SEARCH

SEARCH BY CITATION

Keywords:

  • mitochondrial activity;
  • bortezomib;
  • multiple myeloma;
  • cyclophilin D;
  • superoxide dismutase 2

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Multiple myeloma (MM) is an incurable hematological malignancy that causes most patients to eventually relapse and die from their disease. The 20S proteasome inhibitor bortezomib has emerged as an effective drug for MM treatment; however, intrinsic and acquired resistance to bortezomib has already been observed in MM patients. We evaluated the involvement of mitochondria in resistance to bortezomib-induced cell death in two different MM cell lines (bortezomib-resistant KMS20 cells and bortezomib-sensitive KMS28BM cells). Indices of mitochondrial function, including membrane potential, oxygen consumption rate and adenosine-5′-triphosphate and mitochondrial Ca2+ concentrations, were positively correlated with drug resistance of KMS cell lines. Mitochondrial genes including CYPD, SOD2 and MCU were differentially expressed in KMS cells. Thus, changes in the expression of these genes lead to changes in mitochondrial activity and in bortezomib susceptibility or resistance, and their combined effect contributes to differential sensitivity or resistance of MM cells to bortezomib. In support of this finding, coadministration of bortezomib and 2-methoxyestradiol, a SOD inhibitor, rendered KMS20 cells sensitive to apoptosis. Our results provide new insight into therapeutic modalities for MM patients. Studying mitochondrial activity and specific mitochondrial gene expression in fresh MM specimens might help predict resistance to proapoptotic chemotherapies and inform clinical decision-making.

Strategies for multiple myeloma (MM) treatment have focused on the development of anticancer drugs that reduce the number of abnormal plasma cells in patients. Since the introduction of melphalan and prednisone therapy, numerous multidrug chemotherapies—including dexamethasone, thalidomide and nitrosoureas—have been used to treat MM.[1, 2]

Bortezomib is a first-in-class proteasome inhibitor approved by the United States Food and Drug Administration for the treatment of relapsed or refractory MM.[2] The efficacy and safety of bortezomib alone or in combination with dexamethasone or with cytotoxic agents (e.g., doxorubicin, melphalan and mitoxantrone) or with thalidomide have been established in several studies, and bortezomib has given much hope to MM patients.[3] The cytotoxic effect of bortezomib seems to arise partially from inhibition of the antiapoptotic transcription factor nuclear factor–κB.[4, 5] In addition, bortezomib has pleiotropic effects on MM biology, dysregulating Ca2+ homeostasis.[6] Furthermore, bortezomib may be involved in endoplasmic reticulum (ER) stress–induced apoptosis[7] and in the balance of the MCL-1 (antiapoptotic protein) and NOXA (proapoptotic protein) ratio.[8, 9] Bortezomib therapy, however, can cause both intrinsic and acquired resistance in MM patients.[10, 11] Additionally, the drug is associated with possible off-target toxicities.[12-15] To date, no therapy exists to overcome bortezomib resistance. Therefore, recent studies have focused on developing other proteasome inhibitors and evaluating the combination of bortezomib with other novel or conventional agents to overcome resistance in MM.

In neoplasia, cancer cells often have altered cell-death pathways and mitochondrial functions that allow them to escape cell-death programs.[16] Moreover, cancer cells tend to synthesize adenosine-5′-triphosphate (ATP) mainly through glycolysis, even under aerobic conditions, although glycolysis is a less-efficient system than oxidative phosphorylation for generating ATP.[17, 18] Hence, mitochondria-targeted compounds represent a promising approach to eradicate chemotherapy-refractory cancer cells.

