Aug 31, 2010 - A Simple Method to Sort ESC-Derived Adipocytes ... Email: a.navarrete-santos@medizin. ..... M: 100-bp marker; ntc: no template control.
Technical Note
A Simple Method to Sort ESC-Derived Adipocytes Kristina Schaedlich,1 Julia M. Knelangen,1Alexander Navarrete Santos,2 Bernd Fischer,1 Anne Navarrete Santos1*
Additional Supporting Information may be found in the online version of this article. 1
Department of Anatomy and Cell Biology, Martin Luther University, Faculty of Medicine, Halle, Germany
2
Department of Cardiothoracic Surgery, Martin Luther University, Faculty of Medicine, Halle, Germany
Received 10 March 2010; Revision Received 15 June 2010; Accepted 2 July 2010 Kristina Schaedlich and Julia M. Knelangen contributed equally to this work. Grant sponsor: European Community’s Seventh Framework Program; Grant number: 212885; Grant sponsor: German Research Council; Grant number: DFG NA418; Grant sponsor: German National Academic Foundation. *Correspondence to: Dr. Anne Navarrete Santos, Martin Luther University, Faculty of Medicine, Department of Anatomy and Cell Biology, Grosse Steinstrasse 52, D-06097 Halle (Saale), Germany Email: a.navarrete-santos@medizin. uni-halle.de Published online 31 August 2010 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cyto.a.20953 © 2010 International Society for Advancement of Cytometry
Cytometry Part A 77A: 990 995, 2010
Abstract Because of the increasing incidence of worldwide obesity, cell culture models which enable the study of adipose tissue development are of particular importance. The murine embryonic stem cell (ESC) line CGR8 differentiates into adipocytes with a differentiation efficiency of up to 15%. A critical step for the analysis of stem cell-derived adipogenesis is the reliable separation of adipocytes. Here we report on how to (i) gently separate the cells of embryoid bodies (EBs) and (ii) identify and sort adipocytes from the rest of the heterogeneous cell mixture. Up to the present, no adipocyte specific surface marker is known for fluorescence activated cell sorting (FACS). After separation we employed two independently existing FACS methods for adipocyte cell sorting. These methods are based on Nile red staining and granularity. For stem cell-derived adipocytes only the combination of both methods led to a reliable, efficient, and highly reproducible FACS analysis, as shown by the presence and absence of adipocyte specific markers in positively and negatively sorted cells. ' 2010 International Society for Advancement of Cytometry
Key terms adipogenic differentiation; CGR8 cells; flow cytometry
THE pluripotent murine embryonic stem cell (ESC) line CGR8 has the potential to differentiate into various derivates of the three germ layers, including the differentiation into adipocytes in vitro. An early treatment with retinoic acid and subsequent culture with insulin and triiodthyronine result in functional adipocytes within 21 days (1). This protocol recapitulates the adipogenic differentiation during embryonic development and offers the opportunity to investigate the effects of endogenous and exogenous factors on adipogenesis. Although ESC can be differentiated into many cell types, receiving pure populations of a specific cell type is still a challenge. With the protocol mentioned above the differentiation efficiency is in the range of 15%, i.e., resulting in a heterogeneous cell population including numerous other cell types like mesenchymal myocytes and neuronal cells. The aim of this study was to develop a method to prepare a proper single-cell suspension and to sort efficiently mature adipocytes for further analysis. Fluorescence activated cell sorting (FACS) is a well established method for the analysis and sorting of cells. A common tool for selective cell sorting is the labeling of specific cell surface antigens by fluorescence-dye conjugated antibodies. Even though there are many known adipocyte-specific proteins, adipocyte-specific surface markers which could be used for FACS analysis via labeled antibodies are not available. This is not only the case for mature adipocytes but also for their progenitors. Zimmerlin et al. (2) solved this problem by using at least an eight-color flow cytometric analysis to resolve four histological discrete subpopulations of the adipose stromal vascular populations. To use eight-colors in FACS is a highly sophisticated approach, requiring a high technical standard. We decided to employ granularity and lipid staining, which have been demonstrated before to be adipocyte specific parameters (3,4). In 3T3-L1 cells Lee et al. used granularity, an indicator for fat storage, to collect mature adipocytes via flow cytometric side-scatter (3). Gimble et al. have used the ability of
TECHNICAL NOTE the lipophilic dye Nile red to pass the cell membrane and to accumulate in fat vesicles to specifically pick adipocytes from the bone marrow stromal cell line BMS2 (4). Adipocytes differentiated from preadipocytes are rich in cell numbers and have a differentiating efficiency of 70% (5). Under these circumstances granularity or Nile red accumulation are elevated and sufficient for FACS analysis. However, in the case of embryonic stem cells, a more compelling FACS method is required due to the low differentiation efficiency and high number of other cell types. Therefore we combined the two parameters and analyzed the expression of target genes in sorted adipocytes. High levels of adipocyte specific genes demonstrated the adipogenic character of the sorted cells and underline the reliability of our strategy for the isolation of embryonic stem cell-derived adipocytes.
