SEARCH

SEARCH BY CITATION

Keywords:

  • flow cytometry;
  • side population;
  • CD133;
  • cancer stem cell;
  • medulloblastoma;
  • tumour sphere

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Side population (SP) analyses and CD133 expression have identified cells with stem-like potential in normal and cancerous tissue. Whether stem-like cells exist in cancer cell lines is hotly debated. We have interrogated the DAOY medulloblastoma cell line with respect to stem-like potential. Vital staining for Hoechst 33342 efflux capacity and CD133 immunophenotyping were performed on DAOY cells to assess the presence of the SP and the CD133 stem cell markers, respectively. SP/non-SP and CD133+/CD133 DAOY cells were sorted into separate fractions for limiting dilution analysis (tumor sphere assay) and asymmetric division assessment. SP/non-SP cells were also sorted separately for viability (XTT assay), cell size, cell cycle status, and proliferative capacity (carboxyfluorescein succinimidyl ester (CFSE)) evaluation. A minor proportion of cells displayed either the SP or the CD133+ phenotypes. CD133 expression mapped to both the SP and non-SP compartments, with CD133+ cells being enriched almost fourfold within the non-SP gate. The SP, non-SP, CD133+, and CD133 fractions were all capable of reconstituting the original parental DAOY population. Slight clonogenic enrichment was observed in only the SP fraction; however, both CD133+ and CD133 cells displayed equivalent stem cell-like frequencies. SP cells were resistant to Hoechst 33342-mediated toxicity relative to the parental population and differed from the non-SP cells with respect to increased cell size, decreased S-phase, and slightly decreased proliferative capacity. The multiparametric strategy described in this study revealed that the SP and CD133+ subset may be two independent compartments. Our results highlight the need for new reliable specific cancer stem cell marker(s) as Hoechst 33342 efflux and CD133 expression might not be suitable for selectively isolating cancer stem-like cells from cell lines, as shown for the DAOY cells. As such, care must be used in interpreting therapeutic studies targeting the stem cell compartment of cancer cell lines. © 2008 International Society for Advancement of Cytometry

DESPITE the observation that malignant cellular characteristics closely phenocopy the stem cell-like features of self-renewal and multilineage differentiation, the outstanding feature separating the malignant cell from the normal stem cell is the lack of homeostatic balance between self-renewal and differentiation (1). Furthermore, cancer tissue, much like normal tissue, is hierarchically organized according to stage of differentiation and proliferative potential (2). Such close phenotypic parallels imply that the normal stem cell could transform malignantly; however, it is also possible that lineage-restricted or differentiated progeny transform, through a process of de-differentiation, into malignant cells displaying stem-like features (2). Yet, the target stem cell may require fewer alterations to initiate tumorigenesis compared with its downstream progeny (3). This so-called “cancer stem cell” has been isolated and characterized from a variety of haematological malignancies and solid tumors including medulloblastomas (4).

However, it must be understood that the term “cancer stem cell” is an operational term for a tumor cell that displays the ability to self-renew and also divide to generate another stem cell and a progenitor cell, which generates the multitude of cell types comprising the bulk of the tumor (5). Definitive proof for the existence of cancer stem cells was first elucidated in acute myeloid leukaemia (AML) whereby xenotransplantation of human AML cells into immunodeficient mice revealed the frequency of leukaemic stem cells to be about 0.1–1% of all the tumor cells; specifically, only the CD34+CD38 cells were able to reconstitute the leukaemic phenotype whereas the CD34+CD38+ and CD34 cell were unable to do so (6). More recently, using the neural stem cell marker CD133, a putative brain tumor stem cell has been isolated from medulloblastoma, that is capable of self-renewal and multilineage differentiation (4). Although immunophenotyping has demonstrated that many malignancies may be organized hierarchically, functional assays using vital fluorescent dyes have also revealed that the cancer stem cell phenotype may be defined by high expression of various types of ATP-binding cassette (ABC) transporters. Indeed, seminal experiments using the Hoechst 33342 dye, which can be effluxed by ABCG2 transporters, have identified an unlabeled side population (SP) that is enriched for stem cells in the bone marrow (7). These observations have been extended by an increasing number of studies that have identified the SP phenotype in a variety of primary tumor tissues and cancer cell lines.

