Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Molecular Basis of Constitutive Production of Basement Membrane Components

2003; Elsevier BV; Volume: 278; Issue: 50 Linguagem: Inglês

10.1074/jbc.m304985200

ISSN

1083-351X

Autores

Sugiko Futaki, Yoshitaka Hayashi, Megumi Yamashita, Ken Yagi, Hidemasa Bono, Yoshihide Hayashizaki, Yasushi Okazaki, Kiyotoshi Sekiguchi,

Tópico(s)

Chromatin Remodeling and Cancer

Resumo

Engelbreth-Holm-Swarm (EHS) tumors produce large amounts of basement membrane (BM) components that are widely used as cell culture substrates mimicking BM functions. To delineate the tissue/organ origin of the tumor and the mechanisms operating in the BM overproduction, a genome-wide expression profile of EHS tumor was analyzed using RIKEN cDNA microarrays containing ∼40,000 mouse cDNA clones. Expression profiles of F9 embryonal carcinoma cells that produce laminin-1 and other BM components upon differentiation into parietal endoderm-like cells (designated F9-PE) were also analyzed. Hierarchical clustering analysis showed that the gene expression profiles of EHS and F9-PE were the most similar among 49 mouse tissues/organs in the RIKEN Expression Array Database, suggesting that EHS tumor is parietal endoderm-derived. Quantitative PCR analysis confirmed that not only BM components but also the machineries required for efficient production of BM components, such as enzymes involved in post-translational modification and molecular chaperones, were highly expressed in both EHS and F9-PE. Pairs of similar transcription factor isoforms, such as Gata4/Gata6, Sox7/Sox17, and Cited1/Cited2, were also highly expressed in both EHS tumor and F9-PE. Time course analysis of F9 differentiation showed that up-regulation of the transcription factors was associated with those of BM components, suggesting their involvement in parietal endoderm specification and overproduction of the BM components. Engelbreth-Holm-Swarm (EHS) tumors produce large amounts of basement membrane (BM) components that are widely used as cell culture substrates mimicking BM functions. To delineate the tissue/organ origin of the tumor and the mechanisms operating in the BM overproduction, a genome-wide expression profile of EHS tumor was analyzed using RIKEN cDNA microarrays containing ∼40,000 mouse cDNA clones. Expression profiles of F9 embryonal carcinoma cells that produce laminin-1 and other BM components upon differentiation into parietal endoderm-like cells (designated F9-PE) were also analyzed. Hierarchical clustering analysis showed that the gene expression profiles of EHS and F9-PE were the most similar among 49 mouse tissues/organs in the RIKEN Expression Array Database, suggesting that EHS tumor is parietal endoderm-derived. Quantitative PCR analysis confirmed that not only BM components but also the machineries required for efficient production of BM components, such as enzymes involved in post-translational modification and molecular chaperones, were highly expressed in both EHS and F9-PE. Pairs of similar transcription factor isoforms, such as Gata4/Gata6, Sox7/Sox17, and Cited1/Cited2, were also highly expressed in both EHS tumor and F9-PE. Time course analysis of F9 differentiation showed that up-regulation of the transcription factors was associated with those of BM components, suggesting their involvement in parietal endoderm specification and overproduction of the BM components. Basement membranes (BMs) 1The abbreviations used are: BMbasement membraneECMextracellular matrixEHS tumorEngelbreth-Holm-Swarm tumorF9-PEF9 differentiated into parietal endoderm-like cellsF9-Sundifferentiated F9 cellsRAall-trans-retinoic acidBt2cAMPdibutyryl cAMPCREBcAMP-response element-binding proteinE17.5whole mouse embryo at 17.5 days of gestation.1The