Abstract
Leukemia cells in the bone marrow (BM) must meet the biochemical demands of increased cell proliferation and also survive by continually adapting to fluctuations in nutrient and oxygen availability. Thus, targeting metabolic abnormalities in leukemia cells located in the BM is a novel therapeutic approach. In the present study, we investigated the metabolic role of BM adipocytes in supporting the growth of leukemic blasts. Prevention of nutrient starvation-induced apoptosis of leukemic cells by BM adipocytes, as well as the metabolic and molecular mechanisms involved in this process, were investigated using various analytical techniques. In acute monocytic leukemia (AMoL) cells, the prevention of spontaneous apoptosis by BM adipocytes was associated with an increase in fatty acid β-oxidation (FAO) along with the upregulation of PPARγ, FABP4, CD36, and BCL2 genes. In AMoL cells, BM adipocyte co-culture increased adiponectin receptor gene expression and its downstream target stress response kinase AMPK, p38 MAPK with autophagy activation, and upregulated antiapoptotic chaperone heat shock proteins. Inhibition of FAO disrupted metabolic homeostasis, increased reactive oxygen species production, induced the integrated stress response mediator ATF4, and apoptosis in AMoL cells co-cultured with BM adipocytes. Our results suggest that BM adipocytes support AMoL cell survival by regulating their metabolic energy balance, and that the disruption of FAO in BM adipocytes may be an alternative, novel therapeutic strategy for AMoL therapy.
Keywords: bone marrow, microenvironment, adipocyte, acute monocytic leukemia, fatty acid β-oxidation
INTRODUCTION
Acute myeloid leukemia (AML) comprises a biologically-heterogeneous group of hematopoietic disorders, and it is primarily a disease of the elderly since 75% of AML patients that are > 60-years-old at diagnosis (1). Disease prognosis worsens with every decade of age past the age of 30 years primarily because of the decreased intensity of the chemotherapy administered to these patients (1). Consequently, there is an urgent need for novel therapeutic strategies that are not only effective, but can be well tolerated by elderly AML patients.
Research has long known of metabolic abnormalities in cancer cells, but metabolic modulation is now evolving as a novel therapeutic approach. Cancerous cells are constantly adjusting their metabolic state in response to extracellular signaling and/or nutrient availability (2), and recent evidence suggests that metabolic signals play a critical role in transcriptional regulation (3). Furthermore, metabolic enzymes are often present in transcriptional complexes and they provide a local supply of substrates/cofactors to these complexes (4). As such, although metabolic modulation is evolving as a viable AML therapeutic approach, the mechanisms underlying the environmental regulation of leukemia cell survival remain unclear (5).
The bone marrow (BM) microenvironment contributes to leukemogenesis and resistance to chemotherapy by increasing leukemic cell adhesion, providing growth factors, and promoting immunosuppression (6). Thus, the BM microenvironment represents a potentially attractive target for novel therapeutic interventions (6). Adipocytes, which arise from BM-resident mesenchymal stromal cells (MSC), are one of the major components of BM stroma (6). With advancing age, the number of adipocytes in the BM increases dramatically. Indeed, there is a shift in the differentiation of MSCs toward adipogenesis that is accompanied by a reduction of osteogenic and chondrogenic differentiation potential (7). In a 20-year-old, approximately 15% of the BM is comprised of adipocytes, which increases to approximately 60 % in a 65-year-old (8). This is in line with observations from mouse models of aging where increased adiposity is also associated with an upregulation of the key adipogenesis regulator PPARγ2 (peroxisome proliferator associated receptor γ2) (9).
Adipocytes reportedly support survival and growth of various types of tumor cells by stimulating mitochondrial metabolism that is associated with high-energy lipid transfer (10–12). A recent study demonstrated that co-culture of leukemia cells with adipocytes significantly impaired the antileukemia efficacy of chemotherapeutic agents, and that the rate of disease relapse after chemotherapy was higher in obese mice compared to their normal-weight counterparts (13). Although most tumor cells are programmed to increase glucose uptake, they generally reduce the proportion of glucose oxidized in the Krebs cycle (2,3) and become more dependent on fatty acid β-oxidation (FAO) (14). Indeed, we previously reported that BM-derived MSCs promote leukemic cell survival via a metabolic shift from pyruvate oxidation to FAO, which caused uncoupling of mitochondrial oxidative phosphorylation and regulated anti-apoptotic machinery in these cells (14). We also demonstrated that adipogenic differentiation of MSCs in the BM was associated with enhanced leukemia engraftment in a mouse leukemia model (15). These findings suggest that the increased adipocyte content of the aging BM may promote leukemogenesis and drug resistance.
A key factor implicated in adipocyte-tumor cell interaction is a lipid chaperone fatty acid binding protein 4 (FABP4), and its expression correlates with PPARγ activation under the transcriptional control by fatty acids, and it is predominantly expressed in adipocytes and macrophages(16). Adipocytes are known to be associated not only with metabolic contributions to tumor cells, as a source of energy but also with pro-inflammatory activation and differentiation induction in monocytes (17). Monocyte differentiation caused by inflammatory response has been shown to prevent apoptosis by induction of autophagy(18). A distinct monocytic subtype of AML, acute monocytic leukemia (AMoL), occurs predominantly in elder adults (i.e., ≥ 40 years old). The disease is associated with poor prognosis, and an average three-year overall survival rate of 31% (19). The majority of AMoL patients die of disease progression after relapse following therapy.
In the present study, we focused on this monocytic subtype of AML (AMoL) to investigate the genetic and metabolic networks in these cells that are regulated by BM adipocytes. We show that BM adipocytes activate a transcriptional network in AMoL cells, including PPARγ-FABP4 interactions and heat shock protein (HSP) upregulation, which are associated with FAO metabolism. We also provide evidence that the metabolic activity of AMoL cells is regulated by the energy sensor AMPK in coordination with p38, which mediates autophagy. We finally report that the inhibition of FAO disrupts the metabolic homeostasis, along with induction of integrated stress response (ISR) and apoptosis in AMoL cells, thus providing a novel therapeutic strategy for this disease.
MATERIALS AND METHODS
Cell lines, primary samples, and culture conditions
The AMoL cells U937, THP-1, and MV4;11 were purchased from ATCC, Manassas, VA. MOLM13 AMoL cells were purchased from, and acute myelomonocytic leukemia (AMMoL) cells, OCI-AML3, were purchased from DSMZ, Braunschweig, Germany. Cells were cultured in RPMI1640 medium containing 10% heat-inactivated FBS, 1% L-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin at 37°C in a 5% CO2 atmosphere. Primary AMoL cells from BM samples of five AMoL patients (Supplementary Table S1) were obtained after informed consent in accordance with Institutional guidelines set forth by Aichi Medical University per Declaration of Helsinki principles. MSCs were obtained from healthy donor BM (20− to 54− year-old) after informed consent in accordance with Institutional guidelines set forth by the University of Texas MD Anderson Cancer Center per Declaration of Helsinki principles. Both protocols were approved by the respective Institutional ethics committees.
MSCs were cultured at a density of 5,000 to 6,000 cells/cm2 in minimum essential medium alpha supplemented with 20% FBS, 1% L-glutamine, and 1% penicillin-streptomycin. Cultured MSCs isolated at passage 2 or 3 comprised a single phenotypic population, as previously described (12). Passage two MSCs that reached 90% confluence were allowed to differentiate to adipocytes which were identified by morphology and by the presence of lipid droplets that stained with oil red O (12). In this study, we utilized the monolayer culture composed of more than 50 % of BM adipocytes with lipid droplets obtained by this in vitro differentiation method.
shRNA-mediated stable knockdown of target genes was performed by lentiviral infection. shRNA targeting isoform α1 of the AMPK catalytic subunit (clone TRCN0000000859, targeting residues 1621–1641 of RefSeq NM_006251.5) was obtained from GE Healthcare Biosciences (Pittsburgh, PA). Transduced cells were selected by puromycin treatment, and knockdown was confirmed by immunoblot analysis of target proteins.
