Characterization of Tumor-Derived Pancreatic Stellate Cells

Journal of Surgical Research 157, 96–102 (2009) doi:10.1016/j.jss.2009.03.064

ASSOCIATION FOR ACADEMIC SURGERY Characterization of Tumor-Derived Pancreatic Stellate Cells Buckminster Farrow, M.D.,*,1 David Rowley, Ph.D.,† Truong Dang, B.S.,† and David H. Berger, M.D.* *Michael E. DeBakey Department of Surgery, Michael E. DeBakey, Houston, Texas; and †Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas Submitted for publication January 6, 2009

INTRODUCTION Background. Pancreatic stellate cells (PSCs) are key mediators of the desmoplastic reaction that characterizes pancreatic adenocarcinoma. We sought to isolate and characterize tumor-derived pancreatic stellate (TDPS) cells to further understand how these stromal cells influence pancreatic cancer behavior. Methods. We established a stable line of non-immortalized PSCs from a patient with pancreatic adenocarcinoma using a modified prolonged outgrowth method. Cell staining for cytokeratin, vimentin, and alpha smooth muscle actin (aSMA) was performed. Total RNA was harvested from TDPS and panc-1 cells and gene expression determined by microarray analysis. Results. TDPS cells contain lipid droplets in the cytoplasm, and later stain positive for both vimentin and aSMA, indicative of activated myofibroblasts. Microarray analysis revealed a distinct gene expression profile compared with pancreatic cancer cells, including expression of proteases that facilitate cancer cell invasion and growth factors known to activate pancreatic cancer cells. Additionally, TDPS cells expressed many of the key components of the pancreatic tumor stroma, including collagen, fibronectin, and S100A4, confirming their importance in the tumor microenvironment. Conclusions. Characterization of tumor-derived PSCs will facilitate further studies to determine how the tumor microenvironment promotes the aggressive behavior of pancreatic cancer. Published by Elsevier Inc. Key Words: pancreatic ductal carcinoma; stromal cells; cytoskeleton, cell lines.

Pancreatic cancer remains the most lethal abdominal malignancy, despite recent improvements in multimodal therapy. Surgical resection is the only curative treatment for pancreatic cancer, however, only a fraction of patients are candidates for complete removal of the tumor. Surgery is often precluded due to local invasion of adjacent critical structures. Those patients who do survive a potentially curative pancreaticoduodenectomy have a 5-y survival rate of less than 25% due to the high rate of local recurrence [1]. Adjuvant therapy is only of marginal benefit in extending the disease-free interval or preventing recurrence [2]. The observation that most pancreatic cancer patients eventually die with local recurrence or regional (hepatic) metastases suggests either the nature of pancreatic cancer cells or the regional environment that surrounds them fosters the progression of locoregional microscopic disease. In epithelial-derived cancers, the overwhelming majority of experimental investigation focuses on tumor cells alone, excluding the other cellular and extracellular elements that form the tumor microenvironment [3, 4]. Recently investigators have shown that stromal signals from cells and other extracellular matrix proteins are critical determinants of tumor behavior regulating growth, invasion, and angiogenesis, thus confirming a critical role for the tumor microenvironment in the clinical manifestations of adenocarcinoma [5, 6]. Despite the abundance of desmoplasia and stromal cells found in pancreatic cancer specimens, and the fact that fibrosis within the pancreas increases the risk of pancreatic cancer [7], studies of the tumor–stroma interactions in pancreatic adenocarcinoma have lagged behind [8–10]. The preponderance of evidence now suggests that the pancreatic cancer microenvironment is

1

To whom correspondence and reprint requests should be addressed at Michael E. DeBakey VA Medical Center, OCL 112,2002 Holcombe Blvd., Houston, TX 77030. E-mail: [email protected].

0022-4804/09 $36.00 Published by Elsevier Inc.

