Annals of Surgical Oncology

, Volume 14, Issue 12, pp 3629–3637

Development and Characterization of Gemcitabine-Resistant Pancreatic Tumor Cells

Authors

  • Ami N. Shah
    • Department of Surgical OncologyUniversity of Texas M.D. Anderson Cancer Center
    • Department of Cancer BiologyUniversity of Texas M.D. Anderson Cancer Center
  • Justin M. Summy
    • Department of Cancer BiologyUniversity of Texas M.D. Anderson Cancer Center
  • Jing Zhang
    • Department of Cancer BiologyUniversity of Texas M.D. Anderson Cancer Center
  • Serk In Park
    • Department of Cancer BiologyUniversity of Texas M.D. Anderson Cancer Center
  • Nila U. Parikh
    • Department of Cancer BiologyUniversity of Texas M.D. Anderson Cancer Center
    • Department of Cancer BiologyUniversity of Texas M.D. Anderson Cancer Center
Laboratory Research Original Papers

DOI: 10.1245/s10434-007-9583-5

Cite this article as:
Shah, A.N., Summy, J.M., Zhang, J. et al. Ann Surg Oncol (2007) 14: 3629. doi:10.1245/s10434-007-9583-5

Abstract

Background

Pancreatic cancer is an exceptionally lethal disease with an annual mortality nearly equivalent to its annual incidence. This dismal rate of survival is due to several factors including late presentation with locally advanced, unresectable tumors, early metastatic disease, and rapidly arising chemoresistance. To study the mechanisms of chemoresistance in pancreatic cancer we developed two gemcitabine-resistant pancreatic cancer cell lines.

Methods

Resistant cells were obtained by culturing L3.6pl and AsPC-1 cells in serially increasing concentrations of gemcitabine. Stable cultures were obtained that were 40- to 50-fold increased in resistance relative to parental cells. Immunofluorescent staining was performed to examine changes in β-catenin and E-cadherin localization. Protein expression was determined by immunoblotting. Migration and invasion were determined by modified Boyden chamber assays. Fluorescence-activated cell sorting (FACS) analyses were performed to examine stem cell markers.

Results

Gemcitabine-resistant cells underwent distinct morphological changes, including spindle-shaped morphology, appearance of pseudopodia, and reduced adhesion characteristic of transformed fibroblasts. Gemcitabine-resistant cells were more invasive and migratory. Gemcitabine-resistant cells were increased in vimentin and decreased in E-cadherin expression. Immunofluorescence and immunoblotting revealed increased nuclear localization of total β-catenin. These alterations are hallmarks of epithelial-to-mesenchymal transition (EMT). Resistant cells were activated in the receptor protein tyrosine kinase, c-Met and increased in expression of the stem cell markers CD (cluster of differentiation)24, CD44, and epithelial-specific antigen (ESA).

Conclusions

Gemcitabine-resistant pancreatic tumor cells are associated with EMT, a more-aggressive and invasive phenotype in numerous solid tumors. The increased phosphorylation of c-Met may also be related to chemoresistance and EMT and presents as an attractive adjunctive chemotherapeutic target in pancreatic cancer.

Keywords

Pancreatic cancerEMTGemcitabine chemoresistancec-MetCancer stem cells

One of the major challenges in the treatment of pancreatic cancer is the frequent failure of chemotherapy. Gemcitabine, the current standard of care, has only a 5.4% partial response rate1 and imparts a progression-free survival interval ranging from only 0.9 to 4.2 months.2 The poor response rate and short progression-free interval suggests that pancreatic cancer either rapidly develops or has intrinsic gemcitabine chemoresistance. The mechanisms by which chemoresistance arises in pancreatic cancer are unknown; thus a better understanding of how resistance arises and what molecular alterations cause or correlate with resistance is likely to lead to novel therapeutic strategies for pancreatic cancer.