In this study, we hypothesized that different sensitivities to bortezomib may be related to altered mitochondrial activity because cell death is also strictly regulated by mitochondria. To understand their roles in bortezomib-treated MM cell lines (such as KMS20 and KMS28BM), we evaluated several parameters related to mitochondrial activity and cell-death pathways in bortezomib-induced MM cell death. Both cell lines exhibited differences in mitochondrial Ca2+ concentrations, mitochondrial membrane potential (ΔΨm) and oxygen consumption rate, under nontreated and bortezomib-treated conditions. These functional differences correlate positively with drug susceptibility in the MM cell lines. In addition, the cyclophilin D (CYPD), superoxide dismutase 2 (SOD2) and mitochondrial calcium uniporter (MCU) genes were differentially expressed in KMS20 and KMS28BM. Thus, the combined effect of mitochondrial-specific genes such as CYPD, SOD2 and MCU may contribute to different sensitivities or resistance of MM cells to bortezomib.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Reagents

Bortezomib was purchased by Selleck Chemicals (Houston, TX). The fluorescent dyes tetramethylrhodamine, ethyl ester, perchlorate (TMRE), Mitotracker Green and rhod-2 AM were purchased from Invitrogen. Carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) and 2-methoxyestradiol were purchased from Sigma.

Cell culture

Human MM cell lines, KMS20 and KMS28BM are established from 65- and 77-year-old women at the Kawasaki Medical School (Kurashiki, Japan).[19-21] The cells were maintained in RPMI 1640 (Lonza). The culture medium was supplemented with 10% fetal bovine serum (Lonza) and 1% Antibiotic-Antimycotic solution (Invitrogen) in a 5% CO2 incubator. As shown in Supporting Information Table 1, we represented the characterization of two myeloma cell lines established at the Kawasaki Medical School.

Measurement of mitochondrial ROS

Mitochondrial reactive oxygen species (ROS) generation was assessed using Mito-Sox red (Molecular Probes). KMS cells were seeded on 30-mm culture dishes at a density of 3 × 105 cells, incubated with 1 µM Mito-Sox for 20 min at 37°C. For quantitative analysis of ROS generation, Mito-Sox-treated cells were analyzed by flow cytometry on a FACSCalibur instrument.

siRNA transfection

Small interfering RNA (siRNA) duplexes that target human CYPD (5′-CTGCTAAATTGTGCGTTAT-3′) and SOD2 (5′-GTGGTCATATCAATCATAG-3′) were synthesized (Sigma). siRNA targeting firefly luciferase (5′-CGTACGCGGAATACTTCGA-3′) was used as a negative control.

Flow cytometry for detection of cell death

To estimate cell death, fluorescein isothiocyanate (FITC)-conjugated annexin V–specific antibody was labeled with propidium iodide (PI), according to the manufacturer's instructions (BD Biosciences). PI- and/or annexin V-positive cells were analyzed by FACSCalibur flow cytometry (BD Biosciences). Cell-cycle distribution was determined by staining DNA with PI (Sigma).

Quantitative polymerase chain reaction (qPCR)

All reactions were performed in triplicate, and quantification of relative gene expression was analyzed using the 2−ΔΔCT method. β2-Microgrobulin (B2M) was used as an endogenous control. The primers for qPCR were designed and are summarized in Supporting Information Table 2.

Mitochondrial membrane potential and Ca2+ concentration assay

Mitochondrial transmembrane potential (ΔΨm) and mitochondrial Ca2+ concentration ([Ca2+]M) were measured with confocal microscope and FACScalibur, using specific fluorescence probes, TMRE and rhod-2AM (Invitrogen), respectively.

Mitochondria oxygen consumption assay and ATP assay

Mitochondrial oxygen consumption was determined polarographically using the fiber-optic oxygen monitor NeoFox (Ocean Optics), and the firefly luciferin-luciferase assay was used for determination of mitochondrial ATP concentration. The assay was done following the guidelines of the ATP bioluminescent assay kit (Sigma).

Statistical analysis

Data were analyzed using Student's t-test with SigmaPlot 8.0 software; the p-value was derived to assess the statistical significance, indicated as follows: *, p < 0.05; **, p < 0.01; ***, p < 0.001. Data for all figures are expressed as the means ± standard deviation (SD) of three independent experiments. Full methods are described in the Online Supporting Information.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Bortezomib induces differential mitochondria-mediated cell death in MM cell lines

