Mitochondrial-Targeted Decyl-Triphenylphosphonium Enhances 2-Deoxy-D-Glucose Mediated Oxidative Stress and Clonogenic Killing of Multiple Myeloma Cells

Abstract

Therapeutic advances have markedly prolonged overall survival in multiple myeloma (MM) but the disease currently remains incurable. In a panel of MM cell lines (MM.1S, OPM-2, H929, and U266), using CD138 immunophenotyping, side population staining, and stem cell-related gene expression, we demonstrate the presence of stem-like tumor cells. Hypoxic culture conditions further increased CD138low stem-like cells with upregulated expression of OCT4 and NANOG. Compared to MM cells, these stem-like cells maintained lower steady-state pro-oxidant levels with increased uptake of the fluorescent deoxyglucose analog. In primary human MM samples, increased glycolytic gene expression correlated with poorer overall and event-free survival outcomes. Notably, stem-like cells showed increased mitochondrial mass, rhodamine 123 accumulation, and orthodox mitochondrial configuration while more condensed mitochondria were noted in the CD138high cells. Glycolytic inhibitor 2-deoxyglucose (2-DG) induced ER stress as detected by qPCR (BiP, ATF4) and immunoblotting (BiP, CHOP) and increased dihydroethidium probe oxidation both CD138low and CD138high cells. Treatment with a mitochondrial-targeting agent decyl-triphenylphosphonium (10-TPP) increased intracellular steady-state pro-oxidant levels in stem-like and mature MM cells. Furthermore, 10-TPP mediated increases in mitochondrial oxidant production were suppressed by ectopic expression of manganese superoxide dismutase. Relative to 2-DG or 10-TPP alone, 2-DG plus 10-TPP combination showed increased caspase 3 activation in MM cells with minimal toxicity to the normal hematopoietic progenitor cells. Notably, treatment with polyethylene glycol conjugated catalase significantly reduced 2-DG and/or 10-TPP-induced apoptosis of MM cells. Also, the combination of 2-DG with 10-TPP decreased clonogenic survival of MM cells. Taken together, this study provides a novel strategy of metabolic oxidative stress-induced cytotoxicity of MM cells via 2-DG and 10-TPP combination therapy.

Fig 1. Flow cytometric analysis of stem-like cells in HMCLs.

A. Representative dot plots for CD138 vs. side scatter and B. quantification of % CD138^low fractions. C. Hoechst 33342 staining for SP with or without verapamil. Gate represents the % SP fractions, MP = main population. D. Quantification of % SP cells in MM.1S and OPM-2 cell lines ± verapamil. Bars represent mean of three independent runs ± SEM, *p < 0.05 vs. control.

Fig 2. Hypoxia increases CD138^low population and alters transcriptional profile of HMCLs.

Cell were cultured at either normoxia (21% O₂) or hypoxia (1% O₂) for 3 days, labeled with CD138-APC antibody and the percentage of CD138^low and CD138^high cells were analyzed by flow cytometry. A. Representative dot plots of different HMCLs and B. quantification of % CD138^low fractions under normoxia or hypoxia. C. qRT-PCR analysis of SDC1, stem cell genes (NANOG, OCT4), and VEGF-A. For panels B and C, bars represent mean of three independent runs ± SEM. *p < 0.05 vs. normoxia.

Fig 3. MM stem-like cells and HPCs maintain lower steady-state ROS levels.

HMCLs were stained with APC-CD138 antibody and A. H₂DCF-DA or B. PO-1 oxidation was measured in CD138^low and CD138^high cells by flow cytometry. Results are presented as the fold change relative to CD138^high cells (set to 1). C. MM.1S or OPM-2 cells were stained with Hoechst 33342 and PO-1 oxidation was determined by flow cytometry in SP and MP cells. Results are presented as the fold change relative to MP cells (set to 1). D. Murine HPCs were enriched and H₂DCF-DA oxidation was compared between sca1^-c-Kit^- (set to 1) and sca1⁺c-Kit⁺ cells. As a positive control, menadione or H₂O₂ treatment was used. Bars represent mean of three independent runs ± SEM. *p < 0.01 vs. CD138^high/ MP/ sca1^-c-Kit^- cells.

Fig 4. Glycolytic gene expression correlates with MM patient survival and increased glucose uptake can induce 2-DG-mediated oxidative and ER stress in of CD138^low cells.

