Statins Synergize with Hedgehog Pathway Inhibitors for Treatment of Medulloblastoma Renata E

Published OnlineFirst February 6, 2018; DOI: 10.1158/1078-0432.CCR-17-2923

Cancer Therapy: Preclinical Clinical Cancer Research Statins Synergize with Hedgehog Pathway Inhibitors for Treatment of Medulloblastoma Renata E. Gordon1,2, Li Zhang3, Suraj Peri4, Yin-Ming Kuo5, Fang Du1, Brian L. Egleston4, Jessica M. Y. Ng6, Andrew J. Andrews5, Igor Astsaturov2,7, Tom Curran6, and Zeng-Jie Yang1,3

Abstract

Purpose: The role of cholesterol biosynthesis in hedgehog approved Smoothened antagonist), were utilized to inhibit pathway activity and progression of hedgehog pathway medul- medulloblastoma growth in vivo. loblastoma (Hh-MB) were examined in vivo. Statins, commonly Results: Cholesterol biosynthesis was markedly enhanced in used cholesterol-lowering agents, were utilized to validate cho- Hh-MB from both humans and mice. Inhibition of cholesterol lesterol biosynthesis as a therapeutic target for Hh-MB. biosynthesis dramatically decreased Hh pathway activity and Experimental Design: Bioinformatic analysis was performed reduced proliferation of medulloblastoma cells. Statins effectively to evaluate the association between cholesterol biosynthesis with inhibited medulloblastoma growth in vivo and functioned syn- hedgehog group medulloblastoma in human biospecimens. ergistically in combination with vismodegib. Alterations in hedgehog signaling were evaluated in medulloblas- Conclusions: Cholesterol biosynthesis is required for toma cells after inhibition of cholesterol biosynthesis. The pro- Smoothened activity in the hedgehog pathway, and it is indis- gression of endogenous medulloblastoma in mice was examined pensable for the growth of Hh-MB. Targeting cholesterol bio- after genetic blockage of cholesterol biosynthesis in tumor cells. synthesis represents a promising strategy for treatment of Statins alone, or in combination with vismodegib (an FDA- Hh-MB. Clin Cancer Res; 24(6); 1375–88. 2018 AACR.

Introduction reductase (HMGCR) is rate limiting for cholesterol biosynthesis. HMGCR converts HMG-CoA to mevalonate and is the target Cholesterol is a key component of cellular membranes and a of statins (2). NAD(P) steroid dehydrogenase-like (NSDHL), a precursor for steroid hormones and bile acids. The brain is the 3b-hydroxysterol dehydrogenase, is involved in the removal of most cholesterol-rich organ, containing approximately 20% of two C-4 methyl groups in one of the later steps of cholesterol total body cholesterol. In the brain, cholesterol is primarily biosynthesis. Deficiency in NSDHL dramatically impairs choles- generated by de novo synthesis, as the blood–brain barrier in terol biosynthesis in both humans and mice (3). In particular, vertebrates prevents its uptake from the bloodstream (1). Almost conditional deletion of Nsdhl gene in cerebellar granule neuronal all cholesterol in the brain is present in an unesterified, free form. precursors inhibits cholesterol synthesis and dramatically Cholesterol is synthesized through a series of approximately represses proliferation (4). The final step in cholesterol synthesis 30 enzymatic reactions, starting with the substrate acetyl CoA is catalyzed by the enzyme 24-dehydrocholesterol reductase (Fig. 1A). The enzyme 3-hydroxy-3-methylglutaryl-coenzyme A (DHCR24), which converts desmosterol to cholesterol by satu- rating the C-24,25 double-bond in the side chain. Triparanol, the first synthetic cholesterol-lowering drug, works by antagonizing 1 Cancer Biology Program, Fox Chase Cancer Center, Temple University Health DHCR24 (5). System, Philadelphia, Pennsylvania. 2Institute of Fundamental Medicine and Biology, Kazan Federal University, Kazan, Russia. 3Jiangsu Key Laboratory of In addition to its function as a component of cell membranes, Neuropsychiatric Diseases and College of Pharmaceutical Sciences, Soochow cholesterol regulates many signaling pathways, including the University, Suzhou, Jiangsu, China. 4Biostatistics and Bioinformatics Research hedgehog (Hh) pathway. The Hh pathway plays a critical role in Facility, Fox Chase Cancer Center, Temple University Health System, Philadel- brain development and function. In the absence of Hh ligand, the 5 phia, Pennsylvania. Cancer Epigenetics Program, Fox Chase Cancer Center, antagonizing receptor Patched1 (Ptch1) suppresses the seven- 6 Temple University Health System, Philadelphia, Pennsylvania. Children's transmembrane protein, Smoothened (Smo), thus precluding Research Institute, Children's Mercy Kansas City, Missouri. 7Molecular Thera- peutics Program, Fox Chase Cancer Center, Temple University Health System, downstream signaling. The interaction between Hh and Ptch1 Philadelphia, Pennsylvania. relieves inhibition of Smo, which then triggers a cascade of events culminating in activation of the glioma-associated oncogenes 1 Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). and 2 (Gli1 and Gli2). Gli proteins translocate into cell nuclei and promote transcription of Hh pathway target genes, including Cancer Therapy: Preclinical Ptch1 and Gli1 (6, 7). Extensive evidence suggests that some Corresponding Author: Zeng-Jie Yang, Fox Chase Cancer Center, 333 Cottman exogenous oxysterols, such as 20 (S)-hydroxycholesterol and Avenue, Philadelphia, PA 19111. Phone: 215-214-1545; Fax: 215-728-2741; E-mail: 25-hydroxycholesterol, generated by cholesterol oxidation, can [email protected] activate Smo by binding to a site located in the extracellular doi: 10.1158/1078-0432.CCR-17-2923 cysteine-rich domain (CRD). This indicates that cholesterol deri- 2018 American Association for Cancer Research. vatives are capable of triggering Hh pathway activation in

www.aacrjournals.org 1375

Gordon et al.

