Multitarget hedgehog pathway inhibitors and uses thereof

Abstract

The present invention concerns compounds that selectively inhibit the activity of the Hedgehog (Hh) pathway, their preparation and uses thereof. The compounds of the present invention are useful in treating Hh-dependent tumors, such as medulloblastoma (MB).

Claims

1. A method for the treatment of a Hedgehog (Hh)-dependent tumor pathology comprising: administering to a human subject needing inhibition of the Hedgehog (Hh)-dependent tumor pathology, an effective amount of the compound having general formula I, ##STR00029## wherein: R.sub.1 is OR.sub.A; R.sub.2 is OR.sub.B; wherein each of R.sub.A and R.sub.B is hydrogen or acyclic branched or straight, saturated or unsaturated aliphatic chain having from 1 to 10 carbon atoms; R.sub.3 is OR.sub.C; and R.sub.4 is OR.sub.D; wherein each of R.sub.C and R.sub.D are methyl, ethyl, propyl, isopropyl, prenyl, geranyl, farnesyl or benzyl, wherein the phenyl of the benzyl group can be substituted by halogen, C.sub.1-C.sub.6 haloalkyl, C.sub.1-C.sub.6 haloalkoxy, C.sub.1-C.sub.6 alkoxy, C.sub.1-C.sub.6 alkylthio, amino, C.sub.1-C.sub.6 alkylamino, or C.sub.1-C.sub.6 dialkylamino, wherein the compound is an antagonist of Gli1 and the effective amount inhibits activation of Gli1 associated with the Hh-pathway.

2. The method of claim 1, wherein the Hh-dependent tumor pathology is resistant to a SMO inhibitor.

3. The method of claim 2, wherein the SMO inhibitor is vismodegib or NVP-LDE225.

4. The method of claim 1, wherein the Hh-dependent tumor pathology is selected from the group consisting of: medulloblastoma (MB), esophageal adenocarcinoma, basal cell carcinomas (BCCs), pancreatic, prostate, and small cell lung cancer.

5. The method of claim 1, wherein the Hh-dependent tumor pathology is present in a pediatric human subject.

6. The method of claim 1, wherein the compound is administered by intratumoral injection.

Description

(1) The invention will now be illustrated by non-limiting examples referring to the following figures.

(2) FIG. 1. Representative Type1 (A) and Type2 (B) pharmacophore models used for screening the in house library of natural compounds. Pharmacophoric features are showed as spheres. Hydrogen bond acceptor (HBA); hydrophobic (HYD); Hydrogen bond donor (HBD).

(3) FIG. 2. The best alignment of the most potent compound Glabrescione B toward the representative Type1 pharmacophore. For clarity, non-polar hydrogen atoms were hidden. Heavy atoms are colored as follows: grey=carbon; red=oxygen.

(4) FIG. 3. The predicted binding mode of Glabrescione B (showed as sticks). A: Gli1-ZF (zinc finger) is represented as lines and cartoons. Residues interacting with Glabrescione B are showed as sticks. H-bonds are highlighted by dashed lines. B: Gli1-ZF is showed as transparent surface and Glabrescione B is interacting within ZF4 and ZF5. Zinc fingers 3 and 5 are also labeled. The protein is in the same orientation as in A. T374 is Threonine 374, S378 is Serine 378, K350 is Lysine 350 and F342 is Phenylalanine 342.

(5) FIG. 4. Effect of active natural compounds identified by the pharmacophore screening in a luciferase assay, where the gene reporter was Gli1 and the pathway was activated by the SAG molecule, a potent SMO agonist. Natural compounds were tested at three different concentrations (1, 10 and 20 M) and their activity was compared to that of cyclopamine, the natural compound reference inhibitor of the Hedgehog pathway, which was tested at the same doses (1, 10 and 20 M). On both graphs the activity of cyclopamine is reported as reference standard. (A) cyclopamine, 2,4,5,3,4-penta-Ome chalcone and Auriculasin; (B) cyclopamine, Martinoside, Jaceidin, 2,4,5,3,4-penta-Ome chalcone, 3,4-di-MDO-2,4,5-tri-Ome dihydrochalcone.

(6) FIG. 5. (A) HEK293 cells were transiently transfected with pcDNA-SMO vector and incubated with Bodipy-cyclopamine (BC 5 nM) alone (control) or in the presence of increasing concentrations of KAAD (Cyclopamine-KAAD, Smoothened antagonist) or increasing concentrations of Glabrescione B, as indicated. Bodipy-cyclopamine binding (BC binding) (green) is visualized using fluorescence microscopy in a representative field. (B) The concentration-response curve for Glabrescione B was obtained by quantification of the Bodipy-cyclopamine fluorescence in three photographs for each coverslip. The values are expressed as a percentage of the fluorescence detected in control HEK293 cells incubated with Bodipy-cyclopamine alone. The data shown are the means of triplicates derived from a representative experiment of three. Error bars indicate SD (standard deviation).

(7) FIG. 6. Dose-response curve of Glabrescione B in SAG (Smoothened Agonist, 200 nM) treated NIH 3T3 Shh-light II (Shh-LII) cells. Treatment time was 48 h and normalization was against Renilla luciferase. Shown is the fold increase of Hh reporter activity compared with cells not treated with SAG.

(8) FIG. 7. Inhibition of Gli1-induced transcription in transfected HEK293 cells. HEK293 cells were transfected with 12GliBS-Luc and pRL-TK Renilla (normalization control) plus control (empty vector) or Gli1 vector and treated with Glabrescione B or GANT61 (20 M) or vehicle only (DMSO, dimethyl sulfoxide 0.5%Gli1 bar in the graph) for 24 h and luciferase activity tested. Shown is the mean of three independent experiments. Error bars indicate SD.

(9) FIG. 8. Confluent MEF-Ptch.sup./ cells were treated with DMSO (0.5%, vehicle), KAAD (1 M, as positive control) or different concentrations of Glabrescione B for 48 h. (A) Gli1 mRNA levels were determined by quantitative RT-PCR, normalizing to HPRT (Hypoxanthine-guanine phosphoribosyl transferase) expression. Error bars indicate SD. (B) Western blot analysis shows the decrease of Gli1 protein levels after 48 h of treatment with compound 10 M.

(10) FIG. 9. Confluent MEF-SuFu.sup./ cells were treated with DMSO (0.5%, vehicle), GANT61 (10 M, as positive control) or different concentrations of Glabrescione B for 48 h. Gli1 mRNA levels were determined by quantitative RT-PCR, normalizing to HPRT expression. Error bars indicate SD.

(11) FIG. 10. Cerebellar granule progenitor cells (GCPs) were treated with Shh (Recombinant mouse Sonic Hedgehog, amino-terminal peptide, 3 g/ml) and with different concentrations of Glabrescione B, as indicated, for 48 h. Inhibition of cell proliferation was measured as percentage of BrdU incorporation in comparison to DMSO (0.5%, vehicle) treated sample.

