Patched-2 functions to limit Patched-1 deficient skin cancer growth

Patched-2 functions to limit Patched-1 deficient skin cancer growth Purpose Basal cell carcinoma (BCC) is one of the most common skin cancers, and is typically driven by an aberrantly activated Hedgehog (Hh) pathway. The Hh pathway is regulated by interactions between the Patched-1 (Ptch1) and Smoothened (Smo) receptors. Smo is an activating receptor and is subject to inhibition by Ptch1. Following ligand binding to Ptch1, its inhibitory action is relieved and pathway activation occurs. This receptor interaction is pivotal to restraining uncontrolled cellular growth. Both receptors have been found to be frequently mutated in BCCs. Ptch2 is a Ptch1 paralog that exhibits overlapping functions in both normal development and tissue homeostasis. As yet, its contribution to cancer growth is poorly defined. Here we set out to assess how Ptch2 inhibits BCC growth. Methods We used several in vitro readouts for transcriptional and chemotactic Hh signaling in BCC-derived ASZ001 cells, and a novel xenograft model to assess in vivo BCC tumor growth. Gene editing by TALEN was used to untangle the different Ptch2- dependent responses to its ligand sonic hedgehog (Shh). Results We first defined the signaling competence of Ptch2 in Ptch1-deficient ASZ001 cells in vitro, and found that Ptch2 ligand binding drives their migration rather than eliciting a transcriptional response. We found that subsequent targeting of Ptch2 abrogated the chemotaxic effect. Next, we tested the contribution of Ptch2 to in vivo tumor growth using a xenograft model and found that reduced Ptch function results in increased tumor growth, but that selective pressure appatently acts against complete Ptch2 ablation. Conclusions We conclude that like Ptch1, Ptch2 exerts a tumor-suppressive function in BCC cells, and that after targeting of both paralogs, ligand-independent activation of the Hh pathway contributes to tumor growth. . . . Keywords Basal cell carcinoma Hedgehog signaling Patched Smoothened Electronic supplementary material The online version of this article (https://doi.org/10.1007/s13402-018-0381-9) contains supplementary material, which is available to authorized users. 1 Introduction * Maarten F. Bijlsma The Hedgehog (Hh) pathway is not only crucial to many in- m.f.bijlsma@amc.uva.nl ductive events in developing embryos and to the maintenance Laboratory for Experimental Oncology and Radiobiology, Center for of tissue integrity in adult organisms, but also to the initiation Experimental and Molecular Medicine, Cancer Center Amsterdam and progression of tumors [1, 2]. Hh pathway regulation is and Academic Medical Center, Meibergdreef 9, 1105AZ, primarily mediated by the transmembrane proteins Patched-1 Amsterdam, The Netherlands (Ptch1) and Smoothened (Smo) [3, 4]. Ptch1 is the main re- Oncode Institute, Academic Medical Center, ceptor for the pathway-activating ligand sonic hedgehog (Shh) Amsterdam, The Netherlands [5]. In the absence of this ligand, Ptch1 actively represses Smo Present address: Tumor Immunology Lab, Radboud Institute for to keep the pathway inactive [6]. In the presence of Shh the Molecular Life Sciences, Nijmegen, The Netherlands inhibitory action of Ptch1 is alleviated through its Present address: Biotech Research & Innovation Centre, relocalization, and Smo is free to signal to downstream path- Copenhagen, Denmark way components [7–9]. This intracellular signaling cascade Department of Pathology, Academic Medical Center, can induce a variety of responses such as transcription factor Amsterdam, The Netherlands activation and cytoskeleton remodeling to mediate chemotax- Department of Medical Oncology, Cancer Center Amsterdam and is [10, 11]. In cancer, two different mechanisms responsible Academic Medical Center, Amsterdam, The Netherlands V. Veenstra et al. for aberrant pathway activation can be discerned [12]. The there is little selective pressure on PTCH2. Here, we asked first mechanism entails excessive production of Shh ligand whether absence of repressive PTCH1 action enhances the by tumor cells, which subsequently acts in an autocrine or role of PTCH2 in Shh ligand perception and subsequent path- paracrine manner to provide tumor-promoting signals [13, way activation, rendering cells highly sensitive to Shh ligand, 14]. The second mechanism entails genetic aberrations in Hh or whether the contribution of PTCH2 to tumor growth is pathway components that cause Hh pathway activation solely dependent on its tumor suppressor function via the sup- [15–18]. These aberrations typically include inactivating mu- pression of Smo activity. Another question to be answered is tations in Ptch1, crippling its inhibitory action on Smo, or whether there is selective pressure against ablation of both activating mutations in Smo that render it insensitive to PTCH paralogs, which might explain the low incidence of Ptch1 inhibition. Through these latter activating mutations PTCH2 mutations observed in patient samples. We used pathway activity is induced cell autonomously, rendering the in vitro and in vivo systems, in conjunction with gene editing, cells independent of Shh ligands produced by themselves or to untangle the different responses of BCC cells to Shh ligand their surroundings. Given these considerations, Ptch1 is con- and show that deficiency for both PTCH paralogs accelerates sidered a bona fide tumor suppressor. One human cancer type tumor growth. that relies on mutations in Hh pathway components is basal cell carcinoma (BCC), the most prevalent skin cancer. Oncogenic mutations in Ptch1 and Smo are since long known to drive BCC, and mouse models have underscored the notion 2 Materials and methods that the development of this malignancy relies heavily on these mutations [19, 20]. A recent study on the mutational 2.1 Cell culture profiles of 126 BCCs has revealed PTCH1 mutations in 73% of the cases [21]. PANC-1 cells (ATCC, Manassas, VA) and mouse embry- −/− +/+ The current dogma on Hh pathway regulation holds that onic fibroblasts (Ptch1 and Ptch1 MEFs from Dr. Ptch1 is the principal receptor for Shh, and that other receptors Scott, Stanford University [35]) were cultured in high- involved in Shh binding like Cdon, Boc and Gas1 function as glucose DMEM containing 8% fetal bovine serum coreceptors [22, 23]. A paralog of Ptch1 is Patched-2 (Ptch2) (FBS), L-glutamine, penicillin and streptomycin (all from [24–26], and this paralog is thought to complement some Lonza, Basel, Switzerland) according to routine cell cul- Ptch1 functions [27–29]. It has been found, however, that ture procedures. ASZ001 cells [36, 37]werecultured in Ptch2 does not act as an equally strong regulator of the path- 154CF keratinocyte medium (Life Technologies) supple- −/− way. For instance, Ptch2 embryos have been found to be mented with 50 μM CaCl, penicillin and streptomycin, viable and to develop normally, and that in a genetically and 2% chelex-treated FCS. Cells were screened for my- Ptch1-deficient system Ptch2 cannot fully compensate for loss coplasma monthly by PCR. of the other homolog [29–31]. However, Ptch2 deficiency does exacerbate the skin tumor phenotype in partially Ptch1 deficient mice by deregulating epidermal lineage differentia- 2.2 Quantitative RT-PCR tion, and it has been found that the absence of both paralogs affects skin maintenance [32, 33]. Subsequent detailed analy- Cells were lysed in Trizol (Invitrogen) after which RNA was ses of Hh pathway target expression gradients in the epidermis isolated according to the manufacturer’s protocol. cDNA was revealed that full Ptch deficiency results in a uniformly high synthesized using Superscript III (Invitrogen) and random pathway activation [34]. Recent work in embryonic stem cells primers (Invitrogen). Quantitative real-time RT-PCR (qRT- has shown that Ptch2 is required for ligand perception in the PCR) was performed using SYBR green (Roche, Basel, absence of Ptch1 [27]. Intriguingly, in a Ptch1-deficient mouse Switzerland) on a Lightcycler LC480 II (Roche). Relative model of Hh pathway-driven BCC it was found that the tu- gene expression levels were calculated using the comparative mors preferentially arise from locations close to Shh sources threshold cycle (Ct) method and values were normalized to the [19]. These latter observations imply that in the absence of reference gene Gapdh. The primer sequences used were: Ptch1 at least some responsiveness to Shh remains and, Gapdh 5 ’ CTCATGACCACAGTCCATGC and 3 ’ therefore, that Shh is a likely candidate to mediate Ptch2 CACATTGGGGGTAGGAACAC; Gli1 5’ ATAGGGTC activity. TCGGGGTCTCA and 3’ CGGCTGACTGTGTAAGCAGA PTCH2 mutations are relatively rare events. A large-scale ; Ptch1 5’ GCTACGACTATGTCTCTCACATCAACTand 3’ genetic analysis revealed that only 14 out of 126 BCC cases GGCGACACTTTGATGAACCA; Ptch1 (exon 2) 5’ carried mutations in both PTCH1 and PTCH2 and that only 4 CTGTGGCTGAGAGCGAAGTT and 3’ AGCTCCTC cases exclusively carried PTCH2 mutations [21]. These ob- CACGTTGGTCT; Ptch2 5’ GCGTACACCTCCCA servations suggest that in the absence of functional PTCH1, GATGTT and 3’ GGAACCCCTGATTTGTAGCA. Patched-2 functions to limit Patched-1 deficient skin cancer growth 2.3 FACS analysis 2.6 Cell viability assays Cells were harvested using a trypsin-EDTA solution (Lonza) Cells were seeded at 2000 cells/well in 0.5% or 2% FCS and washed in FACS buffer (PBS containing 1% FBS). and after 3 h treated with the indicated compounds (see also Hybridoma supernatants containing either anti-Shh antibody [42]). After 4 d, MTT was added and dye reduction was 5E1 [38] or anti-Myc antibody 9E10 (isotype) were diluted assessed after a 4 h incubation period. Background 1:5 in FACS buffer and incubated for 30 min at 4 °C. A (10 mM H O treated cells) levels were subtracted and the 2 2 secondary APC labeled anti-mouse (BD, 550826) antibody values obtained from the control treated cells were set to 1, was used at a dilution of 1:500. After washing, the cells were after which the experimental data were normalized to the resuspended in FACS buffer containing 1 μg/ml propidium controls. iodide (PI) (Sigma) and subjected to flow cytometry using a FACSCanto II machine (BD, Franklin Lakes, NJ, USA). The 2.7 Transwell migration assays data obtained were analyzed using FlowJo 7 software (Tree Star, Ashland, OR, USA). Migration assays were performed as previously reported [10]. Briefly, cells were labeled with 10 μM CellTracker Green (Invitrogen) according to the manufacturer’s protocol. After 2.4 Immunofluorescence labeling, the cells were detached using 5 mM EDTA, resus- pended in serum free medium and transferred to FluoroBlok ASZ001 cells were grown on glass coverslips, starved for 2 d, Transwell inserts (BD Falcon) at a density of approximately and fixed using 4% formaldehyde. Following blocking and 5×10 cells per insert. Chemoattractant was added to the permeabilization in 5% goat serum/phosphate-buffered saline bottom compartments of the Transwell plates and GFP- with 0.1% Triton X100 (PBS-T), a primary antibody directed spectrum fluorescence in the bottom compartments was mea- against acetylated α-tubulin (Sigma) was added at 1:2000 and sured using a Synergy HT plate reader (BioTek, Winooski, incubated for 1 h at room temperature or overnight at 4 °C. VT, USA) every 2 min during approximately 3 h. Next, an Alexa 488 conjugated anti-mouse secondary anti- Background fluorescence was measured in time from a well body (Invitrogen) was added at 1:2000 and incubated for 1 h containing only medium and these values were subtracted at room temperature. Finally, the coverslips were mounted from all other measurements. The ‘no attractant’ control was using ProLong Gold (Invitrogen) and images were captured used to measure baseline cell movements for every experi- using a Zeiss AxioVert microscope. For Smo ciliary localiza- mental condition. These values were subtracted from those tion, cells were transfected with Myc-tagged Smo using PEI obtained in the presence of chemoattractants in the bottom 48 h prior to starvation and ShhN stimulation (1:4 diluted compartments of the Transwell inserts. The resulting data supernatant from 293 T cells). yielded specific migration values towards a given attractant. 