL-X L capture antibodies were custom conjugated to Luminex carboxyl beads (bead region 9, 33 and 64, respectively) by Millipore (St. Charles, MO, USA). MCL-1 detection antibody was also conjugated to biotin through a custom service provided by Millipore. BCL-2 and BCL-X L detection antibodies conjugated to biotin were included in the DuoSetIC kits from R&D Systems (Minneapolis, MN, USA). Cells were lysed in MILLIPLEX MAP lysis buffer 1 (Millipore Cat. no. 43-040, Danvers, MA, USA) containing protease inhibitor cocktail (Sigma). Protein expression was determined using a Luminex FlexMap 3D system (Luminex) as described in depth elsewhere. 35 Data are presented as median fluorescent intensity.

Electrochemiluminescent ELISA
Streptavidin multi-array 96-well plates (Meso Scale Discovery (MSD), Gaithersburg, MD, USA) were used to immobilize biotin-labeled anti-BCL-2 (US Biological, catalog no. B0807-067), biotin-labeled anti-BCL-X L (Abcam, catalog no. ab25062), biotin-labeled anti-MCL-1 (NeoMarker, catalog MS-681-B) and biotin-labeled IgG1 (US Biological, catalog no. 11904-6A2). Protein samples (75 μg; extracted with CHAPS buffer containing protease and phosphatase inhibitors; Roche and Sigma, respectively), were subsequently added to each plate in duplicate. The plate was incubated overnight at 4°C to pull down BCL-2. After three washes with PBS-tween, anti-BIM (Epitomics; catalog no. 1036-1) was added and incubated for 1 h in the dark at room temperature with rotation at 650 r.p. m. Subsequently, sulfo-tagged goat anti-rabbit antibody (MSD; Rockville, MD, USA) was added to each well and incubated for a further 30 min as mentioned above and then washed three times with PBS-tween. Finally, 150 μl of 2 × MSD read buffer T was added per well and fluorescence measured with a MSD Sector Imager 6000 (MSD, Gaithersburg, MD, USA).

Statistical analysis
Data are represented as the mean ± s.e.m. In all cases, the number of independent experiments is described within the figure legend. The Mann-Whitney U-test was used to determine statistical significance. Spearman's rank correlation co-efficient was used to determine statistical dependence between two variables. The Bliss independence model was used to evaluate synergy. 36

RESULTS
We recently described the BCL-2-selective inhibitor venetoclax to show superior potency to navitoclax in pre-clinical models of hematological cancers. Venetoclax in vitro potency correlates with the expression of BCL-2 in NHL cell lines. Furthermore, segregation of NHL cell lines into BCL2 High (t(14;18) + and/or high BCL2 CN) and BCL2 Low groups identifies the former group as being particularly sensitive to venetoclax. 20 Here we have characterized additional NHL cell lines for sensitivity to venetoclax and navitoclax as well as To assess the contribution of BCL-X L for survival, we treated all NHL cell lines with BCL-X L -selective inhibitor A-1155463. 31 As expected, BCL2 High cell lines were resistant to the BCL-X L -selective inhibitor A-1155463 (EC 50 45 μM). Although the majority of BCL2 Low NHL cell lines were also resistant to navitoclax, SU-DHL-8 and RCK8 were sensitive. Importantly, navitoclax sensitivity in SU-DHL-8 and RCK8 was driven by BCL-X L inhibition since both cell lines were sensitive to A-1155463 and resistant to venetoclax (Figure 1a and Supplementary Figure 1). We next assessed protein expression of the anti-apoptotic BCL-2 family members in these NHL cell lines using the Luminex FlexMap 3D  Table 1) and segregated according to their BCL2 High or BCL2 Low status. 20 Data are presented as the mean of at least three independent experiments. Cell lines with navitoclax EC 50 s 42 μM were deemed resistant. Data are presented as the mean of at least three independent experiments (a). Expression of anti-apoptotic BCL-2 family proteins was determined by Luminex as described and segregated according to the BCL2 High or BCL2 Low status. The median is shown in red and the Mann-Whitney U-test was used to determine statistical significance. NS, not significant (b). Navitoclax-resistant BCL2 High cells were treated with navitoclax or venetoclax (both 1 μM) for 8 h and the interaction of BIM with BCL-2 or MCL-1 assessed using an Electrochemiluminescent ELISA (MSD) as described in the Materials and Methods section. Data are presented as the mean ± s.e.m. of three independent experiments (c).
