Modeling Colon Adenocarcinomas in Vitro
2011; Elsevier BV; Volume: 179; Issue: 1 Linguagem: Inglês
10.1016/j.ajpath.2011.03.015
ISSN1525-2191
AutoresHelmut Dolznig, Christian Rupp, Christina Puri, Christian Haslinger, Norbert Schweifer, Elisabeth Wieser, Dontscho Kerjaschki, Pilar Garin‐Chesa,
Tópico(s)Genetic factors in colorectal cancer
ResumoActivated tumor stroma participates in tumor cell growth, invasion, and metastasis. Normal fibroblasts and cancer-associated fibroblasts (CAFs) have been shown to display distinct gene expression signatures. This molecular heterogeneity may influence the way tumor cells migrate, proliferate, and survive during tumor progression. To test this hypothesis and to better understand the molecular mechanisms that control these interactions, we established a three-dimensional (3D) human cell culture system that recapitulates the tumor heterogeneity observed in vivo. Human colon tumor cells were grown as multicellular spheroids and subsequently co-cultured with normal fibroblasts or CAFs in collagen I gels. This in vitro model system closely mirrors the architecture of human epithelial cancers and allows the characterization of the tumor cell–stroma interactions phenotypically and at the molecular level. Using GeneChip analysis, antibody arrays, and enzyme-linked immunosorbent assays, we demonstrate that the interaction of colon cancer cells with stromal fibroblasts induced different highly relevant cancer expression profiles. Genes involved in invasion, extracellular matrix remodeling, inflammation, and angiogenesis were differentially regulated in our 3D carcinoma model. The modular setup, reproducibility, and robustness of the model make it a powerful tool to identify target molecules involved in signaling pathways that mediate paracrine interactions in the tumor microenvironment and to validate the influence of these molecular targets during tumor growth and invasion in the supporting stroma. Activated tumor stroma participates in tumor cell growth, invasion, and metastasis. Normal fibroblasts and cancer-associated fibroblasts (CAFs) have been shown to display distinct gene expression signatures. This molecular heterogeneity may influence the way tumor cells migrate, proliferate, and survive during tumor progression. To test this hypothesis and to better understand the molecular mechanisms that control these interactions, we established a three-dimensional (3D) human cell culture system that recapitulates the tumor heterogeneity observed in vivo. Human colon tumor cells were grown as multicellular spheroids and subsequently co-cultured with normal fibroblasts or CAFs in collagen I gels. This in vitro model system closely mirrors the architecture of human epithelial cancers and allows the characterization of the tumor cell–stroma interactions phenotypically and at the molecular level. Using GeneChip analysis, antibody arrays, and enzyme-linked immunosorbent assays, we demonstrate that the interaction of colon cancer cells with stromal fibroblasts induced different highly relevant cancer expression profiles. Genes involved in invasion, extracellular matrix remodeling, inflammation, and angiogenesis were differentially regulated in our 3D carcinoma model. The modular setup, reproducibility, and robustness of the model make it a powerful tool to identify target molecules involved in signaling pathways that mediate paracrine interactions in the tumor microenvironment and to validate the influence of these molecular targets during tumor growth and invasion in the supporting stroma. Carcinomas comprise approximately 90% of all human cancers, including lung, colon, breast, and prostate carcinomas, which together cause approximately 50% of all cancer deaths.1Jemal A. Siegel R. Ward E. Murray T. Xu J. Thun M.J. Cancer statistics, 2007.CA Cancer J Clin. 2007; 57: 43-66Crossref PubMed Scopus (7452) Google Scholar Carcinomas consistently develop as highly complex structures composed of genetically altered tumor cells, together with fibroblasts, endothelial cells, pericytes, and inflammatory cells embedded in an extracellular matrix (ECM) of proteins. The tumor stroma influences the development of malignant tumors2Bhowmick N.A. Neilson E.G. Moses H.L. Stromal fibroblasts in cancer initiation and progression.Nature. 