Recent studies in solid-tumor cell lines have linked chemoresistance to epithelial-to-mesenchymal (EMT)-like transition. Yang et al. recently demonstrated that colon carcinoma cells undergo EMT when chronically exposed to oxaliplatin.3 Also, in breast cancer cell lines a link has been shown between tamoxifen-resistant cell lines and EMT, conferring a more-motile and invasive phenotype.4,5 EMT is a fundamental embryological process characterized by alterations in morphology, cellular architecture, adhesion, and migration potential.6 When EMT occurs in tumor cells it imparts a more migratory, invasive, and metastatic phenotype. Molecular markers for EMT include increased expression of vimentin, a mesenchymal marker, nuclear translocation of β-catenin, and increased production of transcription factors that repress E-cadherin expression, including Twist, Snail, and Slug.6 While the underlying mechanisms for these alterations are unknown, they appear to involve alterations in β-catenin signaling, E-cadherin function, and activation of the transcription factors Twist, Slug, and/or Snail.6,7

Growth factors and their tyrosine kinase receptors have been reported to be important for the induction of EMT.6,8,9 One of the growth factors implicated in EMT is hepatocyte growth factor (HGF), also known as scatter factor.6,10 HGF is primarily produced by mesenchymal cells (such as fibroblasts) and acts in a paracrine or endocrine fashion on epithelial cells through the HGF receptor, c-Met11. Activation of the c-Met tyrosine kinase leads to a wide array of biological activities including motility, migration, growth, angiogenesis, and invasion. These processes are crucial for embryonic development, organ formation, wound healing, and tissue regeneration.1215 However, c-Met activation also significantly enhances the metastatic potential of epithelial tumor cells in vitro.16 In cancer cells, as in normal cells, HGF acts as a motogen, inducing separation and scatter of the cells, leading to the formation of metastases in epithelial tumors.17 Because of these properties, c-Met activation by HGF has been linked to EMT, and in fact addition of HGF to cell cultures is one of the most frequently utilized assays for inducing EMT in vitro.

In this study, we isolated gemcitabine-resistant cells from well-characterized pancreatic tumor cell lines and determined they had many EMT-like properties. We demonstrate that aside from emergence of fibroblastic characteristics, specific activation of the protein tyrosine kinase c-Met consistently occurs as cells acquire gemcitabine resistance. Characterization of these stable pancreatic tumor cells provides new insights into phenotypic changes associated with gemcitabine resistance.

PATIENTS AND METHODS

Cell lines and Culture Conditions

The human pancreatic cancer cell line L3.6pl, a derivative of COLO-357, was obtained from I. J. Fidler, D.V.M., Ph.D. (The University of Texas M.D. Anderson Cancer Center, Houston, TX). The human pancreatic cancer cell line AsPC-1 was obtained from American Type Culture Collection (Manassas, VA). L3.6pl cells were cultured in minimal essential medium (MEM) supplemented with 10% fetal bovine serum (FBS), l-glutamine, and penicillin-streptomycin. AsPC-1 cells were cultured in Roswell Park Memorial Institute (RPMI) medium supplemented with 15% FBS, l-glutamine, and penicillin-streptomycin at 37°C in 5% CO2 and 95% air. In vitro experiments were done at 50–70% confluence. Results from all studies were confirmed in at least three independent experiments.

Pharmacologic Agents

Gemcitabine (Gemzar, Eli Lilly, Indianapolis, IN, USA) was obtained from the M.D. Anderson hospital pharmacy. Stock solutions of gemcitabine were made fresh every seven days in 0.9% normal saline and added to the tissue culture plates upon each media change.

Morphologic Analysis

Cells were grown to 70% confluence in the appropriate gemcitabine-free media for the parental cell lines and media with gemcitabine for the gemcitabine-resistant cell lines and visualized via light microscopy (Nikon Microphot-FX, Japan). Digital pictures were taken from a camera mounted to the microscope (Nikon Cool-Pix 4500, Japan).

