Article pubs.acs.org/jpr
Senescing Human Bone-Marrow-Derived Clonal Mesenchymal Stem Cells Have Altered Lysophospholipid Composition and Functionality Seul Ji Lee,† TacGhee Yi,‡,§,∥ Soo Hyun Ahn,⊥ Dong Kyu Lim,† Ji Yeon Hong,† Yun Kyoung Cho,∥ Johan Lim,⊥ Sun U. Song,*,‡,∥ and Sung Won Kwon*,† †
College of Pharmacy, Seoul National University, 599 Gwanangno, Gwanak-gu, Seoul 151742, Korea Translational Research Center and §Inha Research Institute for Medical Sciences, Inha University School of Medicine, Jungsuck B/D 7-241, 3rd Street Shinheung-Dong, Choong-Gu, Incheon 400-103, Korea ∥ HomeoTherapy, Co. Ltd., 3ga, Sinheung-dong, Incheon 400-711, Korea ⊥ Department of Statistics, Seoul National University, Seoul, Korea ‡
S Supporting Information *
ABSTRACT: Mesenchymal stem cells (MSCs) have been used in a wide range of research and clinical studies because MSCs do not have any ethical issues and have the advantage of low carcinogenicity due to their limited proliferation. However, because only a small number of MSCs can be obtained from the bone marrow, ex vivo amplification is inevitably required. For that reason, this study was conducted to acquire the metabolic information to examine and control the changes in the activities and differentiation potency of MSCs during the ex vivo culture process. Endogenous metabolites of human bone-marrow-derived clonal MSCs (hcMSCs) during cellular senescence were profiled by ultraperformance liquid chromatography/quadrupole time-of-flight mass spectrometry (UPLC/ QTOFMS). To select significant metabolites, we used the linear mixed effects model having fixed effects for batch and time (passage) and random effects for metabolites, determining the mean using a t test and the standard deviation using an F test. We used structural analysis with representative standards and spectrum patterns with different collision energies to distinctly identify eight metabolites with altered expression during senescence as types of lysophosphatidylcholine (LPC) and lysophosphatidylethanolamine (LPE), such as LPC 16:0 and LPE 22:4. The present study revealed changes in endogenous metabolites and mechanisms due to senescence. KEYWORDS: human clonal mesenchymal stem cells, cellular senescence, lysophosphatidylcholine, lysophosphatidylethanolamine
■
for cell-based regenerative therapy.2,3 One of the hurdles in stem cell research is securing a large number of high-quality stem cells for therapeutic uses.4 Because it is very difficult to produce culture-expanded stem cells without alterations in the stem-cell properties, we developed and established a methodology, subfractionation culturing method (SCM), to efficiently separate and culture highly homogeneous human clonal MSCs (hcMSCs).5,6 Using this method, our group was able to secure homogeneous cells and manage the quality of the cells. As follow-up research in high-quality stem-cell production, we conducted this study to control the progression of aging during the ex vivo expansion of MSCs. Aging is an active and universal phenomenon that appears in all living organisms.7 Until now, various studies and hypotheses such as somatic mutation, mitochondrial, genetic dysregulation, gerontogene, wear and tear, and free radical theories have been proposed.8,9 Nonetheless, those theories cannot explain the overall aspects of aging. Further investigations are needed to
INTRODUCTION Stem cells have attracted attention for tissue repair and regeneration due to their multipotency to differentiate into various cell types. Over the past few decades, remarkable advances have been achieved in research and the therapeutic application of three major stem cell types, including embryonic stem cells, adult or somatic stem cells, and induced pluripotent stem cells. Among these, adult stem cells such as hematopoietic stem cells and mesenchymal stem cells (MSCs) are present in small amounts in various tissues of the body, and they serve as building blocks for specific tissues. When cells or tissues are damaged, adult stem cells are activated to generate tissuespecific cells that can repair the injured tissues. Adult stem cells migrating from distant organs such as bone marrow can also be involved in tissue repair.1 Unlike embryonic stem cells, adult stem cells such as MSCs have limited proliferation activity, which may be beneficial for clinical applications due to a very low risk of tumorigenicity. Furthermore, the use of adult stem cells may be relatively free of ethical debates. Current advancements in clinical application have enabled adult stem cells to become a promising resource © 2014 American Chemical Society
Received: October 1, 2013 Published: February 6, 2014 1438
dx.doi.org/10.1021/pr400990k | J. Proteome Res. 2014, 13, 1438−1449
Journal of Proteome Research
Article
passage to minimize the environmental influence on the culture. The hcMSC samples at passages 7, 15, and 20 were independently prepared in triplicate for metabolomics analysis. The samples were frozen immediately and stored at −80 °C until analysis.