2004; 432: 332-337Crossref PubMed Scopus (1846) Google Scholar, 3Bissell M.J. Radisky D. Putting tumours in context.Nat Rev Cancer. 2001; 1: 46-54Crossref PubMed Scopus (1731) Google Scholar, 4Mueller M.M. Fusenig N.E. Friends or foes - bipolar effects of the tumour stroma in cancer.Nat Rev Cancer. 2004; 4: 839-849Crossref PubMed Scopus (1462) Google Scholar; for example, the recruitment of blood vessels is ultimately required for tumor growth.3Bissell M.J. Radisky D. Putting tumours in context.Nat Rev Cancer. 2001; 1: 46-54Crossref PubMed Scopus (1731) Google Scholar, 5Folkman J. Hanahan D. Switch to the angiogenic phenotype during tumorigenesis.Princess Takamatsu Symp. 1991; 22: 339-347PubMed Google Scholar Fibroblasts comprise the major cellular component of the stroma of carcinomas. These cancer-associated fibroblasts (CAFs) have been shown to stimulate tumor progression of initiated nontumorigenic prostate epithelial cells6Olumi A.F. Grossfeld G.D. Hayward S.W. Carroll P.R. Tlsty T.D. Cunha G.R. Carcinoma-associated fibroblasts direct tumor progression of initiated human prostatic epithelium.Cancer Res. 1999; 59: 5002-5011PubMed Google Scholar and to promote the growth of breast and colon cancer in animal models.7Orimo A. Weinberg R.A. Stromal fibroblasts in cancer: a novel tumor-promoting cell type.Cell Cycle. 2006; 5: 1597-1601Crossref PubMed Scopus (645) Google Scholar, 8Yauch R.L. Gould S.E. Scales S.J. Tang T. Tian H. Ahn C.P. Marshall D. Fu L. Januario T. Kallop D. Nannini-Pepe M. Kotkow K. Marsters J.C. Rubin L.L. de Sauvage F.J. A paracrine requirement for hedgehog signalling in cancer.Nature. 2008; 455: 406-410Crossref PubMed Scopus (815) Google Scholar Recently, it was demonstrated that the interaction of colon cancer cells with stromal cells activates the β-catenin pathway in the cancer cells and leads to an increase in colon cancer stem cells.9Vermeulen L. De Sousa E.M.F. van der Heijden M. Cameron K. de Jong J.H. Borovski T. Tuynman J.B. Todaro M. Merz C. Rodermond H. Sprick M.R. Kemper K. Richel D.J. Stassi G. Medema J.P. Wnt activity defines colon cancer stem cells and is regulated by the microenvironment.Nat Cell Biol. 2010; 12: 468-476Crossref PubMed Scopus (1406) Google Scholar Therefore, understanding the molecular mechanisms involved in the paracrine interactions between the tumor and the surrounding stroma will help to identify new molecules that may serve as potential drug targets. The culture of cells on two-dimensional (2D) surfaces has provided groundbreaking insights into basic cell biology and tumorigenesis.10Bodnar A.G. Ouellette M. Frolkis M. Holt S.E. Chiu C.P. Morin G.B. Harley C.B. Shay J.W. Lichtsteiner S. Wright W.E. Extension of life-span by introduction of telomerase into normal human cells.Science. 1998; 279: 349-352Crossref PubMed Scopus (4105) Google Scholar, 11Hahn W.C. Counter C.M. Lundberg A.S. Beijersbergen R.L. Brooks M.W. Weinberg R.A. Creation of human tumour cells with defined genetic elements.Nature. 1999; 400: 464-468Crossref PubMed Scopus (1979) Google Scholar, 12Iyer V.R. Eisen M.B. Ross D.T. Schuler G. Moore T. Lee J.C. Trent J.M. Staudt L.M. Hudson Jr., J. Boguski M.S. Lashkari D. Shalon D. Botstein D. Brown P.O. The transcriptional program in the response of human fibroblasts to serum.Science. 