Contributed equally
Pancreatic ductal adenocarcinoma (PDAC) is often diagnosed when metastatic events have occurred. Cancer stem cells (CSCs) play an important role in tumor initiation, metastasis, chemoresistance and relapse. A growing number of studies have suggested that CSCs exist in a dynamic equilibrium with more differentiated cancer cells via a bi-directional regeneration that is dependent on the environmental stimuli. In this investigation, we obtain, by using a selective medium, PDAC CSCs from five out of nine PDAC cell lines, endowed with different tumorsphere-forming ability. PDAC CSCs were generally more resistant to the action of five anticancer drugs than parental cell lines and were characterized by an increased expression of EpCAM and CD44v6, typical stem cell surface markers, and a decreased expression of E-cadherin, the main marker of the epithelial state. PDAC CSCs were able to re-differentiate into parental cells once cultured in parental growth condition, as demonstrated by re-acquisition of the epithelial morphology, the decreased expression levels of EpCAM and CD44v6 and the increased sensitivity to anticancer drugs. Finally, PDAC CSCs injected into nude mice developed a larger subcutaneous tumor mass and showed a higher metastatic activity compared to parental cells. The present study demonstrates the ability to obtain CSCs from several PDAC cell lines and that these cells are differentially resistant to various anticancer agents. This variability renders them a model of great importance to deeply understand pancreatic adenocarcinoma biology, to discover new biomarkers and to screen new therapeutic compounds.
Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal disease characterized by early spread to local and distant organs, and most patients having an unresectable tumor at the time of diagnosis (
Recently, gene expression analyses of highly aggressive tumors have shown a compelling overlap of their gene expression profiles with those of normal stem cells (
The high level of heterogeneity of CSCs, which originates from genotypic and phenotypic plasticity, and their low presence in cancer sample tissues, make their isolation and identification a strong limitation for the realization of biochemical studies. Thus, in order to obtain valid and reproducible results, the biochemical approach to study CSC pathophysiology can take advantage from the observation that CSCs can be isolated and enriched from several human cancer cell lines (
The present study was conducted to achieve the following three primary goals: i) to test the capacity of several PDAC cell lines to generate CSCs in selective medium; ii) to characterize biological features of the obtained CSCs, both
Gemcitabine (Jemta; Sandoz) and zoledronic acid (Zometa; Novartis) were solubilized in water, sorafenib (BAY43-9006; Bayer AG), tipifarnib (Zarnestra; Johnson & Johnson), and everolimus (RAD001; Novartis) were solubilized in DMSO. Gemcitabine was stored at −80°C, zoledronic acid, everolimus and tipifarnib at −20°C, sorafenib at room temperature until use.
The human cell lines PaCa44, HPAF-II, PT45P1, CFPAC1, PSN1, PC1J, PaCa3, Panc1, MiaPaCa2 (pancreatic adenocarcinoma cell lines), and VIT-1 (normal primary pancreatic mesenchymal cells) were grown in RPMI-1640 supplemented with 10% FBS, 2 mM glutamine, and 50 μg/ml gentamicin sulfate (Gibco/Life Technologies). Adherent cells were maintained in standard conditions for a few passages at 37 °C with 5% CO2. To generate suspension cells and separate stem-like sphere-growing cells, adherent cells were washed twice in 1X PBS (Gibco/Life Technologies) and then cultured in CSC medium, i.e., DMEM/F-12 without glucose (US Biological Life Sciences) supplemented with 1 g/l glucose, B27, 1 μg/ml fungizone, 1% penicillin/streptomycin (all from Gibco/Life Technologies), 5 μg/ml heparin (Sigma), 20 ng/ml epidermal growth factor (EGF), and 20 ng/ml fibroblast growth factor (FGF) (both from PeproTech). Adherent cells were left in CSC medium for at least 1–3 weeks or until the appearance of floating cell aggregates, referred to as tumorspheres. Tumorspheres were cultured in CSC medium for at least three passages before initiating the characterization experiments.