96

FARROW ET AL.: TUMOR-DERIVED PANCREATIC STELLATE CELLS

a key regulator of the rapid growth and invasion that characterizes pancreatic adenocarcinoma [11–15]. Unfortunately, studies of the tumor microenvironment are currently limited by very few in vitro and in vivo models that accurately represent the reactive stroma seen in human disease. Use of well-characterized tumor-derived pancreatic stromal cell lines in vitro and in vivo will greatly enhance the accuracy of experimental models. Pancreatic stellate cells (PSCs) are key mediators of the desmoplastic reaction that characterizes pancreatic adenocarcinoma [16]. PSCs also mediate the deposition of collagen in chronic pancreatitis [17]. These cells are classified as myofibroblasts and express mesenchymal type filaments, such as vimentin and desmin. PSCs can be activated by cytokines and growth hormones, leading to increased expression of another cytosolic filamentous protein, alpha smooth muscle actin (a-SMA) [18]. Whether from rat or human pancreata, these cells can be difficult to isolate and maintain in culture, limiting experimental investigation into how PSCs may facilitate tumor progression. Previously, investigators who have isolated human pancreatic stellate cells have noted senescence within several passages, limiting their usefulness for ongoing experiments [20]. The majority of these studies used either pancreatic tissue from benign pancreatic fibrosis or portions of ‘‘normal’’ pancreas away from a resected pancreatic tumor. We hypothesized that PSCs isolated from pancreatic tumors would replicate more readily, using an outgrowth method we have used previously to isolate prostate myofibroblasts [21]. We demonstrate in this study that these PSCs can be derived from human tumors, maintained in culture for months, and represent a phenotypically and genetically unique component of the pancreatic tumor microenvironment. METHODS Cell Culture Panc-1 cells were purchased from ATCC (Rockville, MD) and grown in IMDM (Gibco/Invitrogen), 10% FBS with penicillin/streptomycin/ mycostatin at 37  C with 5% CO2. The tumor-derived pancreatic stellate (hereafter abbreviated TDPS) cells were also grown in IMDM 10%FBS with penicillin/streptomycin/mycostatin at 37  C with 5% CO2. Unless otherwise noted co-culture experiments were conducted in IMDM with 5% FBS and no antibiotics.

Outgrowth Method of Isolating Tumor-Derived Pancreatic Stellate Cells A small fragment of non-necrotic pancreatic adenocarcinoma was obtained immediately after surgical resection under a protocol approved by the Baylor College of Medicine IRB. The male patient was confirmed to have T3N2M0 pancreatic adenocarcinoma on pathologic diagnosis, and did not receive any preoperative chemotherapy or radiotherapy. The tissue was washed in PBS, then minced into

97

1–2 mm fragments, and placed in polylysine coated plastic dishes with DMEM media containing 10% FBS and pen/strep/mycostatin at 37  C 5% CO2. The media was changed every few days when significant cellular debris was noted, and fragments were separated into individual dishes after 1 wk. Minimal outgrowth was seen after 14 d, and the tissue was left undisturbed excluding media changes. By 21 d, rapid outgrowth was noted from 30% of the tissue fragments; these fragments were placed in new dishes to facilitate further outgrowth. By day 25–28, the cells had become confluent in the 50 mm dishes, and were gently trypsinized and transferred to standard cell culture flasks for further passage and freezing. Cell lines were maintained in IMDM with 10% FBS and penicillin/streptomycin/mycostatin.

Immunohistochemistry Antibodies specific for vimentin, 1:50 (sc-7557; Santa Cruz Biotechnology, Santa Cruz, CA); FITC-conjugated a-SMA,1:200 (F3777; Sigma, St. Louis, MO) and pan-cytokeratin, 1:100 (F0877; Sigma) were used as we have previously described [19]. Fluorescently labeled secondary antibodies were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA), including FITC anti-mouse and Texas Red anti-goat 711-295-152. No significant staining was observed with secondary antibody alone. Pancreatic stellate cells and/or panc-1 cells were grown on coverslips and fixed with 4% paraformaldehyde after being washed with PBS. Immunostaining with protein specific primary and fluorescently labeled secondary antibodies for vimentin and cytokeratin was performed as we have described previously [19]. Briefly, cells were rehydrated then blocked for 30 min in 1% donkey serum in PBS. Primary antibody in 1% BSA was incubated on the coverslip for 1 h at 37  C. After washing, secondary antibody was applied for 45 min in the dark. After a final washing, coverslips were mounted with Vecta-shield containing DAPI (Vector Labs, Burlingame, CA).