Migration and Invasion Assays

Cell migration and invasion were assessed with modified Boyden chamber (Becton Dickinson Labware, Bedford, MA) assays as described by Minard et al. (35). Briefly, 105 L3.6pl parental or gemcitabine-resistant L3.6pl cells in MEM plus 1% FBS were seeded onto 8.0-μm pore size membrane inserts in 24-well plates. MEM plus 10% FBS was added to the bottom wells as a chemoattractant. After 12, 24, and 36 hours, cells that did not migrate were removed from the top side of the inserts with a cotton swab. Cells that had migrated to the underside of the inserts were stained with HEMA 3 (Biochemical Sciences, Swedesboro, NJ, USA) according to the manufacturer’s instructions. The migratory cells were counted under a microscope at 20× magnification. Cell images were obtained using a video camera (SONY DXC-990, Japan) mounted to a microscope (Nikon Microphot-FX, Japan). Cells were counted in five random fields per insert. The invasion assay was done in a similar fashion except the 8.0-μm pore size membrane inserts were coated with Matrigel. Results were expressed as cells migrated per field.

Fluorescent Immunohistochemistry

Immunofluorescence staining was performed as described previously.18 Briefly, 5,000 cells in 0.2 mL of MEM media with 10% FBS were plated per chamber in an eight-chamber slide. After 24 hours incubation at 37°C, cells were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS), permeabilized in PBS containing 0.5% Triton X-100 and 1% bovine serum albumin (BSA), and blocked with 10% fetal bovine serum (FBS) in PBS. Fixed cells were incubated with anti-vinculin (Sigma), anti-E-cadherin (BD Transduction) and/or anti-β-catenin (Upstate Biotechnology Incorporated) antibodies. Cells were washed and incubated with Alexa Fluor 488 (green)- or Alexa Fluor 594 (red)-conjugated anti-mouse or anti-rabbit IgG antibody (Jackson ImmunoResearch Laboratories) diluted 1:100 in blocking buffer. Nuclei were stained with Hoechst 33342 (Molecular Probes, Eugene, OR) at 1 mg/mL. Slides were mounted with mounting media (20 mM Tris pH 8.0, 0.5% N-propyl gallate, 90% glycerol) and examined with an epifluorescence microscope equipped with narrow band-pass excitation filters (Chroma Technology) to individually select for green, red, and blue fluorescence. Cells were observed through a Hamamatsu C5810 camera (Hamamatsu Photonics K.K., Bridgewater, NJ) mounted on a Nikon Microphot-FXA microscope (Nikon, Japan) and images were captured with the Optimas image-analysis software (Media Cybernetics, Silver Springs, MD, USA).

Immunoblotting

Immunoblotting was performed as described previously.19 Briefly, cells in log growth phase at 70% confluence were rinsed twice with ice-cold PBS and then lysed with radioimmunoprecipitation assay (RIPA) B lysis buffer [20 mM sodium phosphate, 150 mM NaCl, 5 mM sodium pyrophosphate, 5 mM ethylenediamine tetraacetic acid (EDTA), 1% Triton X-100, 0.5% sodium deoxycholate 0.1% sodium dodecyl sulfate (SDS)] supplemented with one tablet complete mini-EDTA-free protease inhibitor cocktail (Roche Diagnostic, Manheim, Germany) and 1 mM sodium orthovanadate (pH 7.4). Cells were harvested with the aid of a rubber policeman, clarified by centrifugation at 13,000 rpm for 15 minutes at 4°C, and prepared for immunoblot analysis. Total proteins (50 μg) from clarified cell lysates were separated via 8% SDS polyacrylamide gel electrophoresis (SDS-PAGE) and electroblotted onto polyvinylidene difluoride membranes (Amersham, Chicago, IL, USA). The membranes were blocked with Tris-buffered saline with Tween (0.15%) plus 5% dried non-fat milk for 30 minutes at room temperature and probed with the desired primary antibody diluted 1:1000 in blocking buffer overnight at 4°C. Membranes were probed with antibodies to E-cadherin, β-catenin, phosphorylated c-Met, and c-Met (Cell Signaling Technology, Danvers, MA), Twist (Santa Cruz Biotechnology, Santa Cruz, CA, USA), vinculin, and actin (Sigma-Aldrich, St. Louis, MO). Primary antibody incubation was followed by incubation with a horseradish peroxidase-conjugated secondary antibody (goat anti-mouse, sheep anti-rabbit, Bio-Rad laboratories, Hercules, CA, USA) diluted 1:3000 in blocking buffer for 1 hour at room temperature. Proteins were visualized with chemiluminescence detection reagents (Perkin-Elmer, Boston, MA, USA) and detected by autoradiography. Nuclear fractions were prepared using the Active Motif nuclear fractionation kit (Active Motif, Carlsbad, CA, USA) following the manufacturer’s instructions. Lamin B antibody (as a loading control for nuclear proteins) was also purchased from Active Motif.

