A Novel Mechanism for Activation of GLI1 by Nuclear SMO That Escapes Anti-SMO Inhibitors Muhammad M

Published OnlineFirst February 20, 2018; DOI: 10.1158/0008-5472.CAN-17-2897

Cancer Molecular Cell Biology Research

A Novel Mechanism for Activation of GLI1 by Nuclear SMO That Escapes Anti-SMO Inhibitors Muhammad M. Rahman1, Allon Hazan1, Joanne L. Selway2, Dimalee S. Herath1, Catherine A. Harwood1, Muhammad S. Pirzado1, Ravinder Atkar1, David P. Kelsell1, Kenneth J. Linton1, Mike P. Philpott1, and Graham W. Neill1

Abstract

Small-molecule inhibitors of the Hedgehog (HH) pathway nucleolar localization signal [N(o)LS]. Mutational inactivation of receptor Smoothened (SMO) have been effective in treating some the N(o)LS ablated this increase and suppressed GLI1 induction. patients with basal cell carcinoma (BCC), where the HH pathway is Immunohistologic analysis of human and mouse BCC confirmed often activated, but many patients respond poorly. In this study, we evidence of nuclear SMO, although the pattern was heterogeneous report the results of investigations on PTCH1 signaling in the HH between tumors. In PTCH1-silenced cells, >80% of the genes found pathway that suggest why most patients with BCC respond poorly to be differentially expressed were unaffected by SMO inhibitors, to SMO inhibitors. In immortalized human keratinocytes, PTCH1 including the putative BCC driver gene CXCL11. Our results silencing led to the generation of a compact, holoclone-like mor- demonstrate how PTCH1 loss results in aberrant regulation of phology with increased expression of SMO and the downstream SMO-independent mechanisms important for BCC biology and HH pathway transcription factor GLI1. Notably, although siRNA highlights a novel nuclear mechanism of SMO-GLI1 signaling that silencing of SMO in PTCH1-silenced cells was sufficient to suppress is unresponsive to SMO inhibitors. GLI1 activity, this effect was not phenocopied by pharmacologic Significance: This study describes novel noncanonical Hedgehog inhibition of SMO, suggesting the presence of a second undefined signaling, where SMO enters the nucleus to activate GLI1, a mode pathway through which SMO can induce GLI1. Consistent with this that is unaffected by SMO inhibitors, thus prompting re-evaluation possibility, we observed increased nuclear localization of SMO in of current BCC treatment as well as new potential therapies targeting PTCH1-silenced cells as mediated by a putative SMO nuclear/ nuclear SMO. Cancer Res; 78(10); 2577–88. 2018 AACR.

Introduction factors via a pathway requiring the transmembrane protein Smoothened (SMO; ref. 1). Up to 70% to 80% of sporadic Basal cell carcinoma (BCC) is the most common skin cancer, BCCs have loss-of-function mutations in PTCH1, 6% to 21% with several clinical subtypes. BCC commonly manifests as a harbor activating mutations in SMO and GLI1, and GLI2 is nonaggressive and slow-growing tumor histologically catego- rarely mutated (3, 5, 6). rized as nodular or superficial. Micronodular and infiltrative Our knowledge of HH signaling in BCC stems mainly from subtypes are more aggressive and may be locally destructive mousemodels,butthisisconstrainedasPtch1 / mice are if left untreated (1). BCC incidence is associated with exposure þ embryonic lethal. However, Ptch1 / mice develop follicular to ultraviolet radiation, commonly developing in elderly peo- hamartomas with BCCs arising after ultraviolet irradiation (7), ple especially on the head and neck (1). Hedgehog (HH) whereas conditional Ptch1 knockout mice develop BCC-like signaling was first linked to BCC due to loss-of-function muta- lesions with tumors displaying nuclear GLI1 expression, indic- tions in the Ptch1 gene in patients with nevoid BCC syndrome ative of HH pathway activation (8). In addition, overexpression (NBCCS) or Gorlin syndrome, an autosomal dominant disor- of Gli1 leads to BCC-like epidermal tumor formation, and der predisposing sufferers to early onset BCC, among other targeted expression of an active mutant of Gli2 to the basal developmental defects (2–4). Loss of PTCH1 function leads to layer of murine epidermis is sufficient to sustain BCC-like constitutive activation of the GLI1 and GLI2 transcription lesions (9, 10). Regarding SMO, conditional expression of an active SMO mutant (M2) induces BCC-like tumors when activated in the murine basal layer (11). Intriguingly, reports show 1Centre for Cell Biology and Cutaneous Research, Blizard Institute, Barts and that SMO-M2 induced follicular hamartomas and not BCC-like The London School of Medicine and Dentistry, Queen Mary University of London, K5-GLI2 N 2 lesions, concluding that,bycomparisonwiththe D London, United Kingdom. Buckingham Institute for Translational Medicine, mouse model (10), GLI1 and GLI2 expression was insufficient- University of Buckingham, Buckingham, United Kingdom. ly high to initiate the more malignant tumor form (12). To our Note: Supplementary data for this article are available at Cancer Research knowledge, GLI1 and GLI2 expression levels have not been Online (http://cancerres.aacrjournals.org/). compared between SMO-M2 follicular hamartomas and irra- þ Corresponding Authors: Mike P. Philpott, Blizard Institute, London E12AT, diated Ptch1 / or conditional Ptch1 / BCC-like tumors, but UK. Phone: 02078827162; E-mail: [email protected]; and Muhammad although it is difficult to compare different mouse studies, if M. Rahman, [email protected] loss of PTCH1 alone is sufficient to induce BCC-like lesions (8) doi: 10.1158/0008-5472.CAN-17-2897 while ectopic SMO-M2 is not (albeit at low levels; ref. 12), this 2018 American Association for Cancer Research. suggests that SMO-independent mechanisms may contribute to

www.aacrjournals.org 2577

Rahman et al.

