Alcon et al. Cell Death and Disease (2020)11:634 https://doi.org/10.1038/s41419-020-02887-y Cell Death & Disease ARTICLE Open Access Sequential combinations of chemotherapeutic agents with BH3 mimetics to treat rhabdomyosarcoma and avoid resistance Clara Alcon1, Albert Manzano-Muñoz 1, Estela Prada 2,3, Jaume Mora2,3, Aroa Soriano 4, Gabriela Guillén 4,5 , Soledad Gallego 4, Josep Roma 4, Josep Samitier1,6,7, Alberto Villanueva 8,9 and Joan Montero 1 Abstract Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma in childhood and adolescence. Refractory/ relapsed RMS patients present a bad prognosis that combined with the lack of specific biomarkers impairs the development of new therapies. Here, we utilize dynamic BH3 profiling (DBP), a functional predictive biomarker that measures net changes in mitochondrial apoptotic signaling, to identify anti-apoptotic adaptations upon treatment. We employ this information to guide the use of BH3 mimetics to specifically inhibit BCL-2 pro-survival proteins, defeat resistance and avoid relapse. Indeed, we found that BH3 mimetics that selectively target anti-apoptotic BCL-xL and MCL-1, synergistically enhance the effect of clinically used chemotherapeutic agents vincristine and doxorubicin in RMS cells. We validated this strategy in vivo using a RMS patient-derived xenograft model and observed a reduction in tumor growth with a tendency to stabilization with the sequential combination of vincristine and the MCL-1 inhibitor S63845. We identified the molecular mechanism by which RMS cells acquire resistance to vincristine: an enhanced 1234567890():,; 1234567890():,; 1234567890():,; 1234567890():,; binding of BID and BAK to MCL-1 after drug exposure, which is suppressed by subsequently adding S63845. Our findings validate the use of DBP as a functional assay to predict treatment effectiveness in RMS and provide a rationale for combining BH3 mimetics with chemotherapeutic agents to avoid tumor resistance, improve treatment efficiency, and decrease undesired secondary effects. Introduction neurofibromatosis type 12. Based on histologic criteria, Rhabdomyosarcoma (RMS) is a highly malignant cancer RMS tumors are subdivided into two main groups, that, despite being relatively rare, is the most frequent embryonal (ERMS) and alveolar (ARMS). ERMS account soft-tissue sarcoma in children, accounting for 5% of all for 60% of all RMS, affecting children under the age of 10, pediatric tumors1. RMS tumors are highly aggressive and especially around the head and neck regions2,3. ARMS typically develop from skeletal muscle cells, arising in a represent ~20% of all RMS, occurring mostly in adoles- variety of anatomic sites in the body2,3. There is a slightly cents and frequently localized in the limbs3,4. The current higher prevalence of this disease in males than in females, treatment strategies for RMS include chemotherapy, and it is often associated with genetic disorders, such as radiation, and surgery4. Despite treatment improvement Li–Fraumeni familiar cancer syndrome and for patients with low-risk and intermediate-risk disease, the survival rates for high-risk patients have not advanced in the past decades4. Furthermore, the derived toxicities Correspondence: Joan Montero (jmontero@ibecbarcelona.eu) 1 Institute for Bioengineering of Catalonia (IBEC), Barcelona Institute of Science from current treatments and the lack of biomarkers5, and Technology (BIST), 08028 Barcelona, Spain 2 highlight the need for new therapies to enhance RMS Developmental Tumor Biology Laboratory, Institut de Recerca Sant Joan de clinical outcomes. Déu, 08950 Esplugues de Llobregat, Spain Full list of author information is available at the end of the article Edited by C. Munoz-Pinedo © The Author(s) 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 2 of 13 Therapy causes the death of cancer cells mostly by of BH3 mimetics to overcome therapy-induced apoptosis, a process controlled by the BCL-2 family of resistance7. proteins6. Its members are classified based on their Several publications by Fulda and colleagues elegantly structure, BCL-2 homology (BH) domains, and their demonstrate BH3 mimetics’ therapeutic potential to treat function6,7. In brief, the anti-apoptotic proteins (BCL-2, RMS12,18–20, although sequential combination of anti- BCL-xL MCL-1, and others) present up to four BH cancer agents with BH3 mimetics has not been fully domains (BH1–BH4) and bind to pro-apoptotic proteins. assessed. Here, we report a new strategy that utilizes low- The pro-apoptotic effector proteins BAX and BAK also dose combinations of chemotherapeutic agents with BH3 contain several BH domains and have the capacity to mimetics to increase the efficacy of current treatments, oligomerize and form pores in the mitochondrial outer while decreasing therapy-induced toxicity21 and anti- membrane. Their function is induced by activator pro- apoptotic protection. teins possessing a unique BH3 domain, such as BIM, BID (mostly through the truncated form, tBID), or PUMA. Materials and methods There is a fourth group of BCL-2 family proteins - the Cell lines and treatments so-called sensitizers - also presenting a unique BH3 RMS cell lines (CW9019, RD, and RH4) were kindly domain that cannot directly activate effectors, but can provided by Dr. Oscar Martínez-Tirado and Dr. Cristina inhibit anti-apoptotic members. Sensitizers include Muñoz-Pinedo from the Bellvitge Biomedical Research BAD, HRK, BIK, NOXA, and BMF - among others - and Institute (IDIBELL). C2C12 cells were purchased at exert a pro-apoptotic effect by competing for specific ATCC (ATCC® CRL-1772™, ATCC, Manassas, VI, USA). binding to anti-apoptotic BCL-2 family proteins7. Human skeletal muscle myoblasts (HSMM) were pur- Overall, these proteins regulate mitochondrial outer chased at Lonza (CC-2580, Lonza, Basel, Switzerland). membrane permeabilization (MOMP) and the release of RMS cell lines were cultured in RPMI 1640 medium cytochrome c (and other proteins) that represents the (31870, Thermo Fisher, Gibco, Paisley, Scotland) supple- point of no return for apoptotic cell death. Importantly, mented with 10% heat-inactivated fetal bovine serum MOMP can be prevented by anti-apoptotic proteins (10270, Thermo Fisher, Gibco), 1% of L-glutamine (25030, through direct binding to BAX and BAK or activator Thermo Fisher, Gibco), and 1% of penicillin and strep- BH3-only proteins7. tomycin (15140, Thermo Fisher, Gibco). C2C12 cells were Evasion of apoptosis is a hallmark of human cancers and cultured in DMEM high glucose medium (41965, Thermo can be often explained by the increased expression of anti- Fisher, Gibco) supplemented with 10% heat-inactivated apoptotic proteins8,9. In fact, high levels of BCL-2 and fetal bovine serum (10270, Thermo