Article Cite This: J. Proteome Res. 2019, 18, 2643−2653
pubs.acs.org/jpr
Integrative Omics Analysis Revealed that Metabolic Intervention Combined with Metronomic Chemotherapy Selectively Kills Cancer Cells Chang Shao,† Wenjie Lu,† Ning Wan,† Mengqiu Wu,§ Qiuyu Bao,† Yang Tian,† Gaoyuan Lu,† Nian Wang,† Haiping Hao,*,†,‡ and Hui Ye*,†
Downloaded via NOTTINGHAM TRENT UNIV on August 14, 2019 at 14:58:20 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Key Laboratory of Drug Metabolism and Pharmacokinetics, State Key Laboratory of Natural Medicines, ‡School of Pharmacy, China Pharmaceutical University, Tongjiaxiang No. 24, Nanjing 210009, China § Department of Nephrology, Children’s Hospital of Nanjing Medical University, Guangzhou Road No. 72, Nanjing 210008, China S Supporting Information *
ABSTRACT: Metronomic chemotherapy, a relatively new dosing paradigm for anticancer therapy, is an alternative to traditional chemotherapy that uses maximal tolerated dose (MTD). Although these two dosing regimens both lead to tumor cell death, how cell metabolism is differentially affected during apoptosis remains elusive. Herein, we employed metabolomics to monitor the metabolic profiles of MCF-7 cells in response to the two dosing regimens that mimic MTD and MN treatments using a model chemotherapeutic drug, doxorubicin (Dox), and correlated the changes of metabolic genes examined by PCR array to integratively describe the reprogrammed metabolic patterns. We found glycolysis, amino acid, and nucleotide synthesis-associated metabolic pathways were activated in response to the MN treatment, whereas these pathways were inhibited in a pronounced way in response to the MTD treatment. Direct supplementation of key metabolites and pharmacological modulation of targeted metabolic enzymes can both regulate cell fates. Subsequently, we tested the combined use of MN dosing with targeted metabolic intervention using a normal cell line and found the combined treatment hardly affected its apoptotic rate. Our in vitro findings using MCF-7 and MCF-10A cells thus suggest the promising perspective of combining MN dosing of chemotherapeutic agents with metabolic modulation to selectively kill cancer cells rather than normal cells. KEYWORDS: metronomic chemotherapy, metabolomics, integrative omics, cancer metabolism
■
INTRODUCTION Standard chemotherapy administered chemotherapeutic agents at or close to the maximal tolerated dose (MTD), usually every 2−3 weeks, allowing patients to recover from toxicity.1 Although chemotherapy effectively induces apoptosis of tumor cells, tumor cells can re-initiate growth and proliferation during the intermission of chemotherapy.2 Meanwhile, MTD chemotherapy induces severe adverse effects, such as myelosuppression3 and damage to normal tissue. Besides MTD chemotherapy, metronomic (MN) chemotherapy is an alternative chemotherapeutic strategy.4 MN chemotherapy refers to the frequent, regular administration of chemotherapy agents at low doses. The dose must reach above the minimally effective concentration of chemotherapeutic drugs for a prolonged period without extended drug-free breaks.5 Increasing evidence support that, in comparison to the MTD chemotherapy, MN chemotherapy brings reduced toxicity and improved efficacy. For example, MN chemotherapeutic prevented the MTD-induced stromal activation and enhanced therapeutic response in desmoplastic breast cancer.6 Moreover, a phase III CAIRO3 trial that evaluated the overall survival of 278 patients showed that continuous MN © 2019 American Chemical Society
treatment of orally administered capecitabine with bevacizumab as a maintenance treatment brought reduced toxicity compared to the capecitabine treatment alone.7,8 Besides anticancer treatment with conventional chemotherapeutic agents, cancer metabolism can be alternatively exploited as targetable vulnerabilities.9 It is based on previous knowledge that, different from normal cells, cancer cells are reprogrammed to increase glucose and glutamine uptake and metabolize glucose for anabolic processes rather ATP production.10−12 In the past decade, this field has embraced a renaissance with a myriad of studies bringing to light the contribution of different metabolic pathways, including glycolysis, the TCA cycle, PPP, glutamine carboxylation, and fatty acid oxidation to cancer cell function, survival, and growth.13 Pharmacological control of these pathways thus presents a precise and effective therapeutic strategy for selective killing of cancer cells with less damage brought to normal cells. For instance, CB-839, an inhibitor of glutaminase (GLS), is Received: February 26, 2019 Published: May 16, 2019 2643
DOI: 10.1021/acs.jproteome.9b00138 J. Proteome Res. 2019, 18, 2643−2653
Article
Journal of Proteome Research
data provided comprehensive information for understanding metabolic responses of tumor cells to MTD therapy and MN therapy. Moreover, it sheds light on the important contribution of metabolism in affecting cell apoptosis and suggests that disturbing cancer cell metabolism with MN chemotherapy may be a useful strategy by which to selectively induce apoptosis in cancer cells while reducing severe side effects to benefit patients.