To investigate the connection between mitochondrial activity and bortezomib resistance in MM, we first prepared three MM cell lines—KMS20, KMS26 and KMS28BM—and examined their resistance capacity to bortezomib using FACS analysis. As shown in Supporting Information Fig. S1, KMS28BM cells were sensitive to cell death, whereas KMS20 cells, as compared with KMS26 and KMS28BM cells were resistant to bortezomib-mediated cell death. Next, we evaluated the link of cell death and mitochondrial activity with KMS20 and KMS28BM cells in response to bortezomib. These cells were cultured at various concentrations and times of bortezomib treatment, labeled with PI/annexin V-FITC and subjected to flow cytometry. Bortezomib-treated cells displayed an abrupt dose-dependent (Fig. 1a) and time-dependent (Fig. 1b) increase in cell death. When KMS cell lines were treated with 50 nM bortezomib for 48 hr, KMS28BM cells exhibited marked declines in cell survival compared with that of KMS20 cells (Fig. 1). In contrast, KMS20 cells exhibited negligible bortezomib-mediated cell death. To verify whether differential susceptibility of bortezomib is induced by apoptosis, the sub-G1 phase DNA content of the cell cycle was measured in bortezomib-treated KMS cell lines. DNA content analysis also revealed a population of cells in the sub-G1 hypodiploid phase, indicating that apoptosis increased dose-dependently in bortezomib-treated cells, and a significant difference between KMS20 and KMS28BM cells was observed in apoptotic DNA content (Fig. 1c). We next used Western blot analysis to determine whether bortezomib treatment modified caspase-3, caspase-8 and poly (ADP-ribose) polymerase in a dose-dependent manner. Immunoreactive bands for cleaved caspase-8, caspase-3 and poly (ADP-ribose) polymerase were more intense in lysates from KMS28BM cells than from KMS20 cells (Fig. 1d). Finally, we examined whether mitochondria are involved in bortezomib-induced cell death. In the mitochondrial apoptotic pathway, the signal mediator between the cytosol and the mitochondria is cytochrome c. Thus, cytochrome c release was compared in KMS20 and KMS28BM cells. Cytochrome c release was markedly increased in the cytosol of KMS28BM cells compared with that in KMS20 cells (Fig. 1e). As a result of this release, caspase-3 activation was subsequently enhanced in KMS28BM cells compared with that in KMS20 cells (Fig. 1d). These results indicate that whereas KMS28BM is a bortezomib-sensitive MM cell line, KMS20 cells have acquired resistance to bortezomib. Moreover, our results imply that mitochondria function as apoptosis regulators that direct activation of the caspase cascade by bortezomib.

image

Figure 1. KMS20 cells show drug resistance to bortezomib induced apoptotic cell death. KMS cells were incubated at different doses (a) and for different period of time (b) with bortezomib and cell viability was determined by FACS analysis after PI/annexin V-based staining. (c) The apoptotic cell population was quantified after treatment with bortezomib (Bort) as indicated. (d) KMS cells were treated with bortezomib at the indicated doses for 48 hr and were subjected to Western blotting using indicated antibodies. (e) KMS cells were treated with bortezomib for 48 hr, and cell lysates were separated into cytosolic and mitochondrial fractions. Tubulin and Prx III were used as cytosolic and mitochondrial markers, respectively. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Increased mitochondrial calcium level and membrane potential in bortezomib-resistant KMS20 cells

To investigate the connection between mitochondrial activity and bortezomib resistance in MM, we examined mitochondrial functions such as [Ca2+]M, ΔΨm, mitochondrial ROS, oxygen consumption and ATP production. First, we determined whether divergence of mitochondrial Ca2+ concentrations contributes to bortezomib-mediated cytotoxicity. KMS cell lines were incubated in the presence of a series of different bortezomib concentrations for 48 hr, treated with rhod-2AM and subjected to flow cytometry (Fig. 2a and Supporting Information Fig. S2A) and fluorescence microscopy (Supporting Information Fig. S2B). We noted an increase in [Ca2+]M with bortezomib treatment. The steady state [Ca2+]M was calculated as [Ca2+]M = fluorescence intensities (F) of Ca2+ – fluorescence intensities (Fmin) under Ca2+ free condition by ionomycin treatment. Interestingly, the basal level of [Ca2+]M was inversely measured against the pattern of cell death in KMS cell lines. Specifically, the [Ca2+]M in KMS28BM cells increased more than 10-fold after bortezomib treatment; by comparison, [Ca2+]M in KMS20 cells increased ∼2-fold. In addition, to confirm that the generated Ca2+ is localized in the mitochondria, the KMS cell lines were costained with the mitochondrion-specific dyes Mito-Tracker and rhod-2AM and analyzed with a confocal microscope (Fig. 2b). The Mito-Tracker green fluorescence was well colocalized with the rhod-2AM (red Ca2+) signals. These results show that the highly controlled mitochondrial Ca2+ pool in KMS20 cells minimized bortezomib-induced Ca2+ overload, which consequently prevented Ca2+ overload-induced apoptosis. In addition, the critical determinant of bortezomib cytotoxicity is suggested to be the ratio between the basal and induced [Ca2+]M after stimulation.

