Primary metabolism and medium-chain fatty acid alterations precede

Article pubs.acs.org/jpr

Primary Metabolism and Medium-Chain Fatty Acid Alterations Precede Long-Chain Fatty Acid Changes Impacting Neutral Lipid Metabolism in Response to an Anticancer Lysophosphatidylcholine Analogue in Yeast Nicolas P. Tambellini,†,‡ Vanina Zaremberg,*,† Saikumari Krishnaiah,§ Raymond J. Turner,† and Aalim M. Weljie*,†,‡,§ †

Department of Biological Sciences, University of Calgary, Calgary, Alberta T2N 1N4, Canada Metabolomics Research Centre, University of Calgary, Calgary, Alberta T2N 1N4, Canada § Department of Systems Pharmacology and Translational Therapeutics, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania 19104-5158, United States of America ‡

S Supporting Information *

ABSTRACT: The nonmetabolizable lysophosphatidylcholine (LysoPC) analogue edelfosine is the prototype of a class of compounds being investigated for their potential as selective chemotherapeutic agents. Edelfosine targets membranes, disturbing cellular homeostasis. Is not clear at this point how membrane alterations are communicated between intracellular compartments leading to growth inhibition and eventual cell death. In the present study, a combined metabolomics/ lipidomics approach for the unbiased identification of metabolic pathways altered in yeast treated with sublethal concentrations of the LysoPC analogue was employed. Mass spectrometry of polar metabolites, fatty acids, and lipidomic profiling was used to study the effects of edelfosine on yeast metabolism. Amino acid and sugar metabolism, the Krebs cycle, and fatty acid profiles were most disrupted, with polar metabolites and short−medium chain fatty acid changes preceding long and very long-chain fatty acid variations. Initial increases in metabolites such as trehalose, proline, and γ-amino butyric acid with a concomitant decrease in metabolites of the Krebs cycle, citrate and fumarate, are interpreted as a cellular attempt to offset oxidative stress in response to mitochondrial dysfunction induced by the treatment. Notably, alanine, inositol, and myristoleic acid showed a steady increase during the period analyzed (2, 4, and 6 h after treatment). Of importance was the finding that edelfosine induced significant alterations in neutral glycerolipid metabolism resulting in a significant increase in the signaling lipid diacylglycerol. KEYWORDS: anticancer drug, cell metabolism, drug metabolism, fatty acid metabolism, lipid metabolism, lipidomics, mass spectrometry (MS), metabolomics, multivariate statistical analysis, Saccharomyces cerevisiae

■

INTRODUCTION Selective targeting of cellular membranes provides a nonmutagenic alternative to more traditional chemotherapeutic strategies targeting DNA or the cytoskeleton. Synthetic etherlinked lysophosphatidylcholine (LysoPC) analogues have emerged as compounds that could be used as chemotherapeutic agents and are thought to act by targeting the cell membrane.1−4 Many of these lipids have demonstrated selective antitumor activities in vitro and in vivo2,5 as well as antiparasitic activity.6,7 The alkyl-lysophospholipid 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine, commonly known as edelfosine, is a prototype of this group of lipids. Edelfosine is a LysoPC analog that has been shown to selectively induce apoptosis in tumor cells, and this has been linked to the ability of these cells to uptake the lipid drug.4,8 © 2017 American Chemical Society

Genetic studies performed in the model organism Saccharomyces cerevisiae identified a conserved flippase that facilitates internalization of edelfosine.9 Localization studies using fluorescent edelfosine derivatives have shown that edelfosine accumulates in the endoplasmic reticulum (ER) once inside the cell,10,11 leading to ER stress12 and inhibition of phosphatidylcholine synthesis through the Kennedy pathway.13 In addition, a number of surface receptors, phospholipases, and signaling pathways have been proposed as cellular targets of edelfosine.14 These disparate effects of edelfosine can be reconciled by the fact that edelfosine alters membrane architecture and its biophysical properties, impacting on a Received: June 20, 2017 Published: August 29, 2017 3741

DOI: 10.1021/acs.jproteome.7b00430 J. Proteome Res. 2017, 16, 3741−3752

Article

Journal of Proteome Research

targeted fatty acid analysis, and untargeted lipidomic profiling of yeast treated with edelfosine at concentrations that induce a cytostatic effect identify perturbations induced by the treatment through comparison to untreated yeast. A surprisingly wide-scale metabolic perturbation is identified that suggests a metabolic switch in central carbon metabolism and alterations of neutral lipid reservoirs as part of a cellular attempt to revert the lysolipid burden, which reduces proliferative ability. Integration of the identified metabolic pathways with the results of previous chemogenomic screens11,16 points to mitochondrial dysfunction and oxidative stress as critical mediators of the cellular response to edelfosine. The results open new hypotheses on how cells sense membrane alterations and how different cellular compartments communicate in order to trigger the appropriate response, with diacylglycerol arising as an important signaling lipid in this process.