currently in a phase I clinical trial to treat a broad range of cancers that are sensitive to GLS inhibition.14 TVB-2640, a selective inhibitor of fatty acid synthase (FASN), is in a phase I clinical trial to treat colon cancer and in a phase II clinical trial to treat breast cancer and astrocytoma.15 Enasidenib, a first-inclass inhibitor of mutated isocitrate dehydrogenase 2 (IDH2), was recently approved by the U.S. Food and Drug Administration to treat acute myeloid leukemia (AML).16 Due to the advantages offered by metabolic regulation for anticancer therapy, drugs that target cancer metabolism are also used in combination with chemotherapeutic agents to increase chemo-sensitivity and reduce drug resistance. For example, a chemical analog and an anti-metabolite of folate (Pemetrexed), when used combinatorially with cisplatin, showed superior outcomes compared to standard chemotherapy for the treatment of malignant pleural mesothelioma (MPM) and non-small-cell lung cancer (NSCLC).17,18 Ritonavir, one of the glucose transporter 1 (GLUT1) inhibitors, increased the sensitivity of doxorubicin in multiple myeloma cell lines.19 Additionally, the inhibitor of IDH1, GSK-864, when used in combination with a receptor tyrosine kinase inhibitor (RTKi), Erlotinib, augmented RTKi susceptibility and induced cancer cell apoptosis.20 Recently, mannose was found to reduce tumor growth in tumor-bearing mice by interfering with glucose metabolism and inhibiting enzymes involved in glycolysis without affecting the health of animals. Mannose was also shown to promote apoptosis of cancer cells in response to chemotherapy agents such as cisplatin and Dox in both in vitro and in vivo models.21 Nevertheless, the possibility of combining metabolic intervention with chemotherapeutic agents administered by a MN dosing regimen for anti-tumoral treatment has remained in dark. This combination strategy might bring new hope and clinical perspectives for selectively killing of cancer cells with minimal toxicity brought to normal cells. Therefore, we herein examined the combined use of MN chemotherapy with metabolic inhibitors to treat MCF-7 cells, a breast cancer cell line, and compared the apoptotic rate of MCF-7 cells induced by the combination therapy with that of normal cells, MCF-10A cells, under the same treatment. A prolonged, low-dose treatment of Dox was first employed to mimic the MN treatment, and a transient, high-dose treatment was used to mimic the conventional MTD treatment as we previously described.22 We first employed metabolomics and PCR array as tools to analyze the metabolic changes of MCF-7 cells after the MN treatment of Dox and compared it with the changes induced by Dox administrated with a MTD regimen. Both metabolomics and mRNA data showed overall increases in mRNA and metabolites involved in glycolysis, amino acid, and nucleotide synthesis associated metabolic pathways in cells under the MN treatment, whereas the significant downregulation of these pathways were noted in cells under the MTD treatment. We thus supplemented the MTD-treated cells with metabolites that were markedly decreased and found it reduced the apoptotic rate. In line with the regulatory role of metabolite supplementation, pharmacological inhibition of the activated metabolic pathway in the MN therapy-treated cells was found to exacerbate apoptosis. Both findings indicated that modulating metabolism can affect therapeutic outcomes of chemotherapy in cancer cells. Lastly, we tested the combined treatment in MCF-10A cells and found that such treatment hardly influenced apoptotic rate of this normal cell line induced by the MN treatment alone. Taken together, these
■
EXPERIMENTAL SECTION
Cell Culture
