Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

GLI2 Is Expressed in Normal Human Epidermis and BCC and Induces GLI1 Expression by Binding to its Promoter

2004; Elsevier BV; Volume: 122; Issue: 6 Linguagem: Inglês

10.1111/j.0022-202x.2004.22612.x

ISSN

1523-1747

Autores

Mohammed S. Ikram, Graham W. Neill, Gerhard Regl, Thomas Eichberger, Anna‐Maria Frischauf, Fritz Aberger, Anthony G. Quinn, Michael P. Philpott,

Tópico(s)

Genetic and rare skin diseases.

Resumo

Sonic hedgehog (Shh) binds to its receptor patched (PTCH), leading to the activation and repression of target genes via the GLI family of zinc-finger transcription factors. Deregulation of the Shh pathway is associated with basal cell carcinoma (BCC) due to upregulation of GLI1 and GLI2. We recently demonstrated a positive feedback loop between GLI1 and GLI2, which revealed that GLI1 may be a direct target of GLI2. Using band shift and luciferase reporter assays, we now show that GLI2 binds the GLI-binding consensus sequence in the GLI1 promoter. These data suggest that GLI2 directly activates GLI1 and that retrovirally expressed GLI2 induces expression of endogenous GLI1 in human primary keratinocytes. Finally, using in situ hybridization, we show that GLI2 is expressed in the interfollicular epidermis and the outer root sheath of hair follicles in normal skin as well as in BCC tumor islands. These results suggest an important role for GLI2 in regulating epidermal proliferation and skin tumorigenesis. Sonic hedgehog (Shh) binds to its receptor patched (PTCH), leading to the activation and repression of target genes via the GLI family of zinc-finger transcription factors. Deregulation of the Shh pathway is associated with basal cell carcinoma (BCC) due to upregulation of GLI1 and GLI2. We recently demonstrated a positive feedback loop between GLI1 and GLI2, which revealed that GLI1 may be a direct target of GLI2. Using band shift and luciferase reporter assays, we now show that GLI2 binds the GLI-binding consensus sequence in the GLI1 promoter. These data suggest that GLI2 directly activates GLI1 and that retrovirally expressed GLI2 induces expression of endogenous GLI1 in human primary keratinocytes. Finally, using in situ hybridization, we show that GLI2 is expressed in the interfollicular epidermis and the outer root sheath of hair follicles in normal skin as well as in BCC tumor islands. These results suggest an important role for GLI2 in regulating epidermal proliferation and skin tumorigenesis. sonic hedgehog basal cell carcinoma hedgehog outer root sheath Hedgehog (Hh) signal transduction pathway plays an essential role in a number of Drosophila and vertebrate developmental processes, including cell type specification, pattern formation, and regulation of cell proliferation. Much of our understanding of this pathway comes from studies carried out in Drosophila, in which Hh modulates gene expression through the zinc-finger-containing transcription factor Cubitus interruptus (Ci). Ci can function as either a transcriptional activator or a proteolytically cleaved repressor (Aza-Blanc and Kornberg, 1999Aza-Blanc P. Kornberg T.B. Ci: A complex transducer of the hedgehog signal.Trends Genet. 1999; 15: 458-462Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). In vertebrates, transcriptional responses to Hh signalling are mediated by the three Ci homologues, Gli1, Gli2, and Gli3 (Matise and Joyner, 1999Matise M.P. Joyner A.L. Gli genes in development and cancer.Oncogene. 1999; 18: 7852-7859Crossref PubMed Scopus (138) Google Scholar; Ruiz i Altaba, 1999Ruiz i Altaba A. Gli proteins and Hedgehog signaling: Development and cancer.Trends Genet. 1999; 15: 418-425Abstract Full Text Full Text PDF PubMed Scopus (204) Google Scholar; Ingham and McMahon, 2001Ingham P.W. McMahon A.P. Hedgehog signaling in animal development: Paradigms and principles.Genes Dev. 2001; 15: 3059-3087Crossref PubMed Scopus (2332) Google Scholar), that can act in a combinatorial fashion to modulate target gene expression. Gli1 and Gli2 appear to act primarily as transcriptional activators, whereas Gli3 (and at times GLI2) functions as a repressor (Ingham and McMahon, 2001Ingham P.W. McMahon A.P. Hedgehog signaling in animal development: Paradigms and principles.Genes Dev. 2001; 15: 3059-3087Crossref PubMed Scopus (2332) Google Scholar). Genetic analysis of patients with familial nevoid basal cell carcinoma syndrome (NBCCS) as well as individuals with sporadic BCC led to the identification of the PTCH tumor suppressor gene (Gorlin, 1987Gorlin R.J. Nevoid basal-cell carcinoma syndrome.Medicine (Baltimore). 