(12) FIG. 11. (A) Human MB D283 cells were treated with different concentrations of Glabrescione B as indicated for 48 h. Inhibition of cell proliferation was measured as percentage of BrdU incorporation in comparison to DMSO (0.5%, vehicle) treated sample. (B) Human MB D283 cells were treated for 48 h with increasing concentrations of Glabrescione B as indicated. After 48 h, CellTiter 96 Aqueous One Solution Reagent was added to each well according to the manufacturer's instructions. After one hour in culture the cell viability (expressed in % viability) was determined by measuring the absorbance at 490 nm comparing treatment with DMSO (0.5%, vehicle) control.

(13) FIG. 12. Human MB D283 cells were treated with the DMSO (0.5%, vehicle) or with different concentrations of Glabrescione B as indicated and compared to an untreated sample (NT). After 48 h a tripan blue count was performed to determine the percentage of cell death.

(14) FIG. 13. Cancer stem cells (CSCs) isolated from murine Ptch.sup.+/ MB were treated with DMSO (0.5%, vehicle), KAAD (1 M), used as positive control, or different concentrations of Glabrescione B as indicated for 48 h. Gli1 mRNA levels were determined by quantitative RT-PCR using HPRT as normalizer.

(15) FIG. 14. Suspension of single CSCs isolated from murine Ptch.sup.+/ MB were treated with DMSO (0.5%, vehicle), KAAD (1 M) or different concentrations of Glabrescione B as indicated. After seven days of treatment the number of secondary neurospheres derived from a known number of single cells was counted. The self-renewal CSCs capability is expressed as percentage of neurospheres forming cells.

(16) FIG. 15. Inhibition of Gli-dependent BCC tumor cell growth. ASZ001 BCC cells were treated with Glabrescione B (GlaB) 5 M or DMSO only (A) or with Glabrescione B 1, 5 and 10 M (GlaB) or DMSO only (B). After the indicated times, a trypan blue count was performed to determine the growth rate (A) or the percentage of cell death (B). (C) Gli1 mRNA expression levels were determined by qRT-PCR after treatment of ASZ001 BCC cells with Glabrescione B (GlaB) or DMSO only for the indicated times. Results are expressed in arbitrary units as relative quantification normalized with endogenous control 2-microglobulin and HPRT. In all experiments data show the mean of three independent experiments. Error bars indicate SD. *P<0.05 vs DMSO.

(17) FIG. 16. qRT-PCR shows Hh targets mRNA expression levels determined in 6-day old mouse cerebellar progenitors after s.c. injections of Glabrescione B. Results are expressed in arbitrary units as relative quantification normalized with endogenous control 2-microglobulin and HPRT. Data show the meanSD of three independent experiments. *P<0.05 vs CTR

(18) FIG. 17. Glabrescione B inhibits Gli1-dependent MB tumor growth in vivo (Ptch1.sup.+/ MB allografts). (A) Change of tumor volume during Glabrescione B (GlaB) or vehicle (CTR) treatment period (days 18). (B) Representative flank allografts average volumes (upper panel); representative H&E staining of tumors (middle panel); immunohistochemistry shows Ki67 staining (lower panel). (C) qRT-PCR of Gli1 and Ptch1 mRNA expression levels normalized with endogenous control 2-microglobulin and HPRT. Data show the meanSD of tumor (n=6) for each treatment. *P<0.05 vs CTR.

(19) FIG. 18. Glabrescione B inhibits Gli1-dependent BCC tumor growth in vivo. (ASZ001 BCC allografts). (A) Change of tumor volume during Glabrescione B (GlaB) or vehicle (CTR) treatment period (18 days). (B) Representative flank allografts average volumes (upper panel); representative H&E staining of tumors (middle panel); immunohistochemistry shows Ki67 staining (lower panel). (C) qRT-PCR of Gli1 and Ptch1 mRNA. Shown is the meanSD of tumor (n=6) for each treatment. *P<0.05 vs CTR.

(20) FIG. 19. Effect of derivatives NT8, NT9, NT10, NT11, NT12, NT13, NT14 and NT15 on the Hedgehog pathway, measured at 5 M dose in a luciferase assay where the gene reporter was Gli1 and the pathway was activated by the SAG molecule, a potent SMO agonist.

DETAILED DESCRIPTION OF THE INVENTION

(21) To discover natural compounds as inhibitors of the Hh pathway (targeting the SMO receptor and/or the Gli1 protein), two distinct computational screening protocols were established, based on the availability of structural or molecular information. A ligand-based approach was followed to search for antagonists of the SMO receptor, based on the chemical structure of potent SMO antagonists described in the literature. Moreover, a receptor-based approach was used to target Gli1, based on the availability of a crystallographic structure of Gli1-ZF in complex with DNA (PDB ID: 2GLI) (Pavletich and Pabo 1993).

(22) In Silico Screening

(23) Pharmacophore Generation

(24) A training set of 9 active SMO antagonists, retrieved from the literature and from patents, was chosen to generate ligand-based pharmacophore models. Since multiple alignment schemas are allowed for selected compounds in the 3D space, multiple pharmacophores were generated accordingly by means of the Common Feature Pharmacophore Generation protocol implemented in the molecular modeling suite Discovery Studio 2.5 from Accelrys (Barnum, Greene et al. 1996). Pharmacophores were then scored and ranked based on their ability to map the training set and the six top-ranking pharmacophores were selected. They belong to two different groups: Type1three hydrogen bond acceptor (HBA) and three hydrophobic (HYD) features are necessary to represent the interaction pattern of potent SMO antagonists with the SMO receptor (FIG. 1A); Type2three HBA, two HYD and a hydrogen bond donor (HBD) features are necessary to represent the interaction pattern of potent SMO antagonists with the SMO receptor (FIG. 1B). Coordinates of pharmacophoric features and inter-feature distances are reported in Tables 1-4. It is to be specified that this training set has been already used for similar purposes (Manetti, Faure et al. 2010). However, the use of multiple pharmacophores represents a unique feature of the present invention, since in the previous attempt a single pharmacophore was used for filtering chemical libraries in searching for SMO antagonists (Manetti, Faure et al. 2010). The use of multiple pharmacophores for representing ligand binding to SMO better accounts for receptor flexibility, as well as enhances the research of chemically diverse candidate lead molecules with respect to the training set. Moreover, a HBD feature was used for the first time to represent the ligand interaction to SMO, as several known SMO antagonists are endowed with at least one HBD group.

(25) Pharmacophoric Screening

(26) The six pharmacophores described above were used as 3D query to filter the unique library of natural compounds, whose features are described in the example n.10. Ligand conformational analysis was performed by the CAESAR algorithm. Pharmacophore screening was performed with Discovery Studio 2.5. The Search 3D Database and the Ligand Pharmacophore Mapping protocols were subsequently used to filter the virtual library and to calculate the FitValue for each ligand, as a measure of how well a ligand fits a pharmacophore. Based on pharmacophore screening results, ligands were divided into three groups: 1) ligands that map all pharmacophores; 2) ligands that map only Type1 pharmacophores and 3) ligands that map only Type2 pharmacophores. A consensus rank by rank procedure was applied to merge results obtained by filtering through each pharmacophore model. Since the library included natural products with molecular dimension spanning from small to very large (for example, molecular volume ranges from 465.9 to 2981.8 .sup.3, Polar Surface Area from 8.2 to 265.6 ), with the aim of prioritizing small molecular compounds the Ligand Efficiency (LE) was further calculated for each selected compound as the ratio between the FitValue and the number of ligand heavy atoms (LE=FitValue/no. heavy atoms). 16 molecules endowed with the highest LE values were deemed top priority and selected for in vitro studies. The alignment of one exemplified compound, Glabrescione B, to a pharmacophore model is showed in FIG. 2.

