- Email: [email protected]
Pancreatic carcinoma cells stimulate proliferation and matrix synthesis of hepatic stellate cells
Journal of Hepatology 51 (2009) 307–314 www.elsevier.com/locate/jhep
Pancreatic carcinoma cells stimulate proliferation and matrix synthesis of hepatic stellate cellsq Yu-Wen Tien1,*, Yao-Ming Wu1, Wei-Chou Lin2, Hsuan-Shu Lee3, Po-Huang Lee1 1
Department of Surgery, National Taiwan University Hospital and National Taiwan University College of Medicine, 7 Chung-Shan S. Rd., Taipei 10002, Taiwan, ROC 2 Department of Pathology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan, ROC 3 Department of Internal Medicine, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan, ROC
Background/Aims: Pancreatic ductal carcinoma cells induce fibrosis by stimulating pancreatic stellate cells to proliferate and synthesize matrix. Desmoplastic reaction has also been observed in liver metastases of pancreatic carcinoma. Hepatic stellate cells are similar to pancreatic stellate cells and may contribute to the desmoplasia associated with liver metastases of pancreatic cancer. The aim of this study was to determine the role of hepatic stellate cells in metastasis. Methods: Markers of the desmoplastic reaction in tumors induced in nude mice (n = 6) by subcutaneously injecting pancreatic carcinoma cells with and without hepatic stellate cells were monitored immunohistochemically. Paracrine stimulation was studied by measuring matrix synthesis (collagen type I and c-fibronectin protein) and cell proliferation. Results: Supernatants of pancreatic carcinoma cells stimulated proliferation of cultured hepatic stellate cells and synthesis of collagen I and c-fibronectin. Preincubation of the supernatants with neutralizing antibodies against fibroblast growth factor 2, transforming growth factor-b1, and platelet-derived growth factor significantly reduced these stimulatory effects. Subcutaneous injection of hepatic stellate cells induced earlier onset and faster-growth of subcutaneous fibrotic pancreatic tumors in nude mice. Conclusions: Hepatic stellate cells enhance tumor growth in nude mouse and may play an important role in metastasis formation. Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Hepatic stellate cell; Desmoplastic reaction; Pancreatic cancer
1. Introduction Worldwide, adenocarcinoma of the pancreas is one of the top 10 leading causes of cancer death with an overall 5-year survival of less than 5% [1–3]. In spite of the recent tremendous progress in molecular genetic study Received 28 November 2008; received in revised form 13 February 2009; accepted 6 March 2009; available online 22 April 2009 Associate Editor: A. Geerts q The authors who have taken part in this study declared that they do not have anything to disclose regarding funding from industry or conflict of interest with respect to this manuscript. * Corresponding author. Tel.: +886 2 23123456x65330; fax: +886 2 23568810. E-mail address: [email protected] (Y.-W. Tien).
of cancer, no unique oncogene or tumor suppressor gene has been identified to account for the dismal prognosis of pancreatic adenocarcinoma. Pathologically, one of the most characteristic hallmarks of pancreatic adenocarcinoma other than adenocarcinoma from other organs is desmoplastic stromal tissue [4–8]. The mean collagen content is 3-fold higher in samples of pancreatic tumors and tumor-forming chronic pancreatitis tissue than in samples of normal pancreas tissue [9]. Although pancreatic carcinoma cell (PCC) lines can produce collagen (col) (types I, III, and IV), fibronectin, laminin, vitronectin, and undulin in vitro and in vivo [10], most reports indicate that stromal cells produce the fibrotic extracellular matrix (ECM) associated with pancreatic adenocarcinoma [8,11].
0168-8278/$36.00 Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2009.03.016
308
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
Many reports have provided strong evidence that pancreatic stellate cells (PSCs) play a central role in fibrogenesis associated with pancreatic ductal adenocarcinoma [12–14]. In addition, PSCs strongly support tumor growth in the nude mouse model [15]. A similar distribution of collagens was observed in human pancreatic carcinoma liver metastasis (PCLM) [4]. Presumably, only pancreatic adenocarcinoma cells and not PSCs leave the primary tumor and metastasize to the liver. Thus, theoretically, PSCs are not the progenitors of the desmoplastic stromal cells in the microenvironment of liver metastases. PSCs play an important role in pancreatic fibrosis induced by PCCs or chronic pancreatitis and hepatic stellate cells (HSCs) have similar roles in liver fibrosis and cirrhosis. Both of them store retinyl palmitate and will transit to an a-SMA-positive matrix-producing myofibroblast-like cell when stimulated by transforming growth factor (TGF)-b1 or tumor necrosis factor a [12,13,16–18]. Moreover, an assessment of the expression of 21,329 genes, genome-wide, identified only 29 that were differentially expressed between cultures of primary HSCs and PSCs [19]. However, Nakamura et al. showed that microenvironment significantly influenced gene expression in PCCs [20]. Therefore, it remains unknown whether metrastatic PCCs can drive HSCs to play a role and contribute to the increased connective tissue content in PCLM. To study the role of HSCs in PCLM, we examined (1) paracrine stimulation of cultured HSCs by PCC supernatants (SNs) and the effect of HSCs on the growth of tumors induced in nude mice by subcutaneously injecting PCC lines (MiaPaCa2, AsPC-1, and SU86.86) with and without HSCs.
2. Materials and methods 2.1. Isolation and culture of human HSCs It was almost impossible to get enough human PCLM tissue to isolate cells because resection was not indicated in these patients. Instead, HSCs were isolated by outgrowth from explants of normal human liver tissue obtained after surgical resection of colonic liver metastasis, as described previously [21]. These cells were likewise cultured in DMEM containing 10% fetal bovine serum (FBS). Using a staining kit from Dako, more than 95% of isolated cells were positive anti-a-SMA antibody.
2.2. Production of conditioned media Conditioned medium were harvested form cultured MiaPaCa-2, AsPC-1, and SU86.86 cells after incubation in Dulbecco’s modified Eagle’s medium (DMEM)/Ham’s F12 for 24 h and centrifuged (800 rpm, 5 min, 4 °C) to remove cell debris. Harvested conditioned medium was then stored at 80 °C until use. To determine which medium is responsible for the PCC-stimulated HSC proliferation, conditioned media from MiaPaCa2, AsPC-1, and SU86.86 were preincubated with neutralizing antibodies to platelet-derived growth factor (PDGF) (AA) and PDGF (BB) (each 1 lg/mL), fibroblast growth factor (FGF)-2 (1 lg/mL), or transforming growth factor (TGF)-b1 (10 lg/mL) before being added to cultured HSCs.
2.3. Determination of cell proliferation Proliferation of HSC was measured by bromodeoxyuridine (BrdU) incorporation. Twenty-four hours after stimulation, BrdU (final concentration, 5 10 5 mol/L) was added for another 24 h and BrdU incorporation was determined. DNA was also measured with fluorometry using bisbenzimide and calf thymus DNA as a standard.
2.4. Determination of extracellular matrix synthesis COL I and c-fibronectin concentrations in HSC-SNs were measured in culture supernatants 24 h after stimulation using commercially available ELISA kits (Union Biomed Inc., Taipei, Taiwan). Cells were cultured in DMEM/HAM’s F12 with 0.1% FBS in the presence of ascorbic acid (100 lg/ml) and b-aminopropionitrile (100 lg/ml) to increase proline hydroxylation and to inhibit fibrillar collagen deposition. COL I and c-fibronectin concentrations were expressed in relation to DNA content in the corresponding culture well. All measurements (standards, controls, and samples) were performed in triplicate. Variations in triplicate measurements were between 0.5% and 5% and did not exceed 10%.
2.5. Animal studies To determine whether PCCs stimulate HSC proliferation and matrix synthesis in vivo, we compared the growth of tumors produced by injection of 2 106 MiaPaCa-2 cells (group A), AsPC-1 cells (group B), or SU86.86 cells (group C) on the left side of the back with tumors produced by co-injection of 2 106 HSCs and 2 106 MiaPaCa-2 (group D), AsPC-1 (group E), or SU86.86 cells (group F) on the right side of the same animal’s back. The tumor size was determined using a caliper, and the tumor volume was calculated using the formula (a b c)/2, where a and b were the shorter and longer diameters of each tumor and c was the thickness. The animals were killed after 28 days. Tumors were removed and frozen for sectioning, conventional staining with H&E, and immunostaining of a-SMA, GFAP, COL I, c-fibronectin, anticytokeratin, and CD-31. All animals received humane care according to National Taiwan University guidelines for animal experimentation and the experimental protocol was approved by National Taiwan University College of Medicine and College of Public Health Institutional Animal Care and Use Committee.