2.5 GBS-GFP reporter construct and cell line 2.8 Gene editing and transfections establishment The pCTIGTALEN expression vector was used as described A concatemerized 8 × 3’GLI binding site sequence (GBS) before [27]. A pair of TALEN constructs was modified so that was isolated from the pδ51 GBS-luciferase reporter construct one construct co-expressed GFP and the other tdTomato, [39] by PCR and cloned into a lentiviral pRRL TOP-d2GFP allowing for selection for both 5′ and 3′ targeting constructs reporter vector [40] as reported before [41]. MEFs and by FACS sorting. The constructs were designed using Golden ASZ001 cells were transduced with the GBS-GFP reporter Gate cloning into pCTIG employing the following variable construct, starved in 0.5% FCS containing medium and stim- domain architectures: 5’ NN NN HD NG NG HD NN NI ulated for 4 d with ShhN conditioned supernatant from 293 T NN HD NG NG NI HD NG NG HD; 3’ NG HD NG NN cells or 200 nM Smo agonist (SAG, EMD Millipore, Billerica, NN NI NG HD HD NG NN HD NI HD HD HD HD. See also MA, USA). The resulting cells were sorted on a BD Supplementary Fig. S4a. FACSAria for GFP expression, after which GFP positive cells ASZ001 cells grown in 3 × 12-well plates were transfected were grown under 8% FCS (MEFs) or 2% FCS (ASZ001) with paired TALEN constructs using PEI. 3 days after trans- + + conditions, resulting in the expected loss of GFP activity. fection GFP /tdTomato cells cells were selected by flow For subsequent analyses, cells were seeded in 24-well plates sorting, seeded in bulk and allowed to recover for 7 d in and treated as indicated in the figure panels. Following treat- T25 flasks. To obtain monoclonal cultures, cells were seed- ment, cells were harvested and the percentages of GFP cells ed at single cell/well densities in 2 × 96-well plates. In 35 of were determinedby flow cytometryonaFACSCantoII the 192 wells cells grew out. These cells were genotyped machine. through the sequencing of PCR products spanning the V. Veenstra et al. TALEN binding sites (see also Supplementary Fig. S5). cilium, was confirmed in nearly all ASZ001 cells analyzed Screening was performed on genomic DNA by PCR using (Fig. 1c, right panel) [7]. The primary cilium is an antenna- primers flanking the TALEN binding sites: 5 ’ like protrusion from the cell membrane that is shaped by the AAGGCACAGGGAAAGAGAGTT; 3’ ACTTGCCT microtubule cytoskeleton. A dynamic localization of Hh AGCTTGCACAATG and subsequent digestion of the pathway components in and out of this organelle is required PCR products with AccI. Genomic DNA from monoclonal for pathway regulation, and the primary cilium is required lines that exhibited loss of the restriction site were TOPO for appropriate ligand perception. Thus, at least part of the cloned (Thermo Fisher) and Sanger sequenced. immediate signaling machinery for Shh is intact in these cells. Overxpression of Hedgehog pathway components was ac- The ciliary localization of exogenously overexpressed Smo in complished using wild-type Smo (SmoWT) and ciliary locali- response to the addition of Shh was subsequently assessed by zation domain mutated Smo (SmoCLD) constructs in pCS107 microscopy. Despite a high baseline percentage of cells with obtained from Dr. Jeremy Reiter [7, 43]and awild-type Ptch1 Smo localized in the cilium, as expected from a Ptch1-deficient construct in pcDNA3.1 obtained from Dr. Matthew Scott (see system and overexpression of Smo, we found that the addition loop2 also ref. [44]). mPtch1Δ is a Ptch1 form that lacks the of ligand resulted in a moderate increase in this number (Fig. second extracellular loop required for ligand binding and only 1d). As a control, a form of Smo that cannot localize to the exerts Smo-repressive functions [45]. cilium (SmoCLD), was used [7]. Taken together, our results imply that in the absence of Ptch1, the Shh ligand is still per- 2.9 Xenografting of ASZ001 cells ceived by the cells. scid tm1Wjl NOD.Cg-Prkdc Il2rg /SzJ (NSG) mice were bred in- 3.2 ASZ001 cells respond to hedgehog by chemotaxis 5 5 house. The animals were grafted with 1 × 10 or 5 × 10 ASZ001 cells in Matrigel [46]. All experiments were per- Typically, relocalization of Smo to the primary cilium results formed according to procedures approved by the animal ex- in activation of the downstream pathway leading to transcrip- periment ethical committee (LEX237). Tumor growth was tional responses. However, when Gli transcription factor ac- monitored weekly and the experiments were ended when ul- tivity was measured in these cells using a stably integrated ceration was observed. Of note, ulceration was typically ob- GLI-binding site GFP reporter construct, only minimal re- served before the tumors reached a humane endpoint volume sponses to exogenous pathway activators were observed, al- (1000 mm ). Finally, the tumors were harvested and processed though the baseline pathway activity was high (Fig. 2a,b). In for paraffin embedding and subsequent histopathological ex- comparison, MEFs proficient for Ptch1 showed robust re- amination, as well as for RNA extraction. sponses to the pathway activators. These results were con- firmed by qRT-PCR analysis of target genes, through which minor responses were observed in ASZ001 cells compared to 3 Results MEFs (Fig. 2c,d, for baseline pathway activity see Fig. 1a). Under high serum proliferation conditions, a strong reduction 3.1 Ptch1-deficient basal cell carcinoma cells perceive in basal reporter activity in the ASZ001 cells was observed hedgehog signalling (4% GFP cells, data not shown), indicating that the Hh path- way in these cells is amenable to at least some degree of As a model for Ptch1-deficiency driven skin cancer, we used regulation, most likely through cell cycle progression and the ASZ001 cell line. This cell line has been established from subsequent primary cilium loss. Switching these cells back +/- an irradiation-induced tumor in a Ptch1 mouse, and has to low-serum conditions resulted in a regain of GFP reporter subsequently undergone Ptch1 loss of heterozygosity ( [36] expression to the levels shown in Fig. 2a,b. In order to test and Fig. 1a). The current working model for Hedgehog (Hh) whether the mitogenic Hh response can be driven by exoge- pathway regulation dictates that loss of Ptch1 suffices for full nously added ligand in ASZ001 cells, we treated them with Hh pathway activation and that no additional ligand is re- ShhN or control (Ctrl/GFP) supernatants. By doing so, we quired for this. Indeed, the Hh pathway was found to be acti- observed only marginal proliferative responses to Shh com- vated in these cells as evident from abundant transcription of paredtothe control(SupplementaryFig.S2a). In addition, the remaining exon on the targeted Ptch1 allele, which itself is we found that inhibitors of Hh pathway components and a target of the activated pathway (Fig. 1a). We also found that related signaling molecules were relatively ineffective in the cells did not produce ligand that could cell autonomously restraining ASZ001 cell proliferation, as evident from their activate the pathway (Fig. 1b; SHH expressing pancreatic overall high IC50 values (Supplementary Fig. S2b,c) [42, cancer PANC-1 and Shh transfected ASZ001 cell controls 44, 47]. are shown in Supplementary Fig. S1). One prerequisite As shown previously in fibroblasts and neural develop- for Hh ligand responsiveness, i.e., the presence of a primary mental models, Shh ligand is also able to induce chemotaxis, Patched-2 functions to limit Patched-1 deficient skin cancer growth a b c d 100 GFP 100 0.3 mPtch1 1 ASZ001 α-SHH ShhN mPtch1 exon2 isotype 75 75 ** 0.2 α-acetylated tubulin 50 50 0.1 25 25 DAPI 0 0 0.0 0 12 3 4 ASZ001 SHH (APC-A log10) expressing Fig. 1 Ptch1-deficient BCC cells perceive Shh ligand. a RNA was were grown on coverslips, transfected with Myc-tagged forms of Smo, isolated from the cells indicated on the X-axis after which qRT-PCR- starved and treated with ShhN (or control; GFP) supernatant diluted 1:4 based expression analysis was performed relative to mGapdh using a from 293 T cells for 1 h. Next, the cells were fixed and stained for Myc Ptch1 exon upstream of the targeted exon (grey bars) and the targeted and acetylated α-tubulin, after which the percentage of transfected (Myc- exon 2 (dark blue bars). Bars indicate means ± SEM (n = 4), n.d. indicates positive) cells with ciliary Smo was quantified. SmoCLD; ciliary locali- not detected (no signal). b ASZ001 cell surface levels of Hedgehog li- zation domain mutated form of Smo [7, 43]. For assessment of the frac- gands were determined by FACS using a 5E1 anti-Shh hybridoma anti- tion of ciliated cells, quantifications from both transfections were pooled body or an isotype control. c ASZ001 cells were grown to confluence on and depicted in the separate grey bar graph. Bars indicate means ± SEM coverslips after which primary cilia were visualized by acetylated α- (n > 50 cells quantified from 2 separate experiments). **p =0.0015 by tubulin staining. Nuclei were counter-stained with DAPI. d ASZ001 cells Mann-Whitney U test irrespective the presence of Hh pathway transcription factors. 3.3 Ptch2 mediates a chemotactic response This response appears to be relatively resilient to the pertur- to hedgehog ligand bation of pathway components [43]. To assess whether the perception of Shh by ASZ001 cells may function to activate Due to the catalytic inhibitory activity of Ptch on Smo, a low this chemotactic response, we assessed the migratory capacity number of Ptch molecules suffices to suppress Hh pathway of these cells. Pathway activators were used in a modified activity [6] and, therefore, highly effective targeting is re- Boyden chamber assay to quantitatively measure chemotaxis. quired to study the consequences of Ptch loss. We found that A robust migratory response was observed to both recombi- lentiviral shRNA-mediated Ptch2 silencing using five differ- nant ShhN (5 nM) and SAG (200 nM; Fig. 2e). This response ent sequences and puromycin selection did not result in effec- could be blocked with cyclopamine (5 μM), confirming that tive targeting (Supplementary Fig. S3). Therefore, we turned this form of chemotaxis requires Smo. The subsequent use of to transcription activator-like effector nucleases (TALENs) to a Smo agonist (SAG) further underscored the Smo dependen- edit the Ptch2 locus in ASZ001 cells (strategy shown in cy of this migratory response. After exogenous Ptch1 overex- Supplementary Fig. S4). Following transfection, cells were pression, we observed an increased relative responsiveness of FACS-sorted in bulk to ensure the survival of cells expressing the cells to ligand as evident from an increased net migratory both TALENs (Supplementary Fig. S5a). After recovery, the response to Shh (Fig. 2f). A form of Ptch1 that is unable to cells were seeded at single cell density in 2 × 96 well plates Δloop2 bind Shh (Ptch1 ) was found to be inhibitory to the Shh after which 35 of the clones grew out (18%). These single cell stimulated migration [45]. This suggests that the migratory clones were analyzed for editing events and, by doing so, we response, although not strictly dependent on Ptch1, can be found that in 5 of the clones the digest pattern was indicative modulated by the inhibitory activity of Ptch1. Together, these of an editing event (Supplementary Fig. S5b; lines 1, 4, 14, 16, data indicate that despite the absence of Ptch1, ASZ001 cells 17). These lines were TOPO cloned after which 20 clones per have retained Shh chemotactic responsiveness. This notion line were Sanger sequenced to verify the efficiency of editing suggests that other receptors may mediate the chemotactic (Supplementary Fig. S5c). Surprisingly, none of these lines response to Shh. A likely candidate to mediate this response was completely devoid of wild type Ptch2 sequences. In is its paralog Ptch2. cell line 4, one third of the sequences were mutant, whereas expression relative to mGapdh Ptch1+/+ MEFs Ptch1-/- MEFs n.d. ASZ001 n.d. normalized to mode fraction of cells with ciliary Smo (%) SmoWT SmoCLD fraction of cells with a primary cilium (%) V. Veenstra et al. a b c d 50 control 4 100 ASZ001 cells ASZ001 cells +/+ +/+ SAG Ptch1 fibroblasts Ptch1 fibroblasts ShhN 0 0 0 0.1 0.5 0.5 -9 -6 -3 [ShhN] (sup) [SAG] (M log10) e f 4 ASZ001 cells *** 2.5 SAG rShhN cyclopamine 2.0 rShhN + cycl. 1.5 *** 1.0 0.5 *** -1 0 0 0 10 20 30 40 50 Cycle (min x 3) Fig. 2 ASZ001 cells mediate a chemotactic response to Shh. a GBS-GFP presence of 5 μM cyclopamine in both the upper and the bottom com- transduced ASZ001 and Ptch1-proficient fibroblasts were starved and partments of the Transwells. Shaded curves represent the SEM. The treated with the indicated dilutions of ShhN supernatants produced by curves were plotted using a hyperbola function. Measurements from at 293 T cells. For the highest concentration, GFP transfected 293 T super- least 3 replicates are shown. For details see Materials and methods sec- natant was included as a control (Ctrl/GFP 0.5). After 3 d, the GFP tion. The data are plotted as average RFU in bar graphs. Statistical test percentage was assessed by FACS (n = 4 for the ASZ001 cells; n =2 compares rShhN + cyclopamine versus rShhN. The difference in migra- for the MEFs). b As for panel a, using SAG. At the two highest tested tion between no-attractant control and ShhN/SAG is statistically signifi- doses, SAG was toxic to the fibroblasts (n = 3 for the ASZ001 cells; n =4 cant (p <0.001). f ASZ001 cells were transfected with indicated con- for the MEFs). c-d Cells were treated as for panels a-b using 200 nM SAG structs, and after ~24 h the migration response to 5 nM recombinant ShhN or 1:4 diluted ShhN supernatant. After treatment, RNA was isolated and was assessed. Bars indicate means ± SEM from approximately 70 mea- qRT-PCR was performed for mGli1 and mPtch1 (n = 5). *p < 0.05; surements from 2 separate experiments. Differences to the vector control **p < 0.01; ***p < 0.001; determined by t-test, (e) ASZ001 cells were were tested for the mPtch1 transfected condition, and for the mPtch1 Δloop2 seeded in a modified Boyden chamber after which net migration to 5 nM condition against the mPtch1 condition recombinant ShhN or 200 nM SAG was assessed in the absence or LOW in cell lines 1, 14, 16 and 17 two-thirds of sequences were 14, 16 and 17 (denoted Ptch2 ), of which we continued mutant. In addition, we found that the patterns of gene with clone 1. The failure to establish cell lines with com- editing in cell lines 1, 14, 16 and 17 were nearly identical, plete Ptch2 gene editing from otherwise successfully suggesting that amongst the initial 35 cell lines available for transfected cells implies that selective pressure exists restriction digest analysis, only two parental clones showed against full