SEARCH

SEARCH BY CITATION

Keywords:

  • Liver cancer;
  • MDR1;
  • BCRP1;
  • Microarray;
  • Chemosensitivity assay;
  • Invasion assay

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

A subset of stem cells, termed “side population” (SP) cells, has been identified and characterized in several mammalian tissues and cell lines. However, SP cells have never been identified or isolated from gastrointestinal cancers. We used flow cytometry and the DNA-binding dye Hoechst 33342 to isolate SP cells from various human gastrointestinal system cancer cell lines. Fifteen of sixteen cancer cell lines from the gastrointestinal system contained 0.3%–2.2% SP cells. Next, we used an oligonucleotide microarray to analyze differentially expressed genes between SP and non-SP cells of hepatoma HuH7. The expression of GATA6, which is associated with embryonic development and hepatocytic differentiation, was significantly upregulated in HuH7 SP cells. The expression of ABCG2, ABCB1, and CEACAM6, which are associated with chemoresistance, was also significantly increased in SP cells. In addition, some epithelial markers and mesenchymal markers were overexpressed in SP cells. Reverse transcription-polymerase chain reaction and immunocytochemical staining validated these results and suggested a multilineage potential for HuH7 SP cells. In hepatoma HuH7 and colorectal SW480 cell lines, SP cells showed evidence for self-renewal, generating both SP and non-SP cells. Finally, chemoresistance to anticancer agents, including doxorubicin, 5-fluorouracil, and gemcitabine, were compared between HuH7 SP and non-SP cells using an ATP bioluminescence assay. The HuH7 SP cells expressed a higher resistance to doxorubicin, 5-fluorouracil, and gemcitabine compared with non-SP cells. These findings demonstrate that cancers of the gastrointestinal system do contain SP cells that show some characteristics of so-called stem cells.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

The cancer stem cell (CSC) hypothesis suggests that neoplastic clones are maintained exclusively by a small subpopulation of cells that give rise to phenotypically diverse cancer cells [13]. To date, the possible existence of CSCs has been shown in leukemia [46] and some solid tumors, including breast cancer [7] and brain tumors [812]. Even a small population of CSCs is potentially very important because they may be responsible for recurrence after cancer treatments, such as chemotherapy, even when most of the cancer cells appear to be killed. Various types of ATP-binding cassette (ABC) transporters have been shown to contribute to drug resistance in many cancers by pumping drugs out of cells. Interestingly, some ABC transporters are expressed by many kinds of stem cells. For example, breast cancer-resistance protein 1 (BCRP1) pumps out the fluorescent dye Hoechst 33342, identifying an unlabeled side population (SP) of cells, which is enriched for stem cells [13]. Recent studies provide evidence that cancer SP cells in a mouse glioma cell line and human neuroblastoma exhibited the capacity for both self-renewal and multilineage proliferation and were largely responsible for the in vivo malignancy [8, 9].

Cancers of the gastrointestinal system are a leading cause of death and, worldwide, are much more prevalent compared with breast cancer and brain tumors. In Japan, cancers of seven gastrointestinal system sites, including the esophagus, stomach, colorectum, liver, and pancreas, are listed among the top 10 causes of cancer death. In this study, we tried to identify cancer SP cells in human gastrointestinal system cancer cell lines. Such SP cells were isolated in 15 of 16 examined cancer cell lines. Moreover, we evaluated the capacity for self-renewal, multilineage potential, and resistance to chemotherapeutic agents in the SP cells of representative hepatoma and colorectal cell lines, HuH7 and SW480.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Cell Culture

