Stable Isotope-Assisted Lipidomics Combined ... - ACS Publications

We developed and evaluated a novel strategy for the in-depth analysis of the ... (7) Recent publications show that there is a huge potential for using...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/ac

Stable Isotope-Assisted Lipidomics Combined with Nontargeted Isotopomer Filtering, a Tool to Unravel the Complex Dynamics of Lipid Metabolism Jia Li,†,# Miriam Hoene,‡,§,∥,# Xinjie Zhao,† Shili Chen,† Hai Wei,⊥ Hans-Ulrich Har̈ ing,‡,§,∥ Xiaohui Lin,⊥ Zhongda Zeng,† Cora Weigert,‡,§,∥ Rainer Lehmann,*,‡,§,∥ and Guowang Xu*,† †

CAS Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China ‡ Division of Clinical Chemistry and Pathobiochemistry, Department of Internal Medicine IV, University Hospital Tuebingen, Tuebingen, Germany § Institute for Diabetes Research and Metabolic Diseases of the Helmholtz Centre Munich at the University of Tuebingen (Paul Langerhans Institute Tuebingen), Tuebingen, Germany ∥ German Center for Diabetes Research (DZD), Tuebingen, Germany ⊥ School of Computer Science and Technology, Dalian University of Technology, Dalian, China S Supporting Information *

ABSTRACT: Investigations of complex metabolic mechanisms and networks have become a focus of research in the postgenomic area, thereby creating an increasing demand for sophisticated analytical approaches. One such tool is lipidomics analysis that provides, a detailed picture of the lipid composition of a system at a given time. Introducing stable isotopes into the studied system can additionally provide information on the synthesis, transformation and degradation of individual lipid species. However, capturing the entire dynamics of lipid networks is still a challenge. We developed and evaluated a novel strategy for the indepth analysis of the dynamics of lipid metabolism with the capacity for high molecular specificity and network coverage. The general workflow consists of stable isotope-labeling experiments, ultrahigh-performance liquid chromatography (UHPLC)/high-resolution Orbitrap-mass spectrometry (MS) lipid profiling and data processing by a software tool for global isotopomer filtering and matching. As a proof of concept, this approach was applied to the network-wide mapping of dynamic lipid metabolism in primary human skeletal muscle cells cultured for 4, 12, and 24 h with [U−13C]-palmitate. In the myocellular lipid extracts, 692 isotopomers were detected that could be assigned to 203 labeled lipid species spanning 12 lipid (sub)classes. Interestingly, some lipid classes showed high turnover rates but stable total amounts while the amount of others increased in the course of palmitate treatment. The novel strategy presented here has the potential to open new detailed insights into the dynamics of lipid metabolism that may lead to a better understanding of physiological mechanisms and metabolic perturbations.

L

metabolism reflected by manifold alteration in the synthesis, transformation, and degradation of individual lipid species. Progress to capture these dynamic processes in lipid metabolism has been made by the introduction of stable isotope-labeled precursors in the experimental settings.7 Initially, gas chromatography/isotope ratio mass spectrometry (GC/IRMS) or GC/MS-based mass isotopomer distribution (MID) analysis8 were applied to study isotope incorporation into or degradation from selected lipid pools. However, those

ipids are a vital class of metabolites. In mammals, they serve not only as energy stores and important components of cellular membrane structure but also as key regulators in various signaling processes.1,2 Consequently, dysregulation of lipid metabolism is associated with a variety of perturbations on the cellular level and also with the onset and progression of diseases. Investigations of pathophysiological mechanisms of lipid metabolism as well as screening for disease-associated lipid biomarkers are more and more performed by lipidomics analyses.2−5 The promising perspective of lipidomics as a useful tool in basic and applied medical research is highlighted by applications in diabetes research, cancer, neurodegenerative diseases, cystic fibrosis, and other respiratory diseases.5,6 Up to now, however, most lipidomics analyses are focused on “snapshots” rather than capturing the dynamics of lipid © 2013 American Chemical Society

Received: January 29, 2013 Accepted: March 28, 2013 Published: March 28, 2013 4651

