Aldehyde Dehydrogenase 1, a Potential Marker for Cancer Stem Cells in Human Sarcoma

Tumors contain a small population of cancer stem cells (CSC) proposed to be responsible for tumor maintenance and relapse. Aldehyde dehydrogenase 1 (ALDH1) activity has been used as a functional stem cell marker to isolate CSCs in different cancer types. This study used the Aldefluor® assay and fluorescence-activated cell sorting (FACS) analysis to isolate ALDH1high cells from five human sarcoma cell lines and one primary chordoma cell line. ALDH1high cells range from 0.3% (MUG-Chor1) to 4.1% (SW-1353) of gated cells. Immunohistochemical staining, analysis of the clone formation efficiency, and xCELLigence microelectronic sensor technology revealed that ALDH1high cells from all sarcoma cell lines have an increased proliferation rate compared to ALDH1low cells. By investigating of important regulators of stem cell biology, real-time RT-PCR data showed an increased expression of c-Myc, β-catenin, and SOX-2 in the ALDH1high population and a significant higher level of ABCG2. Statistical analysis of data demonstrated that ALDH1high cells of SW-982 and SW-1353 showed higher resistance to commonly used chemotherapeutic agents like doxorubicin, epirubicin, and cisplatin than ALDH1low cells. This study demonstrates that in different sarcoma cell lines, high ALDH1 activity can be used to identify a subpopulation of cells characterized by a significantly higher proliferation rate, increased colony forming, increased expression of ABC transporter genes and stemness markers compared to control cells. In addition, enhanced drug resistance was demonstrated.


Introduction
The cell population of most tumors is heterogeneous with regard to its proliferation capacity and the ability to initiate tumor formation in immune-deficient mice. A cancer stem cell (CSC) is defined as a cell within a tumour that possesses the capacity to selfrenew and to generate the heterogeneous lineages of cancer cells that comprise the tumor [1,2]. Numerous investigations have provided evidence that CSCs exist in a variety of human tumors such as hematopoietic malignancies, brain tumors, breast cancer, and gastroenterological cancer [3,4,5,6].
Cytosolic aldehyde dehydrogenases (ALDHs) are a group of enzymes involved in oxidizing a wide variety of intracellular aldehydes into their corresponding carboxylic acids [7]. Among theses enzymes, ALDH1 is throught to have an important role in oxidation of alcohol and vitamin A and in cyclophosphamide chemoresistance. Ginestier et al. [8] showed that ALDH1 was a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome of breast cancer patients. High ALDH1 activity has been used to define stem cell populations in many cancer types including human multiple myeloma, acute myeloid leukemia [8], pancreatic cancer [9], and breast cancer [10]. Therefore, ALDH1 activity might be usable as a common marker for malignant stem cell populations [11].
Failure of cancer chemotherapy can occur through increased efflux of chemotherapeutic agents, leading to the reduction of intracellular drug levels and consequent drug insensitivity. ABC transporters have the capacity to export many cytotoxic drugs and recent evidence suggests that the cancer stem cell phenotype is associated with high-level expression of the ABCG2 transporter [12,13,14].
In this study, we used the AldefluorH assay and fluorescenceactivated cell sorting (FACS) analysis to isolate ALDH1 high cells from five human sarcoma cell lines and one recently established chordoma cell line. We analyzed ALDH1 high cells in vitro for their repopulation capacity, clonogenicity, cell proliferation properties, the expression of stem cell markers and ABC transporters, and their multidrug resistance capacities.

AldefluorH Assay and Separation of the ALDH1 + Cell Population by FACS Analysis
Aldehyde dehydrogenase (ALDH) enzyme activity in viable cells was determined using a fluorogenic dye based AldefluorH assay (Stem Cell Technologies, Grenoble, France) according to the manufacturer's instructions. 1610 6 /ml cells were suspended in AldefluorH assay buffer containing ALDH substrate (Bodipy-Aminoacetaldehyde) and incubated for 45 min at 37uC. As a reference control, the cells were suspended in buffer containing AldefluorH substrate in the presence of diethylaminobenzaldehyde (DEAB), a specific ALDH1 enzyme inhibitor. The brightly fluorescent ALDH1-expressing cells (ALDH1 high ) were detected in the green fluorescence channel (520-540 nm) of FACSAria (BD Biosciences, San Diego, CA) and the data was analyzed using FACS DIVA software (BD Biosciences). To exclude nonviable cells propidium iodide (PI; Sigma Aldrich, Vienna, Austria) was added at a final concentration of 2 mg/ml.

Repopulation Assay
To compare the repopulation ability of sarcoma ALDH1 high cells with ALDH1 low cells in vitro, freshly sorted cells were cultured separately under the same culture condition. After 2 weeks, cells were re-stained with the AldefluorH assay and reanalyzed via FACSAria (BD Biosciences).