Remarkably, the presence of stem-like cells has also been reported in long-established cultured cell lines (8–10). Thus, the major objective of this study was to characterize the established DAOY medulloblastoma cell line for cancer stem-like cells using both immunophenotypic (CD133) and physiological substrate (Hoechst 33342) markers and to determine the cellular properties of the isolated putative stem-like fraction with respect to proliferative capacity, cell cycle status, and reconstitution of the original phenotype. Our results demonstrate a weak relationship between the expression of these stem cell markers and clonogenicity, indicating that these markers may not target the stem cell compartment in established cancer cell lines.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Cell Culture

The DAOY human medulloblastoma cell line was obtained from the American Type Culture Collection (Manassas, VA). DAOY cells were routinely propagated in Dulbecco's Modified Eagle's Medium (DMEM; Invitrogen, Carlsbad, CA) supplemented with 10% (v/v) foetal bovine serum (FBS; Sigma Corp., Oakville, ON) and penicillin-streptomycin (1× final concentration; Invitrogen, Carlsbad, CA). Cultures were grown at 37°C in a humidified 5% CO2 incubator and passaged when 70–80% confluent.

Hoechst 33342 Staining

DAOY cells were detached with 0.5 mM EDTA/PBS to better preserve the integrity of cell surface molecules, and resuspended at 0.5–1.0 × 106 cells/mL in 0.1% BSA/PBS. Hoechst 33342 was added at a final concentration of 0.5, 1.5, 2.0, 3.0, 5.0, 6.0, 7.5, and 10.0 μg/mL. In selected samples, verapamil was additionally added to a final concentration of 100 μM. Cells were then incubated for 90 min at 37°C/5% CO2 with intermittent agitation. After incubation, cells were washed twice with cold PBS and resuspended at 2 × 106 cells/mL supplemented with propidium iodide (PI) (2 μg/mL). The cell suspensions were then analyzed with a flow cytometer equipped with an ultraviolet (UV) laser (described below).

Immunophenotyping

Cells, grown in 10-cm tissue culture dish (Sarstedt, Montreal, QC), were detached with 0.5 mM EDTA/PBS for 15 min at 37°C/5% CO2 prior to washing and reconstituting to a final concentration of 0.5–1.0 × 106 cells/mL in 0.1% BSA/PBS. After additional washing in PBS, cells were reconstituted in 0.1 mL of 0.1% BSA/PBS and labeled with 10 μL of anti-CD133-PE (Miltenyi Biotec, Auburn, CA) for 30 min at 4°C; the isotype IgG1-PE was added at a similar concentration. Cells were then washed twice with PBS and resuspended in 0.3 mL of 0.1% BSA/PBS and analyzed by flow cytometry.

Cell Sorting Strategy

Cells were sorted on a MoFlo™ High Performance Cell Sorter (Dako, Fort Collins, CO) either into tubes or into 96-well microwell plates using the Cyclone system. The machine was equipped with three excitation lines which included a water cooled Coherent Innova 90C Argon 488 nm (blue) laser (150 mW), a Spectra-Physics Helium-Neon 635 nm (red) laser (35 mW) and a 351 nm (UV) water-cooled Innova 90 Krypton laser. Five fluorescent channels (FL1 to FL5) can pick up the 488 nm laser, two channels (FL6, FL7) the 351 nm laser, and one the 635 nm laser (FL8). Hoechst fluorescence was measured at both 424/44 nm and above 670 nm (split by a 510 nm long-pass dichroic mirror) resulting from UV excitation. PI fluorescence, measured at 610 nm in the FL3 channel, excluded dead cells during sorting. For cell cycle distribution, 5 × 104 cells from the SP and non-SP (NSP) fractions were sorted, using Summit software version 4.3 (Dako), into Modified Krishan's buffer (0.1% sodium citrate, 0.3% NP-40, 0.05 mg/mL PI, 0.02 mg/mL RNase) and incubated on ice for at least 15 min prior to flow cytometric analysis using the BD FACScan (Becton Dickinson, San Jose, CA), which is equipped with a single argon 488 laser and can read from three different fluorescent channels (FL1, FL2, FL3) and two scatter gates (FSC, forward scatter; SSC, side scatter) simultaneously.