abbreviations used are: BMbasement membraneECMextracellular matrixEHS tumorEngelbreth-Holm-Swarm tumorF9-PEF9 differentiated into parietal endoderm-like cellsF9-Sundifferentiated F9 cellsRAall-trans-retinoic acidBt2cAMPdibutyryl cAMPCREBcAMP-response element-binding proteinE17.5whole mouse embryo at 17.5 days of gestation. are thin sheets of extracellular matrix (ECM) underlying the basal side of epithelial/parenchymal cells and consist of laminins, collagen IV, nidogens, and perlecan, all of which are specifically present in BMs but not in other types of ECMs. Through interactions with cell surface receptors such as integrins, BMs not only provide a structural basis for epithelial cells but also regulate their proliferation, migration, differentiation, and survival. Many growth factors (e.g. fibroblast growth factors, transforming growth factor-β, and hepatocyte growth factor) are also incorporated into the BMs through binding to various ECM molecules, further modifying the biological functions of BMs (1.Erickson A.C. Couchman J.R. J. Histochem. Cytochem. 2000; 48: 1291-1306Crossref PubMed Scopus (241) Google Scholar, 2.Hagedorn H.G. Bachmeier B.E. Nerlich A.G. Int. J. Oncol. 2001; 18: 669-681PubMed Google Scholar). Despite their biological importance, biochemical and cell biological studies of BMs have been hampered by difficulties in preparing BM components on a large scale, since BMs are constantly maintained as very thin sheets in most tissues and organs. An exceptional source of BM components is the murine Engelbreth-Holm-Swarm (EHS) tumor, which produces extraordinary amounts of ECM that are readily extracted under nondenaturing conditions (3.Kleinman H.K. McGarvey M.L. Liotta L.A. Robey P.G. Tryggvason K. Martin G.R. Biochemistry. 1982; 21: 6188-6193Crossref PubMed Scopus (953) Google Scholar, 4.Kleinman H.K. McGarvey M.L. Hassell J.R. Star V.L. Cannon F.B. Laurie G.W. Martin G.R. Biochemistry. 1986; 25: 312-318Crossref PubMed Scopus (1189) Google Scholar). A crude extract prepared from EHS tumor contains nearly 1% (w/v) laminins and reconstitutes BM-like gels in vitro, which have been widely used as two- or three-dimensional cell culture substrates that support the differentiated functions of various cell types (5.Lang S.H. Stark M. Collins A. Paul A.B. Stower M.J. Maitland N.J. Cell Growth Differ. 2001; 12: 631-640PubMed Google Scholar, 6.Levenberg S. Golub J.S. Amit M. Itskovitz-Eldor J. Langer R. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4391-4396Crossref PubMed Scopus (774) Google Scholar, 7.Kragh M. Hjarnaa P.J. Bramm E. Kristjansen P.E. Rygaard J. Binderup L. Int. J. Oncol. 2003; 22: 305-311PubMed Google Scholar, 8.Xu C. Inokuma M.S. Denham J. Golds K. Kundu P. Gold J.D. Carpenter M.K. Nat. Biotechnol. 2001; 19: 971-974Crossref PubMed Scopus (1567) Google Scholar). basement membrane extracellular matrix Engelbreth-Holm-Swarm tumor F9 differentiated into parietal endoderm-like cells undifferentiated F9 cells all-trans-retinoic acid dibutyryl cAMP cAMP-response element-binding protein whole mouse embryo at 17.5 days of gestation. basement membrane extracellular matrix Engelbreth-Holm-Swarm tumor F9 differentiated into parietal endoderm-like cells undifferentiated F9 cells all-trans-retinoic acid dibutyryl cAMP cAMP-response element-binding protein whole mouse embryo at 17.5 days of gestation. The molecular mechanisms operating in the overproduction of BM components in EHS tumor are poorly understood. Moreover, the tissue/organ origin of the tumor is obscure. EHS tumor spontaneously arose in an ST/Eh mouse strain and was initially designated as a chondrosarcoma based on its histological appearance (9.Swarm R.L. J. Natl. Cancer Inst. 1963; 31: 953-974PubMed Google Scholar). Later, biochemical analysis demonstrated that it did not produce cartilaginous proteins but secreted the major BM components (i.e. laminin-1, nidogen-1, collagen IV, and perlecan) (10.Orkin R.W. Gehron P. McGoodwin E.B. Martin G.R. Valentine T. Swarm R. J. Exp. Med. 1977; 145: 204-220Crossref PubMed Scopus (434) Google Scholar). EHS tumor is still often called a “sarcoma,” and it is also referred to as a yolk sac tumor (3.Kleinman H.K. McGarvey M.L. Liotta L.A. Robey P.G. Tryggvason K. Martin G.R. Biochemistry. 