Fluorescence-Activated Cell Sorting (FACS)

Cells were grown to 70% confluence. Cells were then trypsined and washed with FACS buffer (1× PBS, 5% FBS, 0.1% sodium azide) twice. The cells were then resuspended in FACS buffer. Antibodies were added and incubated for 1 hour on ice, and the sample was washed twice with FACS buffer. When needed, a secondary antibody was added by resuspending the cells in FACS buffer followed by a 30-minute incubation. After another washing, cells were resuspended in FACS buffer. The antibodies used were anti-CD44 allophycocyanin, anti-CD24 (phycoerythrin), and anti-H2K (PharMingen, Franklin Lakes, NJ) as well as anti-ESA-FITC (Biomeda, Foster City, CA), each at a dilution of 1:40. Flow cytometry was done using a FACSAria (BD Immunocytometry Systems, Franklin Lakes, NJ). Side scatter and forward scatter profiles were used to eliminate cell doublets. Cells were routinely sorted twice.

RESULTS

Development of Gemcitabine-Resistant Cell Lines

To create stable pancreatic cancer cell lines chronically resistant to gemcitabine, L3.6pl and AsPC-1 cells were exposed to increasing concentrations of gemcitabine. Specifically, cells in log phase were first exposed to 25 nM gemcitabine for L3.6pl cells and 50 nM gemcitabine for AsPC-1 cells, which resulted in greater than 95% cell death. Once surviving cells reached 80% confluence, they were passaged twice in this same concentration of gemcitabine, after which the process was repeated at increasing doses of gemcitabine until a cell population was selected that demonstrated at least a 50-fold greater IC 50 to gemcitabine than the parental cell lines. Resultant L3.6pl and AsPC-1 cells were resistant to 1000 nM gemcitabine. The two resulting cell lines were designated L3.6pl GR and AsPC-1 GR. The gemcitabine-resistant phenotype has been stable over 20 passages.

Gemcitabine Chemoresistant Cells Have Undergone the Morphologic Changes Consistent with EMT or HGF Stimulation

As shown in Fig. 1, the L3.6pl and AsPC-1 resistant cells are morphologically distinct from their respective parental cell lines. The resistant cells demonstrate a loss of cell-cell adhesion, spindle shaped morphology, and increased formation of pseudopodia (Fig. 1— see black arrows). These changes are typical of the scattering of cells treated with HGF,20 in which a more-mesenchymal phenotype arises. However, unlike the transient morphologic alterations observed with HGF, the gemcitabine-resistant cells have maintained this phenotype through more than 20 passages.
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-007-9583-5/MediaObjects/10434_2007_9583_Fig1_HTML.gif
FIG. 1.

Morphologies of gemcitabine-sensitive and gemcitabine-resistant cells. Cells were grown to 60% confluency and then photographed under 40× magnification.