Fisher, Gibco) and 1% MCL-1 have been reported in RMS patients as a pro- of penicillin and streptomycin (15140, Thermo Fisher, survival mechanism10,11. Therefore, targeting anti- Gibco). HSMM cells were cultured in SKBM-2 medium apoptotic proteins represents a promising therapeutic (CC-3246, Lonza) supplemented with its specific Single- approach to treat high-risk or relapsed RMS patients9,12. QuotsTM and growth factors (CC-3244, Lonza). All cells In this regard, BH3 mimetics - a novel class of ther- were tested for mycoplasma and maintained at 37 °C in a apeutics that mimic the action of sensitizer BH3-only humidified atmosphere of 5% CO2. Drug treatments were proteins and selectively inhibit anti-apoptotic BCL-2 performed directly in the culture media at the doses and family members7 - could be used to overcome apoptotic time points indicated in every single experiment. All drugs resistance. There is an increasing interest in BH3 were purchased at Selleckchem (Munich, Germany). mimetics due to their therapeutic potential alone or in combination with other treatments, but the main ques- Dynamic BH3 profiling tions that clinicians must face are when and how to use DBP experiments were performed as previously BH3 mimetics as anti-cancer therapies in the clinic7. On described22,23. In brief, 3 × 104 cells/well in a 96-well plate this subject, the functional assay dynamic BH3 profiling were used for cell lines. 25 μL of BIM BH3 peptide (final (DBP) can determine in <24 h how effective a treatment concentration of 0.01, 0.03, 0.1, 0.3, 1, 3, and 10 μM), will be to engage apoptosis13. This technology uses syn- 25 μL of BAD BH3 peptide (final concentration of 10 μM), thetic BH3 peptides derived from BCL-2 family proteins 25 μL of HRK BH3 peptide (final concentration of to measure how close cells are to the apoptotic threshold 100 μM), and 25 μL of MS1 BH3 peptide24 (final con- (or how primed for death). DBP has been successfully centration of 10 μM) in MEB (150 mM mannitol, 10 mM used to predict - from days to weeks in advance - treat- hepes–KOH pH 7.5, 150 mM KCl, 1 mM EGTA, 1 mM ment effectiveness in cell lines, murine models, and EDTA, 0.1% BSA, 5 mM succinate) with 0.002% digitonin patient samples13–17. In addition to overall susceptibility were deposited per well in a 96-well plate (3795, Corning, to apoptosis, DBP can identify the selective dependence of Madrid, Spain). Single cell suspensions were stained with cancer cells on anti-apoptotic proteins, and guide the use the viability marker Zombie Violet (423113, BioLegend, Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 3 of 13 Koblenz, Germany) and then washed with PBS and for 5 min and stored at −80 °C as unbound fractions. The resuspended in MEB in a final volume of 25 μL. Cell rest of the supernatant was discarded. The resulting pellet suspensions were incubated with the peptides for 1 h was washed and mixed with 40 μL 4× SDS–PAGE sample following fixation with formaldehyde and staining with buffer and heated for 10 min at 70 °C to allow the dis- cytochrome c antibody (Alexa Fluor® 647 anti- sociation between the purified target proteins and the Cytochrome c—6H2.B4, 612310, BioLegend). Individual beads–antibody complex. The sample was magnetized DBP analyses were performed using triplicates for DMSO, and the supernatant was collected and stored at −80 °C as alamethecin (BML-A150-0005, Enzo Life Sciences, Lor- IP fractions for further immunoblotting (Western blot) rach, Germany), the different BIM BH3 concentrations analyses. used, BAD, HRK, and MS1 BH3 peptides. The expressed values stand for the average of three different readings Immunoblotting performed with a high-throughput flow cytometry SONY Proteins were separated by SDS–PAGE (Mini-Protean instrument (SONY SA3800, Surrey, UK). % priming TGX Precast Gel 12%, 456-1045, Bio-Rad) and transferred stands for the % of cytochrome c release obtained from to PVDF membranes (10600023, Amersham Hybond, different BH3 peptides, and Δ% priming represents the Pittsburgh, PA, USA). Membranes were blocked with dry difference between treated cells minus non-treated cells milk dissolved in Tris buffer saline with 1% Tween 20 for a given peptide. (TBST) for 1 h and probed overnight at 4 °C with the primary antibodies of interest directed against: rabbit anti- Cell death analysis BCL-2 (CST4223, Cell Signaling), rabbit anti-BCL-xL Cells were stained with fluorescent conjugates of (CST2764, Cell Signaling), rabbit anti-MCL-1 (CST5453, Annexin V (FITC Annexin V, 640906 or Alexa Fluor® 647 Cell Signaling), rabbit anti-BIM (CST2933, Cell Signal- Annexin V, 640912, BioLegend) and propidium iodide ing), rabbit anti-BID (CST2002, Cell Signaling), rabbit (PI) (1056, BioVision, Milpitas, CA, USA) or DAPI anti-BAK (CST12105, Cell Signaling), rabbit anti-BAX (62248, Thermo Fisher) and analyzed on a flow cytometry (CST2772, Cell Signaling), rabbit anti-Actin (CST4970, Gallios instrument (Beckman Coulter, Nyon, Switzer- Cell Signaling) followed by anti-rabbit IgG HRP-linked land). Viable cells are Annexin V negative and PI or DAPI secondary antibody (CST7074, Cell Signaling) in 3% BSA negative, and cell death is expressed as 100%-viable cells. in TBST for 1 h at room temperature. Immunoblots were developed using Clarity ECL Western substrate (1705060, Protein extraction and quantification Bio-Rad). When necessary, immunoblots were stripped in Proteins were extracted by lysing the cells for 30 min at 0.1 M glycine pH 2.5, 2% SDS for 40 min and washed in 4 °C using RIPA buffer (150 mM NaCl, 5 mM EDTA, TBS. Bands were visualized with LAS4000 imager (GE 50 mM Tris–HCl pH = 8, 1% Triton X-100, 0.1% SDS, Healthcare Bio-Sciences AB, Uppsala, Sweden) and Ima- EDTA-free Protease Inhibitor Cocktail (4693159001 geJ was then used to measure the integrated optical Roche, Mannkin, Germany)) followed by a centrifugation density of bands. at 16,100 × g for 10 min. The supernatant was stored at −20 °C for protein quantification performed using Pier- Development of RMS orthoxenograft mouse model ceTM BCA Protein Assay Kit (23227, Thermo Fisher). Six-week-old male athymic nu/nu mice (Envigo, Bar- celona, Spain) weighing 18–22 g were used in this study. Immunoprecipitation Animals were housed in a sterile environment, in cages Cells were lysed in Immunoprecipitation buffer with autoclaved bedding, food, and water. Mice were (150 mM NaCl, 10 mM Hepes, 2 mM EDTA, 1% Triton, maintained on a daily 12 h light, 12 h dark cycle. The 1.5 mM MgCl2, 10% glycerol, and EDTA-free Protease Institutional Ethics Committees approved the study pro- Inhibitor Cocktail (4693159001 Roche)) and centrifuged tocol, and the animal experimental design was approved at 14,000 × g, 15 min at 4 °C. The resulting supernatants by the IDIBELL