PSN1, PC1J, PaCa3, Panc1, and MiaPaCa2 CSCs were plated in 96-well cell culture plates (3×103 cells/well) and incubated at 37°C with 5% CO2 in CSC medium. Tumorspheres were counted after 5 days.
PSN1, PC1J, PaCa3, Panc1, and MiaPaCa2 parental cell lines, CSCs, and ex-CSCs were plated in 96-well cell culture plates (5×103 cells/well) and incubated at 37°C with 5% CO2. After 24 h the cells were treated with increasing dose of five drugs (gemcitabine, zoledronic acid, sorafenib, tipifarnib, and everolimus). After 72 h, cell viability was measured by Cell Proliferation Reagent WST-1 (Roche Diagnostics).
For proliferation cell assay, Panc1 parental cell line and CSCs were plated in 96-well cell culture plates (5×103 cells/well) and incubated at 37°C with 5% CO2. Viable cells were counted by Trypan Blue dye exclusion after 1, 2, 3, 4, and 7 days. The doubling time was calculated using the formula T = (T2-T1) × log 2/log (Q2/Q1), where: T1, day 3; T2, day 7; Q1, cell number at day 3; and Q2, cell number at day 7. For CSCs, this experiment was performed using CSC medium containing 1 or 3 g/l glucose and similar results were obtained.
Panc1 parental cell line and CSCs were harvested, washed, resuspended in 1X PBS and stained for 30 min on ice with FITC-conjugated monoclonal anti-CD326 (anti-EpCAM, no. 130-098-113), PE-conjugated monoclonal anti-CD133/2 (no. 130-090-853) (both from Miltenyi Biotech), PE-Cy7-conjugated polyclonal anti-CD66c (no. bs-6032R; Bioss, Inc.), APC-H7-conjugated monoclonal anti-CD44 (no. 560532; BD Biosciences), FITC-conjugated monoclonal anti-CD44v6 (no. ab30437; Abcam), and PE-Cy7-conjugated monoclonal anti-CD24 (no. 311119; BioLegend). Isotype- matched irrelevant antibodies were used as negative controls.
Approximately 20,000 gated events were acquired for each sample on a BD FACSCanto (BD Biosciences) and analyzed using FlowJo software (TreeStar, Inc.). Dead cells and debris were excluded based upon forward scatter and side scatter measurements.
Total RNA was extracted from 106 cells using TRIzol Reagent (Life Technologies) and 1 μg of RNA was reverse transcribed using first-strand cDNA synthesis. Real-time quantification was performed in triplicate samples by SYBR-Green detection chemistry with Power SYBR-Green PCR Master Mix (Applied Biosystems) on a 7000 Sequence Detection System. The primers used were: E-cadherin forward, 5′-GAC ACC AAC GAT AAT CCT CCG A-3′ and reverse, 5′-GGC ACC TGA CCC TTG TAC GT-3′; ribosomal protein large P0 (RPLP0) forward, 5′-ACA TGT TGC TGG CCA ATA AGG T-3′ and reverse, 5′-CCT AAA GCC TGG AAA AAG GAG G-3′.
The following cycling conditions were used: 95°C for 10 min, 40 cycles at 95°C for 15 sec, 60°C for 1 min, 95°C for 15 sec, and 60°C for 15 sec. The average of cycle threshold of each triplicate was analyzed according to the 2−ΔΔCt method.