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657

Analytical Chemistry

Article

trypsinization. Cell pellets from one cell culture dish (15 cm diameter) were resuspended in 200 μL of PBS and sonicated. Lipids were extracted with methyl tert-butyl ether as described previously.20 The lipid extract was freeze-dried followed by reconstitution in ACN/IPA/H2O (65:30:5) for UHPLC/ Orbitrap-MS analysis. Profiling and Identification of Lipid Species by UHPLC/Orbitrap-MS. UHPLC/Orbitrap-MS analysis of intracellular lipid profiles was performed according to a method established by Hu et al.21 with slight modifications. Briefly, for LC separation a Thermo Accela UHPLC system (Thermo Fisher, San Jose, CA) and a C8 AQUITY column (2.1 mm × 100 mm × 1.7 μm, Waters, Milford, MA) were utilized. The mobile phases were A (ACN/H2O = 60:40, 10 mM AmFm) and B (IPA/ACN = 9:1,10 mM AmFm). The gradient elution started with 32% B for 1.5 min followed by a linear increase to 97% B from 1.5 to 21 min, which was kept for 4 min and then equilibrated at 32% B. The UHPLC system was coupled with a LTQ-Orbitrap XL mass spectrometer (Thermo Fisher, San Jose, CA). The analytes were ionized in the positive mode in an electrospray source with source voltage of 4.5 kV, sheath gas flow of 35 arb, and auxiliary gas flow of 5 arb. Capillary temperature was set at 325 °C. Before sample analysis, external mass calibration was applied to ensure high-resolution and accurate mass measurement. Quality control (QC) samples22 were prepared from pooled samples by mixing 10 μL from each sample and repeatedly analyzed during the whole run. The reproducibility and robustness of the applied approach were evaluated by investigation of QC and biological replicate samples. Details are given in Figures S1-A and S1-B for QC samples and Figures S1-C and S1-D for biological replicate samples (Supporting Information). Lipid identification was based on a combination of database query using exact mass measurement (mass error < 5 ppm) and MS/MS pattern. The databases used were HMDB (http:// www.hmdb.ca/),23 Metlin (http://metlin.scripps.edu/index. php),24 and LIPID MAPS (http://www.lipidmaps.org/).25 Except for sphingolipids, collision-induced dissociation (CID) fragmentation was performed on a LTQ-Orbitrap-MS with collision energy of 35 V and Activation Q of 0.18. CID of sphingolipids was performed on a triple-quadrupole LC/MS system (Agilent 6460, Agilent, Santa Clara, CA) due to abundance losses of their molecular ions on ion-trap instruments.26 Isotope labeling in conjunction with MS/MS also allows the determination of the labeling position, as shown in Figure S2 of the Supporting Information using ceramides as an example. Data Processing and Statistics. All UHPLC/OrbitrapMS raw data files were time-aligned and framed using the software SIEVE (V1.2, Thermo Fisher, San Jose, CA). A m/z width of 0.005 and a retention time width of 0.5 min were defined as frame settings. After applying the 80% rule,27 the obtained peak list with tR, m/z, and intensity was subjected to isotopomer filtering using our in-house-developed software “filter 1” to screen all ion pairs that showed a m/z difference of 16.0536 (= 16 × mass of [13C−12C]) ± 0.001 and a tR drift tolerance within ±0.05 min. These filtered ion doublets (M, M + 16) or triplets (M, M + 16, M + 32) or quadruplets (M, M + 16, M + 32, M + 48) were defined as potential isotopomers. Aiming to achieve an extensive isotope labeling pattern, the peak list was evaluated further by the software “filter 2” to detect other possible forms originating from metabolized [U−13C]-palmitate (M + 12, M + 14, M + 18, M + 20, M +

approaches are not capable to provide dynamic information on the intact lipid molecules. Liquid chromatography (LC)/MS approaches substantially broadened the analytical range, enabling the identification of isotope incorporation into intact lipids by MS/MS.7 Recent publications show that there is a huge potential for using LC/ MS and stable isotopes to investigate diet- or drug-associated alterations of lipid biosynthesis. Isotope incorporation could be measured into various lipid classes such as cholesterol esters, phospholipids, and triacylglycerols,9−14 into sphingolipids12,15 and acyl-CoA16 or by following free fatty acid carbon-chain elongation and degradation,17 thus significantly increasing the current knowledge about lipid metabolism. So far, however, there have been no attempts to our knowledge to generate a comprehensive pattern covering as many stable isotope-labeled lipid species as possible. Such a strategy would be the prerequisite to identify unexpected alterations in lipid metabolism. Given the large number of lipid species, such a nontargeted strategy must rely on a sophisticated data evaluation approach. To gain even broader insights into the complex dynamics of lipid metabolism in cells, body fluids, or tissues, we here introduce a novel approach that combines stable isotopeassisted ultrahigh-performance liquid chromatography (UHPLC)/Orbitrap-MS lipidomics with data evaluation by a software tool to extract as many isotopomers as possible. As a proof of concept, we applied this approach to the investigation of lipid extracts from primary human skeletal muscle cells exposed for 4, 12, and 24 h to 250 μM [U−13C]-palmitate in comparison to control cells not treated with palmitate.