1987; 66: 98-113Crossref PubMed Scopus (609) Google Scholar; Hahn et al., 1996Hahn H. Wicking C. Zaphiropoulous P. et al.Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome.Cell. 1996; 85: 841-851Abstract Full Text Full Text PDF PubMed Scopus (1611) Google Scholar; Johnson et al., 1996Johnson R.L. Rothman A.L. Xie J. et al.Human homolog of patched, a candidate gene for the basal cell nevus syndrome.Science. 1996; 272: 1668-1671Crossref PubMed Scopus (1571) Google Scholar); this subsequently implicated aberrant sonic hedgehog (Shh) signalling in BCC formation. Indeed, upregulation of GLI1 and GLI2 is frequently observed in BCCs (Gailani et al., 1996Gailani M.R. Stahle-Backdahl M. Leffell D.J. et al.The role of the human homologue of Drosophila patched in sporadic basal cell carcinomas.Nat Genet. 1996; 14: 78-81Crossref PubMed Scopus (631) Google Scholar; Dahmane et al., 1997Dahmane N. Lee J. Robins P. Heller P. Ruiz i Altaba A. Activation of the transcription factor Gli1 and the Sonic hedgehog signalling pathway in skin tumours.Nature. 1997; 389: 876-881Crossref PubMed Scopus (519) Google Scholar; Unden et al., 1997Unden A.B. Zaphiropoulos P.G. Bruce K. Toftgard R. Stahle-Backdahl M. Human patched (PTCH) mRNA is overexpressed consistently in tumor cells of both familial and sporadic basal cell carcinoma.Cancer Res. 1997; 57: 2336-2340PubMed Google Scholar; Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). Further evidence for BCC development as a result of activation of Hh-signalling pathway comes from transgenic mouse models and skin grafting techniques. Heterozygous Ptch+/- mice develop BCC-like features upon UV irradiation, although sporadic BCC formation does not normally occur in these mice (Aszterbaum et al., 1999Aszterbaum M. Epstein J. Oro A. Douglas V. LeBoit P.E. Scott M.P. Epstein Jr., E.H. Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous knockout mice.Nat Med. 1999; 5: 1285-1291Crossref PubMed Scopus (326) Google Scholar). Human keratinocytes expressing Shh form BCC-like structures when grafted onto the back of nude mice (Fan et al., 1997Fan H. Oro A.E. Scott M.P. Khavari P.A. Induction of basal cell carcinoma features in transgenic human skin expressing sonic hedgehog.Nat Med. 1997; 3: 788-792Crossref PubMed Scopus (230) Google Scholar) and overexpression of mediators of Hh-signalling including an oncogenic form of SMO, Shh, GLI1, and GLI2 in epidermal cells of transgenic mice leads to the induction of BCC-like tumors (Fan et al., 1997Fan H. Oro A.E. Scott M.P. Khavari P.A. Induction of basal cell carcinoma features in transgenic human skin expressing sonic hedgehog.Nat Med. 1997; 3: 788-792Crossref PubMed Scopus (230) Google Scholar; Oro et al., 1997Oro A.E. Higgins K.M. Hu Z. Bonifas J.M. Epstein Jr., E.H. Scott M.P. Basal cell carcinomas in mice overexpressing sonic hedgehog.Science. 1997; 276: 817-821Crossref PubMed Scopus (580) Google Scholar; Xie et al., 1998Xie J. Murone M. Luoh S.M. et al.Activating smoothened mutations in sporadic basal-cell carcinoma.Nature. 1998; 391: 90-92Crossref PubMed Scopus (1049) Google Scholar; Grachtchouk et al., 2000Grachtchouk M. Mo R. Yu S. Zhang X. Sasaki H. Hui C.C. Dlugosz A.A. Basal cell carcinomas in mice overexpressing Gli2 in skin.Nat Genet. 