in vitro as well as tumor growth in vivo. Furthermore, we show that Translational Relevance cholesterol is necessary for maintaining Smo activity in the Medulloblastoma is the most common malignant brain absence of Ptch1. Statins, antagonists of cholesterol biosynthesis, tumor in children. Approximately 30% of human medullo- exhibited promising therapeutic effects in mouse medulloblas- blastoma arise from aberrant activation of hedgehog (Hh) toma models. Moreover, statins synergized with the clinically pathway. Vismodegib and sonidegib are two FDA-approved approved Smo antagonist, vismodegib, to prevent medulloblas- Hh pathway inhibitors that have demonstrated efficacy in toma progression in vivo. These findings elucidate an important clinical trials of Hh pathway–associated medulloblastoma role for endogenous cholesterol in medulloblastoma tumorigen- (Hh-MB). Despite a dramatic response to these inhibitors in esis and demonstrate that cholesterol biosynthesis represents a initial clinical trials, drug resistance and developmental toxi- novel therapeutic target for the treatment of medulloblastoma cities arise in humans as predicted by mouse models. Here, we and, potentially, other Hh pathway–associated malignancies. demonstrate that cholesterol is required for Hh pathway signal transduction in tumor cells. Statins, inhibitors of cholesterol Materials and Methods biosynthesis, block hedgehog pathway activity in medulloblastoma and synergize with vismodegib to inhibit tumor Animals growth. Given the excellent safety profile, low cost and ready All mice were maintained under barrier conditions in the availability of cholesterol inhibitors, targeting cholesterol Laboratory Animal Facility at Fox Chase Cancer Center (Phila- biosynthesis represents a promising strategy for treatment of delphia, PA), and all experiments were performed in accordance with procedures approved by the Fox Chase Cancer Center Hh-MB and potentially other malignancies caused by Hh fl/fl pathway activation. Animal Care and Use Committee. Nsdhl mice were kindly provided by Dr. Gail Herman (The Ohio State University, T2 Columbus, OH). Math1-Cre mice, Math1-CreER mice, T2 fl/fl þ/ GFAP-CreER mice, Ptch1 mice, Ptch1 mice, R26R-eYFP mice, and SmoM2 (R26SmoM2)micewerepurchasedfromThe recipient cells (8–11). Two recent studies revealed that cholesterol Jackson Laboratory. CB17/SCID mice were generated by the itself synergizes with Hh ligand to activate Smo through binding animal facility laboratory at Fox Chase Cancer Center. to the CRD (12, 13). The physical interaction between Smo and Tamoxifen (Sigma-Aldrich) was dissolved in corn oil and cholesterol was further confirmed in studies of the crystal structure administrated by oral gavage at a dosage of 200 mg/kg daily of Smo (14). These findings suggest that cholesterol functions as a during 3 subsequent days. Animals were sacrificed for further native ligand for Smo activation and that Hh signaling may be analyses after 7 days from initial tamoxifen treatment. influenced by the local availability of cholesterol. Insufficient Hh signaling can lead to birth defects, whereas Compounds excessive activity of the Hh pathway is associated with the for- The following compounds were used for our in vitro and in vivo mation of human malignancies, including medulloblastoma, the studies: simvastatin (Cayman Chemicals), atorvastatin (Cayman most common malignant brain tumor in children. The Hh Chemicals), triparanol (Sigma-Aldrich), vismodegib (Chemlea- pathway is activated in approximately 30% of human medullo- der Biochemical), and water-soluble cholesterol (cholesterol- blastoma. Although the cellular and molecular basis for medul- methyl-beta-cyclodextrin, Sigma-Aldrich). loblastoma tumorigenesis has been widely studied, the role of Hh conditioned media (Hh-CM) was prepared by transfection cholesterol in medulloblastoma tumorigenesis has yet to be of pcDNA3.1 Shh-N (Addgene) plasmid into 293FT cells. Super- elucidated. natant collected at 48 hours after transfection was used at 20% for Previously, several studies claimed that inhibition of choles- treating NIH 3T3 cells. terol synthesis promoted apoptosis in medulloblastoma cells, and this effect was greater during cotreatment with the naturally Subcutaneous allografts occurring Hh pathway inhibitor cyclopamine (15–18). However, CB17/SCID male mice of age 6 to 8 weeks were used for these effects required the use of statins at levels 1,000-fold greater subcutaneous allograft establishment. Cells isolated from than the physiologic concentrations required to block cholesterol tumor-bearing mice in a single-cell suspension in PBS at concen- biosynthesis (15–18). In addition, these studies relied on the use tration 2 106 cells were mixed with reduced growth factor of established medulloblastoma cell lines in which the Hh path- Matrigel (Corning) in a 20:80 ratio and injected subcutaneously way is now known to be downregulated (19). An additional on the flank of CB17/SCID mice. Four weeks following injection, confounding problem with these studies is that they utilized tumor-bearing mice that satisfied initial tumor volume require- cyclopamine to inhibit the Hh pathway in cultured tumor cells, ment (200–400 mm3) were randomized into experimental and it has now been demonstrated that, at the concentrations groups. Simvastatin DMSO stock was dissolved in a mixture of used, cyclopamine causes apoptosis by promoting ceramide ethanol, 0.1 mol/L NaOH, and PBS and administrated in desig- production independently of any effects on the Hh pathway nated dose by intraperitoneal injection. Atorvastatin and vismo- (20). Thus, the question remains open as to whether cholesterol degib DMSO stock was dissolved in 0.5% methylcellulose 0.2% pathway inhibitors, alone or in combination with Hh pathway Tween-80 and administrated per os. Mixture of DMSO, ethanol, inhibitors, have potential as therapeutic agents in the treatment of 0.1 mol/L NaOH, and PBS or DMSO and 0.5% methylcellulose Hh-medulloblastoma. 0.2% Tween-80 was used as a vehicle substance. Compounds were Here, we demonstrate that there is enhanced cholesterol bio- administrated in half of the indicated dose twice a day and tumor synthesis in Hh pathway–associated medulloblastoma (Hh-MB) size measurements were taken every 2 days and tumor volume and that cholesterol is required for proliferation of tumor cells calculated as described previously (21).

1376 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Effectively Inhibit Medulloblastoma Growth

Figure 1. Cholesterol biosynthesis is enhanced in Hh subtype of medulloblastoma. A, A schematic diagram of cholesterol biosynthesis pathway with genes encoding corresponding enzymes listed along arrows. Enzymes: HMGCS1 3-hydroxy-3-methylglutaryl-CoA synthase 1; HMGCR 3-hydroxy-3-methylglutaryl-CoA reductase; MVD mevalonate diphosphate decarboxylase; FDPS farnesyl diphosphate synthase; FDFT1 farnesyl-diphosphate farnesyltransferase 1; SQLE squalene epoxidase; LSS lanosterol synthase; NSDHL NAD(P) dependent steroid dehydrogenase-like; DHCR7 7-dehydrocholesterol reductase; DHCR24 24-dehydrocholesterol reductase. Statins inhibit conversion of HMG-CoA to mevalonate through competitive inhibition of HMG-CoA reductase (HMGCR). Triparanol blocks the last step of cholesterol biosynthetic pathway through inhibition of 24-dehydrocholesterol reductase (DHCR24). B and C, Boxplots showing the enrichment scores from GSVA. x and y-axes illustrate enrichment statistics across 4 different subtypes of medulloblastoma. Positive and negative scores indicate positive and negative enrichment, respectively. Pink dots represent the enrichment score for each sample in that subtype. Black bar in the middle of the box indicates median. D, Heatmaps showing expression of representative genes in the Hh and cholesterol synthesis pathways. z-scores calculated for each gene are plotted on a red (higher expression) and blue (low expression) scale. Top color bar, subtype. E–G, IHC analysis of NSDHL (E), ABCA1 (F), and SREBP2 (G) proteins in mouse Ptch1þ/ medulloblastoma. Normal adjacent tissue represents a cerebellar lobe with differentiated granule neurons used as a control. Scale bar, 100 mm; inset, 10 mm.

Cell culture pyruvate, 2 mmol/L L-glutamine, 1 g/L glucose, penicillin/strep- Medulloblastoma cells were suspended in NB-B27 culture tomycin 100 U/mL, and 1% nonessential amino acids (all from medium (Neurobasal, 1 mmol/L sodium pyruvate, 2 mmol/L Life Technologies, unless noted otherwise). L-glutamine, B27 supplement, and 1% penicillin/streptomycin, NIH 3T3 and HEK293T cell lines were acquired from ATCC and all from Invitrogen) and plated on poly-D-lysine–coated cover- cultured in DMEM supplemented with 10% FBS, 1% penicillin/ slips (Millipore). For bromodeoxyuridine (BrdUrd) incorpo- streptomycin, and 2 mmol/L L-glutamine (all from Life Technol- ration assay, cells were pulsed with 10 mmol/L BrdUrd (Millipore) ogies). All of the above cell lines were authenticated and tested for for 2 hours, fixed in 4% PFA, and incubated in 4.2 mmol/L MgCl2/ mycoplasma by the Cell Culture Facility at Fox Chase Cancer 0.15 mol/L NaCl pH 5.0 buffer in the presence of 50 U/mL of Center. NIH 3T3 cells and MEFs were starved in serum-free media DNase I (Sigma-Aldrich) at 37C for 30 minutes. for 24 hours before virus infection or Hh treatment. All in vitro Primary mouse embryonic fibroblasts (MEF) were isolated studies were performed in serum-free environment. fl/fl from E13.5 of Ptch1 and SmoM2 mouse embryos. Dissociated For synergistic effects experiments, NIH 3T3 cells were starved MEFs were cultured in DMEM supplemented with 15% FBS for 24 hours before treatment with 20% Hh-CM. Hh-treated NIH þ/ (SH30071.03, HyClone, General Electric), 1 mmol/L sodium 3T3 cells and freshly isolated Ptch1 medulloblastoma cells

www.aacrjournals.org Clin Cancer Res; 24(6) March 15, 2018 1377

Gordon et al.

were then incubated with 0.001 to 1 mmol/L water-soluble cho- 6 mL of 1:2 (v/v) chloroform:methanol was added to the sample. 2 lesterol (WSC) for 18 to 20 hours to saturate Smo, and 1 nmol/L H6-26,26,26,27,27,27-cholesterol (Sigma-Aldrich) was used as vismodegib for additional 48 hours. the internal standard spiked in individual samples. The cholesterol-containing chloroform layer was collected and evaporated Lentiviral production under a gentle, dried nitrogen stream at 35C. The dried samples Nsdhl Lentiviral production was performed by utilizing shRNAs were redissolved in 200 mL methanol and analyzed by ultraper- and scrambled shRNA (Genocopiea, MSH029806), pLUT Gli1- formance liquid chromatography-tandem mass spectrometry HA, pMD2.G, and psPAX2 (Addgene) plasmids in accordance (UPLC-MS/MS). The peaks of samples were processed and ana- with the standard protocol. lyzed using Xcalibur, version 2.1.