For co-culture experiments, leukemia cell lines and primary AMoL cells were cultured, at a density of 5 × 105/mL or 1 × 106/mL, respectively, with or without MSCs or BM adipocytes in serum-free conditions. Leukemia cells were co-cultured by plating them on top of MSCs or BM adipocytes, and were separated from the MSCs or BM-adipocyte monolayer by careful pipetting with ice-cold PBS, repeated twice. The purity of the leukemic cells separated from MSCs or adipocytes was confirmed by the absence of CD90 mRNA expression as determined by PCR.
Analyses of cell viability, apoptosis, cell-cycle, FAO and ROS production
Cell viability and proliferation were assessed by using a Vi-Cell XR (Beckman-Coulter, Brea, CA) cell counter using the Trypan blue exclusion method, or the CellTiter 96 AQueous One Solution Cell Proliferation Assay (MTS, Promega, Madison, WI) according to the manufacturer’s protocols. Apoptotic cell death was analyzed by annexin V staining as described elsewhere (20). Annexin V fluorescence was determined by a FACScan flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, CA). The flow cytometric data were analyzed by Cell Quest software (Becton Dickinson). The extent of drug-specific apoptosis was assessed by the formula: % specific apoptosis = (test – control) x 100 / (100 – control). Cell cycle distribution was analyzed by flow cytometric analysis of PI-stained nuclei as previously described (20). DNA content was determined by a FACScan flow cytometer system and CellQuest acquisition and analysis programs. Gating was set to exclude cell debris, cell doublets, and cell clumps.
Analysis of FAO was performed using the fatty acid oxidation human flow cytometry kit (Abcam, Cambridge, UK). Levels of the FAO cycle enzyme 3-hydroxyacyl-CoA dehydrogenase (HADHA) were determined by a FACScan flow cytometry and CellQuest acquisition and analysis programs. Fatty acid uptake was determined by a fluorometric fatty acid uptake kit (Abcam) according to the manufacturers’ protocols.
Analysis of ROS production was performed using CellROX deep red flow cytometry assay kit (Life Technologies, Carlsbad, CA). Flow cytometric data was acquired using a Canto II flow cytometer (BD Biosciences) and analyzed using FlowJo Version 9.5 software (TreeStar, Ashland, OR).
Fatty acid, ketone body and adiponectin measurement
Free fatty acid (FFA) levels in each conditioned medium in the co-culture experiments were measured by an enzymatic colorimetric method (NEFA-SS EIKEN kit; Eiken Chemical, Tokyo, Japan), and ketone bodies were measured by an enzymatic cycling method (Kainos ketone kit; Kainos Laboratories, Tokyo, Japan). These measurements were performed by SRL (Tokyo, Japan), an independent testing laboratory. Concentrations of adiponectin were determined with adiponectin human ELISA kit (Abcam) according to the manufacturers’ protocols.
mRNA quantification by real-time reverse-transcriptase PCR
Total RNAs were extracted from cells with the RNeasy Mini Kit (Qiagen, Hilden, Germany). First-strand cDNA was synthesized with oligo(dT) as primer (Superscript II System; Invitrogen, Carlsbad, CA). Real-time reverse-transcriptase PCR (RT-PCR) was performed by the Model 7500 Real-time PCR System (Applied Biosystems, Foster City, CA). Expression of PPARγ, CD36, FABP4 (fatty acid binding protein 4), HP, CFD, BCL2, CPT-1, ADIPOR1 (adiponectin receptor 1), ADIPOQ, and GAPDH mRNAs was detected by TaqMan Gene Expression Assays (PPARγ: Hs00234592_m1, CD36: Hs00169627_m1, FABP4: Hs01086177_m1, HP: Hs00978277_m1, CFD: Hs00157263_m1, BCL2: Hs00608023_m1, CPT1: Hs00912671_m1, ADIPOR1: Hs01114951_m1, ADIPOQ: Hs00605917_m1, GAPDH: Hs99999905_m1; Applied Biosystems). The PCR cycle number that generated the first fluorescence signal above a threshold value (the threshold cycle; Ct) was determined. The abundance of each gene transcript relative to that of GAPDH was calculated as follows: relative expression=100 × 2 exp [−ΔCt], where ΔCt is the mean Ct of the transcript of interest minus the mean Ct of the transcript for GAPDH. The Ct data from duplicate PCRs were averaged for calculation of relative expression.
Microarray analysis
The GeneSQUARE Multiple Assay DNA Microarray system (Kurabo Industries, Okayama, Japan) used for the gene expression analysis. The intensity of expression of each gene was normalized to that of GAPDH gene expression. Two independent experiments were carried out.
RNA-seq transcriptome analysis
Poly-A+ RNA was prepared using oligo(dT) magnetic beads. The library for mRNA-seq was prepared using the TruSeq RNA Sample Prep Kit v2 (Illumina, San Diego, CA) according to the manufacturer’s instructions. After the quality of the library was validated using the Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA), the samples were subjected to sequencing with a next-generation sequencer (GAIIx, Illumina). Gene expression was mapped and quantified by Tophat and Cufflinks (21). Functional networks and pathways were generated by Ingenuity Pathways Analysis (IPA; Ingenuity Systems, www.ingenuity.com).
Cap analysis gene expression
The cap analysis gene expression (CAGE) protocol was described previously, and individual tag libraries were normalized by using the common power-law distribution approach (22). Expression data for annotated coding or noncoding genes (according to GencodeV10) were extracted by collecting normalized tag counts in regions −500 to +200 relative to all annotated transcription start sites associated with a single gene ID. Normalized data were subjected to digital gene expression analysis by edgeR, then functional networks and pathways analyses were generated through the use of IPA and REVIGO analysis (23).
iTRAQ sample labeling, mass spectrometry analysis, and peptide identification
Isobaric tags for relative and absolute quantification (iTRAQ), a chemical labeling mass spectrometry (MS) method, was performed according to the manufacturer’s protocol (AB SCIEX, Framingham, MA). Proteins were identified and quantified (relatively) by using ProteinPilot Software Version 5.0 (AB SCIEX). Function definitions of the variable protein contents were searched against the UniProt database (1/20/2016 release). Protein ratios were normalized by using the overall median ratio for all the peptides in the sample for each separate ratio in every individual experiment. A confidence cutoff of > 95% for protein identification was applied. Proteins whose expression was statistically significantly changed by treatment were subjected to functional analysis by KEGG pathway enrichment analysis performed by using DAVID Bioinformatics Resources or IPA.
Immunoblot analysis
Immunoblot analysis was performed as previously described (20). For immunoblotting, the following antibodies were used: α-tubulin (Sigma-Aldrich), LC-3 (MBL, Nagoya, Japan), HSP70, phosphorylated- (p-)4E-BP1 Thr37/Thr46, 4E-BP1, p-Akt Ser473, Akt, p-AMPKα Thr172, AMPKα, p-p38MAPK Thr180/Tyr182, p38MAPK, ATF4, and horseradish peroxidase–linked anti-mouse and anti-rabbit IgG (all from Cell Signaling Technology).