Gemcitabine-Resistant Cells Have Increased Migratory and Invasive Potential

A characteristic of mesenchymal-like cells is increased migratory and invasive potentials. To test these properties, modified Boyden chamber assays were performed on parental and gemcitabine-resistant cells as described in Patients and Methods. As shown in Fig. 2 for L3.6pl cells, the gemcitabine-resistant cells demonstrate a 10-fold increase at 12 hours, a fivefold increase at 24 hours, and a 20-fold increase at 36 hours. The L3.6pl GR cells were increased twofold, eightfold, and sevenfold in invasion at 12, 24, and 36 hours respectively relative to parental L3.6pl cells. Similar results were obtained for AsPC-1 and AsPC-1 GR cells (data not shown).
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-007-9583-5/MediaObjects/10434_2007_9583_Fig2_HTML.gif
FIG. 2.

Migration and invasion of gemcitabine-sensitive and gemcitabine-resistant L3.6pl cells. For these assays, 100,000 cells in 1% FBS were placed on each 8.0-μm pore size membrane insert in 24-well plates without (for migration assays) or with (for invasion assays) matrigel. MEM with 10% FBS was placed in the bottom wells as chemoattractants. (A) Cells per 20× high-power field migrating at indicated time periods. (B) Photomicrographs of HEMA-3-stained migrated cells. (C) Cells per 20x high-power field invading through matrigel at indicated time periods. (D) Photomicrographs of HEMA-3-stained migrated cells. P-values were all less than 0.001 for migration and invasion assays. Similar results were obtained for AsPC-1 and AsPC-1 GR cell lines.

Gemcitabine-Resistant cells Are Altered in Localization of E-Cadherin and β-Catenin

Two of the hallmarks of EMT are β-catenin nuclear translocation and localization of E-cadherin away from the plasma membrane. To examine these properties, we performed immunofluorescence on parental and gemcitabine-resistant cells (Fig. 3). Changes in β-catenin localization are also evident (Fig. 3A). In parental cells, β-catenin is localized both in the plasma membrane and the cytoplasm. In contrast, in gemcitabine-resistant cells β-catenin is located almost exclusively in the nucleus where it can promote transcription of factors that increase migration, invasion, and some of the molecular signals important for EMT. E-cadherin in the parental cells is located at the cell–cell junctions that are important for cell-cell linkage and inhibition of cell migration (Fig. 3B) and in the gemcitabine-resistant cell lines no longer located at the plasma membrane and is relocated to the cytoplasm where it can no longer function as an adhesion molecule.
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-007-9583-5/MediaObjects/10434_2007_9583_Fig3_HTML.gif
FIG. 3.

Confocal microscopy of gemcitabine-sensitive and gemcitabine-resistant L3.6pl cells. (A). Immunofluorescence for β-catenin and actin and their colocalization. (B) Immunofluorescence for E-cadherin and α actin and their colocalization. All photographs were taken with identical exposure times.

Gemcitabine-Resistant Cells Express EMT Markers

Immunoblotting was performed to examine expression of E-cadherin, β-catenin, and the transcription factors Snail, Slug, and Twist. In both L3.6pl and AsPC-1 resistant cells, E-cadherin expression was greatly reduced (Fig. 4A). β-catenin translocation to the nucleus is also evident. In both parental cell lines β-catenin is found primarily in the cytosolic compartment. However, β-catenin is localized primarily in the nucleus in both L3.6pl and AsPC-1 gemcitabine-resistant cell lines. No difference was noted in the total pool of β-catenin, demonstrating that the differences in localization were not due to degradation. No obvious differences in expression of total Snail, Slug and Twist were evident (data not shown); however, elevated levels of Twist were detected in the nuclear fraction of gemcitabine-resistant cell lines (Fig. 4A), suggesting that expression of this transcription factor may be important to gemcitabine-induced EMT in pancreatic cancer.
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-007-9583-5/MediaObjects/10434_2007_9583_Fig4_HTML.gif
FIG. 4.