wide range of membrane associated proteins. Unbiased genetic studies conducted in budding yeast support this view, where choline-containing lysolipid analogues induce biophysical modifications of the plasma membrane15 leading to alteration of pH homeostasis.16 In addition, defects in vesicle trafficking pathways resulted in resistance to edelfosine, by allowing the plasma membrane architecture to be repaired.11 A highly enriched cluster of genes related to mitochondria function was also identified in these studies, but the role of this organelle in modulating sensitivity to edelfosine is not yet clear.11 Interestingly, a metabolic flux study using 13C-labeled glucose in cancer cells17 was able to identify progressive alterations of the TCA cycle induced by edelfosine, leading to ROS-induced apoptosis. Altogether these results point to a mechanism by which changes in cellular membranes by edelfosine lead to a shift in metabolism involving mitochondria, changes in the TCA cycle, and oxidative stress. A gap remains however in the understanding of how cellular membranes and organelles communicate and what kind of checkpoints operate in order to trigger the appropriate response upon sensing alterations in membranes. The combined use of unbiased approaches from the omics fields was favored in this study to elucidate the temporal changes induced by edelfosine and to identify pathways and metabolites involved in the cellular response to the treatment. Metabolomics allows for the metabolic status to be monitored and gives a snapshot of the processes occurring in a cell. As metabolomics has emerged as a viable tool, it has been used extensively in drug discovery and development in addition to many other fields.18 Recent advances in metabolomics technologies and data processing have allowed for studies that encompass the global cellular metabolism in contrast to previous studies that targeted specific metabolite classes or metabolic pathways. This untargeted approach has recently been used with great success in our laboratories to study the effects of metal toxicity on bacteria19,20 and cancer hypoxia.21 Metabolomic studies have been successfully used to uncover changes in the fatty acid and lipid profile of yeast cells during different phases of growth22 or caused by defective lipid biosynthesis,23 as well as the response to furfural, phenol, acetic acid,24 and copper.25 Here a combined metabolomics and lipidomics study is reported identifying comprehensive metabolic changes induced by edelfosine in the model organism Saccharomyces cerevisiae (budding yeast). Yeast is a powerful model for the identification of pathways involved in drug action26−29 and has been extensively used as an experimental system to investigate the mode of action of edelfosine and other medically relevant lysophosphatidylcholine analogues.11,15,16,30−32 Yeast is susceptible to edelfosine at similar concentrations to those used to induce apoptosis in tumoral cells.15 One additional advantage of using budding yeast for this study is that it possesses a relatively simple lipidome with the main fatty acid species consisting of 16:0, 16:1, 18:0, and 18:1.23 Yeast has been at the vanguard of the development of functional genomic, transcriptomic, proteomic, and metabolomic approaches and currently represents the organism with the most comprehensive and well curated experimental data set available to the research community.33,34 New levels of understanding of the physiological roles of lipids and their remodeling as part of the cell response to drugs and environmental changes is emerging as a consequence of the use of these systems’ biology endeavors. In this work, untargeted profiling of the polar metabolites,

■

EXPERIMENTAL PROCEDURES

Yeast and Edelfosine Growth Curves

Yeast strain BY4741 (MATa his3 leu2 met15 ura3) was obtained from Euroscarf (Frankfurt, Germany) and grown in 50 mL cultures in liquid synthetic medium containing 0.67% Yeast Nitrogen Base with ammonium sulfate (MP Biomedical, Solon OH, USA), 2% glucose, and histidine, leucine, methionine, and uracil to fulfill auxotrophies, as described previously.35 Two biological replicates were started to generate technical replicates, and this was repeated twice. Each culture was grown to an initial A600 of 0.2 from a start point of approximately 0.05, and edelfosine was then added. Edelfosine was dissolved in anhydrous ethanol and added at the indicated concentrations while 0.2% (v/v) ethanol only was added to the control culture. A600 readings were taken every 90 min after edelfosine addition for the indicated times. A final reading was taken at 32 h to determine if recovery of the culture had occurred. Two sets of triplicate readings were taken for each concentration of edelfosine and the control culture. The mean and standard error for each concentration were then graphed using GraphPad Prism 3.03 software (La Jolla, CA, USA). Yeast Sample Growth and Sample Harvesting

Wild type strain BY4741 was grown in 1.75 L cultures to log phase (A600 of 0.2). At this point the culture was split and 2 μg/ mL edelfosine or vehicle (0.2% v/v anhydrous ethanol) was added to each flask (treated and control culture, respectively). Cells (10 optical density units as 600 nm) were harvested in multiples of six at each time point of 0, 2, 4, and 6 h after edelfosine addition from both the edelfosine treated and untreated culture. Each harvested sample was washed twice with water to remove all growth media, flash frozen in liquid nitrogen to prevent further growth and metabolite turnover, and stored at −80 °C until extraction. Sample Extraction and Derivatization

A modified version of a cell extraction method previously described by McCombie et al.36 and described in detail by Tambellini et al.35 was used. Briefly, yeast cell pellets were resuspended in CHCl3/CH3OH (1:2, v/v) and cells disrupted with acid-washed glass beads at 4 °C using a BioSpec MultiBead Beater (Bartlesville, OK, USA). Next CHCl3 and H2O were added to facilitate phase separation and extraction of lipids and polar metabolites. Seventy-five microliter (75 μL) aliquots from the organic phase of the edelfosine treated and untreated 3742


Article


B for 3 min, followed by a ramp of 10% solvent A and 90% solvent B for 8 min, and a final ramp to 6% solvent A and 94% solvent B for 8 min. The column eluent was introduced directly into the mass spectrometer. Mass spectrometry was performed on a quadrupole time-of-flight instrument (Xevo G2-S, Waters, Milford, MA) operating in positive ion, sensitive mode with a capillary voltage of 3000 V and sampling cone temperature of 40 °C. The desolvation gas flow was set to 800 L/h, and the temperature was set to 450 °C. The source temperature was set at 80 °C. Accurate mass was maintained by introduction of a lock-spray interface of leucine-enkephalin (556.2771 m/z) at a concentration of 0.5 ng/μL in 50% aqueous acetonitrile and a rate of 5 μL/min. Data was acquired in centroid MSe mode from 50 to 1200 m/z mass ranges for both MS and MSe modes. Low energy or fragmented data was collected without collision energy, while high energy or fragmented data was collected by using a collision energy ramp from 15 to 40 eV. The entire set of triplicate sample injections was bracketed with a test mix of standard metabolites at the beginning and at the end of the run to evaluate instrument performance with respect to sensitivity and mass accuracy. The sample queue was randomized to remove bias. Lipid analysis and identifications were done using Progenesis QI software (Waters, Milford, MA).

samples were then transferred to 1.5 mL microcentrifuge tubes after extraction had occurred. Isolated organic fractions containing the lipids and fatty acids and aqueous fractions containing the polar metabolites were dried down in a fume hood or speed vacuum (Eppendorf 5415 C, Hamburg, Germany) and stored at −80 °C. Aqueous samples were prepared for GC-MS analysis by derivatization with methoxyamine and MSTFA (N-methyl-N(trimethylsilyl) trifluoroacetamide) as previously described.19 Briefly 50 μL of 20 mg/mL solution of methoxylaminehydrochloride in pyridine was added to each dry sample with shaking at 37 °C for 2.5 h. After the addition of 50 μL of MSTFA, 45 min of additional shaking at 37 °C followed. Organic samples were prepared for GC-MS fatty acid methyl ester (FAME) analysis by derivatization with BF3/methanol as previously described.36 Briefly the dried down organic fractions were dissolved in 750 μL of 1:1 (CHCl3/CH3OH) under sonication for 15 min. 125 μL of BF3/methanol was then added, and the samples were incubated in glass vials at 80 °C for 90 min. After cooling, 300 μL of H2O and 600 μL of hexane were added to each sample and phase separation was allowed. The organic layer was isolated and evaporated to dryness overnight in a fume hood. GC-MS Data Acquisition