CFSE Labeling

Cell division tracking was carried out in the presence of the vital dye carboxyfluorescein succinimidyl ester (CFSE, Invitrogen, Carlsbad, CA), which is converted intracellularly into a membrane impermeable fluorescent dye allowing the monitoring of cell division through a decrease in its intensity in subsequent cell populations. DAOY cells were detached with 0.5 mM EDTA/PBS and resuspended in 0.1% BSA/PBS (106 cells/mL). A stock solution of 10 mM CFSE was adjusted to a final concentration of 10 μM in the cell suspension and incubated for 10 min at 37°C/5% CO2. Subsequently, the staining was quenched by addition of five volumes of ice-cold DMEM/10% FBS and incubated for 5 min on ice in the dark. The cells were washed thrice in PBS and cultured in 10-cm dishes for about 36 h at 37°C/5% CO2. Cells were then harvested by EDTA dissociation, resuspended to about 106 cells/mL and stained with Hoechst 33342, as described above. Data from over 20,000 cells were collected for each fraction (SP and non-SP) with aggregates and cell doublets excluded from the analysis using an FL2-Area versus FL2-width histogram plot. SP and non-SP dye dilution profiles were analyzed using ModFit LT™ 3.2 software packages (Verity Software House, Topsham, ME).

XTT Viability Assay

The XTT assay measures the viability of living cells by quantitatively assessing the production of soluble formazan product resulting from the mitochondrial dehydrogenase-mediated cleavage of the tetrazolium sodium salt XTT. Briefly, after the indicated experimental treatments, buffer exchange into OPTIMEM was performed. Sterile XTT was prepared at 1 mg/mL in pre-warmed (37°C) OPTIMEM whereupon 50 μL of the XTT solution was added to each 200 μL of culture. It must be noted that PMS was prepared at 5 mM (1.53 mg/mL) in PBS and was added to the XTT solution at a final concentration of 25 μM (5 μL of 5 mM PMS added to 1 mL of XTT (1 mg/mL)) prior to addition to the culture media; PMS acts as an electron coupling agent thereby potentiating XTT reduction. After incubating at 37°C for 1.5 h, the optical density (OD) of each well was measured at both 450 nm (reference 620 nm) for quantification of formazan production using an EAR 400AT 96-well plate reader (SLT Labinstruments, Gröding/ Salzburg, Austria). XTT assays were always performed in triplicate with the one-way analysis-of-variance (ANOVA) and the Bonferroni post-test determining statistical significance.

Limiting Dilution Analysis—Neurosphere Assay

Cellular suspensions were labeled with propidium iodide (PI) with PI-negative (viable) cells being sorted by the MoFlo™ flow cytometer (Dako, Fort Collins, CO) into 96-well microwell plates (Sarstedt, Montreal, QC) containing neural stem cell (NSC) media (epidermal growth factor (EGF; 20 ng/mL), basic fibroblast growth factor (bFGF; 20 ng/mL), heparin (2 μg/mL), 2 mM L-glutamine, B-27 (1×), penstrep (1×), and DMEM/F12); final cell numbers ranged from 10 cells per well to 1 cell per well in 0.2 mL volume. Cultures were incubated for 7 days prior to counting the fraction of wells without spheres as plotted against the number of cells plated per well. The proportion of stem-like cells (or the number of cells required to form one sphere) in the DAOY cell line was determined from the Poisson distribution of stem-like cells where F0 = ex (F0 represents the fraction of wells without spheres and x is the average number of stem-like cells per well); F0 = 0.37 corresponds to the number of cells required to be plated in order for one stem-like cell to be present (11).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Stem Cell Marker Expression in the DAOY Medulloblastoma Cell Line

Immunophenotypic stem cell markers or physiological markers of stem cell function can address the existence of putative stem-like cells in the DAOY cell line. To this end, the SP protocol was first established on the MoFlo™ flow cytometer using DAOY medulloblastoma cells titrated with varying concentrations (0.5–10.0 μg/mL) of the Hoechst 33342 dye (data not shown). Incubation at a Hoechst concentration of 5.0 μg/mL for 90 min at 37°C/5% CO2 was utilized to ensure consistency amongst the various experiments. Control incubations were performed with the ABCB1 inhibitor verapamil to verify the specificity of the SP gate. As shown in Figure 1a, the flow cytometric profile of cells labeled with Hoechst 33342 demonstrates that a minor proportion of DAOY cells (mean: 21%; range: 12.4%–39.1%; n = 13 independent experiments) can exclude the Hoechst 33342 dye; the SP was also sensitive to the ABCB1 inhibitor verapamil (Fig. 1a).