Ptch-deficiency. This notion is in agreement MED significant Ptch2 editing, i.e., clone 4 (denoted Ptch2 ) with the lentiviral shRNA Ptch2 silencing results (see and an additional parental clone which gave rise to clones 1, Supplementary Fig. S3). GFP cells (%) migration (RFUx10 ) Ctrl/ GFP average migration (RFUx10 ) rShhN SAG cyclopamine mGli1 relative to control treated rShhN + cyclopamine migration to rShhN (RFUx10 ) ASZ001 * RFP +/+ Ptch1 *** fibroblasts mPtch1 ** Δloop2 mPtch1 mPtch1 relative to control treated ASZ001 +/+ Ptch1 ** fibroblasts Cells targeted for Ptch2 were found to be transcriptionally 3.4 Ptch2 deficiency accelerates tumor growth unresponsive to ShhN ligand as determined by qRT-PCR (Fig. 3a). Surprisingly, we also found that Ptch2 gene editing Having established that (1) in the absence of Ptch1 Hh ligand resulted inanincrease inbasal transcriptional pathway activity is perceived by Ptch2 only to mediate chemotaxis and (2) that +/+ in vitro when comparing wild-type Ptch2 parental cells with additional targeting of Ptch2 impedes the ability of the cells to +/+ wild-type Ptch2 clone 35, arguing for the use of gene edited perceive Hh ligand, we proceeded to test the consequences of but wild-type cells as controls. If Ptch2 mediates chemotaxis in these signaling outputs for in vivo tumor growth. To this end, response to Hh ligand as was concluded from the data presented ASZ001 cells were injected subcutaneously in immunodefi- in Fig. 2, its targeting should inhibit the chemotactic response. cient mice after which tumor growth was monitored. We Indeed, we observed a strong reduction in chemotaxis to ShhN found that tumors grown from cells targeted for both Ptch LOW MED LOW MED in the Ptch2 and Ptch2 cells,ascomparedtothe paralogs (i.e., Ptch2 and Ptch2 ASZ001 cells) expand- +/+ Ptch2 clone 35 cells (Fig. 3b-c). Baseline chemokinesis ed faster than those grown from fully Ptch2-proficient cells +/+ (i.e., the movement of cells in the absence of attractant; these (i.e., ASZ001 Ptch2 clone 35 cells). The former mice were values are typically subtracted from migration to Shh to yield sacrificed sooner based on ulceration (Fig. 4a). Subsequent net chemotaxis as shown in Fig. 3b)did not differbetween the immunohistochemical analysis for the proliferation marker genotypes (Supplementary Fig. S6a). The residual chemotactic Ki67 confirmed a higher proliferative index in the Ptch2 responsiveness of Ptch2-edited ASZ001 cells was sensitive to gene-edited tumors (Fig. 4b, quantification in Fig. 4c). cyclopamine, showing that the regulation of Shh chemotaxis is Assessment of tumor histology by a pathologist confirmed dependent on Smo (Supplementary Fig. S6b). a cutaneous source of the tumor cells, but also revealed a Also, receptors other than Ptch1 and Ptch2 have been im- squamous rather than basal histology for both genotypes, pos- plicated in the perception of Shh for chemotactic responses sibly owing to extensive in vitro culturing of the cells prior to LOW MED [23, 48, 49]. In order to test whether putative Hh receptors in grafting. The Ptch2 and Ptch2 derived tumors showed addition to Ptch2 may mediate Shh chemotaxis in ASZ001 more keratin depositions (eosin rich pink areas in Fig. 4d; cells, we targeted the Cdon and Boc receptors by lentiviral quantification in Fig. 4e). To determine whether the in vivo shRNA delivery and found that, despite a modest knockdown accelerated proliferation of these Ptch2 gene-edited cells efficiency, migration was hampered (Supplementary Fig. could be explained by additional activation of the Hh pathway S6c,d). This implies that in addition to Ptch2, other candidate over the already high level caused by the Ptch1-deficiency, the receptors for Shh may elicit chemotactic responses. expression of Hh pathway target genes was measured by qRT- LOW Nevertheless, given that Ptch2 is incapable of mediating a PCR. Indeed, these were found to be elevated in the Ptch2 MED robust transcriptional response to ShhN (Fig. 2a-d and 3a), and Ptch2 ASZ001 tumors (Fig. 4f)and arelikelyrespon- we conclude that the ligand-dependent contributions of sible for the observed increase in tumor growth rates. We Ptch2 in Ptch1-deficient cells in vitro are confined to the che- hypothesize that the discrepancy in Hh pathway activation motactic response. in vitro and in vivo following Ptch2 targeting results from +/+ a b Ptch2 clone 35 c MED Ptch2 clone 4 4 LOW control 1.1 10.0 Ptch2 clone 1 ShhN p<0.0001 *** 1.0 7.5 0.9 NS ** 2 0.8 NS 5.0 NS NS *** NS 0.7 NS NS 2.5 0.6 0.5 0 0 0 25 50 75 100 Cycle (min x 3) +/+ Fig. 3 Ptch2 is required for Shh chemotaxis. a ASZ001 Ptch2 parental parental cells. *p <0.05; **p <0.01;***p <0.001; determined by t-test, +/+ MED LOW cells and TALEN genome edited ASZ001 cells were stimulated with (b) Ptch2 , Ptch2 , and Ptch2 cells were seeded in a modified ShhN as depicted in Fig. 2c-d. Target gene (mPtch1) transcript analysis Boyden chamber after which specific migration (net chemotaxis) to 5 nM was performed. Genotypes and clones are indicated on the X-axis and by recombinant ShhN was assessed as in Fig. 2e. Measurements from 4 the blue-shaded (ShhN) bars. Blue asterisks denote statistical compari- replicates in 2 experiments are shown. The indicated p-values were de- sons of ShhN-treated cells and control treated cells of the same genotype. termined by one-way ANOVA. c Schematic representation of Fig. 3b. +/+ Grey asterisks indicate significance compared to ASZ001 Ptch2 Significances were determined using Mann-Whitney U test mPtch1 expression relative to control treated parental cells +/+ Ptch2 parental +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 LOW Ptch2 clone 14 LOW Ptch2 clone 16 net chemotaxis to ShhN (RFUx10 ) normalized migration (RFU) +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 *** *** V. Veenstra et al. 5 5 1000 1x10 cells grafted 800 5x10 cells grafted +/+ +/+ ASZ001 Ptch2 clone 35 ASZ001 Ptch2 clone 35 MED MED ASZ001 Ptch2 clone 4 ASZ001 Ptch2 clone 4 800 LOW LOW ASZ001 Ptch2 clone 1 ASZ001 Ptch2 clone 1 0 0 01234567 01234567 time post-graft (weeks) time post-graft (weeks) +/+ ASZ001 Ptch2 clone 35 MED b c e ASZ001 Ptch2 clone 4 1.0 30 ** * LOW *** ASZ001 Ptch2 clone 1 *** ** *** *** ns -5 0.5 ns -10 0.0 0 -15 Fig. 4 Patched deficiency accelerates tumor growth. a 1×10 and 5 × of positive nuclei per optical field (n ≥ 7). The indicated p-values were 10 ASZ001 cells of the indicated genotypes were subcutaneously grafted determined using Mann-Whitney U test. d Histology of tumors grown in immune deficient mice in PBS/Matrigel. Subsequent tumor growth from the indicated genotypes revealed by hematoxylin and eosin staining. was monitored and tumors were harvested after signs of ulceration. In Scale bar: 1 mm. e Automated quantification of pink eosin-stained fields each group 4 mice were grafted with 2 tumors, yielding sample sizes of 8. shown in d. f After RNA extraction from the harvested tumors transcript b After harvesting, the tumors were processed for Ki67 analyses for the indicated genes were performed (n = 16). Significances immunohistochemistry. Scale bars: 200 μm. c Automated quantification were determined using Mann-Whitney U test (see Fig. 3) environmental signals and/or mechanical properties that are of ligand in developmental models [27]. Also, Ptch1 and only present in vivo, which feed into the signaling cascade Ptch2 have been shown to exhibit overlapping functions in downstream of Smo. From the increased baseline pathway Hh pathway-dependent skin development and maintenance activity and associated tumor growth following Ptch2 [32, 34]. The exact contribution of Ptch2 to an established targeting in vivo, we conclude that Ptch2 exerts ligand- cancer type such as basal cell carcinoma (BCC) has, however, independent pathway inhibitory functions that make it a tumor remained unclear, despite clinical data that suggest a tumor suppressor, but one that is also required and essential for can- suppressor function [50]. Whether such a putative tumor sup- cer cell viability. pressor function of Ptch2 may be uncoupled from its ligand binding function is currently unknown and is possibility com- plicated by the fact that these functions are connected [45]. 4 Discussion Here, we have untangled these functions by first delineating the signaling capabilities of Ptch2 in BCC cells in vitro and The current model of Hedgehog (Hh) pathway regulation subsequently testing the consequences of Ptch2 perturbation for in vivo BCC tumor growth. Using migration assays we holds that Ptch1 acts as the main receptor for Shh, and that it −/− serves as the master switch for downstream pathway regula- found that in the parental Ptch1 ASZ001 cells the only response to Hh ligand was chemotactic. After Ptch2 targeting tion [3, 6]. Recent work has shown, however, that the concept of Ptch1 as key Shh receptor needs revision. In the absence of by TALEN we found that this chemotactic response was se- verely hampered. Therefore, we conclude that the increased Ptch1, Ptch2 has been found to be required for the perception tumor size (mm ) LOW MED +/+ Ptch2 Ptch2 Ptch2 clone 1 clone 4 clone 35 Ki67+ nuclei (fraction) +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 LOW MED +/+ Ptch2 Ptch2 Ptch2 clone 1 clone 4 clone 35 tumor size (mm ) keratin surface (%) +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 expression relative to mGapdh (log2) Gli1 Ptch1 Ptch2 Patched-2 functions to limit Patched-1 deficient skin cancer growth tumor growth observed following targeting of Ptch2 is likely is warranted. In addition, it remains to be established whether unrelated to ligand perception and that the tumor suppressive the identified tumor suppressive action of Ptch2 is also rele- action of Ptch2 very likely depends on its baseline inhibitory vant for non- or pre-malignant cultured keratinocytes. activity towards Smo, which for an as yet unknown reason Previously, we have shown that ciliary relocalization of only becomes apparent in vivo. Smo in response to Shh is not required for Shh chemotaxis, As mentioned above, Ptch2 mutations are uncommon in and suggested that chemotactic and transcriptional responses BCC [21]. This is at apparent odds with its similarities and to Shh may represent separate phenomena resulting from dis- overlapping signaling roles with Ptch1. However, as we find tinct intracellular mechanisms [43]. Our current data indicate that Ptch2 has a very limited ligand-perceiving role in BCC, that these responses are indeed not mutually exclusive, and we assume that its main function as a tumor suppressor is that Smo localization to the cilium occuring in the absence of −/ overshadowed by the loss of Ptch1, which is a stronger inhib- robust downstream transcriptional signaling (i.e., in Ptch1 − +/+ itor of the Hh pathway [31, 33]. This notion is also supported ;Ptch2 cells) may initiate a chemotactic response. In fact, by the high incidence of mutations in for instance TP53. TP53 the convergence of both pathways at several points within the is a much stronger tumor suppressor than Ptch2, and it is more signaling cascade (Smo, the primary cilium) seems to suggest easily perturbed given that only one allele requires a mutation that they are both part of a relatively conserved pathway. This to yield an oncogenic event. Mutations in TP53 are, therefore, holds promise for the application of currently available drugs more advantageous and more regularly achieved than PTCH2 against BCC such as erivedge/vismodegib [54]. Given the fact LOH which, in turn, could explain the low mutation rate of that vismodegib acts to inhibit Smo, rather than downstream PTCH2. Another explanation for the paucity of PTCH2 mu- signaling components or transcription factors, it may be effec- tations in human BCC follows directly from our observation tive against both the chemotactic and the transcriptional ligand that full Ptch2 gene ablation could not be achieved, despite responses, as well as against baseline pathway activity. This selection. It thus appears that a low level of Ptch2 activity is drug may, therefore, turn out to be effective against BCC required for cancer cell viability, at least in vitro. Modest per- irrespective the mutational status of other Hh ligand receptors. turbations of Ptch2 are apparently tolerated, resulting in suffi- Acknowledgements We would like to thank Dr. Ervin Epstein and Dr. cient Hh pathway upregulation in vivo to boost tumor growth, Heidi Hahn for the ASZ001 cells, and Dr. Matthew Scott for providing as evident from the enhanced tumor growth found for the the (Ptch1) MEFs. We are thankful to Brock Roberts and Henk Roelink MED Ptch2 cells. How these two counteracting activities are for assisting in the TALEN genome editing, Tom van Leusden for tech- balanced remains to be determined, but it is possible that they nical assistance, and Dr. Louis Vermeulen for proofreading. are driven by distinct signaling functions of Ptch2 (Smo an- tagonist, chemotaxis receptor, dependence receptor), or by a Funding This work was supported by a KWF Dutch Cancer Society cancer-specific context. In addition, it is possible that rare Research Grant (UVA 2012–5607) to MFB and HVL. tumors that are currently not recognized to be Hh-driven may rely on mutations in Ptch2 if that serves as the dominant Compliance with ethical standards Hh receptor in the tissue of origin. For instance, the testis- specific expression of Desert Hedgehog (Dhh) and Ptch2 Conflicts of interest MFB and HVL have received research funding from Celgene. This party was not involved in drafting the manuscript. and their role in glioblastoma-endothelium crosstalk hints to this possibility [24, 51, 52]. We found that some chemotactic responsiveness was retained by the Ptch-edited ASZ001 cells. Open Access This article is distributed under the terms of the Creative We hypothesize that this responsiveness may result from can- Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, didate coreceptors for Shh other than Ptch2, such as Cdon, distribution, and reproduction in any medium, provided you give Boc or Gas1, or possibly Smo itself [53]. Indeed, we found appropriate credit to the original author(s) and the source, provide a link that the Cdon and Boc receptors may also contribute to che- to the Creative Commons license, and indicate if changes were made. motactic Shh signaling. Given the fact that the only experi- mental variable in the xenografting experiments was the level of Ptch2 we are, however, reluctant to draw firm conclusions References on its signaling contributions in vivo. One caveat of our study is that it is yet unclear to what 1. J. Briscoe, P.P. Therond, The mechanisms of hedgehog signalling extent our findings in murine cells can be translated to the and its roles in development and disease. Nat Rev Mol Cell Biol 14, 416–429 (2013) human situation. The ASZ001 cell line represents a powerful 2. S. Teglund, R. Toftgard, Hedgehog beyond medulloblastoma and exerimental tool given that these cells are unambiguously Hh- basal cell carcinoma. Biochim Biophys Acta 1805,181–208 (2010) dependent, which is important for our experiments, and that 3. D.M. Stone, M. Hynes, M. Armanini, T.A. Swanson, Q. Gu, R.L. many murine-specific genetic tools to study Hh signaling are Johnson, M.P. Scott, D. Pennica, A. Goddard, H. Phillips, M. Noll, available. Nevertheless, validation in a human-derived model J.E. Hooper, F. de Sauvage, A. Rosenthal, The tumour-suppressor V. Veenstra et al. gene patched encodes a candidate receptor for sonic hedgehog. 