GC-MS Data Processing and Multivariate Statistical Projection and Analysis

GC-MS acquisition was carried out as previously described35 using a Waters Micromass GCT Premier GC-TOF-MS (Waters, Mississauga Ontario, Canada) coupled to a 7683 B Series Injector and Autosampler (Agilent Technologies, Mississauga Ontario, Canada). A DB-5MS 30 m × 0.25 mm column with a 0.25 μm filament size (Agilent Technologies, Mississauga Ontario, Canada) was used for polar metabolite analysis, and a DB-23 60 m × 0.25 mm column with a 0.15 μm filament size (Agilent Technologies, Mississauga Ontario, Canada) was used for FAME (fatty acid methyl ester) analysis.

Raw GC-MS data from polar metabolite analysis was imported to a MetaboliteDetector37 for peak detection and identification using an untargeted approach. Peak detection and identification for FAME analysis was done with AMDIS/MetIdea using a targeted approach with a 37 FAME standard (Supelco, Bellefonte, PA, USA). After peak detection and identification, the data were normalized using Excel 2010 (Microsoft, Redmond, WA, USA) first to the internal standard, D-25 Tridecanoic Acid, in the case of FAME profiling, followed by integral normalization for both the FAME and the polar metabolite profiling. After normalization and peak detection, univariate scaling and mean centering were applied before the model construction and validation steps, and this was all done using SIMCA-P (Version 13) (MKS Umetrics, Umeå, Sweden), a multivariate statistical analytical software. To avoid overfitting of the data, the autofit routine of SIMCA-P was used during model construction, and for 7-fold cross-validation of the OPLS-DA (orthogonal partial least-squares-discriminant analysis) models and CV-ANOVA (cross validated-analysis of variance) p-values were used.

UPLC-QTOF-MS Based Data Acquisition for Untargeted Lipidomic Profiling

Four-hour time point edelfosine treated and untreated samples were thawed on ice. Lipids were extracted by a 2:1 (CH3OH/ CHCl3) mixture (300 μL), and then vortexed and sonicated for 15 min. Chloroform and water were added (100 μL each) and the samples vortexed. Organic and aqueous layers were separated by centrifugation for 7 min at 10,000g at 4 °C. Internal lipid standards were added to the organic layer and the samples dried in a fume hood. All lipid standards PC (17:0/ 14:1), PC (17:0/20:4), PC (21:0/22:6), Lyso PC (13:0/0:0), Lyso PC (17:1/0:0), PE (17:0/14:1), D5-16:1 DAG, D5-18:1 DAG, and D5-18:2 DAG were purchased from Avanti Polar Lipids (Alabaster, AL, USA). The dried lipid samples were redissolved in a mixture of 60% solvent A (40% H2O, 60% ACN, 10 mM Ammonium formate) and 40% solvent B (90% IPA, 10% ACN, 10 mM ammonium formate). The samples were centrifuged at 10000g for 5 min to remove fine particulates. The supernatant was transferred to a glass vial for Ultraperformance liquid chromatography on qTOF Xevo G2S (Waters Corporation, Milford, MA, USA) for highthroughput LC-MS lipidomics. For each sample, 10 μL was injected onto a reverse-phase HSS T3 1.8 μm 2.1 × 150 mm column using an Aquity H-class UPLC system (Waters Corporation, Milford, MA, USA). Each sample was resolved for 20 min, at a flow rate of 0.3 mL/min. The UPLC gradient consisted of 75% solvent A and 25% solvent B for 1 min, a ramp of 50% solvent A and 50% solvent

Data Analysis and Multivariate Projection Modeling

Raw LC-MS data files were converted to mzXML format using masswolf. The data from 12 edelfosine treated and 10 untreated yeast samples was uploaded to XC-MS online.38 Analysis and peak detection was then carried out using the HPLC/Waters TOF parameters built into the server with [M + H], [M + NH4], [M + Na], [M + H − H20], [M + K], [M + ACN + H], and [M + ACN + Na] adducts detected. Upon initial examination of the resulting data, it was observed that edelfosine treated samples had approximately 5-fold higher total ion counts than untreated samples for an unknown reason even though the internal standard concentrations were verified to be statistically the same using a t test. Figure S1 is a cloudplot demonstrating the extent of this disparity. As a result the data was exported out of XC-MS online and normalized using total integral normalization to allow for compositional 3743


Article

Journal of Proteome Research comparison of treated and untreated samples using multivariate projection modeling and statistical analysis in SIMCA-P13 (Umetrics AB, Umeå Sweden). Edelfosine (523.73 g/mol) and presumed downstream metabolite adducts and isotope peaks were identified and removed before normalization as they could affect the normalization and modeling steps. These peaks were determined by a series of criteria including similar mass and retention times, as well as detection only in the treated samples through the first phase of XCMS peak detection and fold changes >10 between the treated and untreated samples following the peakfilling procedure. The complete list of 51 removed peaks is provided in Table S1. After import into SIMCA-P Version 13, mean centering and pareto scaling were applied to the data followed by PCA and OPLS-DA modeling. Additionally an S-plot was constructed to identify significantly increased or decreased lipids as the use of VIP (variable influence on projection) scores for screening is not ideal with a high number of variables, in this case thousands. S-plots combine the modeled variance and modeled correlation from the OPLS-DA plot using the magnitude of each variable (p) and reliability of each variable (p(corr)) to screen out metabolite peaks that have low magnitudes of change or intensities.39 Waters Progenesis software (Waters Corporation, Milford, MA, USA) was used to identify the lipid features found to be noticeably increased or decreased by edelfosine treatment.

Figure 1. Yeast and edelfosine treatment growth curves. Wild type yeast strain BY4741 was to an initial OD600 (optical density at 600 nm) of approximately 0.2 from a start point of approximately 0.05. Edelfosine was then added at 2, 4, 8, and 16 μg/mL. Two sets of triplicate readings were taken, and the mean and standard error of the mean (SEM) for each concentration were calculated.