Metabolite measurements
Metabolic methanol extracts, spiked with an internal standard solution (Human Metabolome Technologies, Inc., Tsuruoka, Japan), were prepared from 2 to 5 × 106 cells and analyzed using a capillary electrophoresis (CE)-connected ESI-time of flight (TOF)-MS and CE-MS/MS system (CARCINOSCOPE: Human Metabolome Technologies, Inc,) according to the manufacturer’s instructions. Cationic compounds were analyzed in the positive mode of CE-TOF-MS, and anionic compounds were analyzed in the positive and negative modes of CE-MS/MS according to the methods developed by Soga et al. (24). To obtain peak information, including m/z, migration time, and peak area, peaks detected by CE-TOF-MS and CE-MS/MS were extracted by automatic integration software (MasterHands, Keio University, Tsuruoka, Japan and MassHunter Quantitative Analysis B.04.00, Agilent Technologies, respectively). The peaks were annotated with putative metabolites from the Human Metabolome Technologies, Inc., metabolite database based on their migration times in CE, and m/z values determined by TOF-MS. The tolerance range for the peak annotation was configured at ±0.5 min for migration time and ±10 ppm for m/z. Concentrations of metabolites were calculated by normalizing the peak area of each metabolite with respect to the area of the internal standard and using standard curves, which were obtained by three-point calibration.
Statistical analyses
Differences between groups were assessed by a two-tailed Student t-test or a Wilcoxon matched pair test. A p-value ≤ 0.05 was considered statistically significant. Where indicated, the results are expressed as the mean ± SD of triplicate samples.
RESULTS
Antiapoptotic effects of BM adipocytes on AMoL cells in co-culture
We first investigated if BM adipocytes have a protective effect on AMoL cells in co-culture. AMoL cells and MSCs or BM adipocytes were co-cultured in serum-starved conditions to mimic the nutrient-deprived leukemic BM microenvironment. BM adipocytes supported viability of U937, MOLM13, and MV4;11 AMoL cells and OCI-AML3 AMMoL cells in these conditions showed by low rates of annexin positivity as effectively, or more, as undifferentiated MSCs (Figure 1A). Similarly, co-culture with BM adipocytes protected primary AMoL cells (Supplementary Table S1) from spontaneous apoptosis to a greater extent than those co-cultured with MSCs as assessed by annexin V/PI assay (Figure 1B, Supplementary Figure S1A) and confirmed by viable cell counts using the Trypan blue exclusion method. (Supplementary Figure S1B).
Figure 1. BM adipocytes protect acute monocytic leukemia cells from apoptosis.
(A) U937, MOLM13, and MV4;11 AMoL cells and OCI-AML3 AMMoL cells were cultured 48 h in the presence or absence of MSCs or MSC-derived BM adipocytes under serum-starved conditions. Percentages of cell death were determined by cell counts using the Trypan blue exclusion method. Graphs show the mean ± SD of the results from three independent experiments. *p < 0.05; **p < 0.01. (B) Primary cells from five AMoL patients were cultured 48 h in the presence or absence of MSCs or MSC-derived BM adipocytes. Percentage of apoptotic cells (annexin V positive) was detected by flow cytometry.
BM-adipocyte co-culture induces metabolic alterations in AMoL cells
Next, we investigated whether fatty acid metabolism is involved in AMoL cell survival in our BM adipocyte co-culture conditions. The pharmacologic inhibition of FAO by etomoxir, which impairs fatty acid entry into the mitochondria by blocking carnitine palmitoyltransferase 1 (CPT1; the ratelimiting enzyme regulating mitochondrial import of fatty acid derivatives of the amino acid carnitine), abolished the antiapoptotic effects of BM adipocytes on U937 cells (p = 0.02; Figure 2A). Etomoxir did not induce significant cytotoxicity of BM adipocytes (Trypan blue (+) cell %; etomoxir (−) 6.7±3.4, etomoxir (+) 11.7±3.0, p = 0.13) as determined by viable cell counts using the Trypan blue exclusion method.
Figure 2. Co-culture with BM adipocytes and FAO inhibition induce metabolic alterations in U937 cells.
(A) U937 cells were treated with FAO inhibitor etomoxir (50 µM) for 48 h in the presence or absence of MSC-derived BM adipocytes under serum-starved conditions. Percentage of apoptotic cells (annexin V positive) was detected by flow cytometry. Graphs show the mean ± SD of results of three independent experiments. Cont, controls; *p < 0.05. (B) U937 cells were treated with FAO inhibitor etomoxir (50 µM) for 24 h in the presence or absence of MSC-derived BM adipocytes under serum-starved conditions. The representatives of histogram of CellROX in the viable cells of indicated conditions of three independent experiments are shown. Mean fluorescence intensity (MFI) are the mean ± SD of results of three independent experiments. (C) Quantified levels of metabolites in U937 cells co-cultured with BM adipocytes in the presence or absence of etomoxir were determined by CE-TOF-MS analysis. The quantification data are superimposed on a metabolic pathway map that includes the glycolysis and Krebs pathways. Results shown are representative of three independent CE-TOF-MS experiments. Bars, SD. All p-values were determined by the Wilcoxon matched pair test. *p < 0.05; **p < 0.01.
Stressors such as nutrient deprivation and ROS production are known to lead to progressive redox damage and ultimately cell death. FAO inhibition by etomoxir has been reported to induce ROS production (25). We therefore measured ROS generation in U937 cells cultured in the presence or absence of BM adipocytes and with and without etomoxir treatment. As shown in Figure 2B, co-culture with BM adipocytes decreased constitutive ROS production by the leukemia cells. Conversely, etomoxir treatment enhanced ROS production and increased the number of dead U937 cells when co-cultured with BM adipocytes. This was not observed in AML cells that were cultured alone, which was consistent with our previous observations (14). These results suggested that the cytotoxic effect of etomoxir on leukemia cells co-cultured with BM-adipocyte was, at least in part, dependent on enhanced ROS production and oxidative stress.
We next analyzed the concentrations of free fatty acids (FFA) and ketone bodies in the media in that U937 cells, BM adipocytes, or both were cultured for 24 hours, with or without etomoxir treatment (Table 1). Compared to the FFA concentration in control media (fresh media), the FFA concentration was significantly decreased in the media in which U937 cells that were cultured alone, and increased in media in which normal BM adipocytes were cultured and decreased in the media in which BM adipocytes were co-cultured with U937 cells for the same period. Etomoxir did not significantly change FFA levels in culture media of BM adipocytes, indicating that etomoxir did not affect FFA release from BM adipocytes.
Table 1.
BM-adipocyte co-culture increases production of free fatty acids and ketone bodies in U937 cells a
| cell type/culture condition | treatment | FFA (µM) |
acetoacetate (µM) |
3-hydroxybutyric acid (µM) |
total ketone body (µM) |
|
|---|---|---|---|---|---|---|
| U937 | ||||||
| culture alone | - | 5.0 ± 1.0 ** |
<2 | 3.3 ± 0.6 | 3.3 ± 0.6 | |
| culture alone co-culture with adipocyte |
etomoxir - |
5.7 ± 0.6** 8.3 ± 2.1 |
<2 6.7 ± 0.6 |
<2 8.0 ± 1.0 |
<2 14.7 ± 1.5 |
|
| co-culture with adipocyte | etomoxir | 9.1 ± 1.2 | <2 | 3.0 | 3.0 | |
| adipocyte | ||||||
| culture alone | - | 11.7 ± 1.5* | <2 | <2 | <2 | |
| culture alone | etomoxir | 10.6 ± 0.6* | <2 | <2 | <2 | |
| medium only (without FBS) |
8.7 ± 0.6 | <2 | <2 | <2 | ||
Cells were cultured in serum-free medium (without FBS) for 24 hours. Data shown are mean ± SD of three independent experiments.
FFA, free fatty acids
p <0.05,
p <0.01, compared to medium only
In contrast, ketone bodies that are produced by FAO, including acetoacetate and 3-hydroxybutyric acid, were higher in the media in which U937 and BM adipocytes had been co-cultured compared to the control media, and greatly reduced by etomoxir treatment. Levels of ketone bodies were not increased in media in which BM adipocytes were cultured alone. These results indicated that FAO in U937 cells and/or BM adipocytes is activated by co-culture.