20. K.K. Youssef, A. Van Keymeulen, G. Lapouge, B. Beck, C. Michaux, Y. Achouri, P.A. Sotiropoulou, C. Blanpain, Nature 384,129–134 (1996) 4. J. Alcedo, M. Ayzenzon, T. Von Ohlen, M. Noll, J.E. Hooper, The Identification of the cell lineage at the origin of basal cell carcino- ma. Nat Cell Biol 12,299–305 (2010) Drosophila smoothened gene encodes a seven-pass membrane pro- tein, a putative receptor for the hedgehog signal. Cell 86,221–232 21. X. Bonilla, L. Parmentier, B. King, F. Bezrukov, G. Kaya, V. Zoete, (1996) V.B. Seplyarskiy, H.J. Sharpe, T. McKee, A. Letourneau, P.G. 5. V. Marigo, R.A. Davey, Y. Zuo, J.M. Cunningham, C.J. Tabin, Ribaux, K. Popadin, N. Basset-Seguin, R. Ben Chaabene, F.A. Biochemical evidence that patched is the hedgehog receptor. Santoni, M.A. Andrianova, M. Guipponi, M. Garieri, C. Verdan, Nature 384,176–179 (1996) K. Grosdemange, O. Sumara, M. Eilers, I. Aifantis, O. Michielin, 6. J. Taipale, M.K. Cooper, T. Maiti, P.A. Beachy, Patched acts cata- F.J. de Sauvage, S.E. Antonarakis, S.I. Nikolaev, Genomic analysis lytically to suppress the activity of smoothened. Nature 418,892– identifies new drivers and progression pathways in skin basal cell 897 (2002) carcinoma. Nat Genet 48,398–406 (2016) 7. K.C. Corbit, P. Aanstad, V. Singla, A.R. Norman, D.Y. Stainier, J.F. 22. B.L. Allen, T. Tenzen, A.P. McMahon, The hedgehog-binding pro- Reiter, Vertebrate smoothened functions at the primary cilium. teins Gas1 and Cdo cooperate to positively regulate Shh signaling Nature 437,1018–1021 (2005) during mouse development. Genes Dev 21,1244–1257 (2007) 8. J.P. Incardona, J. Gruenberg, H. Roelink, Sonic hedgehog induces 23. T. Tenzen, B.L. Allen, F. Cole, J.S. Kang, R.S. Krauss, A.P. the segregation of patched and smoothened in endosomes. Curr McMahon, The cell surface membrane proteins Cdo and Boc are Biol 12,983–995 (2002) components and targets of the hedgehog signaling pathway and 9. J.P. Incardona, J.H. Lee, C.P. Robertson, K. Enga, R.P. Kapur, H. feedback network in mice. Dev Cell 10,647–656 (2006) Roelink, Receptor-mediated endocytosis of soluble and membrane- 24. D. Carpenter, D.M. Stone, J. Brush, A. Ryan, M. Armanini, G. tethered sonic hedgehog by Patched-1. Proc Natl Acad Sci U S A Frantz, A. Rosenthal, F.J. de Sauvage, Characterization of two 97,12044–12049 (2000) patched receptors for the vertebrate hedgehog protein family. Proc 10. M.F. Bijlsma, K.S. Borensztajn, H. Roelink, M.P. Peppelenbosch, Natl Acad Sci U S A 95, 13630–13634 (1998) C.A. Spek, Sonic hedgehog induces transcription-independent cy- 25. I. Smyth, M.A. Narang, T. Evans, C. Heimann, Y. Nakamura, G. toskeletal rearrangement and migration regulated by arachidonate Chenevix-Trench, T. Pietsch, C. Wicking, B.J. Wainwright, metabolites. Cell Signal 19,2596–2604 (2007) Isolation and characterization of human patched 2 (PTCH2), a pu- 11. A. Ruiz, i. Altaba, C. Mas, B. Stecca, The Gli code: An information tative tumour suppressor gene inbasal cell carcinoma and medullo- nexus regulating cell fate, stemness and cancer. Trends Cell Biol 17, blastoma on chromosome 1p32. Hum Mol Genet 8, 291–297 438–447 (2007) (1999) 12. S.J. Scales, F.J. de Sauvage, Mechanisms of hedgehog pathway 26. P.G. Zaphiropoulos, A.B. Unden, F. Rahnama, R.E. Hollingsworth, activation in cancer and implications for therapy. Trends R. Toftgard, PTCH2, a novel human patched gene, undergoing Pharmacol Sci 30,303–312 (2009) alternative splicing and up-regulated in basal cell carcinomas. 13. H. Tian, C.A. Callahan, K.J. DuPree, W.C. Darbonne, C.P. Ahn, Cancer Res 59,787–792 (1999) S.J. Scales, F.J. de Sauvage, Hedgehog signaling is restricted to the 27. A.C. Alfaro, B. Roberts, L. Kwong, M.F. Bijlsma, H. Roelink, stromal compartment during pancreatic carcinogenesis. Proc Natl Ptch2 mediates the Shh response in Ptch1−/− cells. Development Acad Sci U S A 106,4254–4259 (2009) 141,3331–3339 (2014) 14. R.L. Yauch, S.E. Gould, S.J. Scales, T. Tang, H. Tian, C.P. Ahn, D. 28. B. Roberts, C. Casillas, A.C. Alfaro, C. Jagers, H. Roelink, Marshall, L. Fu, T. Januario, D. Kallop, M. Nannini-Pepe, K. Patched1 and Patched2 inhibit smoothened non-cell autonomously. Kotkow, J.C. Marsters, L.L. Rubin, F.J. de Sauvage, A paracrine elife 5 (2016) requirement for hedgehog signalling in cancer. Nature 455,406– 29. O. Zhulyn, E. Nieuwenhuis, Y.C. Liu, S. Angers, C.C. Hui, Ptch2 410 (2008) shares overlapping functions with Ptch1 in Smo regulation and limb 15. H. Hahn, C. Wicking, P.G. Zaphiropoulous, M.R. Gailani, S. development. Dev Biol 397,191–202 (2015) Shanley, A. Chidambaram, I. Vorechovsky, E. Holmberg, A.B. 30. A.M. Holtz, K.A. Peterson, Y. Nishi, S. Morin, J.Y. Song, F. Unden, S. Gillies, K. Negus, I. Smyth, C. Pressman, D.J. Leffell, Charron, A.P. McMahon, B.L. Allen, Essential role for ligand- B. Gerrard, A.M. Goldstein, M. Dean, R. Toftgard, G. Chenevix- dependent feedback antagonism of vertebrate hedgehog signaling Trench, B. Wainwright, A.E. Bale, Mutations of the human homo- by PTCH1, PTCH2 and HHIP1 during neural patterning. log of Drosophila patched in the nevoid basal cell carcinoma syn- Development 140,3423–3434 (2013) drome. Cell 85,841–851 (1996) 31. E. Nieuwenhuis, J. Motoyama, P.C. Barnfield, Y. Yoshikawa, X. 16. R.L. Johnson, A.L. Rothman, J. Xie, L.V. Goodrich, J.W. Bare, Zhang, R. Mo, M.A. Crackower, C.C. Hui, Mice with a targeted J.M. Bonifas, A.G. Quinn, R.M. Myers, D.R. Cox, E.H. Epstein mutation of patched2 are viable but develop alopecia and epidermal Jr., M.P. Scott, Human homolog of patched, a candidate gene for the hyperplasia. Mol Cell Biol 26,6609–6622 (2006) basal cell nevus syndrome. Science 272,1668–1671 (1996) 32. C. Adolphe, E. Nieuwenhuis, R. Villani, Z.J. Li, P. Kaur, C.C. Hui, 17. J. Svard, K. Heby-Henricson, M. Persson-Lek, B. Rozell, M. Lauth, B.J. Wainwright, Patched 1 and patched 2 redundancy has a key A. Bergstrom, J. Ericson, R. Toftgard, S. Teglund, Genetic elimi- role in regulating epidermal differentiation. J Invest Dermatol 134, nation of suppressor of fused reveals an essential repressor function 1981–1990 (2014) in the mammalian hedgehog signaling pathway. Dev Cell 10,187– 33. Y. Lee, H.L. Miller, H.R. Russell, K. Boyd, T. Curran, P.J. 197 (2006) McKinnon, Patched2 modulates tumorigenesis in patched1 hetero- . J. Xie, M. Murone, S.M. Luoh, A. Ryan, Q. Gu, C. Zhang, J.M. zygous mice. Cancer Res 66,6964–6971 (2006) Bonifas, C.W. Lam, M. Hynes, A. Goddard, A. Rosenthal, E.H. 34. C. Adolphe, J.P. Junker, A. Lyubimova, A. van Oudenaarden, B. Epstein Jr., F.J. de Sauvage, Activating smoothened mutations in Wainwright, Patched receptors sense, interpret, and establish an sporadic basal-cell carcinoma. Nature 391,90–92 (1998) epidermal hedgehog signaling gradient. J Invest Dermatol 137, 19. S.C. Peterson, M. Eberl, A.N. Vagnozzi, A. Belkadi, N.A. 179–186 (2016) Veniaminova, M.E. Verhaegen, C.K. Bichakjian, N.L. Ward, A.A. Dlugosz, S.Y. Wong, Basal cell carcinoma preferentially 35. L.V. Goodrich, L. Milenkovic, K.M. Higgins, M.P. Scott, Altered arises from stem cells within hair follicle and mechanosensory neural cell fates and medulloblastoma in mouse patched mutants. niches. Cell Stem Cell 16,400–412 (2015) Science 277, 1109–1113 (1997) Patched-2 functions to limit Patched-1 deficient skin cancer growth 36. M. Aszterbaum, J. Epstein, A. Oro, V. Douglas, P.E. LeBoit, M.P. C. Hauser-Kronberger, A.N. Ermilov, M.E. Verhaegen, C.K. Bichakjian, A.A. Dlugosz, W. Nietfeld, M. Sibilia, H. Lehrach, C. Scott, E.H. Epstein Jr., Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous Wierling, F. Aberger, Hedgehog-EGFR cooperation response genes knockout mice. Nat Med 5,1285–1291 (1999) determine the oncogenic phenotype of basal cell carcinoma and 37. J. Xie, M. Aszterbaum, X. Zhang, J.M. Bonifas, C. Zachary, E. tumour-initiating pancreatic cancer cells. EMBO Mol Med 4, Epstein, F. McCormick, A role of PDGFRalpha in basal cell carci- 218–233 (2012) noma proliferation. Proc Natl Acad Sci U S A 98, 9255–9259 47. J.K. Chen, J. Taipale, M.K. Cooper, P.A. Beachy, Inhibition of (2001) hedgehog signaling by direct binding of cyclopamine to smooth- 38. J. Ericson, S. Morton, A. Kawakami, H. Roelink, T.M. Jessell, Two ened. Genes Dev 16, 2743–2748 (2002) critical periods of sonic hedgehog signaling required for the speci- 48. B.L. Allen, J.Y. Song, L. Izzi, I.W. Althaus, J.S. Kang, F. Charron, fication of motor neuron identity. Cell 87,661–673 (1996) R.S. Krauss, A.P. McMahon, Overlapping roles and collective re- 39. H. Sasaki, C. Hui, M. Nakafuku, H. Kondoh, A binding site for Gli quirement for the coreceptors GAS1, CDO, and BOC in SHH path- proteins is essential for HNF-3beta floor plate enhancer activity in way function. Dev Cell 20,775–787 (2011) transgenics and can respond to Shh in vitro. Development 124, 49. L. Izzi, M. Levesque, S. Morin, D. Laniel, B.C. Wilkes, F. Mille, 1313–1322 (1997) R.S. Krauss, A.P. McMahon, B.L. Allen, F. Charron, Boc and Gas1 40. T. Reya, A.W. Duncan, L. Ailles, J. Domen, D.C. Scherer, K. each form distinct Shh receptor complexes with Ptch1 and are re- Willert, L. Hintz, R. Nusse, I.L. Weissman, A role for Wnt signal- quired for Shh-mediated cell proliferation. Dev Cell 20,788–801 ling in self-renewal of haematopoietic stem cells. Nature 423,409– (2011) 414 (2003) 50. K. Fujii, H. Ohashi, M. Suzuki, H. Hatsuse, T. Shiohama, H. 41. H. Damhofer, V.L. Veenstra, J.A. Tol, H.W. van Laarhoven, J.P. Uchikawa, T. Miyashita, Frameshift mutation in the PTCH2 gene Medema, M.F. Bijlsma, Blocking hedgehog release from pancreatic can cause nevoid basal cell carcinoma syndrome. Familial Cancer cancer cells increases paracrine signaling potency. J Cell Sci 128, 12,611–614 (2013) 129–139 (2015) 51. W.A. O'Hara, W.J. Azar, R.R. Behringer, M.B. Renfree, A.J. Pask, 42. R.B. Corcoran, M.P. Scott, Oxysterols stimulate sonic hedgehog Desert hedgehog is a mammal-specific gene expressed during tes- signal transduction and proliferation of medulloblastoma cells. ticular and ovarian development in a marsupial. BMC Dev Biol 11, Proc Natl Acad Sci U S A 103,8408–8413 (2006) 72 (2011) 43. M.F. Bijlsma, H. Damhofer, H. Roelink, Hedgehog-stimulated che- 52. S. Azzi, L. Treps, H.M. Leclair, H.M. Ngo, E. Harford-Wright, J. motaxis is mediated by smoothened located outside the primary Gavard, Desert hedgehog/Patch2 Axis contributes to vascular per- cilium. Sci Signal 5, ra60 (2012) meability and angiogenesis in glioblastoma. Front Pharmacol 6, 44. M.F. Bijlsma, C.A. Spek, D. Zivkovic, S. van de Water, F. Rezaee, 281 (2015) M.P. Peppelenbosch, Repression of smoothened by patched- 53. A. Okada, F. Charron, S. Morin, D.S. Shin, K. Wong, P.J. Fabre, M. dependent (pro-)vitamin D3 secretion. PLoS Biol 4, e232 (2006) Tessier-Lavigne, S.K. McConnell, Boc is a receptor for sonic 45. J. Briscoe, Y. Chen, T.M. Jessell, G. Struhl, A hedgehog-insensitive hedgehog in the guidance of commissural axons. Nature 444, form of patched provides evidence for direct long-range morphogen 369–373 (2006) activity of sonic hedgehog in the neural tube. Mol Cell 7,1279– 54. N. Basset-Seguin, H.J. Sharpe, F.J. de Sauvage, Efficacy of hedge- 1291 (2001) hog pathway inhibitors in basal cell carcinoma. Mol Cancer Ther 46. M. Eberl, S. Klingler, D. Mangelberger, A. Loipetzberger, H. 14,633–641 (2015) Damhofer, K. Zoidl, H. Schnidar, H. Hache, H.C. Bauer, F. Solca, http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Cellular Oncology Springer Journals

Patched-2 functions to limit Patched-1 deficient skin cancer growth

Free
11 pages

Loading next page...