2.6. Statistical analysis All assays were performed in triplicate and results are presented as means ± SD of three independent experiments; each condition was performed using three cultures. Statistical analysis was done with SPSS 10.0.1 for Windows 98/NT (SPSS Inc., Chicago, IL, USA). Comparison of parameters between different groups was made using analysis of variance (Scheffe´ test). An a level of P < 0.05 was considered to be statistically significant. To compare tumor volume between groups A and D, groups B and E, or groups C and F, paired Student t-tests with twotailed comparisons were used.
3. Results 3.1. Stimulation of HSC proliferation by conditioned media from PCC lines The addition of PCC-SN (100–500 lL/mL of medium) to HSCs cultured in the presence of 0.1% FBS increased proliferation dose-dependently (Fig. 1) and DNA content per culture (100 lL of MiaPaCa-2-SN increased DNA by 26%, 100 lL of AsPC-1-SN
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
309
Fig. 1. Effect of pancreas carcinoma cell line supernatants on proliferation of cultured HSCs. (A) Forty-eight hours after stimulation, cultures were stopped, and HSC proliferation was measured by fluorometric DNA quantification. DNA content per culture increased significantly after the addition of PCC-SN, (B) twenty-four hours after stimulation, BrdU (final concentration, 5 10 5 mol/L) was added for another 24 h and BrdU incorporation was determined. The BrdU labeling indices increased significantly after the addition of PCC-SN. (C) Carcinoma cell supernatants were preincubated for 1 h with PDGF-neutralizing antibody [anti-PDGF(AA) and anti-PDGF(BB) before the conditioned media were added to cultured PSCs. Both anti-PDGF-A and anti-PDGF-B significantly reduced the mitogenic activity of MiaPaCa2-SN, AsPC1-SN, and SU86.86-SN. Statistically significant difference (P < 0.05) compared with control.
increased DNA by 43%, and 100 lL of SU86.86 increased DNA by 20%; Fig. 1A). The BrdU labeling indices of HSCs unstimulated or stimulated by 100 lL of MiaPaCa-2-SN, AsPC-1, SU86.86-SN or 10% FBS were 0.122, 0.328, 0.364, 0.426, and 0.31, respectively (Fig. 1B). Anti-PDGF-A and anti-PDGF-B but not anti- TGF-b1 and anti-FGF-2 significantly reduced the mitogenic activity of MiaPaCa2-SN, AsPC1-SN, and SU86.86-SN (Fig. 1C).
3.2. Stimulation of extracellular matrix synthesis in HSCs by conditioned media from PCC lines The addition of MiaPaCa-2-, AsPC-1-, and SU86.86conditioned media (100, 2400, and 500 lL/mL) to cultured HSCs caused a dose-dependent increase in COL I (Fig. 2A) and c-fibronectin (Fig. 2B) synthesis. To identify the mediators of matrix synthesis, PCCSNs were preincubated for 1 h with neutralizing
310
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
antibodies to PDGF (each 1 lg/mL), FGF-2 (1 lg/mL), and TGF-b1 (10 lg/mL) before being added to cultured HSCs. All neutralizing antibodies except anti-PDGF (A and B) significantly reduced the stimulatory effect of
MiaPaCa-2-SN, AsPC-1-SN, and SU86.86-SN on cfibronectin synthesis in cultured HSCs (Fig. 2C–E). 3.3. Animal studies When tumor growth on the left side (injected with PCCs alone) was compared to tumor growth on the right side (PCCs plus HSCs), tumors on the right developed earlier and grew faster than tumors on the left (Fig. 3). Tumors developed on post-injection day 14 at 100% of sites injected with MiaPaCa-2 plus HSC, 50% of sites injected with MiaPaCa-2 alone (P = 0.045), 83% of sites injected with AsPC-1 plus HSC, 50% of sites injected with AsPC-1 alone (P = 0.22), 83% of sites injected with SU86.86 plus HSC, and 0% of sites injected with SU86.86 alone (P = 0.003). Subcutaneous tumors grew significantly earlier after injection of PCCs plus HSCs than after injection of PCCs alone. As shown in Fig. 3, tumor volume on pos-tinjection day 28 was enhanced 1.76-fold at sites injected with MiaPaCa-2 plus HSCs compared with sites injected with MiaPaCa-2 alone (Fig. 3A), 1.4-fold at sites injected with AsPC-1 plus HSCs compared with sites injected with AsPC-1 alone (Fig. 3B), and 4.2-fold at sites injected with SU86.86 plus HSCs compared with sites injected with SU86.86 alone (Fig. 3C). In contrast, injection of 6.5 106 HSCs alone did not result in tumor development within 4 weeks. We conclude that HSCs induce significantly faster growth of subcutaneous PCCs in nude mice. Immunohistologically, numbers of a-SMA-positive cells (Fig. 4A) were much higher in the tumors growing at sites injected with PCCs plus HSCs than at sites injected with PCCs alone (Fig. 4B). Furthermore, immunohistochemical staining of COL I and c-fibronectin showed that the desmoplastic reaction was more intense in tumors at sites injected with PCCs plus HSCs (Fig. 4C and E) than
3 Fig. 2. Stimulation of collagen type I and c-fibronectin synthesis in cultured human HSCs by pancreatic carcinoma cell supernatants and inhibition of stimulated c-fibronectin synthesis by neutralizing antibodies to PDGF, FGF-2, and TGF-b1. Two days after passage, HSCs were stimulated for 24 h with 100, 200, and 500 lL/mL MiaPaCa-2, AsPC-1, and SU86.86-conditioned media, respectively. The addition of MiaPaCa2-, AsPC-1-, and SU86.86- conditioned media to cultured HSCs caused a dose-dependent increase in COL I (A) and c-fibronectin (B) synthesis. Carcinoma cell line supernatants were preincubated for 1 h with neutralizing antibodies to PDGF(AA) and PDGF(BB) (each 1 lg/mL), FGF-2 (1 lg/mL), and TGF-b1 (10 lg/mL) before the supernatants (300 lL/mL) were added to cultured HSCs. After 24 h, cultures were stopped, and c-fibronectin was measured in HSC supernatants by immunoassay. All neutralizing antibodies except anti-PDGF significantly reduced the stimulatory effect of MiaPaCa-2-SN, AsPC-1-SN, and SU86.86-SN on c-fibronectin synthesis in cultured HSCs (C, D, E). * Statistically significant difference (P < 0.05) compared with the stimulation with 300 lL/mL carcinoma cell supernatant without specific antibody preincubation. Neutralizing antibodies anti-FGF-2 and antiTGFb1 significantly reduced the stimulatory effect of MiaPaCa-2, AsPC-1, and SU86.86 on c-fibronectin synthesis of cultured HSCs.
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
Fig. 3. Growth rate of subcutaneous tumors induced in nude mice. A total of 2 106 MiaPaCa2 cells (A), 2 106 AsPC-1 cells (B), or 2 106 SU86.86 cells (C) were injected alone and together with 2 106 HSCs subcutaneously on to nu/nu mice. Tumor volume at day 28 after injection of MiaPaCa-2 with HSCs was 1.76-fold higher compared with that after injection of MiaPaCa-2 alone (A); tumor volume at day 28 after injection of AsPC-1 with HSCs was 1.4-fold higher compared with that after injection of AsPC-1 alone (B); and the tumor volume at day 28 after injection of SU86.86 with HSCs was 4.2-fold higher compared with that after injection of SU86.86 alone (C). Subcutaneous injection of PCCs and HSCs induced faster-growing subcutaneous tumors in nude mice than injection of PCCs alone.
311
Fig. 4. Immunohistochemistry of a-SMA, collagen type I and cfibronectin in tissue sections of subcutaneous tumors induced in nude mice. A total of 2 106 AsPC-1 cells were injected with HSCs (A, C, E) and alone (B, D, F). Animals were killed after 28 days. Tumors were excised and immediately frozen in liquid nitrogen until sectioning and immunohistochemistry for a-SMA (A and B), collagen type I (C and D), c-fibronectin (E and F). High numbers of a-SMA cells were found in tumors arising from the combined injection of carcinoma cells and HSCs (A). In contrast, no or few a-SMA -positive cells were found in tumors arising from the injection of carcinoma cells alone (B). Immunohistochemistry of collagen type I and c-fibronectin showed that the desmoplastic reaction was more intense in the tumors developing after injection of carcinoma cells together with HSCs (C and E) compared with tumors arising from single injection of AsPC-1 cells (D and F).
tumors (68.5 ± 10.5% vs. 73.5 ± 13.2%). Microvessel density assessed by immunohistochemical staining with anti-CD31 antibody was also similar between MiaPaCa2 and MiaPaCa-2-HSC tumors (83.2 ± 12.6 vs. 75 ± 14.5 per 200 field ) or between AsPC-1 and AsPC-1-HSC tumors (74.6 ± 11.3 vs. 69.8 ± 12.1 per 200 field).