animal facility committee (AAALAC were incubated with magnetic beads (161-4021, Bio-Rad, Unit1155). All experiments were performed in accordance Madrid, Spain) conjugated to 5 μg of rabbit anti-MCL-1 with the guideline for Ethical Conduct in the Care and antibody (CST94296, Cell Signaling, Leiden, The Neth- Use of Animals as stated in The International Guiding erlands) or 5 μg of rabbit IgG antibody (CST2729, Cell Principles for Biomedical Research Involving Animals, Signaling) at 4 °C overnight. A fraction of the supernatant developed by the Council for International Organizations (30 μL) was removed and mixed with half volume of 4× of Medical Sciences. SDS–PAGE sample buffer, heated at 96 °C for 5 min and An embryonal rhabdomyosarcoma (ERMS) orthoxeno- stored at −80 °C as cell lysate fractions. After magneti- graft was generated from a small biopsy of a metastatic zation, a part of the supernatant was mixed with half case taken at diagnostic from the primary tumor located volume of 4× SDS–PAGE sample buffer, heated at 96 °C in the child gluteus. The patient gave written consent to Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 4 of 13 participate in the study. The primary tumor did not weighed. Representative fragments were either frozen in receive radiotherapy or chemotherapy prior to surgery. nitrogen or fixed and then processed for paraffin Under isoflurane anesthesia, a subcutaneous pocket was embedding. made with surgical scissors. Then, a small incision was made in the muscle and the tumor was fixed with syn- PDX cell isolation thetic monofilament, non-absorbable polypropylene Primary tumors from PDX animals were exposed to an suture (Prolene 7.0) to the muscle of the upper thigh enzymatic digestion after mechanical disaggregation in (orthotopic implantation). After implantation, tumor 2.5 mL of DMEM media with 125 units of DNAse I formation was checked weekly by palpation. Orthotopic (DN25, Sigma-Aldrich, Buchs, Switzerland), 100 units of tumor (named RMSX1) became apparent 1–3 months Hyaluronidase (H3506, Sigma-Aldrich), and 300 units of after engraftment. Once orthotopic tumors had reached a collagenase IV (17104-019, Thermo Fisher, Gibco). The volume of around 1500 mm3, mice were sacrificed and tissue suspension was processed using gentleMACS Dis- tumors were passed to another three animals in order to sociator (Miltenyl Biotec, Madrid, Spain) using the obtain a sufficient quantity of tumor material. After each hTUMOR 1 program. The suspension was incubated at passage tumors were frozen, paraffin-embedded, and 37 °C for 30 min with constant agitation. Then, the pro- cryopreserved in (10% DMSO + 90% non-inactivated fetal gram hTUMOR 1 was ran again and repeated the 30 min bovine serum (10270, Thermo Fisher, Gibco)) to provide a incubation. We filtered the suspension with a 70 micron source of viable tissue for future experiments. filter into a 50 mL conical and cells were spun down at 500 × g for 5 min. To lyse the residual red blood cells, Drug treatment in ERMS RMSX1 orthoxenograft tumor 100 μL of ice cold water was added for 15 s and then model diluted to 50 mL with PBS, then cells were spun down The orthoxenograft procedure was approved by the again. Cells were finally resuspended in RPMI media, campus Animal Ethics Committee and complied with counted by trypan blue exclusion and plated in a 12-well Association for Assessment and Accreditation of plate, 3 × 104 cells/well and treated with DMSO or vin- Laboratory Animal Care International (AAALAC) pro- cristine 1 nM. After a 16 h incubation at 37 °C in a cedures. A mouse harboring RMSX1 tumor - orthotopi- humidified atmosphere of 5% CO2, DBP analyses were cally growing, at passage#2 - was sacrificed, tumors were performed. harvested and cut into small fragments 4 × 4 mm3, and the tumor fragments were grafted in 20 young mice. When Statistical analysis tumors reached a homogeneous size (1200–1500 mm3), Statistical significance of the results was analyzed using mice were randomly allocated into the different treatment Student’s t-tail test. *p < 0.05 and **p < 0.01 were con- groups (n = 4/group): (i) Placebo; (ii) ABT-199 (100 mg/ sidered significant. SEM stands for Standard Error of the kg); (iii) vincristine (1 mg/kg); (iv) S63845 (20 mg/kg); and Mean. For ROC curve analysis, cell lines were considered (v) combined vincristine (1 mg/kg) plus S63845 (20 mg/ responsive to treatment when Δ% cell death > 20%. Drug kg). Vincristine was intravenously administrated via tail synergies were established based on the Bliss Independent vein injection (i.v.) once per week for 3 consecutive weeks model as previously described25. Combinatorial index (CI) (days 0, 7, and 14). ABT-199 was daily administered (q.d.) was calculated CI = ((DA + DB) − (DA*DB))/DAB, where D by oral gavage (p.o.) for 21 days and S63845 was i.v. represents cell death of compound A or B or the com- administered 3 consecutive days per week for 2 weeks. All bination of both. Only the combination of drugs with a CI the animals/groups were sacrificed at day 21. To minimize < 1 were considered synergies. GraphPad Prism8 was used in combined treatments the risk of developing drug- to generate the graphs and to perform the statistical induced toxicity, drugs were administered spaced in time. analyses. Vincristine was administered first and S63845 2 h later. Vincristine from Eli Lilly (1 mg/ml) was purchased at the Results hospital pharmacy of the Catalan Institute of Oncology Novel chemotherapy combinations with BH3 mimetics to (ICO) and diluted in saline before use. ABT-199 and increase RMS cytotoxicity S63845 were purchased at Selleckchem. ABT-199 was Chemotherapeutic agents are commonly used in clinical diluted in 10% Ethanol/30% PEG 400/60% Phosal 50 PG protocols for RMS treatment4. However, they negatively (v/v/v), while S63845 was diluted in 10% DMSO/40% PEG impact patients with short-term and long-term therapy 300/5% Tween 80/saline. After treatment initiation, toxicities26, and often treatment resistance is acquired by tumors were measured using a caliper every 2–3 days and cancer cells27. Therefore, we focused on reducing che- tumor volume was calculated using the formula v = (w² × motherapeutic dosage to decrease therapy-associated l)/2, where l is the longest diameter and w the width. At undesired effects. First, we used DBP to analyze the the moment of sacrifice, tumor was dissected out and increase in priming after incubation with four standard of Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 5 of 13 Fig. 1 Dynamic BH3 profiling predicts chemotherapy sensitivity in different RMS cell lines. a Results from the DBP assay after 36 h incubation with the treatments in CW9019, RD, and RH4 cells. Results expressed as Δ% priming