 
/lp/springer_journal/patched-2-functions-to-limit-patched-1-deficient-skin-cancer-growth-aeXnUbKH04
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Biomedicine; Cancer Research; Biomedicine, general; Pathology; Oncology
ISSN
2211-3428
eISSN
2211-3436
D.O.I.
10.1007/s13402-018-0381-9
Publisher site
See Article on Publisher Site

Abstract

Purpose Basal cell carcinoma (BCC) is one of the most common skin cancers, and is typically driven by an aberrantly activated Hedgehog (Hh) pathway. The Hh pathway is regulated by interactions between the Patched-1 (Ptch1) and Smoothened (Smo) receptors. Smo is an activating receptor and is subject to inhibition by Ptch1. Following ligand binding to Ptch1, its inhibitory action is relieved and pathway activation occurs. This receptor interaction is pivotal to restraining uncontrolled cellular growth. Both receptors have been found to be frequently mutated in BCCs. Ptch2 is a Ptch1 paralog that exhibits overlapping functions in both normal development and tissue homeostasis. As yet, its contribution to cancer growth is poorly defined. Here we set out to assess how Ptch2 inhibits BCC growth. Methods We used several in vitro readouts for transcriptional and chemotactic Hh signaling in BCC-derived ASZ001 cells, and a novel xenograft model to assess in vivo BCC tumor growth. Gene editing by TALEN was used to untangle the different Ptch2- dependent responses to its ligand sonic hedgehog (Shh). Results We first defined the signaling competence of Ptch2 in Ptch1-deficient ASZ001 cells in vitro, and found that Ptch2 ligand binding drives their migration rather than eliciting a transcriptional response. We found that subsequent targeting of Ptch2 abrogated the chemotaxic effect. Next, we tested the contribution of Ptch2 to in vivo tumor growth using a xenograft model and found that reduced Ptch function results in increased tumor growth, but that selective pressure appatently acts against complete Ptch2 ablation. Conclusions We conclude that like Ptch1, Ptch2 exerts a tumor-suppressive function in BCC cells, and that after targeting of both paralogs, ligand-independent activation of the Hh pathway contributes to tumor growth. . . . Keywords Basal cell carcinoma Hedgehog signaling Patched Smoothened Electronic supplementary material The online version of this article (https://doi.org/10.1007/s13402-018-0381-9) contains supplementary material, which is available to authorized users. 1 Introduction * Maarten F. Bijlsma The Hedgehog (Hh) pathway is not only crucial to many in- m.f.bijlsma@amc.uva.nl ductive events in developing embryos and to the maintenance Laboratory for Experimental Oncology and Radiobiology, Center for of tissue integrity in adult organisms, but also to the initiation Experimental and Molecular Medicine, Cancer Center Amsterdam and progression of tumors [1, 2]. Hh pathway regulation is and Academic Medical Center, Meibergdreef 9, 1105AZ, primarily mediated by the transmembrane proteins Patched-1 Amsterdam, The Netherlands (Ptch1) and Smoothened (Smo) [3, 4]. Ptch1 is the main re- Oncode Institute, Academic Medical Center, ceptor for the pathway-activating ligand sonic hedgehog (Shh) Amsterdam, The Netherlands [5]. In the absence of this ligand, Ptch1 actively represses Smo Present address: Tumor Immunology Lab, Radboud Institute for to keep the pathway inactive [6]. In the presence of Shh the Molecular Life Sciences, Nijmegen, The Netherlands inhibitory action of Ptch1 is alleviated through its Present address: Biotech Research & Innovation Centre, relocalization, and Smo is free to signal to downstream path- Copenhagen, Denmark way components [7–9]. This intracellular signaling cascade Department of Pathology, Academic Medical Center, can induce a variety of responses such as transcription factor Amsterdam, The Netherlands activation and cytoskeleton remodeling to mediate chemotax- Department of Medical Oncology, Cancer Center Amsterdam and is [10, 11]. In cancer, two different mechanisms responsible Academic Medical Center, Amsterdam, The Netherlands V. Veenstra et al. for aberrant pathway activation can be discerned [12]. The there is little selective pressure on PTCH2. Here, we asked first mechanism entails excessive production of Shh ligand whether absence of repressive PTCH1 action enhances the by tumor cells, which subsequently acts in an autocrine or role of PTCH2 in Shh ligand perception and subsequent path- paracrine manner to provide tumor-promoting signals [13, way activation, rendering cells highly sensitive to Shh ligand, 14]. The second mechanism entails genetic aberrations in Hh or whether the contribution of PTCH2 to tumor growth is pathway components that cause Hh pathway activation solely dependent on its tumor suppressor function via the sup- [15–18]. These aberrations typically include inactivating mu- pression of Smo activity. Another question to be answered is tations in Ptch1, crippling its inhibitory action on Smo, or whether there is selective pressure against ablation of both activating mutations in Smo that render it insensitive to PTCH paralogs, which might explain the low incidence of Ptch1 inhibition. Through these latter activating mutations PTCH2 mutations observed in patient samples. We used pathway activity is induced cell autonomously, rendering the in vitro and in vivo systems, in conjunction with gene editing, cells independent of Shh ligands produced by themselves or to untangle the different responses of BCC cells to Shh ligand their surroundings. Given these considerations, Ptch1 is con- and show that deficiency for both PTCH paralogs accelerates sidered a bona fide tumor suppressor. One human cancer type tumor growth. that relies on mutations in Hh pathway components is basal cell carcinoma (BCC), the most prevalent skin cancer. Oncogenic mutations in Ptch1 and Smo are since long known to drive BCC, and mouse models have underscored the notion 2 Materials and methods that the development of this malignancy relies heavily on these mutations [19, 20]. A recent study on the mutational 2.1 Cell culture profiles of 126 BCCs has revealed PTCH1 mutations in 73% of the cases [21]. PANC-1 cells (ATCC, Manassas, VA) and mouse embry- −/− +/+ The current dogma on Hh pathway regulation holds that onic fibroblasts (Ptch1 and Ptch1 MEFs from Dr. Ptch1 is the principal receptor for Shh, and that other receptors Scott, Stanford University [35]) were cultured in high- involved in Shh binding like Cdon, Boc and Gas1 function as glucose DMEM containing 8% fetal bovine serum coreceptors [22, 23]. A paralog of Ptch1 is Patched-2 (Ptch2) (FBS), L-glutamine, penicillin and streptomycin (all from [24–26], and this paralog is thought to complement some Lonza, Basel, Switzerland) according to routine cell cul- Ptch1 functions [27–29]. It has been found, however, that ture procedures. ASZ001 cells [36, 37]werecultured in Ptch2 does not act as an equally strong regulator of the path- 154CF keratinocyte medium (Life Technologies) supple- −/− way. For instance, Ptch2 embryos have been found to be mented with 50 μM CaCl, penicillin and streptomycin, viable and to develop normally, and that in a genetically and 2% chelex-treated FCS. Cells were screened for my- Ptch1-deficient system Ptch2 cannot fully compensate for loss coplasma monthly by PCR. of the other homolog [29–31]. However, Ptch2 deficiency does exacerbate the skin tumor phenotype in partially Ptch1 deficient mice by deregulating epidermal lineage differentia- 2.2 Quantitative RT-PCR tion, and it has been found that the absence of both paralogs affects skin maintenance [32, 33]. Subsequent detailed analy- Cells were lysed in Trizol (Invitrogen) after which RNA was ses of Hh pathway target expression gradients in the epidermis isolated according to the manufacturer’s protocol. cDNA was revealed that full Ptch deficiency results in a uniformly high synthesized using Superscript III (Invitrogen) and random pathway activation [34]. Recent work in embryonic stem cells primers (Invitrogen). Quantitative real-time RT-PCR (qRT- has shown that Ptch2 is required for ligand perception in the PCR) was performed using SYBR green (Roche, Basel, absence of Ptch1 [27]. Intriguingly, in a Ptch1-deficient mouse Switzerland) on a Lightcycler LC480 II (Roche). Relative model of Hh pathway-driven BCC it was found that the tu- gene expression levels were calculated using the comparative mors preferentially arise from locations close to Shh sources threshold cycle (Ct) method and values were normalized to the [19]. These latter observations imply that in the absence of reference gene Gapdh. The primer sequences used were: Ptch1 at least some responsiveness to Shh remains and, Gapdh 5 ’ CTCATGACCACAGTCCATGC and 3 ’ therefore, that Shh is a likely candidate to mediate Ptch2 CACATTGGGGGTAGGAACAC; Gli1 5’ ATAGGGTC activity. TCGGGGTCTCA and 3’ CGGCTGACTGTGTAAGCAGA PTCH2 mutations are relatively rare events. A large-scale ; Ptch1 5’ GCTACGACTATGTCTCTCACATCAACTand 3’ genetic analysis revealed that only 14 out of 126 BCC cases GGCGACACTTTGATGAACCA; Ptch1 (exon 2) 5’ carried mutations in both PTCH1 and PTCH2 and that only 4 CTGTGGCTGAGAGCGAAGTT and 3’ AGCTCCTC cases exclusively carried PTCH2 mutations [21]. These ob- CACGTTGGTCT; Ptch2 5’ GCGTACACCTCCCA servations suggest that in the absence of functional PTCH1, GATGTT and 3’ GGAACCCCTGATTTGTAGCA. Patched-2 functions to limit Patched-1 deficient skin cancer growth 2.3 FACS analysis 2.6 Cell viability assays Cells were harvested using a trypsin-EDTA solution (Lonza) Cells were seeded at 2000 cells/well in 0.5% or 2% FCS and washed in FACS buffer (PBS containing 1% FBS). and after 3 h treated with the indicated compounds (see also Hybridoma supernatants containing either anti-Shh antibody [42]). After 4 d, MTT was added and dye reduction was 5E1 [38] or anti-Myc antibody 9E10 (isotype) were diluted assessed after a 4 h incubation period. Background 1:5 in FACS buffer and incubated for 30 min at 4 °C. A (10 mM H O treated cells) levels were subtracted and the 2 2 secondary APC labeled anti-mouse (BD, 550826) antibody values obtained from the control treated cells were set to 1, was used at a dilution of 1:500. After washing, the cells were after which the experimental data were normalized to the resuspended in FACS buffer containing 1 μg/ml propidium controls. iodide (PI) (Sigma) and subjected to flow cytometry using a FACSCanto II machine (BD, Franklin Lakes, NJ, USA). The 2.7 Transwell migration assays data obtained were analyzed using FlowJo 7 software (Tree Star, Ashland, OR, USA). Migration assays were performed as previously reported [10]. Briefly, cells were labeled with 10 μM CellTracker Green (Invitrogen) according to the manufacturer’s protocol. After 2.4 Immunofluorescence labeling, the cells were detached using 5 mM EDTA, resus- pended in serum free medium and transferred to FluoroBlok ASZ001 cells were grown on glass coverslips, starved for 2 d, Transwell inserts (BD Falcon) at a density of approximately and fixed using 4% formaldehyde. Following blocking and 5×10 cells per insert. Chemoattractant was added to the permeabilization in 5% goat serum/phosphate-buffered saline bottom compartments of the Transwell plates and GFP- with 0.1% Triton X100 (PBS-T), a primary antibody directed spectrum fluorescence in the bottom compartments was mea- against acetylated α-tubulin (Sigma) was added at 1:2000 and sured using a Synergy HT plate reader (BioTek, Winooski, incubated for 1 h at room temperature or overnight at 4 °C. VT, USA) every 2 min during approximately 3 h. Next, an Alexa 488 conjugated anti-mouse secondary anti- Background fluorescence was measured in time from a well body (Invitrogen) was added at 1:2000 and incubated for 1 h containing only medium and these values were subtracted at room temperature. Finally, the coverslips were mounted from all other measurements. The ‘no attractant’ control was using ProLong Gold (Invitrogen) and images were captured used to measure baseline cell movements for every experi- using a Zeiss AxioVert microscope. For Smo ciliary localiza- mental condition. These values were subtracted from those tion, cells were transfected with Myc-tagged Smo using PEI obtained in the presence of chemoattractants in the bottom 48 h prior to starvation and ShhN stimulation (1:4 diluted compartments of the Transwell inserts. The resulting data supernatant from 293 T cells). yielded specific migration values towards a given attractant. 