2000; 24: 216-217Crossref PubMed Scopus (313) Google Scholar; Nilsson et al., 2000Nilsson M. Unden A.B. Krause D. et al.Induction of basal cell carcinomas and trichoepitheliomas in mice overexpressing GLI-1.Proc Natl Acad Sci USA. 2000; 97: 3438-3443Crossref PubMed Scopus (340) Google Scholar). The predominance of either GLI1 or GLI2 activation with regard to BCC formation is unclear; however, we recently identified a positive feedback loop between these two proteins, which suggested that GLI1 is an early target of GLI2, whereas GLI2 is likely to be an indirect target of GLI1 (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). We now present data showing that GLI2 specifically binds to the GLI1 promoter, suggesting that GLI1 is a direct target of GLI2. Furthermore, we show that retrovirally expressed GLI2 induces endogenous GLI1 expression in human primary keratinocytes. Finally, using in situ hybridization, we show that GLI2 is expressed in the outer root sheath (ORS) of the hair follicle as well as in BCC tumor islands. These findings suggest that GLI2 directly activates GLI1 expression and may play an important role in skin tumorigenesis. We have previously identified a positive feedback loop between GLI2 and GLI1 in primary human keratinocytes. In particular, time-course analysis of GLI1 mRNA expression in GLI2 expressing keratinocytes provided preliminary evidence for GLI1 possibly being a direct target of GLI2 (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). To investigate this in more detail, we have measured the increase in GLI1 mRNA and protein in a sub-line of the human keratinocyte cell line HaCaT (GLI2-HaCaT), which expresses N-terminally HIS-tagged GLI2 protein (NHIS-GLI2) under control of the tetracycline repressor. Furthermore, we have related the increase in GLI1 mRNA and protein to levels of NHIS-GLI2 protein. Figure 1a shows that following tetracycline treatment, there was a gradual increase in GLI1 mRNA. In addition, activation of the direct target gene PTCH was slightly delayed compared with induction of GLI1 mRNA. Western blot analysis Fig 1b showed that NHIS-GLI2 protein was detected 1 h after tetracycline treatment, whereas GLI1 protein was not detected until 6 h after treatment. Moreover, the increase in GLI2 protein is closely paralleled by the gradual increase in GLI1 mRNA. This suggests that GLI1 represents an early GLI2 target gene. Figure 1a also shows that GLI1 mRNA induction is closely followed by a gradual increase in endogenous GLI2 mRNA. This confirms our previous observation that a positive feedback loop exists between GLI1 and GLI2 (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). In addition to studies in immortalized HaCat cells, we have also shown in this study Fig 2a, b that endogenous GLI1 expression is induced in retrovirally transduced GLI2 primary keratinocytes, thus confirming that GLI1 is a target of GLI2 in primary as well as immortalized cell lines. This suggests that there is also a positive feedback between GLI1 and GLI2 in human primary keratinocytes and this may be important for tumor formation. To assess whether GLI2 may be capable of directly activating GLI1 transcription by binding to the GLI1 promoter, we first performed in silico analysis of the 5′ flanking region of the human GLI1 core promoter region (Liu et al., 1998Liu C.Z. Yang J.T. Yoon J.W. Villavicencio E. Pfendler K. Walterhouse D. Characterization of the promoter region and genomic organization of GLI, a member of the sonic hedgehog-patched signaling pathway.Gene. 1998; 209: 1-11Crossref PubMed Scopus (33) Google Scholar; Villavicencio et al., 2002Villavicencio E.H. Yoon J.W. Frank D.J. Fuchtbauer E.M. Walterhouse D.O. Iannaccone P.M. Cooperative E-box regulation of human GLI1 by TWIST and USF.Genesis. 2002; 32: 247-258Crossref PubMed Scopus (32) Google Scholar) to identify putative GLI-binding sites. Using the ScanAce tool (Roth et al., 1998Roth F.P. Hughes J.D. Estep P.W. Church G.M. Finding DNA regulatory motifs within unaligned noncoding sequences clustered by whole-genome mRNA quantitation.Nat Biotechnol. 1998; 16: 939-945Crossref PubMed Scopus (744) Google Scholar; Hughes et al., 2000Hughes J.D. Estep P.W. Tavazoie S. Church G.M. Computational identification ofcis-regulatory elements associated with groups of functionally relatedgenes in Saccharomyces cerevisiae.J Mol Biol. 2000; 296: 1205-1214Crossref PubMed Scopus (791) Google Scholar), we analyzed a region of approximately 1 kb upstream of the transcriptional start site of GLI1 (Liu et al., 1998Liu C.Z. Yang J.T. Yoon J.W. Villavicencio E. Pfendler K. Walterhouse D. Characterization of the promoter region and genomic organization of GLI, a member of the sonic hedgehog-patched signaling pathway.Gene. 1998; 209: 1-11Crossref PubMed Scopus (33) Google Scholar) for motifs identical to or closely matching