Site-Specific Epithelial-Mesenchymal Interactions in Digestive Neuroendocrine Tumors
2000; Elsevier BV; Volume: 156; Issue: 2 Linguagem: Inglês
10.1016/s0002-9440(10)64771-2
ISSN1525-2191
AutoresJérôme Dumortier, Christelle Ratineau, Jean–Yves Scoazec, Céline Pourreyron, Wena Anderson, Marie‐France Jacquier, Martine Blanc, Christine Bernard, Claire Bellaton, L Rémy, J.A. Chayvialle, Colette Roche,
Tópico(s)Lung Cancer Research Studies
ResumoLittle is known about the functional interactions between digestive neuroendocrine tumor cells and their stromal microenvironment. The focus of our study is whether mesenchymal cells modulate peptide expression, cell proliferation, and invasiveness in digestive neuroendocrine tumor cells. We designed an experimental in vivo and in vitro study using the mouse enteroendocrine cell line STC-1. In vivo, STC-1 cells were injected subcutaneously in 18 immunosuppressed newborn rats. At day 21, all animals presented poorly differentiated neuroendocrine tumors with lung metastases. Subcutaneous tumors were usually limited by a capsule containing basement membrane components and myofibroblasts that presented a low mitotic index. Lung tumors were devoid of capsule and poor in myofibroblasts, and their mitotic index was high. The profile of peptide expression in STC-1 tumors was different from that of cultured STC-1 cells. In vitro, STC-1 cells were cultured with fibroblasts of different origins, including dermis, lung, digestive tract, and liver. Based on their origin, myofibroblasts differentially modulated hormone synthesis, proliferation, spreading, and adhesion of STC-1 cells. In conclusion, our results show that site-specific functional interactions between mesenchymal and neuroendocrine cells may contribute to modulating the behavior of digestive neuroendocrine tumors, depending on their growth site. Little is known about the functional interactions between digestive neuroendocrine tumor cells and their stromal microenvironment. The focus of our study is whether mesenchymal cells modulate peptide expression, cell proliferation, and invasiveness in digestive neuroendocrine tumor cells. We designed an experimental in vivo and in vitro study using the mouse enteroendocrine cell line STC-1. In vivo, STC-1 cells were injected subcutaneously in 18 immunosuppressed newborn rats. At day 21, all animals presented poorly differentiated neuroendocrine tumors with lung metastases. Subcutaneous tumors were usually limited by a capsule containing basement membrane components and myofibroblasts that presented a low mitotic index. Lung tumors were devoid of capsule and poor in myofibroblasts, and their mitotic index was high. The profile of peptide expression in STC-1 tumors was different from that of cultured STC-1 cells. In vitro, STC-1 cells were cultured with fibroblasts of different origins, including dermis, lung, digestive tract, and liver. Based on their origin, myofibroblasts differentially modulated hormone synthesis, proliferation, spreading, and adhesion of STC-1 cells. In conclusion, our results show that site-specific functional interactions between mesenchymal and neuroendocrine cells may contribute to modulating the behavior of digestive neuroendocrine tumors, depending on their growth site. Gastrointestinal neuroendocrine tumors contain a distinctive fibrovascular stroma, characterized by a dense network of capillary vessels embedded in variable amounts of extracellular matrix.1Solcia E Capella C Fiocca R Cornaggia M Bosi F The gastroenteropancreatic endocrine system and related tumours.Gastroenterol Clin North Am. 1989; 18: 671-693PubMed Google Scholar, 2Lechago J Gastrointestinal neuroendocrine cell proliferations.Hum Pathol. 