To investigate the change of FAO levels in U937 cells after co-culture with BM adipocytes, we examined FAO cycle enzyme HADHA expression which is responsible for FAO upregulation (26). The significant increase of HADHA expression levels was detected in U937 cells co-cultured with BM adipocytes by flow cytometry (% positive cells; control 28.9±1.1, adipocyte co-culture 46.3±10.5, p = 0.04).
We next investigated metabolic changes in U937 cells co-cultured with BM adipocytes in the presence or absence of etomoxir. CE-TOF-MS cation and anion analyses identified 94 and 80 metabolites in two independent experiments (Supplementary Table S2). Concentrations of metabolites that are involved in glycolysis and Krebs cycle are illustrated on a metabolic pathway map in Figure 2C. U937 cells co-cultured with BM adipocytes showed higher levels of glucose 6-phosphate, fructose 6-phosphate, and pyruvate, the dominant product of aerobic glycolysis, than controls. Since aerobic glycolysis can support mitochondrial FAO (14), these data are consistent with the activation of anaplerotic pathways to maintain Krebs cycle efficiency by utilizing fatty acid–derived acetyl CoA (14).
Interestingly, ATP production was not increased in AMoL cells co-cultured with BM adipocytes. FAO is known to support oxygen consumption in leukemia cells, and this process can be uncoupled from oxidative phosphorylation translating into decreased efficiency of ATP synthesis (27). This inefficiency of FAO energy metabolism could explain our observation that BM adipocyte co-cultures upregulated FAO without increasing ATP generation in AMoL cells.
Notably, inhibition of FAO by etomoxir diminished these metabolites and markedly increased lactate, indicating a stress-related hyperlactataemia. Lactate accumulation can cause local tissue acidosis that potentially modulates the activity of proteases that decompose extracellular matrix, thereby liberating amino acids that are consumable for energy generation (28). Indeed, although etomoxir treatment strikingly depleted citrate, succinate, fumarate, and malate in the Krebs cycle, these losses were balanced by an increase of most amino acids, including glutamine. Monitoring of the cellular energy status by measuring ADP/ATP and AMP/ATP ratios showed that FAO inhibition by etomoxir increased both ADP/ATP (control 0.9, adipocyte co-culture 0.9, etomoxir 1.4) and AMP/ATP (control 0.4, adipocyte co-culture 0.4, etomoxir 1.1) via significant increases of ADP and AMP (p < 0.01) and decrease of ATP (p < 0.01) (Supplementary Table S2). These changes signify the reduction in cellular respiration associated with FAO inhibition by etomoxir under adipocyte co-culture condition (29).
Gene expression changes in U937 cells co-cultured with BM adipocytes
To investigate gene expression changes in AMoL cells potentially associated with FAO metabolic alterations, we performed microarray analysis of 339 genes related to lipid metabolism (GeneSQUARE) using RNA from U937 cells co-cultured with BM adipocytes or MSCs. Genes that changed frequently in the same direction (13 genes upregulated, four genes downregulated, by > 2-fold) in two experimental replicates are shown in Table 2. The genes upregulated by BM-adipocyte co-culture included nuclear receptor PPARγ-encoding PPARG and its target genes scavenger receptor CD36 and fatty acid carrier protein FABP4. These findings were confirmed by quantitative RT-PCR assay (Figure 3A). Moreover, the upregulation of cell surface CD36 expression by co-culture with BM adipocytes was detected in U937 cells and in four of the five primary AMoL samples by flow cytometry (Figure 3B). These data suggested that perhaps upregulation of scavenger receptor CD36 in AMoL cells by BM adipocyte co-culture stimulates FFA uptake, and thereby activation of nuclear receptor PPARγ, which, in turn, results in the transcriptional induction of CD36 and FABP4 in AMoL cells. We confirmed the FFA uptake after co-culture with BM adipocytes in U937 cells (Figure 3C). These findings were paralleled by increased levels of CPT-1 and antiapoptotic BCL2 mRNAs (Figure 3A).
Table 2.
Changes in gene expression in U937 AMoL cells by co-culture with BM adipocytes or MSCs a
| gene name | adipocyte co-culture vs control (fold change) |
MSC co-culture vs control (fold change) |
||
|---|---|---|---|---|
| microarray 1 | microarray 2 | microarray 1 | microarray 2 | |
| upregulated | ||||
| FABP4 | 207.03 | 243.24 | 1.52 | 2.17 |
| SLC17A3 | 10.92 | 2.54 | 2.79 | 1.94 |
| APOD | 6.52 | 2.69 | 1.52 | 0.68 |
| CCR2 | 4.65 | 3.67 | 1.93 | 1.77 |
| CRP | 3.93 | 6.69 | 1.26 | 1.97 |
| HP | 3.45 | 3.15 | 2.51 | 3.21 |
| HSD11B2 | 3.37 | 2.87 | 2.25 | 2.17 |
| CCND1 | 3.06 | 2.42 | 2.07 | 2.26 |
| CFD | 2.81 | 2.44 | 0.92 | 1.14 |
| CD36 | 2.77 | 2.41 | 1.29 | 1.49 |
| TCF7L1 | 2.46 | 2.45 | 0.37 | 1.35 |
| SPOCK3 | 2.28 | 2.66 | 0.93 | 2.37 |
| TEK | 2.05 | 2.15 | 0.93 | 2.00 |
| downregulated | ||||
| HK2 | 0.46 | 0.31 | 2.13 | 2.18 |
| AGTRL1 | 0.37 | 0.35 | 0.28 | 0.45 |
| GHRL | 0.23 | 0.20 | 0.48 | 0.67 |
| CCL2 | 0.06 | 0.05 | 2.28 | 2.69 |
Detected by microarray. Values shown represent gene expression intensity normalized to that of the GAPDH gene.
Figure 3. Co-culture with BM adipocytes induces gene and protein expression changes and AMPK activation in AMoL cells.
Cells were cultured with or without BM adipocytes for 24 h. (A) Quantitative RT-PCR analysis showing mRNA expression of PPARγ, CD36, FABP4, CPT-1, and BCL2 in U937 cells after 24 h co-culture with MSCs or MSC-derived BM adipocytes; Co, controls. The abundance of transcripts of each gene relative to the abundance of GAPDH transcripts was determined as described in Materials and Methods. Graphs show representative data from three independent experiments. (B) Cell surface CD36 expression was determined by flow cytometry in AMoL primary samples or U937 cells co-cultured with MSCs or MSC-derived BM adipocytes for 48 h. For U937 cells, graph shows the mean ± SD from three independent experiments. *p < 0.05. (C) Fatty acid uptake by U937 cells was evaluated under conditions of co-culture with or without BM adipocytes under serum-starved conditions for 16 h. Cells were plated at 50,000 cells / well, after which fatty acid mixture (dodecanoic acid fluorescent fatty acid substrate) was added and incubated for 1 h. Fluorescence signal was measured with a plate reader using bottom read mode. Graphs show the mean ± SD of the results from three independent experiments. **p < 0.01. (D) Expression levels of ATF4 protein in U937 cells after co-culture with BM adipocytes with or without etomoxir (EX) were detected by immunoblotting. α-tubulin was used as a loading control. The results shown are representative of three independent experiments. (E, F) Expression levels of HSP70, p38MAPK, p-p38MAPK, AMPK, p-AMPK, AKT, p-AKT, 4EBP1, p-4EBP1, and LC3 proteins in U937 cells after co-culture with BM adipocytes for 24 h with or without etomoxir (EX) were detected by immunoblotting; Cont, controls. Results shown are representative of three independent experiments. The intensity of the immunoblot signals compared to that of α-tubulin was quantified using Image J software. (G) Immunoblot analysis of total and phosphorylated AMPK in OCI-AML3 cells transfected with control short hairpin RNA (shRNA) or shRNA against AMPK, and effects of AMPK knockdown on cell survival under serum-starved conditions in the presence or absence of MSCs or MSC-derived BM adipocytes cultured 48 h. Percentages of cell death were determined by cell counts using the Trypan blue exclusion method. Graphs show the mean ± SD of the results from three independent experiments. *p < 0.05; **p < 0.01, shC; control shRNA, shAMPK; shRNA against AMPK. (H) Expression of ADIPOR1 mRNA in U937 cells co-cultured with MSCs or MSC-derived BM adipocytes for 24 h, with or without etomoxir (Ex) as indicated, shown by quantitative RT-PCR analysis. The abundance of transcripts of ADIPOR1 gene relative to the abundance of GAPDH transcripts was determined as described in Materials and Methods. Graphs show representative data from three independent experiments. (I) Expression of ADIPOQ mRNA in MSCs and MSC-derived BM adipocytes was analyzed by quantitative RT-PCR analysis. The abundance of transcripts of ADIPOQ gene relative to the abundance of GAPDH transcripts was determined. Graphs show representative data from three independent experiments. Culture supernatants were collected after 24 h incubation and analyzed for adiponectin concentration. Data sets are mean values ± SD of three independent experiments. *p < 0.05.