represents the increase in priming compared to control cells. b Cell death results from Annexin V and propidium iodide/DAPI staining FACS analyses after 96 h incubation with the chemotherapeutic agents in CW9019, RD, and RH4 cells. c Correlation between Δ% priming at 36 h and % cell death at 96 h. d Receiver operating characteristic curve analysis. Values indicate mean values ± SEM from at least three independent experiments. **p < 0.01 and *p < 0.05. care RMS chemotherapeutic agents: the microtubule- treatment and even nearly complete elimination of cells destabilizing agent vincristine, the alkylating molecule with doxorubicin, but no effect with cyclophosphamide or cyclophosphamide, the anthracycline doxorubicin, and etoposide, confirming DBP predictions (Fig. 1b). Similar the topoisomerase inhibitor etoposide28. We performed results were obtained in the other two RMS cell lines, RD DBP on three different RMS cell lines to account for the and RH4 (Fig. 1a, b). When we statistically compared Δ% disease heterogeneity: two ARMS cell lines (CW9019 and priming and % cell death in all three cell lines, we RH4) and an ERMS cell line (RD). In CW9019 cells, we observed a significant correlation (Fig. 1c). To further observed an increase in apoptotic priming upon treatment determine how good DBP is as a binary predictor for (Δ% priming) after a short incubation with vincristine and RMS, we performed a Receiver Operating Characteristic doxorubicin, but not with cyclophosphamide or etoposide (ROC) curve analysis29. We observed that the area under (Fig. 1a). Using Annexin V and PI or DAPI staining, we the curve (AUC) for our experiments was 0.81 (Fig. 1d), analyzed by flow cytometry cell death after 96 h of indicating that DBP presents a good predictive capacity exposure to the same chemotherapeutic agents as a proof for chemotherapy cytotoxicity in the RMS cell lines tested. of principle to evaluate the correlation between DBP As mentioned above, one of the hallmarks of cancer is predictions and later cell death. We observed high levels treatment adaptation and resistance to anti-cancer of cell death (between 40% and 80%) after vincristine drugs27. This resistance can be acquired by different Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 6 of 13 Fig. 2 Dynamic BH3 profiling predicts synergistic combinations with vincristine and BH3 mimetics in the CW9019 cell line. a Results from the contribution of each anti-apoptotic protein: BCL-2/BCL-xL dependence BAD peptide; BCL-xL dependence HRK peptide; and MCL-1 dependence MS1 peptide in acquiring resistance to vincristine 1 nM treatment. Results expressed as Δ% priming represents the increase in priming compared to control cells. HRK and MS1 BH3 peptides showed a significant increase, indicating BCL-xL and MCL-1 adaptation, respectively. b Cell death from Annexin V and propidium iodide staining FACS analyses after 96 h incubation of CW9019 cells with single agents or the combination of vincristine (1 nM) with the corresponding BH3 mimetics S63845 (1 µM) and A-1331852 (0.1 µM) for 96 h. Values indicate mean values ± SEM from at least three independent experiment. **p < 0.01, *p < 0.05 compared to single agents and # indicates CI < 1. mechanisms such as drug target alterations (mutations), fact, the combination index (CI) calculations25 indicated drug export transporters’ gain, increased DNA damage that S63845 addition to vincristine is synergistic (CI = repair, altered proliferation and - as we further investi- 0.916) while A-1331852 is additive (CI = 1.028). Obtain- gated - through anti-apoptotic BCL-2 proteins30. Using ing synergy between two agents is an important goal to specific synthetic BH3 peptides, that mimic sensitizer decrease treatment toxicity and to avoid undesired side BCL-2 family proteins, with DBP we can identify which is effects associated with high doses of chemotherapy, a the anti-apoptotic protein that cancer cells rely on to constant challenge for pediatric cancer25. We repeated acquire resistance to a given treatment7. In this regard, we these experiments with another RMS standard che- can precisely evaluate the contribution of three main pro- motherapeutic agent, doxorubicin, and we observed an survival BCL-2 family members: BCL-2/BCL-xL depen- increase in Δ% priming with DBP (Fig. 1a) and a high dence with the BAD BH3 peptide, BCL-xL dependence percentage of cell death with Annexin V and DAPI with the HRK BH3 peptide, and MCL-1 dependence with staining (Fig. 1b). Like vincristine, we could detect an the MS1 BH3 peptide7,22–24,31. First, we observed that increase in priming with BAD, HRK, and MS1 BH3 CW9019 cells present low initial BCL-2, BCL-xL, and peptides in CW9019 cells (Supplementary Fig. 1a) indi- MCL-1 dependence (Fig. 2a) based on the basal priming cating that cancer cells also acquired resistance to dox- observed in control cells6. However, we identified that orubicin treatment through BCL-xL and MCL-1. these cells experimented an increase in % priming, Doxorubicin is already a potent chemotherapeutic drug as referred as Δ% priming with BAD, HRK, and MS1 BH3 a single agent and exerts an extensive cytotoxicity after peptides upon vincristine treatment (Fig. 2a), pointing to 96 h (Fig. 1b), but also causes cardiotoxicity in the clinic21. BCL-xL and MCL-1 mediated pro-survival adaptation. In Therefore, we sought to reduce doxorubicin dosing by consequence, we decided to pharmacologically exploit exploring synergistic sequences with the anti-apoptotic this anti-apoptotic dependence utilizing two new selective inhibitors A-1331852 and S63845. Hereof, doxorubicin BH3 mimetics: S63845 (MCL-1 inhibitor)32 and A- combined with both BH3 mimetics was highly cytotoxic 1331852 (A-133) (BCL-xL inhibitor)33 and test their at 96 h for RMS cells, even when reducing ten-fold its cytotoxic effect in combination with vincristine. We concentration (Supplementary Fig. 1B). Both combina- observed that sequentially adding S63845 or A-1331852 tions of doxorubicin with S63845 or A-1331852 were after 36 h of exposure to vincristine significantly increased synergistic as we observed a CI = 0.522 and CI = 0.836, cell death at 96 h compared to single agents (Fig. 2b). In respectively. These combinations were effective in RMS Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 7 of 13 Fig. 3 Vincristine treatment subtly affects BCL-2 family proteins. a Anti-apoptotic protein levels analysis by Western blot from control CW9019 cells and after the treatment with 1 nM vincristine for 36 h. b Effector and activator BCL-2 family protein levels analysis by Western blot from control CW9019 cells, after vincristine (1 nM) and the sequential combination with S63845 (1 µM) for 36 h. Quantification of the optical density of each protein