carcinogenesis in tumors associated with loss of PTCH1 func- were cloned into pEGFP-C2 (Clontech Laboratories). To generate tion. Indeed, such a hypothesis was proposed to explain why, the EGFP-NLS-SMO construct, the predicted N(o)LS sequence of þ unlike irradiated Ptch1 / mice, SMO-M2 tumors do not arise SMO was cloned into pEGFP-C2. SMO-NLS-F-50 GAGGCAGA- from the hair follicle (13). However, the same study also GATCTCCCCAGA 30 and SMO-NLS-R-50 ATATGGATCCGTGTG- concluded that tumors may not arise in ears or tails of irradi- GAGGAAGAAGGAG 30 oligonucleotides were used. The puri- þ ated Ptch1 / mice because of low SMO expression, and thus, fied SMO N(o)LS DNA was ligated into pEGFP-C2. Site-direct- the canonical signal cannot be transduced. ed mutagenesis using the QuikChange II XL Site-Directed The importance of HH signaling in BCC development is Mutagenesis Kit (Agilent Technologies) was utilized to modify supported by FDA approval of the anti-SMO compounds the N(o)LS of SMO according to the manufacturer's instruc- vismodegib and sonidegib to treat advanced/metastatic BCC. tions. Mutagenesis primers used were as follows: ESMOmNLS- Complete responses have been documented, and the overall F-50 CTTCTTCCGGGCCAGGGCCGCCTGCAGTCGAGATCTGA response rate for locally advanced and metastatic BCC is 30 and ESMOmNLS-R-50 TCAGATCTCGACTGCAGGCGGCCC- around 50% (14). However, this also suggests that in the TGGCCCGGAAGAAGA 30 that mutate K679A and K680A, absence of a drug bioavailability issue, many BCCs are driven respectively (Supplementary Fig. S1B). GLI1 fireflyluciferase by SMO-independent mechanisms or that some tumors are reporter pGL3-6GBS (15) combined with a pCMV-Renilla lucif- heterogeneous such that their progression is not solely driven by erase normalization vector was transfected into keratinocytes SMO-dependent signaling. using FuGENE 6. Cells were harvested 24 hours after transfec- To address these issues and to investigate the HH signaling tion and RNA extracted for qPCR analysis. pathway further in BCC, we analyzed the effects of suppressing the PTCH1 gene in immortalized human keratinocytes. Our results RNA, cDNA extraction, and qPCR analysis show that the majority of transcriptome changes observed upon RNA was extracted from cells using the RNeasy Mini Kit (Qia- PTCH1 suppression are independent of SMO activity. In addition, gen). cDNA was then synthesized using the Superscript VILO we present evidence that the increase of GLI1 transcription in cDNA Synthesis kit (Invitrogen). Note that 100 ng of cDNA per PTCH1 knockdown cells is associated with nuclear SMO function, reaction was used for qPCR analysis using the Rotor-Gene SYBR a mechanism insensitive to pharmacologic inhibition. Nuclear Green PCR Kit (Qiagen). The Rotor-Gene 2000 machine (Qiagen) SMO was also observed in human and, notably, mouse BCCs, was used for the analyses with the GAPDH reference gene. Primer although the pattern of expression was heterogeneous. In sum- sequences for qPCR were: AGR2: F-50 GGGATGGAGAAAATTC- mary, this study identifies a novel mode of SMO-GLI1 signaling as CAGTG 30, R-50 GGGTACAATTCAGTCTTCAG 30; AMOT: F-50 well as revealing the presence of SMO-independent signaling CATGGAGGGCAGGATTAAGA 30, R-50 TCGTCTCGCTTTTCT- mechanisms downstream of PTCH1 that may be relevant to TCCAT 30; CXCL11: F-50 GTGCTACAGTTGTTCAAGGC 30, R-50 BCC biology. CTAGGTTTTTCAGATGCTCT 30; FBN2: F-50 AATGTGGGTCT- CAACCTTCG 30, R-50 CTGTAGCCACCCAGGATGTT 30; GAPDH Materials and Methods F-50 GTGAACCATGAGAAGTATGACA 30, R-50 CATGAGTCCTTC- CACGATACC 30; GFP: F-50 TATATCATGGCCGACAAGCA 30, R-50 Retroviral transduction of immortalized keratinocytes and cell GAACTCCAGCAGGACCATGT 30; GLI1: F-50 GAAGACCTCTC- culture CAGCTTGGA 30, R-50 GGCTGACAGTATAGGCAGAG 30; GLI2: Two immortalized keratinocyte cell lines were used for this F-50 TGGCCGCTTCAGATGACAGATGTTG 30, R-50 CGTTAGC- study: NEB1 and N/Tert cells (obtained in 2009). Cells were CGAATGTCAGCCGTGAAG 30; luciferase: F-50 AGTGCTCAT- cultured in Alpha MEM and 10% (v/v) FBS (Lonza), 2 nmol/L CATCGGGAATC 30, R-50 CATCCAACATTTTCGTGTCG 30; MMP2: L-glutamine, 2% (v/v) penicillin–streptomycin (PAA Laborato- F-50 AGGGCACATCCTATGACAGC 30, R-50 ATTTGTTGCCCAG- ries), and keratinocyte growth medium supplement consisting of GAAAGTG 30; MMP9: F-50 TTGACAGCGACAAGAAGTGG 30, R-50 10 ng/mL EGF, 0.5 mg/mL hydrocortisone, 5 mg/mL insulin, 0 0 TCACGTCGTCCTTATGCAAG 3 ; PTCH1: F-5 ACTCGCCAGAA- 1.8 10 4 mol/L adenine, and 1 10 10 mol/L cholera toxin GATTGGAGA 30, R-50 TCCAATTTCCACTGCCTGTT 30; SMO: F-50 (Sigma). SMO inhibitors KAAD-Cyc and SANT-1 (Merck) were GGCTGCTGAGTGAGAAG 30, R-50 CTGGTTGAAGAAGTCGTA- used at 100-nmol/L and 1-mmol/L concentrations from 24 to 96 GAAG 30; SNAI1: F-50 TTTACCTTCCAGCAGCCCTA 30, R-50 hours. Cells were negative for Mycoplasma, tested using MycoAlert CCCACTGTCCTCATCTGACA 30; THY1: F-50 CCCAGTGAA- (Lonza). Cells were retrovirally transduced with PTCH1 small GATGCAGGTTT 30, R-50 GACAGCCTGAGAGGGTCTTG 30; VIM: hairpin RNAs (shRNA) targeting exon 3 (AAGGTGCTAA- F-50 CCCTCACCTGTGAAGTGGAT 30, R-50 CAACCAGAGGGAG- TGTCCTGACCA) and exon 24 (AAGGAAGGATGTAAAGTGGT). TGAATCC 30. A nontargeting control sequence was used (GCGCGATATATA- GAATACG). The sequences were cloned into the pSUPERIOR. retro.puro vector (OligoEngine). Retrovirally transduced cells Immunocytochemistry and ImageJ staining quantification were then selected with 1-mg/mL puromycin (Sigma) for a week A total of 15,000 cells per well were seeded in 12-well plates on and cultured normally thereafter, from which clonal cell lines 18-mm diameter glass coverslips (VWR International) for 3 days. were derived. Cells were fixed with 4% (w/v) paraformaldehyde and permea- bilized with 0.1% (w/v) Triton X-100 (Sigma). Note that 3% EGFP-SMO/M2, EGFP-NLS-SMO, and GLI1-luciferase reporter (w/v) BSA (Fisher Scientific) was used for blocking and primary vectors antibodies diluted to 1:1,000 in 3% (w/v) BSA. Primary anti- To generate the EGFP-SMO/M2 constructs, SMO-M2 was bodies used for this study were fibrillarin-C13C3 (Cell Signaling excised from the PRK-SMO vector (6). SMO-M2 contains an Technology), GLI1-C18, GLI1-H300, PTCH1-C20, SMO-C17, activating point mutation of W535L (6). Purified DNA products SMO-N19 (Santa Cruz Biotechnology), and SMO-ab72130