system. 35 As expected, BCL-2 protein expression was significantly higher in BCL2 High cell lines relative to BCL2 Low cell lines. However, MCL-1 and BCL-X L protein levels were approximately the same in either population (Figure 1b).  Figure 1c). We hypothesize that MCL-1 functions as a sink in reserve that sequesters free BIM and prevents subsequent BAX activation. We therefore treated a panel of navitoclax-resistant BCL2 High cell lines with the recently described MCL-1-specific inhibitor A-1210477 (refs 28,32) in combination with navitoclax. A-1210477 alone had minimal effect on cell viability but substantially sensitized resistant BCL2 High NHL cell lines to navitoclax (Figure 2a). To determine whether the synergy between navitoclax and A-1210477 was driven by BCL-2 or BCL-X L inhibition, we treated these cell lines with the BCL-2-selective inhibitor venetoclax or the BCL-X L -selective inhibitor A-1155463 as adjuvants to A-1210477. A-1210477sensitized BCL2 High NHL cells to venetoclax but not A-1155463 (Figure 2b and c). The synergy observed with A-1210477 was often greater than that obtained with navitoclax and A-1210477 ( Figure 2c). Navitoclax and venetoclax but not A-1155463 also synergized with A-1210477 in more sensitive BCL2 High NHL cells such as DB (venetoclax EC 50 is 166.5 nM) and OCI-Ly2 (venetoclax EC 50 is 958.3 nM; Figure 2c). Conversely, chemical dissection of synergy between navitoclax and A-1210477 in BCL2 Low lines revealed cell death to be primarily driven by BCL-X L inhibition (Figure 3a). Treatment of navitoclax-sensitive or resistant BCL2 Low NHL cell lines (SU-DHL-8, SU-DHL-10 and OCI-Ly-7) with navitoclax or A-1155463 resulted in dissociation of BIM from BCL-X L and enhanced MCL-1:BIM interactions (Figure 3b).
The CDK inhibitor flavopiridol has been assessed clinically in NHL patients [37][38][39][40] and is a transcriptional repressor of MCL-1 expression. 41,42 We have demonstrated that A-1210477 induces a cellular phenotype similar to that observed following repression of MCL-1 expression through inhibition of transcriptional elongation by flavopiridol. 27 Flavopiridol therefore represents a clinically relevant surrogate for inhibiting MCL-1 function. Similarly to A-1210477, flavopiridol co-treatment sensitized BCL2 High NHL cell lines to navitoclax-or venetoclax-induced cell death but not A-1155463 in a synergistic manner (Figure 4a and b). In fact, synergy between navitoclax, venetoclax or A-1155463 and A-1210477 was predictive of synergy with flavopiridol in all NHL cell lines analyzed (Figure 4c). Flavopiridol reduced the expression  (Figure 5a and b). To further validate the mechanism of cell death and synergy, we evaluated annexin-v/7-AAD staining by flow cytometry in the presence or absence of the broad spectrum caspase inhibitor z-VAD-fmk. A-1210477 potentiated the degree of annexin-v/7-AAD staining induced by either navitoclax or venetoclax. This staining was inhibited by z-VAD-fmk, indicating that caspase activity is required for this apoptotic phenotype. Similarly, flavopiridol-sensitized BCL2 High NHL cells to venetoclax-mediated apoptosis as evidenced by high annexin-v/7-AAD staining that was caspase-dependent (Figure 5c and d).