(27) Pharmacophore Refinement

(28) Based on biological results from a luciferase assay where the gene reporter was Gli1 and the pathway was activated by the SAG molecule, a potent SMO agonist, 2 molecules were classified as highly active Hh inhibitors, showing more than 50% inhibition at 5 M (Glabrescione B and 2,4,5,3,4-penta-OMe Chalcone), 5 as moderately active, showing up to 30% inhibition at 30 M (Jaceidin, Auriculasin, 3,4-di-MDO-2,4,5-tri-OMe dihydrochalcone, 2,3,4,6,3,4-hexa-OMe Chalcone, Martinoside,) and 9 were not active (Barbinervic Acid, Kuwanol-E, Myricetin, Sorocein-A, Sorocein-B, Isosophoranon, Veratrin, Hesperidin, Naringin). Notably, a highly comparable average molecular weight for highly active (average MW=404.45) and moderately active Hh inhibitors (average MW=390.80) was observed, whereas a significantly higher molecular weight is showed by inactive compounds (average MW=503.51). Since tested compounds were composed only by C, O and H atoms, molecular weight was considered in this case as proportional to the molecular dimension. Therefore, the inventors envisioned that compounds having a high MW were not active probably because the ligand binding site on the SMO receptor was not large enough to accommodate these substances, or they encounter steric hindrance within the SMO binding site. Reducing the steric accessibility of pharmacophore models may improve computational results. Therefore, inactive compounds were used for the steric refinement of pharmacophores by means of the Steric Refinement with Excluded Volumes protocol of Discovery Studio 2.5 that places excluded volumes to pharmacophore models.

(29) Structure-Based Virtual Screening

(30) Based on the availability of a crystal structure of the zing-finger domain of Gli1-ZF in complex with DNA, a structure-based virtual screening protocol was established to screen the unique library of natural compounds, whose features are described in the example n.10. First, MD simulations were performed to relax atoms coordinates in explicit solvent and to sample the conformational space. Four different replicas of unrestrained MD were simulated for 20 ns each, starting from slightly different initial coordinates. A representative structure was extracted from MD trajectories at the convergence and further relaxed by means of energy minimization after removal of DNA. Subsequently, molecular docking was used to predict the theoretical binding mode and affinity of natural compounds towards the representative Gli1-ZF structure refined by MD. Based on literature data, the binding site was centered on T374 that is within ZF4 and ZF5 (Sheng, et al. 2006). Compounds of the unique library were docked to Gli1-ZF by means of the GoldScore function of the GOLD program (version 5.0.1) (Verdonk, et al. 2003), which is generally recommended for binding sites particularly solvent-exposed or accounting for several H-bonding ligand-protein interactions, such as the ZF4. Further, a rescoring procedure consisting on the calculation of the ligand delta energy of binding by means of the MM-GBSA methods was applied, with the aim of decreasing as much as possible the number of false positive identified by docking. After docking and rescoring, the LE was calculated as the ratio between the delta energy of binding and the number of heavy atoms of each ligand. Interestingly, Glabrescione B that was already evaluated as SMO antagonist was found within the top ranking positions of the structure-based virtual screening. For this reason, as well as to exploit the possibility to develop a multitarget inhibitor of the Hh pathway, Glabrescione B was deemed top priority for in vitro testing. The docking-based binding mode of Glabrescione B to Gli1-ZF is showed in FIG. 3.

(31) Pharmacology

(32) Glabrescione B potency was evaluated by measuring its ability to inhibit Hh pathway in a cellular context of Hh signaling hyper-activation. This condition, leading to constitutive activation of Gli transcription factors, can be triggered by overexpressing Gli1 or treating Hh-responsive cell lines with Shh ligand or SAG, a potent SMO agonist or can occur in cells following mutations of key components of the pathway.

(33) The effects of Glabrescione B as SMO and Gli1 antagonist have been examined by biochemical and in various cell-based assays.

(34) Biochemical Assay.

(35) Affinity of Glabrescione B and its direct binding to SMO receptor was quantified in a displacement assay. The assay is based on the use of the Bodipy-cyclopamine (BC), a fluorescent derivative of cyclopamine, which interacts with SMO at the level of its heptahelical bundle. To this end, HEK293 cells were transfected with a vector expressing SMO protein and then incubated with BC in the absence or presence of various concentrations of Glabrescione B. This assay revealed that Glabrescione B blocked BC binding to SMO in a dose-dependent manner with an IC.sub.50 of 1 M (FIG. 5). KAAD, used as positive control, abrogated BC binding to cell expressing SMO. Together, these results show that displacement assay is specific and demonstrate that Glabrescione B binds to SMO receptor at the level of the Bodipy-cyclopamine binding site.

(36) Cellular Assay.

(37) To investigate the inhibitory properties of Glabrescione B on Hh signaling, the inventors examined its effects in NIH 3T3 Shh-light II (Shh-LII) cells stably incorporating an Hh-responsive (Gli-RE) reporter, in which induction of the pathway occurs following treatment with the SMO agonist SAG. This in vitro test revealed that Glabrescione B significantly reduced luciferase activity in cells treated with SAG in a dose-dependent manner (FIG. 6) and suppressed the signaling. The IC.sub.50 in this assay was 0.8 M.

(38) Efficacy of Glabrescione B to directly target Gli1 protein, the final effector of Hh pathway, was also tested. HEK293 cells transiently expressing Gli1 and a Gli-dependent luciferase reporter, showed that Glabrescione B was capable of reducing Gli1-mediated transcription in a dose-dependent manner with an IC.sub.50 of 15 M (FIG. 7). GANT61, the only small-molecule described to inhibit Gli1 function (Lauth et al., PNAS 104: 8455-8460), was included as positive control.

(39) To analyze the antagonist properties of Glabrescione B under more physiological conditions, the inventors used MEF-Ptch.sup./ cells, embryo fibroblasts derived from Ptch.sup./ mouse, in which the activation of Hh signaling is consequence of Ptch1 deletion. Since Patched acts as upstream repressor of SMO, inhibition of the Hedgehog pathway from Patched is prevented and the Hedgehog pathway is constitutively activated in this cell lines, which turned out to be a reliable cellular model for studying the effect of Hedgehog inhibitors. These cells showed strongly reduced mRNA and protein levels of Gli1 (a readout of Hh signaling) when treated with Glabrescione B at different concentrations (FIGS. 8A and B). Further, authors tested the inhibitory activity of Glabrescione B on MEF-SuFu.sup./ cells, which represent another cellular model of activated Hedgehog pathway. Indeed, in this cell line aberrant Hh pathway activation is due to genetic ablation of the downstream negative regulator SuFu. Treatment of MEF-SuFu.sup./ cells with Glabrescione B led to significant reduction of the high expression levels of Gli1, as indicated by quantitative RT-PCR (FIG. 9). These results confirm that Glabrescione B is a multitarget inhibitor of Hh signaling, able to act both upstream by its direct binding to SMO receptor and downstream of SMO and SuFu by suppressing Gli1 transcription functions.