Gene Array All reagents and supplies were purchased from Superarray (Frederick, MD) unless otherwise specified. Total RNA (1 mg) was amplified and used to synthesize cDNA probes for gene array analysis using the True Labeling-AMP 2.0 kit. To compare differences in gene expression between tumor and stromal components, RNA was isolated from Panc-1 cells and TDPS cells, respectively. Microarray analysis was performed using the Human Cancer Pathway Finder Gene Array according to the manufacturer’s protocol. Briefly, biotinylated cDNA probes were hybridized to membranes that contain complementary sequences from 96 genes related to cancer growth, cell cycle regulation, and metastases. Membranes were washed after overnight hybridization, and signals were detected using a nonradioactive chemiluminescent detection system. Images were obtained by exposing the membranes to Kodak BioMax film (Kodak, Rochester, NY) then digitizing the images from all the membranes simultaneously on a scanner. Comparisons of expression were performed after background levels were subtracted and values were normalized to housekeeping gene expression (GAPDH and b-actin) on each array using the GE Array Expression Analysis Suite. Probe synthesis and hybridization were performed in duplicate to ensure consistency of results. To determine whether TDPS cells express genes in the tumor microenvironment that stimulate the progression of pancreatic cancer, microarray analysis was again performed. Total RNA from TDPS cells was subjected to analysis with the Human Chemokine and Receptor Array, the Human Tumor and Metastasis Array, and the Human Extracellular Matrix and Adhesion Molecules Array as described above.

Ethics Informed consent was obtained from each patie, and the study protocol conforms to the ethical guidelines of the World Medical Association

98

JOURNAL OF SURGICAL RESEARCH: VOL. 157, NO. 1, NOVEMBER 2009

FIG. 1. Outgrowth of TDPS cells from a human pancreatic tumor. Tumor fragments from a pancreatic adenocarcinoma were subjected to standard cell culture conditions, resulting in the outgrowth of mesenchymal like cells. These cells demonstrate lipid droplets in their cytoplasm, characteristic of pancreatic stellate cells. Declaration of Helsinki, Ethical Principles for Medical Research Involving Human Subjects, adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964, as revised in Tokyo, 2004, as reflected in a priori approval by the appropriate institutional review committee.

RESULTS Tumor-Derived Pancreatic Stellate Cells Can Be Isolated and Maintained Under Standard Cell Culture Conditions

After 21 d of incubation, pancreatic stellate cells were found growing from 30% of the tissue fragments (Fig. 1). Several of these cultures were discarded due to fungal contamination. After 28 d and removal of any contaminated samples, three separate cell populations were obtained, hereafter designated TDPS cells. The cell population was morphologically homogeneous, i.e., no epithelial appearing cells were seen upon harvesting the TDPS cells. After an initial passage, these TDPS cells would grow to confluence in 5–7 d under standard culture conditions as described in Methods, (Fig. 2). We have successfully maintained one of these cell lines for at least 15 passages with only minimal diminution in the growth rate (data not shown). Additionally, TDPS cells can be frozen in liquid nitrogen and fully recovered in growth media, although this does appear to induce earlier senescence. Tumor-Derived Pancreatic Stellate Cells are Phenotypically Myofibroblasts

Mesenchymal cells can be distinguished from epithelial cells by the expression of cytoskeletal filaments such as vimentin or desmin. Further characterization into myofibroblasts is determined when markers of both mesenchymal cells and smooth muscle cells, such as calponin and aSMA, are expressed simultaneously

FIG. 2. TDPS cells in monolayer. Following initial isolation TDPS cells were grown in standard polystyrene coated flasks in IMDM with 10% FBS and antibiotics. Note the lipid droplets still present in the cytoplasm of some cells indicating a quiescent stellate cell phenotype.