Human gastrointestinal system cancer cell lines included esophageal (TE1, TE2 and TE13), gastric (NUGC3, MKN1, MKN7, and MKN28), colorectal (WiDr, SW480, HSC15, and CCK81), pancreas (PK9 and PK45H), and liver (HuH7, Hep3B and HepG2) cell lines. CCK81 was cultured in minimal essential medium (Invitrogen, Carlsbad, CA, http://www.invitrogen.com), Hep3B was maintained in GIT (Nihon Pharmaceutical, Tokyo, http://www.nihon-pharm.co.jp), and the remaining cell lines were cultured in RPMI 1640 (Invitrogen). Each medium was supplemented with 10% fetal bovine serum (FBS; Equitech-Bio, Kerrville, TX, http://equitech-bio.com), 100 units per ml penicillin G, and 100 μg/ml streptomycin (Gibco, Grand Island, NY, http://www.invitrogen.com). The cells were cultured at 37°C in a humidified atmosphere containing 5% CO2. NUGC3, MKN1, MKN7, MKN28, WiDr, SW480, HSC15, and CCK81 were obtained from the Japanese Collection of Research Bioresources Cell Bank (Tokyo, http://cellbank.nibio.go.jp), and the remaining cell lines were from the Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer (Tohoku University, Sendai, Japan).

Flow Cytometry

To identify and isolate SP and non-SP fractions, cells were removed from the culture dish with trypsin and EDTA, pelleted by centrifugation, washed with phosphate-buffered saline (PBS), and resuspended at 37°C in Dulbecco's modified Eagle's medium (DMEM) containing 2% FBS and 1 mM HEPES. The cells were then labeled with Hoechst 33342 (Invitrogen) at a concentration of 5 μg/ml for TE2, TE13, MKN7, MKN28, PK9, and PK45H and 10 μg/ml for TE1, NUGC3, MKN1, WiDr, CCK81, HuH7, Hep3B, and HepG2. The labeled cells were incubated for 70–90 minutes at 37°C, either alone or with 50 μM verapamil (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com). After staining, the cells were suspended in Hanks' balanced saline solution (HBSS; Invitrogen) containing 2% FBS and 1 mM HEPES, passed a through 40 μm mesh filter, and maintained at 4°C until flow cytometry analysis.

Cells were counterstained with 1 μg/ml propidium iodide to label dead cells. Then, 1 × 106 viable cells were analyzed and sorted by an EPICS ALTRA (Beckman Coulter, Fullerton, CA, http://www.beckmancoulter.com). The Hoechst dye was excited at 350 nm, and its fluorescence was measured at two wavelengths using a 450 DF10 (450/20 nm band-pass filter) and a 675LP (675 nm long-pass edge filter) optical filter. The gating on forward and side scatter was not stringent, and only debris was excluded [4].

RNA Extraction and Oligonucleotide Microarray

We collected both SP and non-SP cells of HuH7, 1 × 104 cells each, in a microcentrifuge tube with 350 μl RLT buffer containing 1% 2-mercaptoethanol. We extracted total RNA from these cells using an RNeasy Mini Kit (Qiagen, Valencia, CA, http://www1.qiagen.com) according to the manufacturer's protocol. The purity and concentration of RNA were determined by a Nano Drop (NanoDrop Technologies, Wilmington, DE, http://www.nanodrop.com) and Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, http://www.agilent.com) as described previously [14, 15]. We used the commercially available Human Whole Genome Oligo Microarray Kit (Agilent Technologies), which contains more than 41,000 features, including 36,866 characterized human genes, listed at http://www.chem.agilent.com/scripts/generic.asp?lpage=5175&indcol=Y&prodcol=Y. Cyanine (Cy)-labeled cRNA was prepared using T7 linear amplification as described in the Agilent Low RNA Input Fluorescent Linear Amplification Kit Manual (Agilent Technologies). Briefly, 50 ng of purified total RNA was reverse-transcribed to generate double-stranded cDNA using an oligo dT T7 promoter primer and Moloney murine leukemia virus reverse transcriptase. Next, cRNA was synthesized using T7 RNA polymerase, which simultaneously incorporated Cy3- or Cy5-labeled cytidine triphosphate. During this process, the samples of SP and non-SP cells were labeled with Cy5 whereas the Human Universal Reference Total RNA (Clontech, Palo Alto, CA, http://www.clontech.com) was labeled with Cy3 as control. Quality of the cRNA was again checked using the Agilent 2100 Bioanalyzer. One microgram aliquot each of Cy3-labeled cRNA and Cy5-labeled cRNA were combined and then fragmented in a hybridization cocktail (Agilent Technologies). Then, the labeled cRNAs were hybridized to a 60-mer probe oligonucleotide microarray and incubated for 17 hours at 60°C. The fluorescent intensities were determined by an Agilent DNA Microarray Scanner and were analyzed by G2567AA Feature Extraction Software Version A.7.5.1 (Agilent Technologies), which used the locally weighted linear regression curve fit normalization method. We performed these procedures, from flow cytometry to the oligonucleotide microarray analysis, in triplicate. This microarray study followed MIAME (minimum information about a microarray experiment) guidelines issued by the Microarray Gene Expression Data group [16].

