Sonic Hedgehog Acts as a Negative Regulator of β-Catenin Signaling in the Adult Tongue Epithelium
2010; Elsevier BV; Volume: 177; Issue: 1 Linguagem: Inglês
10.2353/ajpath.2010.091079
ISSN1525-2191
AutoresFabian T. Schneider, Anne Schänzer, Cathrin J. Czupalla, Sonja Thom, Knut Engels, Mirko H. H. Schmidt, Karl H. Plate, Stefan Liebner,
Tópico(s)Wnt/β-catenin signaling in development and cancer
ResumoWnt/β-catenin signaling has been implicated in taste papilla development; however, its role in epithelial maintenance and tumor progression in the adult tongue remains elusive. We show Wnt/β-catenin pathway activation in reporter mice and by nuclear β-catenin staining in the epithelium and taste papilla of adult mouse and human tongues. β-Catenin activation in APCmin/+ mice, which carry a mutation in adenomatous poliposis coli (APC), up-regulates Sonic hedgehog (Shh) and Jagged-2 (JAG2) in the tongue epithelium without formation of squamous cell carcinoma (SCC). We demonstrate that Shh suppresses β-catenin transcriptional activity in a signaling-dependent manner in vitro and in vivo. A similar regulation and function was observed for JAG2, suggesting that both pathways negatively regulate β-catenin, thereby preventing SCC formation in the tongue. This was supported by reduced nuclear β-catenin in the tongue epithelium of Patched+/− mice, exhibiting dominant active Shh signaling. At the invasive front of human tongue cancer, nuclear β-catenin and Shh were increased, suggesting their participation in tumor progression. Interestingly, Shh but not JAG2 was able to reduce β-catenin signaling in SCC cells, arguing for a partial loss of negative feedback on β-catenin transcription in tongue cancer. We show for the first time that the putative Wnt/β-catenin targets Shh and JAG2 control β-catenin signaling in the adult tongue epithelium, a function that is partially lost in lingual SCC. Wnt/β-catenin signaling has been implicated in taste papilla development; however, its role in epithelial maintenance and tumor progression in the adult tongue remains elusive. We show Wnt/β-catenin pathway activation in reporter mice and by nuclear β-catenin staining in the epithelium and taste papilla of adult mouse and human tongues. β-Catenin activation in APCmin/+ mice, which carry a mutation in adenomatous poliposis coli (APC), up-regulates Sonic hedgehog (Shh) and Jagged-2 (JAG2) in the tongue epithelium without formation of squamous cell carcinoma (SCC). We demonstrate that Shh suppresses β-catenin transcriptional activity in a signaling-dependent manner in vitro and in vivo. A similar regulation and function was observed for JAG2, suggesting that both pathways negatively regulate β-catenin, thereby preventing SCC formation in the tongue. This was supported by reduced nuclear β-catenin in the tongue epithelium of Patched+/− mice, exhibiting dominant active Shh signaling. At the invasive front of human tongue cancer, nuclear β-catenin and Shh were increased, suggesting their participation in tumor progression. Interestingly, Shh but not JAG2 was able to reduce β-catenin signaling in SCC cells, arguing for a partial loss of negative feedback on β-catenin transcription in tongue cancer. We show for the first time that the putative Wnt/β-catenin targets Shh and JAG2 control β-catenin signaling in the adult tongue epithelium, a function that is partially lost in lingual SCC. The Wnt/β-catenin and the hedgehog (Hh)/Gli pathways are essential for various developmental processes, and their adult deregulation has been implicated in several pathologies, namely cancer and neurodegenerative diseases (for review see Clevers1Clevers H Wnt/beta-catenin signaling in development and disease.Cell. 2006; 127: 469-480Abstract Full Text Full Text PDF PubMed