1999; 283: 83-87Crossref PubMed Scopus (1724) Google Scholar However, most physiologic parameters of organs or tumors, such as tissue architecture, cell-to-cell and cell-to-matrix interaction, mechanical properties, and biochemical networks are lost under these simplified conditions. Cells grown in three-dimensional (3D) scaffolds or as 3D aggregates (multicellular spheroids) much better recapitulate the in vivo structure of tissues13Bell E. Ehrlich H.P. Buttle D.J. Nakatsuji T. Living tissue formed in vitro and accepted as skin-equivalent tissue of full thickness.Science. 1981; 211: 1052-1054Crossref PubMed Scopus (854) Google Scholar, 14Runswick S.K. O'Hare M.J. Jones L. Streuli C.H. Garrod D.R. Desmosomal adhesion regulates epithelial morphogenesis and cell positioning.Nat Cell Biol. 2001; 3: 823-830Crossref PubMed Scopus (204) Google Scholar, 15Gudjonsson T. Ronnov-Jessen L. Villadsen R. Rank F. Bissell M.J. Petersen O.W. Normal and tumor-derived myoepithelial cells differ in their ability to interact with luminal breast epithelial cells for polarity and basement membrane deposition.J Cell Sci. 2002; 115: 39-50Crossref PubMed Google Scholar and tumors.16Kunz-Schughart L.A. Heyder P. Schroeder J. Knuechel R. A heterologous 3-D coculture model of breast tumor cells and fibroblasts to study tumor-associated fibroblast differentiation.Exp Cell Res. 2001; 266: 74-86Crossref PubMed Scopus (106) Google Scholar, 17Fischbach C. Chen R. Matsumoto T. Schmelzle T. Brugge J.S. Polverini P.J. Mooney D.J. Engineering tumors with 3D scaffolds.Nat Methods. 2007; 4: 855-860Crossref PubMed Scopus (673) Google Scholar, 18Okawa T. Michaylira C.Z. Kalabis J. Stairs D.B. Nakagawa H. Andl C.D. Johnstone C.N. Klein-Szanto A.J. El-Deiry W.S. Cukierman E. Herlyn M. Rustgi A.K. The functional interplay between EGFR overexpression, hTERT activation, and p53 mutation in esophageal epithelial cells with activation of stromal fibroblasts induces tumor development, invasion, and differentiation.Genes Dev. 2007; 21: 2788-2803Crossref PubMed Scopus (148) Google Scholar, 19Takagi A. Watanabe M. Ishii Y. Morita J. Hirokawa Y. Matsuzaki T. Shiraishi T. Three-dimensional cellular spheroid formation provides human prostate tumor cells with tissue-like features.Anticancer Res. 2007; 27: 45-53PubMed Google Scholar, 20Wolf K. Wu Y.I. Liu Y. Geiger J. Tam E. Overall C. Stack M.S. Friedl P. Multi-step pericellular proteolysis controls the transition from individual to collective cancer cell invasion.Nat Cell Biol. 2007; 9: 893-904Crossref PubMed Scopus (758) Google Scholar, 21Pickl M. Ries C.H. Comparison of 3D and 2D tumor models reveals enhanced HER2 activation in 3D associated with an increased response to trastuzumab.Oncogene. 2009; 28: 461-468Crossref PubMed Scopus (305) Google Scholar The 3D culture systems have been used to address the functional interaction between human tumor and stromal cells embedded in the ECM in different cancer models.22Ronnov-Jessen L. Van Deurs B. Nielsen M. Petersen O.W. Identification, paracrine generation, and possible function of human breast carcinoma myofibroblasts in culture.In Vitro Cell Dev Biol. 1992; 28A: 273-283Crossref PubMed Scopus (59) Google Scholar, 23Nielsen H.L. Ronnov-Jessen L. Villadsen R. Petersen O.W. Identification of EPSTI1, a novel gene induced by epithelial-stromal interaction in human breast cancer.Genomics. 2002; 79: 703-710Crossref PubMed Scopus (50) Google Scholar, 24Kunz-Schughart L.A. Knuechel R. Tumor-associated fibroblasts (part I): active stromal participants in tumor development and progression.Histol Histopathol. 