the GLI-consensus-binding site TGGGTGGTC (Kinzler and Vogelstein, 1990Kinzler K.W. Vogelstein B. The GLI gene encodes a nuclear protein which binds specific sequences in the human genome.Mol Cell Biol. 1990; 10: 634-642Crossref PubMed Scopus (398) Google Scholar) Fig 3a. An identical putative-binding site is also present in the cis-regulatory sequence of the mouse Gli1 gene, suggesting a possible functional relevance of this motif in the regulation of Gli1 expression Fig 3b. To determine whether NHIS-GLI2 protein can bind to the putative GLI-binding site in the GLI1 core promoter, we performed electrophoretic mobility shift assays using bacterially expressed NHIS-GLI2-(332) protein and radioactively labelled double-stranded oligonucleotides containing either the putative GLI-binding site (Bs) or a mutated control oligonucleotide with two base pair exchanges in the core region (Bsm). As shown in Fig 4, recombinant NHIS-GLI2-(332) protein specifically binds to the Bs sequence, but not to the Bsm sequence, suggesting that NHIS-GLI2 protein may be activating GLI1 transcription by directly binding to the 5′ flanking region of the GLI1 promoter at a position close to the transcriptional start site. To analyze whether the GLI-binding site in the GLI1 promoter is involved in activation of GLI1 transcription by GLI2, we performed luciferase reporter assays using wild-type and mutated GLI1 promoter constructs. Co-transfection of GLI2 expression plasmid with the reporter construct containing the 5′ upstream region of human GLI1 (-1400 to +93,Liu et al., 1998Liu C.Z. Yang J.T. Yoon J.W. Villavicencio E. Pfendler K. Walterhouse D. Characterization of the promoter region and genomic organization of GLI, a member of the sonic hedgehog-patched signaling pathway.Gene. 1998; 209: 1-11Crossref PubMed Scopus (33) Google Scholar) resulted in a 3-fold increase in luciferase activity compared with controls. By contrast, no increase in reporter activity was observed when the Gli-binding site at position -56 was changed from CGGGTGGTC to CGCCTGGTC by site-directed mutagenesis Fig 5. The GLI-binding and luciferase reporter assay data suggest that HIS-GLI2 activates transcription of GLI1 by interacting directly with the GLI-binding site at positions -56 to -48 of the GLI1 promoter. The expression profile of GLI1 is well characterized in BCC and normal skin (Ghali et al., 1999Ghali L. Wong S.T. Green J. Tidman N. Quinn A.G. Gli1 protein is expressed in basal cell carcinomas, outer root sheath keratinocytes and a subpopulation of mesenchymal cells in normal human skin.J Invest Dermatol. 1999; 113: 595-599Crossref PubMed Scopus (83) Google Scholar). Having shown that GLI1 is a putative direct target of GLI2, we investigated the expression pattern of GLI2 in BCC and normal skin by in situ hybridization Fig 6. We report that GLI2 mRNA was expressed in BCC tumor islands of all BCC samples analyzed (n=8) and in normal skin (n=3), GLI2 was expressed in the ORS of the hair follicle, sebaceous glands, and in the interfollicular epidermis. No staining was observed with a GLI2 sense probe in adjacent serial sections. Only samples showing a positive signal for keratin 14 (to confirm RNA integrity) were analyzed. Members of the GLI family of zinc-finger transcription factors are key mediators of Hh-signalling and in BCC activation of the Hh pathway results in overexpression of GLI1 and GLI2 (Green et al., 1998Green J. Leigh I.M. Poulsom R. Quinn A.G. Basal cell carcinoma development is associated with induction of the expression of the transcription factor Gli-1.Br J Dermatol. 1998; 139: 911-915Crossref PubMed Scopus (55) Google Scholar; Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). In primary human keratinocytes, retrovirally expressed GLI1 induces GLI2 expression and in HaCaT cells it has been shown that a positive feedback loop between GLI1 and GLI2 exists (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). In this study, we have shown by RT-PCR and western blot analysis that retrovirally expressed GLI2 induces endogenous GLI1 expression in primary keratinocytes, confirming that a positive feedback loop between GLI1 and GLI2 exists. Time-course analysis of GLI1 expression in GLI2-inducible cell line suggested