understand the exact molecular mechanisms underlying aging and aging-related phenomena. Aging can be defined as “the sum of primary restrictions in regenerative mechanisms of multicellular organisms”.10 These restrictions can be induced by senescence, which is the concept of aging with a strong emphasis on the cellular level.11 On the basis of the definition quoted by Campisi,12 the meaning of cellular senescence is equal to the replicative senescence,13 which stands for a state of irreversibly arrested cell division and growth. Cellular senescence is a complex phenotype that is caused by alterations in function and replicative capacity. This phenomenon is regulated by DNA, protein, and metabolites. Therefore, through the examination of the endogenous expression of cellular products, it is possible to identify the mechanism of senescence and to understand the mechanism of aging phenomena in complex organisms. This approach may be more simple and efficient than the direct aging induction model in the complex organism.14 Our work focused on analyzing endogenous metabolite during the cellular senescence of hcMSCs to describe the aging phenomena,15 although these metabolites are positioned in the lowest subnetwork of cellular expression hierarchy and thus have the most complex network. We isolated hcMSCs from a single cell-derived single colony and subcultured them repeatedly to reach senescence to observe the change in endogenous metabolite production before and after cellular senescence. This approach enabled us to measure the metabolite changes that were only due to the aging progression by excluding the effects of other external factors. To interpret the data reliably, optimal statistical algorithms for experimental models were created. Metabolites whose expression was altered during the cellular senescence of hcMSCs were selected and analyzed to reveal the underlying mechanism. In this study, we aimed to provide significant molecular markers associated with cellular senescence to monitor the ex vivo culturing process of MSCs. This would complement existing biomolecular techniques and provide a new path to MS-based automated methods for an effective large-scale production of MSCs. In addition, suggested candidate molecules can be used as quality-control indicators for quantitative measurement of senescence to enable high-quality stem cell production that is appropriate for clinical uses and related research. Lastly, it will be possible to enhance the understanding of the mechanism of senescence through metabolite markers that are directly related to mechanism.
■
hcMSC Characterization
For MSC identification, cell surface marker expression and differentiation potential were evaluated. The expression of cell surface marker in hcMSC lines was analyzed by flow cytometry at passage 6. The antibodies used were anti-CD14, anti-CD34, anti-CD44, anti-CD73, anti-CD90, anti-CD105, anti-HLA class I, anti-HLA-DR, and anti-PODXL antibodies (BD Biosciences Pharmingen, San Diego, CA). The stem-cell marker expression was analyzed in a FACSCalibur flow cytometer (BD Biosciences). Isotype-matched control antibodies were used as controls. To determine the differentiation potential of hcMSCs, the differentiation of the hcMSCs into three mesenchymal cell types was induced. hcMSCs were differentiated into adipocytes by seeding in a four-well plate at a density of 6 × 104 cells/well. The following day, the subconfluent cells were incubated in an adipogenic medium composed of DMEM with high glucose, 10% calf serum, 10−7 M dexamethasone (DEX; Sigma, St. Louis, MO), 10 μg/mL insulin, 0.5 mM 1-methyl-3isobutylxanthine (IBMX; Sigma), and 50 μg/mL indomethacine (Sigma). The cells were differentiated into adipocytes for 4 to 5 days. The cells were fixed with 4% formaldehyde and then stained with oil red O for 30 min, followed by counterstaining with hematoxylin for 10 min. For osteogenic induction, hcMSCs were seeded in a four-well plate at a density of 6 × 104 cells/well and cultured in an osteogenic medium (α-MEM containing 10% FBS, 50 μg/mL ascorbic acid (Sigma), 10−8 M DEX, and 10 mM β-glycerophosphate). The osteogenic medium was changed every 3 days for 3 weeks. The cells were fixed using 4% paraformaldehyde and stained with alizarin red S. A pellet culture system was used for chondrogenic differentiation. Then, 2.0 × 105 hcMSCs were put in a 15 mL conical tube and pelleted by spin-down. The pellet was cultured in 500 μL of serum-free chondrogenic medium (α-MEM supplemented with 10 ng/mL TGF-β1 (R&D Systems, Minneapolis, MN), 10 ng/mL TGF-β3 (R&D Systems), and 1% insulin-transferrin-selenious acid premix (BD Biosciences). The chondrogenic medium was changed every 3 days for 3 weeks. The cell pellet was embedded in an OCT compound (Sakura Finetek, Torrance, CA), frozen, sectioned into 8 mm slices, and stained with safranin O.