(40) A biological assay was performed in cerebellar neural progenitors cells obtained from 4-days-old mice, when they are actively proliferating under Sonic Hedgehog (Shh) stimulus. As expected, treatment of these cells with Shh increased the proliferation rate, as evaluated by BrdU-incorporation assay. Importantly, Glabrescione B antagonized this effect (FIG. 10). To further evaluate the antiproliferative effect of Glabrescione B, human MB cell lines D283 were treated with Glabrescione B at different concentrations. MTS and BrdU incorporation assay demonstrated a dose-dependent inhibitory effect on cell proliferation following 48 h of exposure to the dose range of 1-50 M (FIGS. 11A and B). The biomolecule was active, as shown by the growth rate inhibition (70-95%). Notably, the inhibitory effect of Glabrescione B on MB D283 cell proliferation was coupled to an increase of the percentage of cell death, in a dose-dependent manner (FIG. 12). More important, Glabrescione B was ineffective on unrelated signal transduction pathways (Wnt/Bcatenin and Jun/AP1activation; 0% inhibition at 30 M of Glabrescione B), demonstrating a high degree of selectivity for Hh signaling.

(41) Hh activity preferentially associates to stemness features. For this reason, the Hh pathway antagonist Glabrescione B has also been investigated for its ability to modulate the behavior of MB cancer stem cells (CSCs). Analysis by quantitative RT-PCR revealed an inhibitory effect of Glabrescione B on Hh activity, as showed by reduction of Gli1 mRNA levels (FIG. 13). Further, treatment with Glabrescione B had a suppressive effect on MB CSCs self-renewal, in particular on the percentage of neurospheres-forming cells derived from murine Ptch.sup.+/ MB (condition of Hh pathway constitutive activation). Notably, this inhibitory effect was significantly higher than the effect achieved with known Hh antagonist, as KAAD (FIG. 14).

(42) Chemical Synthesis of the Bioactive Substances

(43) Abbreviations used in the description of the chemistry and in the examples that follow are: MD: Molecular Dynamics; MM-GBSA: Molecular Mechanics Generalized Born Surface Area; Gli1-ZF: zinc finger domain of Gli1; BrdU: 5-bromo-2Ldeoxy-uridine; DMSO: dimethyl sulfoxide; MeOH: methanol; EtOH: ethanol; EtOAc: ethyl acetate; EDTA: Ethylenediaminetetraacetic acid; HBSS: Hank's Balanced Salt Solution; HPRT: Hypoxanthine-guanine phosphoribosyl transferase; cyclopamine-KAAD: 3-Keto-N (aminoethyl-aminocaproyl-dihydrocinnamoyl)-cyclopamine; M: molar; min: minutes; h: hour(s); g (grams); L (microliters); mL (milliliters); mmol (millimoles); nm (nanometers); M (micromolar); r.t.: room temperature; RT-QPCR: quantitative real time PCR; Gli-luc: Gli-dependent luciferase reporter, ESI (Electron Spray Ionization).

(44) Except where indicated otherwise, all temperatures are expressed in C. (degrees centigrade) or K (Kelvin).

(45) .sup.1H NMR and .sup.13C NMR spectra were recorded using a Bruker 400 Ultra Shield spectrometer (operating at 400 MHz for .sup.1H and 100 MHz for .sup.13C) using tetramethylsilane (TMS) as internal standard. Chemical shifts are reported in parts per million (ppm). Signals for NCH3 carbon are not present in the .sup.13C NMR data.

(46) The chemical shifts are expressed in parts per million (ppm, units). The coupling constants are expressed in Hertz (Hz) and the splitting patterns are described as s (singlet), bs (broad signal), d (doublet), t (triplet), q (quartet), quint (quintet), m (multiplet).

(47) Mass spectrometry was performed using a Thermo Finnigan LXQ linear ion trap mass spectrometer, equipped with an electrospray ionization (ESI) ion source. High-resolution mass spectra (HRMS) were obtained using a Bruker BioApex Fourier transform ion cyclotron resonance (FT-ICR) spectrometer fitted with an ESI source.

Example n. 1: Displacement Assay with Bodipy-Cyclopamine

(48) To investigate the binding properties of Glabrescione B to SMO receptor, a competition assay was carried out. In this experiment it was analyzed whether Glabrescione B could compete with Bodipy-cyclopamine (BC; 5 nM USBiological, cat. # B2527), a fluorescent derivative of cyclopamine, which is known to interact with the SMO receptor at the level of its heptahelical bundle. To this aim, HEK293 cells (ATCC, cat. # CRL-1573) were transfected with SMO expression vector (Chen et al, 2002) by using Lipofectamine 2000 (Invitrogen, cat #11668-019); 24 h after transfection the cells were fixed in paraformaldehyde 4% and incubated with 5 nM BC alone or in presence of increasing amounts of Glabrescione B for 2 h at 37 C. The cells were also treated with the steroidal alkaloid SMO antagonist cyclopamine-KAAD (3-Keto-N-(aminoethyl-aminocaproyl-dihydrocinnamoyl)-cyclopamine, 1 M Calbiochem, cat. #239804) a chemical analogous of cyclopamine, included as positive control. Cells were analyzed with a Carl Zeiss microscope (Axio Observer ZI). Cell cultures images were taken with a 20 objective for the analysis. Bodipy-cyclopamine (green) and DAPI (blue) signals were analyzed in 3-4 representative fields for coverslips. The result showed in FIG. 5 indicates that Glabrescione B abrogated BC binding to cells expressing SMO with an IC.sub.50 of 1 M by interacting with SMO receptor.

Example n. 2: Effects of Glabrescione B on Hh Signaling

(49) The ability of Glabrescione B to suppress Hh signaling was assessed in a cellular context of Hh pathway induction. To this end, the antagonist properties of Glabrescione B were determined in Shh Light II (ATCC, cat. # CRL-2795), a murine fibroblast NIH 3T3 cell line. These cells, stably incorporating the Gli-luc reporter and the pRL-TK Renilla reniformis (as described in Taipale, Chen et al. 2000) are a standard tool used to assay Hh pathway activity. The cells that reached the confluence were treated for 48 h with the SMO agonist SAG (200 nM, Alexis, cat. #ALX-270-426-M001) in absence or in presence of Glabrescione B at the amount indicated in FIG. 6. SAG is a potent cell-permeable benzothiophene compound that promotes the coupling of SMO with its downstream effector by interacting with the SMO heptahelical domain (K.sub.d=59 nM) and shown to induce Hedgehog pathway activation. Control cells were treated with only DMSO (0.5%, SERVA, cat. #39757.02) Determination of luciferase activities was carried out using the Dual-Luciferase assay according to manufacturer's instructions (Dual-Luciferase Reporter assay system; Promega, cat. # E1980). All presented data are firefly luciferase activity reported to the Renilla control activity. As shown in FIG. 6, Glabrescione B inhibits Hh signaling upon SAG activation, as evidenced by the dose-dependent decrease in luciferase reporter activity. IC.sub.50 of 0.8 M, represents the concentration of Glabrescione B necessary to have fifty percent of inhibition.