thumbnail image

Figure 1. Stem cell marker expression and co-localization in the DAOY cell line. DAOY cells were first labeled with Hoechst 33342 for 1.5 h at 37°C/5% CO2 followed by labeling with anti-CD133-PE for 30 min at 4°C. The cells were analyzed by a flow cytometer and the following regions were gated on R1 (parental cell population excluding cellular debris): (a) R3 (SP) and R6 (non-SP); (b) R5 (total CD133+ cells); (c) R7 (proportion of CD133+ cells that are SP) and R8 (proportion of CD133+ cells that are non-SP); (d) R2 (proportion of SP cells that are CD133+) and R4 (proportion of non-SP cells that are CD133+).

Download figure to PowerPoint

Expression of the primitive stem and progenitor cell marker, CD133, was also evaluated in the DAOY cell line. As illustrated in Figure 1b, the DAOY cell line contains a minor proportion of CD133+ cells (mean: 0.5%; range: 0.36%–0.58%; n = 3 independent experiments); control IgG1 antibody staining did not reveal any significant staining. Furthermore, co-localization of the two stem cell markers, CD133 and SP, was also investigated in the DAOY cell line. After Hoechst 33342 staining, DAOY cells were labeled with anti-CD133 and processed for flow cytometry. Interestingly, CD133+ cells are present in both the SP and non-SP fractions and the proportion of CD133+ cells is almost fourfold greater in the non-SP relative to the SP (Figures 1c and 1d).

Cell Cycle Parameters of DAOY SP and NSP

Given the relatively large proportion of SP cells present within the DAOY cell line, characteristics such as cell size, cell cycle status, and proliferative capacity were evaluated within both the SP and non-SP fractions. Cell size data were captured during cell acquisition and the forward scatter (FSC) profiles of both the SP and non-SP cells were compared and analyzed (Fig. 2). These studies revealed that SP cells are larger than the non-SP cells as evidenced by the significant increase of SP cells in the larger FSC region relative to non-SP cells (Figs. 2a and 2b). Subsequently, the cell cycle status of the SP and non-SP fractions was analyzed through isolation and staining of the individual fractions with Krishan's modified cell cycle buffer for about 15 min prior to cell cycle analysis. Figure 2c demonstrates that the only significant difference in the cell cycle parameters between the two fractions lay in the increased number of non-SP cells residing in the S phase; one-way ANOVA failed to reveal any significant differences in the G0/G1 and G2/M phases. Finally, the proliferative potential of each fraction was assessed by first labeling the cells with 10 μM CFSE and culturing the CFSE-labeled cells for 72 h prior to Hoechst 33342-staining to evaluate CFSE dilution in SP and non-SP cells. Both Figure 3 and Table 1 illustrate that the CFSE dye appears to be slightly more diluted in non-SP cells, thereby potentially reflecting increased cell divisions within this fraction relative to SP cells. These analyses revealed that in addition to the expression of ABC transporters, the SP cells also had different cell cycle characteristics than non-SP cells.

thumbnail image

Figure 2. Phenotypic and cell cycle analysis of SP and non-SP cells isolated from the DAOY cell line. (a) DAOY cells were labeled with Hoechst 33342 and then propidium iodide staining was used to define the gate R3 for live cells. The R1 and R2 gates were used to define the SP and non-SP cells. Forward scatter and side scatter density plots of all live cells (R3) and either SP cells (R1) or non-SP cells (R2) were used to define R6, R7, R8, and R9 (small SP, large SP, small non-SP and large non-SP, respectively). (b) Graphical representation of the proportion of SP and non-SP cells in either the small (R6, R8) or large cell (R7, R9) size gates. (c) SP and non-SP cells were sorted separately into modified Krishan's buffer for analysis of cell cycle phase distribution. *P < 0.05 as determined by Student's t-test.; NSP: non-SP.

