Portal de Periódicos da CAPES

Activation of Expression of Hedgehog Target Genes in Basal Cell Carcinomas

2001; Elsevier BV; Volume: 116; Issue: 5 Linguagem: Inglês

10.1046/j.1523-1747.2001.01315.x

1523-1747

Jeannette M. Bonifas, Ervin H. Epstein, Sally D. Pennypacker, Pao-Tien Chuang, Andrew P. McMahon, Mickey Williams, Arnon Rosenthal, Frederic J. de Sauvage,

Cancer and Skin Lesions

Mutations in hedgehog signaling pathway genes, especially PTC1 and SMO, are pivotal to the development of basal cell carcinomas. The study of basal cell carcinoma gene expression not only may elucidate mechanisms by which hedgehog signaling abnormalities produce aberrant tumor cell behavior but also can provide data on in vivo hedgehog target gene control in humans. We have found, in comparison with normal skin, that basal cell carcinomas have increased levels of mRNA for PTC1, GLI1, HIP, WNT2B, and WNT5a; decreased levels of mRNA for c-MYC, c-FOS, and WNT4; and unchanged levels of mRNA for PTC2, GLI2, WNT7B, and BMP2 and 4. These findings suggest that mutations in hedgehog signaling pathway genes may exert both cell autonomous and indirect effects and indicate that basal cell carcinoma tumor cells have a phenotype that at least in some aspects resembles that of epidermal stem cells. Mutations in hedgehog signaling pathway genes, especially PTC1 and SMO, are pivotal to the development of basal cell carcinomas. The study of basal cell carcinoma gene expression not only may elucidate mechanisms by which hedgehog signaling abnormalities produce aberrant tumor cell behavior but also can provide data on in vivo hedgehog target gene control in humans. We have found, in comparison with normal skin, that basal cell carcinomas have increased levels of mRNA for PTC1, GLI1, HIP, WNT2B, and WNT5a; decreased levels of mRNA for c-MYC, c-FOS, and WNT4; and unchanged levels of mRNA for PTC2, GLI2, WNT7B, and BMP2 and 4. These findings suggest that mutations in hedgehog signaling pathway genes may exert both cell autonomous and indirect effects and indicate that basal cell carcinoma tumor cells have a phenotype that at least in some aspects resembles that of epidermal stem cells. basal cell carcinoma human epidermal keratinocyte hedgehog ribonuclease protection assay Basal cell carcinoma (BCC), although the commonest human cancer, had been subjected to relatively little molecular analysis until several years ago, when heritable mutations in PATCHED1 (PTC1) were found to underlie the basal cell nevus syndrome (MIM no. 109400) (Hahn et al., 1996Hahn H. Wicking C. Zaphiropoulos P.G. et al.Mutations of the human homologue 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), a rare autosomal dominant affliction characterized by multiple phenotypic abnormalities, including the development of multiple BCC starting in youth. PTC1 inhibits signaling by the membrane protein Smoothened (Smo), and this inhibition is relieved by binding Sonic Hedgehog (SHH) to PTC. Unrestricted Smo signaling can affect target gene transcription via the GLI family of transcription factors (GLI1, GLI2, and GLI3). PTC1 is itself a target for HH signaling, thus forming a feedback loop in which decreased PTC1 protein function leads to the production of more PTC1 mRNA. In BCC, the feedback loop is disrupted because the increased mRNA encodes a nonfunctional PTC1 protein, and indeed PTC1 mRNA, as well as SMO, GLI1, and GLI3 mRNA, have been reported to accumulate in BCC. Other genes whose expression has been described to be activated by HH signaling in various tissues include members of the WNT/Wg and BMP/TGFβ families of signaling molecules and the gene encoding the cell-surface HH-binding protein HIP (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 tumors.Nature. 1997; 389: 876-881https://doi.org/10.1038/39918Crossref PubMed Scopus (519) Google Scholar;Ingham, 1998Ingham P.W. The patched gene in development and cancer.Curr Opin Genet Dev. 1998; 8: 88-94Crossref PubMed Scopus (95) Google Scholar;Chuang and McMahon, 1999Chuang P.-T. McMahon A.P. Vertebrate hedgehog signalling modulated by induction of a hedgehog binding protein.Nature. 