1994; 25: 1114-1122Abstract Full Text PDF PubMed Scopus (19) Google Scholar, 3Capella C Heitz PU Höffler H Solcia E Klöppel G Revised classification of neuroendocrine tumors of the lung, pancreas and gut.Virchows Arch. 1995; 425: 547-560Crossref PubMed Scopus (500) Google Scholar In many instances, the stroma of neuroendocrine tumors is formed by a delicate connective matrix, intimately associated with tumor cells. However, in some tumors, such as midgut carcinoids, the amount of extracellular matrix may be very high, resulting in a so-called desmoplastic appearance. Histopathological studies have shown that neuroendocrine tumor cells are able to synthesize several growth factors known to promote mesenchymal cell migration and proliferation and to stimulate extracellular-matrix synthesis and deposition.4Funa K Papanicolaou V Juhlin C Rastad J Akerström G Heldin C Öberg K Expression of platelet-derived growth factor β-receptors on stromal tissue cells in human carcinoid tumors.Cancer Res. 1990; 50: 748-753PubMed Google Scholar, 5Chaudhry A Papanicolaou V Oberg K Heldin CH Funa K Expression of platelet-derived growth factor and its receptors in neuroendocrine tumors of the digestive system.Cancer Res. 1992; 52: 1006-1012PubMed Google Scholar, 6Chaudhry A Funa K Oberg K Expression of growth factor peptides and their receptors in neuroendocrine tumors of the digestive system.Acta Oncol. 1993; 32: 107-114Crossref PubMed Scopus (98) Google Scholar, 7Chaudhry A Oberg K Transforming growth factor α and epithelial growth factor receptor expression in neuroendocrine tumors of the digestive system.Diagn Oncol. 1993; 3: 81-85Google Scholar, 8Chaudhry A Oberg K Gobl A Heldin CH Funa K Expression of transforming growth factors β 1, β 2, β 3 in neuroendocrine tumors of the digestive system.Anticancer Res. 1994; 14: 2085-2091PubMed Google Scholar, 9Nilsson O Wangberg B McRae A Dahlstrom A Ahlman H Growth factors and carcinoid tumours.Acta Oncol. 1993; 32: 115-124Crossref PubMed Scopus (48) Google Scholar, 10Nilsson O Wangberg B Kolby L Schultz GS Ahlman H Expression of transforming growth factor α and its receptor in human neuroendocrine tumours.Int J Cancer. 1995; 60: 645-651Crossref PubMed Scopus (81) Google Scholar, 11Terris B Scoazec JY Rubbia L Brégeaud L Pepper MS Ruszniewski P Belghiti J Fléjou JF Degott C Expression of vascular endothelial growth factor in digestive neuroendocrine tumors.Histopathology. 1998; 32: 133-138Crossref PubMed Scopus (246) Google Scholar In vitro studies have confirmed that some neuroendocrine cell lines, such as the BON cell line, may induce cell proliferation and extracellular matrix protein synthesis in fibroblastic cells.12Townsend Jr, CM Ishizuka J Thompson JC Studies of growth regulation in a neuroendocrine cell line.Acta Oncol. 1993; 32: 125-130Crossref PubMed Scopus (44) Google Scholar, 13Beauchamp RD Coffey Jr, RJ Lyons RM Perkett EA Townsend Jr, CM Moses HL Human carcinoid cell production of paracrine growth factors that can stimulate fibroblast and endothelial cell growth.Cancer Res. 1991; 51: 5253-5260PubMed Google Scholar It is, therefore, likely that neuroendocrine tumor cells regulate stromal cell proliferation and activity through paracrine interactions. The possible reciprocal influence of stromal cells on the growth and differentiation of neuroendocrine tumor cells has received little attention. Clinical and experimental data show that, in various types of cancer, mesenchymal cells are able to modulate the proliferative activity, the invasive and metastatic properties, and the differentiation state of neoplastic cells. It can therefore be hypothesized that, as for other types of tumors, the biological characteristics of digestive neuroendocrine tumors are modulated by mesenchymally derived factors. This hypothesis is supported by embryological14Gittes GK Studies of early events in pancreatic organogenesis.Ann NY Acad Sci. 1994; 733: 68-74Crossref PubMed Scopus (4) Google Scholar and experimental15Montesano R Mouron P Amherdt M Orci L Collagen matrix promotes reorganization of pancreatic endocrine cell monolayers into islet-like organoids.J Cell Biol. 1983; 97: 935-939Crossref PubMed Scopus (153) Google Scholar, 16Ratineau C Plateroti M Dumortier J Blanc M Kedinger M Chayvialle JA Roche C Intestinal-type fibroblasts selectively influence proliferation rate and peptide synthesis in the murine entero-endocrine cell line STC-1.Differentiation. 1997; 62: 139-147Crossref PubMed Scopus (13) Google Scholar data, showing the role of extracellular matrix proteins and mesenchymal factors in the normal differentiation process of digestive neuroendocrine cells. Moreover, tissue-specific mesenchymal influences may help to explain the differences in hormone content, stromal characteristics, and local behavior observed between primary neuroendocrine tumors originating from the different segments of the digestive tract (foregut, midgut, hindgut), and between the primary and secondary lesions of the same tumors. Recent experimental evidence has underlined the marked functional differences existing between fibroblasts originating from the various segments of the digestive tract.17Plateroti M Rubin DC Duluc I Singh R Foltzer-Jourdainne C Freund JN Kedinger M Subepithelial fibroblast cell lines from different levels of gut axis display regional characteristics.Am J Physiol. 