Download figure to PowerPoint

thumbnail image

Figure 3. Comparison of CFSE profiles of SP and non-SP cells. CFSE-labeled DAOY cells were incubated at 37°C/5% CO2 for 36 h and subsequently labeled with Hoechst 33342 for 1.5 h under the same conditions prior to analysis by flow cytometry. Over 20,000 events were acquired for both SP and non-SP fractions. The reduced χ2 was 2.19 and 1.78 for SP cells and non-SP cells, respectively.

Download figure to PowerPoint

Table 1. Tabulated summary of the generational history distribution of CFSE-labeled SP and non-SP cells
GENERATIONSP (%)NON-SP (%)
Parent0.000.52
29.155.34
331.4819.43
438.6541.65
511.3217.33
68.3112.79
70.712.13
80.300.65
90.060.14
100.020.02

Since Hoechst-labeled cells (primarily non-SP) may undergo apoptosis when cultured for extended periods of time by virtue of binding to DNA, a bias in favor of improved SP cell viability relative to non-SP cells could potentially exist. Prior to undertaking further studies, we tested this hypothesis by determining the effects of the Hoechst 33342 dye on cellular viability (Fig. 4). Standard gating facilitated sorting of SP and non-SP cells (Fig. 4a) into regular culture media containing 10% FBS; parental cells (+/− Hoechst 33342 labeling) were also sorted into separate wells. Furthermore, additional gating divided both SP and non-SP fractions into lower and upper regions (Fig. 4b). As seen in Figure 4c Hoechst labeling significantly decreased the viability of parental DAOY cells; non-SP cells also displayed a significant reduction in viability compared to SP cells. No significant differences in viability were observed between the lower SP and upper SP cells; interestingly, the converse was true for lower and upper non-SP cells (Fig. 4d). Therefore, during subsequent studies, if long-term viability were a concern, SP cells were compared to non-SP cells as well as parental cells that were isolated through flow cytometry in the absence of Hoechst 33342.

thumbnail image

Figure 4. Cell viability of defined Hoechst 33342 fractions. DAOY cells were first labeled with Hoechst 33342 and then propidium iodide was used to define live cells from which further gates were defined. Cells from each gate, including parental cells (non-gated), were sorted into separate 96-microwell plates and cultured at 37°C/5% CO2 for 4 days at which point viability was measured by the XXT assay as described in the Methods section. (a) SP (R1) and non-SP (R2). (b) Lower SP (LSP; R2), upper SP (USP; R7), lower non-SP (LNSP; R1), and upper non-SP (UNSP; R6) gates. (c). XTT viability of SP, non-SP, and total (parental) cells +/− Hoechst 33342, P < 0.0001 from one-way ANOVA. (d). XTT viability of LSP, USP, LNSP, and UNSP cells, P < 0.01 from one-way ANOVA. P = parental cells.

Download figure to PowerPoint

Asymmetric Division Capacity and Tumor Sphere Formation of DAOY Cell Fractions

In the classical view, through cell division a stem cell may generate a daughter stem cell as well as a progenitor cell that can then proceed through lineage restriction to generate a mature differentiated cell. Therefore, we compared the ability of SP/non-SP and CD133+/CD133 cells to regenerate the entire cellular population. DAOY cells were cultured in DMEM/10% FBS, stained with the fluorescent Hoechst 33342 dye and sorted into SP and non-SP fractions by flow cytometry and then further expanded in the same medium for an additional 2 weeks. Subsequently, upon restaining with Hoechst 33342 and reanalysis by flow cytometry, it was revealed that both fractions have the capability of regenerating each other (Fig. 5). A similar finding was observed for DAOY cells stained with anti-CD133 whereby upon isolation, culture and restaining, both the CD133+ and CD133 fractions were able to regenerate each other (Fig. 5).

thumbnail image

Figure 5. Asymmetric divisional capacity of (a) SP/non-SP and (b) CD133+/CD133 fractions. DAOY cells were labeled with Hoechst 33342 or immunolabeled with anti-CD133-PE (with propidium iodide defining the gate for live cells). The R1 and R3 gates defined the SP and non-SP cells, respectively and R2 defined the CD133+ cells. One thousand events from each gate were sorted separately into separate wells of a 24-well culture plate; parental cells gated from the forward scatter and side scatter plot were also sorted separately. The sorted fractions were incubated at 37°C/5% CO2 and repeatedly passaged for about 2 weeks. The separate fractions were then relabeled with Hoechst 33342 and anti-CD133-PE and subjected to flow cytometric analysis.