Western Blot Analysis
For total protein analysis, cells were re-suspended in lysis buffer (50 mM Tris-HCL pH 7.4, 150 mM NaCl, 50 mM NaF, 1 mM EDTA, 10% NP-40, 1% Triton-X and protease inhibitors), incubated on ice for 10 min and centrifuged at 15,000 rpm for 15 min. Aliquots of protein extracts (20 mg) were separated on 12% SDS-PAGE and electro-blotted onto 0.45 mm Hybond ECL nitrocellulose membrane (Amersham Biosciences, Little Chalfont, UK). The membrane was blocked with 3% milk blocking buffer for 1 h and then incubated with the primary antibodies for 2 h at room temperature. As the primary antibody, rabbit polyclonal ALDH1/2 antibody (#sc50385; Santa Cruz Biotechnology, Santa Cruz, CA) was used. The major liver isoform ALDH1 localized to cytosolic space, while ALDH2 localized to the mitochondria. The blots were developed using horseradish peroxidase-conjugated secondary antibodies (Dako, Vienna, Austria) at room temperature for 1 h and the SuperSignalH West Pico Chemoluminescent Substrate (Thermo Scientific, Rockford, IL), in accordance with the manufacturers' protocol.

Immunohistochemistry
Each 1610 4 ALDH high and ALDH low cells were seeded in polystyrene culture slides (BD Biosciences), fixed with 4% formalin/PBS solution, and dehydrated in an ascending series of alcohol. Immunohistochemical (IHC) studies using the streptavidin-biotin peroxidase complex method were carried out employing antibody against the anti-Ki-67 (clone 30-9) rabbit monoclonal primary antibody (Ventana Medical Systems, Tucson, AZ) using the BenchMark Ultra instrument (Ventana Medical Systems). Cells were imaged using an Olympus BX51 microscope with Olympus DP71 microscope digital camera. The stained slides were digitally scanned and positive and negative cells were quantified using the ImageScope software (ImageScope Virtual Slide, version 6.25, Aperio Technol.,Vista, CA). The positivity = N positive cells/N total cells. The xCELLigence DP device from Roche Diagnostics (Mannheim, Germany) can be used to quantitatively and dynamically monitor cell proliferation in real-time [15]. Respectively 1610 4 freshly sorted ALDH1 high and ALDH1 low cells were seeded in electronic microtiter plates (E-Plate TM ; Roche Diagnostic) and measured for 72 h with the xCELLigence system according to the instructions in the user's manual. Application of a low-voltage (less

Colony Formation Assay
To determine the clone formation efficiency (CFE) of sorted cells in vitro, ALDH1 high , ALDH1 low cells and unstained cells (control) were counted and 200 cells per well were seeded in six well plates. Triplicate wells were used for each group. Cells were cultured in DMEM-F12 with supplements for 14 days, fixed in methanol for 10 min and stained with crystal violet (Sigma Aldrich, Hamburg, Germany). The clone's number which consisted of more than 50 cells was counted. The CFE was calculated according to the formula: (the clone number/the plated cell number)6100.

Real-Time RT-PCR
Real-time RT-PCR was performed according to MIQE criteria [16] to determine the relative expression of the ABC transporter genes ABCG2/BCRP1, ABCA2, and ABCB1/MDR1 and the stemness markers c-Myc, b-catenin, and SOX-2. Total RNA was isolated with RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturers' recommended protocol. DNA was digested with 1 U DNase (Fermentas, St.Leon-Rot, Germany) per mg RNA. 1 mg RNA was reverse transcripted using RevertAid cDNA Synthesis Kit (Fermentas). Real-time PCR reactions were performed in triplicates using the Platinum SYBR Green Super Mix with ROX (Invitrogen) on AB7900HT (Applied Biosystems, Invitrogen). The housekeeping genes glyceraldehyde 3-phosphate dehydrogenase (GAPDH), b-actin (ACTB) and hypoxanthine phosphoribosyltransferase (hprt-n) served as an internal control due to their stable expression in different tissues. Table 1 lists the primers used for real-time PCR. The expression levels were calculated based on the 2 2DDCT method [17].

Drug Sensitivity Assay
Sorted cells were adjusted to a density of 5610 3 cells/100 ml and incubated in 96-well microplates. The cells were exposed to

Statistical Analysis
The outcome variables were expressed as mean 6SD. Student's unpaired t-test and the exact Wilcoxon's test was used to evaluate differences between groups with the PASW statistics 18 software (IBM Corporation, Somers, NY). Two-sided P-values below 0.05 were considered statistically significant. IC 50 curves were fitted according the Hill equation (sigmoid, 3 parameters). Graphic data were prepared with SigmaPlotH (Systat Software Inc., San Jose, CA).