MATERIALS AND METHODS Cell Culture and Reagents The murine embryonic stem cell line CGR8 was a gift from Prof. Anna Wobus, IPK Gatersleben, Germany, originating from the laboratory of Prof. Austin Smith (6). CGR8 cells were cultured on 0.1% gelatine-coated tissue culture dishes in Glasgow MEM BHK 21 medium (Invitrogen, Germany) supplemented with 10% heat inactivated foetal calf serum (selected batches, Gibco, Germany), 2 mM L-glutamine (Invitrogen, Germany), 100 mM b-mercaptoethanol (Serva, Germany), nonessential amino acids (stock solution diluted 1:100; Invitrogen, Germany), 1 mM sodium-pyruvate (Invitrogen, Germany), streptomycin/penicillin (5,000 U ml21 penicillin, 5,000 lg ml21 streptomycin; Invitrogen, Germany), and 10 ng ml21 leukaemia inhibitory factor (LIF) to maintain the pluripotent undifferentiated state. The differentiation medium was comprised of Dulbecos MEM (Invitrogen, Germany), 20% heat inactivated foetal calf serum, 2 mM L-glutamine, 100 mM b-mercaptoethanol, essential amino acids (stock solution diluted 1:50; Invitrogen, Germany), 1 mM sodium-pyruvate, and streptomycin/penicillin (5,000 U ml21 penicillin, 5,000 lg ml21 streptomycin).
Differentiation of CGR8 Cells into Adipocytes In Vitro The adipogenic differentiation of the embryonic stem cells CGR8 was performed in a hanging culture as described previously by Dani et al. (1). Drops of differentiation medium (20 ll) with 1,000 cells were placed onto the lids of Petri dishes filled with PBS and cultured for 2 days. The CGR8 cells aggregated and formed so-called embryoid bodies (EBs). The EBs were transferred into nonadhesive bacteriological grade Petri dishes and maintained for 3 days in suspension in differentiation medium containing 0.1 lM retinoic acid, which was replaced every day, followed by 2 days without retinoic acid supplementation. The 7-day-old EBs were placed onto 0.1% gelatine-coated dishes and cultured for another 14 days in differentiation medium supplemented with 85 nM insulin and 2 nM triiodothyronine. Medium was replaced every 2 days and adipocytes appeared at day 21. To prove successful adipogenic Cytometry Part A 77A: 990 995, 2010
differentiation EBs were stained with Red Oil O (0.5% in isopropanol; Sigma–Aldrich, Germany). Preparation of a Single-Cell Suspension from EBs For the separation of the EBs to single-cell suspension the Gentle MACS (gentle MACS dissociator, order No: 130-093235) and the gentleMACS-m-neural-tissue kit (Neural Tissue Dissociation Kit (P), order No: 130-092-628) have been used. A preheated papain enzyme mix provided with the kit was given directly to the cell culture dishes containing the differentiated EBs. The EBs were carefully mechanically removed and transferred to a provided tube (C tubes, order No: 130-093237). All further steps were performed according to the manufacture’s protocol, except for the temperature. All incubation steps were performed at room temperature. To remove cell aggregates from the cell suspension a preseparation filter with a pore diameter of 40 lm (BD biosciences, USA) was used. After a washing step, the cells were resuspended in 0.753 phosphate buffered saline (PBS), stained with Trypan blue (Biochrom, Germany) to examine the viability, and counted. Until FACS analysis the cell suspension was kept on ice.