(A) Lysates from cells grown to 70% confluence were prepared and immunoblotting with the antibodies indicated was performed as described in Patients and Methods. Vinculin was used as a loading control. Nuclei were isolated, cell lysates were prepared, and immunoblotting was performed with the indicated antibody as described in Patients and Methods. Lamin B was used as a loading control for nuclear proteins. These experiments were representative of blots of at least three independent experiments. (B) Total cell lysates were prepared and immunoblotting was performed with antibodies to tyrosine phosphorylated c-Met and total c-Met. Vinculin was performed as total protein loading control. These experiments were representative of blots of at least three independent experiments.

Expression and Phosphorylation of c-Met Receptor

As activation of c-Met is known to drive EMT in several cell culture systems, we examined the activity of c-Met in gemcitabine-resistant clones and their respective parental cell lines via phosphorylation specific immunoblot analysis. Interestingly, there was a marked increase in the tyrosine phosphorylation of c-Met consistent with constitutive activation of the receptor in both gemcitabine-resistant cell lines (Fig. 4B). This was not due to elevated c-Met expression, as total c-Met levels were unchanged. Increased c-Met phosphorylation also did not appear to be due to elevated HGF production, as no differences in HGF levels from cell culture supernatants of parental or gemcitabine-resistant cells were detected (data not shown).

Expression of Stem Cell Markers in Gemcitabine-Resistant Cells

A newly evolving area in cancer research is stem cell populations. Li et al.21 have demonstrated that in pancreatic cancer a side population of cells that are CD24, CD44, and ESA positive are highly tumorigenic and may be cancer stem cells. In our study, it is possible that during our gemcitabine-resistant cell selection process that we have selected for pancreatic cancer stem cells. Cancer stem cells have been shown to be inherently chemoresistant and we may have selected for this cell type.22,23 Cancer stem cells may also have a plastic morphology that allows them to undergo EMT in order to invade and metastasize. Flow cytometry data from our laboratory suggests that we do have an increased amount of cancer stem cells in gemcitabine-resistant cell population (Table 1).
TABLE 1.

Percent positive stem cell markers by flow cytometry

 

CD24 +

CD44 +

ESA +

Triple positive

L3.6pl

0.2%

28.9%

1.8%

0.2%

L3.6pl GR

15.1%

34.9%

15.2%

14.2%

AsPC-1

0.4%

16.7%

2.4%

0.4%

AsPC-1 GR

18.6%

35.2%

14.3%

13.1%

DISCUSSION

Pancreatic adenocarcinoma is a highly aggressive disease and patients with this malignancy have only a 4% overall five-year survival rate.24 These dismal survival rates highlight the need for new treatment options for surgically unresectable patients. Many different chemotherapeutic agents have failed to demonstrate any survival advantage in patients with pancreatic adenocarcinoma. Gemcitabine, the current standard of care, imparts improvement only in symptoms of the disease, with little effect on tumor burden. Improved therapeutic treatment will require a better understanding of the mechanisms by which these tumors become chemoresistant, and development of strategies to overcome this resistance. To further study these mechanisms, we developed two gemcitabine-resistant pancreatic cancer cell lines. Both stable cell lines had two characteristic properties: (1) morphologic and molecular alterations consistent with EMT and (2) activation of the c-Met tyrosine kinase. In addition, there was an increase in stem cell markers in these cells.