1998; 274: G945-G954PubMed Google Scholar In turn, gut-associated fibroblasts are functionally different from the various organ-specific mesenchymal cell subsets identified so far, such as those of the liver18Pinzani M Hepatic stellate (ITO) cells: expanding roles for a liver-specific pericyte.J Hepatol. 1995; 22: 700-706Abstract Full Text PDF PubMed Scopus (111) Google Scholar and the lung,19Ohmichi H Koshimizu U Matsumoto K Nakamura T Hepatocyte growth factor (HGF) acts as a mesenchyme-derived morphogenic factor during fetal lung development.Development. 1998; 125: 1315-1324Crossref PubMed Google Scholar which represent the most frequent metastatic sites of human digestive neuroendocrine tumors. Such site-specific differences in their mesenchymal environment may contribute to modulating the behavior of neuroendocrine tumor cells. To test these hypotheses, we 1) evaluated whether mesenchymal cells may modulate the hormone content, cell proliferative activity, and invasive capacities of digestive neuroendocrine tumor cells and 2) searched for site-specific differences in mesenchymal interactions with digestive neuroendocrine tumor cells. To address our questions, we designed an experimental in vivo and in vitro study using the enteroendocrine mouse cell line STC-1.20Rindi G Grant SG Yiangou Y Ghatei MA Bloom SR Bautch VL Solcia E Polak JM Development of neuroendocrine tumors in the gastrointestinal tract of transgenic mice: heterogeneity of hormone expression.Am J Pathol. 1990; 136: 1349-1363PubMed Google Scholar The intestinal STC-1 plurihormonal cell line, a gift from Dr. D. Hanahan through the courtesy of Dr. A. Leiter (New England Medical Center, Boston, MA), is derived from an endocrine tumor that developed in the small intestine of a double transgenic mouse expressing the rat insulin promoter linked to the simian virus 40 large-T antigen and to the polyomavirus small-t antigen, respectively.20Rindi G Grant SG Yiangou Y Ghatei MA Bloom SR Bautch VL Solcia E Polak JM Development of neuroendocrine tumors in the gastrointestinal tract of transgenic mice: heterogeneity of hormone expression.Am J Pathol. 1990; 136: 1349-1363PubMed Google Scholar. The standard culture medium consisted of Dulbecco's modified Eagle's Medium (DMEM) supplemented with 5% fetal calf serum (FCS), 2 mmol/L glutamine, and antibiotics (100 UI/ml penicillin plus 50 mmol/L streptomycin). Mesenchyme-derived intestinal cell lines (MICs), a gift from M Plateroti, Institut National de la Santé et de la Recherche Médicale (INSERM) U381, Strasbourg, France, were isolated from 8-day postnatal rats. Clonal cell lines that were derived from mixed subepithelial fibroblast parental cell lines were characterized as myofibroblasts: MIC 101–1, MIC-219, and MIC-316, respectively, from jejunum, ileum, and colon.17Plateroti M Rubin DC Duluc I Singh R Foltzer-Jourdainne C Freund JN Kedinger M Subepithelial fibroblast cell lines from different levels of gut axis display regional characteristics.Am J Physiol. 1998; 274: G945-G954PubMed Google Scholar All of these cell lines were maintained in DMEM supplemented with 10% FCS, 2 mmol/L glutamine, antibiotics, and 0.25 U/ml insulin. All cultures were done in a humidified 8% CO2/92% air incubator at 37°C. The antibodies used in this study are listed in Table 1.Table 1Antibodies Used in the StudyReactivitymAb or pAb*mAb, monoclonal antibody; pAb, polyclonal antibody.OriginApplication (working dilution)Chromogranin ARabbit pAbIncstar BioRad, Ivrysur Seine (France)IH (1:100)VimentinMouse mAb clone V9Dako, Glostrup (Denmark)IC (1:25)α-Smooth muscle actinMouse mAbSigma, St Louis, MOIH, IC (1:400)Rat PECAM-1Mouse mAbR&D Systems, Oxon (UK)IH (1:500)Laminin-1Rabbit pAb 1219JC Lissitzky, INSERM U387, Marseille (France)IH, IC (1:100)Type IV collagenRabbit pAbChemicon, Temecula, CA (USA)IH, IC (1:100)CholecystokininRabbit pAbNovocastra, Nottingham (UK)IH (1:100)GlucagonRabbit pAbNovocastra, Nottingham (UK)IH (1:1000)SomatostatinRabbit pAb (56D)Our laboratory (23)IH (1:500)RIA (1:250000)CholecystokininRabbit pAb (39A)Our laboratory (24)RIA (1:300000)Truncated glucagon-like peptideRabbit pAb (199D)Our laboratory (25)RIA (1:250000)Antibodies used for immunohistochemistry (IH), immunocytochemistry (IC) and radioimmunoassay (RIA).