Table 1. Summary of Parameters for the Assessment of the Quality of OPLS-DA Models Comparing Edelfosine Treated and Untreated Yeast Samples Analyzed by GC-MSa Time After Edelfosine Treatment 0h

Metabolite Modeling and Pathway Analysis

■

Polar Metabolites FAME’s Polar Metabolites FAME’s Polar Metabolites FAME’S Lipidomics Polar Metabolites FAME’S

0h 2h

Polar metabolites and fatty acids identified to have VIP scores greater than one through OPLS-DA modeling of edelfosine treated and untreated yeast cells were subjected to pathway analysis using the MetaboAnalyst 3.040 server. Pathway analysis was done with 17 polar metabolites and 8 fatty acids using the S. cerevisiae pathway library and hypergeometric test and relative-betweeness centrality algorithms for the over representation analysis and pathway topology analysis portions, respectively.

2h 4h 4h 4h 6h 6h

RESULTS

a 2

Model Group

CV-ANOVA p-value

R2X

R2Y

Q2

0.237

0.388

0.005

1.000

0.897 0.607

0.489 0.992

0.156 0.847

0.825 0.004

0.768 0.696

0.655 0.989

0.293 0.918

0.401 7.078 e−5

0.751 0.722 0.202

0.830 0.963 0.573

0.772 0.874 0.464

0.002 1.43 e−5 0.003

0.59

0.762

0.661

4.101 e−4

2

R X and R Y are the modeled variation in the X and Y matrix, respectively, Q2 is the predicted variation, and the CV-ANOVA p-value is obtained from the cross-validated analysis of variance of the OPLSDA model.

Edelfosine Has a Rapid Impact on the Polar Metabolome at Cytostatic Concentrations

In order to minimize biological variation introduced through cell death, a cytostatic concentration of edelfosine was determined (Figure 1). Dose-dependent growth of wild type yeast exposed to edelfosine was recorded, and 2 μg/mL of edelfosine achieved the desired effect as evidenced by the halting of growth during the first 6 h followed by culture recovery overnight. Higher concentrations did not recover, and as such were deemed overly cytotoxic for the purpose of this study. Based on the known effects of edelfosine on ER fuction and lipid interaction, the metabolome and fatty acid composition of the model organism S. cerevisiae was assessed using comparative metabolomics and lipidomics. Yeast treated with the sublethal concentration of edelfosine were compared to vehicle control yeast at 0, 2, 4, and 6 h. Polar metabolites were extracted and analyzed using GC-MS, followed by orthogonal partial leastsquares-discriminant analysis (OPLS-DA) to elucidate global patterns of change (Table 1) as OPLS-DA is a pattern recognition approach. The OPLS-DA models for polar metabolites show a significant difference between the treated and untreated samples at 2, 4, and 6 h after the addition of edelfosine as evidenced by the predictive ability (Q2 > 0.4) and

low CV-ANOVA (cross-validated analysis of variance) p-values 0.6 and significant CV-ANOVA p-value 1, with a higher score indicating a greater influence on the separation of the two sample groups. These time points displayed Q2 > 0.5 and CV-ANOVA p-values 0.1 and raw pvalues 0.05 at nonzero time points (#, p < 0.1; *, p < 0.05). (B) Fold change of selected polar metabolites representing the various changes observed, with citrate and glycine elevated in WT; alanine, myo-inositol, glucose, and glucose-6-phosphate elevated in edelfosine treated yeast; and 6 h differences in lactate and fumarate. Significance noted as in part A. Error bars indicate the SEM.

An interesting and significant finding of our study is that eight fatty acids (FAs) were found to be different in treated and untreated samples with a delayed temporal profile as compared to the polar disruptions. The overall trend in FA perturbation points to an increase in C12−C14 FAs at the expense of C10:0 at early time points (2 and 4 h) followed by a significant decrease in saturated long (C16:0, C20:0) and very long FAs (C22:0, C24:0) at the latest time point analyzed (6 h). Of note is a steady increase of myristoleic acid (C14:1) across all time points. This suggests that edelfosine treatment induces acyl chain remodeling. The shortening of fatty acids and the increment in the degree of unsaturation resembles the adaptation described for yeast mutants defective in acyl-CoAbinding protein, Acb1p.56,57 Furthermore, a reduction in inositol-mediated repression of UASINO containing genes has been described for acb1 mutants.56 Inositol is the master regulator of glycerophospholipid biosynthesis in yeast58 making it of particular interest that myo-inositol was found to be increased in edelfosine treated compared to untreated yeast samples (Table S2). Insight on how edelfosine could be inducing shortening of acyl chains could be linked to the fact that yeast strains defective in genes involved in mitochondrial medium chain fatty acid synthesis (etr1, eht1, lip5) as well as in long and very long fatty acid elongation (elo2, elo3) were identified in a genetic screen for edelfosine hypersensitivity in yeast.16 It is therefore speculated that mitochondrial FA synthesis could be involved in providing medium long chain

3B). Trehalose is a well-known oxidative stress protectant of yeast membranes and proteins48−50 while increased levels of proline and inositol were shown to confer tolerance to various stress conditions by reducing reactive oxygen species levels in yeast.51,52 Another metabolite that increased with the treatment was γ-aminobutyrate (GABA), which has also been linked to oxidative stress protection53 and thermotolerance54 in yeast. Therefore, these metabolites could be part of an early response to oxidative stress induced by edelfosine which has been previously observed in yeast32 and tumoral cells.55 In summary, the emerging map of pathways altered by edelfosine suggests a metabolic reprogramming that favors gluconeogenesis and glucose storage in the form of trehalose (Figure 7). In addition, pyruvate seems to be diverted toward alanine synthesis impacting also glutamate levels which significantly dropped by 6 h of treatment. A slowdown of the TCA cycle accompanies these metabolic changes as reflected by the decrease in citrate and fumarate, possibly indicative of a deleterious effect of edelfosine on mitochondrial performance. It is interesting to note that a large enrichment of mitochondrial related genes was found in a recent yeast chemogenomic screen using cytotoxic concentrations of edelfosine.11 Considering mitochondria are mostly responsible for the generation of reactive oxygen species, altogether our results on primary metabolism support the notion that edelfosine induces mitochondrial dysfunction, contributing to oxidative stress. 3747


Article


Figure 5. Changes in the fatty acid and polar metabolite profile of yeast treated with edelfosine over a 6 h period. (A) Heatmap of polar metabolite profiles with clustering performed on both metabolites and time points. Heatmaps were constructed in R, and data is shown as the z-score normalized distribution (Red, increased response; Green, decreased). (B) Fatty acid metabolite profiles during 6 h of edelfosine treatment. (C) Heatmap of p-values from Wilcoxon rank sum and signed rank tests demonstrating temporal deviation of individual metabolites in treated vs control samples.