Histology and IHC Statistical analysis For histologic analyses, cerebella were removed from mice and All studies were performed at least in three independent repli- fi xed in 10% neutral-buffered formalin overnight. Samples were cates with outcomes reflected on photographs, graphs, or film fi m embedded in paraf n; 5 m sagittal sections were prepared and scans. Material collection time points were designed in accor- stained for ABCA1 (Abcam, ab18180, 1:500), NSDHL (Protein- dance to previously published literature. During data analysis, tech, 15111-1-AP, 1:100), and SREBP2 (Abcam, ab30682, 1:100). no exclusion criteria were applied. Two-sided Student t tests were Frozen sections were prepared according to standard protocol performed to determine the statistical significance of the dif- fl fi (22). Brie y, tissue was xed in 4% PFA overnight, cryoprotected ference in means between samples in the experiments reported. in 30% sucrose overnight, and frozen in O.C.T. (Thermo Fisher P < 0.05 was considered statistically significant. Error bars fi m Scienti c). Sagittal sections (12 m) were prepared with a cryostat represent the SEM. Data handling and statistical processing fi (Leica). Cultured medulloblastoma cells were xed in 4% PFA for was performed using Microsoft Excel. 15 minutes. The raw gene expression data used for analyzing subtype- Tissue sections or cultured cells were permeabilized, blocked specific pathway alterations in medulloblastoma were obtained with 10% normal goat serum in PBS, 0.1% Triton X-100, and from GEO (GSE21140) and normalized using RMA. Gene set incubated overnight in primary antibody mix against Ki67 variation analysis (GSVA; ref. 23) on the RMA normalized expres- (Abcam, ab15580, 1:1,000), NeuN (Abcam, ab104224, 1:200), sion data was applied to identify pathways that are enriched in cleaved caspase-3 (Cell Signaling Technology, CST9661, 1:500), a single sample (method arguments: function¼'gsva'; mx. BrdUrd (Sigma-Aldrich, B8434, 1:500), mCherry (Abcam, diff¼'FALSE'; verbose¼FALSE). Over 4,900 gene sets compiled ab167453, 1:500), Zic1 [generous gift from Dr. Rosalind Segal from all gene sets available in MsigDB, including KEGG and b at Dana Farber Research Institute (Boston, MA), 1:200], S100 Biocarta pathways, were used for enrichment analysis. To identify (Sigma-Aldrich, S2532, 1:500), GFP (Life Technologies, A11122, pathways that differ significantly among subtypes, Kruskal–Wallis 1:500), GFAP (BD, 556330, 1:500). After 2 hours of incubation in test was used to analyze the enrichment scores resulting from secondary antibodies, samples were counterstained with DAPI GSVA. fi and mounted with Fluoromount-G. For llipin staining experi- To compare the progression rates (Figs. 5A and 6A and C), we ments, medulloblastoma sections were incubated with 2 mg/mL used generalized linear regression models assuming gamma fam- fi lipin (Sigma-Aldrich) for 2 hours, counterstained with DRAQ5, ily and log links. We used the following growth curve model to test and mounted with Fluoromount-G. for synergy: logðÞ¼ E½ YjV; S; D b þ b V þ b S þ b V S Western blotting 0 1 2 4 þ ðÞb þ b þ b þ b ; Cells and tissues were lysed in RIPA buffer (Thermo Fisher 3 5V 6S 7V S D Scientific) supplemented with protease (Roche) and phosphatase where Y ¼ tumor volume, D ¼ day, V ¼ vismodegib, S ¼ statin, (Thermo) inhibitors cocktail. Total protein lysates were separated and E[Y|V,S,D] ¼ conditional expectation (i.e., average). b and b in 8% SDS-PAGE and transferred on PVDF membrane. Mem- 5 6 test for additive drug growth effects, and the interaction term, b , branes were probed with antibodies against Gli1 (Cell Signaling 7 tests for synergistic/antagonistic drug growth effects. Synergism Technology, CST2534, 1:1,000), NSDHL (Proteintech, 15111-1- was concluded if the direction of b enhanced drug effects and AP, 1:1,000), Ptch1 (Novus Biologicals, MAB41051, 1:1,000), 7 P < 0.05 (24). GAPDH (Sigma-Aldrich, G8795, 1:1,000). Western blot signals were detected with SuperSignal West Pico Substrate and exposed on films. Results Hh-MB exhibit dysregulated cholesterol metabolism qPCR Human medulloblastoma comprises at least four subgroups, RNA was isolated using TRI reagent (Sigma-Aldrich) in RNase- including Hh, Wnt, group 3, and group 4, based on distinct gene free conditions. cDNA was synthesized using oligo(dT) and Super- expression profiles (25, 26). Hh-MB arises from aberrant activa- script II reverse transcriptase (Invitrogen). Quantitative PCR reac- tion of the Hh pathway in the developing cerebellum. To examine tions were performed in triplicates using iQ SYBR Green Supermix pathway alterations among medulloblastoma subgroups, we (Bio-Rad) and the Bio-Rad iQ5 Multicolor Real-Time PCR Detec- analyzed transcriptomes of 103 primary human medulloblasto- tion System. Primer sequences are available upon request. ma by GSVA (23). This analysis uncovered 363 gene sets and pathways (GSP) that exhibit selective enrichment across the four Mass spectrometry and fluorometric cholesterol assay subtypes. The significant GSPs ranking high in the list included the Cells or tissue samples were washed with ice-cold DPBS, Hh and Netrin pathways known to play roles in Hh-MB tumor- and lipids were extracted using a Bligh/Dyer procedure, in which igenesis (Supplementary Table S1; Fig. 1B–D; ref. 27), In addition,

1378 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Effectively Inhibit Medulloblastoma Growth