1982; 21: 6188-6193Crossref PubMed Scopus (953) Google Scholar, 11.Baldwin C.T. Silbert J.E. Humphries D.E. Cogburn J.N. Smith B.D. Matrix. 1989; 9: 389-396Crossref PubMed Scopus (5) Google Scholar, 12.Fowler K.J. Mitrangas K. Dziadek M. Exp. Cell Res. 1990; 191: 194-203Crossref PubMed Scopus (21) Google Scholar, 13.Wewer U.M. Albrechtsen R. Hassell J.R. Differentiation. 1985; 30: 61-67Crossref PubMed Scopus (13) Google Scholar). The primary aims of the present study were to elucidate the origin of EHS tumor and the molecular basis of the BM overproduction. To this end, we utilized genome-wide gene expression profiling using RIKEN mouse 20K chip-1 and -2 containing ∼40,000 cDNA clones corresponding to 24,000 nonredundant genes and expressed sequence tags (14.Miki R. Kadota K. Bono H. Mizuno Y. Tomaru Y. Carninci P. Itoh M. Shibata K. Kawai J. Konno H. Watanabe S. Sato K. Tokusumi Y. Kikuchi N. Ishii Y. Hamaguchi Y. Nishizuka I.I. Goto H. Nitanda H. Satomi S. Yoshiki A. Kusakabe M. DeRisi J.L. Eisen M.B. Iyer V.R. Brown P.O. Muramatsu M. Shimada H. Okazaki Y. Hayashizaki Y. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2199-2204Crossref PubMed Scopus (191) Google Scholar, 15.Bono H. Yagi K. Kasukawa T. Nikaido I. Tominaga N. Miki R. Mizuno Y. Tomaru Y. Goto H. Nitanda H. Shimizu D. Makino H. Morita T. Fujiyama J. Sakai T. Shimoji T. Hume D.A. Hayashizaki Y. Okazaki Y. Genome Res. 2003; 13: 1318-1323Crossref PubMed Scopus (71) Google Scholar). To further elucidate the regulatory mechanisms of BM production, we also analyzed the gene expression profiles of murine F9 embryocarcinoma cells that differentiate into parietal endoderm-like cells (designated F9-PE) and produce large amounts of BM components upon treatment with all-trans-retinoic acid (RA) and dibutyryl cAMP (Bt2cAMP) (16.Strickland S. Smith K.K. Marotti K.R. Cell. 1980; 21: 347-355Abstract Full Text PDF PubMed Scopus (591) Google Scholar, 17.Verheijen M.H. Defize L.H. Int. J. Dev. Biol. 1999; 43: 711-721PubMed Google Scholar). Utilizing these approaches, we characterized EHS tumor as a parietal yolk sac-derived tumor. In both EHS tumor and F9-PE, not only secretory/ECM molecules, but also enzymes and chaperones involved in the post-translational modification of ECM molecules, were highly expressed, suggesting that parietal endoderm cells are an optimized “factory” producing the BM. Engelbreth-Holm-Swarm Tumor—EHS tumor was maintained by intramuscular implantation in the hind limbs of C57BL/6J mice in the animal experiment facility of Aichi Medical University under approval of the Animal Care Committee of Aichi Medical University. Solid tumors that developed 3-4 weeks after transplantation were excised and used for RNA preparation. Cell Culture—Murine F9 embryonic carcinoma cells were obtained from the Health Science Research Resource Bank (Osaka, Japan; available on the World Wide Web at www.jhsf.or.jp/English/index_e.html) and cultured on gelatin-coated culture dishes (Asahi Techno Glass Corp, Chiba, Japan) in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37 °C under an atmosphere of 95% air, 5% CO2, and 100% humidity. For induction of differentiation into parietal endoderm-like cells, F9 cells were treated with 1 μm RA and 1 mm Bt2cAMP for 96 h (16.Strickland S. Smith K.K. Marotti K.R. Cell. 1980; 21: 