2578 Cancer Res; 78(10) May 15, 2018 Cancer Research

GLI1 Is Activated by Nuclear SMO in PTCH Knockout Cells

(Abcam). Fluorescence dye-labeled Alexa 488/568 (Invitrogen) construct targets downstream of the STOP codon at the end of secondary antibodies were diluted to 1:800 in PBS. Cell nuclei exon 23. Ectopic expression of the coding sequence of only the were stained with 0.1 mg/mL of DAPI (Sigma). Coverslips were PTCH1-1B (PTCH1L) isoform suppressed increased GLI1 mounted onto microscope slides using VECTASHIELD mounting mRNA expression by approximately 50% in NEB1-shPTCH1 medium (Vector Laboratories). Images were taken on the Zeiss cells (Fig. 1D); higher suppression was probably hampered LSM 510 META confocal microscope and quantified using by PTCH1-1B appearing to induce an apoptotic response in ImageJ software. some cells, as described in 293T cells (17). Morphologically, shPTCH1 cells formed holoclone-like colonies, characteristic of Immunohistochemistry nodular/micronodular BCC tumor islands with smaller nuclei, SMO protein expression was examined in 29 biopsies of compared with shCON cells also containing more nucleoli þ human BCC, 14 invasive and 15 noninvasive samples. Ptch1 / (Fig. 1E and F). g-irradiated mice tumors were kindly provided by Anna Saran (ENEA). Tissue sections (4 mm) were cut from paraffin blocks and GLI1 is activated by nuclear SMO transferred to microscope slides. Slides were run on the Ventana In addition to the increase in SMO mRNA (Fig. 1B), an increase Discovery platform (Ventana Molecular Discovery Systems, Roche of SMO protein expression was observed by immunocytochem- Diagnostics) with an automated protocol, including deparaffini- istry in shPTCH1 cells compared with shCON using a commercial zation, antibody incubations, DAB application, hematoxylin N-terminal antibody (N-19; Santa Cruz Biotechnology). and bluing reagent counterstaining, and finally slide cleaning. More specifically, SMO displayed cytoplasmic and nuclear local- Following dehydration in a series of solvents, the slides were ization in shCON cells, with a comparable but more intense mounted with coverslips. fluorescent pattern in shPTCH1 cells (Fig. 2A). A similar pattern of fluorescence was observed using another antibody targeting the Gene expression microarray C-terminus (C-17; Santa Cruz Biotechnology), which provides Gene expression microarray profiling was performed on further evidence for the presence of nuclear SMO in human NEB1-shCON and NEB1-shPTCH1 cells lines using Human keratinocytes (Fig. 2A). SMO N-19 antibody specificity was con- Gene 1.1 ST Array Strip (Affymetrix). Each sample was repeated firmed by a reduction of the fluorescent signal in cells transfected in triplicate, and the data normalized using the MetaCore with SMO siRNA (Fig. 4D). pathway analysis software (GeneGo). Three Affymetrix To our knowledge, nuclear SMO has not been described chips were used for each sample, and the raw data were before; however, in contrast to the Drosophila homolog, processed using the multiarray average (RMA) method. Prob- human and mouse SMO contain a C-terminal region pre- abilities of gene expression between the experimental and dicted to harbor a monopartite nuclear/nucleolar localization control group were generated using the Wilcoxon signed-rank signal, or N(o)LS (Supplementary Fig. S1B). Tagging EGFP test. The Tukey biweight method was used to obtain log ratios withtheputativeN(o)LSsequencetargetedtheproteintothe that were then antilogged to generate fold-change values com- nucleolus based on fluorescent colocalization with fibrillarin. paring genes between samples. Genes with a P value at or below A full-length EGFP–SMO fusion protein also localized to the 5% and a fold change greater than 2foldareconsideredas nucleus and occasionally to the nucleolus. Mutational inacti- differentially expressed genes (DEG). DEGs were analyzed in vation of residues within the N(o)LS region mNLS2 (Supple- MetaCore using its GeneGo database of cellular pathways and mentary Fig. S1B) localized EGFP-SMO exclusively to the process networks. cytoplasm (Fig. 2B) and impaired its ability to increase GLI1 mRNA levels (Fig. 2C). In addition to the full-length SMO above, we also expressed Results the constitutively active SMO-M2 mutant which showed PTCH1 gene suppression leads to increased GLI1 activity much higher levels of GLI1 expression (Supplementary To understand the role of PTCH1 in keratinocyte biology Fig. S1C). Mutation of the NLS2 in SMO-M2 also significantly and BCC formation, we used shRNA to suppress PTCH1. downregulated GLI1. Interestingly, mutation at another site, Human skin keratinocyte cell lines (NEB1 and N/Tert) were mNLS1, within the N(o)LS (Supplementary Fig. S1B) did retrovirally transduced with PTCH1 shRNA targeting exons 3 or not influence the ability of EGFP-SMO or EGFP-SMO-M2 to 24 and normalized against a nontargeting control vector activate GLI1 expression (thus providing an internal negative (shCON). Clonal cell lines were generated from heterogeneous control). The fact that GLI1 levels were reduced with mNLS2 populations of the PTCH1 knockdown cells and validated by in both shCON and shPTCH1 cells and for both EGFP- qPCR; exon 24 clone 1 (C1) of NEB1 cells displayed the SMO and EGFP-SMO/M2 provides very strong evidence for a strongest level of PTCH1 suppression and the highest level of specific role of nuclear SMO in GLI regulation (Supplementary GLI1 mRNA expression (Supplementary Fig. S1A), which was Fig. S1C). confirmed at the protein level as well as by increased GLI reporter activity (Fig. 1A and C). PTCH1 protein expression was Nuclear SMO expression is observed in mammalian skin and predominantly nuclear in shCON cells, which may represent BCCs the C-terminal fraction previously described (16). In addition SMO protein expression was examined in a panel of BCCs to to the upregulation of GLI1, suppression of PTCH1 reduced determine its subcellular localization: The SMO-N19 antibody SHH and IHH mRNA levels, whereas SMO was increased, but was employed, having confirmed its specificity for SMO by there was no significant effect upon GLI2 expression (Fig. 1B). immunocytochemistry (Figs. 2A and 4D). In human BCCs, the To confirm that increased GLI1 resulted directly from reduced staining pattern was predominantly cytoplasmic and membra- PTCH1 activity, we made use of the fact that the shPTCH1 nous with some evidence of nuclear expression (Fig. 3A). In

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2579

Rahman et al.

shPTCH1 shCON A NEB1-shCON NEB1-shPTCH1 B 7 ** 6 5 4

- DAPI 3 * 2

mRNA fold change * 1 *

PTCH1-C20 0 GLI1 GLI2 SMO IHH SHH

C EGFP Luciferase shCON

4 *

1 mRNA fold change

GLI1-H300 - DAPI 0 shPTCH1 + 6xGLI-BS-Firefly D Exon 23 Exon 23 Exon 23

STOP shPTCH1

PTCH1 GLI1 shCON E NEB1-shCON NEB1-shPTCH1 8 7 ** ** 6 5 ** 4 * 3 2 mRNA fold change * * 1 0 shCON + PTCH1B shPTCH1 shPTCH1 + PTCH1B

F NEB1-shCON NEB1-shPTCH1 8 **

2 Fibrillarin - DAPI Number Of nucleoli 0 shCON shPTCH1

Figure 1. PTCH1 gene suppression leads to increased GLI1 activity. A, NEB1-shPTCH1 cells show reduced PTCH1 and increased GLI1 protein expression, and both PTCH1-C20 and GLI1-H300 antibodies display a nuclear staining pattern. B, Suppression of PTCH1 reduces mRNA expression of SHH and IHH ligands with increased GLI1 and SMO but no significant change in GLI2 levels. C, NEB1-shPTCH1 transfected with a 6x GLI binding site firefly luciferase reporter construct (6xGLI-BS-firefly) and EGFP as a transfection control confirm increased GLI1 activity based on a 3-fold increase in luciferase mRNA expression. D, Ectopic expression of the PTCH1B isoform (shRNA targets exon 24 beyond the STOP codon in exon 23) suppressed the increased GLI1 mRNA expression in NEB1-shPTCH1 cells. E, NEB1-shPTCH1 cells form densely packed colonies similar in appearance to nodular or micronodular BCC tumors. F, Due to tight packing of NEB1-shPTCH1 cells, the size of the nuclei appears smaller, with a greater average number of nucleoli present (average taken from 50 cells). , P 0.05; , P 0.01 as calculated by the Student t test; error bars, SD; n ¼ 6.