DISCUSSION
Although navitoclax has shown encouraging activity in hematologic malignancies as a single agent and in the adjuvant setting, its clinical utility is limited by thrombocytopenia driven by inhibition of BCL-X L . [14][15][16][17][18][19] We recently described the development of the BCL-2-selective inhibitor venetoclax that shows superior affinity for BCL-2 and excellent selectivity over BCL-X L . This translates into increased potency and efficacy in pre-clinical models of lymphoid malignancies that are dependent on BCL-2 for survival. 20 Importantly, its lack of affinity for BCL-X L circumvents thrombocytopenia, a dose-limiting toxicity associated with navitoclax. 20 Subsequently, objective responses have been obtained in clinical trials of venetoclax in chronic lymphocytic leukemia 43 and NHL 44 patients. MCL-1 and BCL-X L are intrinsic and acquired resistance factors that limit the efficacy of navitoclax or ABT-737 (refs 25-29) and therefore may impact the clinical utility of venetoclax. We therefore sought to understand the functional roles of MCl-1 and BCL-X L in NHL cell lines with intrinsic resistance to venetoclax.
We have segregated NHL cell lines into two populations; BCL2 High represents lines with high BCL2 CN and/or the BCL2 translocation t(14;18), whereas lines without these lesions were defined as BCL2 Low . As a population, BCL2 High NHL cell lines are largely sensitive to navitoclax-or venetoclax-induced apoptosis. Despite this, some BCL2 High NHL cell lines are relatively resistant to venetoclax, with EC 50 s 42 μM in vitro, a facet we hypothesized here to be a consequence of MCL-1 function and not simply expression. Expression of MCL-1 at the protein level does not directly correlate with resistance to navitoclax or venetoclax in NHL cell lines herein, or at the gene level with ABT-737 in chronic lymphocytic leukemia. 24 Treatment of resistant NHL cell lines with navitoclax, venetoclax or A-1155463 resulted in enhanced MCL-1: BIM interactions that we hypothesized to inhibit BAX/BAK activation and subsequently limit the efficacy of these compounds. This capacity of MCL-1 to function as a 'sink' for additional free or displaced BIM serves as a survival response to cellular stress mediated by BCL-2 and/or BCL-X L inhibition. However, this process also primes these cells for death by agents that inhibit MCL-1 function.
Loss of MCL-1 function through gene silencing or indirect pharmacological inhibition sensitizes many tumor types to navitoclax. [24][25][26][27][28] A-1210477 is a MCL-1-specific inhibitor that induces apoptosis in a phenotypically identical fashion to MCL-1 siRNA. 27,28,32 Herein, we have used A-1210477 and other selective BCL-2 family inhibitors to define the contributions of MCL-1, BCL-2 and BCL-X L in maintaining the survival of various NHL cell lines. These 'chemical parsing' experiments 45 demonstrated that sensitization of BCL2 High NHL cell lines to navitoclax in response to direct MCL-1 inhibition with A-1210477 or indirectly through loss in MCL-1 expression mediated by flavopiridol, is driven by BCL-2 inhibition with no contribution from BCL-X L . The BCL-2-selective inhibitor venetoclax was equivalent, if not slightly superior, to navitoclax in inducing caspase-dependent cell death in synergy with A-1210477. The BCL-X L -selective inhibitor A-1155463 showed no single agent activity and did not synergize with A-1210477 in BCL2 High NHL cells in vitro, further emphasizing the importance of BCL-2 inhibition in the BCL2 High subtype.
Synergy between A-1210477 and navitoclax, venetoclax or A-1155463 correlated with that observed with flavopiridol ( Figure 4). However, the degree of synergy observed with flavopiridol and venetoclax or navitoclax is less than that observed with A-1210477; perhaps because loss in MCL-1 function achieved through cellular exposure to A-1210477 is mechanistically distinct from that of flavopiridol. A-1210477 binds to the BH3-binding groove of MCL-1 and results in stabilization of MCL-1 protein levels. 28,32 This is analogous to that observed with BIM BH3 peptides. 46 In contrast, flavopiridol treatment results in a loss in MCL-1 protein expression (Figure 5a and b) through transcriptional repression of MCL-1. 41,42 However, CDK9 regulates several other genes that dictate cellular survival, 47 and flavopiridol's effect on their expression may also contribute to synergy with navitoclax, venetoclax or A-1155463.