Scopus (4491) Google Scholar and Ingham2Ingham PW Hedgehog signalling.Curr Biol. 2008; 18: R238-R241Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). The key protein in “canonical” Wnt signaling is β-catenin, which acts as a transcription factor with members of the lymphoid enhancer factor (Lef)/T-cell factor family.3Harris TJ Peifer M Decisions, decisions: beta-catenin chooses between adhesion and transcription.Trends Cell Biol. 2005; 15: 234-237Abstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar On Wnt ligand-mediated activation of a receptor complex formed by frizzled and low-density lipoprotein receptor-related protein 5/6, proteasomal degradation of β-catenin is inhibited by inactivating the destruction complex assembled by glycogen synthase kinase 3β (GSK3β), adenomatous poliposis coli (APC), and axin.1Clevers H Wnt/beta-catenin signaling in development and disease.Cell. 2006; 127: 469-480Abstract Full Text Full Text PDF PubMed Scopus (4491) Google Scholar Consequently, β-catenin circumvents ubiquitinylation and subsequent destruction, favoring its transcriptional activity together with Lef/T-cell factor. In the adult, Wnt signaling participates in the maintenance of several tissues, such as crypt cells of the colon, endosteal cells of the bone marrow, the hair bulge stem cell niche, and the hippocampal dentate gyrus.4Scheller M Huelsken J Rosenbauer F Taketo MM Birchmeier W Tenen DG Leutz A Hematopoietic stem cell and multilineage defects generated by constitutive beta-catenin activation.Nat Immunol. 2006; 7: 1037-1047Crossref PubMed Scopus (337) Google Scholar, 5Harada N Tamai Y Ishikawa T Sauer B Takaku K Oshima M Taketo MM Intestinal polyposis in mice with a dominant stable mutation of the beta-catenin gene.EMBO J. 1999; 18: 5931-5942Crossref PubMed Scopus (968) Google Scholar, 6Maretto S Cordenonsi M Dupont S Braghetta P Broccoli V Hassan AB Volpin D Bressan GM Piccolo S Mapping Wnt/beta-catenin signaling during mouse development and in colorectal tumors.Proc Natl Acad Sci USA. 2003; 100: 3299-3304Crossref PubMed Scopus (687) Google Scholar In these tissues activation of the Wnt pathway likely plays a role in stem or progenitor cell maintenance.7Reya T Clevers H Wnt signalling in stem cells and cancer.Nature. 2005; 434: 843-850Crossref PubMed Scopus (3013) Google Scholar Overactivation of “canonical” Wnt signaling in the adult is an initial step to neoplasia, as known to be the case in familial adenomatous poliposis. In familial adenomatous poliposis and in APCmin/+ mice, the corresponding mouse model, one allel of the tumor suppressor gene APC is mutated, leading to diminished β-catenin degradation and, therefore, to increased signaling.1Clevers H Wnt/beta-catenin signaling in development and disease.Cell. 2006; 127: 469-480Abstract Full Text Full Text PDF PubMed Scopus (4491) Google Scholar Hence the Wnt pathway requires tight regulation. Interestingly, the Wnt and the Sonic hedgehog (Shh) pathways were described to have distinct and reciprocal, repressive function in various tissues.8Borycki A Brown AM Emerson Jr, CP Shh and Wnt signaling pathways converge to control Gli gene activation in avian somites.Development. 2000; 127: 2075-2087Crossref PubMed Google Scholar In mammals there are three members of the Hh family: Indian hedgehog, Desert hedgehog, and Shh of which Shh has the most prominent developmental effect.9Jia J Jiang J Decoding the Hedgehog signal in animal development.Cell Mol Life Sci. 2006; 63: 1249-1265Crossref PubMed Scopus (87) Google Scholar In the absence of Hh ligand, patched (Ptch), a 12-transmembrane spanning protein, represses smoothened (SMO) thereby inhibiting Hh signaling. SMO, like the frizzleds, is a member of the 7-transmembrane spanning G-protein-coupled