2580 Cancer Res; 78(10) May 15, 2018 Cancer Research

GLI1 Is Activated by Nuclear SMO in PTCH Knockout Cells

A NEB1-shCON NEB1-shPTCH1 NEB1-shCON NEB1-shPTCH1 C17 - SMO SMO-N19

24.5 41.5 30.5 43.9

75.5 58.5 69.5 56.1

% Nuclear staining % Cytoplasmic staining

B NEB1-EGFP EGFP-NLS-SMO EGFP-SMO EGFP-SMO-mNLS DAPI - Fibrillarin – GFP

NLS SMO NLS SMO NLS

EGFP EGFP EGFP EGFP

SMO N(o)LS SMO NLS Site directed mutagenesis

C shCON shPTCH1 EGFP-control

4 ** 2 GLI1 mRNA fold change * 0 EGFP-SMO mNLS

Figure 2. GLI1 is activated by nuclear SMO. A, Staining using two SMO antibodies (N- and C-terminal protein) revealed distinct nuclear localization of SMO in shPTCH1 cells, with a higher percentage of nuclear SMO in shPTCH1 cells (calculated by ImageJ; n ¼ 20). B, NEB1 cells transfected with EFP containing the predicted N(o)LS sequence from SMO direct GFP to the nucleolus (Supplementary Fig. S1C). EGFP–SMO fusion protein is expressed in both the nucleus and nucleolus, and by mutating the N(o)LS, nucleolar GFP expression is lost and the protein is retained in the cytoplasm. C, shCON cells and, to a greater extent, shPTCH1 cells transfected with EGFP-SMO show upregulation of GLI1 mRNA expression that is then reduced to below basal levels upon mutation of N(o)LS sequences. , P 0.05; , P 0.01 as calculated by the Student t test; error bars, SD; n ¼ 6.

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2581

Rahman et al.

A SMO-N19–Morphoeic BCC SMO-N19–Nodular BCC x200 x400

Sample 1 Sample 2 Sample 3 Sample 4

B SMO-N19–γ-irradiated Ptch1+/- mouse model

Morphoeic Nodular x400

Sample 1 Sample 2 Sample 1 Sample 2

Figure 3. Nuclear SMO expression is observed in mammalian skin and BCCs. A, SMO-N19 staining of human morphoeic BCCs shows cytoplasmic expression and variable staining intensities, whereas nodular BCCs also express cytoplasmic SMO; certain tumor nodules also express nuclear SMO protein (arrows). þ B, Mouse BCC tumors derived from Ptch1 / g-irradiated mice show distinct nuclear SMO expression, although there is heterogeneity between tumors.

þ contrast, mouse BCC tumors derived from Ptch1 / g-irradiated Again, elevated GLI1 expression was suppressed by neither mice showed clear evidence of nuclear SMO, although hetero- KAAD-Cyc nor SANT-1, indicating that the increase of GLI1 geneous staining was also observed (Fig. 3B). observed upon PTCH1 suppression in NEB1 and N/Tert human keratinocytes is insensitive to SMO pharmacologic inhibition GLI1 expression is unresponsive to anti-SMO inhibition in (Supplementary Fig. S1D and S1E). To investigate the SMO– shPTCH1 cells GLI1 signaling axis in more detail, shPTCH1 cells were treated We assessed the canonical PTCH–SMO–GLI1 signaling axis with siRNA-targeting SMO (siSMO). Intriguingly, GLI1 protein in shPTCH1 cells by employing SMO pharmacologic inhibi- expression was reduced in both shCON and shPTCH1 cells, tors. Twenty-four hours after exposure, GLI1 protein and mRNA suggesting that SMO is required for GLI1 signaling (Fig. 4D). expression were strongly suppressed in shCON cells, whereas When comparing the effects of KAAD-Cyc and siSMO in no changes were observed in shPTCH1 cells treated with 100 shPTCH1 cells, GLI1 mRNA expression was reduced by siSMO, nmol/L KAAD-Cyc (Fig. 4A and B). The same lack of response but not KAAD-Cyc, whereas GLI2 expression was reduced by was observed in shPTCH1 cells treated with 100 nmol/L SANT- both methods (Fig. 4E). Although GLI2 expression was not 1 (Fig. 4C); in addition, no change in GLI1 expression was elevated in shPTCH1 cells, GLI2 levels were suppressed by observed upon prolonged exposure to either inhibitor (up to KAAD-Cyc in both shCON and shPTCH1 cell lines, thus sup- 96 hours). To validate these results, other NEB1 and N/Tert- porting the concept of a canonical SMO–GLI2 signaling axis shPTCH1 clones were examined, targeting exons 3 and 24. independent of PTCH1 control.

2582 Cancer Res; 78(10) May 15, 2018 Cancer Research

GLI1 Is Activated by Nuclear SMO in PTCH Knockout Cells

A NEB1-shCON NEB1-shPTCH1 B shCON shPTCH1

8 10.2% 7 6

0 h 5 4 3 2 56.4% 1 * * GLI1 mRNA fold change 0 0 h 4 h 8 h 24 h 24 h KAAD-Cyc C GLI1-H300 shCON KAAD-Cyc shPTCH1 KAAD-Cyc shCON SANT-1 shPTCH1 SANT-1 3 D siCON siSMO

SMO-N19 1

0 Relative GLI1 staining density 0 h 24 h 48 h 72 h 96 h NEB1-shCON GLI1-H300 E GLI1 GLI2 SMO shCON

8 7 6 5 4 SMO-N19 * 3 * 2

mRNA fold change * 1 * ** * * ** ** ** 0 NEB1-shPTCH1 shCON shCON shPTCH1 shPTCH1 shPTCH1 siSMO KAAD-Cyc siSMO KAAD-Cyc GLI1-H300

Figure 4. PTCH1-suppressed cells are unresponsive to anti-SMO inhibitors. Twenty-four hours of 100 nmol/L KAAD-Cyc is able to suppress GLI1 protein (A)and mRNA expression (B) in NEB1-shCON cells; however, NEB1-shPTCH1 cells show no GLI1 reduction. C, The same lack of response in NEB1-shPTCH1 cells was seen with 100 nmol/L SANT-1 and also with extended treatment up to 96 hours, where GLI1 protein remained constant. D, Although shPTCH1 cells did not respond to SMO inhibitors, SMO siRNA (siSMO) did suppress GLI1 protein expression compared with control siRNA (siCON), suggesting that SMO is required for GLI1 signaling. E, qPCR conﬁrms that GLI1 expression is reduced by siSMO, but not KAAD-Cyc in shPTCH1 cells, whereas GLI2 expression is suppressed by both siSMO and KAAD-Cyc. , P 0.05; , P 0.01 as calculated by the Student t test; error bars, SD; n ¼ 6.