1999; 397: 617-621https://doi.org/10.1038/17611Crossref PubMed Scopus (587) Google Scholar;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: 595-599https://doi.org/10.1046/j.1523-1747.1999.00729.xCrossref PubMed Scopus (83) Google Scholar). So far, it remains unclear how mutations in PTC1 and SMO and consequent changes in HH target gene expression produce the abnormal proliferation, differentiation, and adhesion characteristic of BCC tumor cells. As part of an effort to investigate this question, we have surveyed BCC for in vivo quantitative abnormalities of mRNA encoding hedgehog target genes, hemidesmosomal proteins, and a selection of genes that encode proteins important in apoptosis and cell cycle regulation. Aliquots of nodular BCC removed from the skin by curettage and normal human skin (obtained as part of breast reduction surgery) or epidermis (the roof of suction blisters of the volar forearm) were frozen immediately in liquid nitrogen and were stored at -70°C until assayed or were fixed and processed for routine microscopic examination. We used a Qiagen RNAeasy kit (Qiagen, Valencia, CA) to extract RNA from these tissues and from cultured normal cells [preconfluent epidermal keratinocytes (HEK) grown in 0.07 mM Ca2+ and fibroblasts] and cell lines (HaCaT, HeLa, and D259MG). mRNA for PTC1, PTC2, GLI1, c-MYC, HIP, and GAPDH was quantitated by TaqMan. Primer and probe sequences are available as follows: Patched 1: forward (5′-TCT TCA TGG CCG CGT TAA TC-3′), reverse (5′-TTG CAG GAA AAA TGA GCA GAA C-3′), probe (5′-FAM-AAT TCC CGC TCT GCG GGC G-TAMARA-3′); Patched 2: forward (5′-CCA GCA CCC CCT CAT CAG-3′), reverse (5′-GAC CCG AAG ACC AAT TCA GC-3′), probe (5′-FAM-CAC AAG GAG CGC CAC TGT CTG GAC-TAMARA-3′); GLI1: forward (5′-TGA GGC CCT TCA AAG CCC-3′), reverse (5′-GTA TGA CTT CCG GCA CCC TTC-3′), probe (5′-FAM-TGC TGG TGG TTC ACA TGC GCA GA-TAM ARA-3′); HIP: forward (5′-TGC TAA GCC TCG CAT TCC A-3′), reverse (5′-ACA ACC CTA AGA ATG TGG TCA TGA-3′), probe (5′-FAM-TGT ATG TGT CCT ATA CCA CCA ACC AAG AAC GG-TA-3′); c-Myc: forward (5′-CAG CTG CTT AGA CGC TGG ATT-3′), reverse (5′-GAG GTC ATA GTT CCT GTT GGT GAA-3′), probe (5′-FAM-CTC CCG CGA CGA TGC CCC T-3′); GADPH: forward (5′-GAA GGT GAA GGT CGG AGT CA-3′), reverse (5′-GAA GAT GGT GAT GGG ATT TC 3′), probe (5′-FAM-CAA GCT TCC CGT TCT CAG CC–TAMARA-3′). Ribonuclease protection assays (RPA) were performed using the Ambion RPAII kit according to the manufacturer's recommendations (Ambion, Austin, TX). Probes used for RPA were produced by polymerase chain reaction (PCR) amplification of published sequences (GLI2, INTEGRIN β4), were purchased from Pharmingen or Ambion, or were generous gifts (PTC1, GLI1, INTEGRIN α6, and BPAG1 and BPAG2). DNA was linearized to give anti-sense probes of 150–300 bp and was labeled with a Promega (Madison, WI) transcription kit. Gels were scanned for quantitative densitometry using a BioRad (Hercules, CA) Model GS-670 imaging densitometer. The relative levels of three controls (“housekeeping” genes – GAPDH, ACTIN, and CYCLOPHILIN) were consistent in each sample, suggesting that none of the three changed significantly in these samples. We identified WNT family members expressed in HEK and in BCC by reverse transcription–PCR amplification using consensus WNT family primers: forward (5′-GGG GAA TTC CAR GAR TGY AAR TGY CAY-3′, reverse (5′-AAA ATC ATC TAG ARC ARC ACC ART GRA A-3′) (Gavin et al., 1990Gavin B.J. McMahon J.A. McMahon A.P. Expression of multiple novel Wnt-1/int-1-related genes during fetal and adult mouse development.Genes Dev. 1990; 4: 2319-2332Crossref PubMed Scopus (394) Google Scholar). Levels of PTC1 mRNA in all 27 BCC studied were elevated markedly compared with controls (epidermis, skin, or HEK) with a 15-fold variation among individual tumors. Levels were 10–500-fold higher by TaqMan and up to 1600-fold higher by RPA; however, PTC1 expression in the controls was very low, and consequently precise quantitation by RPA was difficult. Like those of PTC1, levels of PTC2 mRNA were high in BCC and were low in HEK; however, unlike those of PTC1, levels of PTC2 mRNA also were high in normal epidermis. Cultured dermal fibroblasts and HaCaT cells had uniformly low levels of PTC1 (albeit higher than the levels in HEK) and of PTC2 mRNA. D259MG cells, the cell line from which amplified GLI1 originally was isolated, had low