receptor-like superfamily and transduces the Hh signal form the plasma membrane to the cytoplasm. On Hh binding to Ptch its repressive function on SMO is released,10Cohen MM The hedgehog signaling network.Am J Med Genet A. 2003; 123A: 5-28Crossref PubMed Scopus (379) Google Scholar thereby activating Gli1/2 dependent transcription. In the repressive state Gli1/2 proteins are phosphorylated by GSK3β, CK1, and PKA, promoting the ubiquitination and thereby the degradation of Gli1/2.2Ingham PW Hedgehog signalling.Curr Biol. 2008; 18: R238-R241Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar Similar to the Wnt/β-catenin pathway, Shh signaling requires tight regulation as it has been implicated in tumorigenesis in particular in the skin and the cerebellum of the brain.11Goodrich LV Milenkovic L Higgins KM Scott MP Altered neural cell fates and medulloblastoma in mouse patched mutants.Science. 1997; 277: 1109-1113Crossref PubMed Scopus (1429) Google Scholar Although the Wnt and Shh pathways were implicated in the formation of various cancers, their role in the oral tongue epithelium and hence in lingual squamous cell carcinoma (LSCC) remains elusive. Expression of Shh and its receptor Ptch was detected in the developing tongue taste papilla,12Hall JM Hooper JE Finger TE Expression of sonic hedgehog, patched, and Gli1 in developing taste papillae of the mouse.J Comp Neurol. 1999; 406: 143-155Crossref PubMed Scopus (97) Google Scholar, 13Nakayama A Miura H Shindo Y Kusakabe Y Tomonari H Harada S Expression of the basal cell markers of taste buds in the anterior tongue and soft palate of the mouse embryo.J Comp Neurol. 2008; 509: 211-224Crossref PubMed Scopus (29) Google Scholar and subsequently the importance of a functional Shh signal during embryonic mouse taste bud development was shown.14Mistretta CM Liu HX Gaffield W MacCallum DK Cyclopamine and jervine in embryonic rat tongue cultures demonstrate a role for Shh signaling in taste papilla development and patterning: fungiform papillae double in number and form in novel locations in dorsal lingual epithelium.Dev Biol. 2003; 254: 1-18Crossref PubMed Scopus (92) Google Scholar Recently, β-catenin/Wnt signaling was identified in developmental placode and taste papilla formation in the tongue, establishing this pathway as an upstream factor of Shh, Sox2, and BMP4 in this tissue.15Iwatsuki K Liu HX Grunder A Singer MA Lane TF Grosschedl R Mistretta CM Margolskee RF Wnt signaling interacts with Shh to regulate taste papilla development.Proc Natl Acad Sci USA. 2007; 104: 2253-2258Crossref PubMed Scopus (130) Google Scholar, 16Liu F Thirumangalathu S Gallant NM Yang SH Stoick-Cooper CL Reddy ST Andl T Taketo MM Dlugosz AA Moon RT Barlow LA Millar SE Wnt-beta-catenin signaling initiates taste papilla development.Nat Genet. 2006; 39: 106-112Crossref PubMed Scopus (124) Google Scholar, 17Okubo T Pevny LH Hogan BL Sox2 is required for development of taste bud sensory cells.Genes Dev. 2006; 20: 2654-2659Crossref PubMed Scopus (149) Google Scholar Here we show activation of the canonical Wnt pathway during tongue epithelial turnover in the healthy mouse and human. We provide evidence that the increase in β-catenin transcriptional activity in APCmin/+ mice up-regulates Shh, Jagged-2 (JAG2), and Notch2 without affecting epithelial morphology and homeostasis. Shh and JAG2 are involved in down-regulating β-catenin signaling, thus participating in counterbalancing activation of the canonical Wnt pathway. In samples of human LSCC (hLSCC), nuclear β-catenin and cytoplasmic Shh are frequently detected at the invasive front, correlating with tumor grade and proliferation. Although Shh inhibits β-catenin signaling in an hLSCC cell line, JAG2-mediated down-regulation of β-catenin