Download figure to PowerPoint

To determine the stem cell-like capacity of each fraction in terms of tumor sphere expansion, the limiting dilution analysis method was used (12). DAOY cells were stained with Hoechst 33342 and gated by flow cytometry into separate SP and parental cell fractions (no Hoechst 33342 treatment) that were sorted into separate 96-well microplates containing neural stem cell media at different cell densities. Similarly, DAOY cells were stained with anti-CD133 and sorted by flow cytometry into CD133+ and CD133 populations. Both the CD133+ and CD133 fractions demonstrated nearly equivalent stem cell-like frequencies (Fig. 6) whereas stem-like cells appeared to be slightly enriched in the SP fraction as compared with the parental cell population (Fig. 6). These results indicate that the capacity for tumor sphere formation is not restricted to CD133+ cells alone.

thumbnail image

Figure 6. Tumour sphere assay of (a) SP/parental cell fractions and (b) CD133+/CD133 fractions. DAOY cells were labeled with Hoechst 33342 or anti-CD133 mAb and subjected to flow cytometric analysis. SP/total cell or CD133+/CD133 fractions were identified and sorted into 96 well culture plates containing neural stem cell media (20 ng/mL EGF and 20 ng/mL bFGF); cells were incubated at 37°C/5% CO2 for 7 days to produce spheres as depicted in (c). The fraction of wells without spheres was plotted against the number of cells plated per well. The proportion of stem-like cells (or the number of cells required to form one sphere) in the DAOY cell line was determined as described in the Methods section.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The technique for stem cell labeling and sorting based on cells displaying low red and blue fluorescence subsequent to incubation with the Hoechst 33342 dye has been appreciated for quite some time (7). With resurgence of the cancer stem cell hypothesis, the Hoechst-based technique has now been recently applied to assessing the stem cell frequency of cancer cell lines and primary tumors. However, a recent study has demonstrated that not all cancer cell lines display an SP phenotype (10). An explanation for this discrepancy may lie in re-visiting the basic premise or assumptions underlying the association between the SP phenotype and stem cell activity. In addition to displaying the classical features of self-renewal and quiescence, stem cells also express high levels of certain members of the ABC transporter family, which includes ABCB1 (P-glycoprotein), ABCC1 (multidrug-associated protein 1 or MRP1), and ABCG2 (breast cancer-related protein or BCRP) (13). Interestingly, integrity of the stem cell compartment is not affected in mice with disruptions in Abcb1, Abcc1, Abcg2; however, there is increased sensitivity to chemotherapeutics such as mitoxantrone, vinblastine, and topotecan suggesting that drug transporters may not play an essential role in stem cell development but only subserve xenobiotic resistance (14–16). Clinically, the SP phenotype may reflect selection of drug transporter expression leading to evolution of treatment-resistant cells in a variety of cancers (13). Previous studies have marked putative stem cells with fluorescent dyes (e.g. Hoechst 33342, rhodamine 123) and cytotoxic compounds (e.g. mitoxantrone, methotrexate) and characterized self-renewal properties (7, 15, 17). However, our results have demonstrated Hoechst 33342-mediated toxicity of DAOY cells and that only SP cells display resistance to this toxicity (Fig. 4), presumably attributed to Hoechst 33342 efflux. Furthermore, DNA-binding affinity of Hoechst 33342 may interfere with cellular replication and differentiation thus confounding the ability to detect relevant biological differences between the SP and non-SP fractions. Indeed, an early study revealed that Hoechst 33342 can induce the F9 embryonal carcinoma cell line to differentiate along the endodermal pathway (18). Likewise, a recent study reported that nuclear Hoechst 33342 staining can have a dramatic impact on C2C12 myogenic differentiation and PC12 neuritic differentiation (19). Regarding our results, it is possible that Hoechst 33342 may interfere with differentiation and thus affect non-SP parameters; interestingly, this may explain the slight stem-like cell enrichment in the SP fraction whereas both CD133+ and CD133 fractions have equivalent stem-like cell frequencies. The tumor sphere limiting dilution assay revealed that SP cells were more clonogenic when compared with the parental unlabeled tumor cell population (Fig. 6). Although not presented here, preliminary experiments did reveal that non-SP cells were also able to generate tumor spheres. As well, the cancer stem cell hypothesis would predict that the SP fraction should be able to regenerate both the SP and non-SP fractions whereas the non-SP fraction should only be able to regenerate itself. Interestingly, the results presented here demonstrate that both the SP and the non-SP fraction have the capacity to completely regenerate both fractions (Fig. 5). A recent article has also observed this same finding in the C6 glioma cell line where either fraction was capable of reconstituting the parental cellular population (20).