4. Discussion in tumors at sites injected with PCCs alone (Fig. 4D and F). However, no or few desmin-positive or GFAP-positive cells were noted in tumors arising from injection of carcinoma cells combined with HSCs or not. Cancer cells were stained with anticytokeratin antibody and Cancer cell density per area of section of the subcutaneous tumor was determined by morphometry. Morphometric analysis showed no differences in PCC density between MiaPaCa-2 and MiaPaCa-2-HSC tumors (62.5 ± 12.5% vs. 67.5 ± 11.2%) or between AsPC-1 and AsPC-1-HSC
Tumor metastasis consists of a series of discrete biological processes that result in tumor formation at sites distant from the primary neoplasm. In a melanoma experimental metastasis model, the majority (>80%) of injected tumor cells survived the circulation and successfully extravasated into the liver. Only 1 in 40 cells formed micrometastases by day 3, however, and only 1 in 100 micrometastases grew into macroscopic metastases 10 days later [22]. Thus, in this model, colonization
312
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
after extravasation is the rate-limiting step of metastasis. Successful colonization crucially depends on interaction with the microenvironment or ‘‘soil” of the distant tissue. Liver is the most common metastatic site of pancreatic cancer. Since HSCs and PSCs have common features, and PCLMs and primary pancreatic tumors have similar collagen distributions, it is reasonable to postulate that HSCs play an important role in the metastasis of pancreatic cancer cells to liver. In our study, culture medium from our three PCC lines stimulated the proliferation and matrix synthesis of HSCs. We further attempted to identify the mediators of this stimulation. Since (1) PCC lines express PDGF [23,24], FGF-2 [24–31], and TGF-b1 [24,32]; (2) PDGF and FGF-2 accelerate proliferation of cultured HSCs[33–35]; and (3) FGF-2, TGF-b, and PDGF are fibrogenic mediators that stimulate ECM synthesis in activated HSCs [36–40], it is likely that these growth factors in supernatants from cultured PCC lines are fibrogenic. Using neutralizing antibodies to PDGF, we showed that PDGF from cultured PCC lines stimulate proliferation of activated HSCs. These results are in accordance with earlier reports showing (1) expression of PDGF in pancreatic carcinoma [24,25], (2) expression of PDGFRa and PDGFRb in HSCs [41,42], (3) mitogenic [35] and chemotactic [35,43] properties of PDGFs on HSCs, and (4) association of PDGF with hepatic fibrosis [39,40,43]. Furthermore, neutralizing antibodies to PDGF, FGF-2, and TGF-b1 identified FGF-2 and TGF-b1 but not PDGF as the mediators in carcinoma SNs responsible for stimulating ECM synthesis in cultured HSCs. The central role of TGF-b in human pancreas and liver fibrogenesis is well documented [44]. In addition, previous studies have shown that TGF-b1 increases synthesis and deposition of collagens and fibronectin in cultured HSCs [36,39]. Furthermore, TGF-b1 induces HSC activation, namely, a phenotype change from quiescent retinoid-containing fat-storing cells to highly active myofibroblast-like cells [45]. During HSC activation, the number and size of perinuclear fat droplets decreases and cells lose their cellular retinoids, develop a prominent endoplasmic reticulum, express a-SMA [46,47], and increase their expression of PDGF receptors [48]. Also, a stimulating effect of TGF-b1 directly on PDGF-receptor expression and PDGF synthesis was reported [49]. These findings suggest that TGF-b released by PCCs might induce desmoplasia in PCLM via several mechanisms, including stimulation of matrix synthesis, stimulation of HSC proliferation (through autocrine PDGF loops), and activation. We also identified FGF-2 as a fibrogenic mediator produced by PCCs and a stimulator of matrix synthesis in HSCs. Our data are consistent with previous data showing that pancreatic cancer cells overexpress FGF2 (and the FGF-2 receptor) [24,29–31] and induce prolif-
eration of adjacent fibroblasts in vitro [8]. The addition of recombinant FGF-2 to cultured HSCs stimulated ECM synthesis in these cells [50]. Our results using the pancreatic carcinoma xenograft model confirm our results using the pancreatic carcinoma cell culture model. In the presence of HSCs, PCC lines stimulate fibrogenic processes, e.g., deposition of collagen type I and fibronectin in dense fibrous bundles around islets of PCCs. Although PCCs are producers of matrix proteins [10], our immunofluorescence micrographs (showing a close association of a-SMApositive with fibrillar collagens and fibronectin) indicate that HSCs are the primary producers of matrix. The sparse collagens and fibronectin in tumors developing from PCCs injected in the absence of HSCs are most probably generated by desmin-negative and a-SMAnegative fibroblasts of the nude mice. Of particular interest is our present finding that tumor growth occurs earlier and is accelerated after co-injection of PCCs and HSCs. Morphometric analysis showed no differences in cancer cell density between tumors arising from injection of carcinoma cells alone or combined with HSCs. Angiogenesity assessed by immunohistochemical staining with anti-CD31 antibody was also similar between tumors arising from injection of carcinoma cells alone or combined with HSCs. Therefore, we believed that subcutaneous tumors grow faster in the presence of HSCs because of increased amount of fibrous tissue. Our study showed that human pancreatic cancer cell lines are able to stimulate the proliferation and fibrogenic activities of human liver-derived HSCs and hepatic stellate cells enhance tumor growth in nude mouse. Although the presented data did not show metastatic pancreatic cancer cells can recruit HSCs into hepatic metastasis, we believe that hepatic stellate cells may contribute to postextravasation growth of metastatic pancreatic cancer cells in liver because: 1. hepatic stellate cells represent 5–8% of liver cells in normal liver, 2. desmoplastic reaction has also been observed in liver metastases of pancreatic carcinoma, and 3. only pancreatic adenocarcinoma cells and not PSCs leave the primary tumor and metastasize to the liver. In conclusion, in culture, PCC lines (MiaPaCa-2, AsPC-1, and SU86.86) stimulated proliferation of HSCs via PDGF, and ECM synthesis via FGF-2 and TGF-b1. In addition, after subcutaneous injection of PCCs (MiaPaCa-2, AsPC-1, and SU86.86) with HSCs into immunodeficient mice, tumors developed earlier, grew faster, and became more fibrotic than tumors induced by subcutaneous injection of PCC cells alone. Acknowledgement Grant support: National Council of Science (NSC95-2314-B-002-160).
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
References [1] Warshaw AL, Ferandez-Del Castillo C. Pancreatic carcinoma. N Engl J Med 1992;326:455–465. [2] DiMagno EP, Reber HA, Tempero MA. AGA technical review on the epidemiology, diagnosis, and treatment of pancreatic ductal adenocarcinoma. American Gastroenterologic Association. Gastroenterology 1999;117:1464–1484. [3] Sener SF, Fremgen A, Menck HR, Winchester DP. Pancreatic cancer: a report of treatment and survival trends for 100313 patients diagnosed from 1985 to 1995, using the National Cancer Database. J Am Coll Surg 1999;189:1–7. [4] Mollenhauer J, Roether I, Kern HF. Distribution of extracellular matrix proteins in pancreatic ductal adenocarcinoma and its influence on tumor cell proliferation in vitro. Pancreas 1987;2:14–24. [5] Korc M. Pancreatic cancer-associated stroma production. Am J Surg 2007;194:S84–S86. [6] Klo¨ppel G, Lingenthal G, Bu¨low MV, Kern HF. Histological and fine structural features of pancreatic ductal adenocarcinomas in relation to growth and prognosis: studies in xenografted tumours and clinco-histopathological correlation in a series of 75 cases. Histopathology (England) 1985;9:841–856. [7] Cubilla AL, Fitzgerald PJ. Morphological lesions associated with human primary invasive non-endocrine pancreas cancer. Cancer Res 1976;36:2690–2698. [8] Kuniyasu H, Abbruzzese JL, Cleary KR, Fidler IJ. Induction of ductal and stromal hyperplasia by basic fibroblast growth factor produced by human pancreatic carcinoma. Int J Oncol 2001;19:681–685. [9] Imamura T, Iguchi H, Manabe T, Ohshio G, Yoshimura T, Wang ZH, et al. Quantitative analysis of collagen and collagen subtypes I, III, and V in human pancreatic cancer, tumor-associated chronic pancreatitis, and alcoholic chronic pancreatitis. Pancreas 1995;11:357–364. [10] Lo¨hr M, Trautmann B, Gottler M, Peters S, Zauner I, Maillet B, et al. Human ductal adenocarcinomas of the pancreas express extracellular matrix proteins. Br J Cancer 1994;69:144–151. [11] Gress TM, Menke A, Bachem MG, Muller-Pillasch F, Ellenrieder V, Weidenbach H, et al. Role of extracellular matrix in pancreatic diseases. Digestion 1998;59:625–637. [12] Apte MV, Haber PS, Applegate TL, Norton ID, McCaughan GW, Korsten MA, et al. Periacinar stellate-shaped cells in rat pancreas: identification, isolation and culture. Gut 1998;43:128–133. [13] Bachem MG, Schneider E, Groß H, Weidenbach H, Schmid RM, Menke A, et al. Identification, culture, and characterization of pancreatic stellate cells in rats and humans. Gastroenterology 1998;115:421–432. [14] Apte MV, Wilson JS. Alcohol-induced pancreatic injury. Best Pract Res Clin Gastroenterol 2003;17:593–612. [15] Bachem MG, Schu¨nemann M, Ramadani M, Siech M, Beger H, Buck A, et al. Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells. Gastroenterol 2005;128:907–921. [16] Apte MV, Haber PS, Darby SJ, Rodgers SC, McCaughan GW, Korsten MA, et al. Pancreatic stellate cells are activated by proinflammatory cytokines: implications for pancreatic fibrogenesis. Gut 1999;44:534–541. [17] Haber PS, Keogh GW, Apte MV, Moran CS, Stewart NL, Crawford DH, et al. Activation of pancreatic stellate cells in human and experimental pancreatic fibrosis. Am J Pathol 1999;155:1087–1095. [18] Schneider E, Schmid-Kotsas A, Zhao J, Weidenbach H, Schmid RM, Menke A, et al. Identification of mediators stimulating proliferation and matrix synthesis of rat pancreatic stellate cells. Am J Physiol Cell Physiol 2001;281:C532–C543.