Example n. 3: Inhibition of Gli1-Mediated Transcription in HEK293 Cells

(50) To investigate the ability of Glabrescione B to block downstream Hh signaling by targeting Gli1 protein functions, the inventors performed a luciferase assay in HEK293 cells transiently expressing Gli1. The cells were seeded in 24-multiwell plates and the day after transfected with Gli1 expression vector together with the reporter plasmids 12GliBS-Luc and pRL-TK Renilla reniformis (Kogerman P et al, 1999; Everett L et al, 1999). Twenty-four hours later, Glabrescione B was added at different concentration as indicated and GANT61 (included as positive control, Enzo Life Science, cat. # ALX-270-482) was added at the final concentration of 20 M in DMSO (0.5%, SERVA, cat. #39757.02). After 24 h of treatment cells were lysed and analyzed by using the Dual Luciferase kit according to manufacturer's instructions (Dual-Luciferase Reporter assay system; Promega, cat. # E1980). All presented data are firefly luciferase activity reported to the Renilla control activity. As shown in FIG. 7, Glabrescione B was capable of interfering with Gli1-induced transcription in a dose-dependent manner and similarly to GANT61.

Example n. 4: Effects of Glabrescione B on Hh Responsive Cell Line

(51) In order to determine whether Glabrescione B regulates cells with a constitutive Hh pathway activation, the MEF-Ptch.sup./ cell line (a kind gift by M. Scott, Stanford University, CA, in which the activation of Hh signaling is consequence of Ptch1 deletion, Goodrich, L. V. et al, 1997) was treated with Glabrescione B at different concentrations and the effect on expression of Hh/Gli1 pathway was determined by quantitative RT-PCR and western blot analysis. Total RNA was isolated with TRI Reagent (Invitrogen, cat. # AM9738) and reverse transcribed with Superscript II reverse transcriptase (Invitrogen, cat. # PN100004925) and random hexamers (Invitrogen, cat. # PN58875). Quantitative PCR (QPCR) analysis of Gli1 mRNA expression was performed on each cDNA sample using the ABI Prism 7900 Sequence Detection System employing Assay-on-Demand Reagents (Life Technologies). A reaction mixture containing cDNA template, TaqMan Universal PCR master mix (ABI) and primer probe mixture was amplified using standard QPCR thermal cycler parameters. Each amplification reaction was performed in triplicate and the average of the three threshold cycles was used to calculate the amount of transcript in the sample (using SDS version 2.3 software). All values were normalized with HPRT (Hypoxanthine-guanine phosphoribosyl transferase) endogenous controls (Life Technologies, cat. # Mm 00494645_m1 Gli1; Mm 01545399_m1). For western blot analysis, cells were lysed in Tris-HCl (pH 7.6) 50 mM, NaCl 150 mM, NP-40 1%, EDTA 5 mM, deoxycholic acid sodium salt 0.5%, NaF 100 mM and protease inhibitors. The lysates were centrifuged at 13,000 g for 20 minutes, separated by a 8% SDS-PAGE and transferred to a nitrocellulose membrane. Immunoblotting, was performed with a mouse monoclonal antibody against Gli1 (Cell Signaling, cat. # L42B10) or goat polyclonal antibody against Actin (Santa Cruz Biotechnology, cat. #11012), used as loading control. HRP-conjugated secondary antibody anti-mouse or anti-goat IgG (Santa Cruz Biotechnology, cat. # SC-2005, # SC-2020) were used, and immunoreactive bands were visualized by enhanced chemiluminescence (Perkin Elmer, cat. # NEL105001EA). KAAD (1 M, Calbiochem, cat. #239804) was used as control. As shown in FIG. 8, Glabrescione B down-regulates Gli1 expression compared to the untreated control.

(52) To further elucidate the concept of downstream pathway inhibition by Glabrescione B, the inventors used MEF-SuFu.sup./ cells (a kind gift by R. Toftgard, Karolinska Institutet, Sweden, Svrd, J. et al, 2006), in which the downstream Hh pathway activation is consequence of SuFu genetic ablation. Confluent cells were treated with Glabrescione B at the amount indicated in the FIG. 9 and Gli1 expression levels were determined by quantitative RT-PCR as previously described. As shown in FIG. 9, Glabrescione B promoted a significant downregulation of mRNA Gli1 levels. Data were normalized with HPRT endogenous controls, as described above. These findings confirm that Glabrescione B is indeed inhibitor of Hh signaling downstream of SMO and SuFu.

(53) Another experiment to assess whether Glabrescione B also regulates Hh pathway in a physiological context was carried out in cerebellar granule cells precursors (GCPs) isolated from 4-days-old mice according to established protocols (Wechsler-Reya and Scott, 1999). Briefly, cerebella were removed aseptically, cut into small pieces, and incubated at room temperature for 15 min in digestion buffer [Dulbecco's PBS (Invitrogen, Gaithersburg, Md.) with 0.1% trypsin, 0.2% EDTA, and 100 g/ml DNase]. Tissues were then triturated with fire-polished Pasteur pipettes to obtain a single-cell suspension. Cells were centrifuged, resuspended in Neurobasal medium supplemented with B27, penicillin-streptomycin, and L-glutamine (2 mM) (Invitrogen) and plated at a density of 810.sup.5 cells/cm.sup.2 on tissue-culture dishes or eight-well Lab-Tek chamber slides (Permanox slide; Nunc, cat. #177445) coated with 1 mg/ml poly-L-lysine. These cells were treated with Shh alone (Recombinant mouse Sonic Hedgehog, amino-terminal peptide; 3 g/ml, R&D system, cat. #461-SH) or in combination with different concentration of Glabrescione B for 48 h and growth inhibition IC.sub.50 doses were determined by BrdU (bromodeoxyuridine, 5-bromo-2Ldeoxy-uridine, Roche, cat. #11 296 736 001) labeling assay according to standard methods (FIG. 10). After BrdU incorporation, GCPs were fixed in 4% paraformaldehyde for 20 min at room temperature, incubated in 0.2% Triton X-100 to permeabilize cells and treated with 2N HCl to denature DNA. BrdU detection was performed according to the manufacturer's instructions.

(54) In order to determine whether Glabrescione B also inhibits the proliferation of cancer cells, human MB D283 cells (ATCC, cat. #HTB-185) were treated with Glabrescione B and proliferation/viability rate were measured by BrdU incorporation and MTS assays, respectively. Cells, as shown in FIG. 11, displayed a dose-dependent inhibitory effect on cell proliferation (FIG. 11A) and viability (FIG. 11B) following 48 h of exposure to the dose range of 1-50 M. More important, Glabrescione B was able to promote cell death of D283 MB cells, as reported by Trypan Blue count assay (FIG. 12). Cell proliferation was evaluated by a BrdU-labelling assay (8 h pulse; Roche, cat. #11 296 736 001) as described above; cell viability was evaluated by MTS (a tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt; MTS; Promega, cat. #G3580). Cell death was measured by Trypan Blue (Sigma, cat. #T8154) exclusion counting at least 200 cells, in duplicate samples. All described assays were evaluated at 48 h after treatment.