FLOW CYTOMETRY Nile Red Staining Nile red stock solution (100 lg ml21, Sigma–Aldrich, Germany) was prepared in DMSO and stored in the dark (7). Following trypsination of the undifferentiated CGR8 cells or separation of the EBs to a single-cell suspension the cells were once washed with ice-cold 0.753 PBS. After that the dye was added at a final concentration of 0.1 lg ml21. The cells were then incubated on ice for 5 min, centrifuged, and washed once with 0.753 PBS. Afterward they were resuspended in an appropriate volume of 0.753 PBS and kept on ice prior to flow cytometric analysis. Instrumentation, Fluorescence Activated Cell Sorting, and Data Analysis All measurements were performed on a BD FACS Vantage (BD Biosciences) equipped with three lasers (excitation wavelengths: 488, 633, and 351 nm). Details about the instrument configuration according to the MIFlowCyt guideline are shown in Table 1S and 2S in the Supporting Information. The Nile red dye was excited by the 488-nm laser and the signal was assured by the band pass filter BP-585/42. Cell debris, doublets, and aggregates were excluded from analysis based on a dual parameter dot plot in which the pulse ratio (signal area/ signal high; y-axis) versus signal area (x-axis) was displayed. Undifferentiated CGR8 cells served to define the threshold of the Nile red fluorescence and the scatter signals. For this purpose, scatter signals and PMT voltage of the Nile red signal were adjusted using Nile red stained undifferentiated CGR8 cells. In a dot plot of the scatter signals and the Nile red signal (x-axis) versus the side scatter signal (y-axis) the cells were located on the lower left corner of the plot. These sample-specific instrument settings were then applied to the differentiated CGR8-derived cells. Compensation settings have not 991
TECHNICAL NOTE
Figure 1. Strategy for gating and sorting of differentiate adipocytes. Left panel: Undifferentiated CGR8 cells were analyzed on a dot plot of FSC versus SSC. The voltage of the scatter signals was reduced to values where the cell population was located on the lower left corner of the plot. A region P1 representing the undifferentiated cells (debris and cell doublets excluded) was positioned. Above region P1 a region P2 was selected. The criterion for the positioning of P2 was that no more than 0.5% of the undifferentiated cells were located in this region. Cells in the P2 region evidenced an enhanced granularity. On a dot plot of the Nile red signal versus SSC the PMT of the Nile red signal was adjusted the same way as the scatter signals (cell population located in the lower left corner of the plot). Region P3 in the plot denoted an enhanced uptake of the dye and no more than 0.5% of undifferentiated cells were present in the region. Middle panel: Under the instrument setting determined, using undifferentiated cells, differentiated CGR8 cells emerged in the dot plot FSC versus SSC as a wide cell population with about 25% of the cells located in the region P2 (increased granularity) and 30% in the region P3 (enhanced Nile red uptake). The mean of the SSC and the Nile red signals in the differentiated cell population was increased by 0.5 and 6 times, respectively in comparison to the undifferentiated cells. Right panel: For final sorting the three different populations were combined. The region P2 and P3 but not P1 (P2 1 P3 and not P1; [enhanced granularity and Nile red uptake] but not [undifferentiated cells]) was selected. The mean of the SSC and Nile red signals in the selected region increased by 2.2 and 1.35, respectively in comparison to the mean of the signals without region combination. Under the selected conditions about 20% of the total cells analyzed were sorted.