DAG, and TAG was detected in our untargeted lipidomic profiling herein. Diacylglycerol (DAG) is not only an intermediate in glycerolipid metabolism but is also a lipid second messenger with a complex strategic role in the regulation of biochemical networks.60,61 Eleven out of 50 identified lipids perturbed by edelfosine treatment were DAG species, including DAG 28:1, DAG 32:0, DAG 32:2, DAG 34:0, DAG 34:2, and DAG 36:0, which were all increased by treatment. Although an increase in DAG could be the result of the known inhibitory effect of edelfosine on the Kennedy pathway for synthesis of PC,13,15 this explanation is not favored, as these experiments were conducted in the absence of exogenous choline and did not show evidence of changes in PC levels other than an increase in PC 30:2 probably arising from acylation of the LysoPC pools which decreased alongside. Judging from the observed increases in several MAG (16:0, 18:1, and 24:1) and TAG (44:2, 46:2, 48:3, 50:3) species and a decrease in TAG (46:3), the increment in DAG may be the result of neutral lipid turnover and resynthesis. It is worth noting four lysolipid species with choline and ethanolamine headgroups decreased while PC and PE species increased concomitantly. This reflects a cellular effort aimed at decreasing the presence of lysolipids through acylation, in response to the burden imposed by edelfosine. Neutral lipid turnover would result in an increase in fatty acid availability for such acylations. In addition, synthesis of several new PI species must be consuming PA and inositol while also increasing the synthesis of complex sphingolipid species such as

Figure 6. Changes in the lipidome of yeast 4 h post-treatment with edelfosine. 26 known yeast lipids were identified as altered across all replicates with statistical significance. Lipid class abbreviations as indicated in text and DG, diacylglycerol; MG, monoacylglycerol; and TG, triacylglycerol.

fatty acids (C12−14) for remodelling of glycerolipids. A recent study has described alterations of neutral glycerolipid metabolism in the mitochondrial FA synthesis mutant etr1.59 Similarly, a striking effect of edelfosine on the levels of MAG, 3748


Article


Figure 7. Schematic diagram of pathway connections. Open circles indicate the metabolite was not detected, green filled circles indicate the metabolite has lower levels in edelfosine treated samples compared to untreated samples, and red filled circles indicate the metabolite has higher levels in treated compared to untreated samples.

neutral lipid metabolism has been proposed as a means to rapidly regulate acyl-CoA pools and acyl chain remodeling of phospholipids in response to cellular needs.65 A primary aim of this work was to identify metabolic changes triggered by edelfosine which could be used to advance our understanding of not only how this lipid induces cell death but also how cells respond to a LysoPC burden. A possible link between DAG accumulation and mitochondrial metabolism emerging from this study is intriguing. Further work will aim to determine the contribution of each of these pathways to the toxic properties of edelfosine and other LysoPC analogues, like the naturally occurring platelet activating factor (PAF). Recent work using yeast as model organism has led to the discovery that intraneuronal accumulation of PAF alone is sufficient to disrupt mitochondrial function and increase ROS in cultured human neurons, which could be relevant in development of Alzheimer’s disease.30 Future research efforts will be directed at studying the molecular mechanisms by which edelfosine induces the metabolic changes observed, with focus on mitochondria and neutral lipid metabolism.

inositol-phospho-ceramide (IPC). Biosynthesis of IPC produces DAG, possibly contributing to the DAG increment observed. A similar effect on neutral lipid metabolism and specifically on DAG levels has been described for the treatment of tumor cell lines with the alkylphospholipid miltefosine.41 DAG is a cone shaped lipid with a small area occupied by the polar headgroup compared to that of the apolar tails. Thus, the presence of DAG in membranes induces negative curvature and favors nonbilayer inverted phase transitions.62 This is in great contrast with the positive curvature induced by inverted cone shaped lysolipids such as edelfosine, so DAG could potentially correct the curvature stress introduced by edelfosine owing to its complementary shape.63 Therefore, reducing the levels of LysoPC and LysoPE pools while increasing the levels of DAG could be part of a novel cellular strategy aimed at counteracting the stress inflicted by edelfosine. While our methodology was not able to conclusively identify cardiolipiin (CL) species, three accurate mass results were consistent with CLs and all three were significantly elevated with edelfosine treatment. Such changes would be consistent with mitochondrial stress and the proposed mechanism described above. A functional crosstalk between mitochondria and neutral lipid metabolism has been proposed, where oxidative damage due to mitochondrial dysfunction could be ameliorated by lipid droplets, with neutral lipids acting as a “sink” to terminate propagation of radicals.64 Our experimental findings combined with those from previous genetic screens are consistent with this model. Furthermore, the transit of fatty acids through