Example n. 5: Effects of Glabrescione B on MB Cancer Stem Cells and Cancer-Derived Stem-Like Cells

(55) It has been described that Hh activity preferentially associates to stemness features. In order to determine whether Glabrescione B also modulates the behavior of MB cancer stem cells (CSCs), neurospheres isolated from murine Ptch.sup.+/ MB were treated with Glabrescione B for 48 h and the levels of Gli1 were determined by quantitative RT-PCR. Murine MBs were isolated from Ptch1.sup.+/ mice (Goodrich, Milenkivic et al. 1997). Tissues were collected in HBSS supplemented with 0.5% glucose and penicillin-streptomycin, grossly triturated with serological pipette and treated with DNAse I to a final concentration of 0.04% for 20 min. Finally, cells aggregates were mechanically dissociated using pipettes of decreasing bore size to obtain a single cell suspension. CSCs were cultured as neurospheres in selective medium after centrifugation, DMEM/F12 supplemented with 0.6% glucose, 25 mg/ml insulin, 60 mg/ml N-acetyl-L-cystein, 2 mg/ml heparin, 20 ng/ml EGF, 20 ng/ml bFGF, 1 penicillin-streptomycin and B27 supplement without vitamin A. Neurospheres were treated with the steroidal alkaloid SMO antagonist, cyclopamine-KAAD (1 M Calbiochem, cat. #239804). RNA extraction and mRNA expression analysis were performed as described above. As shown in FIG. 13, Gli1 expression was down-regulated compared to the untreated control. All values were normalized with HPRT internal controls.

(56) Another experiment was carried out to determine the ability of Glabrescione B to suppress MB CSCs self-renewal. To this aim neurospheres derived from murine Ptch.sup.+/ MB were treated with Glabrescione B at different concentrations and the percentage of neurospheres-forming cells was measured. For the neurosphere forming assay, cells were plated at clonal density (1-2 cells/mm2) into 96-well plates and cultured in selective medium as described above.

(57) The results shown in FIG. 14 indicate that Glabrescione B has an inhibitory effect on self-renewal of cancer stem cells. Notably, this inhibitory effect was significantly higher than the effect achieved with known Hh antagonist KAAD, thus reinforcing that Glabrescione B is a potent inhibitor of tumor growth via inhibition of the Hh signaling pathway.

(58) Similar results were observed in ASZ001 basal cell carcinoma cells (BCC), previously characterized as an Hh/Gli-dependent tumor cell line harbouring Ptch1 deletion (Aszterbaum et al, 1999). Total RNA was isolated with Trizol (Invitrogen, Eugene, Or, USA) and reverse transcribed with Superscript II reverse transcriptase and random hexamers (Invitrogen, Eugene, Or, USA). Quantitative real-time PCR (qRT-PCR) analysis of Gli1, Gli2, Ptch1, Ptch2, Nanog, Oct-4, Cyclin D1, Cyclin D2, Nmyc, HIP1, Sfrp1, Bmp2, IGF2, -2 microglobulin, HPRT, mRNA expression was performed on each cDNA sample using the ABI Prism 7900HT Sequence Detection System employing Assay-on-Demand Reagents (Applied Biosystems, Foster City, Calif., USA). A reaction mixture containing cDNA template, TaqMan Universal PCR master mix (Applied Biosystems, Foster City, Calif., USA) and primer probe mixture was amplified using standard qRT-PCR thermal cycler parameters. Each amplification reaction was performed in triplicate and the average of the three threshold cycles was used to calculate the amount of transcript in the sample (using SDS version 2.3 software). mRNA quantification was expressed, in arbitrary units, as the ratio of the sample quantity to the quantity of the calibrator. All values were normalized with two endogenous controls, -2 microglobulin and HPRT, which yielded similar results.

(59) BCC cell proliferation was impaired by in vitro treatment with Glabrescione B together with a suppression of Gli1 mRNA before a drug-induced cell death occurred (FIG. 15). Notably, no effect was observed in Hh-independent HepG2 hepatocellular carcinoma cells or Jurkat T leukemia cells, which display undetectable levels of Gli1 (Lauth et al, 2007).

(60) In summary, the inhibitory activity of Glabrescione B on the subset of both normal and tumor progenitor/stem cells as well as the whole tumor cell populations, is restricted to Hh/Gli-dependent cells.

Example n. 6: Glabrescione B Inhibits Gli1-Dependent Tumor Growth In Vivo

(61) To assess the in vivo activity of Glabrescione B, we first tested its ability to suppress Hh signaling in 6-day old mouse cerebellar progenitors, considered the cell of origin of MB. 6-day old CD1 mice were randomly divided into two groups (n=6) and injected s.c. with solvent only (2-hydroxypropyl--cyclodextrin:ethanol, 3:1) or Glabrescione B in solvent (100 mol/Kg) for 2 days (2-hydroxypropyl--cyclodextrin was purchased from Sigma Aldrich, St. Louis, Mo., USA). Cerebella were collected and mRNA levels were determined by qRT-PCR. Glabrescione B treatment reduced significantly the cerebellar levels of Hh target genes (FIG. 16).

(62) Therefore, to verify the Glabrescione B efficacy to inhibit Hh-dependent tumor cell growth in vivo, we turned to an allograft model of MB cells. Spontaneous Medulloblastomas from Ptch1.sup.+/ mice were isolated, minced, pipetted to obtain a single-cell suspension and grafted s.c. at the posterior flank of female BALB/c nude mice (nu/nu) (Charles River Laboratories, Lecco, Italy). Tumors were grown until a median size of 100 mm.sup.3. Animals were randomly divided into two groups (n=6) and treated with solvent only (2-hydroxypropyl--cyclodextrin:ethanol, 3:1) or Glabrescione B in solvent (75 mol/Kg) for 18 days (2-hydroxypropyl--cyclodextrin was purchased from Sigma Aldrich, St. Louis, Mo., USA). 210.sup.6 ASZ001 BCC cells were resuspended in an equal volume of 154CF medium and Matrigel (BD Biosciences, Heidelberg, Germany) and injected s.c. at the posterior flank of female NOD/SCID mice (Charles River Laboratories, Lecco, Italy), as previously described (Eberl et al, 2012). Tumors were grown until a median size of 200 mm.sup.3. Animals were randomly divided into two groups (n=6) and treated with solvent only (2-hydroxypropyl--cyclodextrin:ethanol, 3:1) or Glabrescione B in solvent (100 mol/Kg) for 18 days. Tumor volumes change was calculated by the formula lengthwidth0.5(length+width) (Lauth et al, 2007).

(63) Nude mice were grafted with spontaneous primary MB from Ptch1.sup.+/ mice and treated every second day with s.c. injections of Glabrescione B at a concentration of 75 mol/Kg or solvent only (n=6 for each group). During an 18-day treatment period Glabrescione B significantly suppressed tumor mass compared with controls, as confirmed by in vivo decreased Ki67 staining in Glabrescione B-treated tumors together with a reduction of Gli1 mRNA levels (FIG. 17). Notably, in vivo Glabrescione B-induced tumor growth inhibition was also observed in BCC s.c. allografts. A significant reduction of tumor growth, as well as Gli1 mRNA levels were observed 18 days after administration of Glabrescione B (100 mol/Kg) compared to solvent alone (FIG. 18).