image

Figure 2. Increased mitochondrial calcium concentration and membrane potential in bortezomib-resistant KMS20 cells. (a) Comparisons of mitochondrial Ca2+ concentrations in KMS cell lines before and after treatment with bortezomib (48 hr) using a FACSCalibur (upper). Values are presented as means ± SDs (lower). (b) KMS cells were costained with the mitochondrion-specific dyes MitoTracker and rhod2-AM and then analyzed with a confocal microscope. (c) ΔΨm in KMS cell lines before and after treatment with bortezomib (48 hr) using a FACSCalibur (upper). Values are presented as means ± SDs (lower). (d) Cells were treated as in (c), subjected to flow cytometry after staining with Mito-Sox, and quantified. (e) The difference in mitochondrial oxygen consumption was measured in digitonin-permeabilized KMS cells. (f) mtATP was determined based on luciferase-catalyzed ATP-dependent oxidation of luciferin.

Download figure to PowerPoint

We next examined the effect of bortezomib on the basal level of ΔΨm in the two cell lines. The ΔΨm was measured using TMRE, which accumulates in the mitochondria via a ΔΨm-driven process. Because uptake of TMRE into mitochondria depends on ΔΨm, a collapse of ΔΨm is associated with a decrease in the fluorescence intensity of TMRE. Flow cytometric (Fig. 2c) and fluorescence microscopic (Supporting Information Fig. S2C) analyses of red fluorescence indicated that bortezomib-treated KMS20 and KMS28BM cells exhibited a depolarization of ΔΨm and that this response is differently induced by bortezomib. Consistent with the [Ca2+]M finding, ΔΨm was greater in bortezomib-resistant KMS20 cells than in bortezomib-sensitive KMS28BM cells (Fig. 2c). Especially, ΔΨm of KMS28BM cells were decreased more than 80% by treatment with bortezomib. Next, we measured the mitochondrial ROS level using an oxidant sensitive fluorescent dye, Mito-Sox red, while the bortezomib treatment was not trigger induction of superoxide anion production in KMS20 cells, the KMS28BM cells resulted in a remarkable increase of mitochondrial ROS level by bortezomib treatment. Taken together, these data suggest that increased-mitochondrial ROS is possible cause of bortezomib induced cell death in KMS28BM cells.

To better characterize the mitochondrial energy metabolism underlying the susceptibility of KMS cells to bortezomib, we analyzed their oxygen consumption rate and ATP concentrations. The basal amounts of oxygen consumption (Fig. 2e) and mtATP (Fig. 2f) were significantly greater in KMS20 cells than in KMS28BM cells. Consistent with previous data, the mtATP concentrations in KMS28BM cells, but not in KMS20 cells, were also decreased by bortezomib treatment. Our results suggest that mitochondrial functional differences including ΔΨm, ROS production, oxygen consumption rate, ATP production and Ca2+ retention capacity were involved in the susceptibility of MM cells to undergo bortezomib-mediated cell death.

Mitochondrial ΔΨm and ROS related to the susceptibility of cell death by bortezomib in KMS cells

To examine whether the cell death were induced by mitochondrial ROS in KMS28BM cells, KMS28BM cells were pretreated with membrane-targeted-Catalase (mt-CAT) adenoviruses prior to bortezomib treatment. Since the catalase is the enzyme that dismutase H2O2 into water without the help of coupling system, the subcellular targeting of catalase has often been used for studying the local redox event.[22] The modified catalase was efficiently expressed in the KMS28BM cells by adenoviral infection (Supporting Information Fig. S3A). When KMS28BM cells were infected with the adenoviruses following the bortezomib treatment, mitochondria-targeted catalase effectively eliminated the ROS elevated by the bortezomib treatment (Fig. 3a) and the scavenging of mitochondrial ROS inhibited the cell death of KMS28BM cells by bortezomib treatments (Fig. 3b).