* mAb, monoclonal antibody; pAb, polyclonal antibody. Open table in a new tab Antibodies used for immunohistochemistry (IH), immunocytochemistry (IC) and radioimmunoassay (RIA). STC-1 cells in exponential growth were detached by trypsinization. They were suspended in culture medium, counted, centrifuged (10 minutes, 1500 rpm), and resuspended in phosphate buffered saline (PBS). Eighteen Wistar newborn rats received a subcutaneous injection in the abdominal region of 1.2 × 106 cells suspended in 100 μl of PBS. All of the rats were subsequently immunosuppressed by dorsal subcutaneous injections of antithymocyte serum21Bailly M Bertrand S Dore JF Human tumor spontaneous metastasis in immunosuppressed newborn rats. I. Characterization of the bioassay.Int J Cancer. 1991; 49: 457-466Crossref PubMed Scopus (20) Google Scholar on days 0, 2, 7, and14 and maintained in a specific pathogen-free environment throughout the experiment. Three weeks after cell inoculation, subcutaneous tumors and lung metastases were counted, excised, and processed for morphological, immunohistochemical, and biochemical analyses. Tissue Processing: Tissue samples were divided into three parts. The first part was processed for light-microscopical examination. Tissue samples were fixed in formalin and embedded in paraffin. Three-μm-thick sections were stained with hematoxylin and eosin. Another part of the tissue samples was processed for ultrastructural examination. Tissue samples were fixed with 2.5% glutaraldehyde in 0.1 mol/L sodium cacodylate buffer for 60 minutes. These samples were subsequently rinsed overnight in the same buffer, then postfixed with 1% osmium tetroxide. After dehydration in graded ethanols and propilene oxide, they were embedded in epoxy resin. Ultrathin sections were prepared, stained with uranyl acetate and lead citrate, and examined with a Jeol 100 CX 2 electron microscope (JEOL, Tokyo, Japan). The remaining tissue samples were immediately snap-frozen in Freon that had been prechilled in liquid nitrogen. Immunohistochemistry: For detection of cell markers and extracellular-matrix proteins, an indirect immunofluorescence technique was applied to cryostat sections of frozen tissue. Briefly, acetone-fixed 5-μm-thick cryostat sections were incubated sequentially, with the primary antibody diluted in PBS, then with polyclonal fluorescein isothiocyanate-conjugated species-specific anti-mouse or -rabbit immunoglobulin antibodies (Immunotech, Marseille-Luminy, France). For detection of endogenous peptides, a streptavidin-biotin-peroxidase technique was applied to sections of formalin-fixed, paraffin-embedded tissue. As a brief description of the procedure, 3-μm-thick sections were allowed to dry on silane-coated slides. After rehydration, they were incubated for 60 minutes with the primary antibody at the appropriate dilution. Antigen-antibody complexes were revealed by the streptavidin-biotin technique (Dako, Glostrup, Denmark), using diaminobenzidine as chromagen. In all techniques, negative controls were included and obtained by either omission of the primary antibody, substituted for by PBS alone, or incubation with isotypic immunoglobulins. Quantitation of Morphological Data: The following data were quantitated: mitotic index, apoptotic index, extent of basement membrane deposition at the periphery of tumor nodules, intratumoral vascular density, and number of peptide-containing cells. For quantitation of the mitotic index and of the apoptotic index, the respective numbers of mitotic and apoptotic bodies within tumor nodules were counted in 10 consecutive fields with a ×40 objective lens, as previously described.22La Rosa S Sessa F Capella C Riva C Leone BE Klersy C Rindi G Solcia E Prognostic criteria in nonfunctioning pancreatic endocrine tumours.Virchows Arch. 1996; 429: 323-333PubMed Google Scholar Results were expressed as the mean number of mitoses or apoptotic bodies