313
[19] Buchholz M, Kestler HA, Holzmann K, Ellenrieder V, Schneiderhan W, Siech M, et al. Transcriptome analysis of human hepatic and pancreatic stellate cells: organ-specific variations of a common transcriptional phenotype. J Mol Med 2005;83:795–805. [20] Nakamura T, Fidler IJ, Coombes KR. Gene expression profile of metastatic human pancreatic cancer cells depends on the organ environment. Cancer Res 2007;67:139–148. [21] Lee HS, Miau LH, Chen CH, Chiou LL, Huang GT, Yang PM, et al. Differential role of p38 in IL-1a induction of MMP-9 and MMP-13 in an established liver myofibroblast cell line. J Biomed Sci 2003;10:757–776. [22] Luzzi KJ, Mcdonald IC, Schmidt EE, Kerkvliet N, Morris VL, Chambers AF, et al. Multistep nature of metastatic ineffiencyDormancy of solitary cells after successful extravasation and limited survival of early micrometastases. Am J Pathol 1998;153:865–873. [23] Levy MT, McCaughan GW, Abbott CA, Park JE, Cunningham AM, Muller E, et al. Fibroblast activation protein: a cell surface dipeptidyl peptidase and gelatinase expressed by stellate cells at the tissue remodeling interface in human cirrhosis. Hepatology 1999;29:1768–1778. [24] Beauchamp RD, Lyons RM, Yang EY, Coffey Jr RJ, Moses HL. Expression of and response to growth regulatory peptides by two human pancreatic carcinoma cell lines. Pancreas 1990;5: 369–380. [25] Ebert M, Yokoyama M, Friess H, Kobrin MS, Buchler MW, Korc M. Induction of platelet-derived growth factor A and B chains and over-expression of their receptors in human pancreatic cancer. Int J Cancer 1995;62:529–535. [26] Kuniyasu H, Abbruzzese JL, Cleary KR, Fidler IJ. Induction of ductal and stromal hyperplasia by basic fibroblast growth factor produced by human pancreatic carcinoma. Int J Oncol 2001;19:681–685. [27] Kornmann M, Ishiwata T, Beger HG, Korc M. Fibroblast growth factor-5 stimulates mitogenic signaling and is overexpressed in human pancreatic cancer: evidence for autocrine and paracrine actions. Oncogene 1997;15:1417–1424. [28] Kornmann M, Beger HG, Korc M. Role of fibroblast growth factors and their receptors in pancreatic cancer and chronic pancreatitis. Pancreas 1998;17:169–175. [29] Leung HY, Gullick WJ, Lemoine NR. Expression and functional activity of fibroblast growth factor and their receptors in human pancreatic cancer. Int J Cancer 1994;59:667–675. [30] Leung HY, Hughes CM, Klo¨ppel G. Localization of expression of fibroblast growth factors and their receptors in pancreatic adenocarcinoma by in situ hybridization. Int J Oncol 1994;4:1219–1223. [31] Ohta T, Yamamoto M, Numata M, Iseki S, Tsukioka Y, Miyashita T, et al. Expression of basic fibroblast growth factor and its receptor in human pancreatic carcinomas. Br J Cancer 1995;72:824–831. [32] Friess H, Yamanaka Y, Buchler M, Ebert M, Beger HG, Gold LI, et al. Enhanced expression of transforming growth factor beta isoforms in pancreatic cancer correlates with decreased survival. Gastroenterology 1993;105:1846–1856. [33] Wong L, Yamasaki G, Johnson RJ, Friedman SL. Induction of beta-platelet-derived growth factor receptor in rat hepatic lipocyte during cellular activation in vivo and in culture. J Clin Invest 1994;94:1563–1569. [34] Marra F, Gentilini A, Pinzani M, Choudhury GG, Parola M, Herbst H, et al. Phosphatidyllinositol 3-kinase is required for platelet-derived growth factor’s action on hepatic stellate cells. Gastroenterology 1997;112:1297–1306. [35] Failli P, Ruocco C, De Franco R, Caligiuri A, Gentilini A, Giotti A, et al. The mitogenic effect of platelet-derived growth factor in human hepatic stellate cells requires calcium influx. Am J Physiol 1995:C1133–39.
314
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
[36] Gong W, Roth S, Michel K, Gressner AM. Isoforms and splice variant of transforming growth factor beta-binding protein in rat hepatic stellate cells. Gastroenterology 1998;114:352–363. [37] Yu C, Wang F, Jin C, Huang X, Miller DL, Basilico C, et al. Role of fibroblast growth factor type 1 and 2 in carbon tetrachlorideinduced hepatic injury and fibrogenesis. Am J Pathol 2003;163:1653–1662. [38] Hellerbrand C, Stefanovic B, Giordano F, Burchardt ER, Brenner DA. The role of TGF beta1 in initiating hepatic stellate cell activation in vivo. J Hepatol 1999;30:77–87. [39] Czochra P, Klopcic B, Meyer E, Herkel J, Garcia-Lazaro JF, Thieringer F, et al. Liver fibrosis induced by hepatic overexpression of PDGF-b in transgenic mice. J Hepatol 2006;45:419–428. [40] Henderson NC, Iredale JP. Liver fibrosis: cellular mechanisms of progression and resolution. Clin Sci 2007;112:265–280. [41] Iwamoto H, Nakamuta M, Tada S, Sugimoto R, Enjoji M, Nawata H. Platelet-derived growth factor receptor tyrosine kinase inhibitor AG1295 attenuates rat hepatic stellate cell growth. J Lab Clin Med 2000;135:406–412. [42] Brady LM, Fox ES, Fimmel CJ. Polyenylphosphatidylcholine inhibits PDGF-induced proliferation in rat hepatic stellate cells. Bioch Biophysi Res Comm 1998;248:174–179. [43] Scott F. Cytokines and fibrogenesis. Semin Liver Dis 1999;19:129–140.
[44] Tsukada S, Parsons CJ, Rippe RA. Mechanisms of liver fibrosis. Clin Chimica Acta 2006;364:33–60. [45] Okuno M, Moriwaki H, Imai S, Muto Y, Kawada N, Suzuki Y, et al. Retinoids exacerbate rat liver fibrosis by inducing the activation of latent TGFbeta in liver stellate cells. Hepatology 1997:913–921. [46] Mak KM, Leo MA, Lieber CS. Alcoholic liver injury in baboobn: transformation of lipocytes to transitional cells. Gastroenterology 1984;87:188–200. [47] Mak KM, Lieber CS. Lipocytes and transitional cells in alcoholic liverc disease: a morphometric study. Hepatology 1988;8:1027–1033. [48] Marra F, Choudhury GG, Pinzani M, Abboud HE. Regulation of platelet-derived growth factor secretion and gene expression in human fat-storing cells. Gastroenterology 1994;107:1110–1117. [49] Pinzani M, Gentilini A, Caligiuri A, De Franco R, Pellegrini G, Milani S, et al. Transforming growth factor-beta 1 regulates platelet-derived growth factor receptor beta subunit in human liver fat-storing cells. Hepatology 1995;21:232–239. [50] Rosenbaum J, Blazejewski S, Preaux AM, Mallat A, Dhumeaux D, Mavier P. Fibroblast growth factor 2 and trandforming growth factor beta 1 interactions in human liver myofibroblasts. Gastroenterology 1995;109:1986–1996.