[22]. Additionally due to potential transdifferentiation or aberrant expression of mesenchymal cytoskeletal filaments by pancreatic cancer cells, it is important to confirm that TDPS cells are not epithelial in origin, i.e., do not express cytokeratins. Morphologically, the TDPS cells appear as myofibroblasts with lipid droplets in the cytoplasm, and stain positive for vimentin (Fig. 3A). Cells also stain positive for aSMA, a marker of activated myofibroblasts [18] (Fig. 3B), and do not express cytokeratin compared with adjacent panc-1 cells (Fig. 3C), distinguishing TDPS cells as distinct from epithelial tumor cells. TDPS Cells are Genetically Distinct from Pancreatic Cancer Cells

Previously, we have reported that the cellular component of the pancreatic tumor stroma is genetically distinct from the epithelial component of the pancreatic tumor [10]. To further define the genetic differences between epithelial and stromal components, and to provide additional evidence for the unique pattern of gene expression seen in pancreatic stellate cells, we isolated total RNA from both TDPS and panc-1 cells. Microarray analysis was then performed, and genes expressing more than 3-fold higher or lower in the TDPS cells compared with the panc-1 cells were considered to be significant differences in expression (Fig. 4). Notably, we discovered that several proteases (TIMP-1, MMP-1) and extracellular matrix proteins (thrombospondin) were more highly expressed in TDPS cells, consistent with previous reports that PSCs are the primary source of proteases in pancreatic tumors [23]. Conversely, expression of genes related to apoptosis resistance (bcl-2), invasion (MTA-2), and downstream targets of ras (phosphoinositide-3-kinase, catalytic subunit) were more than 3-fold higher in panc-1 cells. Also

FARROW ET AL.: TUMOR-DERIVED PANCREATIC STELLATE CELLS

99

cellular elements of pancreatic cancer stroma. Additionally TDPS should express genes shown to be expressed in pancreatic stellate cells isolated by other investigators. To further characterize TDPS cells, microarray analysis was performed to determine the expression of cytokines, growth factors, ECM proteins, adhesion molecules, and proteases in these cells. The results of the microarrays were compared with known expression profiles of pancreatic stellate cells and cellular elements of the pancreatic cancer stroma found in the literature. TDPS cells expressed unique elements of the pancreatic cancer stroma, including S100A4 and connective tissue growth factor (Table 1). Additionally, TDPS cells expressed collagen, fibronectin, and laminin transcripts, which are ECM proteins found in the pancreatic cancer stroma, and were shown previously to be secreted by pancreatic stellate cells. These results further confirm that TDPS cells are pancreatic stellate cells that express the same gene products found in pancreatic cancer stroma.

DISCUSSION

FIG. 3. TDPS cells express vimentin, a-smooth muscle actin, but not cytokeratin. After passage 5-7 TDPS cells were grown on glass coverslips, immunostained for filamentous proteins, and nuclei counterstained with DAPI. Vimentin expression (A) was found in all TDPS cells while alpha smooth muscle actin expression (B) was found in most cells. TDPS cells were also grown together with panc-1 cells and immunostained for cytokeratins (C). All panc-1 cells (small arrows) demonstrated cytokeratin expression while TDPS cells did not (large arrow).

of note was higher expression of PDGF in panc-1 cells, which is a known activator of PSCs [24]. Genes Found in Pancreatic Cancer Stroma and Pancreatic Stellate Cells are Expressed by TDPS Cells

If TDPS cells are pancreatic stellate cells, they should have a phenotypic profile similar to that described for

The role of the tumor microenvironment in the progression of pancreatic cancer has recently been further defined. Both cellular and extracellular elements contribute to the rapid growth and aggressive invasiveness that characterizes this lethal disease. Success in targeting pancreatic cancer cells in clinical trials has overall been disappointing, suggesting a novel approach may be necessary to find more effective treatments for pancreatic adenocarcinoma. Targeting the tumor stroma in other epithelial cancers has been successful in preclinical models [25, 26], thus similar approaches may be effective against pancreatic cancer. To investigate potential treatments and design preclinical models of the human disease, cellular elements of the pancreatic cancer microenvironment must be isolated and characterized. In this study, we demonstrate that a tissue-specific myofibroblast, the pancreatic stellate cell, can be isolated from human pancreatic tumors. Additionally, we show that these TDPS cells express vimentin and aSMA (but not cytokeratins), confirming a myofibroblastic phenotype. Use of these human stromal cells extracted directly from the pancreatic tumor microenvironment will allow investigators to determine how pancreatic stellate cells contribute to the aggressive behavior of pancreatic cancer. Other investigators have defined several important functions for pancreatic stellate cells in pancreatic cancer progression. Rat pancreatic stellate cells can be activated by conditioned medium from pancreatic cancer cells to induce expression of TIMP-1[27] and collagen [16], both of which may facilitate cancer cell invasion

100

JOURNAL OF SURGICAL RESEARCH: VOL. 157, NO. 1, NOVEMBER 2009

FIG. 4. Patterns of gene expression in TDPS and panc-1 cells. TDPS and panc-1 cells were grown under standard conditions and RNA isolated to perform microarray analysis and elucidate differential patterns of gene expression. Fold differences were calculated after background levels of expression were subtracted and intensity normalized for housekeeping gene expression.