Sarcoma Cell Lines Display a Distinctive Fraction of ALDH1 high Cells
The AldefluorH assay system has been developed to detect the activity of the ALDH1 isoform. We used this assay followed by FACS analysis to assess the presence and quantity of ALDH1 high cell populations in five human sarcoma cell lines and one chordoma cell line [18].
To set a marker for ALDH1 high cells DEAB control cells was used to ensure the accuracy of the analysis. The representative SW-1353 cells were treated in the presence of the ALDH1 inhibitor DEAB ( Figure 1A) or stained with AldefluorH reagent, which are defined as ALDH1 high cells ( Figure 1B showed only a small proportion of 1.1160.5% ALDH1 high cells (n = 9) ( Figure 1C). We therefore focused on the three sarcoma cell lines SW-684, SW-982, and SW-1353 in the following experiments. In the repopulation assay the ALDH1 high population generated a statistic significant higher account of 41.73618.5% (*p = 0.0039) ALDH1 high cells in the SW-684 cells, 52.568.75% (*p = 4.39E-07) in the SW-982 cell line, and 31.7611.1% (*p = 0.0025) (n = 5) in SW-1353 chondrosarcoma cells ( Figure 1D). Additional our observations of enhanced ALDH1 expression could be further be substantiated by western blot analysis ( Figure 1E; Table 2).

ALDH1 high Cells Show Higher Proliferation and Clonogenicity
Using the ImageScope software Ki-67 positive and negative cells were quantified after immunohistochemical staining. ALDH1 high cells from all three cell lines have an increased proliferation level compared to ALDH1 low cells. Representative staining of SW-982 ALDH1 high (Figure 2A) and ALDH1 low cells ( Figure 2B) are shown and summarized in Table 2 (n = 5). Furthermore, ALDH1 high and ALDH1 low cells differed significantly in logarithmic growth velocity measured with the xCELLigence-System ( Figure 2C2E). Figure 3A shows the clonogenic activity of ALDH1 high and ALDH1 low cells. Data from five independent experiments represent average colony count/well after 14 days and all values are listed in Table 2. The clone formation efficiency was significantly higher in SW-1353 ALDH1 high cells compared to corresponding ALDH1 low cells (*p = 0.0005). For the other two cell lines these effect could also be demonstrated, however in a smaller extent. The higher number of colonies in the SW-684, SW-982, and SW-1353 ALDH high cells is presented ( Figure 3B).
The mRNA Expression of ABCG2, c-Myc, b-catenin, and SOX-2 are Upregulated in ALDH1 high Cells We investigated whether ALDH1 high cells are enriched for expression of genes that have been postulated to play key roles in stem cell biology, such as c-Myc, b-catenin, and SOX-2 [19]. Quantitative RT-PCR showed increased expression of c-Myc in the ALDH1 high population, while unsorted control cells (ctrl) and ALDH1 low cells had only minimal expression ( Figure 4A). Similarly, a slight but not significant increase in the expression of b-catenin, and SOX-2 in the ALDH1 high fraction could be observed (n = 6) ( Figure 4B-C).
The relative expression of the three major drug transporters ABCG2/BCRP1, ABCA2, and ABCB1/MDR1 was determined by real-time RT-PCR (n = 5). Interestingly the ALDH1 high population of all sarcoma cell lines demonstrated, with statistic significance, increased expression levels of ABCG2 compared to control or ALDH1 low cells ( Figure 4D), whereas the p value for ABCA2 was not significant ( Figure 4E). In addition, in ALDH1 high SW-1353 cells a statistic significant higher expression of ABCB1 (p = 0.0302) could be observed ( Figure 4F). The 2 2DDCT values and the corresponding p-values are listed in Table 2.

ALDH1 high Cells Show Enhanced Drug Resistance
The cancer stem cell hypothesis proposes that the discrepancy between treatment response and patient survival noted in most cancer types reflects an inherent resistance of the cancer stem cells to chemotherapy. To investigate possible differences in drug resistance ALDH1 high and ALDH1 low sorted SW-982 and SW-1353 cells were treated with increasing doses of three commonly used chemotherapeutic agents after a two weeks recovery phase. ALDH1 high SW-982 cells treated for 48 h with 1.0 mM (p = 0.016) and 5.0 mM (p = 0.001) doxorubicin were significantly increased compared with ALDH1 low cells ( Figure 5A). Treatment with 1.0 mM epirubicin (p = 0.045) induced an enhanced drug resis-   Table 2.