signaling was abolished in these cells. This suggests that the concomitant repression of β-catenin signaling by Shh and JAG2 are required to control epithelial homeostasis and to prevent cancer formation. BAT-gal reporter mice have been provided by Stefano Piccolo (Padua, Italy),6Maretto S Cordenonsi M Dupont S Braghetta P Broccoli V Hassan AB Volpin D Bressan GM Piccolo S Mapping Wnt/beta-catenin signaling during mouse development and in colorectal tumors.Proc Natl Acad Sci USA. 2003; 100: 3299-3304Crossref PubMed Scopus (687) Google Scholar and C57BL/6J-ApcMin/J (APCmin/+) by Markus F. Neurath (Mainz, Germany). Offspring was genotyped by genomic PCR from tail biopsy. Ptch+/− mice were kept in a C57BL/6J background as described previously.11Goodrich LV Milenkovic L Higgins KM Scott MP Altered neural cell fates and medulloblastoma in mouse patched mutants.Science. 1997; 277: 1109-1113Crossref PubMed Scopus (1429) Google Scholar Adult mice were anesthetized with isofluorane and subsequently sacrificed by cervical dislocation. Biopsies were collected randomly from 19 patients with LSCC at the Department of Pathology, Frankfurt University Hospital (15 male patients; average age, 56.8 years; four women; average age, 87 years). The age of the patients varied between 41 and 96, with a mean of 66.6 years of age. According to World Health Organization tumor classification, tumors were classified as grade I (n = 3), grade II (n = 13), and grade III (n = 3; Supplemental Table T1, see http://ajp.amjpathol.org). Use of archival specimen was approved by the Ethical Committee of Surgery, Frankfurt University Hospital, Frankfurt/Main, Germany. For analysis of human tongue, papillae postmortem specimens were used. Immunohistochemical staining of the formalin fixed, paraffin embedded tissue was performed according to standard procedures. Nuclei were counterstained with TOTO-3 (Molecular Probes, Eugene, OR; Table 1). Histology and eosin counterstaining on paraffin sections was according to standard procedures. For immunofluorescent staining, unfixed samples were embedded in TissueTek OCT (Sakura Finetech Europe B.V., Zoeterwoude, NL), sectioned, and labeled as described previously.18Liebner S Cattelino A Gallini R Rudini N Iurlaro M Piccolo S Dejana E Beta-catenin is required for endothelial-mesenchymal transformation during heart cushion development in the mouse.J Cell Biol. 2004; 166: 359-367Crossref PubMed Scopus (316) Google ScholarTable 1List of Primary AntibodiesAntigenHostDilutionManufacturerAnti-HA high affinity antibodyRat1:500Roche Diagnostics, Mannheim, GermanyCleaved-caspase-3Rabbit1:100R&D Systems, Minneapolis, MNCytokeratin 5/6Mouse1:50Chemicon International Inc., Temecula, CACytokeratin Endo-A (TROMA-I)Rat1:200Developmental Studies Hybridoma Bank, Iowa City, IAGα-gustducinRabbit1:200Santa Cruz Biotechnology, Santa-Cruz, CAKi67 antigen (M7249)Rat1:50Dako Denmark, Glostrup, DenmarkNotch2Rat1:200Developmental Studies Hybridoma BankPLCγ2Rabbit1:200Santa Cruz Biotechnology, Santa Cruz, CAShhGoat1:50R&D Systems, Minneapolis, MNβ-catenin (clone 14)Mouse1:200BD Biosciences, San Jose, CAβ-galactosidaseRabbit1:1000Abcam, Cambridge, UK Open table in a new tab Fluorescence was observed with an epifluorescence microscope (Nikon 80i; Nikon, Tokyo, Japan) and documented with a digital camera (DS-5Mc; Nikon), or with a confocal microscope (Nikon C1si; Nikon). Images were computer processed by using ImageJ (National Institutes of Health, Bethesda, MD) and Adobe Photoshop CS3 for Macintosh. Tongues from BAT-gal mice were fixed in 4% paraformaldehyde, permeabilized in PBS (0.02% NP-40). Staining in potassium ferri-cyanide solution was performed over night at 4°C and additional 5 hours at room temperature. Samples were photographed and