2002; 17: 599-621PubMed Google Scholar, 25Schmeichel K.L. Bissell M.J. Modeling tissue-specific signaling and organ function in three dimensions.J Cell Sci. 2003; 116: 2377-2388Crossref PubMed Scopus (471) Google Scholar, 26Gaggioli C. Hooper S. Hidalgo-Carcedo C. Grosse R. Marshall J.F. Harrington K. Sahai E. Fibroblast-led collective invasion of carcinoma cells with differing roles for RhoGTPases in leading and following cells.Nat Cell Biol. 2007; 9: 1392-1400Crossref PubMed Scopus (1081) Google Scholar We have combined well-established 3D cellular assays, namely, multicellular spheroids, 3D collagen gel cultures, and co-cultures, into one experimental setup, which we call the "in vitro carcinoma assay." We have focused on colorectal cancer and shown that this model system is physiologically relevant and allows live imaging, histologic examination, biochemical assays, and functional experiments to be performed in parallel. In addition, expression profiling analysis identified genes differentially regulated on tumor stroma interaction, which are relevant for carcinogenesis. Finally, we demonstrated that the assay format is suitable for in vitro drug testing. LS 174T (CL-188), HCT 116 (CCL-147), Colo205 (CCL-222), MCF7 (HTB-22), SW480 (CCL-227), SW620 (CCL-228) CCD-18Co (CTL1459), Caco-2 (HTB-37), HT-29 (HTB-38), and H446 (HTB-171) were from ATCC. Telomerase immortalized normal human fibroblasts (BJ-1) were purchased from Clontech (Mountain View, CA). Tumor cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (Invitrogen, Gibco, Carlsbad, CA) and penicillin/streptomycin (Invitrogen, Gibco) in conventional plastic culture flasks (Iwaki, Fukushima, Japan). The CAFs were isolated from fresh samples of colon adenocarcinomas received at the Institute of Pathology at the Medical University of Vienna, Vienna, Italy, in accordance with the institutional ethical guidelines. Pieces of approximately 0.5 cm3 were minced in the MediMachine (No. 340588, Becton Dickinson, Franklin Lakes, NJ) and seeded in DMEM/10% fetal calf serum, penicillin/streptomycin, and ciprofloxacin (50 μg/mL) onto 5-cm tissue culture plates (No. 150288, Thermo Fisher Scientific, Rochester, NY). After 3 days the remnants of the tissue were carefully washed away with PBS and the attached fibroblasts were further incubated in serum free fibroblast growth medium (FGM; No. C-23010, PromoCell, Heidelberg, Germany) containing 1 ng/mL of basic fibroblast growth factor and 5 ng/mL of insulin for four population doublings, and aliquots were frozen in CryoSFM (No. C-29912, PromoCell). Immunofluorescence staining of whole gels was adapted from an established protocol.27Janda E. Lehmann K. Killisch I. Jechlinger M. Herzig M. Downward J. Beug H. Grunert S. Ras and TGF[beta] cooperatively regulate epithelial cell plasticity and metastasis: dissection of Ras signaling pathways.J Cell Biol. 2002; 156: 299-313Crossref PubMed Scopus (609) Google Scholar Collagen gels were washed once in cold PBS, fixed in 4% paraformaldehyde/250 mmol/L HEPES for 15 minutes at room temperature incubated in Tris-buffered saline 0.1% Tween 20 (TBS-T) for 10 minutes, followed by three washes in PBS/0.1% Tween 20 (PBS-T). After blocking for 30 minutes in PBS-T/0.2% gelatin/0.1 μg/mL of mouse IgG, the collagen gels were incubated overnight at 4°C in 250 μL of PBS-T/0.2% gelatin (blocking solution) containing the appropriate dilution of the first antibody. Then the gels were washed 3 times in PBS-T, and 250 μL of the second antibody in blocking solution was added and incubated for another 2 hours at room temperature. After washing, the procedure was repeated for staining with a second primary antibody of a different source. Thereafter, the gels were incubated for 1 hour in PBS-T and DAPI (0.2 μg/mL), washed, placed on glass slides with Vectashield mounting medium (H-1000, Vector Laboratories, Burlingame, VT), coverslipped, and analyzed on a Zeiss LSM5-Exciter confocal microscope. Immunohistochemical analysis was performed either on 2-μm sections from FFPE samples or on acetone/methanol fixed 5-μm sections from frozen samples embedded in O.C.T. compound (Tissue Tek, No. 4583, Sakura Finetek Europe, Zoeterwoude, the Netherlands). The Envision+R System PolymerHRP method (No. K4000, DakoCytomation, Carpinteria, CA) was used followed by diaminobenzidine and hematoxylin counterstaining. Cason's trichrome and PAS staining were performed following standard procedures. Live imaging was performed on an AxioVert 200M (Zeiss, Oberkochen, Germany) microscope using AxioVision 4.5 software. For long-time in vivo imaging, we used a Pecon incubation chamber (37°C, 5% carbon dioxide) for 24-well plates equipped with a humidifier system. Spheroids and fibroblasts containing collagen gels were lysed in toto in guanidinium isothiocyanate RNA lysis buffer. Total RNA was extracted using the Arcturus Pico Pure RNA Isolation Kit (Arcturus Engineering, Mountain View, CA) following the manufacturer's instructions, including DNase I treatment. RNA quality was assessed with the Agilent Nano LabChip Assay (Agilent, Palo Alto, CA). The obtained RNA was linear amplified and biotinylated with the MessageAmp Premier RNA Amplification Kit (Ambion, Austin, TX) and used for hybridization to the Human Genome U133 Plus 2.0 GeneChip arrays using standard Affymetrix protocols. Microarray data were normalized with the Robust Multi-Array Analysis as implemented in Bioconductor.28Irizarry R.A. Bolstad B.M. Collin F. Cope L.M. Hobbs B. Speed T.P. Summaries of Affymetrix GeneChip probe level data.Nucleic Acids Res. 2003; 31: e15Crossref PubMed Scopus (4014) Google Scholar, 29Gentleman R.C. Carey V.J. Bates D.M. Bolstad B. Dettling M. Dudoit S. Ellis B. Gautier L. Ge Y. Gentry J. Hornik K. Hothorn T. Huber W. Iacus S. Irizarry R. Leisch F. Li C. Maechler M. Rossini A.J. Sawitzki G. Smith C. Smyth G. Tierney L. Yang J.Y. Zhang J. Bioconductor: open software development for computational biology and bioinformatics.Genome Biol. 2004; 5: R80Crossref PubMed Google Scholar All analyses were performed with log2-transformed data. Hypothesis tests were performed using a modified t statistic with an empirical Bayes approach as implemented in the Bioconductor LIMMA package.30Smyth G.K. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments.Stat Appl Genet Mol Biol. 2004; 3 (Article3)PubMed Google Scholar P values were adjusted by the false discovery rate (FDR) method of Benjamini and Hochberg.31Benjamini Y. Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing.J R Stat Soc Ser B. 2005; 57: 289-300Google Scholar Gene Set Enrichment Analysis (GSEA) is a computational method that determines whether an a priori defined set of genes shows statistically significant, concordant differences between two groups of samples. We used two gene set collections from the Molecular Signature Database provided by the Broad Institute: c2, the curated gene set, and c5, the gene ontology (GO) gene set. The core genes of statistically significant genes sets were used to calculate principal component analysis (PCA) plots. Pathway analysis based on literature information extracted by experts (IPA; Ingenuity Systems, Inc., Redwood City, CA) was performed to corroborate GSEA results and to generate gene networks of significantly regulated groups of genes. Collagen gel inserts were washed in PBS and incubated with 200 μL of prewarmed collagenase B (25 mg/mL, Hoffmann-La Roche, Nutley, NJ) at 37°C for 1 minute, 5 mL of ice cold PBS/5% fetal calf serum was added, the released spheroids and fibroblasts were harvested by centrifugation and washed