992
Selection of ESC-Derived Adipocytes by FACS
TECHNICAL NOTE Table 1. Quantification of P-regions used for sorting of adipogenic differentiated CGR8 cells REGION
P1 P2 P3 (P2 1 P3) and not P1
EVIDENCE OF
undifferentiated cells cells with increased granularity cells with enhanced Nile red uptake cells with increased granularity and Nile red uptake but not undifferentiated cells
% OF TOTAL EVENTS IN THE REGION SD
MEAN OF THE SSC SD
MEAN OF NILE RED FLUORESCENCE SD
50.60 1.8 24.70 2.2 23.50 6.1 19.07 4.7
70.2 1.3 118.7 4.3 128.6 1.5 125.4 0.58
950.2 224 1,354 117 1,409 208 1,704 67
n 5 4, SD: standard deviation.
been done. Experiment measurements and data analysis were performed with FACSDiva (Version 5.0.3; BD Biosciences). For sorting a 90-lm nozzle was used and the sorting rate was no more than 1,000 events per second. RNA Isolation and RT-PCR To collect sufficient mRNA amounts from the small number of sorted cells (\500,000 cells) the mRNA was extracted with Dynabeads1 Oligo(dT)25 (Invitrogen, Germnay) following the manufacture’s protocol. To avoid RNA degradation the lysis buffer was supplemented with RNAse Inhibitor (40 U ll21) (Promega, USA). Total RNA from higher cell numbers (500,000 cells) was isolated with the RNeasy Lipid Tissue Mini Kit from Qiagen (Qiagen, Germany). Two micrograms of total RNA were reverse transcribed in a volume of 20 ll containing 0.5 mM dNTPs, 10 mM DTT, 200 U superscript II, 20 U RNAse inhibitor, 1 ll random primer and 2 ll reverse transcriptase buffer at 428C for 1 h, followed by incubation at 908C for 5 min. As a control for DNA contamination, 2 lg RNA was PCR-amplified without reverse transcription reaction. This control reaction was performed for each primer combination and in all PCR amplifications. Amplification was carried out in a 50-ll volume containing 1 ll 10 nM dNTP, 2.5 U Taq polymerase and the specific primer combinations (adipsin fw: TGATGTGCAGAGTGTAGTGCCTCA, rv: CAAC GAGGCATTCTGGGATAGCTT; adiponectin fw: TGTTGG AATGACAGGAGCTG, rv: CGAATGGGTACATTG GGAAC and leptin fw: ATCTATGTGCACCTGAGGGTAGA, rv: TCCT TTTCACAAAGCCACACTAT). Resulting PCR products were separated by electrophoresis on a 1.8% agarose gel and stained with ethidium bromide. Agarose gels were documented by Bio-Capt MW software (LTF, Germany).