Kaplan–Meier graphs from TT2 trial clustered on glycolytic gene signatures (HK2, ALDOA, TPI1, GAPDH, PGK1, PKM2, and LDHA) showing cumulative probabilities of A. OS and B. EFS in MM patients. Glucose uptake assays in HMCLs cultured under C. normoxia or D. hypoxia for 3 days. Cells were incubated with 2-NBDG followed by APC-CD138 staining and flow analysis. For panel C, mean fluorescence values (MFI) values was normalized to CD138^high cells and presented fold change. *p < 0.01 vs. control. Bars represent mean of three independent runs ± SEM. For panel D, 2-NBDG uptake under hypoxia was compared with normoxia for CD138^low and CD138^high separately and depicted as hypoxia-induced fold change. *p < 0.01 vs. CD138^high cells under normoxia, ^#p < 0.01 vs. CD138^low cells under normoxia. E. MM.1S and OPM-2 cells were treated without or with 2-DG for 1.5 h followed by DHE staining and flow analysis. Normalized MFI relative to CD138^high cells are shown. MM.1S and OPM-2 cells were treated with 2-DG and/or mannose or 24 h followed by F. qRT-PCR analysis of BiP or ATF4 G. Western blot analysis for BiP (78 kDa), CHOP (27 kDa) or β-actin (42 kDa, loading control), tunicamycin (TM, 5 μM is used as a positive control. The quantification of BiP after normalization to untreated control is shown below each band. For panels D-F, bars represent mean of three independent runs ± SEM; *p < 0.01 vs. CD138^high or control untreated cells, ^#p < 0.01 vs. CD138^low or 2-DG treated cells.

Fig 5. CD138^low cells have altered mitochondrial properties that can be utilized to induce oxidative stress by 10-TPP treatment.

A. Representative electron micrographs of sorted CD138^low and CD138^high MM.1S and OPM-2 cells. For ultrastructure analysis, all mitochondria were selected (indicated by *), manually scored, and assigned either condensed or orthodox morphology. Low magnification shows the entire cell with inset used for analysis of mitochondria under higher magnification. *p < 0.01 vs. CD138^low cells. HMCLs were co-stained with APC-CD138 antibody and B. MitoTracker Green or C. Rhodamine 123 and analyzed by flow cytometry. Data is presented as the fold change relative to CD138^high cells. *p < 0.01 vs. CD138^high cells. D. Structure of 10-TPP; 10-TPP-induced H₂DCF-DA oxidation in CD138^high and CD138^low cells in MM1.S and OPM-2 cells. Data of three independent runs is presented as the fold change relative to CD138^high cells. *p < 0.01 vs. CD138^high cells, ^#p < 0.01 vs. CD138^low cells. E. HMCLs were transduced with Ad-CMV or Ad-MnSOD, treated with 10-TPP, and MitoSOX oxidation was analyzed by flow cytometry. Data is presented as fold change normalized to control cells expressing Ad-CMV. Antimycin A treatment was used as a positive control. *p < 0.01 vs. control cells (Ad-CMV or Ad-MnSOD), ^#p < 0.05 vs. 10-TPP treated Ad-CMV cells. Representative Western blot of HMCLs transduced with Ad-CMV or Ad-MnSOD (MOI = 50), whole-cell extract was made at 48 h and probed with an antibody against MnSOD (24 kDa) or β-actin. The quantification of MnSOD after normalization to untreated control is shown below each band. F. Structure of 10-TPVP; representative confocal images of MM.1S and OPM-2 cells stained with MitoTracker red and 10-TPVP. The bright field view and merge image of MitoTracker Red and 10-TPVP are also shown. For panels B-E, bars represent mean of three independent runs ± SEM.

Fig 6. 10-TPP and 2-DG treatment induces apoptosis and reduces clonogenic survival of HMCLs.

A. MM1.S and OPM-2 cells or B. HPCs were treated with 10-TPP and/or 2-DG for 24 h and total protein lysate was immunoblotted for caspase 3 (full length, 37 kDa and cleaved fragment, 17 kDa). β-actin was used as a loading control. The quantification of caspase 3 after normalization to untreated control is shown below each band. C. Annexin V and PI staining of MM1.S and OPM-2 cells with 10-TPP and/or 2-DG for 12 h with or without PEG-catalase (PEG-CAT). Percentage of viable, early apoptotic (annexin V⁺ PI^-), and late apoptotic/necrotic (annexin V⁺ PI⁺) for a representative of three independent experiments is shown. *p < 0.05 for annexin V⁺ PI^- fractions vs. 2-DG, TPP or 2-DG+TPP only, ^#p < 0.05 for annexin V⁺ PI⁺ fractions vs. 2-DG, TPP or 2-DG+TPP only. D. Clonogenic assays with 2-DG and/or 10-TPP. 10-TPP low and high concentrations are 0.02/0.2 μM and 0.1/1 μM, respectively. The normalized survival fraction was calculated. Data was normalized to untreated control cells. Bars represent mean of three independent runs ± SEM. *p < 0.01 vs. control, ^#p < 0.01 vs. 2-DG, and ϕp < 0.01 vs. 2-DG or 10-TPP.