EXPERIMENTAL SECTION Materials and Chemicals. HPLC-grade methanol (MeOH) and isopropanol (IPA) were purchased from TEDIA (Fairfield, OH), MS-grade acetonitrile (ACN) was purchased from Merck (Darmstadt, Germany), ammonium formate (AmFm, purity >99.0%), amphotericin B, [U−13C]palmitate, and fatty acid free bovine serum albumin (10% solution, A1595) were purchased from Sigma−Aldrich (St. Louis, MO). Ultrapure water was obtained by a Milli-Q system (Millipore, Billerica, MA). Lipid standards were purchased from Avanti Polar Lipids (Alabaster, AL). Cell culture media and supplements were from Lonza (Basel, Switzerland). Cell Culture Experiments and Sample Preparation. Primary skeletal muscle cells were grown from satellite cells obtained from percutaneous needle biopsies performed on the lateral portion of the quadriceps femoris (vastus lateralis) muscle of a healthy subject. The protocol was approved by the Ethics Committee of the University of Tuebingen, and the subject gave informed written consent to the study. Experiments were performed with first-passage subcultured cells. Cells were grown and differentiated as described elsewhere.18 On the seventh day of myoblast-to-myotube fusion, treatment of the cells with 250 μM [U−13C]-palmitate for 4, 12, and 24 h was performed in Eagle’s modified essential medium (EMEM) containing 5.5 mM glucose with 2% FBS, 2 mM glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin. Bovine serum albumin (BSA)-bound [U−13C]-palmitate was prepared as described recently (molar ratio palmitate/BSA 4:1, end concentration of BSA 0.4%).19 Controls were treated in the same way but incubated solely with BSA for the same periods of time. Lysates were generated after washing the cells with phosphate-buffered saline (PBS) containing 0.5% BSA and 4652

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657

Analytical Chemistry

Article

28, M + 30, M + 34, M + 36, M + 44, M + 46, M + 50, M + 52) with a mass error tolerance of ±0.0015 and a tR drift tolerance of 0.1 min. All filtered isotopomers belonging to one lipid species were then matched by the software. In addition, natural isotope correction was performed. Further details about the inhouse-developed software (“filter 1” and “filter 2”) are given in the Supporting Information. Principal component analysis (PCA) with autoscaling (UV) or pareto scaling28 was performed with SIMCA-P 11.5 (Umetrics AB, Umeå, Sweden) after normalizing to the area of all features. UV and pareto scaling were performed by dividing variables by their standard deviation or square root of standard deviation after centering. Statistical analysis was performed by SPSS 13.0 (SPSS inc., Chicago, IL). Statistical significance between different time points in the cell culture experiments was evaluated by one-way ANOVA. P < 0.05 was set as the statistic significance level. Lipid Nomenclature and Annotation. The lipid nomenclature throughout this manuscript follows the lipid classification and nomenclature system established by LIPID MAPS.29,30 Abbreviations for lipid classes are as follows: PC, phosphatidylcholine; LPC, lyso-phosphatidylcholine; PE, phosphatidylethanolamine; LPE, lyso-phosphatidylethanolamine; PS, phosphatidylserine; LPS, lyso-phosphatidylserine; PI, phosphatidylinositol; LPI, lyso-phosphatidylinositol; PG, phosphatidylglycerol; LPG, lyso-phosphatidylglycerol; PA, phosphatidic acid; LPA, lyso-phosphatidic acid; Cer, ceramide; dHCer, dihydroceramide; HexCer, hexosylceramide; SM, sphingomyelin; DG, diacylglycerol; TG, triacylglycerol; choE, cholesterol ester; CL, cardiolipin; LCL, lyso-cardiolipin. For example, TG (48:1) denotes triacylglycerol with a summed carbon number of all fatty acyl chains of 48 and one double bond. Sphingolipid species were denoted by a sphingoid base moeity and an amidelinked fatty acid moiety. Other abbreviations include the following: G-3-P, glycerol-3-phosphate; CDP-DG, cytidine-5′diphosphate 1,2-diacyl-sn-glycerol.