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jproteome.7b00430. 3749


Article


■

(11) Cuesta-Marbán, Á .; Botet, J.; Czyz, O.; Cacharro, L. M.; Gajate, C.; Hornillos, V.; Delgado, J.; Zhang, H.; Amat-Guerri, F.; Acuña, A. U.; et al. Drug uptake, lipid rafts, and vesicle trafficking modulate resistance to an anticancer lysophosphatidylcholine analogue in yeast. J. Biol. Chem. 2013, 288 (12), 8405−8418. (12) Nieto-Miguel, T.; Fonteriz, R. I.; Vay, L.; Gajate, C.; LópezHernández, S.; Mollinedo, F. Endoplasmic reticulum stress in the proapoptotic action of edelfosine in solid tumor cells. Cancer Res. 2007, 67 (21), 10368−10378. (13) Boggs, K. P.; Rock, C. O.; Jackowski, S. Lysophosphatidylcholine and 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine inhibit the CDP-choline pathway of phosphatidylcholine synthesis at the CTP:phosphocholine cytidylyltransferase step. J. Biol. Chem. 1995, 270 (13), 7757−7764. (14) Arthur, G.; Bittman, R. The inhibition of cell signaling pathways by antitumor ether lipids. Biochim. Biophys. Acta, Lipids Lipid Metab. 1998, 1390 (1), 85−102. (15) Zaremberg, V.; Gajate, C.; Cacharro, L. M.; Mollinedo, F.; McMaster, C. R. Cytotoxicity of an anti-cancer lysophospholipid through selective modification of lipid raft composition. J. Biol. Chem. 2005, 280 (45), 38047−38058. (16) Czyz, O.; Bitew, T.; Cuesta-Marbán, Á .; McMaster, C. R.; Mollinedo, F.; Zaremberg, V. Alteration of plasma membrane organization by an anticancer lysophosphatidylcholine analogue induces intracellular acidification and internalization of plasma membrane transporters in yeast. J. Biol. Chem. 2013, 288 (12), 8419−8432. (17) Selivanov, V. A.; Vizán, P.; Mollinedo, F.; Fan, T. W. M.; Lee, P. W. N.; Cascante, M. Edelfosine-induced metabolic changes in cancer cells that precede the overproduction of reactive oxygen species and apoptosis. BMC Syst. Biol. 2010, 4 (1), 135. (18) Wishart, D. S. Emerging applications of metabolomics in drug discovery and precision medicine. Nat. Rev. Drug Discovery 2016, 15 (7), 473−484. (19) Booth, S. C.; Workentine, M. L.; Wen, J.; Shaykhutdinov, R.; Vogel, H. J.; Ceri, H.; Turner, R. J.; Weljie, A. M. Differences in Metabolism between the Biofilm and Planktonic Response to Metal Stress. J. Proteome Res. 2011, 10 (7), 3190−3199. (20) Booth, S. C.; George, I. F. S.; Zannoni, D.; Cappelletti, M.; Duggan, G. E.; Ceri, H.; Turner, R. J. Effect of aluminium and copper on biofilm development of Pseudomonas pseudoalcaligenes KF707 and P. fluorescens as a function of different media compositions. Metallomics 2013, 5 (6), 723−735. (21) Weljie, A. M.; Bondareva, A.; Zang, P.; Jirik, F. R. (1)H NMR metabolomics identification of markers of hypoxia-induced metabolic shifts in a breast cancer model system. J. Biomol. NMR 2011, 49 (3− 4), 185−193. (22) Casanovas, A.; Sprenger, R. R.; Tarasov, K.; Ruckerbauer, D. E.; Hannibal-Bach, H. K.; Zanghellini, J.; Jensen, O. N.; Ejsing, C. S. Quantitative analysis of proteome and lipidome dynamics reveals functional regulation of global lipid metabolism. Chem. Biol. 2015, 22 (3), 412−425. (23) Ejsing, C. S.; Sampaio, J. L.; Surendranath, V.; Duchoslav, E.; Ekroos, K.; Klemm, R. W.; Simons, K.; Shevchenko, A. Global analysis of the yeast lipidome by quantitative shotgun mass spectrometry. Proc. Natl. Acad. Sci. U. S. A. 2009, 106 (7), 2136−2141. (24) Xia, J.-M.; Yuan, Y.-J. Comparative lipidomics of four strains of Saccharomyces cerevisiae reveals different responses to furfural, phenol, and acetic acid. J. Agric. Food Chem. 2009, 57 (1), 99−108. (25) Farrés, M.; Piña, B.; Tauler, R. LC-MS based metabolomics and chemometrics study of the toxic effects of copper on Saccharomyces cerevisiae. Metallomics 2016, 8 (8), 790−798. (26) Bjornsti, M.-A. Cancer therapeutics in yeast. Cancer Cell 2002, 2 (4), 267−273. (27) Menacho-Marquez, M.; Murguia, J. R. Yeast on drugs: Saccharomyces cerevisiae as a tool for anticancer drug research. Clin. Transl. Oncol. 2007, 9 (4), 221−228.

Figure S1: Cloudplot of raw data from untargeted lipidomics analysis. Table S1: Ions removed from the analysis presumed to be of drug or drug metabolite origin. Table S2. Proposed lipids identified as altered by edelfosine treatment. Table S3: Polar metabolites and fatty acids identified as significantly changed for Edelfosine treated samples compared to untreated samples (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] +1-403-220-4298. *E-mail: [email protected] +1-215-939-4945. ORCID

Aalim M. Weljie: 0000-0002-7145-4494 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding for this work was provided in part by Discovery grants from the National Sciences and Engineering Research Council (NSERC) of Canada to AMW, RJT, and VZ, as well as a Discovery Accelerator Supplement to VZ and an American Cancer Society award to AMW (124268-IRG-78-002-35-IRG).

■

REFERENCES

(1) Berdel, W. E. Ether lipids and derivatives as investigational anticancer drugs. A brief review. Onkologie 1990, 13 (4), 245−250. (2) van Blitterswijk, W. J.; Verheij, M. Anticancer mechanisms and clinical application of alkylphospholipids. Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2013, 1831 (3), 663−674. (3) Wright, M. M.; Howe, A. G.; Zaremberg, V. Cell membranes and apoptosis: role of cardiolipin, phosphatidylcholine, and anticancer lipid analogues. Biochem. Cell Biol. 2004, 82 (1), 18−26. (4) Mollinedo, F.; Gajate, C.; Martín-Santamaría, S.; Gago, F. ET-18OCH3 (edelfosine): a selective antitumour lipid targeting apoptosis through intracellular activation of Fas/CD95 death receptor. Curr. Med. Chem. 2004, 11 (24), 3163−3184. (5) Jaffrès, P.-A.; Gajate, C.; Bouchet, A. M.; Couthon-Gourvès, H.; Chantôme, A.; Potier-Cartereau, M.; Besson, P.; Bougnoux, P.; Mollinedo, F.; Vandier, C. Alkyl ether lipids, ion channels and lipid raft reorganization in cancer therapy. Pharmacol. Ther. 2016, 165, 114−131. (6) Croft, S. L.; Engel, J. Miltefosine–discovery of the antileishmanial activity of phospholipid derivatives. Trans. R. Soc. Trop. Med. Hyg. 2006, 100 (Suppl 1), S4−S8. (7) Varela-M, R. E.; Villa-Pulgarin, J. A.; Yepes, E.; Müller, I.; Modolell, M.; Muñoz, D. L.; Robledo, S. M.; Muskus, C. E.; LópezAbán, J.; Muro, A.; et al. In vitro and in vivo efficacy of ether lipid edelfosine against Leishmania spp. and SbV-resistant parasites. PLoS Neglected Trop. Dis. 2012, 6 (4), e1612. (8) Van Der Luit, A. H.; Budde, M.; Verheij, M.; van Blitterswijk, W. J. Different modes of internalization of apoptotic alkyl-lysophospholipid and cell-rescuing lysophosphatidylcholine. Biochem. J. 2003, 374 (Pt 3), 747−753. (9) Hanson, P. K.; Malone, L.; Birchmore, J. L.; Nichols, J. W. Lem3p is essential for the uptake and potency of alkylphosphocholine drugs, edelfosine and miltefosine. J. Biol. Chem. 2003, 278 (38), 36041− 36050. (10) Quesada, E.; Delgado, J.; Gajate, C.; Mollinedo, F.; Acuña, A. U.; Amat-Guerri, F. Fluorescent phenylpolyene analogues of the ether phospholipid edelfosine for the selective labeling of cancer cells. J. Med. Chem. 2004, 47 (22), 5333−5335. 3750