and normalized with actin. Results expressed as fold increase represents the increase in optical density compared to control cells. Values indicate mean values ± SEM from at least three independent experiments. cell lines but we did not observe any cytotoxic effect in most synergistic combination observed, but taking into non-tumoral myoblast cells (Supplementary Fig. 2), indi- account the cardiotoxicity caused by doxorubicin21, we cating specific toxicity of these treatments for malignant postulated the sequential treatment of vincristine and cells. Additionally, we analyzed different anti-apoptotic S63845 as the most promising and effective therapy for BCL-2 family proteins expression to determine molecular RMS, and we further studied it (Fig. 2b). fluctuations after vincristine and doxorubicin treatments. Surprisingly, we found that upon vincristine treatment Vincristine promotes MCL-1-mediated resistance there were no significant changes in the anti-apoptotic To better understand the molecular adaptation by proteins MCL-1, BCL-xL, or BCL-2 expression, indicating which cells acquire resistance to vincristine and why its that cancer cells’ adaptation to this therapy relies on dif- combination with S63845 is highly effective in vitro, we ferent mechanisms other than increased expression of analyzed MCL-1 interactions with other BCL-2 family these proteins (Fig. 3a). On the other hand, doxorubicin proteins by performing immunoprecipitation assays. First, treatment led to a different adaptation: marked decline in we found that there were no significant expression MCL-1 and BCL-xL levels, an increase in BCL-2, BAD, changes in effectors and activator BCL-2 family proteins and BIM expression and a decrease in the pro-apoptotic in cell lysates when exposed to vincristine or its sequential protein BID (Supplementary Fig. 3). From this first set of combination with S63845 (Fig. 3b). Using CW9019 cells experiments we conclude that we can increase che- we observed a significant MCL-1 decrease in the flow motherapeutic agents’ efficacy by rationally combining through fraction despite similar MCL-1 levels in initial them with specific BH3 mimetics. Despite not being the cell lysates (Fig. 4a), and a good detection in the pulled Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 8 of 13 Fig. 4 Vincristine induces resistance in RMS cells through BID and BAK inhibition by MCL-1. a Left panel: Western blot results of the unbound fraction after MCL-1 immunoprecipitation. Right panel: MCL-1 levels in the initial cell lysates. High efficiency of MCL-1 immunoprecipitation compared to Rabbit IgG control antibody. b Western blot results of the co-immunoprecipitation between MCL-1 and tBID in control conditions and after 1 nM vincristine treatment for 36 h. Results showed a significant increase in tBID and MCL-1 binding after vincristine treatment. c Western blot results of the co-immunoprecipitation between MCL-1 and BAK in control conditions, after 1 nM vincristine treatment and after the sequential combination of 1 nM vincristine and 1 µM S63845 for 36 h. Results showed a significant increase in BAK and MCL-1 binding after vincristine treatment, which was decreased below control levels after the addition of S63845. Values indicate mean values ± SEM from at least three independent experiments. **p < 0.01, *p < 0.05. down samples (Fig. 4c, d). We assessed MCL-1 binding the truncated form of BID (tBID), prevent BAX and BAK with activator proteins BIM and BID after vincristine activation, MOMP, and apoptosis34. Indeed, we could treatment. As previously described, MCL-1 can bind to observe an increase in binding between MCL-1 and tBID Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 9 of 13 after vincristine treatment (Fig. 4b) but not with BIM Overall, these results demonstrate DBP’s efficacy to (Supplementary Fig. 4A). When we studied BCL-2 family design more effective therapeutic strategies to overcome effector proteins we observed that vincristine treatment anti-apoptotic resistance and avoid cancer progression caused a significant increase in MCL-1 binding with BAK using BH3 mimetics. (Fig. 4c) but not with BAX (Supplementary Fig. 4B). These observations correlate with previous reports showing Discussion tBID preferential activation of BAK over BAX35, and There is an urgent medical need to find more effective MCL-1 sequestration of BAK as a resistance mechanism and less toxic treatments for RMS patients, since recur- to anti-cancer treatments36,37. Importantly, this enhanced rent malignancies present poor prognosis and the overall binding could be reversed by S63845 addition after vin- survival after relapse is very low31. There is a growing cristine treatment (Fig. 4c), decreasing MCL-1:BAK evidence that the BCL-2 family of proteins (particularly binding below control levels and recovering CW9019 the anti-apoptotic members) may mediate drug resistance cells’ apoptotic function. Vincristine treatment increases in cancer cells causing disease progression in patients7,8. priming and pro-apoptotic proteins tBID and BAK Therefore, it is key to predict these acquired pro-survival binding to MCL1, which can be released by S63845. These mechanisms and overcome them with anti-apoptotic findings demonstrate why the sequential combination of inhibitors like BH3 mimetics. As mentioned above, vincristine and S63845 is effective against RMS. DBP, beyond measuring a given treatment effectiveness to engage apoptosis, can also detect anti-apoptotic adapta- Effective therapeutic combination in vivo of vincristine tions derived from therapy that promote cancer survival13, with the MCL-1 inhibitor S63845 and guide the use of BH3 mimetics to avoid resistance. Patient-derived xenografts (PDXs) are advantageous in Anti-apoptotic inhibitors such as A-1331852 (BCL-xL pre-clinical research as they recapitulate patients’ ther- selective), ABT-199 (BCL-2 selective), S63845 (MCL-1 apeutic response26. After identifying different effective selective) among others that are now evaluated in the combinations in vitro, we analyzed tumors from RMS clinic, can be used as single agents or especially in com- PDX models. We disaggregated the tumors to obtain a bination with other therapies to enhance cancer elim- single-cell suspension and performed DBP analyses to ination7. In particular, highly potent and selective MCL-1 evaluate different therapies’ effectiveness and possible inhibitors such as S64315 (also named MIK665, similar to anti-apoptotic adaptations. We focused on chemother- S63845)32, AZD-599138, and AMG-17639, are currently apeutic agents, particularly on vincristine as it is used in explored in clinical trials and hold great promise for the clinic to treat RMS and because we already generated cancer treatment. Therefore, in this study we aimed to promising preliminary results in vitro in combination investigate the use of BH3 mimetics to boost RMS sen- with S63845 (Fig. 1). We analyzed an ERMS orthoxeno- sitivity to current chemotherapy. graft generated from a small biopsy from the primary At present, radiotherapy, surgery, and chemotherapy tumor located in the gluteus of a metastatic patient, that are the standard of care for RMS treatment. Regarding the we named RMSX1. We detected an increase in Δ% latter, a three-drug combination is currently utilized: priming after incubating tumor cells with vincristine (Fig. vincristine, actinomycin D, and cyclophosphamide (VAC). 