per field. For quantitation of basement membrane deposition around tumor nodules, a semiautomated image analysis technique was used to evaluate 1) the total length of the perimeter of tumor nodules and b) the fraction of the perimeter length occupied by immunodetectable deposits of laminin 1. The final result was expressed as a percentage of the total perimeter length. For quantitation of intratumoral vascular density, the structures expressing PECAM-1, a specific endothelial cell marker, were counted in five consecutive fields with a ×40 objective. Results were expressed as the mean number of vessels per field. For quantitation of peptide-containing cells, the percentages of cells labeled by the antibodies used for detection of glucagon, somatostatin (STS), and cholecystokinin (CCK) were evaluated in 10 consecutive fields with a ×40 objective. Radioimmunoassay (RIA) for the detection of endogenous peptides was performed only on homogenates of subcutaneous tumors. After peptide extraction, RIAs were performed using the rabbit polyclonal antibodies 56D for STS, 39A for CCK, and 199D for truncated glucagon-like peptide 1 (TGLP-1), as previously described.23Chayvialle JA Miyata M Rayford PL Thompson JC Effects of test meal, intragastric nutrients, and intraduodenal bile on plasma concentrations of immunoreactive somatostatin and vasoactive intestinal peptide in dogs.Gastroenterology. 1980; 79: 844-852PubMed Scopus (109) Google Scholar, 24Chery-Croze S Kocher L Bernard C Chayvialle JA Substance P-, somatostatin-, vasoactive intestinal peptide-, and cholecystokinin-like levels in the spinal cord of polyarthritic rats.Brain Res. 1985; 339: 183-185Crossref PubMed Scopus (23) Google Scholar, 25Abello J Ye F Bosshard A Bernard C Cuber JC Chayvialle JA Stimulation of glucagon-like peptide-1 secretion by muscarinic agonist in a murine intestinal endocrine cell line.Endocrinology. 1994; 134: 2011-2017PubMed Google Scholar Peptide concentrations were calculated as nanograms per milligram wet weight. Cellular CCK, TGLP-1, and STS concentrations were calculated as ng/106cells. DMEM, additive, and FCS were obtained from Life Technologies, Inc. (Cergy-Pontoise, France). α-32P-labeled dCTP, 35S-methionine, and 3H-thymidine were purchased from Amersham (Les Ulis, France). All other reagents were of analytical grade. Dermal, lung, and liver fragments of Wistar newborn rats were taken and washed once with a 30% sodium hypochloride solution and twice with sterile PBS. Each fragment was dissociated first mechanically by using scissors and then enzymatically with collagenase I (Sigma Chemical Co., St. Louis, MO) in culture medium. After incubation at 37°C (15–60 minutes), the tissue fragments were filtrated and centrifuged (15 minutes, 1500 rpm). The cell pellet was resuspended with 5 ml of culture medium supplemented with 10% FCS and was plated in culture flasks. Primary cultured fibroblasts were respectively obtained from subcutis, lung, and liver and were used at passages 1–5. Organ-specific fibroblasts were cultured for 3 days in Lab-Tek tissue culture chambers/slides (Miles, Elkhart, IN). After PBS rinsing, they were fixed and permeabilized with 100% cold methanol for 5 minutes. Cells were incubated for 1 hour with specific antibodies diluted from 1:25 to 1:400. Controls were incubated without specific antibodies. After abundant washing with 10% FCS in PBS, slides were incubated with fluorescein isothiocyanate (1/50)- or rhodamin (1/100)-labeled, species-specific antibodies for 30 minutes. After rinsing, coverslips were mounted with Fluoprep (Biomérieux, Lyon, France), and the slides were examined using an epifluorescence microscope. STC-1 cells were cultured on confluent fibroblastic layers. To describe the procedure briefly, 10 × 104 STC-1 cells that had been suspended in medium supplemented with 5% FCS were seeded onto fibroblast monolayers in 24-well plates (Costar, Cambridge, MA) in medium supplemented with 5% FCS. Direct cocultures were maintained for 5 days. All experiments were performed in triplicate. Ultrastructural analysis of cellular interactions was performed on 3. and 5-day direct cocultures in 60-mm petri dishes (Costar). Cocultures were fixed with 2.5% glutaraldehyde in 0.1 mol/L sodium cacodylate buffer for 15 minutes and 60 minutes, respectively. They were