Discussion
Based on the current cancer stem cell (CSC) hypothesis, only a small subpopulation within the heterogeneous tumor population is capable of initiating and re-initiating tumors. The concept of CSCs was based on the observation that when cancer cells of many different types were assayed for their proliferative potential in various assays in vitro and in vivo, only a minority of cells showed extensive proliferation [20]. CSCs have been identified in a variety of malignancies [21,22,23]. One widly accepted method for identifying CSCs is based on the enzymatic activity of aldehyde dehydrogenase 1 (ALDH1), a detoxifying enzyme responsible for the oxidation of intracellular aldehydes [8,21]. There are different isoforms of ALDH. The AldefluorH assay system has been developed to detect the activity of the ALDH1 isoform. ALDH1 activity showed to be increased in CSCs and has been used to isolate CSCs in different cancers [24,25,26]. Therefore, ALDH1high cells display several features typically seen in CSCs, including the ability for self-renewal, generation of differentiated progeny, and increased expression of stem cell marker genes. The study of CSC biology is predicated on the ability to accurately assess CSC representation within cancer cell populations. As suggested by more recent findings CSC representation may be a function of the cell type of origin, stromal microenvironment, accumulated somatic mutations and stage of malignant progression reached by a tumor [27,28].
To date, the existence of such a stem-like cell population in human osteosarcoma and Ewing's sarcoma cell lines has been based on the expression of stem cell marker genes as well as their ability to form spheroids in vitro [29,30,31]. It has been suggested that identification of the CSC cannot solely rely on side population (SP) sorting using efflux of Hoechst 33342 dye. However the SP phenotype is not presented in all CSCs and there may exist other defensive mechanisms for CSCs to evade drug therapies that cannot be identified by Hoechst dye staining [32]. Therefore, we chose the marker ALDH1. Our results show that all five sarcoma cell lines contained different percentage of ALDH1 high cells, with the highest percentage in fibrosarcoma, synovial sarcoma, and chondrosarcoma cell lines. The small ALDH1 expression of the ALDH low cells in the western blot analysis can be explained by the use of the ALDH1/2 primary antibody. The proliferation rate and clonogenicity of SW-684, SW-982, and SW-1353 ALDH1 high cells in vitro were significantly higher than that of ALDH1 low cells, consistent with the characteristics of the high ALDH1 activity phenotype in other cancer cells [33,34], which may indicate that ALDH1 high cells from sarcoma are partially responsible for tumor metastasis and recurrence and should be focused during the cancer therapy. As c-Myc has been recently recognized as an important regulator of stem cell biology, it may serve as a link connecting malignancy and ''stemness'' [35]. Introduction of c-Myc with other transcription factors (including SOX-2) generates induced pluripotent stem cells from differentiated cells [36]. Wnt/b-catenin signaling plays an important role not only in cancer, but also in cancer stem cells [37]. Our quantitative RT-PCR data showed increased expression of c-Myc, b-catenin, and SOX-2 in the ALDH1 high population, while unsorted control cells (ctrl) and ALDH1 low cells had only minimal expression.
A proposed mechanism of chemotherapy resistance of cancer stem cells is based on the enhanced expression of ATP-binding cassette (ABC) transport proteins, which are responsible for drug efflux. Higher expression of ABC transport proteins in stem cells compared to non-stem cells results in relative resistance of the stem cells to the toxic effects of chemotherapy drugs [12,13]. We analysed the mRNA expression of three major drug transporters (ABCG2/BCRP1, ABCA2, ABCB1/MDR1) of ABC transporter family. In the present study, ABCG2 was upregulated in ALDH1 high cells from all three sarcoma cell lines. Furthermore, another ABC transporter ABCB1/MDR1 was also found with higher mRNA expression level in SW-1353 ALDH1 high cells compared to ALDH1 low cells. These genes may be responsible for multi-drug resistance of cancer cells and should be ideal targets for clinical cancer therapy.
Additional, ALDH1 high cells showed increased resistance to commonly used chemotherapeutic drugs. ALDH1 high cells of SW-982 and SW-1353 showed significantly lower sensitivity to both doxorubicin and epirubicin compared with ALDH1 low cells. The cisplatin treatment showed only slight differences. Together, we successfully isolated ALDH1 high cells from different sarcoma cell lines using the AldefluorH assay. ALDH1 high cells exhibited in vitro a significant higher proliferation rate, increased clone formation efficiency, elevated expression of ABC transporters and stemness marker, as well as increased chemotherapeutic drug resistance compared to ALDH1 low cells.
In conclusion, the presence of stem-like cells with increased ALDH1 expression could be one of the possible contributors to the development of drug resistance in sarcomas. Further study will be required to define the sarcoma stem cells and the mechanisms of drug resistance, but ALDH1 high population may serve as an in vitro model to search for new therapeutic treatment options.