subjected to standard embedding procedures for paraffin or cryostat sectioning. Mouse Jagged-2 cDNA was kindly provided by Verdon Taylor (University of Sheffield, Sheffield, UK), and mouse Shh cDNA was obtained from imaGenes (imaGenes GmbH, Berlin, Germany). In situ hybridization analysis was performed by standard procedures. Briefly, cryo sections were fixed in 4% paraformaldehyde, dehydrated in an ascending alcohol series, air dried, and refrozen at −70°C. To inactivate endogenous alcaline phosphatases and basic proteins, the sections were incubated with 0.2 M HCl, washed, and proteins were denatured by proteinase K (10 μg/ml). Proteinase K was inactivated with 0.1 M Glycin, and slides were rinsed and refixed in 4% paraformaldehyde. Titration of 625 μl acetic anhydride in 250 ml 0.1 M triethanolamine prehybridization in premix (for 10 ml: 5 ml de-ionized formamide, 2 ml 20× standard saline citrate, 1 ml 50% dextransulfate, 1 ml 50× Denhardt’s, 0.05 ml 20% SDS, t-RNA [10 mg/ml] 0.25 ml) for 5 hours at room temperature. The probes were diluted 1:20 in premix and heated to 95°C, cooled down to 48°C, and incubated over night at 68°C. The next day, slides were rinsed twice in standard saline citrate at 70°C and twice standard saline citrate at room temperature. Incubation in dioxygenin-1 (DIG1, Roche, Mannheim, Germany) was followed by blocking in DIG2 (Roche, Mannheim, Germany). Primary antibody was diluted 1:500 in DIG2 and incubated for 1 hour at room temperature, rinsed three times in DIG1, and 1 time in DIG3. Staining solution (polyvenyl alcohol, DIG3, and 4-nitroblue-tetrazolium-[5-bromo-4-chloro-3-indolylphosphate]) was incubated over night at 30°C, followed by 10 washes in PBS. Slides were counterstained by hematoxylin and mounted in Aqua Polymount. The tongue epithelial layer was dissected from 5-month-old APCmin/+ mice and control litters, and total RNA was extracted in guanidium solution (4 M Guanidine Thiocyanate, 0.5% N-Lauroylsarcosine sodium salt, 25 mmol/L tri-Sodium Citrate 2-hydrate, 0.1 mmol/L 2-Mercaptoethanol), centrifuged o.n. over a cesium-chloride gradient (18°C/28,000 rpm/min/16 hours; SW41Ti; Beckman Coulter), precipitated in sodium-acetate, and resuspended in RNase-free water.19Sambrook J Fritsch AF Maniatis T Harbor CS Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, New York1989Google Scholar RT-PCR was performed with the First Strand cDNA Sythesis Kit (Fermentas GmbH, St. Leon-Rot, Germany) by using 1 μg total RNA per reaction. Resulting cDNA was digested with RNase H (Roche Diagnostics GmbH). For quantitative RT-PCR (qRT-PCR), Absolute QPCR SYBR Green Fluorescein Mix (ThermoFisher Scientific, Waltham, MA) was used according to the manufacturer’s protocol. qRT-PCR conditions were as follows: 15 minutes at 95°C, 45 cycles of 30 seconds at 95°C, 30 seconds at 62°C, and 35 seconds at 72°C in a BioRad MyiQ Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA). Raw data were analyzed with iQ5-Standard Edition software (Bio-Rad Laboratories, Hercules, CA). Primers used were as follows: Shh (sense 5′-TGGAAGCAGGTTTCGACTGG-3′; antisense 5′-GGAAGGTGAGGAAGTCGCTGT-3′); Gli1 (sense 5′-CCTTTAGCAATGCCAGTGACC-3′; antisense 5′-GAGCGAGCTGGGATCTGTGTAG-3′); patched homolog 1 (Ptch1; sense 5′-TTGGGATCAAGCTGAGTGCTG-3′; antisense 5′-CGAGCATAGCCCTGTGGTTCT-3′); Jagged-2 (sense 5′-TTCCTGGATGGAGACTGCAAC-3′; antisense 5′-CTGACCAGAGAGCAGGCAAGG-3′); and Notch2 (sense 5′-GCCAACTGCACCTCCACTCTT-3′; antisense 5′-AGCCACACTCCTCGCTGTTG-3′) in murine tongue samples. The expression was analyzed with GraphPad Prism 5.0 for Macintosh (GraphPad Software, LaJolla, CA). Samples were normalized by the expression of RNA-polymerase II (MGI: 98086) and G6PDX (MGI:105979) and shown as percentage of control. Paraffin sections stained