Human breast cancer cells MCF-7 and human mammary epithelial cells MCF-10A were purchased from the American Type Culture Collection (ATCC). MCF-7 cells were cultured in RMPI 1640 medium supplemented with 10% (v/v) fetal bovine serum and 100 U/mL penicillin and streptomycin (Invitrogen, Carlsbad, CA) at 37 °C with 5% CO2. MCF-10A cells were cultured in Dulbecco’s modified Eagle medium supplemented with 10% (v/v) fetal bovine serum and 100 U/ mL penicillin and streptomycin (Invitrogen) at 37 °C with 5% CO2. Cell culture media were changed every other day. TUNEL Assay
The apoptotic rates of MCF-7 cells from the control and Doxtreated groups were determined by FragEL DNA Fragmentation Detection Kit (Calbiochem). Cells were stained according to the manufacturer’s instruction. Briefly, after the Dox treatment, cells were fixed and rehydrated. Subsequently, the 3′-OH ends of DNA fragments generated in apoptotic cells were labeled with fluorescein-conjugated deoxynucleotides in the presence of terminal deoxynucleotidyl transferase (TdT). Labeled cells were analyzed by flow cytometer (Accuri C6, BD Biosciences, San Jose, CA) at 488 nm. Quantitative Real-time PCR Assay and PCR Array Assay
After the Dox treatment, MCF-7 cells were suspended in RNAiso Plus reagent (Takara Bio) to extract the total RNA. Total RNAs was extracted followed by measurement of concentrations and purities. The RNA was then reversely transcripted into cDNA with a PrimeScript RT Reagent Kit (Takara Bio). The resulting cDNA was diluted in DEPCtreated water, mixed with the PowerUp SYBR Green Master Mix (Applied Biosystems, CA), and loaded in a 96-well PCR Array plate. The quantitative real-time PCR (qPCR) was performed using the StepOnePlus Real-Time PCR system (Applied Biosystems). The thermal cycling conditions of qPCR included denaturation at 95 °C for 2 min, followed by 40 cycles of 95 °C for 15 s, annealing at 60 °C for 15 s, and extension at 72 °C for 1 min. Endogenous β-actin gene was used as internal reference to normalize the expression levels of target genes by correcting variations in cDNA content loaded into qPCR reaction wells. The sequences of the primers used in this study are summarized in Table S1. Sample Preparation for Metabolomic Analysis
Sample preparation for metabolomics was performed as previously described.23 Briefly, MCF-7 cells were harvested and permeabilized with cold methanol: water solution (80:20, v/v) followed by a 20 min incubation at −80 °C. A stock solution of 4-chloro-phenylalanine (Sigma-Aldrich, Saint Louis, MO) was pre-spiked into the methanol aqueous buffer to a final concentration of 1.5 μg/mL, which subsequently served as an internal standard (IS). The cell lysates were centrifuged at 14000g for 5 min at 4 °C, and the resultant 2644
DOI: 10.1021/acs.jproteome.9b00138 J. Proteome Res. 2019, 18, 2643−2653
Article
Journal of Proteome Research
Figure 1. Illustration of the established dosing models that simulated maximal tolerated dose (MTD) and metronomic (MN) dosing with Doxorubicin (Dox) using MCF-7 Cells. (A) Cells were continuously incubated with 0.1 μM Dox for 96 h as a model of MN dosing. (B) Cells were incubated with 10 μM Dox for 4 h, followed by culture in drug-free medium for up to 48 h, as a model of MTD treatment. (C) Cells treated with 0.1 μM Dox were subjected to TUNEL assay for apoptosis analysis. (D) Cells treated with 10 μM Dox were subjected to TUNEL assay for apoptosis detection. Different dosing regimens with 0.1 μM Dox (E) and 10 μM Dox (F) transactivated the p53 downstream apoptosis-related gene, BAX, to similar levels. The expression levels of BAX mRNA were normalized to β-actin. Data are derived from three independent experiment and represented as mean ± SD (n = 3).