that GLI1 may be a direct target GLI2 (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). Using gel shift and luciferase assays, we have identified a functional Gli-binding site in the GLI1 promoter that confers activation of GLI1 transcription in response to GLI2. Together with detailed time-course studies of GLI1 transcription in GLI2 expressing cells, these data suggest direct activation of GLI1 by GLI2; however, it is possible that the behavior of the NHIS-GLI2 and the endogenous GLI2 protein may not be identical. Finally, we have shown by in situ hybridization that GLI2 mRNA is expressed in both the interfollicular epidermis and ORS of hair follicles as well as in BCC tumor islands. The importance of Shh-Gli2 signalling in murine hair follicle development has been demonstrated by several groups (St-Jacques et al., 1998St-Jacques B. Dassule H.R. Karavanova I. et al.Sonic hedgehog signaling is essential for hair development.Curr Biol. 1998; 8: 1058-1068Abstract Full Text Full Text PDF PubMed Google Scholar; Chiang et al., 1999Chiang C. Swan R.Z. Grachtchouk M. et al.Essential role for Sonic hedgehog during hair follicle morphogenesis.Dev Biol. 1999; 205: 1-9Crossref PubMed Scopus (403) Google Scholar; Karlsson et al., 1999Karlsson L. Bondjers C. Betsholtz C. Roles for PDGF-A and sonic hedgehog in development of mesenchymal components of the hair follicle.Development. 1999; 126: 2611-2621PubMed Google Scholar; Brewster et al., 2000Brewster R. Mullor J.L. Ruiz i Altaba A. Gli2 functions in FGF signaling during antero-posterior patterning.Development. 2000; 127: 4395-4405PubMed Google Scholar; Bai et al., 2002Bai C.B. Auerbach W. Lee J.S. Stephen D. Joyner A.L. Gli2, but not Gli1, is required for initial Shh signaling and ectopic activation of the Shh pathway.Development. 2002; 129: 4753-4761Crossref PubMed Scopus (87) Google Scholar; Mill et al., 2003Mill P. Mo R. Fu H. Grachtchouk M. Kim P.C. Dlugosz A.A. Hui C.C. Sonic hedgehog-dependent activation of Gli2 is essential for embryonic hair follicle development.Genes Dev. 2003; 17: 282-294Crossref PubMed Scopus (222) Google Scholar) and transgenic mice expressing GLI2 in the basal layer develop BCC (Grachtchouk et al., 2000Grachtchouk M. Mo R. Yu S. Zhang X. Sasaki H. Hui C.C. Dlugosz A.A. Basal cell carcinomas in mice overexpressing Gli2 in skin.Nat Genet. 2000; 24: 216-217Crossref PubMed Scopus (313) Google Scholar). As expected for a target of GLI2, the expression profile of GLI1 correlates with that of GLI2 in both the ORS of the hair follicle and in BCC. Antibodies to GLI2 are not available and, therefore, we have been unable to confirm that the patterns of GLI2 protein mimic those of its mRNA. We have shown that in primary human keratinocytes GLI2 induces GLI1, therefore, the lack of GLI1 expression may indicate that GLI2 is not activated in the interfollicular epidermis and that other factors may be required for transcriptional activity. It has recently been demonstrated in transgenic mice that Gli2 activation in mouse skin is dependent on Shh. However, it has been shown that during embryogenesis, Gli2 can be activated in response to FGF independent of Shh, providing evidence that GLI2 can function independent of Shh. Moreover, it has also been shown that deletion of the N-terminal repressor domain of GLI2 produces a mutant protein that can induce Gli1 and Ptch expression independent of Shh (Sasaki et al., 1999Sasaki H. Nishizaki Y. Hui C. Regulation of Gli2 and Gli3 activities by an amino-terminal repression domain: Implication of Gli2 and Gli3 as primary mediators of Shh signaling.Development. 1999; 126: 3915-3924Crossref PubMed Google Scholar; Brewster et al., 2000Brewster R. Mullor J.L. Ruiz i Altaba A. Gli2 functions in FGF signaling during antero-posterior patterning.Development. 