The global effects of PTCH1 suppression are largely insensitive with shCON cells, expression levels of 213 transcripts were to SMO inhibition altered 2-fold (137 increased and 76 decreased, P < 0.01) in To characterize the global effects of PTCH1 suppression, shPTCH1 cells (Fig. 5A). However, of the 76 downregulated shCON and shPTCH1 cells were subject to cDNA microarray genes, the expression of 41 genes was further reduced by KAAD- proﬁling.Furthermore,todetermineiftheeffectsofPTCH1 Cyc treatment, whereas the expression of 47 of the 137 upre- suppression are sensitive to pharmacologic SMO inhibition, gulated genes was further increased (Fig. 5B). Exposure shPTCH1 cells were additionally exposed to 100 nmol/L to KAAD-Cyc failed to normalize the expression of the majority KAAD-Cyc for 24 hours prior to RNA harvesting. In comparison of transcripts to basal levels, with the inhibitor often

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2583

Rahman et al.

A B

213 DEGs ≥2-fold change

137

76 Downregulated 137 Upregulated

35 47 41 90

41 Downregulated by 90 Upregulated by KAAD-Cyc KAAD-Cyc

C shPTCH1 KAAD-Cyc shCON

25 45 ** ** * 40 ** * 20 35 * * 30 15 * * 25 * 20 10 * 15 * 5 10 mRNA fold change ** ** 5 0 0 AGR2 AMOT FBN2 MMP2 SNAI1 VIM CXCL11 MMP9 THY1

Figure 5. The global effects of PTCH1 suppression are largely insensitive to SMO inhibition. A, List of top and bottom DEGs (2 fold change, P > 0.01) comparing NEB1-shPTCH1 and NEB1-shPTCH1 treated with 100 nmol/L KAAD-Cyc with NEB1-shCON cells reveals various genes reported to be involved in BCC biology are differentially expressed. B, Diagram of identiﬁed DEGs and the proportion of upregulated and downregulated genes further altered by KAAD-Cyc treatment. C, qPCR on NEB1-shPTCH1 cells was performed to validate the results from the microarray, with a number of cancer-related genes all strongly upregulated in NEB1-shPTCH1 and also in cells treated with 100 nmol/L KAAD-Cyc. , P 0.05; , P 0.01 as calculated by the Student t test; error bars, SD; n ¼ 6.

exacerbating the effects of PTCH1 suppression. Only two upre- (#10.2-fold), AMOT (#3.7-fold), and AGR2 (#3.3-fold) have gulated DEGs, CXCL10 and SNORA38B, were reduced by 50% been associated with tumor biology. Excluding MMP7 and upon KAAD-Cyc treatment. CXL10 (not detectable with two primer sets each), the differ- Figure 5A contains transcripts associated with tumor biology, ential expression of the transcripts listed above was conﬁrmed including the cancer stem cell marker THY1 ("6.3-fold), epi- by qPCR in shPTCH1 cells (Fig. 5C). However, elevated SNAI1 thelial–mesenchymal transition (EMT) markers VIM ("3.6- expression was partially suppressed by KAAD-Cyc, whereas fold) and SNAI1 ("2.4-fold), and matrix metalloproteases there was a potent (albeit potentially undesirable) increase of MMP7 ("9.5-fold), MMP2 ("4.1-fold), and MMP9 ("2.2-fold). MMP9. Finally, a similar but reduced pattern of differential In addition, the chemokines CXCL10 ("8.2-fold) and CXCL11 expression was observed for these transcripts in the shPTCH1- ("4.9-fold) proved to be of interest, as both are known to be ex3cloneaswellasinoneN/Tert-shPTCH1 clonal cell line elevated in BCC (18). Of the suppressed transcripts, FBN2 (Supplementary Fig. S1F).

2584 Cancer Res; 78(10) May 15, 2018 Cancer Research

GLI1 Is Activated by Nuclear SMO in PTCH Knockout Cells

Discussion ular BCC, which was predominantly cytoplasmic but with clear The purpose of this study was to identify novel targets areas of nuclear staining (Fig. 3A and B). SMO expression was downstream of PTCH1 relevant to BCC biology and regulated reduced in morphoeic tumors, and although this may appear by SMO-dependent and SMO-independent mechanisms. paradoxical, it correlates with data showing that GLI1 expres- PTCH1 was stably suppressed in immortalized human kerati- sion is reduced in morphoeic BCC (23). Nuclear SMO was most nocytes, and clonal cell lines were generated. The fact that prevalent in mouse BCC-like nodular tumors, which may reflect clonal cell lines with strong PTCH1 suppression adopt a holo- the more homogenous nature of the system, but, as for human clone-like morphology characteristic of BCC nodular islands tumors, expression was also reduced in morphoeic-like BCC supports the validity of the model. We have previously shown (Fig. 3B). that retroviral GLI1 expression promotes tight N/Tert colony BCC tumors that become resistant to vismodegib most formation, and endogenous GLI1 may help promote this commonly acquire point mutations affecting their ability to phenotype in the shPTCH1 model (19). However, the level of bind SMO (24). How and if nuclear SMO expression influences ectopic GLI1 is considerably higher than that induced upon drug effectiveness remains to be determined, but the overall PTCH1 suppression, and it is likely that additional factors response to anti-SMO compounds in clinical trials was approx- regulate colony formation in shPTCH1 cells. Interestingly, imately 50% (14). As such, many BCCs are classified as SMO while validating the shPTCH1 model with regard to canonical independent because they are dependent upon other mechan- PTCH/SMO/GLI signaling, some observations do not correlate isms that correlate with the lack of SMO expression observed in with current dogma, i.e., PTCH1 suppression led to an increase sometumors(Fig.3B).Alternatively,sometumorsmaystillbe of GLI1, but not GLI2, and an increase of GLI1 correlated with SMO dependent, but an anti-SMO compound lacks efficacy increased nuclear SMO, which was unresponsive to anti-SMO because it is influenced by SMO subcellular localization. pharmacologic agents. Indeed, it has been postulated that more selective targeting of With regard to GLI2, it has been shown to activate GLI1, but subcellular compartments may improve G-protein–coupled using an artificial active mutant and not the wild-type protein receptor (GPCR) drug selectivity (25). (20). GLI2 mouse models of BCC also employed the active SMO is a seven-pass transmembrane protein closely related to mutant (21), and although elevated GLI2 levels have been the Frizzled family of GPCR proteins. Drosophila and vertebrate observed in human tumors, the role of the wild-type protein SMO share considerable homology, although the former has a remains unclear. GLI2 was not increased in shPTCH1 cells longer C-terminal tail. The C-terminal has been studied compre- (Fig. 1B), which indicates that GLI2 expression is PTCH indepen- hensively, particularly the region that is phosphorylated, and dent. Despite this, GLI2 was suppressed upon KAAD-Cyc treat- determines the extent of pathway activation (26). However, ment, revealing that it is controlled by SMO (Fig. 4E). It is also unlike human SMO, the Drosophila homolog does not contain possible that PTCH1 suppression permits GLI2 nuclear localiza- a putative NLS (homology with dSMO extends to residue 632 for tion and subsequent activation of GLI1, but this does not correlate human SMO), and although such a region has not been described with the fact that GLI2 was suppressed by anti-SMO inhibitors. for SMO, NLS domains have been recognized in other GPCRs, Another consideration is the nuclear localization of PTCH1 in including the Bradykinin and Apelin receptors (25). Nuclear NEB1-shCON cells (Fig. 1A), which was detected using a C- translocation of GPCRs is mediated by importins and can be terminal antibody. PTCH1 can be cleaved with the C-terminal constitutive or agonist induced; for example, Frizzled2 is cleaved fragment (PTCH1-C) residing in the nucleus to negatively regulate upon Wingless stimulation, with the C-terminus entering the GLI1 (16). Based upon our immunohistochemistry data, it is nucleus to regulate postsynaptic neuron development in Drosoph- likely that suppression of PTCH1-C permits the increase of ila (27). Future work will determine if SMO is similarly cleaved, endogenous GLI1 (Fig. 1A–D). Irrespective of the role of GLI2 but the fact that nuclear SMO is detected with an N-terminal or PTCH1, SMO control of GLI1 was confirmed by a reduction of antibody suggests that this is not the case and not all GPCRs are GLI1 in the presence of SMO siRNA in both shCON and shPTCH1 cleaved before entering the nucleus. Other proteins that regulate cells (Fig. 4D and E). By contrast, anti-SMO inhibitors only GPCR signaling in the nucleus and harbor NLS sequences include suppressed GLI1 levels in shCON cells. It is unlikely that failure b-arrestins and the GPCR kinases (28, 29). Therefore, the iden- to suppress GLI1 was due to the increase of SMO in shPTCH1 cells, tification of an NLS in SMO as a GPCR family member is not as GLI1 levels remained elevated in the presence of high drug without precedent and represents a mode of signaling already concentrations, and therefore, we hypothesize that nuclear SMO established for these proteins. regulates GLI1 through a mechanism that is unresponsive to anti- With regard to the NLS sequences, the SMO region predicted to SMO inhibition. be an N(o)LS targets EGFP to the nucleolus (Fig. 2B). Indeed, Indeed, the most intriguing aspect of the shPTCH1 model although NoLS (nucleolar) are normally distinct to NLS (nuclear), concerns the localization of SMO. SMO expression for both they may occur in the same region, which can make it difficult RNA and protein was increased. The reliability of anti-SMO to delineate the role of specific subdomains. Studies suggest antibodies has been much debated in the HH field; however, a that shared motifs create genetic stability and that a combined similar expression pattern was observed with three antibodies N(o)LS region may facilitate active nuclear import (30). We (SMO-N19, C-17, and ab72130; Supplementary Fig. S1G), have not functionally delineated the SMO N(o)LS region but which was suppressed upon SMO siRNA treatment (Fig. 4D). determined that it is functional and that nuclear import is We are unaware of any studies that have tested anti-SMO important for GLI1 regulation (Fig. 3C). Whether SMO nucle- antibody specificity using siRNA, but nuclear SMO expression olar localization regulates GLI1 requires further investigation, was recently described in cancers of unknown primary origin, but to our knowledge, nucleolar function has not been and this significantly correlated with nuclear GLI1 and nuclear described for other GPCRs. The nucleolus is emerging as a b-catenin (22). Strong SMO expression was observed in nod- target for cancer therapy and is regulated by oncogenic