The mechanisms regulating EMT have yet to be fully defined. A number of factors, including transforming growth factor (TGF)-β, β-integrins, activated growth factor receptors, and activated Src have been associated with EMT in various cell types. The plasticity of the phenotype would suggest that the microenvironment also plays a key role in this process, but is also consistent with the increase in stem cell markers we observe in our gemcitabine-resistant cells. Thus, in our study and that of Yang et al.,3 a chemotherapeutic agent was sufficient to induce (or select for) cells able to undergo this transition. Further studies will be required to determine whether these changes require distinct genetic mutations in tumor cells, epigenetic changes, or both. With respect to gemcitabine resistance in cultured pancreatic tumor cells, alterations in enzymes affecting gemcitabine transport and metabolism have been recently observed.25 Whether these changes were genetic or epigenetic remain to be determined and effects on EMT characteristics were not examined in this study. Nevertheless, it seems likely that multiple mechanisms may lead to gemcitabine resistance, and that sustained resistance to this and other chemotherapeutic agents is associated with EMT.26

HGF activation of c-Met induces many of the cellular alterations associated with EMT. Shibamoto et al.27 reported that HGF induces phosphorylation of β-catenin that decreases the E-cadherin concentration at the plasma membrane and the number of cell-cell adhesion sites. Hiscox et al.28 reported that HGF enhances tyrosine phosphorylation of β-catenin resulting in the dissociation of β-catenin from E-cadherin. The activation of this pathway can also cause the nuclear translocation of β-catenin independent of the Wnt pathway.29As demonstrated in previous studies, the nuclear translocation of β-catenin is fundamentally important for EMT.30,31 As only marginal changes were observed in our promoter/reporter assays for β-catenin, whether it functions as a transcription factor to upregulated mesenchymal genes remains to be determined.

Howe et al.32 reported that another factor that may be transcriptionally upregulated by β-catenin is Twist, a basic helix–loop–helix transcription factor characterized by a basic DNA-binding domain that targets a consensus E-box sequence. The Twist promoter contains an element responsive to β-catenin.32 Elevated Twist expression positively correlates with aggressiveness of cancer and poor survival.33 Twist functions to repress E-cadherin expression, by binding directly or indirectly through the E-box elements on the E-cadherin promoter.30 Additionally, Twist siRNA knockdown in metastatic mammary tumors prevented metastasis.30 Recently Twist has been implicated in paclitaxel resistance in breast cancer and siRNA knockdown of Twist caused a restoration of paclitaxel sensitivity and a decrease in migration and invasion.34

Finally, FACS analysis of stem cell markers demonstrated marked increase in expression of CD24-, CD44-, and ESA-positive cells. Thus, selection may have resulted in enrichment of a minor fraction of pancreatic tumor stem cells in the sensitive population. We are currently isolating the stem cell-like population and its characterization is likely to provide new insights with respect to the association of the stem cell phenotype and EMT.

In summary, our study suggests a link between c-Met activation, gemcitabine resistance, tumor aggressiveness, potential increase in cancer stem cells, and EMT. c-Met activation may lead to a more-invasive, migratory phenotype through multiple signaling pathways, (Shah, unpublished data) including through the induction of β-catenin translocation affecting specific transcriptional programs such as Twist activation, that are associated with EMT and drug resistance. The molecular basis for activation of the HGF/c-Met axis promoting EMT is unclear and the precise genes that this altered program trigger responsible for gemcitabine resistance remain to be investigated.

Whether c-Met activation is necessary and/or sufficient for gemcitabine resistance and EMT is currently under investigation in the laboratory. Regardless, we demonstrate that critical aspects of EMT are associated with chemoresistance to therapeutic agents, and that novel approaches to overcoming such resistance may be important in treatment of pancreatic and other tumors in which EMT plays a prominent role during tumor progression.

Acknowledgments

This research was supported in part by NIH U54 CA 090810 (G.E.G), NIH T32 CA 09599 (A.N.S.), the Eleanor B. Pillsbury Fellowship-University of Illinois Hospital (A.N.S.), the Lockton Fund (G.E.G., J.M.S.), and G.E.G is the Sowell-Huggins Professor and J.Z. the Sowell-Huggins Fellow in Cancer Biology.

Copyright information

© Society of Surgical Oncology 2007