we observed that cholesterol and sterol biosynthesis GSPs were flanked STOP cassette upstream of the eYFP gene (33). After T2 þ associated with Hh-MB but not the other 3 medulloblastoma tumors were established, Math1-CreER /R26R-eYFP/Ptch1 / subgroups. The enrichment scores of both steroid and cholesterol mice were treated with tamoxifen by oral gavage. As shown biosynthesis pathways were higher in Hh-MB than the other 3 in Fig. 2E, most of the tumor cells positive for Zic1 (34) also subtypes (Fig. 1C). Most of the genes in these two pathways are expressed YFP, but no astrocytes (S100b positive) were positive T2 identical, and their expression levels are remarkably elevated in for YFP (Fig. 2F). These data indicate that Math1-CreER mice can Hh-MB (Fig. 1D). These data suggest an unexpected association be utilized to target medulloblastoma cells in vivo. between cholesterol biosynthesis and Hh-MB. To disrupt cholesterol biosynthesis in medulloblastoma cells, þ fl/fl Mice heterozygous for Ptch1 (hereafter referred as Ptch1 / we utilized conditional Nsdhl knockout mice, in which a mice) are widely used in preclinical studies for Hh-MB as 15% portion of the Nsdhl gene is flanked by loxP sites (4), crossed T2 þ T2 to 20% of mice spontaneously develop medulloblastoma in with Math1-CreER mice and Ptch1 / mice. Math1-CreER / fl/fl þ fl/fl þ cerebella after 3 to 4 months of age (28), and this is accelerated Nsdhl /Ptch1 / mice and Nsdhl /Ptch1 / mice were orally to 100% in the absence of p53 (29). Constitutive activation of the treated with tamoxifen after tumors were fully established. At 7 Hh pathway as a consequence of loss of Ptch1 heterozygosity is days after tamoxifen treatment, medulloblastoma tissues from þ T2 fl/fl þ fl/fl þ critical for medulloblastoma formation in Ptch1 / mice. To Math1-CreER /Nsdhl /Ptch1 / mice and Nsdhl /Ptch1 / lit- investigate the possible association between cholesterol biosyn- termates were examined for expression levels of NSDHL and Gli1 thesis and tumorigenesis in Hh-MB, we examined the expression as well as the amount of intracellular cholesterol present in the of proteins involved in cholesterol homeostasis in medulloblas- tumor. As shown in Fig. 2G, after tamoxifen treatment, NSDHL þ toma tissue from Ptch1 / mice by IHC. Cholesterol homeostasis was depleted in medulloblastoma tissues obtained from Math1- T2 fl/fl þ is tightly regulated by a complex protein network responsible for CreER /Nsdhl /Ptch1 / mice. Moreover, Gli1 expression was T2 its import, synthesis, and export (30). As shown in Fig. 1E, decreased in medulloblastoma tissue from Math1-CreER / fl/fl þ expression of NSDHL was significantly upregulated in medullo- Nsdhl /Ptch1 / mice compared with tumor tissue from fl/fl þ blastoma tissue compared with adjacent normal cerebellar tissue, Nsdhl /Ptch1 / mice, suggesting that Hh signaling was com- suggesting there is increased cholesterol synthesis in tumor tissue. promised in Nsdhl-deficient medulloblastoma. In addition, the Transcription of ATP-binding cassette transporter 1 (ABCA1), an amount of cholesterol in tumor tissue was reduced by 50% 10% established regulator of cellular cholesterol homeostasis, is nor- after tamoxifen treatment, as revealed by mass spectrometry mally driven by increased accumulation of intracellular choles- analysis (Fig. 2H), indicating that Nsdhl deletion effectively sup- terol (31) and sterol regulator element-binding protein 2 pressed cholesterol biosynthesis in tumor cells. We next examined (SREBP2) is well known to upregulate genes involved in choles- the proliferation and survival of cholesterol-deficient medullo- terol biosynthesis and uptake (32). The levels of both ABCA1 and blastoma cells in vivo. As shown in Fig. 2I–I00, cell proliferation SREBP2 were substantially elevated in tumors, compared with was reduced by 50% 5% in Nsdhl-deficient medulloblastoma fl/fl adjacent normal cerebellum (Fig. 1F–G). These data imply that tissue compared control medulloblastoma tissue from Nsdhl / þ there is enhanced synthesis of cholesterol in Hh-MB. Ptch1 / mice, as shown by immunostaining for Ki67. Medullo- blastoma cells that remained proliferating were predominately Medulloblastoma growth relies on de novo cholesterol located in regions of tumor tissue in which cholesterol was not biosynthesis completely depleted based on filipin staining (Supplementary Having observed upregulation of cholesterol biosynthesis Fig. S1). It appeared that most of the Nsdhl-deficient tumor cells genes in Hh-MB, we investigated the possible function of cho- differentiated based on the expression of neuronal differentiation lesterol in medulloblastoma tumorigenesis. Medulloblastoma marker NeuN (Fig. 2J–J00). Increased apoptosis was also detected þ cells isolated from Ptch1 / mice were infected with a lentivirus by cleaved caspase-3 (CC3) immunostaining of tumor tissue carrying mCherry-tagged Nsdhl shRNAs or scrambled shRNA. At after Nsdhl deletion compared with that in control tumor tissue 48 hours following viral infection, NSDHL protein expression was (Fig. 2K–K00). Consistent with our findings in NSDHL-deficient effectively repressed by Nsdhl shRNAs (Fig. 2A). Expression of Gli1 medulloblastoma tissues, expression of Hh pathway target genes, was significantly reduced in NSDHL-deficient medulloblastoma including Gli1, Myc-N, and Ptch1, was significantly repressed after compared with scrambled shRNA-infected medulloblastoma Nsdhl deletion (Fig. 2L). The above data demonstrate that de novo cells (Fig. 2A), indicating that Nsdhl knockdown impaired Hh biosynthesis of cholesterol in medulloblastoma cells is required pathway activity. We next examined proliferation of medullo- for proliferation and tumor growth. blastoma cells at 48 hours following viral infection. As shown Astrocytes expressing glial fibrillary acidic protein (GFAP) were in Fig. 2B–D, approximately 15% of scrambled shRNA-infected abundantly present and intermingled with tumor cells in medul- medulloblastoma cells were positive for BrdUrd, compared with loblastoma tissue (Zic1 positive, Supplementary Fig. S2A). To test only 4% to 6% BrdUrd-positive cells among NSDHL-deficient whether cholesterol from astrocytes is required for medulloblas- medulloblastoma cells. These data suggest that cholesterol syn- toma progression, we examined medulloblastoma cell prolifer- thesis is critical for proliferation and Hh pathway activity in ation after disrupting cholesterol synthesis in astrocytes by using T2 medulloblastoma cells. GFAP-CreER mice (35). For this purpose, we generated GFAP- T2 fl/fl þ fl/fl þ We next examined proliferation of medulloblastoma cells CreER /Nsdhl /Ptch1 / mice, and Nsdhl /Ptch1 / mice as a in vivo after genetic ablation of cholesterol biosynthesis in tumor control. These mice were treated with tamoxifen by oral gavage cells. Hh-MB cells specifically express the helix-loop-helix tran- after tumors were established at approximately 30 weeks of age. As scription factor Math1 (22). To establish an appropriate assay shown in Supplementary Fig. S2B–S2D00, no obvious alterations T2 system, Math1-CreER mice, capable of expressing a tamoxifen- in the proliferation, differentiation, or apoptosis were found inducible Cre recombinase only in Math1-expressing cells, were in medulloblastoma tissue after Nsdhl deletion in astrocytes in þ T2 fl/fl þ crossed with Ptch1 / mice and R26R-eYFP mice carrying a loxP- GFAP-CreER /Nsdhl /Ptch1 / mice, compared with that in

www.aacrjournals.org Clin Cancer Res; 24(6) March 15, 2018 1379

Gordon et al.

Figure 2. Deficiency of NSDHL inhibits Ptch1þ/ medulloblastoma cell proliferation. A, Western blot analysis of NSDHL and Gli1 protein expression in Ptch1þ/ medulloblastoma cells following infection with lentivirus carrying Nsdhl shRNA (#3 and #4) or scrambled shRNA and uninfected control after 48 hours in culture. GAPDH was used as a loading control. B and C, Immunofluorescence analysis of BrdUrd among Ptch1þ/ medulloblastoma cells infected with scrambled control shRNA (mCherryþ, B)orNsdhl shRNA #4 (mCherryþ, C) after 48 hours in culture. D, Percentage of BrdUrdþ cells in infected medulloblastoma cells (mCherryþ)inFig.2Band C. E and F, Sagittal sections of medulloblastoma tissue from Math1-CreERT2/R26R-eYFP/Ptch1þ/ mice were immunostained against Zic1 and YFP (E), S100b, and YFP (F). G, Western blot analysis of NSDHL and Gli1 protein expression in medulloblastoma tissue from Math1-CreERT2/Nsdhlfl/fl/Ptch1þ/ and Nsdhlfl/fl/Ptch1þ/ mice after tamoxifen treatment. GAPDH was used as a loading control. H, Mass spectrometry analysis of cholesterol content (mmol/L) in medulloblastoma tissue from Math1-CreERT2/Nsdhlfl/fl/Ptch1þ/ mice and Nsdhlfl/fl/Ptch1þ/ mice after tamoxifen treatment. I–K00, Immunofluorescence analysis of Ki67 (I–I00), NeuN (J–J00), and CC3 (K–K00) in tumor tissue from Nsdhlfl/fl/Ptch1þ/ mice (I–K), Math1-CreERT2/Nsdhlfl/fl/Ptch1þ/ mice (I0–K0) after tamoxifen treatment, and percentage of Ki67, NeuN, and CC3-positive cells in above sections, respectively (I00–K00). Medulloblastoma tissues were counterstained with DAPI. L, qPCR analysis of relative mRNA expression of Hh pathway genes (Gli1, N-Myc, Ptch1) in medulloblastoma tissue from Math1-CreERT2/Nsdhlfl/fl/Ptch1þ/ mice and Nsdhlfl/fl/Ptch1þ/ mice after tamoxifen treatment. Scale bar, 10 mm. Significance asterisk key: P < 0.05 (), P < 0.01 (), P < 0.001 ().

1380 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Effectively Inhibit Medulloblastoma Growth

Figure 3. Cholesterol is required for Smo activation in Hh signal transduction. A, Western blot analysis of Gli1 protein expression in Ptch1þ/ medulloblastoma (MB) cells following treatment with simvastatin alone or in combination with WSC for 48 hours. B, Western blot analysis of Gli1 protein expression in NIH 3T3 cells following treatment with Hh-CM, simvastatin, triparanol, or together with WSC for 48 hours. C and D, Western blot analysis of Gli1 protein expression in Ptch1ﬂ/ﬂ MEFs (C)orSmoM2 MEFs (D) infected with a lentivirus carrying Cre recombinase or control empty vector following treatment with simvastatin, triparanol, or WSC for 48 hours. E, Western blot analysis of Gli1 and Ptch1 proteins in NIH 3T3 cells infected with a lentivirus carrying a Gli1-HA construct or an empty vector following treatment with simvastatin, triparanol, or WSC for 48 hours. GAPDH was used as a loading control. F, A schematic diagram of Hh signaling pathway. In the absence of Hh ligand, its receptor Ptch1 tethers Smo, preventing the downstream activation of Hh pathway. After interaction with Hh, Ptch1 releases Smo, causing activation of Gli1/2 transcription factor and leading to transcription of Hh pathway genes (Gli1, Ptch1, etc.). Cholesterol is required for Smo activation in Hh signal transduction.