levels of PTC2 mRNA but, unlike other cultured cells studied, had elevated levels of PTC1 mRNA—as high as those in BCC (Table I).Table IPTC1, PTC2, and GLI1 expression in BCC mRNA levels were measured by TaqMan analysis and were normalized to the expression levels of GAPDHaValues in the table are (2⁁Delta CT) *100 and, for BCC, represent 12–20 individual tumors for PATCHED 1, PATCHED 2, and GLI1 and five tumors for c-MYC and HIP.PATCHED 1PATCHED 2GLI1c-MYCHIPHEK0–0.140.1–0.50.007–0.2170–0.05Epidermis0–0.59110.194–990–0.01Normal skin0–2.21.6–62 (av. 18)0.7–1.7210.04BCC2–100 (av. 18)2.3–54 (av. 17)7–102 (av. 41)0.8–14 (av. 6)0.1–3.6 (av. 1.4)D259MG7–110.2275NDNDa Values in the table are (2⁁Delta CT) *100 and, for BCC, represent 12–20 individual tumors for PATCHED 1, PATCHED 2, and GLI1 and five tumors for c-MYC and HIP. Open table in a new tab Levels of GLI1 message were elevated (vs that in HEK or skin) in all 12 BCC studied by TaqMan, the average level being approximately 1/6 that in D259MG cells (Table I). In seven BCC studied by RPA, levels of GLI2 varied over a 9-fold range, were absent in HEK and in D259MG cells, but were as high in skin as in BCC. Of five BCC tested by RPA three had mRNA levels 10-fold increased, and two had levels approximately 100-fold increased vs those in HEK. The levels in the two latter BCC were similar to that in D259MG cells. HIP expression was undetectable by RPA in extracts of normal skin (Figure 1). In five BCC tested by TaqMan, HIP expression averaged 40-fold higher than in normal skin and nearly 300-fold higher than in epidermis or HEK (Table I). By cloning reverse transcription–PCR product prepared from RNA, we found five different WNT family members (2, 2b, 4, 5a, and 7b) to be expressed in cultured HEK and/or BCC, of which WNT7b was found the most commonly. By RPA WNT2b and 5a were undetectable to very low in HEK and skin but clearly were expressed in BCC, albeit at varying levels. By contrast, WNT4 was expressed in epidermis and skin but not in BCC (or in HEK). WNT7b was detected readily in all BCC, epidermis, and skin but was undetectable in HEK. WNT2 expression was highly elevated in one of five BCCs studied but was undetectable by RPA in HEK. Of eight BCC tested by RPA all had near-complete loss of c-MYC mRNA as compared with that in HEK, epidermis, or whole skin. D259MG cells also have low c-MYC mRNA, whereas levels in HeLa and HaCaT cells are at least as high as those in HEK (Figure 2). With Taqman, c-MYC expression is very high in epidermis and at least 15-fold lower in BCC (Table I). mRNA for BMP2 and BMP4 were detectable at similar levels in BCC, skin, and epidermis. BMP2 and BMP4 expression levels were absent to very low in HEK. We assessed mRNA levels of the four major known components of hemidesmosomes in BCC. mRNA for BPAG1 and BPAG2 were readily detectable in BCC and were present at levels similar to those in HEK or whole skin. α6 and β4 integrin mRNA are visible as faint bands in RPA of extracts of BCC, normal skin, or epidermis and are expressed at much higher levels in HEK. The single BCC with high α6 integrin expression did not differ clinically or histologically from those with lower expression. Levels of expression of several genes involved in cell cycle regulation (p53, Rb, p21, and cdk4) and in control of apoptosis (bclx, bax, and mcl) varied among individual tumors and were not markedly different from those of skin. One consistent change seen was an at least 100-fold decrease in mRNA for c-FOS in BCC (n = 8) vs in normal skin (Figure 3), a finding that is consistent with previous reports of high levels of c-FOS expression in adult human skin, especially in suprabasal cells, and reduced levels in BCC (Basset-Seguin et al., 1990Basset-Seguin N. Escot C. Blanchard J.M. et al.High levels of c-fos proto-oncogene expression in normal human adult skin.J Invest Dermatol. 