for β-catenin were photographed with a 40× objective. For each tongue (APCmin/+ and APC+/+, Ptch+/− and Ptch+/+ n = 3), six slices were counted, and raw data were analyzed with GraphPad Prism 5.0 for Macintosh. Statistical analysis was performed by using the unpaired, two-sided t-test (Student’s t-test) between the groups. Significant differences were assumed at a level of P < 0.05. The data are expressed as mean values ± SEM. The same procedure was used for the quantification for Ki-67 and terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assays. Luciferase assay was preformed according to the protocol of Dyer et al.20Dyer BW Ferrer FA Klinedinst DK Rodriguez R A noncommercial dual luciferase enzyme assay system for reporter gene analysis.Anal Biochem. 2000; 282: 158-161Crossref PubMed Scopus (197) Google Scholar HEK293 cells were seeded in 24-well plates and transfected by the calcium-phosphate method with the following plasmids for canonical Wnt signaling: super(8×)TOP-FLASH or super(8×)FOP-FLASH (generous gifts from Dr. R. Moon, University of Washington School of Medicine, Seattle, WA) and pRL-CMV (Promega, Madison, WI). For Hh signaling, cells were either transfected with the 8 × 3′Gli-BSδ51LucII or the 8 × 3′GliM3-BSδ51LucII (generous gifts from Dr. Hiroshi Sasaki, Center of Developmental Biology, RIKEN Kobe, Japan) and pRL-CMV. If desired, cells were also transfected with 500 ng mouse (m) Shh-pCMVSport6.1 (imaGenes), 300 ng pcDNA3.1-mNotch2, 600 ng pcDNA3.1-ratJAG2, 250 ng of pcDNA3.1-HisB-Gli1 (generous gift from Dr. Hiroshi Sasaki, Center of Developmental Biology, RIKEN Kobe, Japan), or 50 ng pCS2+-LEFΔN-βCTA (generous gift from Dr. A. Hecht, Freiburg, Germany). The concentrations of all transfections were brought up to the same level of DNA with the vector pTRE2hyg (Clontech-Takara Bio Europe, Saint-Germain-en Laye, France) or pcDNA3.1/Hygro (+) (Invitrogen, Carlsbad, CA). After 24 hours, cells were stimulated with 20 mmol/L LiCl, Shh-conditioned medium (CM), control-CM, 10 μmol/L cyclopamine (Calbiochem, Darmstadt, Germany), 5 to 10 μmol/L dimethyl sulfoxide, or 5 to 10 μmol/L γ-secretase inhibitor N-[N-(3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT) (Sigma, Munich, Germany). Twenty-four hours later, cells were washed in PBS and lysed, and luciferase activity was determined on a Berthold Lumat LB 9507 luminometer (Berthold Technologies GmbH & Co KG, Bad Wildbad, Germany). Luciferase activity was normalized to TK-renilla luciferase, and values were blotted as fold induction to the control values. Three independent experiments are shown as averages with SEM. HEK293 cells (3 × 105) were seeded on a 6-well plate. After 24 hours, cells were transfected with 5.5 μg of pCMV-Sport6.1-Shh, pcDNA3.1-Jagged2, pcDNA3.1-Notch2, or pcDNA3.1/hygro (+) plasmids. Forty-eight hours after transfection, cells were washed two times in ice cold PBS, scraped in ice-cold lysing buffer (50 mmol/L TricHCl; 150 mmol/L NaCl; 1% NP-40 [Igepal Ca-630]; 0.5% Sodium deoxycholate; 0.5% SDS; 1 mmol/L EDTA; and 1 mmol/L phenylmethylsulfonyl fluoride; Complete Mini, Roche Diagnostics GmbH, Mannheim, Germany), and centrifuged for 10 minutes at 13,000 rpm. Total extracts were separated by 7% or 12% SDS-polyacrylamide gel electrophoresis and analyzed in immunoblot with specific antibodies according to standard procedures. To investigate the canonical Wnt pathway in adult mice, we performed LacZ staining in BAT-gal reporter mice, demonstrating activation of the Wnt pathway in some fungiform and, most prominent, in filiform papilla that was corroborated by nuclear β-catenin staining (Supplemental Figure S1, see http://ajp.amjpathol.org). As dominant activation of the canonical Wnt pathway plays a role in tumor formation, we