image

Figure 3. Mitochondrial dysfunction by FCCP pretreatment boosted effect of bortezomib in KMS20 cells. (a) ROS level in KMS28BM cells infected with the mitochondria-targeted encoding viruses before the bortezomib treatment. (b) mt-CAT infected KMS28BM cells were treated with indicated dose for 48 hr. Cell viability was determined by FACS analysis after PI/annexin V-based staining. KMS20 (c) and KMS28BM (d) cells were pretreated with the indicated dose of FCCP for 1 hr and then were treated with 50 nM bortezomib for indicated time period. Cell death was measured using a CCK-8 kit, as described in Supporting Information. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

To determine whether mitochondrial function is a determinant for bortezomib resistance, we next examined the responses of KMS cell lines to bortezomib combined with FCCP, which, acting as a protonophore, induces the dissipation of ΔΨm.[23, 24] We assessed the effect of treatment with FCCP alone and found that single treatment with FCCP did not induce cell death in either cell line (Supporting Information Fig. S3B). Next, we tested the effect of treatment with bortezomib and FCCP in KMS cell lines. Pretreatment with FCCP for 1 hr significantly enhanced bortezomib-induced cell death and overcame the resistance to bortezomib in KMS20 cells (Fig. 3c), which are highly resistant to bortezomib. Specifically, in 0.5 μM FCCP-treated KMS20 cells, cell death was induced more than 50% within 24 hr. Moreover, FCCP treatment dramatically increased cell death in KMS28BM cells within 24 hr as compared with KMS28BM cells treated with bortezomib alone (Fig. 3d). Thus, concomitant treatment with bortezomib and FCCP demonstrated a greater anticancer effect than did treatment with bortezomib alone.

Differential gene expression associated with mitochondrial function results in the divergence of susceptibility to cell death by bortezomib treatment

To determine the regulators of mitochondrial activity, we next analyzed the basal expression of mitochondrial genes associated with the ΔΨm, ROS, energy metabolism and cell death in KMS cell lines (Fig. 4a). Several genes were differentially expressed in both cell lines. Particularly, the expression of CYPD, SOD2 and MCU, which are related to maintenance of the ΔΨm, elimination of ROS and Ca2+ influx into mitochondria, were correlated with the susceptibility of mitochondria-mediated cell death by bortezomib treatment (Fig. 4b). In KMS28BM cells, MCU expression was dramatically increased by bortezomib treatment. Moreover, in Western blot analysis, the pattern of protein expression of these genes demonstrated a similar result (Fig. 4c). CYPD was decreased in KMS20 cells, compared with that found in KMS28BM cells. In addition, SOD2 and pyruvate dehydrogenase (PDH)-E1α proteins levels were diminished in KMS28BM cells. The amounts of most of the detected proteins were not changed by bortezomib treatment, indicating that the bortezomib-resistance of KMS20 cells pre-exists and is not induced by the administration of bortezomib. Thus, we suggest a direct link between the divergence of mitochondrial activity by differential gene expression and the susceptibility to bortezomib in MM cells. That is, the mechanistic basis of bortezomib sensitivity can be explained not only by NOXA and MCL-1—whose protein expression ratios in cells help determine whether a cell survives or undergoes apoptosis—but also by mitochondrial genes, including CYPD, MCU and SOD2. From these results, we assume that chemoresistance of MM patients is regulated according to the specific expression of genes associated with mitochondrial function and that bortezomib can potently sensitize chemoresistant MM cells with mitochondria-damaging chemotherapeutic agents.

image

Figure 4. CYPD, SOD2 and MCU are associated with the mitochondrial mediated cell death by bortezomib treatment. (a) Mitochondrial genes were analyzed using qPCR in the KMS cell lines. (b) CYPD, SOD2 and MCU were measured using qPCR in KMS cells treated with 50 nM bortezomib for 48 hr. (c) After exposure to 50 nM bortezomib for 48 hr, KMS cells were analyzed by Western blotting using indicated antibodies. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