2000; 127: 4395-4405PubMed Google Scholar; Mill et al., 2003Mill P. Mo R. Fu H. Grachtchouk M. Kim P.C. Dlugosz A.A. Hui C.C. Sonic hedgehog-dependent activation of Gli2 is essential for embryonic hair follicle development.Genes Dev. 2003; 17: 282-294Crossref PubMed Scopus (222) Google Scholar). Currently, little is known about processing of GLI2 and it will be important to determine, therefore, whether GLI2 is processed and if so whether differences in processing occur between normal skin compartments, such as interfollicular epidermis, ORS of hair follicle and in BCC. We have recently found that GLI2 represses expression of genes associated with epidermal differentiation, suggesting that GLI2 may have repressor activity in the human epidermis (Regl et al, in press). In BCC, GLI1 and GLI2 are consistently expressed in tumor islands and this may be accounted for by the positive feedback mechanism that we have identified (Green et al., 1998Green J. Leigh I.M. Poulsom R. Quinn A.G. Basal cell carcinoma development is associated with induction of the expression of the transcription factor Gli-1.Br J Dermatol. 1998; 139: 911-915Crossref PubMed Scopus (55) Google Scholar; Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). Autoregulation of Gli1 expression has previously been observed in frog embryos, suggesting that positive feedback loops may be a common feature to sustain pathway activation in the absence of Shh stimulation (Dahmane et al., 2001Dahmane N. Sanchez P. Gitton Y. et al.The sonic hedgehog-Gli pathway regulates dorsal brain growth and tumorigenesis.Development. 2001; 128: 5201-5212PubMed Google Scholar). Whether Gli1 or Gli2 is the initial factor that induces the feedback loop and possibly tumor formation remains to be determined. In contrast to Gli1, expression of Gli2 is not dependent on Shh and during mouse embryogenesis Gli2 is expressed before the onset of Shh and Gli1 transcription (Bai et al., 2002Bai C.B. Auerbach W. Lee J.S. Stephen D. Joyner A.L. Gli2, but not Gli1, is required for initial Shh signaling and ectopic activation of the Shh pathway.Development. 2002; 129: 4753-4761Crossref PubMed Scopus (87) Google Scholar). In addition, Gli2 is required for the expression of Gli1 in murine skin (Mill et al., 2003Mill P. Mo R. Fu H. Grachtchouk M. Kim P.C. Dlugosz A.A. Hui C.C. Sonic hedgehog-dependent activation of Gli2 is essential for embryonic hair follicle development.Genes Dev. 2003; 17: 282-294Crossref PubMed Scopus (222) Google Scholar). This suggests that Gli2 induces Gli1, which in turn further increases Gli2 levels leading to hyperproliferation (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar). In summary, we provide further evidence for a positive feedback loop between GLI1 and GLI2 that potentiates the Shh signal in hair follicle morphogenesis and BCC formation. This indicates a possible important role for human GLI2 in regulating skin tumorigenesis. The study was approved by the local ethical committee (ELCHA Research Ethics Committee, Ref: T/01/037 and adhered to the declaration of Helsinki guidelines for use of human tissue.) HaCaT cells were grown at 37°C in Dulbecco's modified medium (DMEM) adjusted to pH 7.5 with 5 mM HEPES, containing 10% fetal calf serum (FCS), 100 mg per L streptomycin, and 62.5 mg per L penicillin. Double-stable HaCaT lines, expressing N-terminally HIS-tagged human GLI2 (GenBank GI No. 3061315) (NHIS-GLI2) under the control of the tetracycline repressor, were generated using the T-REX system (Invitrogen, Carlsbad, CA). NHIS-GLI2 expression was induced by culturing cells in DMEM-high glucose (Invitrogen), containing 10% FCS and 1 mg per L tetracycline (Invitrogen). Primary human keratinocytes were isolated and cultured according to the methods described in Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar. The retroviral SIN-GLI1-EGFP construct was generated as described in Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar. To generate the retroviral bicistronic GLI2-EGFP expression construct, N-terminally HIS-tagged GLI2 was cloned into pI2E-A, a modified version of the pIRES2-EGFP plasmid (Clontech, Hampshire, UK) (Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar), PI2E-A-GLI2 was digested with SalI and NotI to excise CMV-GLI2-IRES-EGFP. The resulting fragment was cloned into XhoI–NotI digested retroviral SIN-IP plasmid (a gift from Prof. P. Khavari) to create SIN-GLI2-EGFP. Retroviral transduction of primary keratinocytes and cell transfections were carried out as described in Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar. Cells were harvested at the time points indicated in the text. Isolation of total RNA from HaCaT cells and real-time