2.5 GBS-GFP reporter construct and cell line 2.8 Gene editing and transfections establishment The pCTIGTALEN expression vector was used as described A concatemerized 8 × 3’GLI binding site sequence (GBS) before [27]. A pair of TALEN constructs was modified so that was isolated from the pδ51 GBS-luciferase reporter construct one construct co-expressed GFP and the other tdTomato, [39] by PCR and cloned into a lentiviral pRRL TOP-d2GFP allowing for selection for both 5′ and 3′ targeting constructs reporter vector [40] as reported before [41]. MEFs and by FACS sorting. The constructs were designed using Golden ASZ001 cells were transduced with the GBS-GFP reporter Gate cloning into pCTIG employing the following variable construct, starved in 0.5% FCS containing medium and stim- domain architectures: 5’ NN NN HD NG NG HD NN NI ulated for 4 d with ShhN conditioned supernatant from 293 T NN HD NG NG NI HD NG NG HD; 3’ NG HD NG NN cells or 200 nM Smo agonist (SAG, EMD Millipore, Billerica, NN NI NG HD HD NG NN HD NI HD HD HD HD. See also MA, USA). The resulting cells were sorted on a BD Supplementary Fig. S4a. FACSAria for GFP expression, after which GFP positive cells ASZ001 cells grown in 3 × 12-well plates were transfected were grown under 8% FCS (MEFs) or 2% FCS (ASZ001) with paired TALEN constructs using PEI. 3 days after trans- + + conditions, resulting in the expected loss of GFP activity. fection GFP /tdTomato cells cells were selected by flow For subsequent analyses, cells were seeded in 24-well plates sorting, seeded in bulk and allowed to recover for 7 d in and treated as indicated in the figure panels. Following treat- T25 flasks. To obtain monoclonal cultures, cells were seed- ment, cells were harvested and the percentages of GFP cells ed at single cell/well densities in 2 × 96-well plates. In 35 of were determinedby flow cytometryonaFACSCantoII the 192 wells cells grew out. These cells were genotyped machine. through the sequencing of PCR products spanning the V. Veenstra et al. TALEN binding sites (see also Supplementary Fig. S5). cilium, was confirmed in nearly all ASZ001 cells analyzed Screening was performed on genomic DNA by PCR using (Fig. 1c, right panel) [7]. The primary cilium is an antenna- primers flanking the TALEN binding sites: 5 ’ like protrusion from the cell membrane that is shaped by the AAGGCACAGGGAAAGAGAGTT; 3’ ACTTGCCT microtubule cytoskeleton. A dynamic localization of Hh AGCTTGCACAATG and subsequent digestion of the pathway components in and out of this organelle is required PCR products with AccI. Genomic DNA from monoclonal for pathway regulation, and the primary cilium is required lines that exhibited loss of the restriction site were TOPO for appropriate ligand perception. Thus, at least part of the cloned (Thermo Fisher) and Sanger sequenced. immediate signaling machinery for Shh is intact in these cells. Overxpression of Hedgehog pathway components was ac- The ciliary localization of exogenously overexpressed Smo in complished using wild-type Smo (SmoWT) and ciliary locali- response to the addition of Shh was subsequently assessed by zation domain mutated Smo (SmoCLD) constructs in pCS107 microscopy. Despite a high baseline percentage of cells with obtained from Dr. Jeremy Reiter [7, 43]and awild-type Ptch1 Smo localized in the cilium, as expected from a Ptch1-deficient construct in pcDNA3.1 obtained from Dr. Matthew Scott (see system and overexpression of Smo, we found that the addition loop2 also ref. [44]). mPtch1Δ is a Ptch1 form that lacks the of ligand resulted in a moderate increase in this number (Fig. second extracellular loop required for ligand binding and only 1d). As a control, a form of Smo that cannot localize to the exerts Smo-repressive functions [45]. cilium (SmoCLD), was used [7]. Taken together, our results imply that in the absence of Ptch1, the Shh ligand is still per- 2.9 Xenografting of ASZ001 cells ceived by the cells. scid tm1Wjl NOD.Cg-Prkdc Il2rg /SzJ (NSG) mice were bred in- 3.2 ASZ001 cells respond to hedgehog by chemotaxis 5 5 house. The animals were grafted with 1 × 10 or 5 × 10 ASZ001 cells in Matrigel [46]. All experiments were per- Typically, relocalization of Smo to the primary cilium results formed according to procedures approved by the animal ex- in activation of the downstream pathway leading to transcrip- periment ethical committee (LEX237). Tumor growth was tional responses. However, when Gli transcription factor ac- monitored weekly and the experiments were ended when ul- tivity was measured in these cells using a stably integrated ceration was observed. Of note, ulceration was typically ob- GLI-binding site GFP reporter construct, only minimal re- served before the tumors reached a humane endpoint volume sponses to exogenous pathway activators were observed, al- (1000 mm ). Finally, the tumors were harvested and processed though the baseline pathway activity was high (Fig. 2a,b). In for paraffin embedding and subsequent histopathological ex- comparison, MEFs proficient for Ptch1 showed robust re- amination, as well as for RNA extraction. sponses to the pathway activators. These results were con- firmed by qRT-PCR analysis of target genes, through which minor responses were observed in ASZ001 cells compared to 3 Results MEFs (Fig. 2c,d, for baseline pathway activity see Fig. 1a). Under high serum proliferation conditions, a strong reduction 3.1 Ptch1-deficient basal cell carcinoma cells perceive in basal reporter activity in the ASZ001 cells was observed hedgehog signalling (4% GFP cells, data not shown), indicating that the Hh path- way in these cells is amenable to at least some degree of As a model for Ptch1-deficiency driven skin cancer, we used regulation, most likely through cell cycle progression and the ASZ001 cell line. This cell line has been established from subsequent primary cilium loss. Switching these cells back +/- an irradiation-induced tumor in a Ptch1 mouse, and has to low-serum conditions resulted in a regain of GFP reporter subsequently undergone Ptch1 loss of heterozygosity ( [36] expression to the levels shown in Fig. 2a,b. In order to test and Fig. 1a). The current working model for Hedgehog (Hh) whether the mitogenic Hh response can be driven by exoge- pathway regulation dictates that loss of Ptch1 suffices for full nously added ligand in ASZ001 cells, we treated them with Hh pathway activation and that no additional ligand is re- ShhN or control (Ctrl/GFP) supernatants. By doing so, we quired for this. Indeed, the Hh pathway was found to be acti- observed only marginal proliferative responses to Shh com- vated in these cells as evident from abundant transcription of paredtothe control(SupplementaryFig.S2a). In addition, the remaining exon on the targeted Ptch1 allele, which itself is we found that inhibitors of Hh pathway components and a target of the activated pathway (Fig. 1a). We also found that related signaling molecules were relatively ineffective in the cells did not produce ligand that could cell autonomously restraining ASZ001 cell proliferation, as evident from their activate the pathway (Fig. 1b; SHH expressing pancreatic overall high IC50 values (Supplementary Fig. S2b,c) [42, cancer PANC-1 and Shh transfected ASZ001 cell controls 44, 47]. are shown in Supplementary Fig. S1). One prerequisite As shown previously in fibroblasts and neural develop- for Hh ligand responsiveness, i.e., the presence of a primary mental models, Shh ligand is also able to induce chemotaxis, Patched-2 functions to limit Patched-1 deficient skin cancer growth a b c d 100 GFP 100 0.3 mPtch1 1 ASZ001 α-SHH ShhN mPtch1 exon2 isotype 75 75 ** 0.2 α-acetylated tubulin 50 50 0.1 25 25 DAPI 0 0 0.0 0 12 3 4 ASZ001 SHH (APC-A log10) expressing Fig. 1 Ptch1-deficient BCC cells perceive Shh ligand. a RNA was were grown on coverslips, transfected with Myc-tagged forms of Smo, isolated from the cells indicated on the X-axis after which qRT-PCR- starved and treated with ShhN (or control; GFP) supernatant diluted 1:4 based expression analysis was performed relative to mGapdh using a from 293 T cells for 1 h. Next, the cells were fixed and stained for Myc Ptch1 exon upstream of the targeted exon (grey bars) and the targeted and acetylated α-tubulin, after which the percentage of transfected (Myc- exon 2 (dark blue bars). Bars indicate means ± SEM (n = 4), n.d. indicates positive) cells with ciliary Smo was quantified. SmoCLD; ciliary locali- not detected (no signal). b ASZ001 cell surface levels of Hedgehog li- zation domain mutated form of Smo [7, 43]. For assessment of the frac- gands were determined by FACS using a 5E1 anti-Shh hybridoma anti- tion of ciliated cells, quantifications from both transfections were pooled body or an isotype control. c ASZ001 cells were grown to confluence on and depicted in the separate grey bar graph. Bars indicate means ± SEM coverslips after which primary cilia were visualized by acetylated α- (n > 50 cells quantified from 2 separate experiments). **p =0.0015 by tubulin staining. Nuclei were counter-stained with DAPI. d ASZ001 cells Mann-Whitney U test irrespective the presence of Hh pathway transcription factors. 3.3 Ptch2 mediates a chemotactic response This response appears to be relatively resilient to the pertur- to hedgehog ligand bation of pathway components [43]. To assess whether the perception of Shh by ASZ001 cells may function to activate Due to the catalytic inhibitory activity of Ptch on Smo, a low this chemotactic response, we assessed the migratory capacity number of Ptch molecules suffices to suppress Hh pathway of these cells. Pathway activators were used in a modified activity [6] and, therefore, highly effective targeting is re- Boyden chamber assay to quantitatively measure chemotaxis. quired to study the consequences of Ptch loss. We found that A robust migratory response was observed to both recombi- lentiviral shRNA-mediated Ptch2 silencing using five differ- nant ShhN (5 nM) and SAG (200 nM; Fig. 2e). This response ent sequences and puromycin selection did not result in effec- could be blocked with cyclopamine (5 μM), confirming that tive targeting (Supplementary Fig. S3). Therefore, we turned this form of chemotaxis requires Smo. The subsequent use of to transcription activator-like effector nucleases (TALENs) to a Smo agonist (SAG) further underscored the Smo dependen- edit the Ptch2 locus in ASZ001 cells (strategy shown in cy of this migratory response. After exogenous Ptch1 overex- Supplementary Fig. S4). Following transfection, cells were pression, we observed an increased relative responsiveness of FACS-sorted in bulk to ensure the survival of cells expressing the cells to ligand as evident from an increased net migratory both TALENs (Supplementary Fig. S5a). After recovery, the response to Shh (Fig. 2f). A form of Ptch1 that is unable to cells were seeded at single cell density in 2 × 96 well plates Δloop2 bind Shh (Ptch1 ) was found to be inhibitory to the Shh after which 35 of the clones grew out (18%). These single cell stimulated migration [45]. This suggests that the migratory clones were analyzed for editing events and, by doing so, we response, although not strictly dependent on Ptch1, can be found that in 5 of the clones the digest pattern was indicative modulated by the inhibitory activity of Ptch1. Together, these of an editing event (Supplementary Fig. S5b; lines 1, 4, 14, 16, data indicate that despite the absence of Ptch1, ASZ001 cells 17). These lines were TOPO cloned after which 20 clones per have retained Shh chemotactic responsiveness. This notion line were Sanger sequenced to verify the efficiency of editing suggests that other receptors may mediate the chemotactic (Supplementary Fig. S5c). Surprisingly, none of these lines response to Shh. A likely candidate to mediate this response was completely devoid of wild type Ptch2 sequences. In is its paralog Ptch2. cell line 4, one third of the sequences were mutant, whereas expression relative to mGapdh Ptch1+/+ MEFs Ptch1-/- MEFs n.d. ASZ001 n.d. normalized to mode fraction of cells with ciliary Smo (%) SmoWT SmoCLD fraction of cells with a primary cilium (%) V. Veenstra et al. a b c d 50 control 4 100 ASZ001 cells ASZ001 cells +/+ +/+ SAG Ptch1 fibroblasts Ptch1 fibroblasts ShhN 0 0 0 0.1 0.5 0.5 -9 -6 -3 [ShhN] (sup) [SAG] (M log10) e f 4 ASZ001 cells *** 2.5 SAG rShhN cyclopamine 2.0 rShhN + cycl. 1.5 *** 1.0 0.5 *** -1 0 0 0 10 20 30 40 50 Cycle (min x 3) Fig. 2 ASZ001 cells mediate a chemotactic response to Shh. a GBS-GFP presence of 5 μM cyclopamine in both the upper and the bottom com- transduced ASZ001 and Ptch1-proficient fibroblasts were starved and partments of the Transwells. Shaded curves represent the SEM. The treated with the indicated dilutions of ShhN supernatants produced by curves were plotted using a hyperbola function. Measurements from at 293 T cells. For the highest concentration, GFP transfected 293 T super- least 3 replicates are shown. For details see Materials and methods sec- natant was included as a control (Ctrl/GFP 0.5). After 3 d, the GFP tion. The data are plotted as average RFU in bar graphs. Statistical test percentage was assessed by FACS (n = 4 for the ASZ001 cells; n =2 compares rShhN + cyclopamine versus rShhN. The difference in migra- for the MEFs). b As for panel a, using SAG. At the two highest tested tion between no-attractant control and ShhN/SAG is statistically signifi- doses, SAG was toxic to the fibroblasts (n = 3 for the ASZ001 cells; n =4 cant (p <0.001). f ASZ001 cells were transfected with indicated con- for the MEFs). c-d Cells were treated as for panels a-b using 200 nM SAG structs, and after ~24 h the migration response to 5 nM recombinant ShhN or 1:4 diluted ShhN supernatant. After treatment, RNA was isolated and was assessed. Bars indicate means ± SEM from approximately 70 mea- qRT-PCR was performed for mGli1 and mPtch1 (n = 5). *p < 0.05; surements from 2 separate experiments. Differences to the vector control **p < 0.01; ***p < 0.001; determined by t-test, (e) ASZ001 cells were were tested for the mPtch1 transfected condition, and for the mPtch1 Δloop2 seeded in a modified Boyden chamber after which net migration to 5 nM condition against the mPtch1 condition recombinant ShhN or 200 nM SAG was assessed in the absence or LOW in cell lines 1, 14, 16 and 17 two-thirds of sequences were 14, 16 and 17 (denoted Ptch2 ), of which we continued mutant. In addition, we found that the patterns of gene with clone 1. The failure to establish cell lines with com- editing in cell lines 1, 14, 16 and 17 were nearly identical, plete Ptch2 gene editing from otherwise successfully suggesting that amongst the initial 35 cell lines available