Cells were collected, washed in 1X PBS, and resuspended in RIPA buffer, pH 8.0 (150 mM NaCl, pH 8.0; 50 mM Tris-HCl; 1% Igepal; 0.5% Na-Doc; and 0.1% SDS), 1 mM PMSF, 1 mM Na3VO4, 1 mM NaF, 2.5 mM EDTA, and 1X protease inhibitor cocktail (Calbiochem; Merck Millipore) for 30 min on ice. The lysate was centrifuged at 2,300 × g for 10 min at 4°C and the supernatant was used for western blot analysis. Protein concentration was measured with the Bradford Protein Assay Reagent (Thermo Fisher Scientific) using bovine serum albumin as a standard. Thirty micrograms of protein extracts were electrophoresed through a 10% SDS-polyacrylamide gel and electroblotted onto PVDF membranes (Merck Millipore). Membranes were then incubated for 1 h at room temperature with blocking solution, i.e., 3% low-fat milk in TBST (100 mM Tris, pH 7.5, 0.9% NaCl, and 0.1% Tween-20), and probed overnight at 4°C with the monoclonal rabbit E-cadherin primary antibody (1:20,000 in blocking solution, no. ab40772; Abcam). Horseradish peroxidase conjugated IgG polyclonal (1:8,000 in blocking solution, no. 12348; Merck Millipore) was used to detect specific proteins. Immunodetection was carried out using chemiluminescent HRP substrates (Merck Millipore) and recorded using an Amersham Hyperfilm ECL (GE Healthcare). To quantify E-cadherin expression, bands were scanned as digital peaks and the areas of the peaks were calculated in arbitrary units using the public domain NIH Image software (
Panc1 parental cell line and CSCs at three different dilutions (1×104, 1×105 and 1×106 cells/mice) were s.c. injected into the dorsal flank of five nude female mice for each condition (5 weeks of age; Charles River Laboratories, Inc.). For the control groups, mice received 100 μl injections of 1X PBS. Body mass was recorded weekly for each animal. Tumor size was monitored weekly in two perpendicular dimensions parallel to the surface of the mouse using a caliper. Tumor volume was calculated using the formula of V = π/6 × [(w × L)^(3/2)]. Animals were sacrificed at the volume of 2 cm3. Immediately after death, neoplastic masses were collected for flow cytometry and histological assessment. To perform flow cytometry analysis, tumor masses were dissociated enzymatically and mechanically (MACS Dissociation kit; Miltenyi Biotech) and 5×105 of the obtained cells were resuspended in 100 μl of RPMI without phenol red (Gibco/Life Techonologies) and then analyzed with 7-AAD (BD Biosciences) and with the antibodies described in flow cytometry analysis section. To perform histological analysis, tissue samples not used in the cytometry analysis were fixed in 10% (v/v) neutral-buffered formalin for 24–48 h and were processed routinely. Serial histological sections (4–6 μm thick) were obtained from each paraffin block and stained with hematoxylin and eosin (H&E) for histology assessment. Animal studies were approved by the Verona University Review Board.
Mice were randomly allocated into the two groups (n=5 mice/group) and were anesthetized with 1.5% isofluorane-air mixture. As previously described (
ANOVA (post hoc Bonferroni) analysis was performed by GraphPad Prism 5 (GraphPad Software, Inc.). P-values <0.05, 0.01, or 0.001 show significant difference.
In order to evaluate whether PDAC cell lines were able to dedifferentiate into cancer stem-like cells, we cultured nine PDAC cell lines (PaCa44, HPAF-II, PT45P1, CFPAC1, PSN1, PC1J, PaCa3, Panc1, MiaPaCa2), and one normal primary pancreatic mesenchymal cell line (VIT-1) in a selective medium containing EGF, FGF, and low glucose (1 g/l). Only five of them, all of tumoral origin (PSN1, PC1J, PaCa3, Panc1, MiaPaCa2), lost their characteristic epithelial morphology and were able to form tumorspheres after 1–3 weeks of culturing (
To evaluate the effect of the phenotypic transformation from epithelial to sphere morphology on cell growth, we measured the proliferation time of Panc1 cancer stem-like cells compared to the parental cell line and we observed that the doubling time was 3.26 and 1.83 days, respectively (