Example n. 7: Chemical Synthesis of Glabrescione B

(64) The total synthesis of the isoflavone Glabrescione B is a six-step synthetic route, with an overall yield of 7%. It also permits the preparation of numerous derivatives of Glabrescione B. The use of Pd EnCat40, as catalyst, introduces an aspect of green chemistry into the synthesis.

(65) The yields were calculated assuming that products were 100% pure if not stated otherwise.

(66) ##STR00006##

Intermediate 2: 2-hydroxy-4,6-dimethoxyacetophenone (2)

(67) ##STR00007##

(68) A flame-dried flask was charged with 2,4,6-trimethoxyacetophenone (23.8 mmol, 5 g) (SIGMA-ALDRICH 630594) and dry CH.sub.2Cl.sub.2 (75 ml) under argon. BBr.sub.3 (87.3 mmol, 15 ml) was added drop-wise. The resulting solution was stirred at room temperature for 2 h before adding NaOH 4M (90 ml) and allowed to stand for 30 min. The solution was extracted with CH.sub.2Cl.sub.2 and combined organic layers were washed with brine, dried over Na.sub.2SO.sub.4 and finally concentrated under reduced pressure. The resulting solid was recrystallized from EtOH and white crystals of 2-hydroxy-4,6-dimethoxyacetophenone 2 (4.43 g, 98% yield) were obtained.

(69) ##STR00008##

2-hydroxy-4,6-dimethoxyacetophenone (2)

(70) White solid, (4.43 g) 98% yield. Mp: 81-82 C. .sup.1H NMR (400 MHz, Acetone-d6): (ppm) 13.8 (s, 1H, OH), 5.94 (d, 1H, J=2.4 Hz, ArH), 5.91 (d, 1H, J=2.4 Hz, ArH), 3.80 (s, 3H, OCH.sub.3), 3.72 (s, 3H, OCH.sub.3), 2.43 (s, 3H, CH.sub.3); .sup.13C NMR (400 MHz, Acetone-d6): (ppm) 202.5 (s, CO), 167.36 (s, C-4), 166.34 (s, C-6), 163.15 (s, C-2), 105.5 (s, C-1), 93.42 (d, C-3), 90.43 (d, C-5), 55.20, 55.00 (q, 2OCH.sub.3), 31.94 (q, CH.sub.3). ESI-MS (positive): m/z calcd for C.sub.10H.sub.12O.sub.4+H.sup.+: 197.080800 [M+H].sup.+(monoisotopic mass). found: 197.080848.

Intermediate 4: 3-iodo-5,7-dimethoxy-4H-chromen-4-one (4)

(71) ##STR00009##

(72) A mixture of 2-hydroxy-4,6-dimethoxyacetophenone 2 (23.1 mmol, 4.5 g) and N,N-dimethylformamide dimethylacetal (97.1 mmol, 13 ml) was stirred at 95 C. for 3 h, then concentrated in vacuo to give enamino ketone 3 (5.8 g) in quantitative yield. (Biegasiewicz, St Denis et al. 2010) Compound 3 was dissolved in MeOH (450 ml) and I.sub.2 (46.2 mmol, 11.7 g) was added to the solution. The mixture was stirred at room temperature for 1 h, then the solvent was evaporated. To remove residual I.sub.2, the crude was treated with a saturated aqueous Na.sub.2S.sub.2O.sub.3 solution until the mixture became clear. The mixture was then extracted with CHCl.sub.3, the combined organic layers were dried over Na.sub.2SO.sub.4 and concentrated under reduced pressure. The residue was purified by column chromatography using hexane-EtOAc as eluent to obtain 3-iodo-5,7-dimethoxy-4H-chromen-4-one 4 (2.5 g, 33% yield) as a white powder.

(73) ##STR00010##

3-dimethylamino-1-(2-hydroxy-4,6-dimethoxyphenyl)propenone (3)

(74) Red solid, (5.8 g) quantitative yield. Mp: 145-147 C. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 15.65 (s, 1H, OH), 7.92 (d, 1H, J=12 Hz, CHN), 6.25 (d, 1H, J=12.0 Hz, CH(CO)), 6.07 (d, 1H, J=2.4 Hz, ArH), 5.91 (d, 1H, J=2.4 Hz, ArH), 3.84 (s, 3H, OCH.sub.3), 3.80 (s, 3H, OCH.sub.3), 3.15 (s, 3H, NCH.sub.3), 2.92 (s, 3H, NCH.sub.3); .sup.13C NMR (400 MHz, CDCl.sub.3): (ppm) 191.64 (s, CO), 169.00 (s, C-4), 165.54 (s, C-6), 162.86 (s, C-2), 155.83 (d, CHN), 106.58 (s, C-1), 96.27 (d, CH(CO)), 95.59 (d, C-3), 92.06 (d, C-5), 57.09, 56.57 (q, 2OCH.sub.3). ESI-MS (positive): m/z calcd for C.sub.13H.sub.17O.sub.4+H.sup.+: 252.1 [M+H].sup.+(monoisotopic mass). found: 252.1.

(75) ##STR00011##

3-iodo-5,7-dimethoxy-4H-chromen-4-one (4)

(76) White solid, (2.5 g) 33% yield. Mp: 156-157 C. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 8.02 (s, 1H, H-2), 6.37 (d, J=2.0 Hz, 1H, ArH), 6.32 (d, J=2.0 Hz, 1H, ArH), 3.87 (s, 3H, OCH.sub.3), 3.82 (s, 3H, OCH.sub.3); .sup.13C NMR (400 MHz, CDCl.sub.3): (ppm) 183.43 (s, CO), 164.28 (s, C-7), 160.97 (s, C-5), 159.81 (s, C-9), 155.33 (d, C-2), 107.53 (s, C-10), 96.59 (d, C-8), 92.44 (d, C-6), 89.71 (d, CHI), 56.41, 55.79 (q, 2OCH.sub.3). ESI-MS (positive): m/z calcd for C.sub.11H.sub.9O.sub.4I+H.sup.+: 332.961800 [M+H].sup.+(monoisotopic mass). found: 332.961633.

Intermediate 5: 3-(3,4-methylendioxyphenyl)-5,7-dimethoxy-4H-chromen-4-one (5)

(77) ##STR00012##

(78) To a solution of 4 (7.5 mmol, 2.5 g) in 1,2-dimethoxyethane/H.sub.2O=50:50 (150 ml) were added Na.sub.2CO.sub.3 (30 mmol, 3.18 g), 3,4-(methylenedioxy)-phenylboronic acid (11 mmol, 1.8 g), and Pd EnCat40 (937 mg, 5%). The resulting mixture was stirred at 45 C. for 2 h and then filtered. The catalyst was washed with H.sub.2O and CH.sub.2Cl.sub.2. The aqueous phase was extracted with CH.sub.2Cl.sub.2. The combined organic layers were dried over Na.sub.2SO.sub.4 and concentrated under reduced pressure. The crude residue was purified by flash chromatography to give 3-(3,4-methylendioxyphenyl)-5,7-dimethoxy-4H-chromen-4-one 5 (1.4 g, 57% yield) as gray powder.