Despite their BCL2 Low NHL classification, SU-DHL-8 and RCK8 cell lines are sensitive to navitoclax. In this case, chemical parsing 45 experiments revealed that this efficacy was driven by BCL-X L inhibition since the BCL-X L -selective inhibitor A-1155463 was efficacious, whereas the BCL-2-selective inhibitor venetoclax was not. These data are in contrast to a recent finding by Merino et al., 29 who proposed that navitoclax does not bind to BCL-X L with sufficient avidity to kill lymphoid cells efficiently. Indeed, we found that navitoclax was able to significantly perturb the BCL-X L : BIM interactions (Figure 3b). The dependency of SU-DHL-8 and RCK8 on BCL-X L for survival is further exemplified at the protein and gene level. These cell lines are characterized as possessing low BCL-2 protein levels ( Figure 1b) and high BCL-X L (BCL2L1) CN (Supplementary Table 1). Furthermore, apoptosis and synergy between navitoclax and the MCL-1 inhibitor A-1210477 in BCL2 Low cell lines generally required BCL-X L inhibition and not BCL-2. We speculate that the BCL2 Low characterization may therefore represent a NHL patient population that may benefit from navitoclax rather than venetoclax treatment in the combination setting, such as with bendamustine/rituximab. 20,48 Taken together, the data described herein demonstrate that approaches to inhibit MCL-1 function can be combined with venetoclax, a selective BCL-2 inhibitor, in pre-clinical models of Data are presented as the mean ± s.e.m. of three independent experiments. Bliss sums obtained in both BCL2 High and BCL2 Low NHL cell lines treated with A-1210477 in combination with navitoclax, venetoclax or A-1155463 were then correlated with those obtained with flavopiridol and navitoclax, venetoclax or A-1155463 (c). Data points represent the mean of three independent experiments. Spearman rank correlation co-efficient and associated statistical significance was determined using GraphPad Prism.
NHL. Combined treatment of BCL2 High NHL cell lines with venetoclax and A-1210477 or flavopiridol results in the synergistic induction of apoptosis in vitro. Importantly, the BCL-X L -selective inhibitor A-1155463 is not efficacious as a single agent or in combination with MCL-1 inhibitors in BCL2 High NHL cell lines in vitro. Collectively these data emphasize that BCL2 status is predictive of venetoclax efficacy in NHL not only as a single agent, but also in the adjuvant setting with anti-tumorigenic agents that modulate MCL-1 levels. Finally, we demonstrate that the BCL2 Low NHL classification predicts navitoclax combinational efficacy due to a requirement for BCL-X L inhibition and not BCL-2. Elevated levels of MCL-1 have been described in chronic lymphocytic leukemia, MCL and multiple myeloma [49][50][51][52][53] and this study in preclinical models of NHL paves the way to evaluate the consequence of functional inhibition of MCL-1 in combination with venetoclax in these additional hematologic malignancies.

CONFLICT OF INTEREST
DCP, YX, LTL, LR-R, AJS and JDL are AbbVie employees and are stock holders. The design, study conduct and financial support were provided by AbbVie. AbbVie participated in the data generation, interpretation of data, review and approval of this publication. EL is now an employee of Abbott Molecular Inc.

AUTHOR CONTRIBUTIONS
DCP, YX and LTL performed the experiments, they were responsible for experimental design, data discussions and writing the manuscript; EL and LR-R performed the experiments and were responsible for writing manuscript; AJS was responsible for data discussions and writing the manuscript; JDL contributed to experimental design, data discussions and writing the manuscript.