MATERIAL AND METHODS
hcMSC Isolation and Induction of Cellular Senescence
β-Gal Staining
hcMSCs were isolated from bone marrow (BM) aspirates using a subfractionation culturing method.5 BM aspiration from a healthy male donor was performed after obtaining informed consent (approved by the Inha University School of Medicine Institutional Review Board under IRB Registration Number 1051). The isolation and culture of BM-derived hcMSCs was performed as described previously.5 hcMSCs were cultured in Dulbecco’s modified essential medium (DMEM) (Gibco BRL, Life Technologies, Rockville, MD) containing 10% heatinactivated fetal bovine serum (FBS; Gibco BRL) and 1% anti-anti (Gibco BRL) without additional supplements in a 5% CO2 incubator. Cellular senescence was induced by continuous subcultivation up to passage 20. Cell detachment and replating into new culture plates were performed in 1 day after each
The expression of pH-dependent senescence-associated βgalactosidase (SA-β-gal) activity was analyzed simultaneously in different passages of hcMSCs using the SA-β-gal staining kit (Cell Biolabs, San Diego, CA) according to the manufacturer’s instruction. In brief, the cells were fixed in 4% paraformaldehyde for 5 min, followed by thorough washing with phosphatebuffered saline (PBS). The fixed cells were incubated at 37 °C overnight after the addition of freshly prepared cell staining working solution containing potassium ferrocyanide, potassium ferricyanide, citrate-Na2HPO4 buffer (pH 6), 50 mM MgCl2, NaCl, and X-gal. The next day, the cell-staining solution was discarded, and the cells were washed twice with PBS. Bluecolored cells were considered senescent by observation with a light microscope. 1439
dx.doi.org/10.1021/pr400990k | J. Proteome Res. 2014, 13, 1438−1449
Journal of Proteome Research
Article
Immunofluorescence Staining and Immunoblot Analysis
performed independently. The experimental process was adapted from Yanes et al.16
For confocal microscopy, hcMSCs were fixed with 4% paraformaldehyde, permeabilized with 0.1% triton X-100 for 5 min, and plated on a glass slide. The cells were blocked with 5% normal goat serum for 30 min at room temperature, followed by anticaveolin-1 antibody (Cell Signaling Technology, Boston, MA) at 4 °C overnight. The next day, the cells were washed thoroughly with tris-buffered saline containing 0.1% tween 20 and incubated with FITC-conjugated secondary antibody (Molecular Probes, Eugene, OR) for 1 h at room temperature. For F-actin, the cells were labeled with AlexaFluor 597-conjugated phalloidin (Molecular Probes) for 30 min at room temperature. Nuclei were stained with 4′,6′-diamidino-2phenylindole (DAPI; Molecular Probes). After extensive washing, the samples were mounted with mounting solution and subjected to confocal microscopy (Olympus FluoView FV1000, Tokyo, Japan). For immunoblot analysis, whole cell lysates from hcMSC samples at different passages were extracted using a lysis buffer of 10 mM Tris-Cl (pH 7.5), 150 mM NaCl, 1 mM EDTA (pH 8.0), 1% triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM PMSF, and a protease inhibitor cocktail solution. The samples with equal amount of total protein were subjected to SDS-PAGE and transferred to a PVDF membrane. The membrane blot was blocked with 5% skim milk and incubated with each antibody. Chemical luminescence was detected in an image analyzer LAS4000 mini (Fuji PhotoFilm, Tokyo, Japan).