1990; 94: 418-422Abstract Full Text PDF PubMed Google Scholar;Takahashi et al., 1994Takahashi S. Pearse A.D. Marks R. Expression of c-fos proto-oncogene mRNA in non-melanoma skin cancer.J Dermatol Sci. 1994; 7: 54-62Abstract Full Text PDF PubMed Scopus (15) Google Scholar). These data are consistent with the growing evidence that hedgehog target gene activation is the pivotal step in BCC carcinogenesis and is the first study of the levels of expression of GLI2, WNT, BMP, c-MYC, and HIP in BCC. Furthermore, we have generated quantitative data regarding the degree to which the message for three HH-regulated genes accumulate in these tumors: PTC1, GLI1, and HIP mRNA levels are at least 50–100-fold increased in vivo in BCC; levels that we have found to be matched in vitro only by the levels in D259MG cells, in which GLI1 is highly amplified. The uniformity of this upregulation of PTC1 mRNA in BCC indicates that loss of PTC1 protein function rarely is due to deletions of both genomic copies, production of unstable mRNA, or prevention of transcription of PTC1 message. The PTC2 levels that we detected confirm those seen by in situ hybridization (Carpenter et al., 1998Carpenter D. Stone D.M. Brush J. et al.Characterization of two patched receptors for the vertebrate hedgehog protein family.Proc Natl Acad Sci USA. 1998; 95: 13630-13634Crossref PubMed Scopus (200) Google Scholar;Zaphiropoulos et al., 1999Zaphiropoulos P.G. Unden A.B. Rahnama F. Hollingsworth R.E. Toftgard R. PTCH2, a novel human patched gene, undergoing alternative splicing and up-regulated in basal cell carcinomas.Cancer Res. 1999; 59: 787-792PubMed Google Scholar). The lack of upregulation of PTC2 expression by amplified GLI1 in D259MG cells, like the maintenance of PTC2 expression in the lung and skin of Shh –/– mouse embryos (Pepicelli et al., 1998Pepicelli C.V. Lewis P.M. McMahon A.P. Sonic hedgehog regulates branching morphogenesis in the mammalian lung.Curr Biol. 1998; 8: 1083-1086Abstract Full Text Full Text PDF PubMed Scopus (491) Google Scholar;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), is consistent with the conclusion that, at least in these tissues, PTC2 expression is not an HH target gene. Overexpression of GLI1 or GLI2 in the epidermis of transgenic mice can drive the development of BCC (Grachtchouk et al., 2000Grachtchouk M. Mo R. Yu S. et al.Basal cell carcinomas in mice overexpressing Gli2 in skin.Nat Genet. 2000; 24: 216-217https://doi.org/10.1038/73417Crossref 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). Our finding of levels of GLI1 that were upregulated consistently, of levels of GLI2 mRNA that were less consistent and on average were no higher than in normal skin, and of stronger RPA bands with GLI1 than GLI2 all suggest that GLI1 may be more important in vivo in driving the aberrant cell behavior characteristic of BCC. We have no information, however, regarding possible post-translational activation of these proteins in epidermis, and so the primacy of GLI1 must be considered a tentative conclusion. HIP is a gene recently identified in a cell-free screen for murine genes encoding proteins that bind Shh. It is present at the cell surface where it binds Shh with an avidity similar to that of PTC1, is a transcriptional target of Shh signaling during mouse development, and appears to attenuate Shh signaling, most likely by binding Shh and thereby reducing its interaction with Ptc1 (Ingham, 1998Ingham P.W. The patched gene in development and cancer.Curr Opin Genet Dev. 1998; 8: 88-94Crossref PubMed Scopus (95) Google Scholar). Our finding of high levels of HIP mRNA in BCC indicates that its expression is an in vivo target of HH signaling in humans as well as in mice. The repertoire of WNT genes expressed in adult human epidermis has not previously been explored systematically.Saitoh et al., 1998Saitoh A. Hansen L.A. Vogel J.C. Udey M.C. Characterization of Wnt gene expression in murine skin: possible involvement of epidermis-derived Wnt-4 in cutaneous epithelial–mesenchymal interactions.Exp Cell Res. 1998; 243: 150-160https://doi.org/10.1006/excr.1998.4152Crossref PubMed Scopus (41) Google Scholar analyzed WNT expression in adult murine epidermis using a strategy similar to ours: cloning of PCR product amplified from WNT consensus primers followed by RPA quantitation. They found clones of Wnt4, 7a, and 10b, expression only of Wnt4 and 10b in epidermis, and of Wnt4 in newborn mouse epidermal keratinocytes in culture. We cloned sequences corresponding to WNT2b, 4, 5a, and 7b from cDNA prepared from cultured HEK and to WNT2, 2b, 4, 5a, and 7b from cDNA prepared from