in PBS, and the cell pellets were lysed in Laemmli buffer. Rat tail collagen I was from BD (Franklin Lakes, NJ), antibodies against phospho-H3, cleaved caspase 3, phosphorylated retinoblastoma protein (pRb), platelet-derived growth factor receptor β (PDGFRβ) were from Cell Signaling Technology (Danvers, MA), anti–E-cadherin (clone HECD1) was purchased from Abcam plc (Oxford, UK), and cytokeratin 18 and vimentin (clone V9) antibodies were from Sigma-Aldrich (St. Louis, MO). Fibroblast activation protein α antibody was clone F19.32Garin-Chesa P. Old L.J. Rettig W.J. Cell surface glycoprotein of reactive stromal fibroblasts as a potential antibody target in human epithelial cancers.Proc Natl Acad Sci U S A. 1990; 87: 7235-7239Crossref PubMed Scopus (490) Google Scholar Samples were fixed in 2% Karnovsky's paraformaldehyde. After washing and postfixing in 1% OsO4 (resolved in 3% potassium hexacyanoferrate), the material was dehydrated and embedded in EPON Resin 812. Ultrathin sections (approximately 70 nm) were prepared and stained with uranylacetate and lead citrate. Sections were examined under a Jeol JEM 1010 electron microscope. Determination of 120 cytokine levels in cell culture supernatants was performed with the RayBio Cytokine Antibody Arrays as recommended by the supplier (RayBiotech, Norcross, GA). The phosphorylation status of 42 receptor tyrosine kinases was determined with the human Phospho-Receptor Tyrosine Kinase Array Kit according to the instructions (R&D Systems Inc, Minneapolis, MN). Multicellular spheroids were generated essentially as published for endothelial cell spheroids.33Korff T. Augustin H.G. Integration of endothelial cells in multicellular spheroids prevents apoptosis and induces differentiation.J Cell Biol. 1998; 143: 1341-1352Crossref PubMed Scopus (471) Google Scholar In brief, tumor cells were grown on plastic, trypsinized, counted, and resuspended in DMEM/5% fetal calf serum containing 12 mg/mL of methylcellulose (Sigma-Aldrich, St. Louis, MO). The cells were seeded into nonadhesive, U-shaped, 96-well plates for suspension culture (Greiner bio-one, Kremsmuenster, Austria) at a concentration of 150 to 200 cells per well. Compact multicellular spheroids were harvested after 2 to 3 days and washed, and 96 tumor cell spheres each were transferred into 1.5 mL of microcentrifuge tubes. Fibroblasts for co-culture experiments were cultivated on plastic in serum free FGM (Promocell, Heidelberg, Germany). After trypsinization, cells were counted; 2 × 105 cells were transferred into 1.5-mL tubes and centrifuged at 250 × g. In case of co-culture experiments, the cells were centrifuged onto the previously prepared spheroid pellets. The entire supernatant was carefully removed, and the cells/cell-spheroid pellets were kept on ice until further use. Gel casting devices were produced by cutting out 2 × 2-cm squares of a silicone foil (Gel dryer sealing gasket, 1-mm thickness, Biorad, Hercules, CA) into which holes of 1.4-cm diameter were cut with a puncher. Nylon filters were prepared from the nylon mesh inserts of Medicon syringe filters (100 μm, Becton Dickinson), of which a hole of 1-cm diameter was cut out in the center. The silicone casting devices and the nylon meshes were autoclaved and placed onto the inside surface of lids of 10-cm cell culture dishes, where they firmly attached. For the collagen gel preparation, all steps were performed on ice. Collagen solutions were prepared by mixing 0.2 mL of 10× PBS, 0.8 mL of FGM/20% methylcellulose, and 1 mL of collagen I (rat, 3.48 mg/mL, Becton Dickinson) and neutralized with 23 μL of 1 mol/L NaOH. The solution was mixed carefully; air bubbles were removed by centrifugation at 300 × g for 2 minutes. A total of 200 μL each of the collagen