5a), but not with cyclophosphamide or etoposide (Sup- This regimen has become the basis for RMS therapy with plementary Fig. 5A), similarly to what we observed in cell the incorporation of other agents, such as etoposide, lines (Fig. 1). Moreover, we identified by DBP an anti- doxorubicin, ifosfamide, cisplatin, and others for inter- apoptotic adaptation to vincristine mediated by MCL-1 mediate risk patients, with scarce clinical outcome (Fig. 5b) that could diminish its efficacy, as previously improvement40. However, secondary effects derived from observed in vitro (Fig. 2). RMSX1 in vivo single agent chemotherapy administration in children are severe and treatment with vincristine or the MCL-1 inhibitor S63845 may include infertility, cardiomyopathy, or the appear- merely delayed tumor growth after 21 days (Fig. 5c). ance of secondary neoplasia40. One explanation for these Surprisingly, we detected that the sequential combination unbearable therapy-associated pediatric toxicities relies of vincristine followed by S63845 was significantly more on differential apoptotic priming between young and effective than single agents and promoted tumor reduc- adult tissues21. Traditional chemotherapy has reached an tion in vivo (Fig. 5c and Supplementary Fig. 5C). More- efficacy plateau in RMS, making development of new over, we could also observe a significant increase in Δ% therapies that increase efficacy while decreasing toxicity a priming after incubating tumor cells with targeted agents, clear unmet need. Thus, we sought to identify possible such as S63845, ABT-199, and SP2509 (Supplementary mechanisms of resistance to chemotherapeutic agents, Fig. 5A) and we also identified possible anti-apoptotic such as vincristine or doxorubicin that could explain their adaptations to those treatments by BCL-xL and MCL-1 limited clinical efficacy by analyzing anti-apoptotic (Supplementary Fig. 5B) that we will further explore. changes with DBP (Fig. 2 and Supplementary Fig. 1). Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 10 of 13 Fig. 5 Sequential treatment of vincristine and S63845 stops tumor progression in the PDX model of RMS RMSX1. a DBP results of PDX cells from RMS cancer patient showing an increase in Δ% priming after vincristine treatment. Results expressed as Δ% priming represents the increase in priming compared to control cells. n = 2. b DBP results of PDX cells from a RMS cancer patient with the sensitizer peptides. MS1 BH3 peptide showed a significant increase, indicating MCL-1 adaptation. n = 2. c Tumor growth results after 21 days of treatment with vehicle, vincristine, the BH3 mimetics S63845 and ABT-199 and the combination of vincristine and S63845. Day 0 indicates the day animals received the treatments. All values indicate mean values ± SEM. **p < 0.01, n = 3. Indeed, we identified anti-apoptotic adaptations to com- treatments present low effectiveness especially in high risk mon therapies and tested them in combination with BH3 and relapsed RMS patients40. Previous studies in RMS mimetics to achieve a high cytotoxicity, around 80%, while also demonstrated that different BH3 mimetics can decreasing ten-fold their concentration, thus their potentiate chemotherapeutic treatment effectiveness18, potential secondary effects. More precisely, we found when combined with an ATP-competitive mTOR inhi- novel synergistic combinations of vincristine with the bitor20 or a histone deacetylase inhibitor12, which sup- MCL-1 inhibitor S63845, and doxorubicin with the same ports exploring these therapies as new approximations to BH3 mimetic or the BCL-xL inhibitor A-1331852 (Fig. 2 treat pediatric patients. and Supplementary Fig. 1); the effectiveness of this last As previously mentioned, vincristine is currently used for combination was also observed in osteosarcoma41. These RMS treatment40, but we observed that cells acquire resistance three new combinations were synergistic as assessed by CI through the anti-apoptotic protein MCL-1. Therefore, we index (CI < 1) and allowed dosing reduction40. Further- focused our efforts on testing vincristine effectiveness in more, all treatments and combinations with BH3 combination with the MCL-1 inhibitor S63845 in vivo. As mimetics reported in this work do not cause cell death PDXs accurately model patients’ outcome26, we used a RMS neither in the non-tumoral cell line C2C12 (Supplemen- PDX model to test the sequential combination of low dose tary Fig. 2A) nor in HSMM (Supplementary Fig. 2B), vincristine therapy with S63845. First, we confirmed using reinforcing the idea of reducing treatment dosage. These DBP in PDX-isolated cancer cells that vincristine resistance results strengthen the importance of BCL-xL and MCL-1 was mediated through MCL-1 (Fig. 5b), correlating with our as therapeutic targets in pediatric cancer, and specifically previous observations in vitro (Fig. 2). When a combination of in RMS as it has also been recently reported42. The vincristine followed by S63845 was sequentially administered, described treatments have been explored in multiple adult these PDXs showed a significant reduction on tumor growth cancers23,43, but not in pediatric cancers, where current with a tendency to its stabilization (Fig. 5c and Supplementary Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 11 of 13 Fig. 6 Use of S63845 to overcome RMS cells’ resistance to vincristine. When RMS cells are exposed to vincristine, they adapt to therapy by increasing the binding of BID/tBID and BAK to MCL-1. By blocking MCL-1, using the BH3 mimetics S63845, BAK will be displaced from MCL-1, and BAX and BAK will be able to oligomerize and restore vincristine-induced apoptosis. Fig. 5C), in accordance with the high cytotoxicity observed dynamic anti-apoptotic adaptations. To our knowledge, in vitro (Fig. 2b). To further explain this therapeutic strategy this is the first time that multiple effective sequential efficacy, we analyzed MCL-1 and NOXA expression but we combinations of chemotherapeutics with BH3 mimetics could not detect significant changes on those proteins (Fig. 3 are