fl/fl þ Nsdhl /Ptch1 / mice. These data suggest that astrocyte-derived medulloblastoma cells. These data suggest that simvastatin cholesterol is dispensable for medulloblastoma growth. inhibits Hh pathway activity in medulloblastoma cells by disrupting cholesterol biosynthesis. Cholesterol is required for Smo activation To identify which component of the Hh pathway is regulated by To investigate the role of cholesterol in Hh pathway acti- cholesterol, we investigated the effect of cholesterol on Hh ligand þ vation, we treated medulloblastoma cells from Ptch1 / mice signaling in NIH 3T3 cells. As expected, Hh treatment significantly with simvastatin in short-term cultures. Under these condi- upregulated Gli1 expression in NIH 3T3 cells. The induction of tions, the Hh pathway remains active, whereas it is repressed Gli1 by Hh ligand was blocked by simvastatin or triparanol, and in long-term cultures (19). At 48 hours after treatment, cells this was overcome by the addition of exogenous WSC to the were harvested to determine the level of Gli1 expression by culture medium (Fig. 3B). These data confirm that cholesterol is Western blot analysis. As shown in Fig. 3A, the expression necessary for transduction of Hh pathway activity from ligand to levels of Gli1 declined in medulloblastoma cells as the con- target gene expression. fl/fl centration of simvastatin was increased. To confirm that We next generated primary MEFs from Ptch1 mice carrying compromised Hh pathway activity in medulloblastoma exons 1 and 2 of the Ptch1 gene flankedbyloxPsites,andfrom cells after treatment with simvastatin is a consequence of SmoM2 mice carrying loxP-flankedSTOPcassetteplaced reduced cholesterol levels, we added WSC to the tumor cell upstream of constitutively active Smo (36). MEFs were then cultures. As shown in Fig. 3A, the addition of exogenous infected with a lentivirus carrying Cre recombinase to delete cholesterol rescued Gli1 expression in simvastatin-treated Ptch1 or activate Smo, or with a control lentivirus carrying an

www.aacrjournals.org Clin Cancer Res; 24(6) March 15, 2018 1381

Gordon et al.

Figure 4. Cholesterol biosynthesis inhibitors suppress proliferation of Ptch1þ/ medulloblastoma cells. A–K, Medulloblastoma cells isolated from Ptch1þ/ mouse were cultured in the presence of vehicle (A) or simvastatin (B–D), triparanol (E), or together with WSC (F–J) for 48 hours, before cultured cells were immunostained against Ki67 (red) and counterstained with DAPI. Scale bar, 10 mm. K, Percentage of Ki67-positive cells in cultured Ptch1þ/ medulloblastoma cells in the presence of simvastatin, triparanol, or with addition of WSC. Signiﬁcance asterisk key: P < 0.05 (), P < 0.01 (), P < 0.001 (), P < 0.0001 ().

empty vector. At 48 hours after infection with a Cre-bearing Antagonists of cholesterol synthesis inhibit medulloblastoma virus, Hh pathway activity was elevated in both models, as cell proliferation determined by increased levels of Gli1 expression (Fig. 3C and Because cholesterol is necessary for Hh signaling, we hypoth- D). Importantly, treatment with simvastatin or triparanol esized that inhibition of cholesterol synthesis might suppress dramatically inhibited Gli1 expression in Ptch1-deficient medulloblastoma cell proliferation. To test this idea, we exam- MEFs, and this inhibition was overcome by the addition of ined medulloblastoma cell proliferation in the presence or exogenous cholesterol, indicating that cholesterol is required absence of simvastatin or triparanol. Medulloblastoma cells iso- þ for Hh pathway activation resulting from Ptch1 deletion. In lated from Ptch1 / mice were briefly cultured in the presence of contrast, treatment of cells harboring an activated Smo gene cholesterol inhibitors or vehicle for 48 hours, and examined for with simvastatin or triparanol treatment failed to reduce Gli1 proliferation by immunocytochemistry. As shown in Fig. 4A, expression, indicating that cholesterol is dispensable for Hh about 60% of medulloblastoma cells were proliferating (Ki67 pathway signaling resulting from constitutively active Smo. positive) in control cultures, but the percentage of proliferating Consistent with the results described above, we found that cells declined significantly in a dose-dependent manner as the increased levels of Ptch1 caused by overexpression of Gli1 were concentration of simvastatin was increased (Fig. 4B–D). Reduced not altered by treatment with simvastatin or triparanol (Fig. proliferation of medulloblastoma cells was also observed after 3E). Taken together, our data demonstrate that cholesterol is treatment with triparanol (Fig. 4E). As shown in Fig. 4G–J, required for Smo function in the Hh signaling pathway addition of exogenous cholesterol rescued the proliferation (Fig. 3F). repressed by simvastatin or triparanol treatment (Fig. 4K),

1382 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Effectively Inhibit Medulloblastoma Growth

suggesting that reduced proliferation of medulloblastoma cells medulloblastoma resulting from Smo activation, we generated treated with simvastatin or triparanol was indeed due to choles- medulloblastoma using Math1-Cre/SmoM2 mice as described terol deficiency. We also noted that proliferation of vehicle-treated previously (41). As shown in Supplementary Fig. S4A–S4C, cells increased upon exposure to exogenous cholesterol, indicat- proliferation of medulloblastoma cells in vitro was not altered ing that the endogenous levels are rate limiting under the exper- after simvastatin treatment. In the medulloblastoma allograft imental conditions used (Fig. 4F). These data indicate that antago- mouse model generated using medulloblastoma cells from nists of cholesterol biosynthesis significantly inhibit medullo- Math1-Cre/SmoM2 mice, no obvious inhibition of medulloblastoma cell proliferation. We next tested whether pharmaceu- blastoma growth was observed after simvastatin treatment tical inhibition of cholesterol synthesis could be utilized to treat (Supplementary Fig. S4D–S4E). These data indicate that medulloblastoma. For this purpose, we generated a medulloblas- medulloblastoma cells carrying the SmoM2 mutation are resis- toma model by subcutaneous injection of CB17/SCID mice with tant to simvastatin treatment and that the SmoM2 mutation þ medulloblastoma cells isolated from Ptch1 / mice, as described uncouples Smo from cholesterol dependency. previously (19). This model was validated during the development of vismodegib and shown to accurately predict clinical Statins synergize with Smo antagonist in medulloblastoma responses in humans (37, 38). In contrast, human tumor cell treatment lines fail to sustain Hh pathway activity; therefore, they cannot be Previous studies reported that oxysterols and cholesterol bind used to test the efficacy of inhibitors of Hh pathway activity (19). to Smo at CRD-binding site, which is distinct from the binding After the allograft tumor volume reached 200–400 mm3, mice pocket of the conventional Smo antagonist, vismodegib (42). were treated by intraperitoneal injection of 20 mg/kg/day simva- This led us to investigate the possible synergistic effects of statins statin, 40 mg/kg/day simvastatin, or vehicle. No significant dif- and vismodegib in the treatment of medulloblastoma. For this ference in tumor growth was observed between the control and 20 purpose, mice bearing subcutaneous allograft tumors from þ mg/kg/day simvastatin treatment (P ¼ 0.32; Fig. 5A). The volume Ptch1 / mice were treated with vismodegib, simvastatin, or the of tumors from the group of mice treated with 20 mg/kg/day two in combination. Consistent with previous reports (43), simvastatin was comparable with that from the control group treatment with vismodegib at 5 mg/kg/day reduced Hh-MB after 21 days of treatment (Fig. 5B). However, tumor volume was growth (Fig. 6A and B). However, combined treatment with dramatically reduced after treatment with 40 mg/kg/day simva- vismodegib and simvastatin resulted in a complete inhibition of statin, compared with either vehicle (P < 0.001) or with 20 mg/kg/ Hh-MB progression. Although 20 mg/kg/day simvastatin exhib- day (P < 0.001), suggesting that simvastatin at a dose of 40 mg/kg/ ited no inhibitory effect on medulloblastoma growth (P ¼ 0.32), day effectively suppressed medulloblastoma growth in vivo.To 5 mg/kg/day vismodegib in combination with 20 mg/kg/day determine the mechanism responsible for suppressed tumor simvastatin resulted in an additional reduction in tumor growth growth by simvastatin, we examined the proliferation and sur- compared with vismodegib alone (P < 0.001). These data suggest vival of tumor cells following simvastatin treatment by IHC. that simvastatin and vismodegib act synergistically to prohibit Proliferation of medulloblastoma cells (Ki67 positive) was tumor growth. The synergistic effect of simvastatin and vismode- markedly reduced after simvastatin treatment (Fig. 5C–C00) com- gib in medulloblastoma treatment was confirmed statistically as pared with vehicle. Moreover, simvastatin treatment promoted explained in the Materials and Methods section. þ significant differentiation, evidenced by upregulation of NeuN in CB17/SCID mice carrying allografts of Ptch1 / medulloblas- medulloblastoma tissue (Fig. 5D–D00). In addition, an increased toma were also treated with another HMG-CoA inhibitor, ator- number of apoptotic cells (CC3 positive) was observed in tumor vastatin, alone or combined with vismodegib. Atorvastatin treat- tissues after simvastatin treatment, indicating increased apoptosis ment at 10 mg/kg/day (P < 0.001) or 40 mg/kg/day (P < 0.001) (Fig. 5E–E00). On the basis of qPCR and Western blot analyses, significantly repressed tumor growth compared with vehicle Gli1 mRNA and protein expression were significantly downregu- treatment, confirming inhibitory effects of statins on medullo- lated in tumors treated with 40 mg/kg/day simvastatin compared blastoma growth (Fig. 6C and D). Moreover, 10 mg/kg/day with the vehicle control (Fig. 5F–H), suggesting simvastatin atorvastatin dramatically enhanced the inhibitory effects of vis- compromised Hh signal transduction in medulloblastoma cells. modegib on medulloblastoma growth, when compared with þ In addition, Ptch1 / mice with established medulloblastoma vismodegib treatment alone (P < 0.001), further confirming the were treated with 40 mg/kg/day simvastatin for 2 weeks. As shown synergistic effects of statins and vismodegib in inhibiting medul- in Supplementary Fig. S3A–S3D, simvastatin markedly inhibited loblastoma growth. tumor cell proliferation and promoted tumor cell differentiation Having observed the synergistic effects of statins and vismo- in medulloblastoma tissue. Moreover, activation of Hh pathway degib in medulloblastoma treatment, we next investigated in medulloblastoma tissue was repressed following simvastatin whether cholesterol could affect the interaction between Smo treatment. Together, these data demonstrate that simvastatin and vismodegib. For this purpose, we pretreated NIH 3T3 cells þ treatment can reduce medulloblastoma growth in vivo. and Ptch1 / medulloblastoma cells with WSC and then exam- The SmoM2 mutation results in an amino acid substitution ined the inhibitory effects of vismodegibonHhsignalingbased (W535L) in the seventh transmembrane domain, creating a on Gli1 protein expression. As expected, vismodegib signifi- constitutively active form of Smo by disrupting G-protein cantly inhibited Gli1 expression in Hh-treated 3T3 cells (Fig. þ coupling (39). The SmoM2 mutation has been observed in 6E) and Ptch1 / medulloblastoma cells (Fig. 6F). However, human medulloblastoma and in basal cell carcinoma (39, 40). pretreatment with cholesterol abrogated vismodegib-induced Hh signaling in MEF cells carrying a SmoM2 mutation was not inhibition of Hh signaling in a dose-dependent manner. These affected by simvastatin treatment (Fig. 3D), suggesting that datasuggestthatathighconcentrations, binding of cholesterol activation of Hh pathway by SmoM2 is not dependent on to Smo may prevent its interaction with vismodegib, perhaps cholesterol. To test the effect of simvastatin on the growth of through an allosteric effect, providing a potential basis for the