Transcriptional network in AMoL cells co-cultured with BM adipocytes
We performed RNA-seq and CAGE transcriptome analyses to investigate the transcriptional network(s) underlying the antiapoptotic effects induced by BM-adipocyte co-culture and the metabolic alterations induced by FAO inhibition in AMoL cells. CAGE and RNA-seq are complementary technologies that can be used to improve incomplete gene models, and CAGE was used for quantitative modeling of the transcriptional output of transcription start sites (TSSs) by identifying and quantifying the 5’ ends of capped mRNA transcripts (30). RNA-seq detected 20 genes upregulated and 23 genes downregulated by BM-adipocyte co-culture in U937 cells (false discovery rate < 0.05; Supplementary Table S3). Ten of the upregulated genes were significantly decreased by etomoxir treatment, including tyrosine kinase FLT3, chemokine CXCL12 receptor CXCR4, co-chaperone immunophilin protein FKBP5, PI3K negative regulator PIK3IP1, and immunosuppressive transcriptional regulator TSC22D3 (Table 3).
Table 3.
Genes whose expression was upregulated by adipocyte co-culture and downregulated by etomoxir treatment in U937 cells a
| gene name | adipocyte co-culture vs control |
etomoxir effect (under acipocyte co-culture) |
||
|---|---|---|---|---|
| fold change | p-value | fold change | p-value | |
| FLT3 | 2.08 | 0.0001 | −2.62 | 0.0001 |
| PIK3IP1 | 1.54 | 0.0001 | −2.30 | 0.0001 |
| KLF9 | 1.43 | 0.0001 | −2.51 | 0.0001 |
| TSC22D3 | 1.41 | 0.0001 | −1.38 | 0.0001 |
| PTAFR | 1.28 | 0.0001 | −1.10 | 0.0002 |
| FKBP5 | 1.27 | 0.0001 | −1.19 | 0.0001 |
| PER1 | 1.13 | 0.0001 | −1.61 | 0.0001 |
| P2RY2 | 0.99 | 0.0002 | −1.32 | 0.0001 |
| CXCR4 | 0.82 | 0.0002 | −0.85 | 0.0001 |
| PRTN3 | 0.80 | 0.0002 | −1.88 | 0.0001 |
Detected by RNA-Seq.
In U937 and THP1 cells, co-culture with BM adipocytes frequently altered 585 genes in a CAGE-mapped TSS signature: 366 upregulated and 219 downregulated (false discovery rate < 0.05; Supplementary Table S4). IPA characterized the upstream mediators, including transcription factors, associated with the alterations of promoter structures necessary for gene expression in CAGE data; co-culture with BM adipocytes activated the cancer-associated transcription factors MYC and FOXM1 and inhibited the p53 transcription regulator IFI16 and PI3K/AKT signaling positive regulator FLT1 kinase. Notably, IPA of CAGE data highlighted the etomoxir-induced activation of transcription factor ATF4, the master regulator of the integrated stress response (ISR), along with further upregulation of PPARγ (Table 4).
Table 4.
Upstream regulators commonly detected in U937 and THP1 cells under adipocyte co-culture and etomoxir treated conditions b
| Upstream regulators of adipocyte co-cultured cells | |||
|---|---|---|---|
| Upstream Regulator |
Activation z- score |
p- value |
|
| activated | |||
| MYC | 2.755 | <0.001 | |
| FOXM1 | 2.383 | 0.006 | |
| IL2 | 2.373 | 0.001 | |
| NLRP12 | 2.000 | 0.009 | |
| inhibited | |||
| FLT1 | −2.333 | <0.001 | |
| IFI16 | −2.219 | 0.029 | |
| TNFRSF8 | −2.000 | 0.049 | |
| Upstream regulators of etomoxir treated cells under adipocyte co-cultured condition | |||
| Upstream Regulator |
Activation z- score |
p- value |
|
| activated | |||
| IL5 | 2.486 | <0.001 | |
| MYCN | 2.345 | <0.001 | |
| ATF4 | 2.035 | <0.001 | |
| PPARG | 2.03 | <0.001 | |
| EPAS1 | 2 | 0.003 | |
Detected by CAGE and analyzed by IPA.
Induction of ATF4 was confirmed by immunoblot analysis (Figure 3D). The RNA-seq and CAGE analysis of AMoL cells detected other consistent alterations of gene expression by BM-adipocyte co-culture, including upregulation of KLF9, a transcription factor that activates PPARγ2 promoter, and FKBP5, a HSP90-interacting co-chaperone immunophilin protein. Etomoxir treatment in BM-adipocyte co-culture consistently upregulated ISR mediator ATF4 and its target genes TRIB3, ASNS, and lipid accumulation marker PLIN2 (Supplementary Tables S3, S4). Chemokine CXCL12 receptor CXCR4 was also upregulated by BM-adipocyte co-culture and reversed by etomoxir treatment, in accord with our previous observation (15). The calcium-sensitive serine-threonine phosphatase calcineurin B-encoding gene PPP3R1 (31) was depressed by BM adipocytes co-culture and that this effect was reversed by etomoxir, detected by CAGE (control 353,000 tags/million, adipocyte co-culture 139,000, etomoxir-treated adipocyte co-culture 229,000). Calcineurin is shown to play a highly conserved role in the adaptive regulation of energy metabolism via attenuation of AMPK signaling (32,33).To extrapolate the protein signaling changes induced in AMoL cells by BM-adipocyte co-culture, we performed REVIGO analysis (23) utilizing the CAGE-mapped TSS signature of genes frequently altered in U937 and THP1 cells (Supplementary Table S4). The “Tree Map” view of REVIGO highlighted the BM-adipocyte co-culture-induced activation of the p38MAPK cascade and protein folding in AMoL cells (Supplementary Figure S2).
Upregulation of AMPK signaling and HSP in AMoL cells by BM-adipocyte co-culture
In two independent experiments, iTRAQ proteomic analysis detected a total of 1609 and 940 proteins in U937 cells cultured in the absence or presence of adipocytes, respectively, and identified replicated significant expression changes of 13 proteins upregulated (> 1.2-fold) and 19 proteins downregulated (< 0.8-fold) compared to control conditions (p < 0.05; Supplementary Table S5). The IPA platform highlighted activation of HSP chaperone proteins (Supplementary Figure S3), which was confirmed by the immunoblot analysis that confirmed the upregulation of HSP70 expression in U937 cells co-cultured with BM adipocytes (Figure 3E). Immunoblotting further demonstrated that BM-adipocyte co-culture upregulated p-AMPK and p-p38 MAPK and downregulated Akt/mTOR signaling with depletion of p-Akt and p-4EBP1 in U937 cells and that all of these changes were reversed by etomoxir treatment (Figure 3F). AMPK is a crucial cellular energy sensor and metabolic modulator; regulating biosynthetic pathways and gene transcription involved in energy metabolism through mTOR inhibition, and is known to modulate cell functions by regulating autophagy (34).