347-355Abstract Full Text PDF PubMed Scopus (591) Google Scholar). RNA Extraction—Total RNAs were extracted from EHS tumor and F9 cells using RNeasy kits (Qiagen, Valencia, CA) according to the manufacturer's instructions. Poly(A)+ RNAs were purified using Oligotex dT-30 (TAKARA BIO Inc., Otsu, Shiga, Japan). RNAs from whole mouse embryos at 17.5 days of gestation (designated E17.5) were prepared as described (14.Miki R. Kadota K. Bono H. Mizuno Y. Tomaru Y. Carninci P. Itoh M. Shibata K. Kawai J. Konno H. Watanabe S. Sato K. Tokusumi Y. Kikuchi N. Ishii Y. Hamaguchi Y. Nishizuka I.I. Goto H. Nitanda H. Satomi S. Yoshiki A. Kusakabe M. DeRisi J.L. Eisen M.B. Iyer V.R. Brown P.O. Muramatsu M. Shimada H. Okazaki Y. Hayashizaki Y. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2199-2204Crossref PubMed Scopus (191) Google Scholar). Microarray Experiments—The RIKEN full-length enriched cDNA microarrays are composed of two chips, 20K-1 and 20K-2, with each chip containing 19,584 cDNA spots including 288 positive and 1,296 negative control spots. For positive controls, cDNA clones for glyceraldehyde-3-phosphate dehydrogenase and β-actin were spotted, whereas plant cDNAs, mouse Cot-1 DNA, salmon sperm DNA, and oligo(dA) were used as negative controls (14.Miki R. Kadota K. Bono H. Mizuno Y. Tomaru Y. Carninci P. Itoh M. Shibata K. Kawai J. Konno H. Watanabe S. Sato K. Tokusumi Y. Kikuchi N. Ishii Y. Hamaguchi Y. Nishizuka I.I. Goto H. Nitanda H. Satomi S. Yoshiki A. Kusakabe M. DeRisi J.L. Eisen M.B. Iyer V.R. Brown P.O. Muramatsu M. Shimada H. Okazaki Y. Hayashizaki Y. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2199-2204Crossref PubMed Scopus (191) Google Scholar, 15.Bono H. Yagi K. Kasukawa T. Nikaido I. Tominaga N. Miki R. Mizuno Y. Tomaru Y. Goto H. Nitanda H. Shimizu D. Makino H. Morita T. Fujiyama J. Sakai T. Shimoji T. Hume D.A. Hayashizaki Y. Okazaki Y. Genome Res. 2003; 13: 1318-1323Crossref PubMed Scopus (71) Google Scholar). Hybridization probes were prepared as described by Miki et al. (14.Miki R. Kadota K. Bono H. Mizuno Y. Tomaru Y. Carninci P. Itoh M. Shibata K. Kawai J. Konno H. Watanabe S. Sato K. Tokusumi Y. Kikuchi N. Ishii Y. Hamaguchi Y. Nishizuka I.I. Goto H. Nitanda H. Satomi S. Yoshiki A. Kusakabe M. DeRisi J.L. Eisen M.B. Iyer V.R. Brown P.O. Muramatsu M. Shimada H. Okazaki Y. Hayashizaki Y. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2199-2204Crossref PubMed Scopus (191) Google Scholar) with modifications. Briefly, poly(A)+ RNAs were reverse-transcribed in the presence of amino-allyl dUTP (Sigma). After purification of the first strand cDNA, incorporated amino-allyl substrates were coupled with a Cy-3 or Cy-5 monofunctional reactive dye (Amersham Biosciences) in 0.1 m sodium bicarbonate buffer (pH 9.0) by incubating for 1 h at room temperature in the dark. Residual free dyes were removed using MicroSpin S-200 HR columns (Amersham Biosciences). Probes labeled with Cy-3 or Cy-5 were combined in 30 μl of hybridization buffer per slide and competitively hybridized overnight. After hybridization, the slides were washed, dried, and subjected to fluorescence scanning. Hybridizations were performed in duplicate for each combination of samples and references. In duplicate experiments, probes were independently prepared from the same pool of poly(A)+ RNAs. Data Analysis—Following the hybridization, fluorescent images were scanned and analyzed using the GenePix 4000B microarray scanner and GenePixPro 3.0 software (Axon Instruments, Union City, CA). Spots with abnormal appearance or with signal intensities lower than the local background were “flagged” and invalidated in the analysis that followed. Expression ratios were calculated by dividing the Cy-5 intensity by the Cy-3 intensity and normalized as the median of all validated ratios set to 1.0 by using GeneSpring 4.0 (SiliconGenetics, Redwood City, CA). Spots that showed more than a 2-fold discrepancy in the Cy-3/Cy-5 ratios in duplicate experiments were also eliminated. Normalized ratios from duplicate experiments were averaged and used for scatter plots and gene extraction based on the expression ratios. For hierarchical clustering, the ratios were preprocessed by a filtering program, PRIM (18.Kadota K. Miki R. Bono H. Shimizu K. Okazaki Y. Hayashizaki Y. Physiol. Genom. 