Article

Journal of Proteome Research (28) Smith, A. M.; Ammar, R.; Nislow, C.; Giaever, G. A survey of yeast genomic assays for drug and target discovery. Pharmacol. Ther. 2010, 127 (2), 156−164. (29) Botstein, D.; Fink, G. R. Yeast: an experimental organism for 21st Century biology. Genetics 2011, 189, 69510.1534/genetics.111.130765 (30) Kennedy, M. A.; Moffat, T. C.; Gable, K.; Ganesan, S.; NiewolaStaszkowska, K.; Johnston, A.; Nislow, C.; Giaever, G.; Harris, L. J.; Loewith, R.; et al. A Signaling Lipid Associated with Alzheimer’s Disease Promotes Mitochondrial Dysfunction. Sci. Rep. 2016, 6 (1), 40. (31) Mahadeo, M.; Nathoo, S.; Ganesan, S.; Driedger, M.; Zaremberg, V.; Prenner, E. J. Disruption of lipid domain organization in monolayers of complex yeast lipid extracts induced by the lysophosphatidylcholine analogue edelfosine in vivo. Chem. Phys. Lipids 2015, 191, 153−162. (32) Bitew, T.; Sveen, C. E.; Heyne, B.; Zaremberg, V. Vitamin E prevents lipid raft modifications induced by an anti-cancer lysophospholipid and abolishes a Yap1-mediated stress response in yeast. J. Biol. Chem. 2010, 285 (33), 25731−25742. (33) Goffeau, A.; Barrell, B. G.; Bussey, H.; Davis, R. W.; Dujon, B.; Feldmann, H.; Galibert, F.; Hoheisel, J. D.; Jacq, C.; Johnston, M.; et al. Life with 6000 genes. Science 1996, 274 (5287), 546−563−7. (34) Cherry, J. M.; Hong, E. L.; Amundsen, C.; Balakrishnan, R.; Binkley, G.; Chan, E. T.; Christie, K. R.; Costanzo, M. C.; Dwight, S. S.; Engel, S. R.; et al. Saccharomyces Genome Database: the genomics resource of budding yeast. Nucleic Acids Res. 2012, 40 (Database issue), D700−D705. (35) Tambellini, N.; Zaremberg, V.; Turner, R.; Weljie, A. Evaluation of Extraction Protocols for Simultaneous Polar and Non-Polar Yeast Metabolite Analysis Using Multivariate Projection Methods. Metabolites 2013, 3 (3), 592−605. (36) McCombie, G.; Medina-Gomez, G.; Lelliott, C. J.; Vidal-Puig, A.; Griffin, J. L. Metabolomic and Lipidomic Analysis of the Heart of Peroxisome Proliferator-Activated Receptor-γ Coactivator 1-β Knock Out Mice on a High Fat Diet. Metabolites 2012, 2 (2), 366−381. (37) Hiller, K.; Hangebrauk, J.; Jäger, C.; Spura, J.; Schreiber, K.; Schomburg, D. MetaboliteDetector: comprehensive analysis tool for targeted and nontargeted GC/MS based metabolome analysis. Anal. Chem. 2009, 81 (9), 3429−3439. (38) Tautenhahn, R.; Patti, G. J.; Rinehart, D. Siuzdak Gary. XCMS Online: A Web-Based Platform to Process Untargeted Metabolomic Data. Anal. Chem. 2012, 84 (11), 5035−5039. (39) Wiklund, S.; Johansson, E.; Sjöström, L.; Mellerowicz, E. J.; Edlund, U.; Shockcor, J. P.; Gottfries, J.; Moritz Thomas; Trygg, J. Visualization of GC/TOF-MS-based metabolomics data for identification of biochemically interesting compounds using OPLS class models. Anal. Chem. 2008, 80 (1), 115−122. (40) Xia, J.; Sinelnikov, I. V.; Han, B.; Wishart, D. S. MetaboAnalyst 3.0–making metabolomics more meaningful. Nucleic Acids Res. 2015, 43 (W1), W251−W257. (41) Engelmann, J.; Henke, J.; Willker, W.; Kutscher, B.; Nössner, G.; Engel, J.; Leibfritz, D. Early stage monitoring of miltefosine induced apoptosis in KB cells by multinuclear NMR spectroscopy. Anticancer Res. 1996, 16 (3B), 1429−1439. (42) Azzouz, S.; Maache, M.; Sánchez-Moreno, M.; Petavy, A. F.; Osuna, A. Effect of alkyl-lysophospholipids on some aspects of the metabolism of Leishmania donovani. J. Parasitol. 2007, 93 (5), 1202− 1207. (43) Yu, S.-L.; An, Y. J.; Yang, H.-J.; Kang, M.-S.; Kim, H.-Y.; Wen, H.; Jin, X.; Kwon, H. N.; Min, K.-J.; Lee, S.-K.; et al. Alaninemetabolizing enzyme Alt1 is critical in determining yeast life span, as revealed by combined metabolomic and genetic studies. J. Proteome Res. 2013, 12 (4), 1619−1627. (44) Diaz-Ruiz, R.; Rigoulet, M.; Devin, A. The Warburg and Crabtree effects: On the origin of cancer cell energy metabolism and of yeast glucose repression. Biochim. Biophys. Acta, Bioenerg. 2011, 1807 (6), 568−576.