BM-adipocyte co-culture induced autophagosomal marker LC3-II, suggesting the activation of AMPK-dependent cytoprotective autophagy (34) under starvation stress conditions (Figure 3F). Whereas co-culture with BM adipocytes protected cells from serum-starvation-induced cell death in control OCI-AML3 cells, the cells with stable knockdown of AMPK (Figure 3G) increased their sensitivity to spontaneous cell death in the presence of BM adipocytes. Notably, AMPK knockdown OCI-AML3cells were significantly less sensitive to nutrient starvation-induced cell death in the absence of BM adipocytes.
Because C receptor 1 (AdipoR1), a the major receptor for adiponectin, activates the AMPK pathways and plays an important role in the regulation of glucose and lipid metabolism and oxidative stress in vivo (35), we investigated the ADIPOR1 mRNA expression in U937 cells under BM-adipocyte co-culture conditions with or without etomoxir treatment. Quantitative RT-PCR detected, as expected, the upregulation of ADIPOR1 by co-culture with BM adipocytes and its suppression by etomoxir (Figure 3H). In turn, BM adipocytes expressed markedly high levels of the adiponectin-coding gene ADIPOQ and secreted significantly higher level of adiponectin compared to parental MSCs (p=0.02) (Figure 3I).
DISCUSSION
Our results show that BM adipocytes support survival of AMoL cells under nutrient starvation conditions by regulating metabolic energy balance, at least in part, through the stimulation of FAO and modulation of transcription factors. Our findings suggest that BM adipocytes supply fatty acids, the essential ligands of nuclear receptor PPARγ, which can be internalized in leukemic cells via CD36 and transferred to the nucleus by FABP4. Once activated, PPARγ then induced its downstream target genes, including CD36 and FABP4 and antiapoptotic BCL2. This hypothesis is supported by several reports addressing the functional relationship between adipocytes and tumor cells that are mediated by a FABP4-dependent mechanism(16).
AMPK modulates transcription of specific genes involved in energy metabolism, thereby exerting long-term metabolic control, including upregulation of fatty acid uptake and oxidation, as well as autophagy regulation (34,36). BM-adipocytes, the major source of serum adiponectin that increases during caloric restriction as well as during cancer therapy (37), has been shown to contribute to chemotherapy resistance via adipocyte-secreted adipokines and AMPK-dependent autophagy activation in myeloma cells (38). Our co-culture experiments conducted in serum-starved conditions also demonstrated adiponectin production by BM-adipocytes. Adipocyte-produced adiponectin induces extracellular calcium influx through AdipoR1 that is necessary for the subsequent activation of AMPK (32). In our study, the activation of AMPK in AMoL cells co-cultured with BM adipocytes was accompanied with the upregulation of the adiponectin receptor ADIPOR1 gene and downregulation of the calcium-dependent AMPK negative regulator calcineurin -encoding gene PPP3R1 (33). These findings suggest that adipocyte-secreted adiponectin causes activation of AdipoR1 and of the downstream AMPK pathway in AMoL cells.
Adiponectin reportedly induced activation of AMPK that causes apoptosis or cell cycle arrest of multiple myeloma cells cultured alone (38), which is in agreement with our results of AMPK knockdown experiments. We observed that AMPK-dependent stress response of AMoL cells in the presence of BM adipocytes was different from the one in the absence of adipocytes, and that AMPK activation supported AMoL cell survival during nutrient starvation in the presence of BM adipocytes. Whereas mTORC1 activation, which is negatively regulated by AMPK (39), has been thought to induce leukemogenesis (40), Hoshii et al. (41) have reported that a subset of AML cells with an undifferentiated phenotypes survived long-term in the absence of mTORC1 activity. In the BM microenvironment where mTORC1 is suppressed mTORC1 may play a role in the maintenance of leukemia-initiating cell homeostasis, and the role of AMPK-mTOR signaling could be distinct in differentiated versus undifferentiated AML cells. Wu et al. (42) reported that the adiponectin-AMPK signals downregulated Bcr-Abl expression and reversed imatinib resistance in chronic myeloid leukemia (CML) cells. The contradictory results between CML and AMoL could be explained by differences in their maturation stage (i.e., more differentiated CML cells should be more dependent on mTOR signaling for their survival than AMoL cells).
In contrast to findings in leukemia cells co-cultured with MSCs, in which p-Akt was upregulated (48), our results demonstrate that co-cultures with BM adipocytes induces downregulation of Akt phosphorylation along with inhibition of FLT1 kinase, the positive regulator of PI3K/AKT signaling, and upregulation of the PI3K negative regulator PIK3IP1. These results again suggest that distinct mechanism(s) are operational in BM adipocyte and MSCs co-culture systems, which are both involved in apoptosis resistance in leukemia cells.
Our observations also suggest that BM adipocytes induce a HSP response in co-cultured AMoL cells. HSP chaperone proteins that bind to denatured and unfolded proteins and promote protein refolding or degradation are known to be positively regulated by AMPK (34). Our proteomic and transcriptome analyses highlight the activation of HSP-associated transcription factors MYC and FOXM1, which both cooperate with HSPs to promote tumor progression (43,44), and the upregulation of co-chaperone immunophilin protein FKBP5 that interacts with HSP90. Thus, HSP activation is supporting AMoL cell survival under co-culture conditions.
We previously reported that pharmacological FAO inhibitors increased the amount of lactate generation by leukemia cells (14), which is likely an adaptive mechanism to maintain ATP production. In this study, we observed that FAO blockade by etomoxir abolished the antiapoptotic effects of BM adipocytes on AMoL cells with increased ROS accumulation and generation of lactate and amino acids. A possible source of increased amino acids after FAO inhibition might be the degradation of extracellular matrix, particularly by matrix metalloproteinases (45). For the better understanding of the interactions between AMoL cells and BM adipocytes, the detection of the metabolic changes occurring in BM adipocytes upon co-culture with AMoL cells, as well as the comparison of the metabolic effects on co-cultured AMoL cells between BM adipocytes generated from healthy and leukemic patients warrant further investigation.
FAO inhibition by etomoxir upregulated the transcription factor ATF4, the master mediator of the ISR, which is known to repress AMPK activation (46). Recently, involvement of p38 in signal switching from autophagy to apoptosis in AML via the ISR-related PERK/elF2a/ATF4 pathway has been reported (47). Validation of the specific role of ATF4 in AMoL cells will be conducted in a future study.
Overall, BM adipocytes activate a transcriptional network that includes PPARγ - FABP4 interaction and HSP chaperones associated with FAO metabolism in AMoL cells. The metabolic energy balance is regulated by the energy sensor AMPK coordinated with p38, which mediates autophagy (48), and is known to induce cell differentiation, growth inhibition, and cell survival (49) (Figure 4). In conclusion, the metabolic adaptation of AMoL cells and the modulation of fatty acid metabolism may represent a novel strategy for the treatment of hematological malignancies, especially in elderly patients, in whom the aging and increased adiposity BM microenvironment reduces the efficacy of cytotoxic chemotherapy.
Figure 4. BM adipocytes support AMoL antiapoptosis via FAO stimulation, with activation of AMPK and HSP chaperone proteins and modulation of transcription factors in vitro.