2001; 4: 183-188Crossref PubMed Scopus (54) Google Scholar), and analyzed using the clustering software CLUSTER and TREEVIEW developed by Eisen et al. (19.Eisen M.B. Spellman P.T. Brown P.O. Botstein D. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 14863-14868Crossref PubMed Scopus (13079) Google Scholar) (available on the World Wide Web at rana.lbl.gov/) as described (14.Miki R. Kadota K. Bono H. Mizuno Y. Tomaru Y. Carninci P. Itoh M. Shibata K. Kawai J. Konno H. Watanabe S. Sato K. Tokusumi Y. Kikuchi N. Ishii Y. Hamaguchi Y. Nishizuka I.I. Goto H. Nitanda H. Satomi S. Yoshiki A. Kusakabe M. DeRisi J.L. Eisen M.B. Iyer V.R. Brown P.O. Muramatsu M. Shimada H. Okazaki Y. Hayashizaki Y. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2199-2204Crossref PubMed Scopus (191) Google Scholar, 15.Bono H. Yagi K. Kasukawa T. Nikaido I. Tominaga N. Miki R. Mizuno Y. Tomaru Y. Goto H. Nitanda H. Shimizu D. Makino H. Morita T. Fujiyama J. Sakai T. Shimoji T. Hume D.A. Hayashizaki Y. Okazaki Y. Genome Res. 2003; 13: 1318-1323Crossref PubMed Scopus (71) Google Scholar). Gene annotations of cDNA spots were according to the data base of functional annotation of RIKEN mouse cDNA clones (FANTOM DB) (20.Bono H. Kasukawa T. Furuno M. Hayashizaki Y. Okazaki Y. Nucleic Acids Res. 2002; 30: 116-118Crossref PubMed Scopus (49) Google Scholar, 21.The RIKEN Genome Exploration Research Group Phase II team and The FANTOM Consortium Nature. 2001; 409: 685-690Crossref PubMed Scopus (567) Google Scholar, 22.Consortium The FANTOM the RIKEN Genome Exploration Research Group Phase I and II Team Nature. 2002; 420: 563-573Crossref PubMed Scopus (1362) Google Scholar). Northern Blotting, Quantitative PCR, and Criteria for Second Round Selection of Genes Highly Expressed in Parietal Endoderm Cells—For quantitatively stricter estimations of the gene expression levels, aliquots of the RNA were subjected to Northern blotting analyses using a previously described method (23.Futaki S. Takagishi Y. Hayashi Y. Ohmori S. Kanou Y. Inouye M. Oda S. Seo H. Iwaikawa Y. Murata Y. Mamm. Genome. 2000; 11: 649-655Crossref PubMed Scopus (20) Google Scholar). Information on the probes is available on request. Alternatively, gene expression levels were estimated by quantitative PCR. Total RNAs were reverse-transcribed by SuperScriptII (Invitrogen) with random primers. The reverse transcripts were used as templates for analysis of the gene expression levels using SmartCycler (Cepheid, Sunnyvale, CA) and a QuantiTect SYBR Green PCR kit (Qiagen) according to the manufacturer's instructions. In general, PCR primers were designed to amplify 200-300 base pairs of the target sequences. Sequences of the primers are available on request. The kinetics of amplification were monitored by SYBR Green fluorescence intensity, and the numbers of cycles required for the fluorescence level to reach the defined threshold level in the logarithmic increase phase (threshold cycle: Ct) were calculated as exponents of the relative amounts of target cDNA in the templates. The differences in the expression levels between F9-PE and undifferentiated F9 (F9-S) were expressed as ΔCt(F9). Similarly, the expression level in EHS relative to F9-S was expressed as ΔCt(EHS). Typically, a ΔCt of 4 