[23]. Pancreatic stellate cells can also produce cytokines [28] and matrix metalloproteases (MMPs) [29], which can promote the progression of pancreatic cancer [30–33]. We also demonstrate in this study, important differences in gene expression between TDPS and pancreatic TABLE 1 Genes Expressed By Tumor-Derived Pancreatic Stellate Cells Gene

Role in tumor microenvironment

Plasminogen activator, urokinase receptor

Key factor in pancreatic cancer metastasis [36] expressed in stromal fibroblasts [37] Correlated with stroma surrounding colorectal cancer liver metastases [38] Homologue of mouse stromal cell line derivative [39] Abundant ECM protein in pancreatic cancer stroma [40] Expressed by activated pancreatic stellate cells [29] Secreted by pancreatic stellate cells to produce stroma [41] Found in pancreatic cancer stroma [42] Expressed by pancreatic stellate cells in pancreatic tumor stroma [11] Expressed by pancreatic myofibroblasts [28] Expressed in pancreatic cancer stroma [43]

Interleukin 8

Stromal-derived factor 2 Fibronectin 1 Matrix metalloproteinase-2 Collagen type I, IV, V, VI, VIII, IX, XII, XV, XVIII Laminin Connective tissue growth factor

Monocyte chemotactic protein 1 S100A4

cancer cells. Not only do the microarray results further support our hypothesis that TDPS cells are a distinct cell population of mesenchymal origin, the pattern of expression is consistent with the current theories about the role of stromal cells in cancer progression. Specifically, we observed a 19-fold increase in MMP-1 expression and a 65-fold increase in the expression of tissue-inhibitor of matrix metalloproteinases (TIMP)3 in TDPS cells compared with pancreatic cancer cells. MMP-1 can facilitate cancer cell invasion, and pancreatic stellate cell expression of MMP-1 can be increased by treatment from tumor cell conditioned medium [29]. TIMPs are correlated with metastasis of pancreatic cancer cells [34], and expression of other TIMPs (TIMP-1) can be increased in pancreatic stellate cells by exposure to panc-1 conditioned medium [27]. Thrombospondin-1 and thrombospondin-2 expression was increased in TDPS cells, and thrombospondins have been shown to facilitate pancreatic cancer cell invasion [35]. One possible limitation of our study is the outgrowth method we used to isolate human tumor-derived pancreatic stellate cells. The prolonged outgrowth method may select for cells that like to grow on plastic, excluding other populations of mesenchymal cells that may also be important cellular components of the pancreatic cancer microenvironment. Additionally, these TDPS may have developed genetic mutations (similar to those seen in the epithelial compartment of the tumor), which would partially explain the hardiness of these cells, but would also profoundly affect the patterns of gene expression.