BCC. Levels of WNT2 expression measured by RPA were low to undetectable in HEK. Expression of WNT2b, 5a, 7b and 13 were readily detectable in BCC. Human (like mouse epidermis) does express WNT4 but this WNT is undetectable in BCC. As the bulk of the cells in these samples were tumor cells, most likely the WNT messages detected in BCC were derived from the malignant cells rather than from “contaminating” normal stromal cells, but confirmation of this conclusion must await in situ studies. The expression of multiple WNT genes detectable at low levels by RPA is reminiscent of reports of similar studies in breast cancers (Huguet et al., 1994Huguet E.L. McMahon J.A. McMahon A.P. Bicknell R. Harris A.L. Differential expression of human Wnt genes 2, 3, 4, and 7B in human breast cell lines and normal and disease states of human breast tissue.Cancer Res. 1994; 54: 2615-2621PubMed Google Scholar), and different WNT have been described as having varying transforming potencies (Wong et al., 1994Wong G.T. Gavin B.J. McMahon A.P. Differential transformation of mammary epithelial cells by WNT Genes.Mol Cell Biol. 1994; 14: 6278-6286Crossref PubMed Scopus (278) Google Scholar). One intracellular pathway by which Wnt signaling proceeds is via stabilization of β-catenin, which interacts with the Tcf-Lef family of transcription factors to affect gene expression. Stabilization of β-catenin protein by mutations in the β-catenin gene in keratinocytes produce pilomatrixomas, a skin tumor distinct from BCC (Chan et al., 1999Chan E.F. Gat U. McNiff J.M. Fuchs E. A common human skin tumour is caused by activating mutations in beta-catenin.Nat Genet. 1999; 21: 410-413https://doi.org/10.1038/7747Crossref PubMed Scopus (516) Google Scholar). Stabilization of β-catenin protein in colon cancer cells, often by mutations in APC, is associated with upregulation of c-MYC expression (He et al., 1998He T.C. Sparks A.B. Rago C. et al.Identification of c-MYC as a target of the APC pathway [see comments].Science. 1998; 281: 1509-1512Crossref PubMed Scopus (3924) Google Scholar). In contrast, despite their upregulation of WNT2b and 5a and hence potential β-catenin stabilization, we have found that BCC have c-MYC downregulation as compared either to skin or to cultured normal HEK. Different WNT species clearly affect the phenotype of cells differently, as manifested both by their different patterns of expression in normal vs malignant cells and also by their differing abilities to stabilize β-catenin and to transform, e.g., mammary epithelial cells in which specific model WNT4, 5a, and 7b have no effect (Shimizu et al., 1997Shimizu H. Julius M.A. Giarre M. Zheng Z. Brown A.M. Kitajewski J. Transformation by Wnt family proteins correlates with regulation of beta-catenin.Cell Growth Differ. 1997; 8: 1349-1358PubMed Google Scholar), and it is possible that different WNT produce different effects on c-MYC expression as well. BCC have been suggested to arise from epidermal stem cells that are located in the bulge region of the hair follicles. That postulate is based not only on the continued proliferation and resistance to differentiation of BCC cells but also on their frequent expression of antigens similar to those expressed by the cells of the hair follicle bulge region, e.g., keratins 15 and 19 and α2, α3, and β1 integrins. Also consistent with this view is the observation that ptc +/– mice have upregulation of ptc promoter activity in the bulge cells as well as in BCC (Aszterbaum et al., 1999Aszterbaum M.A. Epstein J. Oro A. Scott M.P. Epstein E.H. A mouse model of human basal cell carcinoma: Ultraviolet and gamma radiation enhance basal cell carcinoma growth in patched heterozygote knock-out mice.Nature Med. 1999; 5: 1285-1291Crossref PubMed Scopus (326) Google Scholar). Epidermal stem cells express low levels of c-MYC despite their relatively high levels of β-catenin protein and despite the ability of exogenous stabilized β-catenin to enhance stem cell numbers in vitro (Zhu and Watt, 1999Zhu A.J. Watt F.M. beta-catenin signalling modulates proliferative potential of human epidermal keratinocytes independently of intercellular adhesion.Development. 