www.aacrjournals.org Clin Cancer Res; 24(6) March 15, 2018 1383

Gordon et al.

Figure 5. Simvastatin inhibits growth of Ptch1þ/ medulloblastoma (MB). A, Growth rate of subcutaneous Ptch1þ/ medulloblastoma following treatment with simvastatin or vehicle control. B, Tumor size of subcutaneous Ptch1þ/ medulloblastoma following drug treatment. C–E00, Sections of subcutaneous Ptch1þ/ medulloblastoma after treatment with vehicle (C–E) or 40 mg/kg/day simvastatin (C0–E0) were immunostained for Ki67 (C–C00), NeuN (D–D00), or CC3 (E–E00). Percentage of Ki67þ cells (C00), NeuNþ cells (D00), and CC3þ cells (E00) in medulloblastoma tissue after designated drug treatment. Medulloblastoma tissues were counterstained with DAPI. Scale bar, 10 mm. F–H, mRNA (F) and protein (G) expression of Gli1 in Ptch1þ/ medulloblastoma tissues after drug treatment were analyzed by qPCR and Western blotting, respectively. H, Quantiﬁcation of Gli1 protein expression. Signiﬁcance asterisk key: P < 0.05 (), P < 0.01 (), P < 0.001 ().

synergistic effect of statins and vismodegib in medulloblastoma organization, dynamics, and function. During normal devel- treatment. opment, cholesterol homeostasis is tightly regulated by a complex protein network involving importation, biosynthesis, exportation, metabolism, and esteriﬁcation. Altered expression Discussion levels, and mutations of genes involved in the cholesterol Cholesterol is an essential component of plasma mem- homeostasis pathways, have been identiﬁed in many human branes, and it plays an important role in cell membrane malignancies, including glioblastoma, pancreatic cancer, and

1384 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Effectively Inhibit Medulloblastoma Growth

melanoma (44). Impaired cholesterol homeostasis in tumor activity (8–11). However, it is unclear whether these oxysterols cells influences tumor progression and dictates the sensitivity of are physiologic activators of Smo, given that the endogenous tumor cells to chemotherapy (45), although it is still debated levels are significantly lower than the EC50 for Hh pathway whether cellular cholesterol levels influence tumor initiation. activation. In this study, we found that, upon inhibition of Our studies reveal enhanced expression of genes associated cholesterol biosynthesis, using proximal and distal inhibitors with cholesterol biosynthesis in Hh-MB, suggesting that the (simvastatin and triparanol, respectively), Hh pathway activity intracellular cholesterol levels in tumor cells are elevated. in medulloblastoma cells and fibroblasts was significantly Inactivating mutations of Ptch1, the most common mutation repressed. Proliferation of medulloblastoma cells and tumor in Hh-MB, may also contribute to increased cholesterol levels growth in vivo was repressed after pharmacologic or genetic in tumor cells, in that Ptch1 is involved in cholesterol efflux inhibition of cholesterol biosynthesis. These data suggest that from cells (46). endogenous cholesterol biosynthesis is important for Hh path- Extensive evidence indicates that some exogenous oxysterols, way activity. Compromised Hh pathway activity was observed such as 20 (S)-hydroxycholesterol and 25-hydroxycholesterol in cholesterol-deficient fibroblasts after loss of Ptch1 but not generated by cholesterol oxidation, activate Smo by binding to following constitutive activation of Smo or after overexpression a site located in the extracellular CRD, demonstrating that of Gli1. These data indicate that cholesterol biosynthesis is cholesterol derivatives are capable of triggering Hh pathway required for Smo to function in Hh signaling. Recent findings

Figure 6. Statins synergize with vismodegib in inhibiting Ptch1þ/ medulloblastoma (MB) growth. A and B, Growth rate (A) and size (B) of subcutaneous Ptch1þ/ medulloblastoma treated with 20 mg/kg/day simvastatin, 5 mg/kg/day vismodegib, or in combination. C and D, Growth rate (C) and size (D) of subcutaneous Ptch1þ/ medulloblastoma treated with 10 mg/kg/day, 40 mg/kg/day atorvastatin, 5 mg/kg/day vismodegib, or in combination. E, Western blot analysis of Gli1 protein expression in NIH 3T3 cells treated with vismodegib for 48 hours following overnight treatment with WSC and Hh-CM. F, Western blot analysis of Gli1 protein expression in Ptch1þ/ medulloblastoma cells treated with vismodegib for 48 hours after overnight treatment with WSC.

www.aacrjournals.org Clin Cancer Res; 24(6) March 15, 2018 1385

Gordon et al.