FARROW ET AL.: TUMOR-DERIVED PANCREATIC STELLATE CELLS

In conclusion, we demonstrate that the key mediator of the desmoplastic reaction in pancreatic cancer, the pancreatic stellate cell, can be isolated from human pancreatic tumors. These TDPS cells exhibit a unique phenotype characterizing them as myofibroblasts, distinct from epithelial cancer cells. We expect that use of TDPS cells or cell lines using these techniques by other investigators will facilitate the development of models of pancreatic cancer that accurately mimic the aggressive human disease. These models can then be used to develop and test novel treatments targeting the tumor microenvironment, with the goal to improve the number of patients who can undergo curative pancreatic resection. REFERENCES 1. Riall TS, Nealon WH, Goodwin JS, et al. Pancreatic cancer in the general population: Improvements in survival over the last decade. J Gastrointest Surg 2006;10:1212. 2. Mulcahy MF, Wahl AO, Small W. Jr The current status of combined radiotherapy and chemotherapy for locally advanced or resected pancreas cancer. J Natl Compr Cancer Netw 2005;3:637. 3. Bhowmick NA, Moses HL. Tumor-stroma interactions. Curr Opin Genet Dev 2005;15:97. 4. Shekhar MP, Werdell J, Santner SJ, et al. Breast stroma plays a dominant regulatory role in breast epithelial growth and differentiation: Implications for tumor development and progression. Cancer Res 2001;61:1320. 5. Cheng N, Bhowmick NA, Chytil A, et al. Loss of TGF-beta type II receptor in fibroblasts promotes mammary carcinoma growth and invasion through upregulation of TGF-alpha-, MSP- and HGF-mediated signaling networks. Oncogene 2005;24:5053. 6. Tuxhorn JA, Ayala GE, Rowley DR. Reactive stroma in prostate cancer progression. J Urol 2001;166:2472. 7. Lowenfels AB, Maisonneuve P. Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol 2006; 20:197. 8. Ricci F, Kern SE, Hruban RH, Iacobuzio-Donahue CA. Stromal responses to carcinomas of the pancreas: Juxtatumoral gene expression conforms to the infiltrating pattern and not the biologic subtype. Cancer Biol Ther 2005;4:302. 9. Crnogorac-Jurcevic T, Efthimiou E, Nielsen T, et al. Expression profiling of microdissected pancreatic adenocarcinomas. Oncogene 2002;21:4587. 10. Farrow B, Sugiyama Y, Chen A, et al. Inflammatory mechanisms contributing to pancreatic cancer development. Ann Surg 2004;239:763. discussion 769. 11. Hartel M, Di Mola FF, Gardini A, et al. Desmoplastic reaction influences pancreatic cancer growth behavior. World J Surg 2004; 28:818. 12. Qian LW, Mizumoto K, Maehara N, et al. Co-cultivation of pancreatic cancer cells with orthotopic tumor-derived fibroblasts: Fibroblasts stimulate tumor cell invasion via HGF secretion whereas cancer cells exert a minor regulative effect on fibroblasts HGF production. Cancer Lett 2003;190:105. 13. Sato N, Maehara N, Goggins M. Gene expression profiling of tumor-stromal interactions between pancreatic cancer cells and stromal fibroblasts. Cancer Res 2004;64:6950. 14. Vonlaufen A, Joshi S, Qu C, et al. Pancreatic stellate cells: Partners in crime with pancreatic cancer cells. Cancer Res 2008; 68:2085.