Identification of Differentially Expressed Genes in HuH7 SP Cells

After subtracting local and global background signals, expression values were calculated as the log ratio of the dye-normalized red (Cy5) and green (Cy3) channel signals. The data flagged as being of poor quality according to the Agilent data extraction software were removed from the analysis. All intensity data were imported to the Rosetta Luminator system version 2.0. (Rosetta Biosoftware, Seattle, WA, http://www.rosettabio.com) [15]. Sequences that were twofold upregulated or down-regulated in SP cells compared with non-SP cells and that had a p < .1 (Student's t test) were defined as differentially expressed.

RT-PCR Assay

From the isolated SP and non-SP cells, total RNA was extracted, and T7 linear amplification was applied as described above. Each 200 ng aRNA (amplified RNA) was reverse-transcribed into cDNA using 5× first-strand buffer (Invitrogen), random primers (TaKaRa, Shiga, Japan, http://www.takara-bio.co.jp), dNTP (TaKaRa), DTT (Invitrogen), RNase Inhibitor (Promega, Madison, WI, http://www.promega.com), and M-MLVRT (Invitrogen). To validate the oligomicroarray results, we used the aRNA samples for reverse transcription-polymerase chain reaction (RT-PCR) analysis. The PCR primers used for amplification were as follows: BCRP1, 5′-GGAGGCCTTGGGATACTTT-GAA-3′ and 5′-GAGCTATAGAGGCCTGGGGATTAC-3′ for a 380-bp fragment; multidrug resistance 1 (MDR1), 5′-GCCTG-GCAGCTGGAAGACAAATAC-3′ and 5′-ATGGCCAAAAT-CACAAGGGTTAGC-3′ for a 253-bp fragment; Keratin 19, 5′-TCCCGCGACTACAGCCACTACTAC-3′ and 5′-TTGGCTT-CGCATGTCACTCAGGAT-3′ for a 396-bp fragment; Albumin, 5′-CGGCTTATTCCAGGGGTGTG-3′ and 5′-GGGGGAGGTT-TGGGTTGTC-3′ for a 369-bp fragment; CEACAM6, 5′-GAAATACAGAACCCAGCGAGTGC-3′ and 5′-CAGTGAT-GTTGGGGATAAAGAGC-3′ for a 226-bp fragment; and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5′-TTGG-TATCGTGGAAGGACTCA-3′ and 5′-TGTCATCATATTTG-GCAGGTTT-3′ for a 249-bp fragment. The PCR products were separated by electrophoresis in 2% agarose gel.

Immunocytochemical Staining

To examine the cellular characteristics of HuH7 SP cells, we originally established the appropriate cell culture conditions. We first maintained HuH7 SP and non-SP cells in DMEM/Ham's F-12 (Invitrogen) medium, supplemented with 1 μg/ml insulin (Wako, Osaka, Japan, http://www.wako-chem.co.jp), 1 × 10−7 M dexamethasone (Wako), 10 mM nicotinamide (Wako), 2 mM l-glutamine (Invitrogen), 50 μM 2-mercaptoethanol (Wako), 5 mM HEPES, 10% FBS, 100 units per ml penicillin G, and 100 μg/ml streptomycin. We adopted this medium here because it has been used to maintain mouse hepatic progenitor cells [17]. Although the HuH7 SP cells survived and proliferated in this culture medium, few SP cells could be detected after several weeks. We thus examined which factors could adequately maintain HuH7 SP cells in a serum-free culture media. We tested stem cell factor (Chemicon, Temecula, CA, http://www.chemicon.com), platelet-derived growth factor (Pepro Tech, Rocky Hill, NJ, http://www.peprotech.com), basic fibroblast growth factor (R&D Systems, Minneapolis, http://www.rndsystems.com), and leukemia inhibitory factor (LIF; Chemicon) as candidate factors. We subsequently determined that the above culture medium, supplemented with 20 ng/ml recombinant human LIF, could effectively expand the HuH7 SP cells.

To study the expression of lineage markers, including a hepatocyte marker (albumin: ALBU), a cholangiocyte marker (cytokeratin 19: KRT 19), liver stem cell marker (cytokeratin 14: KRT 14), and hematopoietic and neural stem cell marker (prominin 1: CD133), in the HuH7 SP and non-SP cells, we carried out immunocytochemical staining. The cells were maintained for 24 hours in collagen type 1 precoated chamber slides (Bio-coat; Beckton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com). Cells were fixed with 2% paraformaldehyde/PBS for 30 minutes at 4°C for KRT19, KRT14, and ALBU and 4% paraformaldehyde/PBS for 20 minutes at room temperature for CD133. Then they were washed three times with cold PBS and with cold 0.2% TritonX/PBS for 10 minutes at 4°C and stained with the following antibodies overnight at 4°C: anti-KRT19 (mouse monoclonal immunoglobulin G (IgG) 2b; 1:10; Progen Biotechnik GmbH, Heidelberg, Germany, http://www.progen.de), anti-KRT14 (mouse monoclonal IgG3; 1:100; Chemicon), anti-serum albumin (mouse monoclonal IgG1; 1:100; Zymed Laboratories, San Francisco, http://www.invitrogen.com) and anti-CD133 (mouse monoclonal IgG; 5 μg/ml; Genzyme-Techne, Minneapolis, http://www.rndsystems.com). The chamber slide was washed with cold 0.2% Triton-X/PBS for 5 minutes at 4°C three times and with cold 0.2% Triton-X/PBS containing 3% bovine serum albumin for 5 minutes at 4°C. The primary antibodies were detected with fluorescein isothiocyanate-conjugated anti-mouse IgG1. The cells were counterstained with mounting medium containing 4′, 6-diamidino-2-phenylindole (Molecular Probes, Invitrogen) to identify all nuclei.

Survival Studies for Doxorubicin, Gemcitabine, and 5-Fluorouracil

We isolated the HuH7 SP and non-SP cells and seeded them into 96-well culture plates at 5 × 102 cells per well for each population of cells. We incubated them in culture medium as described above at 37°C in an atmosphere containing 5% CO2 for 24 hours. The cells in both populations were treated with doxorubicin (0.01 and 0.1 μg/ml) or 5-fluorouracil (5-FU; 0.1 and 10 μg/ml) or gemcitabine (0.1 and 100 μg/ml). After 72 hours of exposure to the chemotherapeutic agents, viability of the cells was determined using an ATP bioluminescence assay (CellTiter-Glo Luminescent Cell Viability Assay; Promega), and the luminescence signal was detected by a luminometer (ARVO MX; PerkinElmer, Boston, http://las.perkinelmer.com) according to the manufacturer's protocol [18].

Invasion Assay

The invasiveness of SP and non-SP cells was evaluated by a Matrigel assay. Each cell population was added to 8.0-μm pore FluoroBlok (BD Biosciences, San Diego, http://www.bdbiosciences.com) inserts at a density of 5 × 104 per insert. The culture medium (described above) with 20 ng/ml LIF and 10% FBS was used as chemoattractant in the lower wells and cultured for 72 hours. The cells were then labeled with 4 μg/ml Calcein-AM (Molecular Probes, Inc., Eugene, OR, http://probes.invitrogen.com) in HBSS and incubated at 37°C in a humidified atmosphere containing 5% CO2. Fluorescence of cells that had invaded through the FluoroBlok inserts was measured on a plate reader (ARVO MX; PerkinElmer).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Existence of SP Cells in Gastrointestinal Cancer Cell Lines

To determine whether a selection of cancer cell lines contains SP cells, we stained them with Hoechst 33342, which is actively extruded by verapamil-sensitive ABC transporters. Representative results analyzed by flow cytometry (Fig. 1) are shown. Cancer cell lines from the esophagus contained 0.3–1.4% SP cells; stomach 0.6–2.2% SP cells; colorectal 0.3–0.5% SP cells; liver 0–1.8% SP cells; and pancreas 0.3–1.3% SP cells (Table 1). In each case, by definition, the number of SP cells was decreased greatly by treatment with verapamil (Fig. 1). We could not detect SP cells in just a single liver cancer cell line HepG2. Thus, most of the gastrointestinal system cancer cell lines do contain a small component of cancer SP cells. Because the existence of bipotent human liver progenitor cells with both hepatocyte and cholangiocyte lineages has recently been suggested [19, 20], further experiments were focused on one of the human liver cancer cell lines, HuH7.

Identification of Differentially Regulated Genes in HuH7 SP Cells

Using a Human Whole Genome Oligo Microarray, we determined differentially expressed genes between SP and non-SP lineages of HuH7. We identified 268 upregulated and 170 downregulated genes in the HuH7 SP cells compared with the non-SP cells (supplemental online Table 1). We categorized the specifically upregulated genes in HuH7 SP cells, according to their biologic functions, as summarized in Figure 2A. First, we succeeded in confirming that the expression of ABC transporters, such as BCRP1 (ABCG2, 2.0-fold, p = .032) and multidrug resistance 1: MDR1 (ABCB1, 2.3-fold, p = .011), was upregulated in the HuH7 SP cells. Including these ABC transporters, genes associated with a “transporter” (13%) were the most frequently upregulated (Fig. 2A). The expression of genes related to signaling pathways of normal stem cells, such as BMP2 (2.6-fold, p = .011) and JAG1 (2.5-fold, p = .072), was also increased. The most pronounced (9.8-fold, p = .017) upregulated element was carcinoembryonic antigen-related cell adhesion molecule (CEACAM6), which may serve as a determinant of gemcitabine chemoresistance [21].

Verification of Microarray Data by RT-PCR Analysis

We carried out semiquantitative RT-PCR amplification to verify the data of several differentially expressed genes identified by the oligomicroarray. Sufficient expression of ALBU in both SP and non-SP cells was determined in our microarray analysis (data not shown). The RT-PCR results corresponded very well to the microarray data for all five genes tested (Fig. 2B).

Repopulation of Both SP and Non-SP Fractions by HuH7 SP Cells

To further examine the cellular characteristics of HuH7 SP cells, we examined which factors adequately maintained HuH7 SP cells. We determined that culture medium supplemented with LIF could effectively expand the HuH7 SP cells. To compare the self-renewal ability of SP cells, we maintained them in this supplemented medium for 2 weeks and then sorted them again into SP and non-SP cells by flow cytometry. We found that the cultures initiated with SP cells contained both SP and non-SP cells, whereas those with non-SP cells generated only non-SP cells (Fig. 3A). The percentage of SP cells was increased when analyzed after 2 weeks of culture of the SP cells compared with that of the pre-isolated HuH7 cell line (pre-isolation vs. first sort and 2 weeks of culture: 0.9 vs. 9.1%). In addition, when the sorting and 2-week culture was repeated four times, the percentage of SP cells was greatly increased compared with that of the first sort and 2-week culture (first sort and culture vs. four repeats of sorting and culture: 9.1 vs. 31.6%) (Fig. 3A). Even after a fourth repeat of sorting and culture, when these were subcultured without another repeated sorting step, the percentage of SP cells was decreased and almost equaled that of the untreated HuH7 cell line (data not shown). Comparable results were obtained for colorectal cancer cell line SW480, confirming the pattern seen with HuH7 (Fig. 3B). We performed our analyses in triplicate with similar results in each case.

Immunocytochemical Analysis

We confirmed the expression of ALBU, KRT19, KRT14, and CD133 immunocytochemically in both SP and non-SP cells of HuH7 (Fig. 4). Corresponding to the observed increased gene expression in the microarray analysis, staining for KRT19 (cholangiocyte marker; 2.5-fold, p = .06) and KRT14 (thought to be a liver stem cell marker; 2.6-fold, p = .03) was strongly positive in SP cells and negative in non-SP cells. Conversely, ALBU (mature hepatocyte marker) expression was very weak in SP cells and strong in non-SP cells. The staining for CD133 (hematopoietic and neural stem cell marker) was strongly positive in SP cells and negative in non-SP cells. In addition, SP cells were morphologically quite distinguishable (small and round in shape) from non-SP cells (fibroblast-like).

Chemoresistance of HuH7 SP Cells to Anticancer Agents

The chemoresistant ability of SP cells has been reported to depend mainly on ABC transporters [4, 9, 22]. Indeed, HuH7 SP cells express families of ABC transporters such as MDR1 and BCRP1. We examined the sensitivity of HuH7 SP and non-SP cells to doxorubicin, which is well known as an ABC transporter-dependent anticancer drug [22]. The SP and non-SP cells of HuH7 were cultured in 96-well plates and treated with various concentrations of doxorubicin (Fig. 5). After exposure to doxorubicin, the viability of the SP cells was markedly higher compared with that of the non-SP cells (p < .01 for both 0.01 μg/ml and 0.1 μg/ml concentrations). Interestingly, HuH7 SP cells also expressed high levels of CEACAM6 (9.8-fold, p = .017) and AREG (7.6-fold, p = .023) in our microarray study. The expression of CEACAM6 is associated with resistance to gemcitabine [21] and AREG to gefitinib [23]. To determine whether SP cells resist ABC transporter-independent anticancer drugs more than do non-SP cells, we tested 5-FU and gemcitabine because they are usually used for the treatment of hepatobiliary tracts neoplasms. In the case of 5-FU treatment, viability of SP cells was again significantly higher than that of non-SP cells (p = .02 for 0.1 μg/ml, p < .01 for 10 μg/ml) and similar for gemcitabine (p = .02 for 0.1 μg/ml, p < .01 for 100 μg/ml). We performed these analyses in triplicate.

Invasion Assay

To evaluate any differences in biological behavior between SP and non-SP cells, we performed invasion assays (Fig. 6). Invasiveness of non-SP cells was higher than that of SP cells (2.45-fold, p = .01). We performed these analyses in triplicate.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

In the current study, we were able to identify a small SP component (0.3%–2.2%) of cancer cells from various human gastrointestinal organ cancer cell lines. To the best of our knowledge, this is the first report demonstrating the existence of cancer SP cells in gastrointestinal cancers. The percentages of SP cells detected are similar to most previous reports: 2% in human breast cancer cell line MCF7, 0.4% in rat C6 glioma, and 1.2% in human Hela carcinoma [8] but less than the 4%–37% noted in neuroblastoma cell lines [9].

Stem cells identified from solid tumors usually express organ-specific markers. In the case of breast cancers, a cell surface marker profile CD44+ CD24−/low Lin was reported [7]. The cells isolated from central nervous system tumors express cell surface antigen CD133, which is known as a marker of hematopoietic stem cells [12]. However, distinct regulated molecules characterizing CSCs are largely unknown. The drug transport property conferred by the ABC transporter is an important marker in the isolation and analysis in hematopoietic stem cells and termed “side population” cells. The SP fraction is a useful tool for stem cell studies, especially when specific cell surface markers are unknown. Using a Human Whole Genome Oligo Microarray of 41,000 clones, we tried to identify differentially expressed genes between SP and non-SP lineages of HuH7. The top upregulated element was CEACAM6, a glycosylphosphatidylinositol-linked immunoglobulin superfamily member that is over-expressed in a variety of gastrointestinal cancers. Over-expression of CEACAM6 promotes cell survival under anchorage-independent conditions and protects cells from apoptosis [24]. Furthermore, CEACAM6 over-expression may serve as a determinant of gemcitabine chemoresistance [21]. The expression of AREG (7.6-fold, p = .023), which has recently been associated with resistance to gefitinib [23], was also upregulated. The expression of GATA6 was increased in HuH7 SP cells. This GATA6 is reported to be associated with embryonic development [25] and hepatocytic differentiation [26, 27] and is thought to interact with BMP pathways [28]. Several elements (such as pim-1) that have been linked to the Akt/PI3kinase pathway [29] were also upregulated, and the finding would be interesting from the viewpoint of the HuH7 propagation. We noted increased expression of “structural proteins,” such as VIM (4.3-fold, p = .005), KRT14 (2.6-fold, p = .031), and KRT19 (2.5-fold, p = .055). Over-expression of VIM plays an important role in human liver cancer metastasis [30]. In addition, human liver progenitor cells express VIM [19], KRT14 [19], and KRT19 [19, 20]. Therefore, these molecules could be distinctive markers for liver cancer SP cells. Downregulated genes in SP cells could equally be very important but are not further discussed here.

Stem cells show properties of self-renewal and pluripotential differentiation. The CSC shares many properties with normal stem cells. In central nervous system tumors, stem cells with a capacity for self-renewal and pluripotential development have been isolated [12]. We examined whether the HuH7 SP and SW480 SP cells could indeed reflect the biological characteristics of stem-like cells, such as self-renewal and multilineage proliferation. We found that the SP cells, maintained in appropriately supplemented culture media, reproduced both SP and non-SP cells, whereas non-SP cells generated only non-SP cells. These results indicate that, at least in HuH7 and SW480, a tumor hierarchy exists in which SP cells can generate both SP and non-SP cells; this is in accordance with previous observations that the SP fraction can divide asymmetrically and display a capacity of self-renewal [8, 9, 12]. We have not screened the other cell lines in the panel, and the experiments are ongoing. In addition, it will also be important to confirm the existence of SP cells, which are observed in in vitro assays, in primary clinical tumor cells.

Intriguingly, we noticed a greater expression of KRT19 mRNA in HuH7 SP cells in both oligonucleotide microarray and RT-PCR analysis. We thus evaluated the expression of KRT19 protein immunocytochemically and found its distinct expression only in HuH7 SP cells. ALUB expression was observed in both SP and non-SP cells but was very weak or almost negative in SP cells. Because HuH7 SP cells expressed cell lineage markers for both cholangiocytes (KRT19) and hepatocytes (ALBU), as reported in human liver progenitor cells [19, 20], the liver cancer SP cells may have a capacity for bipotential differentiation. In addition to these findings, liver stem cell marker (KRT14) and hematopoietic and neural stem cell marker (CD133) were strongly positive in SP cells only.

The association of increased expression of ABC transporters in cancer SP cells with resistance to chemotherapeutic agents has been reported [4, 9]. We therefore examined the sensitivity of HuH7 SP cells to doxorubicin, a commonly used agent in the treatment of hepatocellular carcinoma, using an ATP bioluminescence assay. After 72 hours of exposure to doxorubicin, the viability of the SP cells was markedly higher than that of the non-SP cells. The results of the assay also demonstrated the apparent chemoresistance of HuH7 SP cells to the ABC transporter-independent anticancer drugs, 5-FU and gemcitabine. Data from a whole genome DNA microarray analysis provide insight into the existence of key molecules, which may be associated with chemoresistant properties of CSCs not only through ABC transporters but also through other mechanisms, possibly associated with the increased expression of genes such as CEACAM6, AREG, and ABCC2/GSTA1.

In our invasion assay results, non-SP cells showed more invasiveness than did SP cells. It is generally accepted that normal stem cells show properties that provide for a long lifespan, such as relative quiescence, resistance to drugs through the expression of ABC transporters, and anti-apoptosis. These cancer SP cells may also be relatively quiescent compared with non-SP cells, while still playing important roles in maintaining cancer foci (that is, to differentiate and proliferate after chemotherapy and radiotherapy using their self-renewal and chemoresistance capacities). It is difficult to identify CSCs, because they are usually present in very small numbers and specific surface markers are still unknown. A whole genome DNA microarray analysis may be useful for identifying potential candidates for specific CSC markers or even for defining a definitive CSC pattern. If CSC specific surface markers are discovered, that may permit isolation of this important population of cells and cancer therapy and science will be revolutionized.