(45) Fan, Y.; Zhou, X.; Xia, T.-S.; Chen, Z.; Li, J.; Liu, Q.; Alolga, R. N.; Chen, Y.; Lai, M.-D.; Li, P.; et al. Human plasma metabolomics for identifying differential metabolites and predicting molecular subtypes of breast cancer. Oncotarget 2016, 7 (9), 9925−9938. (46) Buas, M. F.; Gu, H.; Djukovic, D.; Zhu, J.; Drescher, C. W.; Urban, N.; Raftery, D.; Li, C. I. Identification of novel candidate plasma metabolite biomarkers for distinguishing serous ovarian carcinoma and benign serous ovarian tumors. Gynecol. Oncol. 2016, 140 (1), 138−144. (47) Hu, S.; Balakrishnan, A.; Bok, R. A.; Anderton, B.; Larson, P. E. Z.; Nelson, S. J.; Kurhanewicz, J.; Vigneron, D. B.; Goga, A. 13Cpyruvate imaging reveals alterations in glycolysis that precede c-Mycinduced tumor formation and regression. Cell Metab. 2011, 14 (1), 131−142. (48) Singer, M. A.; Lindquist, S. Thermotolerance in Saccharomyces cerevisiae: the Yin and Yang of trehalose. Trends Biotechnol. 1998, 16 (11), 460−468. (49) Benaroudj, N.; Lee, D. H.; Goldberg, A. L. Trehalose accumulation during cellular stress protects cells and cellular proteins from damage by oxygen radicals. J. Biol. Chem. 2001, 276 (26), 24261−24267. (50) Pereira, E. de J.; Panek, A. D.; Eleutherio, E. C. A. Protection against oxidation during dehydration of yeast. Cell Stress Chaperones 2003, 8 (2), 120−124. (51) Wang, X.; Bai, X.; Chen, D.-F.; Chen, F.-Z.; Li, B.-Z.; Yuan, Y.-J. Increasing proline and myo-inositol improves tolerance of Saccharomyces cerevisiae to the mixture of multiple lignocellulose-derived inhibitors. Biotechnol. Biofuels 2015, 8 (1), 142. (52) Takagi, H.; Taguchi, J.; Kaino, T. Proline accumulation protects Saccharomyces cerevisiae cells in stationary phase from ethanol stress by reducing reactive oxygen species levels. Yeast 2016, 33 (8), 355− 363. (53) Coleman, S. T.; Fang, T. K.; Rovinsky, S. A.; Turano, F. J.; Moye-Rowley, W. S. Expression of a glutamate decarboxylase homologue is required for normal oxidative stress tolerance in Saccharomyces cerevisiae. J. Biol. Chem. 2001, 276 (1), 244−250. (54) Cao, J.; Barbosa, J. M.; Singh, N. K.; Locy, R. D. GABA shunt mediates thermotolerance in Saccharomyces cerevisiae by reducing reactive oxygen production. Yeast 2013, 30 (4), 129−144. (55) Wagner, B. A.; Buettner, G. R.; Oberley, L. W.; Burns, C. P. Sensitivity of K562 and HL-60 cells to edelfosine, an ether lipid drug, correlates with production of reactive oxygen species. Cancer Res. 1998, 58 (13), 2809−2816. (56) Feddersen, S.; Neergaard, T. B. F.; Knudsen, J.; Faergeman, N. J. Transcriptional regulation of phospholipid biosynthesis is linked to fatty acid metabolism by an acyl-CoA-binding-protein-dependent mechanism in Saccharomyces cerevisiae. Biochem. J. 2007, 407 (2), 219−230. (57) Rijken, P. J.; Houtkooper, R. H.; Akbari, H.; Brouwers, J. F.; Koorengevel, M. C.; de Kruijff, B.; Frentzen, M.; Vaz, F. M.; de Kroon, A. I. P. M. Cardiolipin molecular species with shorter acyl chains accumulate in Saccharomyces cerevisiae mutants lacking the acyl coenzyme A-binding protein Acb1p: new insights into acyl chain remodeling of cardiolipin. J. Biol. Chem. 2009, 284 (40), 27609− 27619. (58) Henry, S. A.; Kohlwein, S. D.; Carman, G. M. Metabolism and regulation of glycerolipids in the yeast Saccharomyces cerevisiae. Genetics 2012, 190 (2), 317−349. (59) Singh, N.; Yadav, K. K.; Rajasekharan, R. ZAP1-mediated modulation of triacylglycerol levels in yeast by transcriptional control of mitochondrial fatty acid biosynthesis. Mol. Microbiol. 2016, 100 (1), 55−75. (60) Wakelam, M. J. Diacylglycerol–when is it an intracellular messenger? Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 1998, 1436 (1−2), 117−126. (61) Ganesan, S.; Shabits, B. N.; Zaremberg, V. Tracking Diacylglycerol and Phosphatidic Acid Pools in Budding Yeast. Lipid Insights 2015, 8 (Suppl 1), 75−85. 3751


Article

Journal of Proteome Research (62) Goñi, F. M.; Alonso, A. Structure and functional properties of diacylglycerols in membranes. Prog. Lipid Res. 1999, 38 (1), 1−48. (63) Basanez, G.; Nieva, J. L.; Rivas, E.; Alonso, A.; Goñi, F. M. Diacylglycerol and the promotion of lamellar-hexagonal and lamellarisotropic phase transitions in lipids: implications for membrane fusion. Biophys. J. 1996, 70 (5), 2299−2306. (64) Handee, W.; Li, X.; Hall, K. W.; Deng, X.; Li, P.; Benning, C.; Williams, B. L.; Kuo, M.-H. An Energy-Independent Pro-longevity Function of Triacylglycerol in Yeast. PLoS Genet. 2016, 12 (2), e1005878. (65) Mora, G.; Scharnewski, M.; Fulda, M. Neutral lipid metabolism influences phospholipid synthesis and deacylation in Saccharomyces cerevisiae. PLoS One 2012, 7 (11), e49269.

3752


Primary metabolism and medium-chain fatty acid alterations precede

Recommend Documents