for transfected cells implies that selective pressure exists restriction digest analysis, only two parental clones showed against full Ptch-deficiency. This notion is in agreement MED significant Ptch2 editing, i.e., clone 4 (denoted Ptch2 ) with the lentiviral shRNA Ptch2 silencing results (see and an additional parental clone which gave rise to clones 1, Supplementary Fig. S3). GFP cells (%) migration (RFUx10 ) Ctrl/ GFP average migration (RFUx10 ) rShhN SAG cyclopamine mGli1 relative to control treated rShhN + cyclopamine migration to rShhN (RFUx10 ) ASZ001 * RFP +/+ Ptch1 *** fibroblasts mPtch1 ** Δloop2 mPtch1 mPtch1 relative to control treated ASZ001 +/+ Ptch1 ** fibroblasts Cells targeted for Ptch2 were found to be transcriptionally 3.4 Ptch2 deficiency accelerates tumor growth unresponsive to ShhN ligand as determined by qRT-PCR (Fig. 3a). Surprisingly, we also found that Ptch2 gene editing Having established that (1) in the absence of Ptch1 Hh ligand resulted inanincrease inbasal transcriptional pathway activity is perceived by Ptch2 only to mediate chemotaxis and (2) that +/+ in vitro when comparing wild-type Ptch2 parental cells with additional targeting of Ptch2 impedes the ability of the cells to +/+ wild-type Ptch2 clone 35, arguing for the use of gene edited perceive Hh ligand, we proceeded to test the consequences of but wild-type cells as controls. If Ptch2 mediates chemotaxis in these signaling outputs for in vivo tumor growth. To this end, response to Hh ligand as was concluded from the data presented ASZ001 cells were injected subcutaneously in immunodefi- in Fig. 2, its targeting should inhibit the chemotactic response. cient mice after which tumor growth was monitored. We Indeed, we observed a strong reduction in chemotaxis to ShhN found that tumors grown from cells targeted for both Ptch LOW MED LOW MED in the Ptch2 and Ptch2 cells,ascomparedtothe paralogs (i.e., Ptch2 and Ptch2 ASZ001 cells) expand- +/+ Ptch2 clone 35 cells (Fig. 3b-c). Baseline chemokinesis ed faster than those grown from fully Ptch2-proficient cells +/+ (i.e., the movement of cells in the absence of attractant; these (i.e., ASZ001 Ptch2 clone 35 cells). The former mice were values are typically subtracted from migration to Shh to yield sacrificed sooner based on ulceration (Fig. 4a). Subsequent net chemotaxis as shown in Fig. 3b)did not differbetween the immunohistochemical analysis for the proliferation marker genotypes (Supplementary Fig. S6a). The residual chemotactic Ki67 confirmed a higher proliferative index in the Ptch2 responsiveness of Ptch2-edited ASZ001 cells was sensitive to gene-edited tumors (Fig. 4b, quantification in Fig. 4c). cyclopamine, showing that the regulation of Shh chemotaxis is Assessment of tumor histology by a pathologist confirmed dependent on Smo (Supplementary Fig. S6b). a cutaneous source of the tumor cells, but also revealed a Also, receptors other than Ptch1 and Ptch2 have been im- squamous rather than basal histology for both genotypes, pos- plicated in the perception of Shh for chemotactic responses sibly owing to extensive in vitro culturing of the cells prior to LOW MED [23, 48, 49]. In order to test whether putative Hh receptors in grafting. The Ptch2 and Ptch2 derived tumors showed addition to Ptch2 may mediate Shh chemotaxis in ASZ001 more keratin depositions (eosin rich pink areas in Fig. 4d; cells, we targeted the Cdon and Boc receptors by lentiviral quantification in Fig. 4e). To determine whether the in vivo shRNA delivery and found that, despite a modest knockdown accelerated proliferation of these Ptch2 gene-edited cells efficiency, migration was hampered (Supplementary Fig. could be explained by additional activation of the Hh pathway S6c,d). This implies that in addition to Ptch2, other candidate over the already high level caused by the Ptch1-deficiency, the receptors for Shh may elicit chemotactic responses. expression of Hh pathway target genes was measured by qRT- LOW Nevertheless, given that Ptch2 is incapable of mediating a PCR. Indeed, these were found to be elevated in the Ptch2 MED robust transcriptional response to ShhN (Fig. 2a-d and 3a), and Ptch2 ASZ001 tumors (Fig. 4f)and arelikelyrespon- we conclude that the ligand-dependent contributions of sible for the observed increase in tumor growth rates. We Ptch2 in Ptch1-deficient cells in vitro are confined to the che- hypothesize that the discrepancy in Hh pathway activation motactic response. in vitro and in vivo following Ptch2 targeting results from +/+ a b Ptch2 clone 35 c MED Ptch2 clone 4 4 LOW control 1.1 10.0 Ptch2 clone 1 ShhN p<0.0001 *** 1.0 7.5 0.9 NS ** 2 0.8 NS 5.0 NS NS *** NS 0.7 NS NS 2.5 0.6 0.5 0 0 0 25 50 75 100 Cycle (min x 3) +/+ Fig. 3 Ptch2 is required for Shh chemotaxis. a ASZ001 Ptch2 parental parental cells. *p <0.05; **p <0.01;***p <0.001; determined by t-test, +/+ MED LOW cells and TALEN genome edited ASZ001 cells were stimulated with (b) Ptch2 , Ptch2 , and Ptch2 cells were seeded in a modified ShhN as depicted in Fig. 2c-d. Target gene (mPtch1) transcript analysis Boyden chamber after which specific migration (net chemotaxis) to 5 nM was performed. Genotypes and clones are indicated on the X-axis and by recombinant ShhN was assessed as in Fig. 2e. Measurements from 4 the blue-shaded (ShhN) bars. Blue asterisks denote statistical compari- replicates in 2 experiments are shown. The indicated p-values were de- sons of ShhN-treated cells and control treated cells of the same genotype. termined by one-way ANOVA. c Schematic representation of Fig. 3b. +/+ Grey asterisks indicate significance compared to ASZ001 Ptch2 Significances were determined using Mann-Whitney U test mPtch1 expression relative to control treated parental cells +/+ Ptch2 parental +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 LOW Ptch2 clone 14 LOW Ptch2 clone 16 net chemotaxis to ShhN (RFUx10 ) normalized migration (RFU) +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 *** *** V. Veenstra et al. 5 5 1000 1x10 cells grafted 800 5x10 cells grafted +/+ +/+ ASZ001 Ptch2 clone 35 ASZ001 Ptch2 clone 35 MED MED ASZ001 Ptch2 clone 4 ASZ001 Ptch2 clone 4 800 LOW LOW ASZ001 Ptch2 clone 1 ASZ001 Ptch2 clone 1 0 0 01234567 01234567 time post-graft (weeks) time post-graft (weeks) +/+ ASZ001 Ptch2 clone 35 MED b c e ASZ001 Ptch2 clone 4 1.0 30 ** * LOW *** ASZ001 Ptch2 clone 1 *** ** *** *** ns -5 0.5 ns -10 0.0 0 -15 Fig. 4 Patched deficiency accelerates tumor growth. a 1×10 and 5 × of positive nuclei per optical field (n ≥ 7). The indicated p-values were 10 ASZ001 cells of the indicated genotypes were subcutaneously grafted determined using Mann-Whitney U test. d Histology of tumors grown in immune deficient mice in PBS/Matrigel. Subsequent tumor growth from the indicated genotypes revealed by hematoxylin and eosin staining. was monitored and tumors were harvested after signs of ulceration. In Scale bar: 1 mm. e Automated quantification of pink eosin-stained fields each group 4 mice were grafted with 2 tumors, yielding sample sizes of 8. shown in d. f After RNA extraction from the harvested tumors transcript b After harvesting, the tumors were processed for Ki67 analyses for the indicated genes were performed (n = 16). Significances immunohistochemistry. Scale bars: 200 μm. c Automated quantification were determined using Mann-Whitney U test (see Fig. 3) environmental signals and/or mechanical properties that are of ligand in developmental models [27]. Also, Ptch1 and only present in vivo, which feed into the signaling cascade Ptch2 have been shown to exhibit overlapping functions in downstream of Smo. From the increased baseline pathway Hh pathway-dependent skin development and maintenance activity and associated tumor growth following Ptch2 [32, 34]. The exact contribution of Ptch2 to an established targeting in vivo, we conclude that Ptch2 exerts ligand- cancer type such as basal cell carcinoma (BCC) has, however, independent pathway inhibitory functions that make it a tumor remained unclear, despite clinical data that suggest a tumor suppressor, but one that is also required and essential for can- suppressor function [50]. Whether such a putative tumor sup- cer cell viability. pressor function of Ptch2 may be uncoupled from its ligand binding function is currently unknown and is possibility com- plicated by the fact that these functions are connected [45]. 4 Discussion Here, we have untangled these functions by first delineating the signaling capabilities of Ptch2 in BCC cells in vitro and The current model of Hedgehog (Hh) pathway regulation subsequently testing the consequences of Ptch2 perturbation for in vivo BCC tumor growth. Using migration assays we holds that Ptch1 acts as the main receptor for Shh, and that it −/− serves as the master switch for downstream pathway regula- found that in the parental Ptch1 ASZ001 cells the only response to Hh ligand was chemotactic. After Ptch2 targeting tion [3, 6]. Recent work has shown, however, that the concept of Ptch1 as key Shh receptor needs revision. In the absence of by TALEN we found that this chemotactic response was se- verely hampered. Therefore, we conclude that the increased Ptch1, Ptch2 has been found to be required for the perception tumor size (mm ) LOW MED +/+ Ptch2 Ptch2 Ptch2 clone 1 clone 4 clone 35 Ki67+ nuclei (fraction) +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 LOW MED +/+ Ptch2 Ptch2 Ptch2 clone 1 clone 4 clone 35 tumor size (mm ) keratin surface (%) +/+ Ptch2 clone 35 MED Ptch2 clone 4 LOW Ptch2 clone 1 expression relative to mGapdh (log2) Gli1 Ptch1 Ptch2 Patched-2 functions to limit Patched-1 deficient skin cancer growth tumor growth observed following targeting of Ptch2 is likely is warranted. In addition, it remains to be established whether unrelated to ligand perception and that the tumor suppressive the identified tumor suppressive action of Ptch2 is also rele- action of Ptch2 very likely depends on its baseline inhibitory vant for non- or pre-malignant cultured keratinocytes. activity towards Smo, which for an as yet unknown reason Previously, we have shown that ciliary relocalization of only becomes apparent in vivo. Smo in response to Shh is not required for Shh chemotaxis, As mentioned above, Ptch2 mutations are uncommon in and suggested that chemotactic and transcriptional responses BCC [21]. This is at apparent odds with its similarities and to Shh may represent separate phenomena resulting from dis- overlapping signaling roles with Ptch1. However, as we find tinct intracellular mechanisms [43]. Our current data indicate that Ptch2 has a very limited ligand-perceiving role in BCC, that these responses are indeed not mutually exclusive, and we assume that its main function as a tumor suppressor is that Smo localization to the cilium occuring in the absence of −/ overshadowed by the loss of Ptch1, which is a stronger inhib- robust downstream transcriptional signaling (i.e., in Ptch1 − +/+ itor of the Hh pathway [31, 33]. This notion is also supported ;Ptch2 cells) may initiate a chemotactic response. In fact, by the high incidence of mutations in for instance TP53. TP53 the convergence of both pathways at several points within the is a much stronger tumor suppressor than Ptch2, and it is more signaling cascade (Smo, the primary cilium) seems to suggest easily perturbed given that only one allele requires a mutation that they are both part of a relatively conserved pathway. This to yield an oncogenic event. Mutations in TP53 are, therefore, holds promise for the application of currently available drugs more advantageous and more regularly achieved than PTCH2 against BCC such as erivedge/vismodegib [54]. Given the fact LOH which, in turn, could explain the low mutation rate of that vismodegib acts to inhibit Smo, rather than downstream PTCH2. Another explanation for the paucity of PTCH2 mu- signaling components or transcription factors, it may be effec- tations in human BCC follows directly from our observation tive against both the chemotactic and the transcriptional ligand that full Ptch2 gene ablation could not be achieved, despite responses, as well as against baseline pathway activity. This selection. It thus appears that a low level of Ptch2 activity is drug may, therefore, turn out to be effective against BCC required for cancer cell viability, at least in vitro. Modest per- irrespective the mutational status of other Hh ligand receptors. turbations of Ptch2 are apparently tolerated, resulting in suffi- Acknowledgements We would like to thank Dr. Ervin Epstein and Dr. cient Hh pathway upregulation in vivo to boost tumor growth, Heidi Hahn for the ASZ001 cells, and Dr. Matthew Scott for providing as evident from the enhanced tumor growth found for the the (Ptch1) MEFs. We are thankful to Brock Roberts and Henk Roelink MED Ptch2 cells. How these two counteracting activities are for assisting in the TALEN genome editing, Tom van Leusden for tech- balanced remains to be determined, but it is possible that they nical assistance, and Dr. Louis Vermeulen for proofreading. are driven by distinct signaling functions of Ptch2 (Smo an- tagonist, chemotaxis receptor, dependence receptor), or by a Funding This work was supported by a KWF Dutch Cancer Society cancer-specific context. In addition, it is possible that rare Research Grant (UVA 2012–5607) to MFB and HVL. tumors that are currently not recognized to be Hh-driven may rely on mutations in Ptch2 if that serves as the dominant Compliance with ethical standards Hh receptor in the tissue of origin. For instance, the testis- specific expression of