Immunophenotyping is an alternative method for assessing the existence of stem-like cells. Indeed, this method may bypass the limitations of current physiological-based dye efflux assays by marking stem-like cells extracellularly thereby potentially avoiding toxicity associated with DNA-dye interacalation. The CD133 extracellular receptor is an established primitive stem cell marker that has now found application for cancer stem cell analysis in a variety of cancer cell lines and primary tumors. A recent report has also shown CD133 expression in all tested medulloblastoma cell lines (8). In the present study, CD133+ cells displayed tumor sphere-like growth (Fig. 6) thereby initially confirming previous findings that CD133+ cells display stem-like activity (4, 21, 22). However, CD133 DAOY cells were able to form tumor spheres with a calculated stem-like frequency comparable to that of CD133+ cells (Fig. 6). Also, similar to the SP and non-SP fractions, either the CD133+ or the CD133 fraction was capable of reconstituting both fractions (Fig. 6). A similar study has also reported that in the C6 glioma cell line CD133 cells displayed self-renewal and tumorigenic features (23). Limiting dilution analysis of the tumor sphere assay revealed that not all cells were capable of clonogenic expansion. Thus, rather than the entire culture being made up of stem cells, it appears that a relatively large minority of DAOY cells display clonogenicity. The present study also suggests that plasticity or stem-like potential of tumor cells, as defined by Hoechst 33342 efflux or CD133 expression, is variable and dependent on environmental factors; indeed, the Hoechst 33342 technique appeared to be highly variable with the percentage of SP cells varying significantly with tumor cell density.

Another novel aspect of this study was the finding that CD133 expression mapped to both the SP and non-SP gates (Fig. 1). Initially, the anticipation was that CD133 expression would be localized primarily to the SP gate; however, CD133 expression appears to be enriched almost fourfold within the non-SP gate. Moreover, given both the very low abundance of CD133 expression in both gates and that non-SP cells constitute the vast majority of the parental cell population, this may explain the similar tumor sphere formation efficiencies that were observed between both the CD133+ and the CD133 fractions (Fig. 6).

In conclusion, this study highlights the novel application of flow cytometric methods in assessing the extension of the cancer stem cell hypothesis into cancer cell lines. Using the established DAOY medulloblastoma cell line, our results do not demonstrate a strong relationship between stem cell marker expression and clonogenicity; both non-SP and CD133 cells also display stem-like characteristics. The DAOY cell line contains a relatively large minority of cells that display clonogenicity independent of stem cell marker expression. Thus, Hoechst 33342 and CD133 expression may not be suitable in selectively isolating stem-like cells in the DAOY cell line. This may be of concern when studies begin to utilize these markers to demonstrate efficacy of targeting the cancer stem cell compartment in cancer cell lines such as DAOY (9). This study provides evidence that cancer cell lines may not recapitulate the hierarchical model of stemness observed in vivo. This may be of fundamental clinical importance as many nonclinical testing strategies utilize established cancer cell lines in modeling tumor biology and in screening anticancer therapies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

V.K.S. was supported by scholarships from the Montreal Center for Experimental Therapeutics in Cancer (MCETC) and the Fonds de la recherche en santé du Québec (FRSQ). J.N. was a National Scholar of the FRSQ. We thank Éric Massicotte and Martine Dupuis of the Flow Cytometry Service/Institut de recherches cliniques de Montréal (IRCM) for their invaluable technical assistance and advice.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED