Report

Ovarian carcinoma tumor-initiating cells have a mesenchymal phenotype

Volume 11, Issue 10   May 15, 2012
Pages 1966 - 1976
http://dx.doi.org/10.4161/cc.20308
Keywords: anticancer agents, mesenchymal phenotype, ovarian tumor, pkh26, tumor-initiating cell
Authors: Francesca Ricci, Sergio Bernasconi, Patrizia Perego, Monica Ganzinelli, Giorgio Russo, Francesca Bono, Costantino Mangioni, Robert Fruscio, Mario Signorelli, Massimo Broggini and Giovanna Damia

View affiliations

Abstract:
Solid tumors appear to contain a subpopulation of cells (tumor-initiating cells, TICs) that not only drives and sustains tumor growth, but is possibly responsible for recurrence. We isolated, after enzymatic digestion of primary ovarian carcinoma samples, a subpopulation of cells propagating as non-adherent spheres in medium suitable for tumor stem cells. These cells were able to self-renew in vitro, as suggested by PKH-26 staining studies, were tumorigenic and acquired an epithelial morphology when grown in FBS-supplemented medium, losing their tumorigenic potential. Interestingly, the tumorigenic potential of PKH-26high- and PKH-26neg-sorted cells was similar. These TIC-enriched cultures showed higher levels of genes involved in stemness than differentiated cells derived from them and were more resistant to the cytotoxic effects of some drugs but equally sensitive to others. The higher level of ABCG2 efflux pump could explain increased resistance to taxol and VP16, and higher levels of genes involved in nucleotide excision repair partially explain the resistance to cisplatin. These cells express mesenchymal markers, and epithelial transition could be induced when cultured in differentiating conditions, with a loss of invasive potential. These data suggest that ovarian cancer is a stem cell disease and should help elucidate the role of these cells in the aggressive phenotype of this tumor and find new therapeutic strategies to reduce resistance to current chemotherapeutic drugs.

Received: January 16, 2012; Accepted: April 9, 2012

Preview:



Full Text

Introduction



Epithelial ovarian cancer (EOC) is the eight most common tumor in western women and the most lethal gynecological cancer. The relatively asymptomatic nature of early-stage disease and the lack of adequate screening tests result in 75% of patients being diagnosed at late FIGO stages (III and IV). Taxol and platinum (DDP) are the standard adjuvant therapy in EOC, having greatly improved the overall survival (OS), with 70% of patients achieving complete remission after first-line platinum-based therapy; unfortunately, almost invariably patients relapse with resistant disease.1

Recent data suggest that one of the mechanisms accounting for resistant and/or relapsing disease is a subpopulation of cells in human tumors with stem-like characteristics2-4 (cancer stem cells, CSCs or tumor initiating cells, TICs), which may contribute to the incomplete efficacy of chemotherapy.5,6 For this reason, identification of this cell population and the development of strategies to target it could make a real difference in improving the response rate of ovarian cancer and survival. In the last few years, several studies have reported the isolation and identification of putative TIC in ovarian carcinoma using general cell surface markers, the ability to generate non-adherent, self-renewing spheres in vitro and tumor formation and propagation over several passages in immunodeficient mice. Bapat et al.7 were the first to isolate two clones from ascites of a patient with advanced ovarian adenocarcinoma with the characteristics of TICs (able to form spheres in culture, tumorigenic when transplanted in mice and with a stem cell phenotype). Zhang et al. provided evidence that CD44+/CD117+ cells isolated from ovarian adenocarcinoma overexpressed stem cell genes and were able to form tumors that represented patients’ tumors.8 In addition, these cells were more resistant to some chemotherapeutics than the CD44-/CD117- isolated cells. CD44 with MyD88 positivity defined a subpopulation of ovarian tumor cells with stem-like characteristics. Moreover, the same authors reported that CD44+/Ve-cadherin-/CD34-cells able to regenerate tumors contributed to tumor vascularisation by a mechanism involving IKK.9,10 Recently Gao et al. identified a subpopulation of CD24+ cells with stem-like characteristics in an ovarian tumor specimen of a patient.11 Data on the role of CD133 and ALDH as markers defining an ovarian cancer stem cell are conflicting.12,13

We report the characterization of TICs isolated from fresh ovarian tumor biopsies, able to form spheres in vitro, to self-renew, to form tumors when transplanted in nude mice and to resist standard chemotherapeutics drugs. These cells have a mesenchymal phenotype.

Results



In vivo tumorigenicity of the sphere-forming cells.

Using the method described to obtain TICs from breast cancers,15 we isolated sphere-forming cells from fresh tumor samples of ovarian cancer patients. More than 100 tumor specimens were processed. In these stem cell-selective conditions, all tumors yielded floating cell aggregates; most of the cultures did not even grow after the first passage; some grew up to five passages but then stopped. Two cell lines, derived from patients number 83 and number 110, could be propagated for at least 20 passages after mechanical disaggregation of the spheres and cultured in low-adherence conditions. Subsequent studies focused on them. Limiting dilution experiments showed that the percentage of clonogenic cells was 16% and 26% in cells derived from patients 83 and 110, respectively, and did not change in subsequent passages. When single-cell spheres were cultured with FBS, they easily adhered, acquired an epithelial morphology and grew as a monolayer (Fig. S2A).

To prove that the sphere-forming cells could be considered TICs, we transplanted different numbers of cells in immunodeficient mice. All the ovarian cell-forming spheres gave rise to tumors, indicating that they were strongly tumorigenic (Table 1). In fact, as few as 1,000 cells were able to form tumors that could be serially propagated both as subcutaneous fragments and as limited numbers of cells. When we injected differentiated cells (derived from disaggregated spheres and grown in FBS-containing medium for at least ten days), no tumors grew from cells from patient 110, and one out of two tumors grew from 100,000 cells transplanted from patient 83, with a latency of 67 d, three times longer than after transplanting the same amount of cells from low adherence cultures. These data indicate that the differentiated cells have a lower tumorigenic potential.

<b>Table 1.</b> In vivo tumorigenicity of cells from patient 83 and 110, from low-adherence cultures and from differentiated cells derived from them
Patient Passage no. Cells dose/
mouse		 Injection site Tumor take/
No. mice		 Latency (days)
83 LA I 1,000 s.c. 1/2 60
100 000 s.c. 2/2 20
II 1,000 s.c. 2/2 26
10,000 s.c. 2/2 7
III fragment s.c. 2/2 10
IV fragment s.c. 6/6 10
83 DC I 1,000 s.c. 0/2*
100,000 s.c. 1/2 67
110 LA I 1,000 s.c. 0/2
100 000 s.c. 1/2 30
II fragment s.c. 2/2 8
III fragment s.c. 2/2 10
IV 1,000 s.c. 1/2 40
100,000 s.c. 2/2 30
110 DC I 1,000 s.c. 0/2*
100 000 s.c. 0/2*

LA, low adherence; DC, differentiated cells. *Observation time more than six months.

Histological analysis confirmed that xenograft tumors were ovarian carcinomas. Moreover, while patients’ tumors displayed different degree of differentiation with glandular, papillary and solid patterns and marked cellular polymorphism, tumors grown in nude mice presented a solid pattern of monomorphic cells (Fig. S3). The expression of different type of cytokeratins was observed in tumor xenografts was even less marked than patients’ samples (data not shown).

 PKH-26 studies on sphere-forming cells.

The formation of spheres could be followed more closely after staining cells with PKH-26, a vital dye that irreversibly binds to the cell membrane. In cells undergoing subsequent divisions, the label is equally partitioned among daughter cells, resulting in a reduction of the fluorescence intensity.16,17 When cells from disaggregated spheres of patients 83 and 110 were stained with PKH-26 and allowed to grow in stem cell culture conditions, a clear red-stained cell was visible in the center of the sphere (Fig. S2B); on the other hand, all the other cells of the sphere were almost negative for the PHK‑26 label. These data strongly suggest that the sphere-forming cells underwent the first division, giving rise to two daughter cells, one of which underwent no further divisions, while the other showed total dye quenching, suggesting consecutive, rapid divisions. Around 15% of spheres contained the red cell. All the other cells, which formed spheres without the bright red cell inside, are probably proliferating cells in which the dye has been portioned and progressively diluted. These data were corroborated by FACS analysis, showing that staining of cells from disaggregated spheres resulted in about 100% positive cells, whose fluorescence was progressively lost over time (data not shown).

The experiments performed with PKH-26 staining suggested that only a minority of these low-adherence cultures were able to undergo one or two divisions and then stop dividing. We therefore checked the ability of both PKH-26high and PKH-26neg cells to be clonogenic in vitro and to be tumorigenic in vivo. To this aim, cells disaggregated from 83 and 110 spheres were labeled with PKH-26 as described in Materials and Methods and were allowed to grow for seven days. Then, cells were sorted based on their PKH-26 fluorescent content in PKH-26high and PKH-26neg cells. The in vitro clonogenic ability was higher in PKH-26neg than in PKH-26high cells derived from both 83 and 110 low adherence cultures (33% vs. 18% and 42% vs. 18%, respectively). However, the tumorigenic potential of PKH-26neg and cells PKH‑26high was similar (Table 2); the tumor appearance latency time seemed to be higher in PKH-26high cells, but when tumors grew, their growth rate was similar to tumors derived from PKH-26neg cells (data not shown).

In vivo tumorigenicity of PKH-26high and PKH-26neg cells sorted from patient 83 and 110 low-adherence cultures.
Patient 83 Patient 110
PKH-26high PKH-26neg PKH-26high PKH-26neg
No injected cells Tumor take/ No mice Median latency (days) Tumor take/ No mice Median latency (days) Tumor take/ No mice Median latency (days) Tumor take/ No mice Median latency (days)
500 7/9 75 14/17 46 6/7 58 4/7 48
5000 5/7 30 17/17 26 6/7 36.5 6/7 53
1 0/4 * 0/4 * 0/4 * 1/4 51
10 0/4 * 0/4 * 0/4 * 2/4 56-78
100 0/4 * 3/4 28 1/4 64 1/4 36
1000 1/4 62 3/4 28 4/4 39.5 4/4 36

Last observation day 112.

 Molecular and phenotypical characterization of low adherence and differentiated cells.

We investigated the expression of genes correlated with stemness (including BmI1, Nestin, Nanog, Oct4, Notch and Hes1) in both low adherence and differentiated cells. All these genes were expressed at higher levels in low-adherence spheres from patient 83 than in the corresponding differentiated cells derived from them (Fig. 1A); in cells from patient 110, only BmI1 and Nestin were more expressed in low adherence cells than in the corresponding differentiated cells, while the other genes showed similar patterns of expression.

Figure 1. Molecular and phenotypic characterization of spheres. (A) Expression analysis of genes involved in self-renewal and stemness. Copy number of genes (mean ± SD) normalized by the copy number of two housekeeping genes (actin and cyclofillin) is reported for patient 83 LA (Low Adherence) cultures (), patient 110 LA cultures (▒), patient 83 DC (differentiated cells) (■) and patient 110 DC (▓), *p < 0.03, **p < 0.01, ***p < 0.001. (B) Immunofluorescence staining (DAPI and FITC fluorescence) on spheres of different markers in LA cultures of both patient 83 and 110 spheres.

CD117, CD133, CD44 and CD24 have been reported to be associated with ovarian cancer stem cells in different studies. None of these papers, however, reported overlapping expression of these markers. We evaluated the expression of these markers in low-adherence cell cultures by immunofluorescence assays. Only CD117 marked the cells grown as spheres (Fig. 1B). Muc1 was the only epithelial marker, among all the others tested (data not shown), that was positive in 83 and 110 patients’ low-adherence cells (Fig. 2A and data not shown); interestingly, Muc1 expression was lost in cells cultured in differentiating conditions (data not shown).

Figure 2. Spheres express mesenchymal markers and display more invasive behavior than differentiated cells. (A) Immunofluorescence staining (DAPI and FITC fluorescence) on the low adherence cell (LA) cultures and on more differentiated cells (DC) of epithelial (Ber4, Muc-1, E-cadherin), mesenchymal (N-cadherin, Vimentin) markers and of c-Met and P-Met. (B) Expression analysis of genes involved in EMT on patient 83 () and patient 110 LA spheres (▒) and on patient 83 (■) and patient 110 DC (▓). (C) Chemotaxis (left panel) and invasion assays (right panel). Numbers of cells migrated into the lower side of the membrane are reported. Data are the mean+ SD of at least three different experiments. *p < 0.01, **p < 0.005, ***p < 0.001, ****p < 0.0001.

We studied a number of markers associated with epithelial-mesenchymal transition (EMT) (Fig. 2A). Spheres expressed N-cadherin and vimentin, while the expression of these proteins was lost in cells growing in FBS-supplemented medium (adherent culture condition), where cells acquired the expression of E-cadherin. Both low-adherence spheres and differentiated cells derived from them stained positive for c-Met, but c-Met was constitutively active only in the low adherence cells, as suggested by its high level of phosphorylation.

Then, we studied mRNAs of genes associated with EMT, and found that Snail, Slug, Twist and Fox2 were higher in low-adherence cultures than in cells from differentiating conditions derived from patient 83; in cells derived from patient 110, only Twist was differentially expressed (Fig. 2B).

We then investigated whether the mesenchymal phenotype was associated with an increase in the ability of cells derived from spheres to migrate and invade compared with cells grown in 10% FBS. Chemotaxic experiments showed that only low-adherence cells from patient 83 had a greater ability to move than the differentiated cells derived from them (Fig. 2C, left-hand part). In addition, the invasive capacity was increased in unstimulated control cells derived from both patient’s 83 and 110 disaggregated spheres (Fig. 2C, right-hand part), suggesting that these latter are more aggressive than the differentiated cells.

 Pharmacological characterization of low adherence and adherent cells.

Having established that the low-adherence culture represents a tumor cell population enriched in TICs, we examined the pharmacological profile of both these cultures and differentiated cells derived from them. Anticancer agents with different mechanisms of action were tested, particularly drugs used in the treatment of ovarian cancer and other new cytotoxic and targeted agents. As reported in Figure 3, compounds such as DDP, taxol and VP16 were much more active in differentiated cells than in enriched TIC cultures, confirming data from other experimental systems.8,9 A number of drugs were equally active on both cell populations: yondelis (ET743, a natural compound whose mechanism of action has yet to be fully elucidated), PS341 (a proteasome inhibitor and molecular targeted compound) and PI303 (a dual PIK3/mTOR inhibitor). These intriguing data suggest that TICs may be differentially targeted.

Figure 3. Pharmacological characterization of spheres and adherent/differentiated cells. Dose/response curves of different agents in patient 83 (upper panels) and patient 110 (lower panels) spheres (-■-) and on patient 83 and patient 110 differentiated cells (-□-). Data are percentages of control cells and are the mean + SD of three different experiments.

In order to understand the molecular mechanism at the basis of the differences in sensitivity, we first examined the expression of mRNA of genes coding for important pump efflux proteins (Fig. 4A). RT-PCR indicated that ABCG2 was more expressed in tumor-initiating enriched culture cells than in the differentiated cells derived from them. MDR1 was higher expressed only in 83 low-adherence cells but not in cells derived from patient 110.

Figure 4. Molecular studies of the mechanisms involved in the resistance to antitumor drugs and cytotoxic effect of the combination with AZD7762 and DDP on low-adherence cultures. (A) Expression analysis of genes involved in multidrug resistance (ABCG2 and MDR1) and in NER pathway on patient 83 () and patient 110 spheres (▒) and on patient 83 (■) and patient 110 (▓) differentiated cells. Data are expressed as the mean+ SD of the copy number of the gene (calculated from the absolute curve of the gene) normalized by the copy number of two housekeeping genes (actin and cyclofillin), *p < 0.03, **p < 0.01, ***p < 0.001. (B) Western Blot analysis of proteins involved in NER. Lane 1: patient 83 spheres; lane 2: patient 83 differentiated cells; lane 3: patient 110 spheres; lane 4: patient 110 differentiated cells. (C) Left panel: Western Blot analysis of the activation of the DNA damage signaling pathway in response to the DDP treatment. Cells were treated or not (lane 1) with 20 μM DDP and proteins were sampled at 2 h (lane 2), 4 h (lane 3), 24 h (lane 4) and 48 h (lane 5) in both patient 83 (upper panel) and in patient 110 (lower panel) spheres. Right panel: dose/ response curve of spheres treated with DDP, DDP plus AZD27762 (-◊- 100 nM) and DDP plus AZD7735 (-Δ- 200 nM) as detailed in Materials and Methods.

We studied the expression of genes involved in the repair of cross-linking agents, such as genes involved in nucleotide excision repair (NER), homologous recombination (HR) and Fanconi anemia (FA). As reported in Figure 4A, the expression of ERCC1 and XPG mRNAs, which are pivotal in NER, was lower in the more differentiated cells than in cells grown in stem cell conditions. ERCC1 and XPG mRNAs were expressed 64- and 14- and 37- and 5.1-fold more in low adherence cells than in the more differentiated, respectively, in cells from patient 83 and patient 110. These findings were corroborated by western blot analysis of the protein levels of these two proteins (Fig. 4B). The levels of genes involved in HR (BRCA) and FA (FANC-A, FANC-C, FANC-D2 and PALB2) were found to be similar (data not shown).

We also examined activation of the DNA damage signaling pathway after DDP treatment in TIC-enriched cultures; a clear activation of Chk1 and Chk2 was observed after an IC50 DDP dose (dose inhibiting growth by 50%) (Fig. 4C), suggesting that these cells respond by activating the DNA damage signaling pathways, as reported for other normal and tumor cells.18,19 The functional integrity of this DNA damage checkpoint is confirmed by the fact that inhibition of the pathway by the Chk1 inhibitor AZD7762 sensitized cells to DDP, as shown by a shift toward the left of the DDP dose-response curve with two different doses of the inhibitor (Fig. 4C).

Discussion



The recent hypothesis of a cancer stem cell, better defined as a TIC, states that only a minority of cells in the tumor has the ability for indefinite self-renewal and to form tumors when injected in immunodeficient mice, while the majority of cells in the tumor mass are more differentiated cells with an high proliferative rate but low or no tumor-initiating activity. By analogy with the normal stem cell, TICs are more resistant to different treatments (radio and chemotherapeutic drugs).3,5,20-23

Evidence of the existence of TIC in ovarian tumors has been provided, but its definitive characterization is still lacking. The markers used to isolate TICs from the ovary were not specific for the ovary but were taken from studies that identified TICs in leukemia, colon cancer (CD133 and aldehyde dehydrogenase),24-26 breast cancer15,27 and from studies defining the normal stem cell in different tissues. The sphere-formation assay enriches in stem cells and in TICs from different normal human tissues and tumors (CNS, melanoma, ovarian, breast, lung).4,15,28 We prospectively processed 100 fresh tumor samples and used the low-adherence experimental conditions obtaining two cultures, whose “ovospheres” could self-renew, grow unlimitedly, and form tumors when nude mice were transplanted with as few as 1,000 cells. Despite the number of samples processed only two stabilized sphere cultures were obtained; we looked for clinical and/or hystological informations that could identify these samples among other cases, such as for patients’ age, residual tumor after surgery, tumor histological type and grade, tumor stage, response to therapy, but found none. A more specific ovarian-selective culture has probably to be employed and studies are indeed ongoing to validate this. These growing xenografts could be serially propagated and were monomorphic, lacking the polymorphic pattern that characterized the patient’s original tumor. However, they were positive for cytokeratins, even the staining was less marked than in the original tumor, suggesting that TICs could indeed give rise to differentiated cells expressing markers typical of epithelial tumors. Staining experiments with PKH-26 fluorescent dye showed that a minority of PHK-26 stained cells (15%) retained the dye and formed a single sphere, suggesting they underwent an asymmetrical division or a symmetrical one followed by the entrance in a quiescent state of one of the two daughter cells. The majority of cells underwent an extensive replication likely through symmetric divisions giving rise to spheres in which the dye has been diluted. However, the tumorigenic potential of sorted PKH-26high and PKH-26neg cells was similar or even higher in the PKH-26neg sorted cells, suggesting that even cells undergo a symmetric division, they do not lose their tumorigenic potential. These data align with the ones published by Cicalese et al.16 that showed that the self-renewing divisions of mammary cancer stem cells (derived from a syngenic mouse model of breast cancer overexpressing ErbB2) were unlimited, symmetric and more frequent than that in normal murine counterpart.

Spheres cultured in differentiating conditions acquired an epithelial-like morphology, but without clear positivity to different epithelial markers. This might be due to the experimental setting (10–20 d in FBS containing medium), suggesting that more time might be needed for the expression of specific epithelial markers. These differentiated cells lost some of their tumorigenic potential, as suggested by a lower engraftment rate in nude mice (Table 1). Spheres expressed higher levels of mRNA of genes involved in stemness than the differentiated cells derived from them. This confirms that cells with tumor initiating properties derived both from ovarian tumors and other type of tumors express higher levels of transcripts of genes involved in stemness.29,30

Epithelial mesenchymal transition (EMT) is an evolutionary conserved developmental process that renders epithelial cells mobile and invasive (reviewed in ref. 31). In cancer, EMT causes cells to lose epithelial features and to acquire a mesenchymal phenotype responsible for invasion and metastasis.24,32-34 Recently EMT induced by TGFβ or by conditional overexpression of transcriptional factor Snail or Twist, was shown to induce a population with stem cell characteristics in both normal and breast cancer cells.33,35,36 In addition, stem cells isolated from normal breast tissues or cancer breast express a number of canonical EMT markers. The tumor initiating enriched cell cultures we isolated displayed a mesenchymal phenotype and were able to undergo an epithelial transition when cultured in differentiating conditions. In fact, low-adherence cultures expressed vimentin and no E-cadherin and some transcriptional factors (Snail, Twist and Fox2); vimentin was downregulated and E-cadherin upregulated when cells were cultured in differentiating conditions. The phenotypical and morphological changes in the differentiated cells were associated with decreased motility and invasion. These data might explain their greater tumorigenic potential, reinforcing the emerging data on extensive cross-talk networks between the mesenchymal and stem cell phenotype.25 The literature suggests that ovarian epithelium is different from other epithelia in that it displays both epithelial and mesenchymal characteristics and its integrity is maintained primarily by N-cadherin.37 Cancer cells undergo aberrant differentiation during tumorigenesis, and unlike most carcinomas which de-differentiated during progression, ovarian carcinomas undergo transition to a more epithelial phenotype early in tumorigenesis and acquire mesenchymal features in advanced stages.25

The pharmacological characterization of these low-adherence tumor initiating-enriched cultures clearly indicated that these cells were more resistant to drugs currently used in the management of ovarian cancer than the differentiated cells derived from them. These data agree with reports on cancer stem cells derived from other tumor types, for example glioblastoma, breast and lung cancers with the findings that spheres-forming cells express high transcripts of genes involved in stemness. While the resistance to taxol and VP16 could be explained by the higher mRNA levels of ABCG2 pump efflux protein, as already reported in reference 38, the resistance to DDP might be partially due to higher NER mRNA and protein levels (i.e., ERCC1 and XPG). NER is a multistep process able to recognize and repair different types of DNA damage, mainly bulky adducts and DNA crosslinks. A lack or low levels of both ERCC1 and XPG have been related to an increased susceptibility to alkylating agents, including DDP39 and in ovarian cancer low XPG levels were correlated with an increase response to platinum-based therapy.40 The fact that TIC enriched cell cultures were less sentitive to taxol and DDP could explain the high rate of recurrence in this disease. We found that a number of drugs, both DNA interfering agents (yondelis) and targeted agents (including PI3K and proteasome inhibitors) were equally active on both cell populations, suggesting that TICs can be hit by clinically available drugs. In addition, we saw that DDP treatment induced Chk1-dependent checkpoint activation and inhibition of this activation could sensitize cells to the cytotoxic effects of DDP. These findings strengthen the concept that Chk1 inhibitors, which sensitize tumor cell lines to the action of antimetabolite and alkylating agents,41 exert the same effect in the subpopulation of tumor cells with stem cells characteristics. Similar results have been recently published in non-small cell lung cancer stem cells42 possibly reinforcing the clinical use of Chk1 inhibitors with chemotherapy for a more effective cancer treatment.

It has been suggested that acquired drug resistance parallels epigenetic changes resulting in modification of the differentiation state of the tumor (EMT) and in the emergence of chemoresistant cells with stem-cell features.43,44 The TICs we isolated came from patients with stage III ovarian tumor undergoing surgery, but not yet given chemotherapy, suggesting that we had not selected a subpopulation of chemoresistant cells with stem-like characteristics, but that the TICs were resistant and displayed a mesenchymal phenotype.

The data presented here suggest that EOC is potentially a stem cell disease. The low adherence enriched cultures we isolated from fresh ovarian tumor samples containing the ovarian TICs and will be instrumental for the setting up experimental systems to elucidate the role of these cells in the aggressive phenotype of the disease and to find new therapeutic strategies to reduce resistance to current chemotherapeutic drugs. All this should translate into clinically relevant management strategies for recurrence and metastasis.

Materials and Methods



Tumor samples and cell culture.

San Gerardo Hospital Division of Gynecology provided the human ovarian samples, whose use was approved by the local scientific ethic committee with patient’s written consent. Within 48 h of surgery, fresh samples were mechanically disaggregated and enzymatically digested with 2,500 U/mL collagenase I (Sigma) for 1 h at 37°C. The cell suspension was placed in low adherence flasks (Corning) under stem-cell conditions: serum-free DMEM/F12 supplemented with 5 µg/mL insulin (Sigma), 20 ng/mL human recombinant epidermal growth factor (EGF, Peprotech), 10 ng/mL basic fibroblastic growth factor (bFGF, Peprotech) and B27 Supplement (Gibco).

 Immunoflurorescence staining.

Cells were fixed in 4% paraformaldehyde, washed in PBS and blocked at room temperature for 1 h with 1% BSA. Primary antibody incubation followed at room temperature. Anti CD24 (SN3), CD44 (DF1485), Muc-1 (0.N.272) and c-Met (C-12) were supplied by Santa Cruz Biotechnology. Anti E-cadherin (36/E-Cadherin), N-cadherin (32/N-Cadherin) and anti-CD117 (YB5.B8) were purchased from BD Pharmigen. Anti CD133/2 (293C3) was provided by MiltenyiBiotec and anti-Tyr1234/1235 Met by Cell Signaling. Cells were further probed with Alexa Fluor 488 secondary antibody (Invitrogen). Nuclei were stained with DAPI 1:3,000 (Sigma).

 PKH-26 labeling and sorting.

Spheres were mechanically disaggregated and single cells were stained for 5 min with 1:500 PKH-26 dye (Sigma), blocked with 1% BSA, washed three times and plated to obtain spheres in low adherence flasks. Suspensions from single-cell stained with PKH-26 were sorted (Fig. S1) after 7 d in culture based on PKH-26 fluorescence intensity by using a FACS Vantage SE cytometer (Becton Dickinson) equipped with a 488 nm laser (Enterprice Coherent) and a band-pass 585/42 Optical filter.

 Cytotoxicity assay.

Cells from dissociated spheres were cultured in differentiating conditions (DMEM/F12 medium with 10% fetal bovine serum). After one week of culture, differentiated cells and those derived from dissociated spheres were plated in 96 well plates at a concentration of 12,500 cells/mL. 96 h later cells were treated with different antitumor drugs at various doses. In drug combination experiments, cells were pre-treated for 2 h with the Chk1 inhibitor (AZD7762, Axon Medchem) at a concentration of 100 and 200 nM. After this pre-treatment, cells were treated with different doses of cisplatinum (DDP). Cell survival was assessed at 72 h by MTS assay and dose-response curves were plotted.

 Western blot analysis.

Total cellular protein extracts were obtained using a lysis buffer containing 10 mM TRIS-HCl pH 7.4, 150 mM NaCl, 0.1% Nonidet NP-40, 5 mM EDTA, 50 mM NaF with the addition of proteinase inhibitors. 50 µg of proteins were loaded and separated on an 8% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane (PROTRAN, Schleicher and Shull), except for XPG detection where 100 µg of proteins were loaded. Immunoblotting was done using antibodies anti-ERCC1 (FL-297), -XPG (E-18), -Chk1 (G-4), -Chk2 (H-300), -p53 (DO-1), -p21 (C-19), -actin (C-11) and -tubulin (H-235) supplied by Santa Cruz Biotechnology. Anti ser317-Chk1, anti Thr-68 Chk2 were provided by Cell Signaling.

 In vivo tumorigenicity.

Subcutaneous injections (with Matrigel, BD Biosciences) of different numbers of cells (sphere cells, differentiated cells or sorted PHK-26 positive/negative fractions) were made in the flanks of six to eight week old female NCr-nu/nu mice, provided by HARLAN S.p.a. The animals were followed for appearance of the tumor once a week and tumor takes were recordered.

 Immunohistochemical staining.

Tissues were formalin-fixed, paraffin embedded and sectioned into slides (1 µm). Paraffin sections were dewaxed in xylene and rehydrated through decreasing ethanol concentrations, then washed in Tris Buffer Saline (TBS) pH 7.6 twice. Antigen was retrieved for 30 min at 98°C in Antigen Retrieval Buffer pH 6. Endogenous peroxidase was inhibited by incubating the slides in 3% H2O2 for 5 min followed by washing in TBS. The slides were incubated for 30 min with primary antibody. They were then washed and underwent a multi-step process that included incubation with LINK, washing in TBS, 20 min incubation with Streptoavidin peroxidase (LSAB + System-HRP, DAKO) and a final washing in TBS. The chromogen diaminobenzidine was then applied for no longer than 5 min to avoid background signal and the reaction was stopped in water. The sections were counterstained for 5 min with Mayer’s hematoxylin diluted 1:4 in distilled water, dehydrated, mounted using an automated instrument and visualized with a BX60 microscope (Olympus).

 In vitro cell invasion/chemotaxis assay and data analysis.

Cell chemotaxis and invasion were measured using a Boyden chamber as already described in reference 14. Briefly, polycarbonate filter with 8 µm pores (Whatman) separated the cells seeded on the upper side of the membrane. The lower compartments of the filter were filled with DMEM. In the chemotaxis assay, the upper compartments were filled with 50,000 cells suspended in DMEM with 0.5% BSA (Sigma). Cultures were incubated for 4 h at 37°C in a humidified atmosphere of 5% (v/v) CO2 in air. In the invasion assay, the upper side of the chamber was filled with an additional layer of matrigel (10 µL, BD Biosciences) and with 30,000 cells suspended in DMEM with 0.5% BSA. After 6 h incubation, the Boyden chamber was turned over and the cells on the membrane were fixed in methanol, stained with Eosin-based solution (Diff-Quik I, Medion Diagnostics) and then with a Thiazine dye-based solution. Cell migration was assessed by counting the number of cells in eight random fields on the underside of each membrane using a microscope (Olympus CX41) at 40x magnification. Statistical analysis was done using a non-parametric ANOVA test on data from four independent experiments.

 RNA extraction and real time PCR.

RNA was extracted with SV Total RNA Purification Kit (Promega) and 1 ∝g RNA was retro-transcribed to cDNA using High Capacity cDNA Archive Kit (Applied Biosystem). cDNA was then pre-amplified (Preamp Master mix, Applied Biosystem) with two different pools of primers in order to perform expression analysis of several genes of interest. Optimal primer pairs were chosen using PRIMER-3 software (http://frodo.wi.mit.edu) and the specificity was verified by detecting single bands and single dissociation peaks of the PCR products. Absolute copy numbers of mRNA were determined by RT-PCR (ABI-7900, Applied Biosystems) with SYBR Green technique using an EP Motion 5075 robot (Eppendorf) in 384 wells plates. Standard curves for each gene were included for absolute quantification of mRNA. Data were expressed as the mean of the copy number of the gene normalized by the housekeeping genes of three replicates (actin and cyclofillin). Statistical analysis was done using an unpaired-two tailed t-test.

Disclosure of Potential Conflicts of Interest



No potential conflicts of interest were disclosed.

Acknowledgments



The generous contributions of AIRC (The Italian Association for Cancer Research) and the Nerina and Mattioli Foundation are gratefully acknowledged. F.R. is a recipient of a Fellowship from the Monzino Foundation. Lorella Riva, Lab Technician at HSG, helped with histopathology. We are very grateful to Prof. G. Cattoretti for his helpful and constructive discussion.

Note



Supplemental materials can be found at: www.landesbioscience.com/journals/cc/article/20185

References



1. Cadron I, Leunen K, Van Gorp T, Amant F, Neven P, Vergote I. Management of borderline ovarian neoplasms. J Clin Oncol 2007; 25:2928-37; PMID: 17617524; DOI: 10.1200/JCO.2007.10.8076.

2. Boman BM, Wicha MS. Cancer stem cells: a step toward the cure. J Clin Oncol 2008; 26:2795-9; PMID: 18539956; DOI: 10.1200/JCO.2008.17.7436.

3. Alison MR, Lim SM, Nicholson LJ. Cancer stem cells: problems for therapy?. J Pathol 2011; 223:147-61; PMID: 21125672; DOI: 10.1002/path.2793.

4. Hermann PC, Bhaskar S, Cioffi M, Heeschen C. Cancer stem cells in solid tumors. Semin Cancer Biol 2010; 20:77-84; PMID: 20371287; DOI: 10.1016/j.semcancer.2010.03.004.

5. Zhou BB, Zhang H, Damelin M, Geles KG, Grindley JC, Dirks PB. Tumour-initiating cells: challenges and opportunities for anticancer drug discovery. Nat Rev Drug Discov 2009; 8:806-23; PMID: 19794444; DOI: 10.1038/nrd2137.

6. Tysnes BB. Tumor-initiating and -propagating cells: cells that we would like to identify and control. Neoplasia 2010; 12:506-15; PMID: 20651980.

7. Bapat SA, Mali AM, Koppikar CB, Kurrey NK. Stem and progenitor-like cells contribute to the aggressive behavior of human epithelial ovarian cancer. Cancer Res 2005; 65:3025-9; PMID: 15833827.

8. Zhang S, Balch C, Chan MW, Lai HC, Matei D, Schilder JM, et al. Identification and characterization of ovarian cancer-initiating cells from primary human tumors. Cancer Res 2008; 68:4311-20; PMID: 18519691; DOI: 10.1158/0008-5472.CAN-08-0364.

9. Alvero AB, Chen R, Fu HH, Montagna M, Schwartz PE, Rutherford T, et al. Molecular phenotyping of human ovarian cancer stem cells unravels the mechanisms for repair and chemoresistance. Cell Cycle 2009; 8:158-66; PMID: 19158483; DOI: 10.4161/cc.8.1.7533.

10. Alvero AB, Fu HH, Holmberg J, Visintin I, Mor L, Marquina CC, et al. Stem-like ovarian cancer cells can serve as tumor vascular progenitors. Stem Cells 2009; 27:2405-13; PMID: 19658191; DOI: 10.1002/stem.191.

11. Gao MQ, Choi YP, Kang S, Youn JH, Cho NH. CD24+ cells from hierarchically organized ovarian cancer are enriched in cancer stem cells. Oncogene 2010; 29:2672-80; PMID: 20190812; DOI: 10.1038/onc.2010.35.

12. Curley MD, Therrien VA, Cummings CL, Sergent PA, Koulouris CR, Friel AM, et al. CD133 expression defines a tumor initiating cell population in primary human ovarian cancer. Stem Cells 2009; 27:2875-83; PMID: 19816957.

13. Kusumbe AP, Mali AM, Bapat SA. CD133-expressing stem cells associated with ovarian metastases establish an endothelial hierarchy and contribute to tumor vasculature. Stem Cells 2009; 27:498-508; PMID: 19253934; DOI: 10.1634/stemcells.2008-0868.

14. Gatto C, Rieppi M, Borsotti P, Innocenti S, Ceruti R, Drudis T, et al. BAY 12-9566, a novel inhibitor of matrix metalloproteinases with antiangiogenic activity. Clin Cancer Res 1999; 5:3603-7; PMID: 10589777.

15. Ponti D, Costa A, Zaffaroni N, Pratesi G, Petrangolini G, Coradini D, et al. Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties. Cancer Res 2005; 65:5506-11; PMID: 15994920; DOI: 10.1158/0008-5472.CAN-05-0626.

16. Cicalese A, Bonizzi G, Pasi CE, Faretta M, Ronzoni S, Giulini B, et al. The tumor suppressor p53 regulates polarity of self-renewing divisions in mammary stem cells. Cell 2009; 138:1083-95; PMID: 19766563; DOI: 10.1016/j.cell.2009.06.048.

17. Kusumbe AP, Bapat SA. Cancer stem cells and aneuploid populations within developing tumors are the major determinants of tumor dormancy. Cancer Res 2009; 69:9245-53; PMID: 19951996; DOI: 10.1158/0008-5472.CAN-09-2802.

18. Carrassa L, Montelatici E, Lazzari L, Zangrossi S, Simone M, Broggini M, et al. Role of Chk1 in the differentiation program of hematopoietic stem cells. Cell Mol Life Sci 2010; 67:1713-22; PMID: 20146081; DOI: 10.1007/s00018-010-0274-1.

19. Harper LJ, Costea DE, Gammon L, Fazil B, Biddle A, Mackenzie IC. Normal and malignant epithelial cells with stem-like properties have an extended G2 cell cycle phase that is associated with apoptotic resistance. BMC Cancer 2010; 10:166; PMID: 20426848; DOI: 10.1186/1471-2407-10-166.

20. Baumann M, Krause M, Hill R. Exploring the role of cancer stem cells in radioresistance. Nat Rev Cancer 2008; 8:545-54; PMID: 18511937; DOI: 10.1038/nrc2419.

21. Garvalov BK, Acker T. Cancer stem cells: a new framework for the design of tumor therapies. J Mol Med (Berl) 2011; 89:95-107; PMID: 20890588; DOI: 10.1007/s00109-010-0685-3.

22. Schmidt P, Abken H. The beating heart of melanomas: a minor subset of cancer cells sustains tumor growth. Oncotarget 2011; 2:313-20; PMID: 21487158.

23. Di J, Duiveman-de Boer T, Figdor CG, Torensma R. Eradicating cancer cells: struggle with a chameleon. Oncotarget 2011; 2:99-101; PMID: 21378413.

24. Iwatsuki M, Mimori K, Yokobori T, Ishi H, Beppu T, Nakamori S, et al. Epithelial-mesenchymal transition in cancer development and its clinical significance. Cancer Sci 2010; 101:293-9; PMID: 19961486; DOI: 10.1111/j.1349-7006.2009.01419.x.