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2585

Rahman et al.

signaling pathways, including RAS/ERK and PI3K/AKT/mTOR, target of GLI1 in the murine epidermis (41). Although we found which influence GLI activity (31). Nucleoli are often more no evidence that SNAI1 or VIM mRNAs are direct GLI1 targets in abundant in tumors compared with normal cells, and this has human keratinocytes, both proteins were more highly expressed been observed in BCC (32). Interestingly, PTCH1 suppression in GLI keratinocytes that survived genotoxic insult (19, 42). results in an increase in the number of nucleoli (Fig. 1F). Indeed, VIM expression is associated with a more aggressive Although nucleoli numbers decrease during keratinocyte myofibroblast phenotype in BCC (43). terminal differentiation, an increase in number is seen in We have not examined the migratory/invasive potential of fibroblasts of patients with X-ray–treated Gorlin syndrome shPTCH1 cells in vitro, but although BCCs may be locally invasive, (33, 34). they are rarely metastatic, and this correlates with the holoclone- Regarding a nuclear role for GPCRs, control of gene regula- like colonies observed in culture (Fig. 1E). How PTCH1 regulates tion has been reported for various proteins that may represent cell–cell adhesion warrants investigation, but AMOT (angiomo- signaling from the nuclear membrane or more directly; for tin) positively regulates cell migration, and increased expression example,itwasshownthatF2rl1formspartofatranscriptional has been described in metastatic breast cancer (44). In addition, complex with Sp1 that modulates Vegfa expression (35). Delin- AGR2 is associated with tumor progression and metastasis (par- eation of the N(o)LS region may help determine if SMO ticularly hormone-dependent cancers), and its suppression localizes to the nuclear membrane, and regarding DNA bind- impairs the migration of non–small cell lung cancer cells (45). ing, it is intriguing that SMO has a putative leucine zipper As such, AMOT and AGR2 suppression may represent a negative region from residues 405 to 426 that was not previously feedback loop to limit tumorigenesis and the spread of trans- predicted. Microarray data highlight the downregulation of formed cells. It is likely that exposure to ultraviolet irradiation or zinc-finger proteins commonly localized to the nucleus/nucle- other form of genetic insult will be required to transform olus, with more genes downregulated upon KAAD-Cyc treat- shPTCH1 cells, and the complexity by which BCCs arise and ment in shPTCH1 cells (Supplementary Table S1). In particular, progress is highlighted by the extent of genomic aberrations ZNF750 is known to be involved in psoriasis, squamous cell identified in these common tumors (46). carcinoma, and esophageal, lung, and cervical cancers (36). The Finally, FBN2 was the most potently suppressed gene in role that SMO plays in promoting nucleolar changes requires shPTCH1 cells. FBN2 is a glycoprotein that forms part of con- further investigation as well as the consequences of reduced nective tissue microfibrils in the extracellular matrix, often with zinc-fingerproteinexpressioninshPTCH1cells. low expression in tumors (47). How reduced FBN2 expression We also aimed to identify novel targets downstream of influences tumor biology is unclear, but it was hypothesized that, PTCH1 that may be important for BCC biology, particularly as for FBN1, this may relate to increased TGFb signaling (48). those that are SMO independent. Microarray analysis identified Indeed, it has been previously reported that integrin-mediated 213 transcripts that were altered 2-fold in shPTCH1 cells TGFb activation modulates the stromal microenvironment in versus the control (137 increased and 76 decreased). Of these, morphoeic BCC (23, 49). less than 20% returned to basal levels in the presence of KAAD- Since the discovery of PTCH1 in BCC (2–5), several models Cyc, revealing that the majority of DEGs are SMO independent have been created attempting to identify how HH signaling in shPTCH1 cells, or that if they are SMO dependent, their contributes to BCC tumorigenesis. To our knowledge, this is the mechanism of control is unresponsive to conventional anti- first in vitro model to sustain PTCH1 suppression in immortalized SMO inhibition. human keratinocytes. The model enables us to show that PTCH1 CXCL11 is reported to increase proliferation of and enrich for regulates SMO expression and that SMO, as for other GPCRs, BCC-derived epithelial cells as well as providing immunopro- localizes to the nucleus, where it regulates GLI1 expression tection by increasing indoleamine 2,3-dioxygenase (IDO) through a mechanism that is the subject of ongoing research. expression through its receptor, CXCR3, which is itself impli- Whether or not nuclear SMO influences primary or acquired cated in tumor biology (18). MMPs are associated with tumor resistance to anti-SMO compounds warrants further investiga- invasion and metastasis by degradation of the extracellular tion, but these data suggest that drugs should be tested for efficacy matrix. MMP2 and MMP9 degrade collagen IV, and the levels against nuclear SMO. In addition, this model complements other of MMP2 and MMP9 mRNA are significantly higher in nodular studies implying that PTCH1 regulates pathways independently and infiltrative BCCs compared with normal adjacent tissue of SMO (50), which may involve transmembrane as well as (37). In addition, MMP9 expression was significantly higher in nuclear control. Future studies aim to further characterize these infiltrative compared with nodular tumors. MMP7 expression SMO-independent signaling pathways to help realize new targets has also been described in BCC for aggressive and recurrent for treating BCC. forms, although we could not confirm the microarray result by qPCR (38). Disclosure of Potential Conflicts of Interest THY1 (CD90) is a GPI-anchored cell-surface protein expressed G.W. Neill is Medical Science Liaison Team Lead UK and Ireland at Sanofi in various tissues and cancers. Although it is a putative cancer Genzyme. No potential conflicts of interest were disclosed by the other stem cell marker associated with multiple oncogenic processes, authors. including proliferation and metastasis, there is evidence support- ing its role as a tumor suppressor (39); therefore, due to the Authors' Contributions cell-specific role of CD90, it is difficult to speculate about its role Conception and design: M.M. Rahman, M.P. Philpott, G.W. Neill in BCC. Development of methodology: M.M. Rahman, A. Hazan, C.A. Harwood, D.P. Kelsell, K.J. Linton, G.W. Neill Increased expression of SNAI1 and VIM was of interest, as both Acquisition of data (provided animals, acquired and managed patients, are associated with the cancer stem cell phenotype including EMT provided facilities, etc.): M.M. Rahman, J.L. Selway, D.S. Herath, C.A. Harwood, (40). SNAI1 mRNA has been described in BCC and shown to be a M.S. Pirzado, R. Atkar, G.W. Neill