Desert Hedgehog (Dhh) and Ptch2 Conflicts of interest MFB and HVL have received research funding from Celgene. This party was not involved in drafting the manuscript. and their role in glioblastoma-endothelium crosstalk hints to this possibility [24, 51, 52]. We found that some chemotactic responsiveness was retained by the Ptch-edited ASZ001 cells. Open Access This article is distributed under the terms of the Creative We hypothesize that this responsiveness may result from can- Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, didate coreceptors for Shh other than Ptch2, such as Cdon, distribution, and reproduction in any medium, provided you give Boc or Gas1, or possibly Smo itself [53]. Indeed, we found appropriate credit to the original author(s) and the source, provide a link that the Cdon and Boc receptors may also contribute to che- to the Creative Commons license, and indicate if changes were made. motactic Shh signaling. Given the fact that the only experi- mental variable in the xenografting experiments was the level of Ptch2 we are, however, reluctant to draw firm conclusions References on its signaling contributions in vivo. One caveat of our study is that it is yet unclear to what 1. J. Briscoe, P.P. Therond, The mechanisms of hedgehog signalling extent our findings in murine cells can be translated to the and its roles in development and disease. Nat Rev Mol Cell Biol 14, 416–429 (2013) human situation. The ASZ001 cell line represents a powerful 2. S. Teglund, R. Toftgard, Hedgehog beyond medulloblastoma and exerimental tool given that these cells are unambiguously Hh- basal cell carcinoma. Biochim Biophys Acta 1805,181–208 (2010) dependent, which is important for our experiments, and that 3. D.M. Stone, M. Hynes, M. Armanini, T.A. Swanson, Q. Gu, R.L. many murine-specific genetic tools to study Hh signaling are Johnson, M.P. Scott, D. Pennica, A. Goddard, H. Phillips, M. Noll, available. Nevertheless, validation in a human-derived model J.E. Hooper, F. de Sauvage, A. Rosenthal, The tumour-suppressor V. Veenstra et al. gene patched encodes a candidate receptor for sonic hedgehog. 20. K.K. Youssef, A. Van Keymeulen, G. Lapouge, B. Beck, C. Michaux, Y. Achouri, P.A. Sotiropoulou, C. Blanpain, Nature 384,129–134 (1996) 4. J. Alcedo, M. Ayzenzon, T. Von Ohlen, M. Noll, J.E. Hooper, The Identification of the cell lineage at the origin of basal cell carcino- ma. Nat Cell Biol 12,299–305 (2010) Drosophila smoothened gene encodes a seven-pass membrane pro- tein, a putative receptor for the hedgehog signal. Cell 86,221–232 21. X. Bonilla, L. Parmentier, B. King, F. Bezrukov, G. Kaya, V. Zoete, (1996) V.B. Seplyarskiy, H.J. Sharpe, T. McKee, A. Letourneau, P.G. 5. V. Marigo, R.A. Davey, Y. Zuo, J.M. Cunningham, C.J. Tabin, Ribaux, K. Popadin, N. Basset-Seguin, R. Ben Chaabene, F.A. Biochemical evidence that patched is the hedgehog receptor. Santoni, M.A. Andrianova, M. Guipponi, M. Garieri, C. Verdan, Nature 384,176–179 (1996) K. Grosdemange, O. Sumara, M. Eilers, I. Aifantis, O. Michielin, 6. J. Taipale, M.K. Cooper, T. Maiti, P.A. Beachy, Patched acts cata- F.J. de Sauvage, S.E. Antonarakis, S.I. Nikolaev, Genomic analysis lytically to suppress the activity of smoothened. Nature 418,892– identifies new drivers and progression pathways in skin basal cell 897 (2002) carcinoma. Nat Genet 48,398–406 (2016) 7. K.C. Corbit, P. Aanstad, V. Singla, A.R. Norman, D.Y. Stainier, J.F. 22. B.L. Allen, T. Tenzen, A.P. McMahon, The hedgehog-binding pro- Reiter, Vertebrate smoothened functions at the primary cilium. teins Gas1 and Cdo cooperate to positively regulate Shh signaling Nature 437,1018–1021 (2005) during mouse development. Genes Dev 21,1244–1257 (2007) 8. J.P. Incardona, J. Gruenberg, H. Roelink, Sonic hedgehog induces 23. T. Tenzen, B.L. Allen, F. Cole, J.S. Kang, R.S. Krauss, A.P. the segregation of patched and smoothened in endosomes. Curr McMahon, The cell surface membrane proteins Cdo and Boc are Biol 12,983–995 (2002) components and targets of the hedgehog signaling pathway and 9. J.P. Incardona, J.H. Lee, C.P. Robertson, K. Enga, R.P. Kapur, H. feedback network in mice. Dev Cell 10,647–656 (2006) Roelink, Receptor-mediated endocytosis of soluble and membrane- 24. D. Carpenter, D.M. Stone, J. Brush, A. Ryan, M. Armanini, G. tethered sonic hedgehog by Patched-1. Proc Natl Acad Sci U S A Frantz, A. Rosenthal, F.J. de Sauvage, Characterization of two 97,12044–12049 (2000) patched receptors for the vertebrate hedgehog protein family. Proc 10. M.F. Bijlsma, K.S. Borensztajn, H. Roelink, M.P. Peppelenbosch, Natl Acad Sci U S A 95, 13630–13634 (1998) C.A. Spek, Sonic hedgehog induces transcription-independent cy- 25. I. Smyth, M.A. Narang, T. Evans, C. Heimann, Y. Nakamura, G. toskeletal rearrangement and migration regulated by arachidonate Chenevix-Trench, T. Pietsch, C. Wicking, B.J. Wainwright, metabolites. Cell Signal 19,2596–2604 (2007) Isolation and characterization of human patched 2 (PTCH2), a pu- 11. A. Ruiz, i. Altaba, C. Mas, B. Stecca, The Gli code: An information tative tumour suppressor gene inbasal cell carcinoma and medullo- nexus regulating cell fate, stemness and cancer. Trends Cell Biol 17, blastoma on chromosome 1p32. Hum Mol Genet 8, 291–297 438–447 (2007) (1999) 12. S.J. Scales, F.J. de Sauvage, Mechanisms of hedgehog pathway 26. P.G. Zaphiropoulos, A.B. Unden, F. Rahnama, R.E. Hollingsworth, activation in cancer and implications for therapy. Trends R. Toftgard, PTCH2, a novel human patched gene, undergoing Pharmacol Sci 30,303–312 (2009) alternative splicing and up-regulated in basal cell carcinomas. 13. H. Tian, C.A. Callahan, K.J. DuPree, W.C. Darbonne, C.P. Ahn, Cancer Res 59,787–792 (1999) S.J. Scales, F.J. de Sauvage, Hedgehog signaling is restricted to the 27. A.C. Alfaro, B. Roberts, L. Kwong, M.F. Bijlsma, H. Roelink, stromal compartment during pancreatic carcinogenesis. Proc Natl Ptch2 mediates the Shh response in Ptch1−/− cells. Development Acad Sci U S A 106,4254–4259 (2009) 141,3331–3339 (2014) 14. R.L. Yauch, S.E. Gould, S.J. Scales, T. Tang, H. Tian, C.P. Ahn, D. 28. B. Roberts, C. Casillas, A.C. Alfaro, C. Jagers, H. Roelink, Marshall, L. Fu, T. Januario, D. Kallop, M. Nannini-Pepe, K. Patched1 and Patched2 inhibit smoothened non-cell autonomously. Kotkow, J.C. Marsters, L.L. Rubin, F.J. de Sauvage, A paracrine elife 5 (2016) requirement for hedgehog signalling in cancer. Nature 455,406– 29. O. Zhulyn, E. Nieuwenhuis, Y.C. Liu, S. Angers, C.C. Hui, Ptch2 410 (2008) shares overlapping functions with Ptch1 in Smo regulation and limb 15. H. Hahn, C. Wicking, P.G. Zaphiropoulous, M.R. Gailani, S. development. Dev Biol 397,191–202 (2015) Shanley, A. Chidambaram, I. Vorechovsky, E. Holmberg, A.B. 30. A.M. Holtz, K.A. Peterson, Y. Nishi, S. Morin, J.Y. Song, F. Unden, S. Gillies, K. Negus, I. Smyth, C. Pressman, D.J. Leffell, Charron, A.P. McMahon, B.L. Allen, Essential role for ligand- B. Gerrard, A.M. Goldstein, M. Dean, R. Toftgard, G. Chenevix- dependent feedback antagonism of vertebrate hedgehog signaling Trench, B. Wainwright, A.E. Bale, Mutations of the human homo- by PTCH1, PTCH2 and HHIP1 during neural patterning. log of Drosophila patched in the nevoid basal cell carcinoma syn- Development 140,3423–3434 (2013) drome. Cell 85,841–851 (1996) 31. E. Nieuwenhuis, J. Motoyama, P.C. Barnfield, Y. Yoshikawa, X. 16. R.L. Johnson, A.L. Rothman, J. Xie, L.V. Goodrich, J.W. Bare, Zhang, R. Mo, M.A. Crackower, C.C. Hui, Mice with a targeted J.M. Bonifas, A.G. Quinn, R.M. Myers, D.R. Cox, E.H. Epstein mutation of patched2 are viable but develop alopecia and epidermal Jr., M.P. Scott, Human homolog of patched, a candidate gene for the hyperplasia. Mol Cell Biol 26,6609–6622 (2006) basal cell nevus syndrome. Science 272,1668–1671 (1996) 32. C. Adolphe, E. Nieuwenhuis, R. Villani, Z.J. Li, P. Kaur, C.C. Hui, 17. J. Svard, K. Heby-Henricson, M. Persson-Lek, B. Rozell, M. Lauth, B.J. Wainwright, Patched 1 and patched 2 redundancy has a key A. Bergstrom, J. Ericson, R. Toftgard, S. Teglund, Genetic elimi- role in regulating epidermal differentiation. J Invest Dermatol 134, nation of suppressor of fused reveals an essential repressor function 1981–1990 (2014) in the mammalian hedgehog signaling pathway. Dev Cell 10,187– 33. Y. Lee, H.L. Miller, H.R. Russell, K. Boyd, T. Curran, P.J. 197 (2006) McKinnon, Patched2 modulates tumorigenesis in patched1 hetero- . J. Xie, M. Murone, S.M. Luoh, A. Ryan, Q. Gu, C. Zhang, J.M. zygous mice. Cancer Res 66,6964–6971 (2006) Bonifas, C.W. Lam, M. Hynes, A. Goddard, A. Rosenthal, E.H. 34. C. Adolphe, J.P. Junker, A. Lyubimova, A. van Oudenaarden, B. Epstein Jr., F.J. de Sauvage, Activating smoothened mutations in Wainwright, Patched receptors sense, interpret, and establish an sporadic basal-cell carcinoma. Nature 391,90–92 (1998) epidermal hedgehog signaling gradient. J Invest Dermatol 137, 19. S.C. Peterson, M. Eberl, A.N. Vagnozzi, A. Belkadi, N.A. 179–186 (2016) Veniaminova, M.E. Verhaegen, C.K. Bichakjian, N.L. Ward, A.A. Dlugosz, S.Y. Wong, Basal cell carcinoma preferentially 35. L.V. Goodrich, L. Milenkovic, K.M. Higgins, M.P. Scott, Altered arises from stem cells within hair follicle and mechanosensory neural cell fates and medulloblastoma in mouse patched mutants. niches. Cell Stem Cell 16,400–412 (2015) Science 277, 1109–1113 (1997) Patched-2 functions to limit Patched-1 deficient skin cancer growth 36. M. Aszterbaum, J. Epstein, A. Oro, V. Douglas, P.E. LeBoit, M.P. C. Hauser-Kronberger, A.N. Ermilov, M.E. Verhaegen, C.K. Bichakjian, A.A. Dlugosz, W. Nietfeld, M. Sibilia, H. Lehrach, C. Scott, E.H. Epstein Jr., Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous Wierling, F. Aberger, Hedgehog-EGFR cooperation response genes knockout mice. Nat Med 5,1285–1291 (1999) determine the oncogenic phenotype of basal cell carcinoma and 37. J. Xie, M. Aszterbaum, X. Zhang, J.M. Bonifas, C. Zachary, E. tumour-initiating pancreatic cancer cells. EMBO Mol Med 4, Epstein, F. McCormick, A role of PDGFRalpha in basal cell carci- 218–233 (2012) noma proliferation. Proc Natl Acad Sci U S A 98, 9255–9259 47. J.K. Chen, J. Taipale, M.K. Cooper, P.A. Beachy, Inhibition of (2001) hedgehog signaling by direct binding of cyclopamine to smooth- 38. J. Ericson, S. Morton, A. Kawakami, H. Roelink, T.M. Jessell, Two ened. Genes Dev 16, 2743–2748 (2002) critical periods of sonic hedgehog signaling required for the speci- 48. B.L. Allen, J.Y. Song, L. Izzi, I.W. Althaus, J.S. Kang, F. Charron, fication of motor neuron identity. Cell 87,661–673 (1996) R.S. Krauss, A.P. McMahon, Overlapping roles and collective re- 39. H. Sasaki, C. Hui, M. Nakafuku, H. Kondoh, A binding site for Gli quirement for the coreceptors GAS1, CDO, and BOC in SHH path- proteins is essential for HNF-3beta floor plate enhancer activity in way function. Dev Cell 20,775–787 (2011) transgenics and can respond to Shh in vitro. Development 124, 49. L. Izzi, M. Levesque, S. Morin, D. Laniel, B.C. Wilkes, F. Mille, 1313–1322 (1997) R.S. Krauss, A.P. McMahon, B.L. Allen, F. Charron, Boc and Gas1 40. T. Reya, A.W. Duncan, L. Ailles, J. Domen, D.C. Scherer, K. each form distinct Shh receptor complexes with Ptch1 and are re- Willert, L. Hintz, R. Nusse, I.L. Weissman, A role for Wnt signal- quired for Shh-mediated cell proliferation. Dev Cell 20,788–801 ling in self-renewal of haematopoietic stem cells. Nature 423,409– (2011) 414 (2003) 50. K. Fujii, H. Ohashi, M. Suzuki, H. Hatsuse, T. Shiohama, H. 41. H. Damhofer, V.L. Veenstra, J.A. Tol, H.W. van Laarhoven, J.P. Uchikawa, T. Miyashita, Frameshift mutation in the PTCH2 gene Medema, M.F. Bijlsma, Blocking hedgehog release from pancreatic can cause nevoid basal cell carcinoma syndrome. Familial Cancer cancer cells increases paracrine signaling potency. J Cell Sci 128, 12,611–614 (2013) 129–139 (2015) 51. W.A. O'Hara, W.J. Azar, R.R. Behringer, M.B. Renfree, A.J. Pask, 42. R.B. Corcoran, M.P. Scott, Oxysterols stimulate sonic hedgehog Desert hedgehog is a mammal-specific gene expressed during tes- signal transduction and proliferation of medulloblastoma cells. ticular and ovarian development in a marsupial. BMC Dev Biol 11, Proc Natl Acad Sci U S A 103,8408–8413 (2006) 72 (2011) 43. M.F. Bijlsma, H. Damhofer, H. Roelink, Hedgehog-stimulated che- 52. S. Azzi, L. Treps, H.M. Leclair, H.M. Ngo, E. Harford-Wright, J. motaxis is mediated by smoothened located outside the primary Gavard, Desert hedgehog/Patch2 Axis contributes to vascular per- cilium. Sci Signal 5, ra60 (2012) meability and angiogenesis in glioblastoma. Front Pharmacol 6, 44. M.F. Bijlsma, C.A. Spek, D. Zivkovic, S. van de Water, F. Rezaee, 281 (2015) M.P. Peppelenbosch, Repression of smoothened by patched- 53. A. Okada, F. Charron, S. Morin, D.S. Shin, K. Wong, P.J. Fabre, M. dependent (pro-)vitamin D3 secretion. PLoS Biol 4, e232 (2006) Tessier-Lavigne, S.K. McConnell, Boc is a receptor for sonic 45. J. Briscoe, Y. Chen, T.M. Jessell, G. Struhl, A hedgehog-insensitive hedgehog in the guidance of commissural axons. Nature 444, form of patched provides evidence for direct long-range morphogen 369–373 (2006) activity of sonic hedgehog in the neural tube. Mol Cell 7,1279– 54. N. Basset-Seguin, H.J. Sharpe, F.J. de Sauvage, Efficacy of hedge- 1291 (2001) hog pathway inhibitors in basal cell carcinoma. Mol Cancer Ther 46. M. Eberl, S. Klingler, D. Mangelberger, A. Loipetzberger, H. 14,633–641 (2015) Damhofer, K. Zoidl, H. Schnidar, H. Hache, H.C. Bauer, F. Solca,

Journal

Cellular OncologySpringer Journals

Published: Jun 4, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off