Pancreatic carcinoma cells stimulate proliferation and matrix synthesis of hepatic stellate cellsq Yu-Wen Tien1,*, Yao-Ming Wu1, Wei-Chou Lin2, Hsuan-Shu Lee3, Po-Huang Lee1 1
Department of Surgery, National Taiwan University Hospital and National Taiwan University College of Medicine, 7 Chung-Shan S. Rd., Taipei 10002, Taiwan, ROC 2 Department of Pathology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan, ROC 3 Department of Internal Medicine, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan, ROC
Background/Aims: Pancreatic ductal carcinoma cells induce fibrosis by stimulating pancreatic stellate cells to proliferate and synthesize matrix. Desmoplastic reaction has also been observed in liver metastases of pancreatic carcinoma. Hepatic stellate cells are similar to pancreatic stellate cells and may contribute to the desmoplasia associated with liver metastases of pancreatic cancer. The aim of this study was to determine the role of hepatic stellate cells in metastasis. Methods: Markers of the desmoplastic reaction in tumors induced in nude mice (n = 6) by subcutaneously injecting pancreatic carcinoma cells with and without hepatic stellate cells were monitored immunohistochemically. Paracrine stimulation was studied by measuring matrix synthesis (collagen type I and c-fibronectin protein) and cell proliferation. Results: Supernatants of pancreatic carcinoma cells stimulated proliferation of cultured hepatic stellate cells and synthesis of collagen I and c-fibronectin. Preincubation of the supernatants with neutralizing antibodies against fibroblast growth factor 2, transforming growth factor-b1, and platelet-derived growth factor significantly reduced these stimulatory effects. Subcutaneous injection of hepatic stellate cells induced earlier onset and faster-growth of subcutaneous fibrotic pancreatic tumors in nude mice. Conclusions: Hepatic stellate cells enhance tumor growth in nude mouse and may play an important role in metastasis formation. Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Hepatic stellate cell; Desmoplastic reaction; Pancreatic cancer
1. Introduction Worldwide, adenocarcinoma of the pancreas is one of the top 10 leading causes of cancer death with an overall 5-year survival of less than 5% [1–3]. In spite of the recent tremendous progress in molecular genetic study Received 28 November 2008; received in revised form 13 February 2009; accepted 6 March 2009; available online 22 April 2009 Associate Editor: A. Geerts q The authors who have taken part in this study declared that they do not have anything to disclose regarding funding from industry or conflict of interest with respect to this manuscript. * Corresponding author. Tel.: +886 2 23123456x65330; fax: +886 2 23568810. E-mail address: [email protected] (Y.-W. Tien).
of cancer, no unique oncogene or tumor suppressor gene has been identified to account for the dismal prognosis of pancreatic adenocarcinoma. Pathologically, one of the most characteristic hallmarks of pancreatic adenocarcinoma other than adenocarcinoma from other organs is desmoplastic stromal tissue [4–8]. The mean collagen content is 3-fold higher in samples of pancreatic tumors and tumor-forming chronic pancreatitis tissue than in samples of normal pancreas tissue [9]. Although pancreatic carcinoma cell (PCC) lines can produce collagen (col) (types I, III, and IV), fibronectin, laminin, vitronectin, and undulin in vitro and in vivo [10], most reports indicate that stromal cells produce the fibrotic extracellular matrix (ECM) associated with pancreatic adenocarcinoma [8,11].
0168-8278/$36.00 Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2009.03.016
308
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
Many reports have provided strong evidence that pancreatic stellate cells (PSCs) play a central role in fibrogenesis associated with pancreatic ductal adenocarcinoma [12–14]. In addition, PSCs strongly support tumor growth in the nude mouse model [15]. A similar distribution of collagens was observed in human pancreatic carcinoma liver metastasis (PCLM) [4]. Presumably, only pancreatic adenocarcinoma cells and not PSCs leave the primary tumor and metastasize to the liver. Thus, theoretically, PSCs are not the progenitors of the desmoplastic stromal cells in the microenvironment of liver metastases. PSCs play an important role in pancreatic fibrosis induced by PCCs or chronic pancreatitis and hepatic stellate cells (HSCs) have similar roles in liver fibrosis and cirrhosis. Both of them store retinyl palmitate and will transit to an a-SMA-positive matrix-producing myofibroblast-like cell when stimulated by transforming growth factor (TGF)-b1 or tumor necrosis factor a [12,13,16–18]. Moreover, an assessment of the expression of 21,329 genes, genome-wide, identified only 29 that were differentially expressed between cultures of primary HSCs and PSCs [19]. However, Nakamura et al. showed that microenvironment significantly influenced gene expression in PCCs [20]. Therefore, it remains unknown whether metrastatic PCCs can drive HSCs to play a role and contribute to the increased connective tissue content in PCLM. To study the role of HSCs in PCLM, we examined (1) paracrine stimulation of cultured HSCs by PCC supernatants (SNs) and the effect of HSCs on the growth of tumors induced in nude mice by subcutaneously injecting PCC lines (MiaPaCa2, AsPC-1, and SU86.86) with and without HSCs.
2. Materials and methods 2.1. Isolation and culture of human HSCs It was almost impossible to get enough human PCLM tissue to isolate cells because resection was not indicated in these patients. Instead, HSCs were isolated by outgrowth from explants of normal human liver tissue obtained after surgical resection of colonic liver metastasis, as described previously [21]. These cells were likewise cultured in DMEM containing 10% fetal bovine serum (FBS). Using a staining kit from Dako, more than 95% of isolated cells were positive anti-a-SMA antibody.
2.2. Production of conditioned media Conditioned medium were harvested form cultured MiaPaCa-2, AsPC-1, and SU86.86 cells after incubation in Dulbecco’s modified Eagle’s medium (DMEM)/Ham’s F12 for 24 h and centrifuged (800 rpm, 5 min, 4 °C) to remove cell debris. Harvested conditioned medium was then stored at 80 °C until use. To determine which medium is responsible for the PCC-stimulated HSC proliferation, conditioned media from MiaPaCa2, AsPC-1, and SU86.86 were preincubated with neutralizing antibodies to platelet-derived growth factor (PDGF) (AA) and PDGF (BB) (each 1 lg/mL), fibroblast growth factor (FGF)-2 (1 lg/mL), or transforming growth factor (TGF)-b1 (10 lg/mL) before being added to cultured HSCs.
2.3. Determination of cell proliferation Proliferation of HSC was measured by bromodeoxyuridine (BrdU) incorporation. Twenty-four hours after stimulation, BrdU (final concentration, 5 10 5 mol/L) was added for another 24 h and BrdU incorporation was determined. DNA was also measured with fluorometry using bisbenzimide and calf thymus DNA as a standard.
2.4. Determination of extracellular matrix synthesis COL I and c-fibronectin concentrations in HSC-SNs were measured in culture supernatants 24 h after stimulation using commercially available ELISA kits (Union Biomed Inc., Taipei, Taiwan). Cells were cultured in DMEM/HAM’s F12 with 0.1% FBS in the presence of ascorbic acid (100 lg/ml) and b-aminopropionitrile (100 lg/ml) to increase proline hydroxylation and to inhibit fibrillar collagen deposition. COL I and c-fibronectin concentrations were expressed in relation to DNA content in the corresponding culture well. All measurements (standards, controls, and samples) were performed in triplicate. Variations in triplicate measurements were between 0.5% and 5% and did not exceed 10%.
2.5. Animal studies To determine whether PCCs stimulate HSC proliferation and matrix synthesis in vivo, we compared the growth of tumors produced by injection of 2 106 MiaPaCa-2 cells (group A), AsPC-1 cells (group B), or SU86.86 cells (group C) on the left side of the back with tumors produced by co-injection of 2 106 HSCs and 2 106 MiaPaCa-2 (group D), AsPC-1 (group E), or SU86.86 cells (group F) on the right side of the same animal’s back. The tumor size was determined using a caliper, and the tumor volume was calculated using the formula (a b c)/2, where a and b were the shorter and longer diameters of each tumor and c was the thickness. The animals were killed after 28 days. Tumors were removed and frozen for sectioning, conventional staining with H&E, and immunostaining of a-SMA, GFAP, COL I, c-fibronectin, anticytokeratin, and CD-31. All animals received humane care according to National Taiwan University guidelines for animal experimentation and the experimental protocol was approved by National Taiwan University College of Medicine and College of Public Health Institutional Animal Care and Use Committee.
2.6. Statistical analysis All assays were performed in triplicate and results are presented as means ± SD of three independent experiments; each condition was performed using three cultures. Statistical analysis was done with SPSS 10.0.1 for Windows 98/NT (SPSS Inc., Chicago, IL, USA). Comparison of parameters between different groups was made using analysis of variance (Scheffe´ test). An a level of P < 0.05 was considered to be statistically significant. To compare tumor volume between groups A and D, groups B and E, or groups C and F, paired Student t-tests with twotailed comparisons were used.