subsequently rinsed overnight in the same buffer and postfixed with 1. osmium tetroxide. After dehydration in graded ethanols and propilene oxide, they were embedded in epoxy resin. Ultrathin sections were prepared, stained with uranyl acetate and lead citrate, and examined with a Jeol 100 CX 2 electron microscope (JEOL). To study the adhesion of endocrine tumor cells onto fibroblast monolayers, fibroblasts were seeded in 24-well plates (Costar) and grown to reach confluence. The wells were then washed twice with PBS, and nonspecific sites were saturated with 0.1% bovine serum albumin (BSA). After washing with PBS, STC-1 cells (4 × 104 cells/ml) prelabeled with 35S-methionine (Amersham) were added onto the fibroblast monolayer in serum-free medium. Control experiments were made in plastic wells after BSA saturation. At different times (30 minutes, 1 hour) after seeding, the unattached cells were washed away. Adherent cells were removed in 100 ml of 1% sodium dodecyl sulfate (SDS), and the radioactivity was counted in a scintillation counter. Similarly, the adhesion of endocrine tumor cells onto extracellular-matrix proteins was performed. The wells were coated with laminin-1 (5 μg/ml), type IV collagen (10 μg/ml), or type I collagen (10 μg/ml) (Becton Dickinson, Mountain View, CA) and washed twice with PBS, and nonspecific sites were saturated with 0.1% BSA. To study the cell spreading on fibroblasts, cultures were initiated as described above and daily were observed under phase microscopy and photographed. The spreading and adhesion of endocrine tumor cells were also studied on extracellular-matrix proteins in 24-well plates (Costar) in a similar manner. The wells were coated with laminin-1, type IV collagen, or type I collagen as described for attachment assays. All experiments were performed in triplicate. In vitro indirect coculture assays were performed using modified Boyden chambers (Falcon). STC-1 cells (105) suspended in medium supplemented with 5. FCS were seeded into the upper wells on polycarbonate filters (0.45-μm porosity). Fibroblast monolayers were in the lower wells in medium supplemented with 5% FCS. Indirect cocultures were maintained for 3 days. Myofibroblast-conditioned media (FCM) were prepared by the method of Cornil et al.26Cornil I Theodorescu D Man S Herlyn M Jambrosic J Kerbel RS Fibroblast cell interactions with human melanoma cells affect tumor cell growth as a function of tumor progression.Proc Natl Acad Sci USA. 1991; 88: 6028-6032Crossref PubMed Scopus (194) Google Scholar Fibroblastic cells were seeded into 150-cm2 culture flasks in 40 ml of culture medium and grown to confluence, at which time the medium was removed, and 40 ml of fresh DMEM containing 5% FCS, glutamine, and antibiotics were added to each flask. After a 48-hour incubation, FCM were collected, centrifuged at 1000 × g for 5 minutes to remove cellular debris, pooled, and stored at −20°C. They were thawed and centrifuged immediately before use. Conditioned media were diluted 1:1 (v/v) with fresh DMEM containing 5% FCS, glutamine, and antibiotics. STC-1 cells were cultured in diluted FCM for 3 days. All experiments were performed in triplicate. Cell proliferation was assessed by measuring the rate of DNA synthesis. This last parameter was determined through 3H-thymidine incorporation into the trichloroacetic acid-insoluble cellular fraction. After a 2-hour incubation with 1 mCi/ml of 3H- thymidine, cells were washed twice in ice-cold PBS and precipitated with ice-cold 5. trichloroacetic acid. Then cells were washed once with ice-cold 95. ethanol and solubilized with 1 N NaOH. An aliquot was neutralized, and the radioactivity was determined in a liquid scintillation counter. All experiments were performed in triplicate. Culture medium was removed and cells were harvested in guanidine thiocyanate. Total RNA was extracted by the method of Chomczynski and Sacchi.27Chomczynski P Sacchi N Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction.Anal Biochem. 1987; 162: 156-159Crossref PubMed Scopus (63087) Google Scholar Northern blot analysis of RNA was performed as previously described,28Ratineau C Roche C Chuzel F Cordier-Bussai M Blanc M Bernard C Cuber JC Chayvialle JA Regulation of intestinal cholecystokinin gene expression by glucocorticoids.J Endocrinol. 