reported for RMS in the same study. The specific and data not shown), pointing to another anti-apoptotic capacity of DBP to predict resistance to treatments could mechanism driving the acquired resistance to vincristine. be key to personalize patient’s therapy and to avoid MCL-1 exerts its anti-apoptotic function by sequestering pro- toxicities derived from ineffective combinations of treat- apoptotic BCL-2 family proteins, such as activators BIM/tBID ments that do not promote cancer cell death but unde- or effectors like BAX/BAK, preventing MOMP and avoiding sired side effects. We demonstrated in pre-clinical models apoptosis34–36. When analyzing RMS cells, we observed a (Figs. 2b and 5c) the synergistic antitumor activity of the significant increase in MCL-1 binding to tBID by co- MCL-1 inhibitor S63845 when sequentially combined immunoprecipitation after vincristine treatment (Fig. 4b). with vincristine as was previously identified by DBP (Figs. tBID promotes apoptosis by preferentially binding to the 2a and 5b). Furthermore, this sequential combination will effector protein BAK35, therefore MCL-1 could decrease allow the reduction of chemotherapeutic dosing, which is treatment effectiveness. But MCL-1 can also bind to BAK to essential to decrease the secondary effects derived from protect cells from apoptosis36,37. Interestingly, we could also therapy. These findings, together with the current efforts observe an increase in MCL-1 binding to BAK after vincristine to target the anti-apoptotic protein MCL-1, currently treatment which could be reversed by the sequential addition explored in clinical trials44, manifest the importance of of S63845 (Fig. 4c). MCL-1 augmented binding to tBID and rationally combining anti-cancer agents with BH3 BAK explains CW9019 acquired resistance to vincristine and mimetics. These novel therapeutic strategies could why the sequential combination of this chemotherapeutic improve treatment of RMS patients in the clinic, espe- agent with S63845 restores cytotoxicity both in vitro and cially for those that relapsed, when guided by a functional in vivo. Thus, our findings demonstrate that MCL-1 inhibition predictive biomarker such as DBP. after vincristine treatment is critical to allow MOMP and restore apoptosis in these cells (Fig. 6). Acknowledgements We would like to thank to Dr. Martinez-Tirado and Dr. Muñoz-Pinedo for In summary, the work that we here present demon- providing the cell lines used in this study. We would also like to thank the strates DBP’s capacity to predict days in advance the Cytometry Facility from the University of Barcelona for assistance with flow cytotoxic effect of specific treatments in RMS cells. More cytometry experiments. This work was supported by the CELLEX foundation and the FERO foundation. The work was partially supported by the Networking interestingly, it can identify how RMS cancer cells acquire Biomedical Research Center (CIBER), Spain. CIBER is an initiative funded by the resistance to therapy and new approaches to overcome VI National R&D&i Plan 2008–2011, Iniciativa Ingenio 2010, Consolider Program, Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 12 of 13 CIBER Actions, and the Instituto de Salud Carlos III (RD16/0006/0012), with the 6. Brunelle, J. K. & Letai, A. Control of mitochondrial apoptosis by the Bcl-2 family. support of the European Regional Development Fund (ERDF). This work was J. Cell Sci. 122, 437–441 (2009). partially funded by the CERCA Program and by the Commission for Universities 7. Montero, J. & Letai, A. Why do BCL-2 inhibitorswork and where should we use and Research of the Department of Innovation, Universities, and Enterprise of them in the clinic? Cell Death Differ. 25, 56–64 (2018). the Generalitat de Catalunya (2017 SGR 1079). J.M. acknowledges the Ramon y 8. Frenzel, A., Grespi, F., Chmelewskij, W. & Villunger, A. Bcl2 family Cajal Program, Ministerio de Economia y Competitividad (RYC-2015-18357). We proteins in carcinogenesis and the treatment of cancer. Apoptosis also thank http://www.servier.com/Powerpoint-image-bank (licensed under 14, 584–596 (2012). Creative Commons Attribution 3.0 Unported License) for allowing us to use 9. Place, A. E. et al. Accelerating drug development in pediatric cancer: a novel some images to elaborate Fig. 6. Phase I study design of venetoclax in relapsed/refractory malignancies. Futur. Oncol. 14, 2115–2129 (2018). Author details 10. Armistead, P. M. et al. Expression of receptor tyrosine kinases and apoptotic 1 molecules in rhabdomyosarcoma: correlation with overall survival in 105 Institute for Bioengineering of Catalonia (IBEC), Barcelona Institute of Science and Technology (BIST), 08028 Barcelona, Spain. 2Developmental Tumor Biology patients. Cancer 110, 2293–2303 (2007). Laboratory, Institut de Recerca Sant Joan de Déu, 08950 Esplugues de 11. Pazzaglia, L. et al. Genetic and molecular alterations in rhabdomyosarcoma: Llobregat, Spain. 3Department of Haematology and Oncology, Hospital Sant mRNA overexpression of MCL1 and MAP2K4 genes. Histol. Histopathol. 24, Joan de Déu Barcelona, 08950 Esplugues de Llobregat, Spain. 4Group of 61–67 (2009). Translational Research in Child and Adolescent Cancer, Vall d’Hebron Research 12. Heinicke, U., Haydn, T., Kehr, S., Vogler, M. & Fulda, S. BCL-2 selective inhibitor Institute (VHIR), Universitat Autònoma de Barcelona (UAB), 08035 Barcelona, ABT-199 primes rhabdomyosarcoma cells to histone deacetylase inhibitor- Spain. 5Department of Surgery, Universitat Autònoma de Barcelona (UAB), induced apoptosis. Oncogene 37, 5325–5339 (2018). 08193 Barcelona, Spain. 6Department of Electronics and Biomedical 13. Montero, J. & Letai, A. Dynamic BH3 profiling-poking cancer cells with a stick. Engineering, University of Barcelona (UB), 08028 Barcelona, Spain. 7Networking Mol. Cell. Oncol. 3, 1–3 (2016). Biomedical Research Center in Bioengineering, Biomaterials and 14. Montero, J. et al. Blastic plasmacytoid dendritic cell neoplasm is dependent on Nanomedicine (CIBER-BBN), 28029 Madrid, Spain. 8Program against Cancer BCL2 and sensitive to venetoclax. Cancer Discov. 7, 156–164 (2017). Therapeutic Resistance (ProCURE), IDIBELL, Catalan Institute of Oncology, 15. Deng, J. et al. Bruton’s tyrosine kinase inhibition increases BCL-2 dependence l’Hospitalet del Llobregat, 08907 Barcelona, Spain. 