RT-PCR was carried out as described previously in Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar. Total RNA from primary keratinocytes was isolated using RNAzol B (Biogenesis Ltd, Poole, Dorset, UK) as per the manufacturer's instructions. Genomic DNA was removed by treatment with DNaseI (Boeringer Mannheim, Roche; Basal, Switzerland) and cDNA was synthesized using AMV Reverse Transcriptase (Promega A5000 kit, Madison, WI) according to the manufacturer's instructions. PCR was carried out using 1 μL cDNA reaction, 0.25 mM each dNTP, 1×(25 mM (NH4)2SO4, 93.75 mM Tris-HCl, pH 8.8, at 25°C, 1.25×10−2 Tween 20) reaction buffer, 1.75 mM MgCl2, 0.125 μM each (forward and reverse) primers with 1 U thermal stable Red Hot DNA polymerase (Advanced Biotechnology, Epsom, UK) in a total reaction volume of 20 μL. The samples were thermo cycled, at 94°C for 1 min (94°C for 1 min, 60°C for 30 s, 72°C for 1 min)×40 cycles, and final extension at 72°C for 10 min. G6PDH was amplified with an annealing temperature set at 55°C in 35 cycles. A negative primer-only control without cDNA was also set up. PCR-primer sequences were as follows: GLI1 forward primer 5′ gaagacctctccagcttgga 3′ corresponding to 1504–1523 nucleotides and reverse primer 5′ ggctgacagtataggcagag 3′ corresponding to 1728–1747 nucleotides of GLI1 (GenBank GI No. 4885278). G6PDH forward primer 5′ gttccgtgaggaccagatctac 3′ corresponding to 584–605 nucleotides, reverse primer 5′ ggctccttgaaggtgaggataa 3′ corresponding to 712–733 nucleotides of G6PDH (GenBank GI No. 182870). Western blot analysis of GLI1 and NHIS-GLI2 protein expression was carried out as described in Regl et al., 2002Regl G. Neill G.W. Eichberger T. et al.Human GLI2 and GLI1 are part of a positive feedback mechanism in basal cell carcinoma.Oncogene. 2002; 21: 5529-5539Crossref PubMed Scopus (157) Google Scholar. GLI1 protein was detected with polyclonal goat anti-GLI1 antibody (C-18, Santa Cruz Biotechnology, Santa Cruz, CA) and NHIS-GLI2 with peroxidase-conjugated monoclonal anti-polyhistidine antibody (Sigma, clone HIS-1, Sigma, Vienna Austria; Dorset, UK). Proteins were visualized by chemiluminescence detection (ECL, Amersham Biosciences, Bucks, UK; Uppsala, Sweden). Proteins levels were quantified by densitometric analysis using the AIDA/2D densitometry software package (RAYTEST). For the production of recombinant GLI2 protein, a fragment encoding amino acids 1–386 of GLI2 was PCR amplified using the following primers: forward: 5′ gagggatccgccctcacctccatcaat 3′, reverse: 5′ gaggaattcctaggtcatcatgttcagg 3′. The amplicon was cloned into the EcoRI–BamHI sites of Escherichia coli expression vector pHIS-Parallel2 (Sheffield et al., 1999Sheffield P. Garrard S. Derewenda Z. Overcoming expression and purification problems of RhoGDI using a family of 'parallel' expression vectors.Protein Exp Purif. 1999; 15: 34-39Crossref PubMed Scopus (508) Google Scholar). To increase the solubility of GLI2 protein, the insert was further truncated at the C-terminus with SmaI and XhoI yielding a protein corresponding to amino acids 1–332, which encompasses the N-terminus and the entire zinc-finger DNA-binding domain (NHIS-GLI2-(332)). NHIS-GLI2-(332) protein expression in E. coli strain BL21 was induced for 60 min by the addition of 1 mM IPTG (Sigma). HIS-tagged GLI2-(332) protein was purified with Ni-NTA Agarose (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Binding reactions for band shift assays were carried out in 1×BS buffer (20 mM Tris, pH7.8, 25 mM KCl, 5 mM MgCl2, 0.5 mM DTT, and 10 μM ZnSO4) containing 0.77 μg poly-(dI–dC) (Sigma) and 10% glycerol. Five micrograms of purified NHIS-GLI2-(332) protein and 8 ng of radioactively labelled double-stranded oligonucleotide were added to the reaction and incubated for 25 min at room temperature. In competition experiments, 400 ng of specific or mutant unlabelled oligonucleotide were used per reaction. For unspecific competition, 1.5 μg poly-dI–dC was added to the reaction. Samples were separated on 6% acrylamide gels. Following electrophoresis, gels were dried, exposed overnight, and scanned with a BAS-1800II phosphoimager (Fuji, Fuji, Japan). The 5′ upstream regulatory region of the human GLI1 gene from -1400 to +93 relative to