that cholesterol synergizes with Hh to activate Smo through approved for use in children (48, 49). So far, there have been no binding to the extracellular CRD (12, 13) are consistent with reports of bone toxicities or other phenotypes that could be the conclusion that cholesterol, but not oxysterols, functions ascribed to Hh pathway inhibition. Nevertheless, further studies as the endogenous ligand responsible for Smo activation are warranted to test the potential for combined toxicities of Smo during Hh pathway signaling, and that Ptch1 inhibits Smo inhibitors and statins in developing animal models prior to their by counteracting its activation by cholesterol. Thus, cholesterol use in children. plays a unique role as a second messenger in vertebrate Hh The potential efficacy of statins in Hh-MB therapy depends signaling, by mediating the functional interaction between on which underlying mutation the tumor harbors in the Hh Ptch1 and Smo. These findings resolve a longlasting question pathway. Among Hh-MB patients, approximately 80% of adult of how Ptch1 suppresses Smo activity, and they identify together with 50% of infant and childhood patients harbor cholesterol as a critical component for Smo activation by mutations in Hh, Ptch1, or Smo, who are most likely to Ptch1-mediated Hh signaling. Simvastatin also inhibits pro- respond to Smo antagonists including statins (40). However, liferation of group 3 medulloblastoma, which are associated medulloblastoma carrying the SmoM2 mutation are predicted with activating mutations in c-Myc and dysfunctional p53 (47). to be resistant to Smo inhibitors as well as cholesterol inhibi- These findings imply that cholesterol may regulate many other tors. Similarly, around 45% of infant and childhood Hh-MB signaling pathways in medulloblastoma in addition to Hh patients frequently harbor mutations downstream of Smo, such signaling. as Sufu or Gli2 (40), rendering these tumors intrinsically The blood–brain barrier efficiently blocks transport of choles- resistant to any drugs targeting Smo. terol from the bloodstream into the brain. Therefore, nearly Our studies reveal the synergistic effects of statin and all brain cholesterol is produced via de novo synthesis (1). Recent- vismodegib in treating medulloblastoma. It has been con- ly, it has been demonstrated that tumor cells in glioblastoma firmed that cholesterol activates Smo through interaction with multiforme, the most malignant brain tumor in adults, exclusive- the CRD region, which is distinct from the binding site of ly acquire cholesterol from astrocytes (45). In our studies, con- vismodegib. Recent structural studies found that binding with ditional inhibition of cholesterol biosynthesis in tumor cells vismodegib precludes the interaction of Smo with cholesterol dramatically inhibited medulloblastoma cell proliferation (14). In our studies, pretreatment with cholesterol overcame and promoted the differentiation of tumor cells, whereas disrup- the capacity of vismodegib to inhibit Hh signaling in NIH 3T3 tion of cholesterol synthesis in astrocytes had no effects in cells and medulloblastoma cells, implying that cholesterol medulloblastoma cell proliferation and survival. These observa- compromises the interaction between vismodegib and Smo. tions suggest that medulloblastoma cells predominately rely on de These data suggest an indirect competition between cholesterol novo synthesis of cholesterol. In our studies, cholesterol was not and vismodegib, possibly through allosteric effects, in binding completely depleted after deletion of the Nsdhl gene in medul- to Smo. T2 fl/fl þ loblastoma cells in Math1-CreER /Nsdhl /Ptch1 / mice fol- Treatment with Smo antagonists results in dramatic regression lowing tamoxifen treatment. On the basis of filipin staining, of tumor burden and reduction of symptoms, but there is a rapid some regions of the tumor tissue were enriched with choles- emergence of drug resistance (50). The resistance often arises as a terol. Such incomplete depletion of cholesterol in tumor tissue consequence of mutations in the transmembrane domain of Smo could be due to insufficient Cre recombination and/or inade- that prevent binding of antagonists. Therefore, cholesterol inhi- quate tamoxifen penetration. However, we cannot completely bitors could not only synergize with vismodegib in treating exclude the possibility that tumor cells at least partially acquire medulloblastoma, but also potentially be effective in treating cholesterol from their microenvironment. Nevertheless, prolif- some drug-resistant Hh-MB. erating medulloblastoma cells predominantly accumulated in T2 fl/fl cholesterol-enriched tumor regions of Math1-CreER /Nsdhl / Disclosure of Potential Conflicts of Interest þ/ Ptch1 mice, further confirming that cholesterol is critical for No potential conflicts of interest were disclosed by the authors. medulloblastoma growth. The observation of Hh pathway dysregulation in tumors has Authors' Contributions caused it to be an attractive pharmacologic target for cancer Conception and design: R.E. Gordon, J.M.Y. Ng, A.J. Andrews, I. Astsaturov, treatment for many years. All the available Hh pathway inhibitors T. Curran, Z.-J. Yang being studied in clinical trials target Smo. Our studies demon- Development of methodology: R.E. Gordon, F. Du, J.M.Y. Ng, Z.-J. Yang strate that cholesterol biosynthesis is essential for medulloblas- Acquisition of data (provided animals, acquired and managed patients, toma progression. Statins effectively inhibited medulloblastoma provided facilities, etc.): R.E. Gordon, L. Zhang, Y.-M. Kuo, J.M.Y. Ng, Z.-J. Yang Analysis and interpretation of data (e.g., statistical analysis, biostatistics, cell proliferation and suppressed tumor growth. The excellent computational analysis): R.E. Gordon, S. Peri, Y.-M. Kuo, B.L. Egleston, fi safety pro le, together with low cost and ready availability of J.M.Y. Ng, T. Curran, Z.-J. Yang statins, makes cholesterol biosynthesis a promising target for Writing, review, and/or revision of the manuscript: R.E. Gordon, L. Zhang, medulloblastoma treatment and perhaps other malignancies Y.-M. Kuo, B.L. Egleston, J.M.Y. Ng, T. Curran, Z.-J. Yang involving Hh signaling. Administrative, technical, or material support (i.e., reporting or organizing Smo inhibitors cause developmental bone abnormalities in data, constructing databases): R.E. Gordon, F. Du, J.M.Y. Ng, T. Curran, Z.-J. Yang young mice (43). Therefore, they are not recommended for use in Study supervision: R.E. Gordon, J.M.Y. Ng, T. Curran, Z.-J. Yang young children. Potentially, the combined use of vismodegib and statins could permit the use of lower doses of Smo inhibitors, Acknowledgments potentially reducing the extent of bone toxicity in the pediatric The authors would like to thank Drs. G.E. Herman and D. Cunningham (The fl/fl Hh-MB patient population. Currently, statins are being used to Ohio State University, Columbus, OH) for providing Nsdhl mice, Drs. T. Yen manage dyslipidemia in some children and are among the drugs and E. Golemis for thoughtful comments, and A. Efimov, E. Nicolas, F. Jin, and J.

1386 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Effectively Inhibit Medulloblastoma Growth

Oesterling for technical assistance. This work was also supported by NCI The costs of publication of this article were defrayed in part by the payment of (CA178380, CA185504), ACS RSG (RSG1605301NEC), PA CURE Health page charges. This article must therefore be hereby marked advertisement in Research Fund, Bucks County Chapter, and National Natural Science Founda- accordance with 18 U.S.C. Section 1734 solely to indicate this fact. tion of China (81572724, 81573449). This research was supported by the subsidy of the Russian Government to support the Program of competitive Received October 5, 2017; revised December 1, 2017; accepted January 9, growth of Kazan Federal University to Renata E. Gordon. 2018; published OnlineFirst February 6, 2018.