Figure 1. Flowchart illustrating the strategy of combining stable isotope-labeling experiments with UHPLC/Orbitrap-MS lipidomics and sophisticated data evaluation, exemplarily shown for a cell culture experiment. (a) Sample preparation and data acquisition: UHPLC/ Orbitrap-MS analysis of lipid extracts from cells not treated with palmitate (= control) and from cells treated with stable isotope-labeled substrate. (b) Mass isotopomer filtering and matching: Evaluation of the acquired data based on four rules as shown in the figure. (c) Mapping of lipid metabolism including dynamic changes: Comparison of the pattern formed by all detected lipids (“total lipid” analysis) with the percentage of 13C isotopomers within one labeled species (“labeled lipid” analysis) to dissect the dynamic changes in lipid metabolism under the studied experimental conditions after mass isotopomer distribution (MID) analysis.

typical m/z shift, LC retention time alignment, isotope pattern features, and nonexistence of the feature in the control cells not treated with [U−13C]-palmitate. The filtering and matching performed in step (b) includes three routines: (i) filtering and matching unlabeled forms and all labeled isotopomers containing one or more intact original stable isotopes (e.g., [U−13C]-palmitate); (ii) filtering and matching all metabolized labeled isotopomers, that is, labeled isotopomers containing isotopes that differ from the original stable isotope substrate (e.g., isotopomers containing labeled carbons different from 16/32/48 when [U−13C]-palmitate was used); (iii) correction for natural isotopes to ensure that the abundance that will be calculated for all labeled isotopomers in the next step originates from externally added isotopes. All filtered lipid species are identified by a combination of exact mass, diagnostic CID fragments shown in Table S1 (Supporting Information), and confirmation with the corresponding standard compound in equivocal cases. In step (c), mass isotopomer distribution (MID) analysis is performed first. Next, an in-depth picture of lipid metabolism is generated that consists of two parts: first, a lipid pattern composed of all detected lipids ([12C+13C], i.e., labeled and unlabeled species); second, the percentage of 13C isotopomers within one labeled species (13C/[12C + 13C]). Thus, the presented approach provides both information on the global cellular response (comparable to classical lipidomics approaches) and comprehensive information on dynamic changes within lipid fingerprints that reflects the rates of degradation, elongation, and transformation of lipids. Software-Based Isotopomer Filtering and Matching. Previous approaches11,12,16 employing isotope-labeled substrates have investigated defined lipid species or classes by targeted analysis of, in the case [U-13C]-palmitate, 13C16-



RESULTS AND DISCUSSION General Workflow of Stable Isotope-Assisted Lipidomics. The major goals of our study were the development and evaluation of a novel strategy for in-depth analysis of lipid metabolism. This approach covers not only the total pattern of lipids but also the dynamics within the detected lipid fingerprints by studying isotope incorporation. The general workflow combines stable isotope-labeling experiments, UHPLC/Orbitrap-MS lipid profiling, and extensive isotopomer filtering and matching. Figure 1 exemplarily shows the strategy for a cell culture experiment, but the approach can also be applied to investigate samples from in vivo studies in plants, animal models, or humans. In step (a), cell culture is performed with/without a stable isotope-labeled substrate. In the study presented in the following sections, primary human skeletal muscle cells were exemplarily treated with/without [U−13C]-palmitate. Next, lipid extraction, UHPLC/Orbitrap-MS analysis, retention time alignment, and feature extraction are performed, and the acquired raw data are transformed to a peak list containing tR, m/z, and intensity of all detected features. In step (b), the in-house-developed programs “filter 1”and “filter 2” (see the Supporting Information) are applied to obtain a complete list of all lipid isotopomers and simultaneously match the different isotopomers belonging to one lipid species. All filtering and matching actions are based on four rules: 4653

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657

Analytical Chemistry

Article

12

Figure 3. Lipid profiles of primary human myotubes treated for 4, 12, and 24 h with 250 μM [U−13C]-palmitate were investigated by PCA. Controls were cultured for the same period of time but without additional palmitate in the medium. (A) PCA score plot of the lipid profile composed of all lipids (unlabeled and 13C-labeled); (B) PCA score plot of the lipid profile composed of percentages of 13C isotopomers of labeled species.

13

Figure 2. (A) Extracted ion chromatogram of quadruplet ( C, C16-, 13 C32-, 13C48-labeled) isotopomers of TG (48:1) detected in cellular lipid extracts of primary human skeletal muscle cells treated for 12 h with [U−13C] palmitate (right panel) in comparison to control cells which were left untreated for the same period of time (left panel). (B) Corresponding mass spectrum of these four isotopomers.

correction. Although these metabolized isotopomers accounted for less than 3% of the overall abundance, they cannot generally be neglected, since some metabolized isotopomers had a signal intensity that was as high as the signals derived from isotopomers containing intact 13C16-forms. The isotopomer filtering programs were applied to a lipidomics data set of lipid extracts from primary human myotubes exposed to [U−13C]-palmitate for three periods of time (4, 12, and 24h). After peak alignment and application of the 80% rule,27 greater than 7000 features were included in the peak list. Subsequent to software filtering and natural isotope correction, 692 different isotopomers remained containing doublets (M, M + 16), triplets (M, M + 16, M + 32) and quadruplets (M, M + 16, M + 32, M + 48) as well as metabolized labeled forms. A typical example illustrating the applied filtering procedure for lipid isotopomers originating from the incorporation of intact 13C16 is shown in Figure 2. The extracted ion chromatogram shows the 12C-, 13C16-, 13C32-, and 13 C48-isotopomers of TG (48:1) (Figure 2A) and the corresponding mass spectrum with characteristic isotopic patterns (Figure 2B). The four isotopomers coeluted, and the mass shift between each pair was 13C16 within an m/z error

containing lipids. However, the applied stable isotopes can also be shortened by β-oxidation, leading, for example, to 13C14myristic acid, or elongated by fatty acid elongases, resulting, for example, in 13C18-stearic acid in the example of [U−13C]palmitate. This metabolization of the initial stable isotope may result in lipid isotopomers differing from, for the example of [U-13C]-palmitate, 13C16-/13C32-/13C48-labeled forms. Thus, to reach the goal of an exhaustive screening, all possible lipid isotopomers should be considered including shortened and elongated labeled forms. Accordingly, in the case of [U−13C]palmitate, our novel algorithm considers the unlabeled forms of lipids, “M” (= mother ion), forms derived from the incorporation of intact [U−13C]-palmitate (M + 16, M + 32, and M + 48) and forms derived from the incorporation of metabolized isotope (M + 12, M + 14, M + 18, M + 20, M + 28, M + 30, M + 34, M + 36, M + 44, M + 46, M + 50, and M + 52). To automatically match and filter both types of isotopomers, two pieces of software, “filter 1” and “filter 2”, were compiled (available in the Supporting Information). Of all detected isotopomers, about 23% were metabolized species detected by the software “filter 2” followed by natural isotope 4654

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657

Analytical Chemistry

Article

Figure 4. Lipid metabolic pathways and dynamic changes of lipid classes in primary human myotubes incubated for 4, 12, and 24 h with 250 μM [U−13C]-palmitate. Classes featuring detected labeled lipids are marked in red. Bars represent the total amount of all detected lipids ([12C + 13C]). The solid line represents the fraction of labeled isotopomers (13C/[12C + 13C]). The dotted line represents the amount of the lipid classes detected in the control cells, i.e., primary human myotubes cultured for 4, 12, and 24 h without addition of palmitate. All data are shown as mean ± SD (n = 4). a, p < 0.05 for the time course of all 12C + 13C lipids in myotubes incubated with [U−13C]-palmitate; b, p < 0.05 for the time course of the labeled lipid species (13C/[12C + 13C]) in myotubes incubated with [U−13C]-palmitate; c, p < 0.05 for the time course of the total amount of lipids in the control group. Statistic significance was determined by one-way ANOVA.

Accurate matching of all homologous isotopomers in an automatic manner enables MID analysis of each individual labeled species, thus providing information concerning isotopic enrichment. For instance, after natural isotope correction, Cer(d18:1/24:0) was shown to be composed of five isotopomers: M, M + 16, M + 18, M + 32, and M + 34 that accounted for 54.2 ± 2.0%, 35.3 ± 1.5%, 4.4 ± 0.2%, 3.0 ± 0.3%, and 3.0 ± 0.1% of the total amount of this species in primary human myotubes treated with [U−13C]-palmitate for 12 h (n = 4). Investigation of Lipidomics Fingerprints of Primary Human Myotubes Treated for 4, 12, and 24 h with 250 μM [U−13C]-palmitate. The lipid fingerprints of the different time points composed of all detected lipids, unlabeled lipids, and isotopomers ([12C + 13C]) of labeled lipids were subjected to multivariate data analysis by PCA. Figure 3A shows an apparent time-dependent shift of the lipid pattern of human myotubes incubated with palmitate for 4, 12, and 24 h, indicating progressive alterations of cellular lipid species triggered by a prolonged influx and metabolization of palmitate. In contrast, extracts of untreated control cells cultivated for the same periods of time showed a clear clustering independent of the time point studied (R2X = 0.931, Q2 = 0.856). On the basis of the finding of distinct changes in the global lipid profile, we proceeded to study the labeled lipid profile, composed of the percentages of isotope-labeled palmitate (and its metabolized forms), (13C/[12C + 13C]) by PCA (Figure 3B).