2586 Cancer Res; 78(10) May 15, 2018 Cancer Research

GLI1 Is Activated by Nuclear SMO in PTCH Knockout Cells

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, Cichon for the shRNA control sequences. The authors also thank the computational analysis): M.M. Rahman, A. Hazan, J.L. Selway, M.S. Pirzado, Dr. Hadwen Trust (DHT) for Humane Research for additional funding to K.J. Linton, M.P. Philpott, G.W. Neill develop human models for human disease and conﬁrm that no funding Writing, review, and/or revision of the manuscript: M.M. Rahman, fromtheDHTwasusedforanyanimalresearchinthisarticle. D.P. Kelsell, M.P. Philpott, G.W. Neill Administrative, technical, or material support (i.e., reporting or organizing The costs of publication of this article were defrayed in part by the data, constructing databases): M.M. Rahman, D.S. Herath, M.P. Philpott payment of page charges. This article must therefore be hereby marked Study supervision: M.P. Philpott, G.W. Neill advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Acknowledgments The authors gratefully acknowledge the funding from the British Skin Foundation. The authors also thank Anna Saran (ENEA, Rome, Italy) for Received September 20, 2017; revised December 10, 2017; accepted February providing mouse BCC tissue for immunohistochemistry and Dr. Monika 16, 2018; published ﬁrst February 20, 2018.

References 1. Epstein EH. Basal cell carcinomas: attack of the Hedgehog. Nat Rev Cancer 19. Neill GW, Harrison WJ, Ikram MS, Williams TD, Bianchi LS, Nadendla SK, 2008;8:743–54. et al. GLI1 repression of ERK activity correlates with colony formation and 2. Gailani MR, Bale SJ, Leffell DJ, DiGiovanna JJ, Peck GL, Poliak S, et al. impaired migration in human epidermal keratinocytes. Carcinogenesis Developmental defects in Gorlin syndrome related to a putative tumor 2008;29:738–46. suppressor gene on chromosome 9. Cell 1992;69:111–7. 20. Regl G, Neill GW, Eichberger T, Kasper M, Ikram MS, Koller J, et al. Human 3. Hahn H, Wicking C, Zaphiropoulous PG, Gailani MR, Shanley S, Chi- GLI2 and GLI1 are part of a positive feedback mechanism in basal cell dambaram A, et al. Mutations of the human homolog of drosophila carcinoma. Oncogene 2002;21:5529–39. patched in the nevoid basal cell carcinoma syndrome. Cell 1996;85: 21. Grachtchouk M, Pero J, Yang SH, Ermilov AN, Michael LE, Wang A, et al. 841–51. Basal cell carcinomas in mice arise from hair follicle stem cells and 4. Gorlin RJ. Nevoid basal cell carcinoma syndrome. Dermatol Clin 1995; multiple epithelial progenitor populations. J Clin Invest 2011;121: 13:113–25. 1768–81. 5. Johnson RL, Rothman AL, Xie J, Goodrich LV, Bare JW, Bonifas JM, et al. 22. Fotopoulos G, Gousia A, Bareta E, Koumpis E, Chrisafi S, Bobos M, et al. Human homolog of patched, a candidate gene for the basal cell nevus Prognostic significance of WNT and hedgehog pathway activation markers syndrome. Science 1996;272:1668–71. in cancer of unknown primary. Eur J Clin Invest 2015;45:1145–52. 6. Xie J, Munroe M, Luoh SM, Ryan A, Gu Q, Zhang C, et al. Activating 23. Moutasim KA, Mellows T, Mellone M, Lopez MA, Tod J, Kiely PC, et al. Smoothened mutations in sporadic basal-cell carcinoma. Nature 1998; Suppression of Hedgehog signalling promotes pro-tumourigenic integrin 391:90–2. expression and function. J Pathol 2014;233:196–208. 7. Mancuso M, Pazzaglia S, Tanori M, Hahn H, Merola P, Rebessi S, et al. Basal 24. Atwood SX, Sarin KY, Whitson RJ, Li JR, Kim G, Rezaee M, et al. Smooth- cell carcinoma and its development: insights from radiation-induced ened variants explain the majority of drug resistance in basal cell carcino- tumors in Ptch1-deficient mice. Cancer Res 2004;64:934–41. ma. Cancer Cell 2015;27:342–53. 8. Adolphe C, Hetherington R, Ellis T, Wainwright B. Patched1 functions as 25. Joyal JS, Bhosle VK, Chemtob S. Subcellular G-protein coupled receptor a gatekeeper by promoting cell cycle progression. Cancer Res 2006;66: signaling hints at greater therapeutic selectivity. Expert Opin Ther Targets 2081–8. 2015;19:717–21. 9. Nilsson M, Unden AB, Krause D, Malmqwist U, Raza K, Zaphiropoulos PG, 26. Arensdorf AM, Marada S, Ogden SK. Smoothened regulation: a tale of two et al. Induction of basal cell carcinomas and trichoepitheliomas in mice signals. Trends Pharmacol Sci 2016;37:62–72. overexpressing Gli-1. Proc Natl Acad Sci U S A 2000;97:3438–43. 27. Mosca TJ, Schwarz TL. The nuclear import of Frizzled2-C by Importins- 10. Grachtchouk M, Mo R, Yu S, Zhang X, Sasaki H, Hui CC, et al. Basal cell beta11 and alpha2 promotes postsynaptic development. Nat Neurosci carcinomas in mice overexpressing Gli2 in skin. Nat Genet 2000; 2010;13:935–43. 24:216–7. 28. Hoeppner CZ, Cheng N, Ye RD. Identification of a nuclear localization 11. Youssef KK, Van Keymeulen A, Lapouge G, Beck B, Michaux C, Achouri Y, sequence in b-arrestin-1 and its functional implications. J Biol Chem 2012; et al. Identification of the cell lineage at the origin of basal cell carcinoma. 287:8932–43. Nat Cell Biol 2010;12:299–305. 29. Johnson LR, Robinson JD, Lester KN, Pitcher JA. Distinct structural features 12. Grachtchouk V, Grachtchouk M, Lowe L, Johnson T, Wei L, Wang A, et al. of G protein-coupled receptor kinase 5 (GRK5) regulate its nuclear local- The magnitude of Hedgehog signaling activity defines skin tumor pheno- ization and DNA-binding ability. PLoS One 2013;8:e62508. type. EMBO J 2003;22:2741–51. 30. Emmot E, Hiscox AH. Nucleolar targeting: the hub of the matter. EMBO 13. Wang GY, Wang J, Mancianti ML, Epstein EH Jr. Basal cell carcinomas Rep 2009;10:231–8. arise from hair follicle stem cells in Ptch1(þ/) mice. Cancer Cell 2011; 31. Quin JE, Devlin JR, Cameron D, Hannan KM, Pearson RB, Hannan RD. 19:114–24. Targeting the nucleolus for cancer intervention. Biochim Biophys Acta 14. Yin VT, Esmaeli B. Targeting the hedgehog pathway for locally advanced 2014;1842:802–16. and metastatic basal cell carcinoma. Curr Pharm Des 2017;23:655–9 32. Egan MJ, Crocker J. Nucleolar organizer regions in cutaneous tumours. J 15.NeillGW,GhaliLR,GreenJL,IkramMS,PhilpottMP,QuinnAG. Pathol 1988;154:247–53. Loss of protein kinase C-alpha expression may enhance the tumor- 33. Gdula MR, Poterlowicz K, Mardaryev AN, Sharov AA, Peng Y, Fessing MY, igenic potential of Gli1 in basal cell carcinoma. Cancer Res 2003; et al. Remodeling of three-dimensional organization of the nucleus during 63:4692–7. terminal keratinocyte differentiation in the epidermis. J Invest Dermatol 16. Kagawa H, Shino Y, Kobayashi D, Demizu S, Shimada M, Ariga H, et al. A 2013;133:2191–201. novel signaling pathway mediated by the nuclear targeting of C-terminal 34. Dezawa M, Fujii K, Kita K, Nomura J, Sugita K, Adachi-Usami E, et al. fragments of mammalian patched 1. PLoS One 2011;6:e18638. Increase in nucleoli after x-radiation of fibroblasts of patients with Gorlin 17. Nagao K, Toyoda M, Takeuchi-Inoue K, Fujii K, Yamada M, Miyashita T. syndrome. J Lab Clin Med 1999;134:585–91. Identification and characterization of multiple isoforms of murine and 35. Bhosle VK, Rivera JC, Zhou TE, Omri S, Sanchez M, Hamel D, et al. Nuclear human tumor suppressor, patched, having distinct first exons. Genomics localization of platelet-activating factor receptor controls retinal neovas- 2005;85:462–71. cularization. Cell Discov 2016;2:16017. 18. Lo BK, Yu M, Zloty D, Cowan B, Shapiro J, McElwee KJ. CXCR3/ligands are 36. Cassandri M, Smirnov A, Novelli F, Pitolli C, Agostini M, Malewicz M, et al. significantly involved in the tumorigenesis of basal cell carcinomas. Am J Zinc-finger proteins in health and disease. Cell Death Discov 2017;3: Pathol 2010;176:2435–46. 17071.