3. Results 3.1. Stimulation of HSC proliferation by conditioned media from PCC lines The addition of PCC-SN (100–500 lL/mL of medium) to HSCs cultured in the presence of 0.1% FBS increased proliferation dose-dependently (Fig. 1) and DNA content per culture (100 lL of MiaPaCa-2-SN increased DNA by 26%, 100 lL of AsPC-1-SN
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
309
Fig. 1. Effect of pancreas carcinoma cell line supernatants on proliferation of cultured HSCs. (A) Forty-eight hours after stimulation, cultures were stopped, and HSC proliferation was measured by fluorometric DNA quantification. DNA content per culture increased significantly after the addition of PCC-SN, (B) twenty-four hours after stimulation, BrdU (final concentration, 5 10 5 mol/L) was added for another 24 h and BrdU incorporation was determined. The BrdU labeling indices increased significantly after the addition of PCC-SN. (C) Carcinoma cell supernatants were preincubated for 1 h with PDGF-neutralizing antibody [anti-PDGF(AA) and anti-PDGF(BB) before the conditioned media were added to cultured PSCs. Both anti-PDGF-A and anti-PDGF-B significantly reduced the mitogenic activity of MiaPaCa2-SN, AsPC1-SN, and SU86.86-SN. Statistically significant difference (P < 0.05) compared with control.
increased DNA by 43%, and 100 lL of SU86.86 increased DNA by 20%; Fig. 1A). The BrdU labeling indices of HSCs unstimulated or stimulated by 100 lL of MiaPaCa-2-SN, AsPC-1, SU86.86-SN or 10% FBS were 0.122, 0.328, 0.364, 0.426, and 0.31, respectively (Fig. 1B). Anti-PDGF-A and anti-PDGF-B but not anti- TGF-b1 and anti-FGF-2 significantly reduced the mitogenic activity of MiaPaCa2-SN, AsPC1-SN, and SU86.86-SN (Fig. 1C).
3.2. Stimulation of extracellular matrix synthesis in HSCs by conditioned media from PCC lines The addition of MiaPaCa-2-, AsPC-1-, and SU86.86conditioned media (100, 2400, and 500 lL/mL) to cultured HSCs caused a dose-dependent increase in COL I (Fig. 2A) and c-fibronectin (Fig. 2B) synthesis. To identify the mediators of matrix synthesis, PCCSNs were preincubated for 1 h with neutralizing
310
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
antibodies to PDGF (each 1 lg/mL), FGF-2 (1 lg/mL), and TGF-b1 (10 lg/mL) before being added to cultured HSCs. All neutralizing antibodies except anti-PDGF (A and B) significantly reduced the stimulatory effect of
MiaPaCa-2-SN, AsPC-1-SN, and SU86.86-SN on cfibronectin synthesis in cultured HSCs (Fig. 2C–E). 3.3. Animal studies When tumor growth on the left side (injected with PCCs alone) was compared to tumor growth on the right side (PCCs plus HSCs), tumors on the right developed earlier and grew faster than tumors on the left (Fig. 3). Tumors developed on post-injection day 14 at 100% of sites injected with MiaPaCa-2 plus HSC, 50% of sites injected with MiaPaCa-2 alone (P = 0.045), 83% of sites injected with AsPC-1 plus HSC, 50% of sites injected with AsPC-1 alone (P = 0.22), 83% of sites injected with SU86.86 plus HSC, and 0% of sites injected with SU86.86 alone (P = 0.003). Subcutaneous tumors grew significantly earlier after injection of PCCs plus HSCs than after injection of PCCs alone. As shown in Fig. 3, tumor volume on pos-tinjection day 28 was enhanced 1.76-fold at sites injected with MiaPaCa-2 plus HSCs compared with sites injected with MiaPaCa-2 alone (Fig. 3A), 1.4-fold at sites injected with AsPC-1 plus HSCs compared with sites injected with AsPC-1 alone (Fig. 3B), and 4.2-fold at sites injected with SU86.86 plus HSCs compared with sites injected with SU86.86 alone (Fig. 3C). In contrast, injection of 6.5 106 HSCs alone did not result in tumor development within 4 weeks. We conclude that HSCs induce significantly faster growth of subcutaneous PCCs in nude mice. Immunohistologically, numbers of a-SMA-positive cells (Fig. 4A) were much higher in the tumors growing at sites injected with PCCs plus HSCs than at sites injected with PCCs alone (Fig. 4B). Furthermore, immunohistochemical staining of COL I and c-fibronectin showed that the desmoplastic reaction was more intense in tumors at sites injected with PCCs plus HSCs (Fig. 4C and E) than
3 Fig. 2. Stimulation of collagen type I and c-fibronectin synthesis in cultured human HSCs by pancreatic carcinoma cell supernatants and inhibition of stimulated c-fibronectin synthesis by neutralizing antibodies to PDGF, FGF-2, and TGF-b1. Two days after passage, HSCs were stimulated for 24 h with 100, 200, and 500 lL/mL MiaPaCa-2, AsPC-1, and SU86.86-conditioned media, respectively. The addition of MiaPaCa2-, AsPC-1-, and SU86.86- conditioned media to cultured HSCs caused a dose-dependent increase in COL I (A) and c-fibronectin (B) synthesis. Carcinoma cell line supernatants were preincubated for 1 h with neutralizing antibodies to PDGF(AA) and PDGF(BB) (each 1 lg/mL), FGF-2 (1 lg/mL), and TGF-b1 (10 lg/mL) before the supernatants (300 lL/mL) were added to cultured HSCs. After 24 h, cultures were stopped, and c-fibronectin was measured in HSC supernatants by immunoassay. All neutralizing antibodies except anti-PDGF significantly reduced the stimulatory effect of MiaPaCa-2-SN, AsPC-1-SN, and SU86.86-SN on c-fibronectin synthesis in cultured HSCs (C, D, E). * Statistically significant difference (P < 0.05) compared with the stimulation with 300 lL/mL carcinoma cell supernatant without specific antibody preincubation. Neutralizing antibodies anti-FGF-2 and antiTGFb1 significantly reduced the stimulatory effect of MiaPaCa-2, AsPC-1, and SU86.86 on c-fibronectin synthesis of cultured HSCs.
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
Fig. 3. Growth rate of subcutaneous tumors induced in nude mice. A total of 2 106 MiaPaCa2 cells (A), 2 106 AsPC-1 cells (B), or 2 106 SU86.86 cells (C) were injected alone and together with 2 106 HSCs subcutaneously on to nu/nu mice. Tumor volume at day 28 after injection of MiaPaCa-2 with HSCs was 1.76-fold higher compared with that after injection of MiaPaCa-2 alone (A); tumor volume at day 28 after injection of AsPC-1 with HSCs was 1.4-fold higher compared with that after injection of AsPC-1 alone (B); and the tumor volume at day 28 after injection of SU86.86 with HSCs was 4.2-fold higher compared with that after injection of SU86.86 alone (C). Subcutaneous injection of PCCs and HSCs induced faster-growing subcutaneous tumors in nude mice than injection of PCCs alone.
311
Fig. 4. Immunohistochemistry of a-SMA, collagen type I and cfibronectin in tissue sections of subcutaneous tumors induced in nude mice. A total of 2 106 AsPC-1 cells were injected with HSCs (A, C, E) and alone (B, D, F). Animals were killed after 28 days. Tumors were excised and immediately frozen in liquid nitrogen until sectioning and immunohistochemistry for a-SMA (A and B), collagen type I (C and D), c-fibronectin (E and F). High numbers of a-SMA cells were found in tumors arising from the combined injection of carcinoma cells and HSCs (A). In contrast, no or few a-SMA -positive cells were found in tumors arising from the injection of carcinoma cells alone (B). Immunohistochemistry of collagen type I and c-fibronectin showed that the desmoplastic reaction was more intense in the tumors developing after injection of carcinoma cells together with HSCs (C and E) compared with tumors arising from single injection of AsPC-1 cells (D and F).
tumors (68.5 ± 10.5% vs. 73.5 ± 13.2%). Microvessel density assessed by immunohistochemical staining with anti-CD31 antibody was also similar between MiaPaCa2 and MiaPaCa-2-HSC tumors (83.2 ± 12.6 vs. 75 ± 14.5 per 200 field ) or between AsPC-1 and AsPC-1-HSC tumors (74.6 ± 11.3 vs. 69.8 ± 12.1 per 200 field).