CYPD and SOD2 are mitochondrial regulators of bortezomib-induced cell death in MM

CypD, one of the proteins that exhibited differential expression in the cell lines in this study, is a well-known regulator of mitochondrial permeability transition (MPT).[25] SOD2 is a critical mitochondrial antioxidant protein that detoxifies superoxide produced by respiration. SOD2 overexpression protects mitochondrial respiratory function and inhibits apoptosis induction.[26] Thus, to determine whether CypD and SOD2 participate in bortezomib-induced cell death, and to investigate whether the ΔΨm and Ca2+ influx are key regulators of cell death, we synthesized and tested siRNAs that target CYPD and SOD2. CYPD and SOD2 genes were knocked-down more than 80% with siRNA treatment (Supporting Information Fig. S4). Next, we analyzed cell death in CYPD-silenced KMS cell lines in the presence of bortezomib. The viability of KMS20 cells was slightly increased with CYPD and 75 nM bortezomib (Fig. 5a). Because CypD expression is sufficiently low under physiological and bortezomib-stimulated condition without knockdown of CYPD in KMS20 cells, the survival rate was barely increased by CYPD knock-down. In the case of KMS28BM cells, cell viability was dramatically increased by depletion of CYPD (Fig. 5b). In contrast, siRNA-mediated SOD2 depletion induced cell death in KMS20 cells (Fig. 5c). This SOD2 knock-down phenomenon was rarely observed in KMS28BM cells because SOD2 is disrupted by chromosomal abnormality (Fig. 5d). In addition, we examined whether the membrane-targeted catalase rescue the SOD2-depleted KMS cells from bortezomib-induced cell death to find a direct link between the SOD2 expression and cell death, which were induced by the absence of SOD2 and bortezomib treatment. As shown in Figure 5e, when KMS cells were infected with the adenoviruses following the SOD2 knockdown, membrane-targeted catalase effectively eliminated the ROS elevated by the SOD2 knockdown (Fig. 5e). Beside, the membrane-targeted catalase rescued the cell survival in KMS20 and KMS28BM cells with the SOD2 knockdown (Fig. 5f). Taken together, these data indicate that the resistance of KMS20 cells arises in part from the stability of ΔΨm by suppressed expression of CypD and the elimination of ROS via SOD2 compared with that observed in KMS28BM cells. These results raise the possibility that the retention capacity of MPT according to MM cell type might be involved in cell death owing to the generation of MPT inducers (i.e., Ca2+ overload and ROS) by bortezomib treatment. Thus, bortezomib and the combined use of mitochondria-targeted agents may be necessary to achieve maximum apoptosis-promoting efficacy.

Because SOD2 depletion induced KMS20 cell death, we examined the effect of combination with bortezomib and 2-methoxyestradiol, a known SOD inhibitor.[27] At the concentrations tested, 2-methoxyestradiol elicited significant cytotoxicity (15–25%) as a single agent (Fig. 6a), but the combination of both agents induced apoptosis more than 2-fold greater than by each compound alone (Fig. 6b). These results show that the combination of 2-methoxyestradiol and bortezomib synergistically activates MPT and apoptosis in KMS20 cells by inhibiting SOD and therefore causing superoxide accumulation.

image

Figure 5. SOD2 knock-down cells are sensitive to bortezomib-induced cell death. KMS20 (a) and KMS28BM (b) cells were transfected with siCYPD or siCont and then were exposed to the indicated dose of bortezomib for 48 hr. Cell death was measured using PI/annexinV. KMS20 (c) and KMS28BM (d) cells were transfected with siSOD2 or siCont and then were treated and subjected in cell death assay as described in (a, b). (e) Mitochondrial ROS level in KMS20 and KMS28BM cells infected with the mitochondria-targeted encoding viruses and then the bortezomib treatment for 48 hr. (f) mt-CAT infected KMS20 cells and KMS28BM cells were treated with indicated dose for 48 hr. Cell viability was determined by FACS analysis after PI/annexin V-based staining. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