25. Ahmed N, Abubaker K, Findlay J, Quinn M. Epithelial mesenchymal transition and cancer stem cell-like phenotypes facilitate chemoresistance in recurrent ovarian cancer. Curr Cancer Drug Targets 2010; 10:268-78; PMID: 20370691; DOI: 10.2174/156800910791190175.

26. Kemper K, Grandela C, Medema JP. Molecular identification and targeting of colorectal cancer stem cells. Oncotarget 2010; 1:387-95; PMID: 21311095.

27. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 2003; 100:3983-8; PMID: 12629218; DOI: 10.1073/pnas.0530291100.

28. Eramo A, Lotti F, Sette G, Pilozzi E, Biffoni M, Di Virgilio A, et al. Identification and expansion of the tumorigenic lung cancer stem cell population. Cell Death Differ 2008; 15:504-14; PMID: 18049477; DOI: 10.1038/sj.cdd.4402283.

29. Bertolini G, Roz L, Perego P, Tortoreto M, Fontanella E, Gatti L, et al. Highly tumorigenic lung cancer CD133+ cells display stem-like features and are spared by cisplatin treatment. Proc Natl Acad Sci U S A 2009; 106:16281-6; PMID: 19805294; DOI: 10.1073/pnas.0905653106.

30. Chiou SH, Yu CC, Huang CY, Lin SC, Liu CJ, Tsai TH, et al. Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and high-grade oral squamous cell carcinoma. Clin Cancer Res 2008; 14:4085-95; PMID: 18593985; DOI: 10.1158/1078-0432.CCR-07-4404.

31. Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest 2009; 119:1420-8; PMID: 19487818; DOI: 10.1172/JCI39104.

32. Yang J, Weinberg RA. Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell 2008; 14:818-29; PMID: 18539112; DOI: 10.1016/j.devcel.2008.05.009.

33. Vazquez-Martin A, Oliveras-Ferraros C, Cufí S, Del Barco S, Martin-Castillo B, Menendez JA. Metformin regulates breast cancer stem cell ontogeny by transcriptional regulation of the epithelial-mesenchymal transition (EMT) status. Cell Cycle 2010; 9:3807-14; PMID: 20890129; DOI: 10.4161/cc.9.18.13131.

34. Cufí S, Vazquez-Martin A, Oliveras-Ferraros C, Martin-Castillo B, Joven J, Menendez JA. Metformin against TGFβ-induced epithelial-to-mesenchymal transition (EMT): from cancer stem cells to aging-associated fibrosis. Cell Cycle 2010; 9:4461-8; PMID: 21088486; DOI: 10.4161/cc.9.22.14048.

35. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, et al. The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 2008; 133:704-15; PMID: 18485877; DOI: 10.1016/j.cell.2008.03.027.

36. Radisky DC, LaBarge MA. Epithelial-mesenchymal transition and the stem cell phenotype. Cell Stem Cell 2008; 2:511-2; PMID: 18522839; DOI: 10.1016/j.stem.2008.05.007.

37. Wong AS, Maines-Bandiera SL, Rosen B, Wheelock MJ, Johnson KR, Leung PC, et al. Constitutive and conditional cadherin expression in cultured human ovarian surface epithelium: influence of family history of ovarian cancer. Int J Cancer 1999; 81:180-8; PMID: 10188716; DOI: 10.1002/(SICI)1097-0215(19990412)81:2<180::AID-IJC3>3.0.CO;2-7.

38. Elliott A, Adams J, Al-Hajj M. The ABCs of cancer stem cell drug resistance. IDrugs 2010; 13:632-5; PMID: 20799146.

39. Gossage L, Madhusudan S. Current status of excision repair cross complementing-group 1 (ERCC1) in cancer. Cancer Treat Rev 2007; 33:565-77; PMID: 17707593; DOI: 10.1016/j.ctrv.2007.07.001.

40. Walsh CS, Ogawa S, Karahashi H, Scoles DR, Pavelka JC, Tran H, et al. ERCC5 is a novel biomarker of ovarian cancer prognosis. J Clin Oncol 2008; 26:2952-8; PMID: 18565881; DOI: 10.1200/JCO.2007.13.5806.

41. Ganzinelli M, Carrassa L, Crippa F, Tavecchio M, Broggini M, Damia G. Checkpoint kinase 1 down-regulation by an inducible small interfering RNA expression system sensitized in vivo tumors to treatment with 5-fluorouracil. Clin Cancer Res 2008; 14:5131-41; PMID: 18698031; DOI: 10.1158/1078-0432.CCR-08-0304.

42. Bartucci M, Svensson S, Romania P, Dattilo R, Patrizii M, Signore M, et al. Therapeutic targeting of Chk1 in NSCLC stem cells during chemotherapy. Cell Death Differ 2012; 19:768-78; PMID: 22117197; DOI: 10.1038/cdd.2011.170.

43. Singh A, Settleman J. EMT, cancer stem cells and drug resistance: an emerging axis of evil in the war on cancer. Oncogene 2010; 29:4741-51; PMID: 20531305; DOI: 10.1038/onc.2010.215.

44. Ahmed N, Abubaker K, Findlay J, Quinn M. Epithelial mesenchymal transition and cancer stem cell-like phenotypes facilitate chemoresistance in recurrent ovarian cancer. Curr Cancer Drug Targets 2010; 10:268-78; PMID: 20370691; DOI: 10.2174/156800910791190175.

Top
Back
  • Introduction
  • Results
  • Discussion
  • Materials and Methods
  • References
Jump to Section
Advertisements