Since it is known that CSCs are more resistant to chemotherapeutic treatments, to compare the chemosensitivity of the parental cell lines and the corresponding cancer stem-like cells, we evaluated the anti-proliferative activity of a panel of five anticancer compounds, gemcitabine (pyrimidine nucleoside analogue, gold standard treatment for PDAC), ti pifarnib (farnesyl transferase inhibitor), sorafenib (multi-kinase inhibitor), everolimus (mTOR inhibitor), and zoledronic acid (farnesyl diphosphate synthase inhibitor). We treated PSN1, PC1J, PaCa3, MiaPaCa2, Panc1 parental cells, the corresponding cancer stem-like cells, and ex-CSCs for 72 h with the compounds and we determined the IC50 values, which are shown in
We characterized Panc1 cancer stem-like cells for the expression of various typical surface stem cell markers. The percentage of cells expressing EpCAM, CD44v6, CD44, CD133, CD66, and CD24 was determined in comparison to the parental cell line (
In order to evaluate whether in the transition non-CSC to CSC the Panc1 cancer stem-like cells were subjected to EMT, we analyzed the expression of E-cadherin, the main marker of the epithelial state (
Re-differentiation of PDAC CSCs into adherent cells (ex-CSCs), morphologically observed in
To evaluate the tumor-initiating capabilities of pancreatic CSCs
To evaluate whether injected CSCs maintained stem cell marker expression
To evaluate the metastatic activity of pancreatic CSCs, we used an
Many studies have identified subpopulations of cells within tumors that drive tumor growth and recurrence, termed CSCs. These cells are resistant to the majority of current cancer treatments, including radiation and chemotherapy, suggesting that most of the cancer therapies, while killing the bulk of tumor cells, may ultimately fail because they do not eliminate CSCs, which survive to regenerate new tumors. Consequently, the development of a reliable model of CSCs becomes crucial for basic and clinical cancer research.
Several techniques have been used to isolate CSCs from cancers (
In the current study, we first established the ability of several pancreatic cell lines to form spheres with the aim to correlate this ability to cell resistance to several chemotherapeutic agents. We found that only five out of nine PDAC cell lines had the capacity to form spheres after 1–3 weeks of culturing in a selective medium and the remaining cell lines were unable to form spheres even in other stem cell media. Panc1 cells, described as particularly resistant to gemcitabine (
Evidence of enhanced therapeutic resistance to CSCs has been reported (
Several surface markers have been used to identify and isolate CSCs, including CD24 (
When subcutaneously injected in nude mice, Panc1 CSCs developed larger tumor masses composed by a homogeneous population of cancer cells characterized by larger size compared to the heterogeneous population generated by parental cells. This result suggests a significant alteration of the mechanisms regulating the metabolic and cell division pathways. Interestingly, the tumor masses originated from CSCs and parental cells show a decreased expression of EpCAM and CD44 compared to the cells analyzed
Taken together, our present study has demonstrated that the pancreatic CSCs isolated from PDAC cell lines have all the characteristics of the clinically relevant tumors. This model will be of great importance to deepen our understanding of the biology of pancreatic adenocarcinoma and will also be employed to early marker discovery and screening of compounds for therapeutic intervention.
This study was supported by AIRC-Fondazione CariPaRo, Padova, Italy; AIRC 5 per mille grant no. 12182; Fondazione Cariverona, Project Verona Nanomedicine Initiative. We thank Dr Dea Filippini for her support and suggestions during the
(A) Phase-contrast microscopy images of the indicated parental cell lines, tumorspheres [cancer stem cells (CSCs)], and reverted tumorspheres (ex-CSCs). (B) Tumorsphere formation assay of PSN1, PC1J, PaCa3, Panc1 and MiaPaCa2 CSCs. Tumorsphere number is the mean ± SEM of three independent experiments each performed 5-fold. *P<0.05.
(A) Proliferation curve of Panc1 parental cells (P) and cancer stem cells (CSCs). Viable cells were counted by Trypan Blue dye exclusion. Values are the means ± SEM of four independent experiments. (B) Panc1 P and CSCs doubling time ± SEM was obtained from the data of (A). **P<0.01, ***P<0.001.