solution was transferred into the tubes containing the spheroid/cell pellets, which were gently resuspended. The spheroid/cell suspensions were transferred into the casting devices; a nylon filter was added and submerged. After polymerization of the collagen solution for 30 minutes at 37°C/5% CO2, the silicone foil was removed, leaving collagen gel cylinders. The gels were transferred into 24-well plates containing 1 mL of FGM supplemented with 2.5% serum. The culture medium was replaced every second day, and conditioned medium was kept, sterile filtered, and snap-frozen in liquid nitrogen. Celltracker Green CMFDA (Invitrogen) was used for long-term tracing of fibroblasts for live cell imaging as recommended by the supplier. The growth rate of the tumor spheroids was determined by calculating the mean volume of 8 to 10 spheroids, using the formula (length × width2 × π/6) as used to determine the size of subcutaneous xenograft tumors in mice.34Tomayko M.M. Reynolds C.P. Determination of subcutaneous tumor size in athymic (nude) mice.Cancer Chemother Pharmacol. 1989; 24: 148-154Crossref PubMed Scopus (1334) Google Scholar, 35Euhus D.M. Hudd C. LaRegina M.C. Johnson F.E. Tumor measurement in the nude mouse.J Surg Oncol. 1986; 31: 229-234Crossref PubMed Scopus (477) Google Scholar The TissueQuest software (TissueGnostics GmbH, Vienna, Austria) was used to quantify fluorescent photomicrographs. Cell detection is based on identification of nuclei by DAPI staining. Fluorescence intensity quantification was performed independently in different channels. All photomicrographs were taken with the same settings (exposure time, signal amplification, and objectives). Seventeen human tumor cell lines from different cancer types were tested for their ability to grow as multicellular spheroids. Eleven cell lines formed spheroids, whereas six cell lines only formed loose cell aggregates under any of the experimental conditions tested. Spheroids from different cell lines behaved differently when placed into collagen gels. Some remained as compact spheres and did not show signs of invasion; others displayed moderate invasion, and other cell types, such as SK-OV-3, showed extensive invasive structures characterized by multicellular astral outgrowth into the collagen gels (Figure 1A; see also Supplemental Table S1 and Supplemental Figure S1 at http://ajp.amjpathol.org). The colon cancer cell line LS174T, defined as noninvasive in our assay, was selected for further studies. LS174T spheroids were prepared, harvested, and mixed with freshly trypsinized fibroblasts. The mixture was centrifuged and the pellet resuspended in collagen I solution and poured into silicone molds (Figure 1B). This gave rise to flat collagen gel cylinders, which were mechanically stabilized by submerging nylon mesh rings before polymerization (Figure 1B, bottom). The gel cylinders harbored either 96 tumor cell spheroids surrounded by 2 × 105 normal fibroblasts or CAFs. In addition, tumor cell spheroids without fibroblasts or fibroblasts alone were prepared in a similar way. The fibroblasts used were normal embryonic colon fibroblasts (NCFs; CCD18-Co), hTERT immortalized skin fibroblasts (SFs; BJ-1), and primary fibroblast cultures isolated from human colon carcinoma specimen (CAF1 and CAF2). For phenotypic characterization, the SFs and CAFs were analyzed for marker expression by immunofluorescence. The cells were positive for the mesenchymal marker vimentin, whereas the epithelial marker cytokeratin 8 was not detectable. In addition, approximately 10% of the cells were α-smooth muscle actin positive (see Supplemental Figure S2A at http://ajp.amjpathol.org). A detailed characterization of these various fibroblast types was performed by gene expression