Ultraperformance Liquid Chromatography/Quadrupole Time-of-Flight Mass Spectrometry (UPLC/QTOFMS)
Analysis was performed using a UPLC system (Waters) equipped with a Acquity UPLC column (BEH C18, 1.7 μm, 2.1 mm × 100 mm) and a micrOTOF-QII mass spectrometer (Bruker Daltonics, Germany). Each sample (2 μL) was injected and separated by a gradient method using Solution A (water + 0.1% formic acid) and Solution B (acetonitrile + 0.1% formic acid) as follows: 0 min 100% A, 5 min 70% A, 15 min 30% A, 25 min 20% A, ∼27 min 0% A. The flow rate was set to 0.2 mL/min, and 10 min of prerun was processed among the sample injection to ensure column equilibration. Column temperature was maintained at 40 °C using the column oven. Mass spectrometry was performed using the ESI positive ion mode, which was included in micrOTOF-QII mass spectrometer. Mass range was set to 50−1000 m/z. Capillary voltage was 4.5 kV, and nebulizer pressure was 1.2 bar. Dry gas flow rate was 8.0 L/min with 200 °C of dry temperature. Furthermore, to optimize the quadrupole and collision cell’s parameters, the quadrupole ion energy was set to 15.0 eV, collision energy to 10.0 eV, transfer time to 55.0 μs, and prepulse storage to 2.0 μs. High-purity nitrogen gas was applied as the collision gas, drying gas, and nebulizer gas. To maintain the accuracy of the m/z value, lithium formate was used as the lock mass and was injected before each sample injection. A random sequence was applied to eliminate external factors that occurred during analysis. For quality control of data, the mixture of 18 samples was used as a quality-control sample and analyzed at regular intervals in the middle of the sequence. The relative standard deviation percent (RSD%) was calculated for each variable.17,18
Reverse-Transcription Polymerase Chain Reaction (RT-PCR)
For mRNA expression evaluation, semiquantitative RT-PCR analysis was conducted in a range of linear amplification. Total RNA was extracted from each sample using easy-Blue RNA extraction reagent (Intron Biotechnology, Sungnam, Korea). cDNA was synthesized from 1 mg of total RNA using the AccuPower cDNA synthesis kit (Bioneer, Daejeon, Korea). PCR was performed using AccuPower PCR premix (Bioneer) in a C1000 thermal cycler (BioRad, Hercules, CA). Amplified products were separated on a 1% agarose gel containing SyberSafe (Invitrogen, Carlsbad, CA) and then detected in LAS4000 mini. The PCR primer sequences used in this study are shown in Supplementary Table 1 in the Supporting Information.
Data Preprocessing
The raw file was analyzed using MZmine version 2.5.19 Mass detection, chromatogram builder, and peak deconvolution facilitated peak detection. Through the mass detection process, each MS spectrum was converted into a pair of m/z and intensity values. (The noise level was set at 103 using a centroid algorithm.) Using the chromatogram builder function, chromatogram construction was performed using the following parameter values: the minimum height was 1.5 × 103, and the m/z tolerance was 0.05. For the peak deconvolution, peak recognition was performed using the “Savitzky-Golay” function (minimum peak height at 1.5 × 103, derivative threshold level at 80%). Peak alignment was performed using the “RANdom SAmple Consensus (RANSAC) aligner” method (m/z tolerance at 0.05, absolute retention time (RT) tolerance at 0.1 min, RANSAC iterations at 10 000). Principal component analysis (PCA) was performed using the calibrated peak data, and the list of resulting peaks was transferred to a .csv file.