used for peak picking and alignment to screen the metabolic biomarkers that displayed significant level changes between the control and the Dox-treated group. Peak areas of these metabolic biomarkers were further integrated and calculated by MultiQuant 2.0 (AB SCIEX). Molecular identification of the assigned biomarkers was accomplished by matching the acquired precursors and fragment ions against several standard metabolome database including the Human Metabolome Database (http://www.hmdb.ca/), MassBank (http://www. massbank.jp/index.html), and METLIN (http://metlin. scripps.edu/index.php) (summarized in Figures S2 and S3). A mass error of 10 ppm was allowed for precursor ions matching and 40 ppm for fragment ion matching. Partial metabolite identification was further confirmed by comparison with available standards (Figure S4). Metabolic pathway enrichment analysis of these identified metabolic biomarkers was carried out by pathway enrichment analysis (http://www. metaboanalyst.ca/faces/ModuleView.xhtml).
supernatant was collected and evaporated to dryness. The samples were reconstituted in 100 μL of water (LC-MS grade, Merck) before analysis on a high-performance liquid chromatography−tandem mass spectrometry (HPLC−MS/ MS) instrument, and 20 μL was used for analysis. For each group of samples subjected to metabolomics analysis, five biological replicates were prepared. High-performance Liquid Chromatography and Mass Spectrometry
The LC-30A Shimadzu LC system (Shimadzu) was coupled to a TripleTOF 5600 system (AB SCIEX, Framingham, MA) for metabolite separation and detection. An XBridge BEH Amide HPLC column (100 mm × 4.6 mm, i.d. 3.5 μm) (Waters, Milford, MA) was used for compound separation at 40 °C. The mobile phase consisted of solvent A (95% 5 mM ammonium acetate buffer, pH adjusted to 9, 5% acetonitrile) and solvent B (acetonitrile). The gradient was set as follows: 0−3 min, 85% B; 3−6 min, 85−30% B; 6−15 min, 30−2% B; 15−18 min, 2% B; 18−19 min, 2−85% B; and 19−26 min, 85% B. The flow rate was set at 0.4 mL/min. The MS data was acquired in the negative ion mode using a data-dependent acquisition approach. The electrospray ionization (ESI) source conditions were used as follows: time-offlight MS scan, m/z 50−1000 Da; product ion scan, m/z 50− 900 Da; ion source gas 1 (gas 1), 50 psi; ion source gas 2 (gas 2), 30 psi; curtain gas, 30 psi; source temperature, 500 °C; ion spray voltage floating, −4500 V; declustering potential (DP), −100 V; collision energy (CE), −35 V; and CE spread, 10 V. MS data acquisition was controlled by Analyst TF 1.6.1 (SCIEX). The accurate mass was calibrated by Calibration Delivery System (SCIEX). Automatic calibration was performed between acquisitions of every five samples.
Cell Viability Assay
Cell viability assay was measured by using a cell-counting kit (CCK-8, Shanghai Yeasen Biotechnology, Shanghai, China). For the CCK-8 assay, cells were seeded into a 96-well plate and exposed to different concentration of Dox, as indicated at 37 °C with 5% CO2. After the Dox treatment, 10 μL of CCK-8 solution was added to each well and incubated for 1 h at 37 °C. The absorbance of each sample was then measured at 450 nm using Synergy TM2 (Bio-Tek Instruments Inc.).