RESULTS AND DISCUSSION Cell Separation for Fluorescence Activated Cell Sorting A single-cell suspension is of high importance for flow cytometric cell sorting. We successfully employed the gentle MACS-m-neural-tissue kit (Miltenyi Biotec, Germany), based on mild enzymatic papain treatment and subsequent mechanical homogenization (Gentle MACS) for the cell separation. The Gentle MACS technique reliably separated EBs to singlecell suspensions with highly viable cells avoiding treatments with strong proteolytic enzymes like trypsin or collagenase and their high risk of cell damage. Cell separation with trypsin Cytometry Part A 77A: 990 995, 2010
or collagenase did not work well in our experiments. Because of the three dimensional structure of the embryoid bodies, outer layers were destroyed by the crude digestion conditions, while the inner layers had not even started to separate. After this procedure no FACS suitable single-cell suspension could be obtained and cell viability was not satisfying. Fluorescence Activated Cell Sorting Pluripotent CGR8 cells were differentiated to adipocytes by cell aggregation in EBs and subsequent EB outgrowth. At Day 21 20 EBs were dissociated as described above and a single cell suspension of about 5 3 106 adipogenic differentiated CGR8 cells was analyzed by flow cytometry in a one step procedure employing the parameters granularity and Nile red staining (Fig. 1). Before sorting, single cell suspensions of undifferentiated CGR8 cells and CGR8 derived differentiated adipocytes have been analyzed by flow cytometry. This approach enabled a proper selection and positioning of gates. The lowest signal channel was adjusted to the PMT of the Nile red signal. Likewise the scatter signals were adjusted. Under these conditions a region P1 in the dot plot FSC versus SSC was created and represented the undifferentiated cells. A region P2 in which cells with increased granularity must be located was positioned above the region P1. The third region P3, on a dot plot SSC versus Nile red, discriminated cells with enhanced dye uptake (Fig. 1, left). The criterion for the selection of the region P2 and P3 was that no more than 0.5% of the undifferentiated cells were located in both regions. Under these instrument settings differentiated cells showed a clear increase in cellular granularity in a dot plot forward scatter (FSC) versus sideward scatter (SSC), reflecting an increased cell diameter and granular structure (Fig. 1 middle). Furthermore, cells with enhanced granularity showed a high uptake of the Nile red dye (Fig. 1, P2 2 P3). For cell sorting a combination of the regions representing different cell populations was used. Only cells with high granularity (P2) and high Nile red uptake (P3) were sorted. At the end the combination of the region P2 1 P3 and the exclusion of P1 (undifferentiated cells) was used for sorting (Fig. 1 right). A sharp selection of the regions P2 and P3 resulted in about 19% of cells fulfilling the criteria for sorting enhanced granularity and Nile red uptake (Table 1). After sorting the obtained cells were examined by light microscopy. The positively sorted cells showed a larger cell diameter, Nile red fluorescence emission, and accumulation of multiple lipid droplets inside the cells. In contrast undifferentiated cells 993
TECHNICAL NOTE had a smaller diameter and did not show lipid droplets or Nile red fluorescence emission (Fig. 2). Characterization of the Sorted Cells RNA isolation from adipocytes provides several technical problems based on the high amount of lipids and the low RNA quantity and often yields unacceptable results when following standard protocols (8). Even though embryonic stemcell derived adipocytes possess lower intracellular lipids, a high-quality RNA extraction from the small cell number was difficult to achieve. Here we used two different strategies to obtain satisfying RNA yields of guaranteed RNA integrity from a small adipose sample (500,000 cells): RNA preparation with Dynabeads1 Oligo(dt)25 (Invitrogen, Germany) offers a fast and simple method to gain mRNA. With the RNeasy Lipid Tissue Mini Kit (Qiagen, Germany) we succeeded in total RNA isolation from sorted adipocytes. Both extraction methods gave sufficient RNA amounts of high quality so that these cells were processed for further analyses. To prove the adipogenic nature of the sorted cells the expression of the adipocyte specific genes adipsin, adiponectin, and leptin (9–11) was analyzed by polymerase chain reaction (PCR). All three adipogenic markers were expressed in the positively sorted adipocytes and undetectable or expressed at lower levels in the negatively sorted cells (Fig. 3).
CONCLUSION The variable efficiency and the heterogeneity of the resulting cell population is a general dilemma of ESC differentiation protocols. Cell characterization via PCR often leads to inconsistent replicates with high standard deviations. Development, differentiation, and the physiology of a specific cell type cannot be studied if the cell population is not pure. Many stem cell researchers face these problems and have developed various approaches to select specific cell populations from ESCs. Recently Hattori et al. developed a non genetic method to purify stem-cell derived cardiomyocytes via mitochondrial
Figure 3. RT-PCR analysis for adipogenic marker genes in sorted cells. A: adipsin 94bp; B: adiponectin 147bp; C: leptin 151bp. 1: adipose tissue as positive PCR control; 2: nonselected cells; 3: positively sorted cells with increased granularity and uptake of Nile Red. M: 100-bp marker; ntc: no template control.
staining and FACS analysis (12). With our method it is now possible to separate a low quantity of adipocytes (500,000 cells) from a small number of cells (5 3 106 cells), which can be characterized and used for a broad range of in vitro studies. The method described here represents a solid and simple option for the investigation of vital adipocytes without interference by other ESC-derived cells.