Table Table 1.. Percentages of SP fractions in various human cancer cell lines of gastrointestinal organ system
Thumbnail image of
thumbnail image

Figure Figure 1.. Analysis of side population (SP) cells in gastrointestinal cancers. Representative SP cell isolation from various human gastrointestinal system cancer cell lines. Cell lines of esophageal (TE13), gastric (MKN28), colon (CCK81), liver (Hep3B), and pancreas cancers (PK9) were labeled with Hoechst 33342 and analyzed by flow cytometry. Results when cells are treated with 50 μM verapamil during the labeling procedure. The SP cells, which disappear in the presence of verapamil (bottom panel), are outlined and shown as a percentage of the total cell population.

Download figure to PowerPoint

thumbnail image

Figure Figure 2.. Identification of differentially expressed genes in HuH7 side population cells. (A): Biologic functions of the upregulated genes determined by the Human Whole Genome Oligo Microarray; 15 of the 268 upregulated genes were spotted on this microarray slide in duplicate—we divided the 253 unique genes into 12 categories, including genes whose functions are unknown (73 genes, 29%). Detailed information on the biological functions of the remaining genes (180 genes, 71%) is available in supplemental online Table 1. (B): Semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) analysis. The oligomicroarray data of five genes were successfully validated by RT-PCR. GAPDH was used as a control. Abbreviation: GAPDH, glyceral-dehyde-3-phosphate dehydrogenase.

Download figure to PowerPoint

thumbnail image

Figure Figure 3.. Side population (SP) cells display a capacity for self-renewal. (A): Self-renewal capacity of HuH7 SP cells. The HuH7 cell line contains 0.9% SP cells (left). After the isolation of both the HuH7 SP and non-SP fractions, we cultured them in media supplemented with leukemia inhibitory factor for 2 weeks and then reanalyzed them by flow cytometry. In these culture conditions, the SP cells expanded sufficiently and maintained the SP subpopulation (9.1%), whereas no SP cells were found in the fraction initiated from non-SP cells (middle). When this sorting and culture were repeated four times, the percentage of SP cells was greatly increased (31.6%). The experiments were repeated in triplicate with similar results. (B): The SW480 (colorectal cancer) cell line showed results similar to HuH7 (liver cancer).

Download figure to PowerPoint

thumbnail image

Figure Figure 4.. Side population (SP) cells express multilineage markers. Immunocytochemical staining to examine the multilineage potential of HuH7 SP cells. Both SP and non-SP cells were cultured as described for 2 weeks and then stained with a hepatocyte marker (ALBU), a cholangiocyte marker (KRT19), a liver stem cell marker (KRT14), and a hematopoietic stem cell marker (CD133). The shape of the cells was observed under differential interference contrast (DIC). Scale bar = 20 μm.

Download figure to PowerPoint

thumbnail image

Figure Figure 5.. Side population (SP) cells show a high resistance to anti-cancer drugs. Chemoresistance of HuH7 SP cells to anticancer drugs. We maintained cells in medium supplemented with leukemia inhibitory factor for 24 hours and incubated them with doxorubicin (left), 5-fluorouracil (middle), and gemcitabine (right) for 72 hours, and then cell viability was determined by an ATP bioluminescence assay (black bar, SP cells; white bar, non-SP cells). The experiments were repeated in triplicate with similar results.

Download figure to PowerPoint

thumbnail image

Figure Figure 6.. Non-SP cells show more invasiveness than do SP cells. Invasion assay of HuH7 SP and non-SP cells (black bar, SP cells; white bar, non-SP cells). The invasiveness of non-SP cells was higher than that of the SP cells (2.45-fold, p = .01). Abbreviation: SP, side population.

Download figure to PowerPoint

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank T. Shimooka, H. Yasunami, M. Oda, and M. Kasagi for excellent technical assistance. This study was supported by the Core Research for Evolutional Science and Technology of Japan, Science and Technology Agency; a grant-in-aid for scientific research (S) (17109013) from the Japan Society for the Promotion of Science; a grant-in-aid for scientific research on priority areas (17015032) from the Ministry of Education, Culture, Sports, Science and Technology of Japan; a grant from the Third-Term Comprehensive Ten-Year Strategy for Cancer Control; and a grant for cancer research from the Ministry of Health, Labor and Welfare of Japan.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
SC050282suppdata1.pdf142KSupplemental Data 1
supplimental_data2.xls348KSupplemental Data 2
sc050282suppdata2.pdf473KSupplemental Data 2

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.