Metabolite Extraction
Initially, around three million cells were kept frozen at −80 °C, and cold acetone (600 μL) was added to cell pellets to extract metabolites. Then, the samples were mixed thoroughly for 1 min and soaked for 1 min in liquid nitrogen. The samples were defrosted for 2 to 3 min and sonicated for 10 min. This procedure was repeated three times before the samples were put at −20 °C for 1 h. Next, the pellet was centrifuged at 13 000 rpm for 10 min. The upper layer was transferred to a clean tube; then, a cold methanol/water/formic acid (86.5:12.5:1.0) mixture (400 μL) was added to the pellet. The samples were mixed thoroughly for 1 min, sonicated for 10 min, and stored at −20 °C for 1 h. The pellet was extracted by centrifugation at 13 000 rpm for 10 min, and the upper layer was combined with the first extraction in a clean tube. Samples were then dehydrated by SpeedVac. After mixing by vortex and sonication, 95% acetonitrile/water solution (100 μL) was added. The samples were centrifuged at 13 000 rpm, and the upper layer was transferred to a clean LC−MS vial. In the autosampler, the samples were stored at 4 °C and analyzed by LC−MS. Three duplicates of cell cultures (n = 3) were
Statistical Analysis
The statistical analysis of this study was performed in two steps. First, we tested whether there was a significant change over time in either the population mean or the population standard deviation of metabolite intensity. To test both the mean and the standard deviation, we used the linear mixed effects model with fixed effects for batch and time (passage) and random effects for metabolites. To test the change in the standard deviation, we used the log of the smoothed variance to estimate both the response of the model and the smoothed variance, as 1440
dx.doi.org/10.1021/pr400990k | J. Proteome Res. 2014, 13, 1438−1449
Journal of Proteome Research
Article
Figure 1. Characterization of hcMSCs isolated from human BM. (A) Cell morphology of hcMSCs. (B) Flow cytometric analysis of MSC markers. (C) Multilineage potential of hcMSCs; adipogenic, osteogenic, and chondrogenic differentiation. Magnification, 40× and 100×.
proposed by Cui et al.20 This method resolved the difficulties arising from the small sample size as well as the frequently occurring zero intensities. Second, we identified metabolites that changed in intensity over time (passage) by implementing a large number of t tests to determine the mean value and F tests to determine the standard deviation.
at different collision energies of 10−50 or 5−25 eV were confirmed. This result was validated by comparison with the results of the targets in the sample. The result was combined with the m/z information of precursor ion, and final identification was completed by searching the LIPID MAPS (http://www.lipidmaps.org) database.
■
Full-Scan MS/MS Acquisition of LPCs and LPEs Standards for Qualitative Annotation
RESULTS
The data were collected using an UPLC system (Waters) with a micrOTOF-QII mass spectrometer (Bruker Daltonics) in positive ion mode. 1-Palmitoyl (C16:0) lysophosphatidylcholine (LPC), 1-stearoyl (C18:0) LPC, 1-palmitoyl (C16:0) lysophosphatidylethanolamine (LPE), and 1-stearoyl (C18:0) LPE were purchased from Advanti Polar Lipids (Alabaster, AL). All lipids were dissolved in chloroform/methanol/water (65:35:8). Each standard was diluted in methanol (1→10) prior to injection. The precursor ions [M+H]+ were selected as LPC 16:0−496.3397, LPC 18:0- 524.3710, LPE 16:0− 454.2928, and LPE 18:0−482.3241. The collision energies (10, 20, 30, 40, and 50 eV) were gradually tested on LPC. Unlike LPC, relatively weak collision energies (5, 10, 15, 20, 25 eV) were tested on LPE.21,22 Isolation width was set to 10.0. All MS and MS/MS base peak chromatograms (BPCs), spectra, and m/z error were calibrated to