■
RESULTS AND DISCUSSION
Establishment of MTD and Metronomic Dosing Models with Doxorubicin
Dox, one of the most common chemotherapeutic drugs for anticancer treatment, can induce cell apoptosis by triggering DNA damage.24 Here, a model to mimic standard MTD regimen was utilized as previously reported by exposing the model cell lines, MCF-7 cells, to 10 μM Dox (high dose, HD) for 4 h followed by subsequent withdrawal of the drug. Meanwhile, MCF-7 cells continuously exposed to a low dosage
Metabolomics Data Analysis
After normalization of the peak areas of each compound to the total ion chromatogram (Figure S1), principal component analysis (PCA) and partial least-squares discrimination analysis (PLS-DA) were performed with EZinfo 3.0 software (Waters). The Progenesis QI (Nonlinear Dynamics, Newcastle, UK) was 2645
DOI: 10.1021/acs.jproteome.9b00138 J. Proteome Res. 2019, 18, 2643−2653
Article
Journal of Proteome Research
Figure 2. Apoptotic cells induced by different dosing regimens of Dox showed distinct metabolomics profiles. (A) PLS-DA scoring plot of metabolic profiles differentiated 0.1 μM Dox-treated cells (MN dosing model) collected at 0 h, 24 h, 48 h, 72 and 96 h. (B) PLS-DA scoring plot of metabolic profiles differentiated 10 μM Dox-treated cells (MTD dosing model) collected at 0 h, 12 h, 24 h, 36 and 48 h. PCA scoring plot (C) and PLS-DA scoring plot (D) of metabolic profiles differentiated the cells treat with 0.1 μM Dox and 10 μM Dox from those of the control groups collected at 96 and 48 h, respectively. Metabolite pathway enrichment analysis of the metabolic biomarkers of the MN dosing model (E) and the MTD treatment (F). Heat maps illustrate the time-dependent changes of abundance levels of intermediate metabolites involved in amino acid metabolism pathway (G), purine and pyrimidine biosynthesis (H), glycolysis and tricarboxylic acid (TCA) cycle (I), and antioxidant defense (J) in response to the MN and MTD dosing regimens. Δrepresents metabolite with no significant changes in abundance. They are included for visual comparison with the other group.
(LD) of Dox at 0.1 μM for 96 h was used to mimic the MN chemotherapy (Figure 1A-B). Both dosing regimens induced MCF-7 cells to apoptosis at an approximately equal rate, 45.7% and 48.3%, respectively (Figure 1C−D). Because Dox is known to induce cytotoxicity in a p53-dependent manner,22,25,26 we examined the transcription level of a p53 target gene involved in apoptosis, Bcl-2-associated X protein (BAX). A significant increase of BAX to ∼4 fold compared to the basal level was observed after the Dox treatment administered according to the two dosing regimens. These results combinatorially demonstrated that the prolonged, LD (0.1 μM) treatment of Dox, which is designed to simulate the MN chemotherapy, managed to induce cell apoptosis of a similar rate to the transient, HD (10 μM) treatment that mimicked the standard MTD chemotherapy.
metabolic patterns in response to the MTD and MN treatment of chemotherapeutic drugs differed. Based on the established models of MTD and MN treatment using Dox, we compared the metabolomics profiles of cells under these two dosing regimens (collected at different time points) with the control group, respectively. As shown in Figure 2A,B, we performed PLS-DA analysis for these samples. The score plot showed a time-dependent shift from the cells under the MN treatment collected at 24, 48, 72, and 96 h post-dosing. Meanwhile, a time-dependent shift was also observed for the cells under the MTD treatment collected at 12, 24, 36, and 48 h post-dosing. Interestingly, the two apoptotic models that mimicked MTD and MN treatment were well-separated in the PCA (Figure 2C) and PLS-DA score plots (Figure 2D) when they reached an equal apoptotic rate, indicating distinct metabolic patterns were induced by the MTD and MN dosing regimen of Dox. The metabolite biomarkers that exhibited significant changes (fold change of >1.5 or