corresponded to a 10-fold higher expression than F9-S. For second round extraction of genes highly expressed in both F9-PE and EHS tumor (Table I), a ΔCt of 2 was used as the cut-off point.Table ITranscripts expressed in F9-PE and EHS tumorGenBank™ accession no.Gene ProductSymbolNormalized ratios by microarray (average ± range of duplicate experiment)Difference of PCR cycles to F9-S by quantitative PCREHS/E17.5F9PE/E17.5EHS/F9SF9PE/F9SΔCt(EHS)ΔCt(F9)Extracellular matrix NM_008480Laminin, α1Lama118.1 ± 1.711.4 ± 1.551 ± 2.245.2 ± 1.23.3710.22 XM_126863Laminin, β1Lamb125.7 ± 2.38.6 ± 0.245.7 ± 3.825.3 ± 0.87.868.2026.3 ± 0.77.7 ± 0.754 ± 5.429.5 ± 4.4 J02930Laminin, γ1Lamc125.3 ± 1.89.7 ± 0.612.0 ± 3.110.2 ± 0.46.425.70 XM_134042Procollagen, type IV, α1Col4a14.1 ± 0.21.7 ± 0.38.8 ± 0.68.5 ± 0.511.9911.5115.2 ± 2.45.9 ± 0.736.2 ± 11.154.2 ± 4.87.6 ± 0.13.1 ± 0.4178.7 ± 43.062.6 ± 3.8 XM_134014Procollagen, type IV, α2Col4a25.6 ± 1.24.0 ± 0.4104.7 ± 5.542.3 ± 1.611.5511.369.4 ± 0.03.8 ± 0.2129.0 ± 0.6177.0 ± 18.18.0 ± 0.04.6 ± 0.1157.5 ± 42.586.2 ± 10.56.3 ± 0.73.3 ± 0.147.8 ± 2.968.1 ± 3.7 NM_010917Nidogen 1Nid117.1 ± 0.11.2 ± 0.086.8 ± 9.810.4 ± 0.16.644.4718.0 ± 0.31.6 ± 0.072.5 ± 11.98.8 ± 0.1 NM_011157SerglycinPrg10.4 ± 1.119.1 ± 3.128.2 ± 1.4190.9 ± 45.17.1812.178.9 ± 0.519.4 ± 2.720.6 ± 7.965.2 ± 14.38.9 ± 1.529.1 ± 4.640.9 ± 4.080.9 ± 22.68.4 ± 0.114.0 ± 1.97.8 ± 4.091.0 ± 9.7 NM_009242Secreted acidic cysteine-rich glycoproteinSparc4.4 ± 0.51.1 ± 0.227.4 ± 5.734.9 ± 0.16.675.533.9 ± 0.21.2 ± 0.229.0 ± 4.661.5 ± 4.7 NM_009262Sparc/osteonectin, cwcy, and kazal-like domains proteoglycan 1Spock 19.3 ± 2.22.5 ± 0.025.2 ± 10.26.5 ± 0.311.0610.266.4 ± 0.42.2 ± 0.17.2 ± 3.13.2 ± 0.1Post-translational modification, protein processing NM_011031Procollagen-proline, 2-oxoglutarate 4-dioxygenase (proline 4-hydroxylase), alpha II polypeptideP4ha210.7 ± 1.27.2 ± 1.616.9 ± 4.117.2 ± 2.55.215.4810.1 ± 2.75.5 ± 0.513.8 ± 3.413.8 ± 1.310.8 ± 0.54.3 ± 0.226.9 ± 4.017.3 ± 0.15.3 ± 0.73.4 ± 0.26.5 ± 0.84.9 ± 0.2 NM_011961Procollagen lysine, 2-oxoglutarate 5-dioxygenase 2Plod210.4 ± 1.65.1 ± 0.514.2 ± 2.58.8 ± 0.82.774.86 NM_011962Procollagen lysine, 2-oxoglutarate 5-dioxygenase 3Plod35.5 ± 0.82.9 ± 0.23.4 ± 0.35.4aNo replicates2.062.19 NM_009825Serine (or cysteine) proteinase inhibitor, clade H (heat shock protein 47), member 1HSP473.1 ± 0.51.6 ± 0.37.7 ± 2.25.3 ± 1.72.523.71 NM_134090Similar to KDEL (LYS-ASP-GLU-LEU) endoplasmic reticulum protein retention receptor 3(KDEL) Mm.296445.7 ± 0.71.4 ± 0.215.8 ± 0.925.5 ± 3.96.367.86 NM_010474Heparan sulfate (glucosamine) 3-O-sulfotransferase 1Hs3st112.2 ± 0.23.5 ± 0.39.8 ± 2.14.2aNo replicates11.128.7551.4 ± 3.010.6 ± 1.6102.6 ± 14.538.1 ± 13.342.9 ± 0.78.3 ± 1.136.1 ± 21.019.3 ± 0.3 NM_029935B cell RAG-associated proteinGalnac4s-6st3.7 ± 1.41.1 ± 0.45.2 ± 0.15.7aNo replicates7.534.58 NM_013792α-N-acetylglucosaminidase (Sanfilippo disease IIIB)Naglu5.0 ± 0.22.6 ± 0.38.3 ± 0.24.8aNo replicates3.852.97 NM_010893Neuraminidase 1Neu15.8 ± 0.84.6 ± 0.53.0 ± 0.53.9 ± 0.33.283.225.8 ± 0.23.2 ± 0.25.6 ± 0.35.2 ± 0.29.8 ± 1.75.9 ± 1.63.8 ± 0.45.4aNo replicates NM_011992Reticulocalbin 2Ren24.2 ± 0.32.1 ± 0.44.4 ± 0.94.2 ± 0.73.493.235.3 ± 1.22.0 ± 0.45.4 ± 1.23.7aNo replicatesTranscription factors NM_009330Transcription factor 2Hnflb/Tcf27.3 ± 0.56.6 ± 0.822.4 ± 7.46.0 ± 0.99.6010.51 NM_010446Forkhead box A2Hnf3b/Foxa221.4 ± 0.16.3 ± 1.628.7 ± 1.04.9 ± 0.04.924.84 XM_128828Transcription factor GATA-6Gata615.7 ± 3.411.3 ± 1.335.7 ± 19.327.3 ± 0.811.9613.2311.5 ± 0.02.8 ± 0.113.5 ± 3.016.1 ± 0.33.8 ± 0.01.6 ± 0.03.9 ± 1.63.5 ± 0.4 NM_011446SRY-box containing gene 7Sox74.1 ± 0.74.3 ± 0.88.2 ± 2.56.1 ± 0.64.077.142.7 ± 0.35.1 ± 0.89.6 ± 1.510.4 ± 2.7 NM_011441SRY-box containing gene 17Sox1715 ± 0.22.2 ± 0.116.8 ± 6.36.9 ± 