asked whether APCmin/+ mice exhibit alterations of epithelial and taste papilla turnover in the tongue. The frequency of nuclear β-catenin in the tongue epithelium of APCmin/+ mice was significantly increased compared with littermate controls (Figure 1A and D; G and J; quantification in M, **P < 0.005). To evaluate proliferation, tongues of APCmin/+ mice were sectioned and stained for the proliferation marker Ki-67. The quantification of Ki-67 positive cells revealed no significant difference between APCmin/+ mice and controls (Figure 1B, E, H, K, and N), also apoptosis did not significantly differ, as evaluated by TUNEL staining (Figure 1C, F, I, L, and N) and staining for active caspase-3 (Supplemental Figure S2, see http://ajp.amjpathol.org). We further addressed the question if the increase in β-catenin signaling affects the morphology and cellular architecture of the tongue epithelium in general and the taste buds (TB) in particular. Confocal analysis of staining against α-Gustducin and the TB specific phospholipase Cβ2, as well as scanning electron microscopy, did not reveal any alterations in fungiform papilla of APCmin/+ mice versus wild-type (Supplemental Figure S3, see http://ajp.amjpathol.org). Furthermore, we investigated tongue histology by H&E staining and evaluated the morphology of epithelial layers by morphometric analysis, which did not reveal significant differences between wild-type and APCmin/+ mice (data not shown). To understand the effect of constitutive β-catenin activation on other pathways involved in tongue epithelial development and maintenance, we investigated the expression and distribution of Shh, which plays an antagonistic role during embryonic tongue development on Wnt signaling. Immunohistochemical staining of Shh was stronger in the epithelium of APCmin/+ mice compared to controls (Figure 2, A–D). In situ hybridization (Figure 2, E and F) and qRT-PCR (Figure 2G) corroborated the up-regulation on the mRNA level, suggesting that the increase in nuclear β-catenin correlates with an up-regulation of Shh. Furthermore, Ptch1, which is a known downstream target of Shh, was also significantly up-regulated in APCmin/+ tongues (**P < 0.005), indicating increased Hh signaling activity (Figure 2G), whereas Gli1 mRNA was not effected. To understand the influence of Shh on β-catenin transcriptional activity when the degradation complex is inhibited, which is the case in APCmin/+ mice, we investigated superTOP-FLASH or mutant superFOP-FLASH reporter activity in HEK293 cells in vitro. Cells were transfected with the superTOP/FOP-FLASH reporter constructs and with increasing amounts of the control vector or mShh (1 to 1000 ng cDNA), respectively. Transcriptional activation of β-catenin was stimulated with 20 mmol/L LiCl, which blocks GSK3β enzymatic activity. Shh inhibited superTOP-FLASH activation in LiCl stimulated HEK293 cells in a dose-dependent manner, reaching maximal inhibition at 250 ng cDNA (Figure 3A). To proof the specific effect of Shh, we confirmed the expression of the protein by Western blot in transfected cells and evaluated activation of canonical Shh signaling by 8 × 3′Gli-LucII reporter assay21Sasaki H Hui C Nakafuku M Kondoh H A binding site for Gli proteins is essential for HNF-3beta floor plate enhancer activity in transgenics and can respond to Shh in vitro.Development. 