image

Figure 6. The mitochondria targeted drug 2-methoxyestradiol used in combination with bortezomib enhances the anticancer effect in KMS20 cells. (a) KMS20 cells were treated with 2-methoxyestradiol in dose-dependent manner. The extent cell death was determined by FACS analysis of annexin V/PI staining. (b) KMS20 cells were incubated at different doses as described figure with bortezomib and 2-methoxyestradiol for 24 hr and then cell viability was determined by FACS analysis after PI/annexin V-based staining.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Bortezomib, a small-molecule inhibitor of the 20S proteasome, has demonstrable anticancer activity in several cell types and has received approval for treatment of relapsed and refractory MM. Bortezomib has been associated, however, with possible off-target toxicities and the development of drug resistance.[12, 15] Here, we have shown in MM cells that mitochondrial activity is a determining factor in the regulation of apoptosis resistance in response to bortezomib. Thus, to develop an effective therapeutic regime for MM, we evaluated the involvement of mitochondria and resistance to bortezomib-mediated apoptosis in MM cells. We demonstrated the different susceptibilities of two different MM cell lines, KMS20 and KMS28BM. These findings allow us to propose a comprehensive model for this differential susceptibility to bortezomib and the role of mitochondria in bortezomib-induced cell death (Fig. 1). In bortezomib-treated cells, procaspase-8 is proteolytically cleaved into activated caspase-8, which promotes cytochrome c release, and activates caspase-3 activation, initiating mitochondria-directed apoptosis. During this process, mitochondria actively participate in bortezomib-dependent activation of the caspase cascade.

We next examined whether the difference of cell death in KMS20 cells and KMS28BM cells was derived from difference in biological characteristics, such as the activity and genomic DNA mutation of 20S proteasome. Mutations on the β5 subunit (PSMB5) of the 20S proteasome have been associated with bortezomib resistance since it reversibly inhibits the chymotrysin-like activity of the β5 subunit of the proteasome.[28, 29] As shown in Supporting Information Fig. S5, KMS20 and KMS28BM cells have possessed identical genomic DNA sequence of PSMB5 without mutation. Moreover, the activity of 20S proteasome in KMS20 cells was pretty to that in KMS28BM cells (Supporting Information Fig. S5E). Therefore, these results suggest that the difference of cell death was not caused by the mutation or activity of 20S proteasome of two cell lines.

Bortezomib-induced apoptosis is associated with increased NOXA expression, proteolytic cleavage of MCL-1,[8, 30] caspase-2 activation by the ER stress-response resulting in the induced expression of CHOP (leucine zipper transcriptional factor) and ERO1 (ER oxidase) genes,[7] and the mitochondrial-mediated dysregulation of Ca2+.[6] Thus, to determine whether the previously proposed mechanism contributes to differential susceptibility of bortezomib in the KMS cell lines, gene expression related to the proposed mechanism was examined using qPCR. Expression of NOXA and MCL-1, however, was similarly increased by bortezomib treatment in KMS20 and KMS28BM cells (Supporting Information Fig. S6A), and MCL-1 accumulation, as a result of bortezomib treatment, was also confirmed using immunocytochemistry in both cell lines (Supporting Information Fig. S6B). KMS20 cells display more resistance to bortezomib-induced cell death than do KMS28BM cells (Fig. 1). Thus, in MM cells, mitochondrial stability may be the critical determinant in response to bortezomib. To evaluate this hypothesis, we examined several mitochondrial functional parameters such as Ca2+ and ATP concentrations, ΔΨm and oxygen consumption in MM cell lines. Mitochondrial oxidative phosphorylation (OxPhos) is regulated by several mechanisms, including substrate availability.[31] The major substrates for OxPhos are oxygen, which is the terminal electron acceptor and pyruvate, which is the primary carbon source. Moreover, inhibition of oxygen consumption precedes activation of the MPT. Calcium overload also decreases mtATP production and causes the MPT pore to open.[32, 33] The mitochondrial activities of each cell line had different steady-state levels, and these levels were inversely related to the pattern of cell death in bortezomib-induced KMS cells. Our work directly shows that this highly maintained ΔΨm and [Ca2+]M is associated with increased resistance to cell death. Although the underlying mechanism is still unclear, hyperpolarization of ΔΨm is frequently found in various cancer cells[34] and may be considered an antiapoptotic mechanism utilized by cancer cells, as depolarization of ΔΨm is a trigger of MPT pore opening and subsequent apoptotic cascades. In our previous study, relatively greater ΔΨm and [Ca2+]M were found in a human gastric cancer cell line compared with that in normal cells.[35] Consistently, in this study, the KMS20 cells, having greater ΔΨm, exhibited more robust resistance to bortezomib treatment. In addition, when bortezomib was combined with FCCP, which induces dissipation of ΔΨm, bortezomib resistance was overcome in KMS20 cells (Fig. 3). Moreover, to support our hypothesis, we randomly selected two more MM cell lines including KMS26 and IM-9 then compared cell death against bortezomib and mitochondrial activities, for example, ΔΨm, ROS generation rate and mitochondrial Ca2+ fluxes (Supporting Information Fig. S7). These results were coincident with our present results, which compared mitochondrial activities of KMS20 vs. KMS28BM. Thus, we thoughtfully suggested that presented correlation between mitochondrial activities and drug resistance were not caused by different genetic background but by specific biological modulation.