tolerance of 0.001. A complete list of all detected isotopomers labeled with intact 13C 16:0, as well as further details on these lipids, is given in Table S2 (Supporting Information). In the following step, isotopomers originating from the same lipid species were matched by the software. These 692 filtered isotopomers were assembled and further assigned to 203 intact labeled lipid species spanning 12 (sub)classes (Tables S2 and S3, Supporting Information). Thus, labeled lipid species could successfully be discovered without a priori knowledge. Noteworthy, essential prerequisites for successful filtering and matching from the original peak list are high mass accuracy and reproducible chromatographic conditions. In addition to the 203 isotope-enriched lipids, we also identified 19 lipids that were still unlabeled after 24 h exposure of the cells to [U−13C]-palmitate (Table S2, Supporting Information). The unlabeled lipids were mainly cholesterol esters (choE) and phospholipids with fatty acyl chains consisting of essential fatty acids such as linoleic acid. These fatty acids cannot be produced by de novo biosynthesis from palmitate. In sum, 222 different lipid species (203 labeled and 19 unlabeled) spanning a wide range of lipid classes were detected in our approach. Thus, by filtering all possible labeled forms, an almost comprehensive profile of labeling patterns can be achieved, which goes beyond the information reached by current stable isotope assisted lipidomics approaches. 4655

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657

Analytical Chemistry

Article

lipid metabolism leading to a better understanding of metabolic perturbations.

These isotope profiles also showed a time-course dependency, with distinct differences between 4, 12, and 24 h (R2X = 0.974, Q2 = 0.96). In-Depth Mapping of Lipid Metabolism and Dynamic Changes in Primary Human Myotubes Exposed to 250 μM [U−13C]-palmitate. Palmitate is often applied in cell culture studies to simulate the chronic overload with fatty acids found in metabolic diseases such as obesity and diabetes.31 However, the lipids involved in palmitate-induced insulin resistance and in particular the dynamics of palmitate incorporation into lipids are only partially understood.32 A more detailed lipidomics view on the effects of palmitate exposure and metabolization including the dynamics may disclose new mechanisms. Figure 4 shows for each lipid class an overlay of the summed lipid patterns ([12C + 13C]) with the natural abundances (as found in the control samples at each time point) and the labeled fractions (13C/[12C + 13C]) of these detected lipid classes. Significant incorporation of [U−13C]-palmitate over time was found for DG, TG, PI, dHCer, Cer, HexCer, SM, PC, PE, PS, LPC, and LPE. Interestingly, some lipid classes such as PC and TG showed no or only moderate increases in the total amount ([12C + 13 C]) despite a pronounced increase in the percentage of labeled PC and TG (70−80%) that indicates a high turnover rate. It has been reported that palmitate is a poor substrate for TG synthesis and that this is one link to palmitate-induced apoptosis.33 Here, we could show that palmitate is incorporated into TG without subsequent increase of TG that explains the former assumption that palmitate could not be efficiently stored in TG. In the lipid classes PI, dHCer, PS, LPC, and LPE, a marked synthesis resulting in more than 2-fold increases in the total amount between 4 and 24 h was detected. Of note, we observed only a minor increase in the total amount of Cer and DG that are discussed as inducers of insulin resistance in cells exposed to an excess of palmitate.34,35 This in-depth analysis of the metabolic fate of palmitate in various lipid species may also contribute to the identification of new candidate mediators of the lipoapoptotic effects of an excess flux of fatty acids. A challenging analytical perspective for the future would be to extend the covered lipid profile to include the acyl-CoA pool, the key node of lipid synthesis and fatty acid oxidation.



ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*(R.L.) Phone: +49 7071 29 83193; fax: +49 7071 29 5348; email: [email protected]. (G.X.) Phone/ fax: +86-411-84379559; e-mail: [email protected]. Author Contributions #

Jia Li and Miriam Hoene contributed equally to this study.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by grants of the 973 Project of the State Ministry of Science and Technology of China (no. 2012CB517506 to G.X.), the Sino-German Center for Research Promotion (GZ 753 by DFG and NSFC to G.X. and R.L. and LE 1391/1-1 by DFG to R.L.), the German Research foundation (GRK 1302/2 to C.W.), the “Stiftung fü r Pathobiochemie und Molekulare Diagnostik” of the German Society of Clinical Chemistry and Laboratory Medicine (to M.H. and R.L.), the German Federal Ministry of Education and Research (BMBF) to the German Centre for Diabetes Research (DZD e. V.), the Kompetenznetz Diabetes mellitus (Competence Network for Diabetes mellitus) funded by the Federal Ministry of Education and Research (FkZ 01 GI 1104A to R.L. and H.U.H.), and the key foundation and creative research group project (nos. 21175132 and 21021004 by NSFC to G.X.). We gratefully acknowledge the technical assistance of Heike Runge.