0.612.9310.04 NM_007709CBP/p300-interacting transactivator with Glu/Asp-rich carboxy-terminal domain 1Cited114.6 ± 0.35.6 ± 1.815.0 ± 1.629.1 ± 6.75.158.1217.0 ± 0.78.7 ± 1.451.9 ± 15.414.7 ± 0.2 NM_010137Endothelial PAS domain protein 1Epas16.9 ± 0.22.1 ± 0.17.9 ± 0.83.4 ± 0.26.233.33 NM_007498Activating transcription factor 3Atf33.7 ± 0.34.1 ± 0.54.1 ± 1.14.5 ± 0.05.079.61Intracellular signaling factors, receptors NM_023118Disabled homolog 2Dab23.0 ± 0.01.8 ± 0.38.4 ± 1.74.0 ± 0.58.559.033.6 ± 0.41.9 ± 0.291.9 ± 41.940.3 ± 2.8 NM_011309S100 calcium binding protein A1S100al1.8 ± 0.13.5 ± 0.15.8 ± 1.911.2 ± 2.13.755.922.0 ± 0.23.6 ± 0.17.8 ± 0.67.8 ± 1.62.1 ± 0.54.0 ± 0.74.2 ± 0.320.4 ± 2.7 NM_019392TYRO3 protein tyrosine kinase 3Tyro316.5 ± 2.14.2 ± 0.510.8 ± 2.95.0 ± 0.25.904.2618.0 ± 0.73.6 ± 0.219.5 ± 2.34.7 ± 1.5 NM_009026RAS, dexamethasone-induced 1Rasd19.4 ± 1.21.8 ± 0.48.1 ± 0.77.2 ± 1.57.616.19 NM_012026Rho-guanine nucleotide exchange factorRgnef4.1 ± 0.63.4 ± 0.63.6 ± 0.62.4 ± 0.24.005.243.8 ± 0.23.7 ± 0.14.1 ± 0.34.1 ± 0.3 NM_010884N-Myc downstream regulated 1Ndr15.7 ± 0.32.4 ± 0.17.8 ± 0.45.6 ± 0.43.384.336.6 ± 1.12.9 ± 0.56.2 ± 1.74.4 ± 0.0 NM_022983Endothelial differentiation, lysophosphatidic acid G-protein-coupled receptor 7Edg75.8 ± 1.94.5 ± 1.19.2 ± 2.95.6 ± 0.38.667.87 XM_141384Fibronectin leucine-rich transmembrane protein 3Flrt322.8aNo replicates6.9 ± 0.475.7aNo replicates20.0 ± 1.07.407.0513.5 ± 0.64.8 ± 0.660.5 ± 5.420.8 ± 7.2 NM_010171Tissue factor precursor (coagulation factor III)F310.9 ± 0.66.8 ± 0.89.2 ± 0.312.5 ± 0.56.006.68 NM_033603Amnionless precursor proteinAmn4.5 ± 0.61.3 ± 0.16.7 ± 2.34.4 ± 0.24.583.25Secreted factors NM_008872Plasminogen activator, tissuePlat12.1 ± 1.45.4 ± 1.629.1 ± 10.124.8 ± 2.19.6910.38 NM_009519Wingless-related MMTV integration site 11Wnt115.3 ± 0.31.1 ± 0.213.2 ± 0.03.2 ± 0.33.964.64 NM_007553Bone morphogenetic protein 2Bmp24.7 ± 0.31.4 ± 0.25.8 ± 0.04.8 ± 1.412.649.843.6 ± 0.31.2 ± 0.16.2 ± 0.23.7 ± 0.5 NM_010784MidkineMdk4.8 ± 0.33.9 ± 0.35.5 ± 1.07.2 ± 0.22.314.31 NM_008007Fibroblast growth factor 3Fgf35.1 ± 3.16.3 ± 0.24.9 ± 0.67.1 ± 0.93.835.96 S79463Semaphorin 4C (semaphorin 1)Semal3.4 ± 0.31.1 ± 0.19.5 ± 4.35.5 ± 0.36.315.29 NM_023476Lipocalin 7Lcn721.1 ± 3.82.6 ± 0.237.6 ± 3.29.1 ± 0.58.766.54 NM_023395Prostate stromal protein PS20/WAP four-disulfide core domain 1PS20/Wfdc112.7 ± 0.13.2 ± 0.731.3 ± 7.92.9 ± 0.514.525.543.7 ± 0.01.8 ± 0.14.2 ± 1.24.7 ± 0.110.7 ± 0.52.8 ± 0.525.8 ± 3.66.6 ± 1.414.1 ± 1.42.4 ± 0.023.7 ± 10.63.7aNo replicates Miscellaneous XM_125842B-cell translocation gene 1, anti-proliferativeBtg14.6 ± 0.01.2 ± 0.16.8 ± 0.63.4 ± 0.33.512.434.6 ± 0.11.4 ± 0.17.1 ± 0.93.1 ± 0.2 NM_010174Fatty acid binding protein 3, muscle and heartFabp33.4 ± 0.24 ± 0.11.5 ± 0.23.5 ± 0.12.562.244.6 ± 0.27 ± 0.21.5 ± 0.33.8 ± 0.34.0 ± 0.06.5 ± 0.11.8 ± 0.43.7 ± 0.2 XM_131373SCP-like extracellular protein containing protein3.3 ± 0.64.5 ± 0.81.4 ± 0.13.2aNo replicates3.033.41 NM_026097FRINGFring5.4 ± 0.22.4 ± 0.34.6 ± 0.33.1 ± 0.24.724.417.8 ± 0.33.0 ± 0.06.6 ± 0.43.3 ± 0.0 NM_020052Similar to CEGP1 PROTEINCegf115.6 ± 0.54.1 ± 0.236.9 ± 7.96.0 ± 0.67.817.93 NM_009292Stimulated by retinoic acid gene 8Stra84.0 ± 0.35.6 ± 1.22.7 ± 0.13.7 ± 1.34.418.29 NM_007474Aquaporin 8Aqp89.4 ± 0.35.3 ± 0.262.6 ± 12.465.7 ± 19.612.6812.11 NM_033314Solute carrier family 21 (prostaglandin transporter), member 2Slc21a27.6 ± 0.21.1 ± 0.150.1 ± 8.34.0 ± 0.19.725.31 NM_031251Cystinosis, nephropathicCtns5.3 ± 0.17.4 ± 1.03.9 ± 1.14.8 ± 0.43.553.86a No replicates Open table in a new tab For the time course analysis, F9 cells were treated with RA, Bt2cAMP, or a combination of both reagents for 96 h in duplicates. Cells were harvested every 24 h, and total RNAs were extracted. The RNAs were also prepared from untreated cells. The expression levels of individual genes were quantified using standard curves drawn with serially diluted reverse transcripts obtained from F9-PE cells treated with RA/Bt2cAMP for 96 h. Relative expression levels from duplicate experiments were averaged and expressed as mean ± range. Western Blotting—F9 cells were differentiated by stimulation with RA/Bt2cAMP for up to 96 h. The medium was collected and replaced with fresh medium containing RA/Bt2cAMP every 24 h during the differentiation. 