ACKNOWLEDGMENTS The authors thank Prof. Andreas Simm for allocating the FACS Vantage (BD Biosciences, USA) and Sabine Schroetter for the excellent technical assistance.
LITERATURE CITED
Figure 2. Light microscopy of sorted cells (magnification 363). A 1 B: The positively sorted cells (A) are, in contrast to the nonselected cells (B), bigger and have an accumulation of lipid droplets, demonstrating the adipogenic cell type C 1 D: Nile red fluorescence emission of the positively sorted cells (C) and nonselected cells (D). The scale bars represent 40 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
994
1. Dani C, Smith A, Dessolin S, Leroy P, Staccini L, Villageois P, Darimont C, Ailhaud G. Differentiation of embryonic stem cells into adipocytes in vitro. J Cell Sci 1997;110:1279–1285. 2. Zimmerlin LDV, Pfeifer ME, Meyer EM, Pe´ault B, Rubin JP, Donnenberg AD. Stromal vascular progenitors in adult human adipose tissue. Cytometry A 2010;77A:406. 3. Lee Y-H, Chen S-Y, Wiesner RJ, Huang Y-F. Simple flow cytometric method used to assess lipid accumulation in fat cells. J Lipid Res 2004;45:1162–1167. 4. Gimble JM, Morgan C, Kelly K, Wu X, Dandapani V, Wang C-S, Rosen V. Bone morphogenetic proteins inhibit adipocyte differentiation by bone marrow stromal cells. J Cell Biochem 1995;58:393–402. 5. Mehra A, Macdonald I, Pillay TS. Variability in 3T3-L1 adipocyte differentiation depending on cell culture dish. Anal Biochem 2007;362:281–283. 6. Mountford P, Zevnik B, Duwel A, Nichols J, Li M, Dani C, Robertson M, Chambers I, Smith A. Dicistronic targeting constructs: Reporters and modifiers of mammalian gene expression. Proc Natl Acad Sci USA 1994;91:4303–4307. 7. Greenspan P, Mayer E, Fowler S. Nile red: A selective fluorescent stain for intracellular lipid droplets. J Cell Biol 1985;100:965–973.
Selection of ESC-Derived Adipocytes by FACS
TECHNICAL NOTE 8. Janke J, Engeli S, Gorzelniak K, Sharma A. Extraction of total RNA from adipocytes. Horm Metab Res 2001;33:213–215. 9. White RT, Damm D, Hancock N, Rosen BS, Lowell BB, Usher P, Flier JS, Spiegelman BM. Human adipsin is identical to complement factor D and is expressed at high levels in adipose tissue. J Biol Chem 1992;267:9210–9213. 10. Maeda K, Okubo K, Shimomura I, Funahashi T, Matsuzawa Y, Matsubara K. cDNA cloning and expression of a novel adipose specific collagen-like factor, apM1 (dioseost Abundant Gene Transcript 1). Biochem Biophys Res Commun 1996;221:286–289.
Cytometry Part A 77A: 990 995, 2010
11. Schulz TJ, Tseng Y-H. Emerging role of bone morphogenetic proteins in adipogenesis and energy metabolism. Cytokine Growth Factor Rev 2009;20:523– 531. 12. Hattori F, Chen H, Yamashita H, Tohyama S, Satoh Y-S, Yuasa S, Li W, Yamakawa H, Tanaka T, Onitsuka T, Shimoji K, Ohno Y, Egashira T, Kaneda R, Murata M, Hidaka K, Morisaki T, Sasaki E, Suzuki T, Sano M, Makino S, Oikawa S, Fukuda K. Nongenetic method for purifying stem cell-derived cardiomyocytes. Nat Methods 2010; 7:61–66.
995