analysis. In addition, the proliferation rate of CAFs, NCFs, and SFs grown on 2D was determined and no differences were found among the various cell types (see Supplemental Figure S2B at http://ajp.amjpathol.org). The ring-shaped nylon meshes facilitated the microscopic analysis by phase contrast of living cells (Figures 1C and 2A) and by fluorescence microscopy of labeled cells. When tumor spheroids and fibroblasts were cultured without the supporting nylon mesh, the free-floating collagen gel cylinders shrunk and appeared as dense, refractive structures within 2 to 3 days due to the contractile forces of the fibroblasts in the gels as described,36Elsdale T. Bard J. Collagen substrata for studies on cell behavior.J Cell Biol. 1972; 54: 626-637Crossref PubMed Scopus (1247) Google Scholar which hindered microscopic evaluation (Figure 1C). In contrast, embedded nylon meshes resisted the contracting force of the fibroblasts and completely abolished shrinking (Figure 1C). The morphology of the co-cultures could thus be monitored easily by live microscopy. In the presence of fibroblasts, LS174T tumor spheroids (Figure 2A) displayed well-organized glandular structures during the early phases (24 to 48 hours) of the co-culture as revealed H&E staining (Figure 2B). The epithelial cells expressed the luminal tight junction protein ZO-1, predominantly membrane-associated β-catenin, and secreted mucin into the glandular structures (Figure 2B). Electron microscopy revealed the characteristic intestinal cell microvillus seam at the luminal side on LS174T cells when grown alone or in the LS174T-NCF co-cultures. Differences in the electron density of the mucus were observed when comparing LS174T spheroids cultured alone versus the co-cultures, most likely reflecting the presence of different type of mucins (Figure 2C). Differences were also noted in the cell-cell interactions, with well-established desmosomes, adherens junctions, and tight junctions observed in the LS174T cells co-cultured with fibroblasts, which were less developed in LS174T cells grown alone (Figure 2D). Fibroblasts, when added to the cultures, formed a well-organized network around the tumor cell spheroids within 2 to 3 days (Figure 3A, left), regardless of the source of fibroblasts. The fibroblasts expressed PDGFRβ, and some were closely attached to the tumor spheroid surface (Figure 3A). Electron microscopy revealed close cell-cell interactions between epithelial cells and fibroblasts in tumor spheroid/fibroblast co-cultures (Figure 3B). Fibroblasts in the collagen gels, like in tissues, displayed spindle-shaped morphologic features (Figure 3, A and C), which did not change on co-cultivation with tumor cell spheroids. They expressed the fibroblast markers fibroblast activation protein α, vimentin, Thy1, PDGFRα and β, and α-smooth muscle actin and were negative for endothelial, pericyte, and epithelial markers, closely resembling activated fibroblasts in tumor stroma in vivo (Figure 3C; see also Supplemental Figure S3 at http://ajp.amjpathol.org). In summary, the in vitro cultures closely mimic the cellular architecture of human colon carcinoma samples at the histologic level (Figure 3C) and the phenotype of activated tumor fibroblasts.32Garin-Chesa P. Old L.J. Rettig W.J. Cell surface glycoprotein of reactive stromal fibroblasts as a potential antibody target in human epithelial cancers.Proc Natl Acad Sci U S A. 1990; 87: 7235-7239Crossref PubMed Scopus (490) Google Scholar Co-cultivation experiments of other human cancer cells, such as colon cancer (HCT116, HT-29) and mammary tumor cell spheroids (MCF7, BT474) with CAFs, displayed similar phenotypes with respect to their in vivo counterparts (data not shown), demonstrating the general
Referência(s)