4. Discussion in tumors at sites injected with PCCs alone (Fig. 4D and F). However, no or few desmin-positive or GFAP-positive cells were noted in tumors arising from injection of carcinoma cells combined with HSCs or not. Cancer cells were stained with anticytokeratin antibody and Cancer cell density per area of section of the subcutaneous tumor was determined by morphometry. Morphometric analysis showed no differences in PCC density between MiaPaCa-2 and MiaPaCa-2-HSC tumors (62.5 ± 12.5% vs. 67.5 ± 11.2%) or between AsPC-1 and AsPC-1-HSC
Tumor metastasis consists of a series of discrete biological processes that result in tumor formation at sites distant from the primary neoplasm. In a melanoma experimental metastasis model, the majority (>80%) of injected tumor cells survived the circulation and successfully extravasated into the liver. Only 1 in 40 cells formed micrometastases by day 3, however, and only 1 in 100 micrometastases grew into macroscopic metastases 10 days later [22]. Thus, in this model, colonization
312
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
after extravasation is the rate-limiting step of metastasis. Successful colonization crucially depends on interaction with the microenvironment or ‘‘soil” of the distant tissue. Liver is the most common metastatic site of pancreatic cancer. Since HSCs and PSCs have common features, and PCLMs and primary pancreatic tumors have similar collagen distributions, it is reasonable to postulate that HSCs play an important role in the metastasis of pancreatic cancer cells to liver. In our study, culture medium from our three PCC lines stimulated the proliferation and matrix synthesis of HSCs. We further attempted to identify the mediators of this stimulation. Since (1) PCC lines express PDGF [23,24], FGF-2 [24–31], and TGF-b1 [24,32]; (2) PDGF and FGF-2 accelerate proliferation of cultured HSCs[33–35]; and (3) FGF-2, TGF-b, and PDGF are fibrogenic mediators that stimulate ECM synthesis in activated HSCs [36–40], it is likely that these growth factors in supernatants from cultured PCC lines are fibrogenic. Using neutralizing antibodies to PDGF, we showed that PDGF from cultured PCC lines stimulate proliferation of activated HSCs. These results are in accordance with earlier reports showing (1) expression of PDGF in pancreatic carcinoma [24,25], (2) expression of PDGFRa and PDGFRb in HSCs [41,42], (3) mitogenic [35] and chemotactic [35,43] properties of PDGFs on HSCs, and (4) association of PDGF with hepatic fibrosis [39,40,43]. Furthermore, neutralizing antibodies to PDGF, FGF-2, and TGF-b1 identified FGF-2 and TGF-b1 but not PDGF as the mediators in carcinoma SNs responsible for stimulating ECM synthesis in cultured HSCs. The central role of TGF-b in human pancreas and liver fibrogenesis is well documented [44]. In addition, previous studies have shown that TGF-b1 increases synthesis and deposition of collagens and fibronectin in cultured HSCs [36,39]. Furthermore, TGF-b1 induces HSC activation, namely, a phenotype change from quiescent retinoid-containing fat-storing cells to highly active myofibroblast-like cells [45]. During HSC activation, the number and size of perinuclear fat droplets decreases and cells lose their cellular retinoids, develop a prominent endoplasmic reticulum, express a-SMA [46,47], and increase their expression of PDGF receptors [48]. Also, a stimulating effect of TGF-b1 directly on PDGF-receptor expression and PDGF synthesis was reported [49]. These findings suggest that TGF-b released by PCCs might induce desmoplasia in PCLM via several mechanisms, including stimulation of matrix synthesis, stimulation of HSC proliferation (through autocrine PDGF loops), and activation. We also identified FGF-2 as a fibrogenic mediator produced by PCCs and a stimulator of matrix synthesis in HSCs. Our data are consistent with previous data showing that pancreatic cancer cells overexpress FGF2 (and the FGF-2 receptor) [24,29–31] and induce prolif-
eration of adjacent fibroblasts in vitro [8]. The addition of recombinant FGF-2 to cultured HSCs stimulated ECM synthesis in these cells [50]. Our results using the pancreatic carcinoma xenograft model confirm our results using the pancreatic carcinoma cell culture model. In the presence of HSCs, PCC lines stimulate fibrogenic processes, e.g., deposition of collagen type I and fibronectin in dense fibrous bundles around islets of PCCs. Although PCCs are producers of matrix proteins [10], our immunofluorescence micrographs (showing a close association of a-SMApositive with fibrillar collagens and fibronectin) indicate that HSCs are the primary producers of matrix. The sparse collagens and fibronectin in tumors developing from PCCs injected in the absence of HSCs are most probably generated by desmin-negative and a-SMAnegative fibroblasts of the nude mice. Of particular interest is our present finding that tumor growth occurs earlier and is accelerated after co-injection of PCCs and HSCs. Morphometric analysis showed no differences in cancer cell density between tumors arising from injection of carcinoma cells alone or combined with HSCs. Angiogenesity assessed by immunohistochemical staining with anti-CD31 antibody was also similar between tumors arising from injection of carcinoma cells alone or combined with HSCs. Therefore, we believed that subcutaneous tumors grow faster in the presence of HSCs because of increased amount of fibrous tissue. Our study showed that human pancreatic cancer cell lines are able to stimulate the proliferation and fibrogenic activities of human liver-derived HSCs and hepatic stellate cells enhance tumor growth in nude mouse. Although the presented data did not show metastatic pancreatic cancer cells can recruit HSCs into hepatic metastasis, we believe that hepatic stellate cells may contribute to postextravasation growth of metastatic pancreatic cancer cells in liver because: 1. hepatic stellate cells represent 5–8% of liver cells in normal liver, 2. desmoplastic reaction has also been observed in liver metastases of pancreatic carcinoma, and 3. only pancreatic adenocarcinoma cells and not PSCs leave the primary tumor and metastasize to the liver. In conclusion, in culture, PCC lines (MiaPaCa-2, AsPC-1, and SU86.86) stimulated proliferation of HSCs via PDGF, and ECM synthesis via FGF-2 and TGF-b1. In addition, after subcutaneous injection of PCCs (MiaPaCa-2, AsPC-1, and SU86.86) with HSCs into immunodeficient mice, tumors developed earlier, grew faster, and became more fibrotic than tumors induced by subcutaneous injection of PCC cells alone. Acknowledgement Grant support: National Council of Science (NSC95-2314-B-002-160).
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
References [1] Warshaw AL, Ferandez-Del Castillo C. Pancreatic carcinoma. N Engl J Med 1992;326:455–465. [2] DiMagno EP, Reber HA, Tempero MA. AGA technical review on the epidemiology, diagnosis, and treatment of pancreatic ductal adenocarcinoma. American Gastroenterologic Association. Gastroenterology 1999;117:1464–1484. [3] Sener SF, Fremgen A, Menck HR, Winchester DP. Pancreatic cancer: a report of treatment and survival trends for 100313 patients diagnosed from 1985 to 1995, using the National Cancer Database. J Am Coll Surg 1999;189:1–7. [4] Mollenhauer J, Roether I, Kern HF. Distribution of extracellular matrix proteins in pancreatic ductal adenocarcinoma and its influence on tumor cell proliferation in vitro. Pancreas 1987;2:14–24. [5] Korc M. Pancreatic cancer-associated stroma production. Am J Surg 2007;194:S84–S86. [6] Klo¨ppel G, Lingenthal G, Bu¨low MV, Kern HF. Histological and fine structural features of pancreatic ductal adenocarcinomas in relation to growth and prognosis: studies in xenografted tumours and clinco-histopathological correlation in a series of 75 cases. Histopathology (England) 1985;9:841–856. [7] Cubilla AL, Fitzgerald PJ. Morphological lesions associated with human primary invasive non-endocrine pancreas cancer. Cancer Res 1976;36:2690–2698. [8] Kuniyasu H, Abbruzzese JL, Cleary KR, Fidler IJ. Induction of ductal and stromal hyperplasia by basic fibroblast growth factor produced by human pancreatic carcinoma. Int J Oncol 2001;19:681–685. [9] Imamura T, Iguchi H, Manabe T, Ohshio G, Yoshimura T, Wang ZH, et al. Quantitative analysis of collagen and collagen subtypes I, III, and V in human pancreatic cancer, tumor-associated chronic pancreatitis, and alcoholic chronic pancreatitis. Pancreas 1995;11:357–364. [10] Lo¨hr M, Trautmann B, Gottler M, Peters S, Zauner I, Maillet B, et al. Human ductal adenocarcinomas of the pancreas express extracellular matrix proteins. Br J Cancer 1994;69:144–151. [11] Gress TM, Menke A, Bachem MG, Muller-Pillasch F, Ellenrieder V, Weidenbach H, et al. Role of extracellular matrix in pancreatic diseases. Digestion 1998;59:625–637. [12] Apte MV, Haber PS, Applegate TL, Norton ID, McCaughan GW, Korsten MA, et al. Periacinar stellate-shaped cells in rat pancreas: identification, isolation and culture. Gut 1998;43:128–133. [13] Bachem MG, Schneider E, Groß H, Weidenbach H, Schmid RM, Menke A, et al. Identification, culture, and characterization of pancreatic stellate cells in rats and humans. Gastroenterology 1998;115:421–432. [14] Apte MV, Wilson JS. Alcohol-induced pancreatic injury. Best Pract Res Clin Gastroenterol 2003;17:593–612. [15] Bachem MG, Schu¨nemann M, Ramadani M, Siech M, Beger H, Buck A, et al. Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells. Gastroenterol 2005;128:907–921. [16] Apte MV, Haber PS, Darby SJ, Rodgers SC, McCaughan GW, Korsten MA, et al. Pancreatic stellate cells are activated by proinflammatory cytokines: implications for pancreatic fibrogenesis. Gut 1999;44:534–541. [17] Haber PS, Keogh GW, Apte MV, Moran CS, Stewart NL, Crawford DH, et al. Activation of pancreatic stellate cells in human and experimental pancreatic fibrosis. Am J Pathol 1999;155:1087–1095. [18] Schneider E, Schmid-Kotsas A, Zhao J, Weidenbach H, Schmid RM, Menke A, et al. Identification of mediators stimulating proliferation and matrix synthesis of rat pancreatic stellate cells. Am J Physiol Cell Physiol 2001;281:C532–C543.