References 1. Dietschy JM, Turley SD. Cholesterol metabolism in the brain. Curr Opin 21. Tomayko MM, Reynolds CP. Determination of subcutaneous tumor Lipidol 2001;12:105–12. size in athymic (nude) mice. Cancer Chemother Pharmacol 1989;24: 2. Goldstein JL, Brown MS. Regulation of the mevalonate pathway. Nature 148–54. 1990;343:425–30. 22. Yang ZJ, Ellis T, Markant SL, Read TA, Kessler JD, Bourboulas M, et al. 3. Liu XY, Dangel AW, Kelley RI, Zhao W, Denny P, Botcherby M, et al. The Medulloblastoma can be initiated by deletion of Patched in lineage- gene mutated in bare patches and striated mice encodes a novel 3beta- restricted progenitors or stem cells. Cancer Cell 2008;14:135–45. hydroxysteroid dehydrogenase. Nat Genet 1999;22:182–7. 23. Hanzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for 4. Cunningham D, DeBarber AE, Bir N, Binkley L, Merkens LS, Steiner RD, microarray and RNA-seq data. BMC Bioinformatics 2013;14:7. et al. Analysis of hedgehog signaling in cerebellar granule cell precursors 24. Leardi R. Experimental design in chemistry: a tutorial. Anal Chim Acta in a conditional Nsdhl allele demonstrates an essential role for cho- 2009;652:161–72. lesterol in postnatal CNS development. Hum Mol Genet 2015;24: 25. Thompson MC, Fuller C, Hogg TL, Dalton J, Finkelstein D, Lau CC, et al. 2808–25. Genomics identifies medulloblastoma subgroups that are enriched for 5. Avigan J, Steinberg D, Vroman HE, Thompson MJ, Mosettig E. Studies of specific genetic alterations. J Clin Oncol 2006;24:1924–31. cholesterol biosynthesis. I. The identification of desmosterol in serum and 26. Northcott PA, Korshunov A, Witt H, Hielscher T, Eberhart CG, Mack S, et al. tissues of animals and man treated with MER-29. J Biol Chem 1960; Medulloblastoma comprises four distinct molecular variants. J Clin Oncol 235:3123–6. 2011;29:1408–14. 6. Fuse N, Maiti T, Wang B, Porter JA, Hall TM, Leahy DJ, et al. Sonic hedgehog 27. Akino T, Han X, Nakayama H, McNeish B, Zurakowski D, Mammoto A, protein signals not as a hydrolytic enzyme but as an apparent ligand for et al. Netrin-1 promotes medulloblastoma cell invasiveness and angio- patched. Proc Natl Acad Sci U S A 1999;96:10992–9. genesis, and demonstrates elevated expression in tumor tissue and 7. Kalderon D. Transducing the hedgehog signal. Cell 2000;103:371–4. urine of patients with pediatric medulloblastoma. Cancer Res 2014;74: 8. Corcoran RB, Scott MP. Oxysterols stimulate Sonic hedgehog signal trans- 3716–26. duction and proliferation of medulloblastoma cells. Proc Natl Acad Sci U S 28. Goodrich LV, Milenkovic L, Higgins KM, Scott MP. Altered neural cell fates A 2006;103:8408–13. and medulloblastoma in mouse patched mutants. Science 1997;277: 9. Nachtergaele S, Mydock LK, Krishnan K, Rammohan J, Schlesinger PH, 1109–13. Covey DF, et al. Oxysterols are allosteric activators of the oncoprotein 29. Wetmore C, Eberhart DE, Curran T. Loss of p53 but not ARF accelerates Smoothened. Nat Chem Biol 2012;8:211–20. medulloblastoma in mice heterozygous for patched. Cancer Res 2001;61: 10. Nedelcu D, Liu J, Xu Y, Jao C, Salic A. Oxysterol binding to the extracellular 513–6. domain of Smoothened in Hedgehog signaling. Nat Chem Biol 2013; 30. van der Wulp MY, Verkade HJ, Groen AK. Regulation of cholesterol 9:557–64. homeostasis. Mol Cell Endocrinol 2013;368:1–16. 11. Myers BR, Sever N, Chong YC, Kim J, Belani JD, Rychnovsky S, et al. 31. Venkateswaran A, Laffitte BA, Joseph SB, Mak PA, Wilpitz DC, Edwards PA, Hedgehog pathway modulation by multiple lipid binding sites on the et al. Control of cellular cholesterol efflux by the nuclear oxysterol receptor smoothened effector of signal response. Dev Cell 2013;26:346–57. LXR alpha. Proc Natl Acad Sci U S A 2000;97:12097–102. 12. Huang P, Nedelcu D, Watanabe M, Jao C, Kim Y, Liu J, et al. Cellular 32. Goldstein JL, DeBose-Boyd RA, Brown MS. Protein sensors for membrane cholesterol directly activates smoothened in hedgehog signaling. Cell sterols. Cell 2006;124:35–46. 2016;166:1176–87. 33. Mao X, Fujiwara Y, Chapdelaine A, Yang H, Orkin SH. Activation of EGFP 13. Luchetti G, Sircar R, Kong JH, Nachtergaele S, Sagner A, Byrne EF, et al. expression by Cre-mediated excision in a new ROSA26 reporter mouse Cholesterol activates the G-protein coupled receptor Smoothened to strain. Blood 2001;97:324–6. promote Hedgehog signaling. Elife 2016;5:e20304. 34. Yokota N, Aruga J, Takai S, Yamada K, Hamazaki M, Iwase T, et al. 14. Byrne EF, Sircar R, Miller PS, Hedger G, Luchetti G, Nachtergaele S, et al. Predominant expression of human zic in cerebellar granule cell lineage Structural basis of Smoothened regulation by its extracellular domains. and medulloblastoma. Cancer Res 1996;56:377–83. Nature 2016;535:517–22. 35. Ganat YM, Silbereis J, Cave C, Ngu H, Anderson GM, Ohkubo Y, et al. Early 15. Bar EE, Chaudhry A, Farah MH, Eberhart CG. Hedgehog signaling postnatal astroglial cells produce multilineage precursors and neural stem promotes medulloblastoma survival via Bc/II. Am J Pathol 2007;170: cells in vivo. J Neurosci 2006;26:8609–21. 347–55. 36. Jeong J, Mao J, Tenzen T, Kottmann AH, McMahon AP. Hedgehog signaling 16. Dimitroulakos J, Ye LY, Benzaquen M, Moore MJ, Kamel-Reid S, Freedman in the neural crest cells regulates the patterning and growth of facial MH, et al. Differential sensitivity of various pediatric cancers and squamous primordia. Genes Dev 2004;18:937–51. cell carcinomas to lovastatin-induced apoptosis: therapeutic implications. 37. Gajjar A, Stewart CF, Ellison DW, Kaste S, Kun LE, Packer RJ, et al. Phase I Clin Cancer Res 2001;7:158–67. study of vismodegib in children with recurrent or refractory medulloblas- 17. Wang W, Macaulay RJ. Mevalonate prevents lovastatin-induced apoptosis toma: a pediatric brain tumor consortium study. Clin Cancer Res 2013; in medulloblastoma cell lines. Can J Neurol Sci 1999;26:305–10. 19:6305–12. 18. Takwi AA, Li Y, Becker Buscaglia LE, Zhang J, Choudhury S, Park AK, et al. A 38. Robinson GW, Orr BA, Wu G, Gururangan S, Lin T, Qaddoumi I, et al. statin-regulated microRNA represses human c-Myc expression and func- Vismodegib exerts targeted efficacy against recurrent sonic hedgehog- tion. EMBO Mol Med 2012;4:896–909. subgroup medulloblastoma: results from phase II pediatric brain tumor 19. Sasai K, Romer JT, Lee Y, Finkelstein D, Fuller C, McKinnon PJ, consortium studies PBTC-025B and PBTC-032. J Clin Oncol 2015;33: et al. Shh pathway activity is down-regulated in cultured medullo- 2646–54. blastoma cells: implications for preclinical studies. Cancer Res 2006; 39. Xie J, Murone M, Luoh SM, Ryan A, Gu Q, Zhang C, et al. Activating 66:4215–22. smoothened mutations in sporadic basal-cell carcinoma. Nature 1998; 20. Meyers-Needham M, Lewis JA, Gencer S, Sentelle RD, Saddoughi SA, Clarke 391:90–2. CJ, et al. Off-target function of the Sonic hedgehog inhibitor cyclopamine 40. Kool M, Jones DT, J€ager N, Northcott PA, Pugh TJ, Hovestadt V, et al. in mediating apoptosis via nitric oxide-dependent neutral sphingomyeli- Genome sequencing of SHH medulloblastoma predicts genotype-related nase 2/ceramide induction. Mol Cancer Ther 2012;11:1092–102. response to smoothened inhibition. Cancer Cell 2014;25:393–405.

www.aacrjournals.org Clin Cancer Res; 24(6) March 15, 2018 1387

Gordon et al.

41. Schuller U, Heine VM, Mao J, Kho AT, Dillon AK, Han YG, et al. Acquisition 46. Bidet M, Joubert O, Lacombe B, Ciantar M, Nehme R, Mollat P, et al. The of granule neuron precursor identity is a critical determinant of progenitor hedgehog receptor patched is involved in cholesterol transport. PLoS One cell competence to form Shh-induced medulloblastoma. Cancer Cell 2011;6:e23834. 2008;14:123–34. 47. Pei Y, Liu KW, Wang J, Garancher A, Tao R, Esparza LA, et al. HDAC and 42. Robarge KD, Brunton SA, Castanedo GM, Cui Y, Dina MS, Goldsmith R, PI3K antagonists cooperate to inhibit growth of MYC-driven medulloblas- et al. GDC-0449-a potent inhibitor of the hedgehog pathway. Bioorg Med toma. Cancer Cell 2016;29:311–23. Chem Lett 2009;19:5576–81. 48. Eiland LS, Luttrell PK. Use of statins for dyslipidemia in the pediatric 43. Kimura H, Ng JM, Curran T. Transient inhibition of the Hedgehog pathway population. J Pediatr Pharmacol Ther 2010;15:160–72. in young mice causes permanent defects in bone structure. Cancer Cell 49. Wiegman A, Hutten BA, de Groot E, Rodenburg J, Bakker HD, Buller€ HR, 2008;13:249–60. et al. Efﬁcacy and safety of statin therapy in children with familial 44. Kuzu OF, Noory MA, Robertson GP. The role of cholesterol in cancer. hypercholesterolemia: a randomized controlled trial. JAMA 2004;292: Cancer Res 2016;76:2063–70. 331–7. 45. Villa GR, Hulce JJ, Zanca C, Bi J, Ikegami S, Cahill GL, et al. An LXR- 50. Rudin CM, Hann CL, Laterra J, Yauch RL, Callahan CA, Fu L, et al. Treatment cholesterol axis creates a metabolic co-dependency for brain cancers. of medulloblastoma with hedgehog pathway inhibitor GDC-0449. N Engl Cancer Cell 2016;30:683–93. J Med 2009;361:1173–8.

1388 Clin Cancer Res; 24(6) March 15, 2018 Clinical Cancer Research

Statins Synergize with Hedgehog Pathway Inhibitors for Treatment of Medulloblastoma

Renata E. Gordon, Li Zhang, Suraj Peri, et al.

Clin Cancer Res 2018;24:1375-1388. Published OnlineFirst February 6, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-17-2923

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2021/02/25/1078-0432.CCR-17-2923.DC1

Cited articles This article cites 50 articles, 20 of which you can access for free at: http://clincancerres.aacrjournals.org/content/24/6/1375.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/24/6/1375.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/24/6/1375. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.