1997; 124: 1313-1322Crossref PubMed Google Scholar (Figure 3B). mGli1 transfection resulted in significant reporter induction and served as a positive control. Transfected Shh showed only a moderate induction after 24 hours, whereas Shh-CM elicited significant reporter activation. As GSK3β inhibition by LiCl may also interfere with the signaling of other pathways such as Shh, we made use of a Lef-1-β-catenin fusion construct (LefΔN-βCTA), conferring dominant activation of β-catenin-dependent signaling without interfering with the endogenous β-catenin pool.22Vleminckx K Kemler R Hecht A The C-terminal transactivation domain of beta-catenin is necessary and sufficient for signaling by the LEF-1/beta-catenin complex in Xenopus laevis.Mech Dev. 1999; 81: 65-74Crossref PubMed Scopus (88) Google Scholar Shh also significantly inhibited superTOP-FLASH reporter activation induced by LefΔN-βCTA, suggesting that Shh represses β-catenin signaling at the transcriptional level, independently of the β-catenin degradation complex (*P < 0.0147; **P < 0.0015; Figure 3C). For Shh conditioned medium, we obtained similar results, indicating that extracellular Shh acts via a receptor-mediated pathway (**P < 0.002; *P < 0.0241; Figure 3D). This interpretation is supported by the finding that inhibition of canonical hedgehog signaling by the alkaloid cyclopamine (10 μmol/L), abrogated the inhibitory effect of Shh on the superTOP-FLASH reporter (Supplemental Figure S4, see http://ajp.amjpathol.org). To understand if the Shh pathway negatively regulates β-catenin signaling in the adult tongue in vivo, we made use of Ptch heterozygous mice (Ptch+/−) that exhibit a dominant activation of Shh signaling and evaluated the incidence of nuclear β-catenin in the tongue epithelium in comparison with wild-type mice (Figure 4, A–D). The number of β-catenin positive nuclei in Ptch+/− tongues was significantly decreased compared to wild-type controls (Figure 4I). Shh ligand appeared to be decreased in immunohistochemical staining (Figure 4, E–H). This was confirmed on the mRNA level by qRT-PCR (***P < 0.0001), whereas Gli1 was significantly up-regulated (*P < 0.0187) and the Shh target gene Ptch1 was not regulated compared to wild-type (Figure 4J). This suggests that dominant active Shh signaling dampened β-catenin nuclear translocation and therefore, its signaling activity in the epithelium of Ptch heterozygous tongues. Whether the differential regulation of β-catenin and Shh signaling plays a role in the human tongue, we investigated postmortem tongue tissue of normal and tumor bearing tongues. First we confirmed that nuclear β-catenin and Shh are also present in the healthy human tongue epithelium (Supplemental Figure S5, see http://ajp.amjpathol.org). To clarify if Wnt/β-catenin signaling and Shh show a similar regulation in human tongue cancer as observed in APCmin/+ and Ptch+/− mice, respectively, we analyzed hLSCC (World Health Organization, grades I to III). Paraffin sections were stained against cytokeratin 5/6 (CK5/6), β-catenin, Shh, and MIB1. Staining against CK5/6 confirmed the epithelial origin of the tumor cells, which infiltrated the tongue tissue in a typical cone-like manner (Figure 5A, E, and I). With increased tumor malignancy, epithelial differentiation of tumor cells was decreased, showing more frequently single CK5/6 positive cells disseminating in the stromal tissue (Figure 5, E and I). β-Catenin showed distinct junctional staining between tumor cells in grade I hLSCC, whereas nuclear staining was regularly, but sparsely observed (Figure 5B). In grade II and grade III tumors, the junctional staining of β-catenin was lost at the invasive front of the tumor, whereas nuclear staining was increased (Figure 5, F and J). Shh staining intensity increased with tumor malignancy, and the localization essentially followed the nuclear β-catenin staining observed at the invading front (Figure 5C, G, and K). To evaluate the proliferative phenotype of the tumors, sections were stained against the proliferation marker MIB1. Proliferation notably increased from grade I to grade III tumors, which parallels the increase in nucle
Referência(s)