101

15. Hwang RF, Moore T, Arumugam T, et al. Cancer-associated stromal fibroblasts promote pancreatic tumor progression. Cancer Res 2008;68:918. 16. Apte MV, Park S, Phillips PA, et al. Desmoplastic reaction in pancreatic cancer: Role of pancreatic stellate cells. Pancreas 2004;29:179. 17. Apte MV, Wilson JS. Mechanisms of pancreatic fibrosis. Dig Dis 2004;22:273. 18. Apte MV, Haber PS, Darby SJ, et al. Pancreatic stellate cells are activated by proinflammatory cytokines: Implications for pancreatic fibrogenesis. Gut 1999;44:534. 19. Tuxhorn JA, Ayala GE, Smith MJ, et al. Reactive stroma in human prostate cancer: Induction of myofibroblast phenotype and extracellular matrix remodeling. Clin Cancer Res 2002; 8:2912. 20. Jesnowski R, Furst D, Ringel J, et al. Immortalization of pancreatic stellate cells as an in vitro model of pancreatic fibrosis: Deactivation is induced by matrigel and N-acetylcysteine. Lab Invest 2005;85:1276. 21. Ayala GE, Wheeler TM, Shine HD, et al. In vitro dorsal root ganglia and human prostate cell line interaction: Redefining perineural invasion in prostate cancer. Prostate 2001;49:213. 22. Bachem MG, Schneider E, Gross H, et al. Identification, culture, and characterization of pancreatic stellate cells in rats and humans. Gastroenterology 1998;115:421. 23. Armstrong T, Packham G, Murphy LB, et al. Type I collagen promotes the malignant phenotype of pancreatic ductal adenocarcinoma. Clin Cancer Res 2004;10:7427. 24. Luttenberger T, Schmid-Kotsas A, Menke A, et al. Platelet-derived growth factors stimulate proliferation and extracellular matrix synthesis of pancreatic stellate cells: Implications in pathogenesis of pancreas fibrosis. Lab Invest 2000;80:47. 25. Loeffler M, Kruger JA, Niethammer AG, et al. Targeting tumorassociated fibroblasts improves cancer chemotherapy by increasing intratumoral drug uptake. J Clin Invest 2006; 116:1955. 26. WillhauckMJ,MiranceaN,VosselerS,etal.Reversionoftumorphenotypein surfacetransplants of skinSCC cells by scaffold-induced stromamodulation.Carcinogenesis 2007;256:595. 27. Yoshida S, Yokota T, Ujiki M, et al. Pancreatic cancer stimulates pancreatic stellate cell proliferation and TIMP-1 production through the MAP kinase pathway. Biochem Biophys Res Commun 2004;323:1241. 28. Andoh A, Takaya H, Saotome T, et al. Cytokine regulation of chemokine (IL-8, MCP-1, and RANTES) gene expression in human pancreatic peri acinar myofibroblasts. Gastroenterology 2000; 119:211. 29. Schneiderhan W, Diaz F, Fundel M, et al. Pancreatic stellate cells are an important source of MMP-2 in human pancreatic cancer and accelerate tumor progression in a murine xenograft model and CAM assay. J Cell Sci 2007;120(Pt 3):512. 30. Campbell AS, Albo D, Kimsey TF, et al. Macrophage inflammatory protein-3alpha promotes pancreatic cancer cell invasion. J Surg Res 2005;123:96. 31. Feurino LW, Fisher WE, Bharadwaj U, et al. Current update of cytokines in pancreatic cancer: Pathogenic mechanisms, clinical indication, and therapeutic values. Cancer Invest 2006;24:696. 32. Kimsey TF, Campbell AS, Albo D, et al. Co-localization of macrophage inflammatory protein-3alpha (Mip-3alpha) and its receptor, CCR6, promotes pancreatic cancer cell invasion. Cancer J 2004;10:374. 33. Yang X, Staren ED, Howard JM, et al. Invasiveness and MMP expression in pancreatic carcinoma. J Surg Res 2001;98:33. 34. Gong YL, Xu GM, Huang WD, et al. Expression of matrix metalloproteinases and the tissue inhibitors of metalloproteinases and

102

35.

36.

37.

38.

JOURNAL OF SURGICAL RESEARCH: VOL. 157, NO. 1, NOVEMBER 2009 their local invasiveness and metastasis in Chinese human pancreatic cancer. J Surg Oncol 2000;73:95. Albo D, Berger DH, Tuszynski GP. The effect of thrombospondin-1 and TGF-beta 1 on pancreatic cancer cell invasion. J Surg Res 1998;76:86. Zhu M, Gokhale VM, Szabo L, et al. Identification of a novel inhibitor of urokinase-type plasminogen activator. Mol Cancer Ther 2007;6:1348. He Y, Liu XD, Chen ZY, et al. Interaction between cancer cells and stromal fibroblasts is required for activation of the uPARuPA-MMP-2 cascade in pancreatic cancer metastasis. Clin Cancer Res 2007;13:3115. Mueller L, Goumas FA, Affeldt M, et al. Stromal fibroblasts in colorectal liver metastases originate from resident fibroblasts and generate an inflammatory microenvironment. Am J Pathol 2007;171:1608.

39. Hamada T, Tashiro K, Tada H, et al. Isolation and characterization of a novel secretory protein, stromal cell-derived factor-2 (SDF-2) using the signal sequence trap method. Gene 1996; 176:211. 40. Mahadevan D, Von Hoff DD. Tumor-stroma interactions in pancreatic ductal adenocarcinoma. Mol Cancer Ther 2007; 6:1186. 41. Grzesiak JJ, Ho JC, Moossa AR, et al. The integrin-extracellular matrix axis in pancreatic cancer. Pancreas 2007;35:293. 42. Imamura T, Manabe T, Ohshio G, et al. Immunohistochemical staining for type IV collagen and laminin in the stroma of human pancreatic cancer. Int J Pancreatol 1995;18:95. 43. Ryu B, Jones J, Blades NJ, et al. Relationships and differentially expressed genes among pancreatic cancers examined by large-scale serial analysis of gene expression. Cancer Res 2002;62:819.