the transcriptional start site (Liu et al., 1998Liu C.Z. Yang J.T. Yoon J.W. Villavicencio E. Pfendler K. Walterhouse D. Characterization of the promoter region and genomic organization of GLI, a member of the sonic hedgehog-patched signaling pathway.Gene. 1998; 209: 1-11Crossref PubMed Scopus (33) Google Scholar) was amplified by PCR using the FailSafe PCR system for high-fidelity PCR cloning (Epicentre, Madison, WI). To mutate the Gli-binding site, the GLI1 promoter was first subcloned into pBluescript II SK+ (Stratagene, La Jolla, CA) and then subjected to site-directed mutagenesis with the QuickChange Kit (Stratagene, La Jolla, CA) according to the manufacturer's instructions. Following mutagenesis, the GLI1 promoter region was transferred back to pGL3 reporter vector. The sequence of the oligonucleotide used for mutagenesis was 5′ GCGCTTCTCGCGCCTGGTCCGGGCTTG 3′. Successful mutagenesis was verified by DNA-sequencing on an ABI 310 automated sequencer. For luciferase reporter assays, 293 cells were co-transfected in 12-well plates with GLI2 expression plasmid (GLI2b in pcDNA4/TO, Invitrogen) or empty pcDNA4/TO, and wild-type or mutated GLI1 reporter plasmid. β-galactosidase expression vector (pcDNA4/TO-LacZ, Invitrogen) was co-transfected for normalization of transfection efficiency. Cells were harvested at 48 h post-transfection and luciferase assays were carried out using Luciferase Assay Substrate (Promega), according to the manufacturer's instructions. Luciferase activity was measured on a Lucy II luminometer (Anthos, Wals, Austria). To normalize results for lacZ activity, 10 μL of cell lysate was mixed with 240 μL Z-buffer (100 mM NaPO4 (pH 7), 10 mM KCl2, 1 mM MgSO4, 50 mM b-Mercaptoethanol), and 50 μL O-nitro-α-D-phenyl-galactopyranoside (4 mg per mL) and incubated at 28°C until yellow staining became visible. Reactions were stopped by adding 250 μL 1 M NaCO3. β-galactosidase activity was quantified by measuring absorbance at 405 nm on an SLT-Spectra plate reader (SLT, Austria). GLI2 and keratin 14 sense and anti-sense probes were generated from GLI2 cDNA corresponding to 28,270–28,566 nucleotides (GenBank GI No. 3061315) and keratin 14 cDNA corresponding to 298–1498 nucleotides (GenBank GI No. 15431309) cloned into pCRII-TOPO vector, using DIG-labelling kit (Roche, Basal, Switzerland; Lewes East Sussex, UK) as described by the manufacturer's instructions. Keratin 14 probes were fragmented and purified using sodium acetate ethanol precipitation method (Wilkinson, 1998Wilkinson D.G. 2nd edn. In Situ Hybridization. Vol. 196. Oxford University Press, Oxford1998Google Scholar). Eight micrometer tissue sections were pre-hybridized with (4×SSC, 1×Denhardt's, 50% formamide (500 μg per mL tRNA and 500 μg per mL Salmon Testes DNA denatured at 100°C for 10 min and placed on ice) before adding to the mix) and incubated at 42°C for 3–4 h. Hybridization was carried out using fresh pre-hybridized solution containing (80–100 ng labelled probe denatured at 65°C for 5 min) at 42°C overnight. The next day, the sections were washed in 2×SSC for 5 min (×2) and in (2×SSC, 1×SSC, 0.5×SSC each containing 50% formamide) at 45–55°C and in 0.1×SSC 50% formamide at 50–60°C for 20 min, followed by a final wash in 2×SSC and rinsed in DIG buffer 1 (100 mM Tris-HCl, 150 mM NaCl, pH 7.5). Sections were blocked with 10% normal sheep serum (NSS) in DIG buffer 1 and incubated with anti-digoxigenin-alkaline phosphatase conjugated (Roche) diluted 1:400 in 1% NSS DIG buffer 1 for 2 h, followed by washing in DIG buffer 1 (×2) and DIG buffer 2 (100 mM Tris-HCl, pH 9.5, 100 mM NaCl, 50 mM MgCl2) for 10 min. The Hybrids were visualized by incubating the section with BCIP/NBT (Sigma) liquid substrate in the dark at 4°C overnight. The color reaction was stopped by immersing the sections in (10 mM Tris-HCl, pH 8, 1 mM EDTA) for 30 min. The developed slides were mounted and examined by microscopy. This project was supported by Medical Research Council UK, the Austrian Science Fund (FWF project P14227-MOB) and the 'Schwerpunkt: Biomedizin und Gesundheit' of the University of Salzburg. Thomas Eichberger was supported by an EMBO short-term fellowship during a stay in the Center for Cutaneous Research in London.

Referência(s)