1999; 126: 2285-2298PubMed Google Scholar). Upregulation of c-MYC expression can drive human epidermal stem cells into a transient amplifying phenotype (Gandarillas and Watt, 1997Gandarillas A. Watt F.M. c-Myc promotes differentiation of human epidermal stem cells.Genes Dev. 1997; 11: 2869-2882Crossref PubMed Scopus (270) Google Scholar), and most but not all studies report that c-MYC expression then is downregulated coincident with the onset of keratinocyte terminal differentiation. Consistent with these findings, c-MYC expression in transgenic mice recently was found to drive epidermal keratinocyte proliferation and to inhibit normal complete differentiation in vivo (Pelengaris et al., 1999Pelengaris S. Littlewood T. Khan M. Elia G. Evan G. Reversible activation of c-Myc in skin: induction of a complex neoplastic phenotype by a single oncogenic lesion.Mol Cell. 1999; 3: 565-577Abstract Full Text Full Text PDF PubMed Scopus (389) Google Scholar;Waikel et al., 1999Waikel R.L. Wang X.J. Roop D.R. Targeted expression of c-Myc in the epidermis alters normal proliferation, differentiation and UV-B induced apoptosis.Oncogene. 1999; 18: 4870-4878Crossref PubMed Scopus (65) Google Scholar). Hence, at least in their failure to switch from proliferation to differentiation and their downregulation of c-MYC, BCC resemble epidermal stem cells. Of note, however, the putative epidermal stem cells of the hair follicle bulge, unlike BCC, do express α6 and β4 integrins and bullous pemphigoid proteins. Thus we speculate that upregulation of expression of WNT2b and/or 5a (and/or downregulation of WNT4 expression) and a possible resultant downregulation of c-MYC expression may be a crucial pathway by which hedgehog target gene dysregulation produces at least some aspects of the BCC phenotype. Our quantitation of hemidesmosomal protein mRNA levels suggests that the well-established loss of hemidesmosomal proteins in BCC may not be controlled by loss of mRNA for integrins α6 and β4 or BPAG1 or 2. However, in contrast to our results, others have found consistent loss of BPAG1 and 2 mRNA by reverse transcription–PCR in BCC (Chopra et al., 1998Chopra A. Maitra B. Korman N.J. Decreased mRNA expression of several basement membrane components in basal cell carcinoma.J Invest Dermatol. 1998; 110: 52-56https://doi.org/10.1046/j.1523-1747.1998.00089.xCrossref PubMed Scopus (20) Google Scholar). The marked differences in gene expression between cultured and in situ keratinocytes complicate the choice of proper controls. Further, these differences are consistent with other data indicating that the response of cultured cells to the hedgehog ligand or to the loss of PTC is blunted (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.Nature Med. 1997; 3: 788-792Crossref PubMed Scopus (230) Google Scholar). Our finding of upregulated PTC1 and HIP in cells with amplified GLI1 (D259MG cells) suggests that the block to the hedgehog response in cultured cells lies upstream of this transcription factor. These findings further validate the concept of the primary role of hedgehog target gene activation in BCC tumorigenesis and are consistent with previously elucidated pathways of hedgehog signaling. Considerably more work, however, will be necessary to elucidate which of these changes described or which other changes yet to be described will be found to be the key to the mechanism by which aberrant hedgehog signaling leads to aberrant BCC cell behavior and which will be found to be epiphenomena. We wish to thank D. Bigner for the D259MG cells and the following individuals for their generous contributions of probes: Luis Diaz and George Giudice (BPAG2), Arthur Mercurio (α6 INTEGRIN), Matthew Scott (BMP2 and 4 and PTC1), Jouni Uitto (BPAG1), and Bert Vogelstein (GL11). We also wish to express our appreciation to Richard Assoian, Roel Nusse, Dennis Roop, Matthew Scott, Aylin Tucker, and Mark Udey for their helpful advice, to Philip LeBoit for review of the BCC histology, and to Morrie Schambelan and Shandi Griffin for assistance with the phosphoimager analysis. Supported by grants AR39959, CA81888 (E.H.E.), AR39448 (S.P.), and NS33642 (A.P.M.) from the National Institutes of Health, the Leukemia Society of America (P.-T.C.), and by generous donations from Patricia Hughes and from the Michael J. Rainen Family Foundation.