313
[19] Buchholz M, Kestler HA, Holzmann K, Ellenrieder V, Schneiderhan W, Siech M, et al. Transcriptome analysis of human hepatic and pancreatic stellate cells: organ-specific variations of a common transcriptional phenotype. J Mol Med 2005;83:795–805. [20] Nakamura T, Fidler IJ, Coombes KR. Gene expression profile of metastatic human pancreatic cancer cells depends on the organ environment. Cancer Res 2007;67:139–148. [21] Lee HS, Miau LH, Chen CH, Chiou LL, Huang GT, Yang PM, et al. Differential role of p38 in IL-1a induction of MMP-9 and MMP-13 in an established liver myofibroblast cell line. J Biomed Sci 2003;10:757–776. [22] Luzzi KJ, Mcdonald IC, Schmidt EE, Kerkvliet N, Morris VL, Chambers AF, et al. Multistep nature of metastatic ineffiencyDormancy of solitary cells after successful extravasation and limited survival of early micrometastases. Am J Pathol 1998;153:865–873. [23] Levy MT, McCaughan GW, Abbott CA, Park JE, Cunningham AM, Muller E, et al. Fibroblast activation protein: a cell surface dipeptidyl peptidase and gelatinase expressed by stellate cells at the tissue remodeling interface in human cirrhosis. Hepatology 1999;29:1768–1778. [24] Beauchamp RD, Lyons RM, Yang EY, Coffey Jr RJ, Moses HL. Expression of and response to growth regulatory peptides by two human pancreatic carcinoma cell lines. Pancreas 1990;5: 369–380. [25] Ebert M, Yokoyama M, Friess H, Kobrin MS, Buchler MW, Korc M. Induction of platelet-derived growth factor A and B chains and over-expression of their receptors in human pancreatic cancer. Int J Cancer 1995;62:529–535. [26] Kuniyasu H, Abbruzzese JL, Cleary KR, Fidler IJ. Induction of ductal and stromal hyperplasia by basic fibroblast growth factor produced by human pancreatic carcinoma. Int J Oncol 2001;19:681–685. [27] Kornmann M, Ishiwata T, Beger HG, Korc M. Fibroblast growth factor-5 stimulates mitogenic signaling and is overexpressed in human pancreatic cancer: evidence for autocrine and paracrine actions. Oncogene 1997;15:1417–1424. [28] Kornmann M, Beger HG, Korc M. Role of fibroblast growth factors and their receptors in pancreatic cancer and chronic pancreatitis. Pancreas 1998;17:169–175. [29] Leung HY, Gullick WJ, Lemoine NR. Expression and functional activity of fibroblast growth factor and their receptors in human pancreatic cancer. Int J Cancer 1994;59:667–675. [30] Leung HY, Hughes CM, Klo¨ppel G. Localization of expression of fibroblast growth factors and their receptors in pancreatic adenocarcinoma by in situ hybridization. Int J Oncol 1994;4:1219–1223. [31] Ohta T, Yamamoto M, Numata M, Iseki S, Tsukioka Y, Miyashita T, et al. Expression of basic fibroblast growth factor and its receptor in human pancreatic carcinomas. Br J Cancer 1995;72:824–831. [32] Friess H, Yamanaka Y, Buchler M, Ebert M, Beger HG, Gold LI, et al. Enhanced expression of transforming growth factor beta isoforms in pancreatic cancer correlates with decreased survival. Gastroenterology 1993;105:1846–1856. [33] Wong L, Yamasaki G, Johnson RJ, Friedman SL. Induction of beta-platelet-derived growth factor receptor in rat hepatic lipocyte during cellular activation in vivo and in culture. J Clin Invest 1994;94:1563–1569. [34] Marra F, Gentilini A, Pinzani M, Choudhury GG, Parola M, Herbst H, et al. Phosphatidyllinositol 3-kinase is required for platelet-derived growth factor’s action on hepatic stellate cells. Gastroenterology 1997;112:1297–1306. [35] Failli P, Ruocco C, De Franco R, Caligiuri A, Gentilini A, Giotti A, et al. The mitogenic effect of platelet-derived growth factor in human hepatic stellate cells requires calcium influx. Am J Physiol 1995:C1133–39.
314
Y.-W. Tien et al. / Journal of Hepatology 51 (2009) 307–314
[36] Gong W, Roth S, Michel K, Gressner AM. Isoforms and splice variant of transforming growth factor beta-binding protein in rat hepatic stellate cells. Gastroenterology 1998;114:352–363. [37] Yu C, Wang F, Jin C, Huang X, Miller DL, Basilico C, et al. Role of fibroblast growth factor type 1 and 2 in carbon tetrachlorideinduced hepatic injury and fibrogenesis. Am J Pathol 2003;163:1653–1662. [38] Hellerbrand C, Stefanovic B, Giordano F, Burchardt ER, Brenner DA. The role of TGF beta1 in initiating hepatic stellate cell activation in vivo. J Hepatol 1999;30:77–87. [39] Czochra P, Klopcic B, Meyer E, Herkel J, Garcia-Lazaro JF, Thieringer F, et al. Liver fibrosis induced by hepatic overexpression of PDGF-b in transgenic mice. J Hepatol 2006;45:419–428. [40] Henderson NC, Iredale JP. Liver fibrosis: cellular mechanisms of progression and resolution. Clin Sci 2007;112:265–280. [41] Iwamoto H, Nakamuta M, Tada S, Sugimoto R, Enjoji M, Nawata H. Platelet-derived growth factor receptor tyrosine kinase inhibitor AG1295 attenuates rat hepatic stellate cell growth. J Lab Clin Med 2000;135:406–412. [42] Brady LM, Fox ES, Fimmel CJ. Polyenylphosphatidylcholine inhibits PDGF-induced proliferation in rat hepatic stellate cells. Bioch Biophysi Res Comm 1998;248:174–179. [43] Scott F. Cytokines and fibrogenesis. Semin Liver Dis 1999;19:129–140.
[44] Tsukada S, Parsons CJ, Rippe RA. Mechanisms of liver fibrosis. Clin Chimica Acta 2006;364:33–60. [45] Okuno M, Moriwaki H, Imai S, Muto Y, Kawada N, Suzuki Y, et al. Retinoids exacerbate rat liver fibrosis by inducing the activation of latent TGFbeta in liver stellate cells. Hepatology 1997:913–921. [46] Mak KM, Leo MA, Lieber CS. Alcoholic liver injury in baboobn: transformation of lipocytes to transitional cells. Gastroenterology 1984;87:188–200. [47] Mak KM, Lieber CS. Lipocytes and transitional cells in alcoholic liverc disease: a morphometric study. Hepatology 1988;8:1027–1033. [48] Marra F, Choudhury GG, Pinzani M, Abboud HE. Regulation of platelet-derived growth factor secretion and gene expression in human fat-storing cells. Gastroenterology 1994;107:1110–1117. [49] Pinzani M, Gentilini A, Caligiuri A, De Franco R, Pellegrini G, Milani S, et al. Transforming growth factor-beta 1 regulates platelet-derived growth factor receptor beta subunit in human liver fat-storing cells. Hepatology 1995;21:232–239. [50] Rosenbaum J, Blazejewski S, Preaux AM, Mallat A, Dhumeaux D, Mavier P. Fibroblast growth factor 2 and trandforming growth factor beta 1 interactions in human liver myofibroblasts. Gastroenterology 1995;109:1986–1996.