9Xenopat S.L., Business and enhances sensitivity to venetoclax in chronic lymphocytic leukemia. Bioincubator, Bellvitge Health Science Campus, l’Hospitalet de Llobregat, 08907 Leukemia 31, 2075–2084 (2017). Barcelona, Spain 16. Townsend, E. C. et al. The public repository of xenografts enables discovery and randomized phase II-like trials in mice. Cancer Cell 29, 574–586 (2016). Author contributions 17. Wu, S. C. et al. Activity of the type II JAK2 inhibitor CHZ868 in B cell acute C.A. performed all in vitro experiments, performed the PDX tumors lymphoblastic leukemia. Cancer Cell 28, 29–41 (2015). disaggregation and their DBP analyses. A.V. and A.S. performed and analyzed 18. Meister, M. T., Boedicker, C., Klingebiel, T. & Fulda, S. Hedgehog signaling the in vivo experiments. G.G., A.S., S.G., and J.R. provided the primary tumors to negatively co-regulates BH3-only protein Noxa and TAp73 in TP53-mutated generate the RMS PDX model for in vivo experiments. E.P. and J.Mora provided cells. Cancer Lett. 429, 19–28 (2018). some reagents used in the article and expertise. J.S. and J.M. supervised the 19. Faqar-Uz-Zaman, S. F., Heinicke, U., Meister, M. T., Vogler, M. & Fulda, S. BCL-xL- work. C.A. and J.M. wrote the manuscript. selective BH3 mimetic sensitizes rhabdomyosarcoma cells to chemother- apeutics by activation of the mitochondrial pathway of apoptosis. Cancer Lett. 412, 131–142 (2018). Conflict of interest 20. Preuss, E., Hugle, M., Reimann, R., Schlecht, M. & Fulda, S. Pan-mammalian J.M. was a consultant for Oncoheroes Biosciences and Vivid Biosciences and is target of rapamycin (mTOR) inhibitor AZD8055 primes rhabdomyosarcoma a board member for The Society for Functional Precision Medicine. A.V. is cells for ABT-737- induced apoptosis by down-regulating Mcl-1 protein. J. Biol. cofounder of the Spin-off of XenOPAT SL and has ownership interests. No Chem. 288, 35287–35296 (2013). potential conflicts of interest were disclosed by the other authors. 21. Sarosiek, K. A. et al. Developmental regulation of mitochondrial apoptosis by c-Myc governs age- and tissue-specific sensitivity to cancer therapeutics. Cancer Cell 31, 142–156 (2017). Publisher’s note 22. Ryan, J., Montero, J., Rocco, J. & Letai, A. IBH3: Simple, fixable BH3 profiling to Springer Nature remains neutral with regard to jurisdictional claims in determine apoptotic priming in primary tissue by flow cytometry. Biol. Chem. published maps and institutional affiliations. 397, 671–678 (2016). 23. Montero, J. et al. Destabilization of NOXA mRNA as a common resistance Supplementary Information accompanies this paper at (https://doi.org/ mechanism to targeted therapies. Nat. Commun. 10, 5157 (2019). 10.1038/s41419-020-02887-y). 24. Foight, G. W., Ryan, J. A., Gullá, S. V., Letai, A. & Keating, A. E. Designed BH3 peptides with high affinity and specificity for targeting Mcl-1 in cells. ACS Received: 14 April 2020 Revised: 4 August 2020 Accepted: 4 August 2020 Chem. Biol. 9, 1962–1968 (2014). 25. Fitzgerald, J. B., Schoeberl, B., Nielsen, U. B. & Sorger, P. K. Systems biology and combination therapy in the quest for clinical efficacy. Nat. Chem. Biol. 2, 458–466 (2006). 26. Adamson, P. C. Improving the outcome for children with cancer: Develop- References ment of targeted new agents. CA Cancer J. Clin. 65, 212–220 (2015). 1. De Giovanni, C., Landuzzi, L., Nicoletti, G., Lollini, P. L. & Nanni, P. Molecular and 27. Melguizo, C. et al. Multidrug resistance and rhabdomyosarcoma (review). cellular biology of rhabdomyosarcoma. Future Oncol. 5, 1449–1475 (2009). Oncol. Rep. 26, 755–761 (2011). 2. Sun, X. et al. Rhabdomyosarcoma: advances in molecular and cellular biology. 28. Skapek, S. X. et al. Rhabdomyosarcoma. Nat. Rev. Dis. Prim. 5, 14–16 (2019). Sarcoma 2015, 232010 (2015). 29. Foucquier, J. & Guedj, M. Analysis of drug combinations: current methodo- 3. Hoang, N. T., Acevedo, L. A., Mann, M. J. & Tolani, B. A review of soft-tissue logical landscape. Pharmacol. Res. Perspect. 3, e00149 (2015). sarcomas: translation of biological advances into treatment measures. Cancer 30. Cree, I. A. & Charlton, P. Molecular chess? Hallmarks of anti-cancer drug Manag. Res. 10, 1089–1114 (2018). resistance. BMC Cancer 17, 1–8 (2017). 4. Belyea, B., Kephart, J. G., Blum, J., Kirsch, D. G. & Linardic, C. M. Embryonic 31. Montero, J. et al. Drug-Induced death signaling strategy rapidly predicts signaling pathways and rhabdomyosarcoma: contributions to cancer devel- cancer response to chemotherapy. Cell 160, 977–989 (2015). opment and opportunities for therapeutic targeting. Sarcoma 2012, 406239 32. Kotschy, A. et al. The MCL1 inhibitor S63845 is tolerable and effective in (2012). diverse cancer models. Nature 538, 477–482 (2016). 5. Hawkins, D. S., Gupta, A. A. & Rudzinski, E. R. What is new in the biology and 33. Leverson, J. D. et al. Exploiting selective BCL-2 family inhibitors to dissect cell treatment of pediatric rhabdomyosarcoma? Curr. Opin. Pediatr. 26, 50–56 survival dependencies and define improved strategies for cancer therapy. Sci. (2014). Transl. Med. 7, 1–12 (2015). Official journal of the Cell Death Differentiation Association Alcon et al. Cell Death and Disease (2020)11:634 Page 13 of 13 34. Clohessy, J. G., Zhuang, J., De Boer, J., Gil-Gómez, G. & Brady, H. J. M. Mcl-1 39. Caenepeel, S. et al. AMG 176, a selective MCL1 inhibitor, is effective in interacts with truncated bid and inhibits its induction of cytochrome c release hematologic cancer models alone and in combination with established and its role in receptor-mediated apoptosis. J. Biol. Chem. 281, 5750–5759 therapies. Cancer Discov. 8, 1582–1597 (2018). (2006). 40. Egas-Bejar, D. & Huh, W. W. Rhabdomyosarcoma in adolescent and young adult 35. Sarosiek, K. A. et al. BID preferentially activates BAK while BIM patients: current perspectives. Adolesc. Health Med. Ther. 5, 115–125 (2014). preferentially activates BAX, affecting chemotherapy response. Mol. Cell 51, 41. Baranski, Z. et al. Pharmacological inhibition of Bcl-xL sensitizes osteosarcoma 751–765 (2013). to doxorubicin. Oncotarget 6, 36113–36125 (2015). 36. Hockings, C. et al. Mcl-1 and Bcl-xL sequestration of Bak confers differential 42. Kehr, S. et al. Targeting BCL-2 proteins in pediatric cancer: Dual inhibition of resistance to BH3-only proteins. Cell Death Differ. 25, 719–732 (2018). BCL-XL and MCL-1 leads to rapid induction of intrinsic apoptosis. Cancer Lett. 37. Willis, S. N. et al. Proapoptotic Bak is sequestered by Mcl-1 and Bcl-x L, but 482, 19–32 (2020). not Bcl-2, until displaced by BH3-only proteins. Genes Dev. 19, 1294–1305 43. Pallis, M. et al. Complementary dynamic BH3 profiles predict co-operativity (2005). between the multi-kinase inhibitor TG02 and the BH3 mimetic ABT-199 in 38. Tron, A. E. et al. Discovery of Mcl-1-specific inhibitor AZD5991 and preclinical acute myeloid leukaemia cells. Oncotarget 8, 16220–16232 (2017). activity in multiple myeloma and acute myeloid leukemia. Nat. Commun. 9, 44. Hird, A. W. & Tron, A. E. Recent advances in the development of Mcl-1 5341 (2018). inhibitors for cancer therapy. Pharmacol. Ther. 198, 59–67 (2019). Official journal of the Cell Death Differentiation Association