REFERENCES

(1) Wymann, M. P.; Schneiter, R. Nat. Rev. Mol. Cell Biol. 2008, 9, 162−176. (2) Shevchenko, A.; Simons, K. Nat. Rev. Mol. Cell Biol. 2010, 11, 593−598. (3) Han, X.; Gross, R. W. J. Lipid Res. 2003, 44, 1071−1079. (4) Wenk, M. R. Nat. Rev. Drug Discovery 2005, 4, 594−610. (5) Hu, C.; van der Heijden, R.; Wang, M.; van der Greef, J.; Hankemeier, T.; Xu, G. J. Chromatogr., B 2009, 877, 2836−2846. (6) Postle, A. D. Curr. Opin. Clin. Nutr. Metab. Care 2012, 15, 127− 133. (7) Postle, A. D.; Hunt, A. N. J. Chromatogr., B 2009, 877, 2716− 2721. (8) Hellerstein, M. K.; Neese, R. A. Am. J. Physiol. 1999, 276, E1146− 1170. (9) Castro-Perez, J. M.; Roddy, T. P.; Shah, V.; McLaren, D. G.; Wang, S.-P.; Jensen, K.; Vreeken, R. J.; Hankemeier, T.; Johns, D. G.; Previs, S. F.; Hubbard, B. K. J. Proteome Res. 2011, 10, 4281−4290. (10) Castro-Perez, J.; Previs, S. F.; McLaren, D. G.; Shah, V.; Herath, K.; Bhat, G.; Johns, D. G.; Wang, S.-P.; Mitnaul, L.; Jensen, K.; Vreeken, R.; Hankemeier, T.; Roddy, T. P.; Hubbard, B. K. J. Lipid Res. 2010, 52, 159−169. (11) McLaren, D. G.; He, T.; Wang, S.-P.; Mendoza, V.; Rosa, R.; Gagen, K.; Bhat, G.; Herath, K.; Miller, P. L.; Stribling, S.; Taggart, A.; Imbriglio, J.; Liu, J.; Chen, D.; Pinto, S.; Balkovec, J. M.; DeVita, R. J.; Marsh, D. J.; Castro-Perez, J. M.; Strack, A.; Johns, D. G.; Previs, S. F.; Hubbard, B. K.; Roddy, T. P. J. Lipid Res. 2011, 52, 1150−1161.



CONCLUSIONS The introduction of a stable isotope-labeled substrate/ precursor pool into experimental settings adds another dimension to an in-depth understanding and exploration of dynamic biological process. However, it also complicates the analysis of raw data acquired by MS-based measurement and hinders effective and comprehensive extraction of useful information. An improved analytical capacity is needed to enable screening of as many isotope-labeled lipid metabolites as possible and quantifying their flux from a labeled precursor pool. Our novel approach, combining stable isotope-assisted lipidomics and software-based isotopomer filtering and matching tools, could expand the analytical capability in this respect. The application of this approach to study primary human myotubes exposed to [U−13C]-palmitate along a time course showed its capability to generate additional information on the metabolic fate of fatty acids and candidate mediators of fatty acid induced metabolic disorders. This strategy may open new perspectives for the in-depth dissection of alterations in 4656