24-h conditioned medium of F9-S was used as 0-h conditioned medium (Undiff.). 5 μl of the conditioned medium were subjected to SDS-PAGE under reducing or nonreducing conditions. Following the transfer onto Immobilon-P membrane (Millipore, Bedford, MA), BM component proteins were detected using specific antibodies for mouse EHS-laminin (Sanbio BV, Uden, The Netherlands), mouse collagen IV (LSL, Tokyo, Japan), or mouse nidogen/entactin (Chemicon, Temecula, CA). Global Gene Expression Profiles of EHS Tumor and Differentiated F9 Cells—To confirm that the overproduction of BM components by EHS tumor is regulated at the level of gene expression, total RNA from tumor cells was subjected to Northern blotting analysis using cDNAs encoding mouse laminin-1 subunits (α1, β1, and γ1) as probes. Total RNAs from F9-S and F9-PE were also analyzed. Laminin α1 subunit (Lama1) transcripts were almost absent in F9-S but were highly expressed in F9-PE and EHS tumor (Fig. 1A). Transcripts for β1 (Lamb1) and γ1 (Lamc1) subunits were also highly expressed in F9-PE and EHS tumor but were barely expressed in F9-S, demonstrating that the overproduction of laminin-1 in EHS tumor is due to the coordinated high expression of mRNAs encoding the three laminin subunits, as was the case with F9-PE. To characterize the global gene expression profiles of EHS tumor and F9-PE, a Cy-3-labeled cDNA probe prepared from either EHS tumor or F9-PE cells was competitively hybridized to the 20K-1 or 20K-2 with a Cy-5-labeled reference probe prepared from E17.5. Gene expression ratios were calculated and normalized as described under “Experimental Procedures.” The variances of the normalized expression ratios of the duplicate experiments were within 2-fold in more than 90% of the 39,166 cDNA spots. The averaged normalized expression ratios of the duplicate experiments were logarithm-transformed in base 10 and compared between EHS/E17.5 and F9-PE/E17.5 (Fig. 1B). Consistent with the results of the Northern blotting analysis, the expression levels of Lama1, Lamb1, and Lamc1 in both EHS tumor and F9-PE (Fig. 1B, red spots) were more than 10-fold higher than those in E17.5. Spots including cDNAs for the other laminin subunits such as α5, β2, γ2, and γ3 (cyan spots in Fig. 1B) showed much lower ratios than those of Lama1, Lamb1, and Lamc1 in both EHS tumor and F9-PE, indicating that laminin-1, composed of α1, β1, and γ1 subunits, is the major laminin isoform expressed in EHS tumor and F9-PE. To compare the global gene expression profiles of EHS tumor and F9-PE, a Pearson's correlation coefficient between the normalized expression ratios of EHS/E17.5 and F9-PE/E17.5 was calculated. The resulting coefficient was 0.72, indicating high similarity between EHS and F9-PE in terms of the gene expression profile. These expression profiles were further analyzed by hierarchical clustering together with the expression profiles of 49 mouse tissues in the READ (RIKEN Expression Array Database; available on the World Wide Web at read.gsc.riken.go.jp) (24.Bono H. Kasukawa T. Hayashizaki Y. Okazaki Y. Nucleic Acids Res. 2002; 30: 211-213Crossref PubMed Scopus (39) Google Scholar), which were analyzed using the RIKEN cDNA microarray with E17.5 as a common reference (14.Miki R. Kado

Referência(s)