(A) Representative flow cytometry curves of Panc1 P, cancer stem cells (CSCs), and ex-CSCs stained with the indicated experimental antibodies (blue) or the negative-control IgG antibody (red). The percentage of positive cells is reported on the flow cytometry curves. (B) Histograms of CD44v6 and EpCAM protein expression in Panc1 P, CSCs, and ex-CSCs. Values are the means ± SEM of four independent experiments and are reported as the ratio between median fluorescence intensity (RMFI) obtained with the indicated experimental antibodies versus the negative-control IgG antibody (fold change relative to Panc1 P). (C) Histograms of E-cad mRNA expression in Panc1 P and CSCs. Real-time PCR values are the means ± SEM of three independent experiments each performed in triplicate and are reported as fold change relative to Panc1 P. (D) Western blot analysis of E-cad performed on Panc1 P and CSCs. The bands were scanned as digital peaks and the areas of the peaks were calculated in arbitrary units, as described in Materials and methods. The value of Ponceau S was used as a normalizing factor. Values are the means of three independent experiments. **P<0.01, ***P<0.001.
Parental and cancer stem cell (CSC) subcutaneous growth in nude mice. (A) The number of mice-developing tumors of total injected mice is indicated. Panc1 P or CSCs (1×106 cells/mouse) were subcutaneously injected into female nude mice. Values are the means ± SEM of (B) mouse tumor volume or (C) body mass measured every week. (D) Hematoxylin and eosin (H&E) staining of representative tumor tissues derived from Panc1 P and CSCs (scale bar, 100 μm). (E) Histograms of the percentage of positive cells for CD44 and EpCAM protein expression in dissociated masses derived from Panc1 P and CSCs. *P<0.05, **P<0.01 and ***P<0.001.
Parental and cancer stem cell (CSC) metastatic growth in nude mice. (A) Optical images of two representative mice injected with MiaPaCa2-RFP P and CSCs at 14, 18 and 22 days after cell inoculation into the spleen. Fluorescence images are superimposed on the photographic images in order to localize the cancer masses. False colors represent the radiance (photons/sec/cm2/sr) of the light emitted as indicated in the color bar on the right. (B) Mean total flux ± SEM expressed in photons/sec measured in the two experimental groups. *P<0.05.
IC50 ± SEM values at 72 h in CSCs, parental cell lines and ex-CSCs.
IC50 | Gemcitabine (μM) | Tipifarnib (μM) | Sorafenib (μM) | Everolimus (μM) | Zoledronic acid (μM) |
---|---|---|---|---|---|
PSN1 | |||||
CSC | >2.5 |
3.1±2.7 |
20±3.9 |
8.2±0.3 | >500 |
Parental cell line | 1±0.13 | 17±2.5 | 2.4±0.05 | 7.9±1.3 | 395±57 |
ex-CSC | 1.1±0.15 | 11.8±0.71 | 4.7±0.34 | 13.8±3.6 | 455±64 |
PC1J | |||||
CSC | >5 | 16±0.6 | 18.9±0.55 |
15.7±0.4 | 87±4.1 |
Parental cell line | >5 | 20.5±3.3 | 10±1.2 | 15.7±0.3 | >250 |
ex-CSC | >5 | 23±0.83 | 13.7±1.3 | 18±1.3 | >250 |
PaCa3 | |||||
CSC | 2.5±0.78 | 9.3±2.5 |
>25 |
17.9±0.9 |
80±31 |
Parental cell line | 3.5±0.24 | 1±1.5 | 12.6±3.1 | 9.6±0.8 | 45±26 |
ex-CSC | 3.4±0.58 | 2.8±0.61 | 13.4±3.2 | 9.9±1.3 | 62±71 |
MiaPaCa2 | |||||
CSC | 4.8±0.82 | 16.8±1.7 | 10.3±1.6 | 11.5±3.3 | 80±17 |
Parental cell line | 5±0.74 | 18.4±1.2 | 8.6±1.1 | 15.3±0.8 | >500 |
ex-CSC | >5 | 16±1.3 | 8.2±0.95 | 15.8±0.26 | >500 |
Panc1 | |||||
CSC | >5 | >25 |
24±0.8 |
20±1.6 |
237±97 |
Parental cell line | >5 | 17.5±0.89 | 8.5±0.7 | 11.6±1.6 | 250±87 |
ex-CSC | >5 | 19.7±2.7 | 15.7±3.0 | 18.8±1.7 | >250 |
P<0.05 CSCs vs. parental cell line.
CSC, cancer stem cell.