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2587

Rahman et al.

37. Gozdzialska A, Wojas-Pelc A, Drag˛ J, Brzewski P, Jas´kiewicz J, Pastuszczak 44. Moleirinho S, Guerrant W, Kissil JL. The Angiomotins–from discovery to M. Expression of metalloproteinases (MMP-2 and MMP-9) in basal-cell function. FEBS Lett 2014;588:2693–703. carcinoma. Mol Biol Rep 2016;43:1027–33. 45. Hu R, Huffman KE, Chu M, Zhang Y, Minna JD, Yu Y. Quantitative 38. Karelina TV, Goldberg GI, Eisen AZ. Matrilysin (PUMP) correlates with secretomic analysis identifies extracellular protein factors that modulate dermal invasion during appendageal development and cutaneous neo- the metastatic phenotype of non-small cell lung cancer. J Proteome Res plasia. J Invest Dermatol 1994;103:482–7. 2016;15:477–86. 39. Kumar A, Bhanja A, Bhattacharyya J, Jaganathan BG. Multiple roles of 46. Stacey SN, Helgason H, Gudjonsson SA, Thorleifsson G, Zink F, Sigurdsson CD90 in cancer. Tumour Biol 2016;37:11611–22. A, et al. New basal cell carcinoma susceptibility loci. Nat Commun 40. Ishiwata T. Cancer stem cells and epithelial-mesenchymal transition: novel 2015;6:6825. therapeutic targets for cancer. Pathol Int 2016;66:601–8. 47. Rasmussen SL, Krarup HB, Sunesen KG, Pedersen IS, Madsen PH, Thorla- 41. Li X, Deng W, Nail CD, Bailey SK, Kraus MH, Ruppert JM, et al. Snail cius-Ussing O, et al. Hypermethylated DNA as a biomarker for colorectal induction is an early response to Gli1 that determines the efficiency of cancer: a systematic review. Colorectal Dis 2016;18:549–61. epithelial transformation. Oncogene 2006;25:609–21. 48. Chaudhry SS, Cain SA, Morgan A, Dallas SL, Shuttleworth CA, Kielty CM. 42. Harrison W, Cochrane B, Neill GW, Philpott MP. The oncogenic GLI Fibrillin-1 regulates the bioavailability of TGFbeta1. J Cell Biol 2007;176: transcription factors facilitate keratinocyte survival and transformation 355–67. upon exposure to genotoxic agents. Oncogene 2014;33:2432–40. 49. Marsh D, Dickinson S, Neill GW, Marshall JF, Hart IR, Thomas GJ. alpha 43. De Rosa G, Barra E, Guarino M, Staibano S, Donofrio V, Boscaino A. vbeta 6 Integrin promotes the invasion of morphoeic basal cell carcinoma Fibronectin, laminin, type IV collagen distribution, and myofibroblastic through stromal modulation. Cancer Res 2008;68:3295–303. stromal reaction in aggressive and nonaggressive basal cell carcinoma. Am J 50. Robbins DJ, Fei DL, Riobo NA. The Hedgehog signal transduction network. Dermatopathol 1994;16:258–67. Sci Signal 2012;5:re6.

2588 Cancer Res; 78(10) May 15, 2018 Cancer Research

A Novel Mechanism for Activation of GLI1 by Nuclear SMO That Escapes Anti-SMO Inhibitors

Muhammad M. Rahman, Allon Hazan, Joanne L. Selway, et al.

Cancer Res 2018;78:2577-2588. Published OnlineFirst February 20, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-17-2897

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2018/02/20/0008-5472.CAN-17-2897.DC1

Cited articles This article cites 50 articles, 11 of which you can access for free at: http://cancerres.aacrjournals.org/content/78/10/2577.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/78/10/2577. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.