1996; 151: 137-145Crossref PubMed Scopus (8) Google Scholar using complementary DNA (cDNA. probes. Probes for CCK (575-bp cDNA fragment), STS (570-bp cDNA fragment), and glucagon (408-bp cDNA fragment) were kindly provided by J Philippe (Geneva, Switzerland). All membranes were prehybridized in 1 mol/L NaCl, 1% SDS, and 10. dextran for 3 hours at 60°C. cDNA probes were labeled with α-32P-dCTP, using a random primer labeling kit (Amersham). Membranes were first hybridized with CCK and glucagon radiolabeled cDNA probe overnight in hybridization buffer at 60°C, washed for 20 minutes in 2× standard saline citrate (SSC) and 2% SDS at 60°C, then twice in 0.2× SSC and 0.2% SDS for 20 minutes each at 60°C, and exposed to Hyperfilm (Amersham) at −70°C with intensifying screens. Membranes were then immersed in a boiling solution of 0.01× SSC and 0.01% SDS maintained at 70°C for 30 minutes. Membranes were hybridized with the STS-radiolabeled cDNA probe, in the hybridization buffer supplemented with salmon sperm DNA (100 mg/ml) to avoid nonspecific staining. The equal loading of the lanes with total RNA was checked by staining 28S and 18S ribosomal RNA with ethidium bromide. Each dehybridization step was verified by exposing the membranes to Hyperfilm. STC-1 cells from indirect cocultures were lysed in lysis buffer. An aliquot of each lysate, containing 300 mg of protein as determined by the Bradford method was resolved on a 10% SDS-polyacrylamide gel. The fractionated proteins were transferred to Protran nitrocellulose transfer membranes (Schleicher & Schüll, Ecquevilly, France). The filter was blocked and incubated with a mouse monoclonal antibody against Ets-1 (Transduction Laboratories, Lexington, KY). Antibody-antigen complexes were visualized by enhanced chemiluminescence (Amersham), according to the manufacturer's instructions. Subcutaneous tumors located at the site of injection were obtained in all 18 animals in the study group. At sacrifice 21 days after injection, tumor sizes ranged from 4 to 7 mm (mean ± SD, 5.2 ± 1.5). At gross examination, there were no significant necrotico-hemorragic lesions. In all of the animals included in the study group, lung metastases were detected. They were usually less than 1 mm in diameter. The histological appearance of STC-1-induced subcutaneous tumors was comparable in all of the animals of the study group. STC-1-induced tumors presented the aspect of poorly differentiated neuroendocrine tumors (Figure 1a). The tumor population was composed of monomorphic, small- to medium-sized cells, characterized by a centrally located, often irregular nucleus and by an abundant, basophilic cytoplasm (Figure 1b). The architecture of the neoplastic population was usually compact. However, characteristic trabecular architecture could be found in all tumors examined. Lung tumors developing in STC-1-injected immunosuppressed newborn rats presented as tiny nodules (Figure 1c) formed by medium- to large-sized neuroendocrine cells. Cells were densely packed, and no evidence of trabecular architecture was observed. In both lung and subcutis, tumor cells retained a neuroendocrine differentiation, as assessed by their constant expression of chromogranin A (data not shown) and the presence of neurosecretory granules at ultrastructural examination (Figure 1d). In subcutaneous tumors, the mitotic index of the endocrine neoplastic population was high (Table 2). It varied by case from 1 to 8 mitoses/10 high-magnification fields. In lung tumors, the mitotic index ranged from 4 to 10 mitoses/10 high-magnification fields (Table 2). The mitotic index was significantly higher in pulmonary lesions than in subcutaneous tumors (Table 2). Comparison between subcutaneous and lung tumors in the same animal confirmed this result (Figure 2).Table 2Comparison of Morphological Parameters in Subcutaneous and Lung TumorsParameterSubcutaneous tumorsLung tumorsMitotic index (mean number of mitoses per field)4.2 ± 2.36.5 ± 2.0P < 0.01Apoptotic index (mean number of apoptotic bodies per field)7.9 ± 2.22.2 ± 1.1P < 0.01Basement membrane deposition (% of the circumference)84.9 ± 8.723.7 ± 11.5P < 0.001Intratumoral vascular density (mean number of vessels per field)17.3 ± 0.93.3 ± 1.8P < 0.001Results are given as mean ± SD. Differences were considered as significa
Referência(s)