BM adipocytes induce upregulation of PPARγ, CD36, and FABP4 gene transcription, which stimulates fatty acid endocytosis. In mitochondria, fatty acids are consumed for FAO, which is accompanied by mitochondrial uncoupling, resulting in diminished formation of mitochondrial ROS and decrease of intracellular oxidative stress. The networks of transcriptional regulation and fatty acid metabolism support AMoL cells in a quiescent state associated with activation of AMPK, p38 with autophagy induction, upregulation of HSP antiapoptotic chaperone proteins and acquisition of chemotherapy resistance. FAO inhibition by etomoxir induces the integrated stress response, which stimulates transcriptional activation of ATF4, FABP4, fatty acid binding protein 4; AMPK, AMP-activated protein kinase; p38, p38 mitogen-activated protein kinase; ADIPOR1, adiponectin receptor 1; ATF4, activating transcription factor 4.
Supplementary Material
Acknowledgments
The authors wish to thank Hideki Hayashi and Masayuki Tanaka, Education and Research support center, Tokai University for statistical support and Yasuhito Hatanaka, Hikari Taka, Naoko Kaga, Takako Ikegami, Tomomi Ikeda, Akemi Kawasaki and Sujan Piya for technical assistance. We thank Research Support Centers of Molecular and Biochemical Research and Cell Biology Research and Division of Proteomics and BioMolecular Science, Juntendo University Graduate School of Medicine for use of facilities.
Financial support for this work: This work was supported in part by a Grant-in-Aid for Scientific Research (C), Japan, a Grant-in-Aid (S1311011) from the Foundation of Strategic Research Projects in Private Universities from the MEXT, Japan (to Y.T.), the National Institutes of Health (Cancer Center Support Grant CA016672), and the Paul and Mary Haas Chair in Genetics (to MA).
Footnotes
Conflict of Interest: The authors declare no competing financial interests.
REFERENCES
- 1.Hassan M, Abedi-Valugerdi M. Hematologic malignancies in elderly patients. Haematologica. 2014;99(7):1124–1127. doi: 10.3324/haematol.2014.107557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science. 2009;324(5930):1029–1033. doi: 10.1126/science.1160809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Jones RG, Thompson CB. Tumor suppressors and cell metabolism: a recipe for cancer growth. Genes Dev. 2009;23(5):537–548. doi: 10.1101/gad.1756509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gut P, Verdin E. The nexus of chromatin regulation and intermediary metabolism. Nature. 2013;502(7472):489–498. doi: 10.1038/nature12752. [DOI] [PubMed] [Google Scholar]
- 5.Zhao Y, Butler EB, Tan M. Targeting cellular metabolism to improve cancer therapeutics. Cell Death Dis. 2013;4:e532. doi: 10.1038/cddis.2013.60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tabe Y, Konopleva M. Advances in understanding the leukaemia microenvironment. Br J Haematol. 2014;164(6):767–778. doi: 10.1111/bjh.12725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Choudhery MS, Badowski M, Muise A, Pierce J, Harris DT. Donor age negatively impacts adipose tissue-derived mesenchymal stem cell expansion and differentiation. J Transl Med. 2014;12:8. doi: 10.1186/1479-5876-12-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Justesen J, Stenderup K, Ebbesen EN, Mosekilde L, Steiniche T, Kassem M. Adipocyte tissue volume in bone marrow is increased with aging and in patients with osteoporosis. Biogerontology. 2001;2(3):165–171. doi: 10.1023/a:1011513223894. [DOI] [PubMed] [Google Scholar]
- 9.Lecka-Czernik B, Rosen CJ, Kawai M. Skeletal aging and the adipocyte program: New insights from an “old” molecule. Cell Cycle. 2010;9(18):3648–3654. doi: 10.4161/cc.9.18.13046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tokuda Y, Satoh Y, Fujiyama C, Toda S, Sugihara H, Masaki Z. Prostate cancer cell growth is modulated by adipocyte-cancer cell interaction. BJU Int. 2003;91(7):716–720. doi: 10.1046/j.1464-410x.2003.04218.x. [DOI] [PubMed] [Google Scholar]
- 11.Dirat B, Bochet L, Dabek M, Daviaud D, Dauvillier S, Majed B, et al. Cancer-associated adipocytes exhibit an activated phenotype and contribute to breast cancer invasion. Cancer Res. 2011;71(7):2455–2465. doi: 10.1158/0008-5472.CAN-10-3323. [DOI] [PubMed] [Google Scholar]
- 12.Tabe Y, Konopleva M, Munsell MF, Marini FC, Zompetta C, McQueen T, et al. PML-RARalpha is associated with leptin-receptor induction: the role of mesenchymal stem cell-derived adipocytes in APL cell survival. Blood. 2004;103(5):1815–1822. doi: 10.1182/blood-2003-03-0802. [DOI] [PubMed] [Google Scholar]
- 13.Behan JW, Yun JP, Proektor MP, Ehsanipour EA, Arutyunyan A, Moses AS, et al. Adipocytes impair leukemia treatment in mice. Cancer Res. 2009;69(19):7867–7874. doi: 10.1158/0008-5472.CAN-09-0800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Samudio I, Harmancey R, Fiegl M, Kantarjian H, Konopleva M, Korchin B, et al. Pharmacologic inhibition of fatty acid oxidation sensitizes human leukemia cells to apoptosis induction. J Clin Invest. 2010;120(1):142–156. doi: 10.1172/JCI38942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Battula VL, Chen Y, Cabreira MaG, Ruvolo V, Wang Z, Ma W, et al. Connective tissue growth factor regulates adipocyte differentiation of mesenchymal stromal cells and facilitates leukemia bone marrow engraftment. Blood. 2013;122(3):357–366. doi: 10.1182/blood-2012-06-437988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Herroon MK, Rajagurubandara E, Hardaway AL, Powell K, Turchick A, Feldmann D, et al. Bone marrow adipocytes promote tumor growth in bone via FABP4-dependent mechanisms. Oncotarget. 2013;4(11):2108–2123. doi: 10.18632/oncotarget.1482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Johnson AR, Milner JJ, Makowski L. The inflammation highway: metabolism accelerates inflammatory traffic in obesity. Immunol Rev. 2012;249(1):218–238. doi: 10.1111/j.1600-065X.2012.01151.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhang Y, Morgan MJ, Chen K, Choksi S, Liu ZG. Induction of autophagy is essential for monocyte-macrophage differentiation. Blood. 2012;119(12):2895–2905. doi: 10.1182/blood-2011-08-372383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Tallman MS, Kim HT, Paietta E, Bennett JM, Dewald G, Cassileth PA, et al. Acute monocytic leukemia (French-American-British classification M5) does not have a worse prognosis than other subtypes of acute myeloid leukemia: a report from the Eastern Cooperative Oncology Group. J Clin Oncol. 2004;22(7):1276–1286. doi: 10.1200/JCO.2004.08.060. [DOI] [PubMed] [Google Scholar]
- 20.Milella M, Kornblau SM, Estrov Z, Carter BZ, Lapillonne H, Harris D, et al. Therapeutic targeting of the MEK/MAPK signal transduction module in acute myeloid leukemia. J Clin Invest. 2001;108(6):851–859. doi: 10.1172/JCI12807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012;7(3):562–578. doi: 10.1038/nprot.2012.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Takahashi H, Lassmann T, Murata M, Carninci P. 5’ end-centered expression profiling using cap-analysis gene expression and next-generation sequencing. Nat Protoc. 2012;7(3):542–561. doi: 10.1038/nprot.2012.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Supek F, Bošnjak M, Škunca N, Šmuc T. REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One. 2011;6(7):e21800. doi: 10.1371/journal.pone.0021800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Soga T, Ohashi Y, Ueno Y, Naraoka H, Tomita M, Nishioka T. Quantitative metabolome analysis using capillary electrophoresis mass spectrometry. J Proteome Res. 2003;2(5):488–494. doi: 10.1021/pr034020m. [DOI] [PubMed] [Google Scholar]