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657

Analytical Chemistry

Article

(12) Blachnio-Zabielska, A. U.; Persson, X.-M. T.; Koutsari, C.; Zabielski, P.; Jensen, M. D. Rapid Commun. Mass Spectrom. 2012, 26, 1134−1140. (13) Enjalbal, C.; Roggero, R.; Cerdan, R.; Martinez, J.; Vial, H.; Aubagnac, J.-L. Anal. Chem. 2004, 76, 4515−4521. (14) Bleijerveld, O. B.; Houweling, M.; Thomas, M. J.; Cui, Z. Anal. Biochem. 2006, 352, 1−14. (15) Haynes, C. A.; Allegood, J. C.; Wang, E. W.; Kelly, S. L.; Sullards, M. C.; Merrill, A. H. J. Lipid Res. 2011, 52, 1583−1594. (16) Blachnio-Zabielska, A. U.; Koutsari, C.; Jensen, M. D. Rapid Commun. Mass Spectrom. 2011, 25, 2223−2230. (17) Kamphorst, J. J.; Fan, J.; Lu, W.; White, E.; Rabinowitz, J. D. Anal. Chem. 2011, 83, 9114−9122. (18) Weigert, C.; Brodbeck, K.; Staiger, H.; Kausch, C.; Machicao, F.; Haring, H. U.; Schleicher, E. D. J. Biol. Chem. 2004, 279, 23942− 23952. (19) Peter, A.; Weigert, C.; Staiger, H.; Machicao, F.; Schick, F.; Machann, J.; Stefan, N.; Thamer, C.; Häring, H.-U.; Schleicher, E. Diabetes 2009, 58, 1757−1765. (20) Matyash, V.; Liebisch, G.; Kurzchalia, T. V.; Shevchenko, A.; Schwudke, D. J. Lipid Res. 2008, 49, 1137−1146. (21) Hu, C.; van Dommelen, J.; van der Heijden, R.; Spijksma, G.; Reijmers, T. H.; Wang, M.; Slee, E.; Lu, X.; Xu, G.; van der Greef, J.; Hankemeier, T. J. Proteome Res. 2008, 7, 4982−4991. (22) Gika, H. G.; Theodoridis, G. A.; Wingate, J. E.; Wilson, I. D. J. Proteome Res. 2007, 6, 3291−3303. (23) Wishart, D. S.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A. C.; Young, N.; Cheng, D.; Jewell, K.; Arndt, D.; Sawhney, S.; Fung, C.; Nikolai, L.; Lewis, M.; Coutouly, M.-A.; Forsythe, I.; Tang, P.; Shrivastava, S.; Jeroncic, K.; Stothard, P.; Amegbey, G.; Block, D.; Hau, D. D.; Wagner, J.; Miniaci, J.; Clements, M.; Gebremedhin, M.; Guo, N.; Zhang, Y.; Duggan, G. E.; Macinnis, G. D.; Weljie, A. M.; Dowlatabadi, R.; Bamforth, F.; Clive, D.; Greiner, R.; Li, L.; Marrie, T.; Sykes, B. D.; Vogel, H. J.; Querengesser, L. Nucleic Acids Res. 2007, 35, D521−526. (24) Smith, C. A.; O’Maille, G.; Want, E. J.; Qin, C.; Trauger, S. A.; Brandon, T. R.; Custodio, D. E.; Abagyan, R.; Siuzdak, G. Ther. Drug Monit. 2005, 27, 747−751. (25) Sud, M.; Fahy, E.; Cotter, D.; Brown, A.; Dennis, E. A.; Glass, C. K.; Merrill, A. H.; Murphy, R. C.; Raetz, C. R. H.; Russell, D. W.; Subramaniam, S. Nucleic Acids Res. 2007, 35, D527−D532. (26) Schuhmann, K.; Herzog, R.; Schwudke, D.; Metelmann-Strupat, W.; Bornstein, S. R.; Shevchenko, A. Anal. Chem. 2011, 83, 5480− 5487. (27) Bijlsma, S.; Bobeldijk, I.; Verheij, E. R.; Ramaker, R.; Kochhar, S.; Macdonald, I. A.; van Ommen, B.; Smilde, A. K. Anal. Chem. 2006, 78, 567−574. (28) van den Berg, R. A.; Hoefsloot, H. C.; Westerhuis, J. A.; Smilde, A. K.; van der Werf, M. BMC Genomics 2006, 7, 142. (29) Fahy, E.; Subramaniam, S.; Brown, H. A.; Glass, C. K.; Merrill, A. H., Jr.; Murphy, R. C.; Raetz, C. R. H.; Russell, D. W.; Seyama, Y.; Shaw, W.; Shimizu, T.; Spener, F.; van Meer, G.; VanNieuwenhze, M. S.; White, S. H.; Witztum, J. L.; Dennis, E. A. J. Lipid Res. 2005, 46, 839−861. (30) Fahy, E.; Subramaniam, S.; Murphy, R. C.; Nishijima, M.; Raetz, C. R. H.; Shimizu, T.; Spener, F.; van Meer, G.; Wakelam, M. J. O.; Dennis, E. A. J. Lipid Res. 2008, 50, S9−S14. (31) Li, L. O.; Klett, E. L.; Coleman, R. A. Biochim. Biophys. Acta 2010, 1801, 246−251. (32) Muoio, D. M.; Neufer, P. D. Cell Metab. 2012, 15, 595−605. (33) Listenberger, L. L.; Han, X.; Lewis, S. E.; Cases, S.; Farese, R. V.; Ory, D. S.; Schaffer, J. E. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 3077− 3082. (34) Amati, F.; Dubé, J. J.; Alvarez-Carnero, E.; Edreira, M. M.; Chomentowski, P.; Coen, P. M.; Switzer, G. E.; Bickel, P. E.; Stefanovic-Racic, M.; Toledo, F. G. S.; Goodpaster, B. H. Diabetes 2011, 60, 2588−2597. (35) Turpin, S. M.; Lancaster, G. I.; Darby, I.; Febbraio, M. A.; Watt, M. J. Am. J. Physiol.: Endocrinol. Metab. 2006, 291, E1341−1350. 4657

dx.doi.org/10.1021/ac400293y | Anal. Chem. 2013, 85, 4651−4657