The divergence of mitochondrial activities in the MM cell lines can be induced by expression of genes associated with the MPT pore complex, Ca2+ channels, antioxidant proteins and the BCL2 family. Such differences in mitochondrial gene expression were confirmed in KMS20 and KMS28BM cells. As shown in Figure 4b, CYPD expression was decreased in KMS20 cells compared with that in KMS28BM cells, SOD2 expression in KMS28BM cells was spontaneously depleted owing to chromosomal abnormality, and MCU expression was diminished in KMS28BM cells but was dramatically elevated by bortezomib treatment. Moreover, CypD was reduced in KMS20 cells; SOD2, BCL2 and PDH-E1α proteins were diminished in KMS28BM cells (Fig. 4c). The hypothesis that specific, altered gene expression promotes apoptotic resistance was further supported by the observation that CYPD and SOD2 are differentially expressed in MM patient specimens (CD45CD138+) from human patients, as compared with normal specimens (CD45+CD138; data not shown). Hence, KMS20 cells were more resistant than KMS28BM cells to bortezomib-induced cell death. CypD-deficient or SOD2-overexpressed cells showed an increased capacity to retain Ca2+ and were no longer susceptible to MPT induced by the addition of Ca2+.[25, 36] Thus, MM therapy may require target selection and analysis of mitochondrial gene expression in each patient to achieve maximum efficacy and overcome bortezomib resistance.

Does our hypothesis provide new insight into therapy to overcome apoptotic resistance in MM cells? As shown in Figure 5, KMS20 cells acquired resistance to bortezomib owing to declines in CYPD expression, and when CYPD was depleted using siRNA knock-down in KMS28BM cells, these cells showed increased survival. In addition, suppression of mitochondrial function using FCCP greatly increased the cytotoxicity of bortezomib and overcame the resistance of KMS20 cells, directly implicating mitochondrial function and drug resistance (Fig. 3). Furthermore, in KMS20 cells, SOD2 knock-down had a dramatic effect on bortezomib-induced cell death and their cell death also rescued by over-expression of mitochondria-targeted catalase (Fig. 5). Proceeding from our gene depletion experiments, we tested the combined effect of bortezomib and 2-methoxyestradiol in KMS20 cells. These results showed that cell death by combination treatment increased more than 2-fold as compared with single treatment with bortezomib or 2-methoxyestradiol. Thus, to induce mitochondria-initiated cell death arising from accumulation of mitochondrial ROS and Ca2+, bortezomib may be combined with mitochondrial-targeted agents to overcome resistance, which occurs because of changes in the expression of important mitochondrial genes such as CYPD, SOD2 and MCU.

In summary, the most interesting feature regarding the genomic instability of MM is its variability. As shown in this study, the expression of multiple mitochondrial genes is different in a variety of MM cell lines, and their levels determine the capacity for Ca2+ and ROS accumulation induced by bortezomib. In addition, mitochondrial gene expression serves as a source of drug resistance under apoptotic or stressed conditions. Therefore, studying mitochondrial activity and genes in fresh MM specimens might be a convenient way to predict resistance to proapoptotic chemotherapies, with important implications for clinical decision-making.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors thank Sang Won Kang for adenoviral particle to membrane-targeted catalase.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
ijc28149-sup-0001-suppinfo01.pdf636KSupplementary Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.