- 25.Cabrero A, Alegret M, Sanchez RM, Adzet T, Laguna JC, Carrera MV. Increased reactive oxygen species production down-regulates peroxisome proliferator-activated alpha pathway in C2C12 skeletal muscle cells. J Biol Chem. 2002;277(12):10100–10107. doi: 10.1074/jbc.M110321200. [DOI] [PubMed] [Google Scholar]
- 26.Hossain F, Al-Khami AA, Wyczechowska D, Hernandez C, Zheng L, Reiss K, et al. Inhibition of Fatty Acid Oxidation Modulates Immunosuppressive Functions of Myeloid-Derived Suppressor Cells and Enhances Cancer Therapies. Cancer Immunol Res. 2015;3(11):1236–1247. doi: 10.1158/2326-6066.CIR-15-0036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mazumder PK, O’Neill BT, Roberts MW, Buchanan J, Yun UJ, Cooksey RC, et al. Impaired cardiac efficiency and increased fatty acid oxidation in insulin-resistant ob/ob mouse hearts. Diabetes. 2004;53(9):2366–2374. doi: 10.2337/diabetes.53.9.2366. [DOI] [PubMed] [Google Scholar]
- 28.Brahimi-Horn MC, Chiche J, Pouysségur J. Hypoxia signalling controls metabolic demand. Curr Opin Cell Biol. 2007;19(2):223–229. doi: 10.1016/j.ceb.2007.02.003. [DOI] [PubMed] [Google Scholar]
- 29.Hardie DG, Ross FA, Hawley SA. AMPK: a nutrient and energy sensor that maintains energy homeostasis. Nat Rev Mol Cell Biol. 2012;13(4):251–262. doi: 10.1038/nrm3311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kawaji H, Lizio M, Itoh M, Kanamori-Katayama M, Kaiho A, Nishiyori-Sueki H, et al. Comparison of CAGE and RNA-seq transcriptome profiling using clonally amplified and single-molecule next-generation sequencing. Genome Res. 2014;24(4):708–717. doi: 10.1101/gr.156232.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Liu L, Zhang J, Yuan J, Dang Y, Yang C, Chen X, et al. Characterization of a human regulatory subunit of protein phosphatase 3 gene (PPP3RL) expressed specifically in testis. Mol Biol Rep. 2005;32(1):41–45. doi: 10.1007/s11033-004-4250-4. [DOI] [PubMed] [Google Scholar]
- 32.Iwabu M, Yamauchi T, Okada-Iwabu M, Sato K, Nakagawa T, Funata M, et al. Adiponectin and AdipoR1 regulate PGC-1alpha and mitochondria by Ca(2+) and AMPK/SIRT1. Nature. 2010;464(7293):1313–1319. doi: 10.1038/nature08991. [DOI] [PubMed] [Google Scholar]
- 33.He H, Liu X, Lv L, Liang H, Leng B, Zhao D, et al. Calcineurin suppresses AMPK-dependent cytoprotective autophagy in cardiomyocytes under oxidative stress. Cell Death Dis. 2014;5:e997. doi: 10.1038/cddis.2013.533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Alers S, Löffler AS, Wesselborg S, Stork B. Role of AMPK-mTOR-Ulk1/2 in the regulation of autophagy: cross talk, shortcuts, and feedbacks. Mol Cell Biol. 2012;32(1):2–11. doi: 10.1128/MCB.06159-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Yamauchi T, Nio Y, Maki T, Kobayashi M, Takazawa T, Iwabu M, et al. Targeted disruption of AdipoR1 and AdipoR2 causes abrogation of adiponectin binding and metabolic actions. Nat Med. 2007;13(3):332–339. doi: 10.1038/nm1557. [DOI] [PubMed] [Google Scholar]
- 36.Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, et al. Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat Med. 2002;8(11):1288–1295. doi: 10.1038/nm788. [DOI] [PubMed] [Google Scholar]
- 37.Cawthorn WP, Scheller EL, Learman BS, Parlee SD, Simon BR, Mori H, et al. Bone marrow adipose tissue is an endocrine organ that contributes to increased circulating adiponectin during caloric restriction. Cell Metab. 2014;20(2):368–375. doi: 10.1016/j.cmet.2014.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Medina EA, Oberheu K, Polusani SR, Ortega V, Velagaleti GV, Oyajobi BO. PKA/AMPK signaling in relation to adiponectin’s antiproliferative effect on multiple myeloma cells. Leukemia. 2014;28(10):2080–2089. doi: 10.1038/leu.2014.112. [DOI] [PubMed] [Google Scholar]
- 39.Gwinn DM, Shackelford DB, Egan DF, Mihaylova MM, Mery A, Vasquez DS, et al. AMPK phosphorylation of raptor mediates a metabolic checkpoint. Mol Cell. 2008;30(2):214–226. doi: 10.1016/j.molcel.2008.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Yilmaz OH, Valdez R, Theisen BK, Guo W, Ferguson DO, Wu H, et al. Pten dependence distinguishes haematopoietic stem cells from leukaemia-initiating cells. Nature. 2006;441(7092):475–482. doi: 10.1038/nature04703. [DOI] [PubMed] [Google Scholar]
- 41.Hoshii T, Tadokoro Y, Naka K, Ooshio T, Muraguchi T, Sugiyama N, et al. mTORC1 is essential for leukemia propagation but not stem cell self-renewal. J Clin Invest. 2012;122(6):2114–2129. doi: 10.1172/JCI62279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Wu S, Zheng C, Chen S, Lin B, Chen Y, Zhou W, et al. Adiponectin signals through Adiponectin Receptor 1 to reverse imatinib resistance in K562 human chronic myeloid leukemia cells. Biochem Biophys Res Commun. 2015;456(1):367–372. doi: 10.1016/j.bbrc.2014.11.089. [DOI] [PubMed] [Google Scholar]
- 43.Dai B, Gong A, Jing Z, Aldape KD, Kang SH, Sawaya R, et al. Forkhead box M1 is regulated by heat shock factor 1 and promotes glioma cells survival under heat shock stress. J Biol Chem. 2013;288(3):1634–1642. doi: 10.1074/jbc.M112.379362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Zappasodi R, Ruggiero G, Guarnotta C, Tortoreto M, Tringali C, Cavanè A, et al. HSPH1 inhibition downregulates Bcl-6 and c-Myc and hampers the growth of human aggressive B-cell non-Hodgkin lymphoma. Blood. 2015;125(11):1768–1771. doi: 10.1182/blood-2014-07-590034. [DOI] [PubMed] [Google Scholar]
- 45.Hirayama A, Kami K, Sugimoto M, Sugawara M, Toki N, Onozuka H, et al. Quantitative metabolome profiling of colon and stomach cancer microenvironment by capillary electrophoresis time-of-flight mass spectrometry. Cancer Res. 2009;69(11):4918–4925. doi: 10.1158/0008-5472.CAN-08-4806. [DOI] [PubMed] [Google Scholar]
- 46.Zhang Q, Yu J, Liu B, Lv Z, Xia T, Xiao F, et al. Central activating transcription factor 4 (ATF4) regulates hepatic insulin resistance in mice via S6K1 signaling and the vagus nerve. Diabetes. 2013;62(7):2230–2239. doi: 10.2337/db12-1050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Jiang Q, Li F, Shi K, Wu P, An J, Yang Y, et al. Involvement of p38 in signal switching from autophagy to apoptosis via the PERK/eIF2α/ATF4 axis in selenite-treated NB4 cells. Cell Death Dis. 2014;5:e1270. doi: 10.1038/cddis.2014.200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Jaganathan S, Malek E, Vallabhapurapu S, Driscoll JJ. Bortezomib induces AMPK-dependent autophagosome formation uncoupled from apoptosis in drug resistant cells. Oncotarget. 2014;5(23):12358–12370. doi: 10.18632/oncotarget.2590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Thornton TM, Rincon M. Non-classical p38 map kinase functions: cell cycle checkpoints and survival. Int J Biol Sci. 2009;5(1):44–51. doi: 10.7150/ijbs.5.44. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.