(79) ##STR00013##

3-(3,4-methylendioxyphenyl)-5,7-dimethoxy-4H-chromen-4-one (5)

(80) Gray solid, (1.4 g) 57% yield. Mp: 155-156 C. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.75 (s, 1H, H-2), 7.1 (d, J=2.0 Hz, 1H, H-2), 6.94 (dd, J=8.0 Hz and 2.0 Hz, 1H, H-6), 6.83 (d, J=8.0 Hz, 1H, H-5), 6.44 (d, J=2.2 Hz, 1H, H-6), 6.37 (d, J=2.2 Hz, 1H, H-8), 5.97 (broad s, 2H, OCH.sub.2O), 3.94 (s, 3H, OCH.sub.3), 3.89 (s, 3H, OCH.sub.3); .sup.13C NMR (400 MHz, CDCl.sub.3): (ppm)=175.09 (s, CO), 164.02 (s, C-7), 161.54 (s, C-5), 160.30 (s, C-9), 150.05 (d, C-2), 147.58 (s, C-3), 147.58 (s, C-4), 126.5 (s, C-3), 126.04 (s, C-1), 122.84 (d, C-6), 110.50 (d, C-2), 110.0 (s, C-10), 108.35 (d, C-5), 101.30 (t, OCH.sub.2O), 96.74 (d, C-6), 92.73 (d, C-8), 56.41, 55.79 (q, 2OCH.sub.3). ESI-MS (positive): m/z calcd for C.sub.18H.sub.14O.sub.6+H.sup.+: 327.086300 [M+H].sup.+(monoisotopic mass). found: 327.086206.

Intermediate 6: 3-(3,4-dihydroxyphenyl)-5,7-dimethoxy-4H-chromen-4-one (6)

(81) ##STR00014##

(82) A mixture of 5 (4.3 mmol, 1.4 g) and Pb(OAc).sub.4 (17 mmol, 7.5 g, freshly recrystallized from AcOH) in dry C.sub.6H.sub.6 (100 ml) was stirred at 80 C. under argon overnight. After, being cooled to room temperature, the reaction mixture was filtered through a pad of Celite, washed with CH.sub.2Cl.sub.2 and concentrated under reduced pressure. The crude was diluted with THF/H.sub.2O=5:1 (50 ml) and CH.sub.3COOH (50 ml) and the resulting mixture was stirred at room temperature for 6 h. (Ye, Koshino et al. 2009) Afterwards, a saturated aqueous NaHCO.sub.3 solution was added until pH 8 and extracted with EtOAc. To the combined organic layers was added a solution of NaOH 0.1 M. Water layer was treated with CH.sub.3COOH and then extracted with EtOAc. The combined organic layers were dried over Na.sub.2SO.sub.4 and concentrated under reduced pressure to obtain 3-(3,4-dihydroxyphenyl)-5,7-dimethoxy-4H-chromen-4-one 6 (570 mg, 42% yield).

(83) ##STR00015##

3-(3,4-dihydroxyphenyl)-5,7-dimethoxy-4H-chromen-4-one (6)

(84) Yellow solid, (570 mg) 42% yield. Mp: 127-129 C. .sup.1H NMR (400 MHz, MeOD): (ppm) 7.86 (s, 1H, H-2), 6.88 (broad s, 1H, H-2), 6.70 (broad s, 2H, H-5 and H-6), 6.50 (d, J=2.0 Hz, 1H, H-6), 6.40 (d, J=2.0 Hz, 1H, H-8), 3.80 (s, 3H, OCH.sub.3), 3.79 (s, 3H, OCH.sub.3); .sup.13C NMR (400 MHz, MeOD): (ppm) 175.00 (s, CO), 165.08 (s, C-7), 161.13 (s, C-5), 159.91 (s, C-9), 151.28 (d, C-2), 145.12 (s, C-3), 145.0 (s, C-4), 126.49 (s, C-3), 123.12 (s, C-1), 120.50 (d, C-6), 116.99 (d, C-2), 110.00 (s, C-10), 114.48 (d, C-5), 95.86 (d, C-6), 92.82 (d, C-8), 55.09 (q, 2OCH.sub.3). ESI-MS (positive): m/z calcd for C.sub.17H.sub.14O.sub.6+H.sup.+: 315.086300 [M+H].sup.+(monoisotopic mass). found: 315.086363.

Compound 1: Glabrescione B: 3-(3,4-bis(3-methylbut-2-enyloxy)phenyl)-5,7-dimethoxy-4H-chromen-4-one (1)

(85) ##STR00016##

(86) To a solution of 6 (61.8 mmol, 570 mg) in acetone (100 ml) was added K.sub.2CO.sub.3 (5.4 mmol, 7.4 g) and, after 10 minutes of stirring at room temperature, was added 3,3-dimethylallyl bromide (6.5 mmol, 968 mg). Then, the mixture was stirred at 80 C. overnight. Afterwards, the solvent was evaporated. The resulting solid was dissolved in EtOAc and extracted with water. The combined organic layers were dried over Na.sub.2SO.sub.4 and finally concentrated under reduced pressure. The crude was purified by column chromatography using hexane-EtOAc as eluent to obtain compound 1 as white powder. The powder was recrystallized from hexane resulting with white crystals.

(87) ##STR00017##

3-(3,4-bis(3-methylbut-2-enyloxy)phenyl)-5,7-dimethoxy-4H-chromen-4-one (1)

(88) White solid, (696 mg) 86% yield. Mp 102-104 C. .sup.1H NMR (400 MHz, Acetone-.sub.6): (ppm) 7.94 (s, 1H, H-2), 7.16 (d, 1H, J=1.6 Hz, H-2), 6.99 (dd, 1H, J=8.0 Hz and 1.6 Hz, H-6), 6.89 (d, 1H, J=8.0 Hz H-5), 6.50 (d, 1H, J=2.0 Hz, H-8), 6.42 (d, 1H, J=2.0 Hz, H-6), 5.43 (m, 2H, 2=CH), 4.51 (d, J=6.8 Hz, 4H, 2OCH.sub.2), 3.86 (s, 3H, OCH.sub.3), 3.81 (s, 3H, OCH.sub.3), 1.70 (s, 6H, 2CH.sub.3), 1.67 (s, 6H, 2CH.sub.3); .sup.13C NMR (400 MHz, CDCl.sub.3): (ppm) 175.59 (s, CO), 164.02 (s, C-7), 161.23 (s, C-5), 159.67 (s, C-9), 150.20 (d, C-2), 148.89 (s, C-3), 148.51 (s, C-4), 136.85 (s, 2C=), 126.50 (s, C-3), 125.00 (s, C-1), 121.44 (d, C-6), 120.65 (d, 2=CH), 115.50 (d, C-2), 110.00 (s, C-10), 114.00 (d, C-5), 96.47 (d, C-6), 92.60 (d, C-8), 65.95 (t, 2OCH.sub.2), 56.44 (q, OCH.sub.3), 55.93 (q, OCH.sub.3), 25.62 (q, 2CH.sub.3), 18.13 (q, 2CH.sub.3). ESI-MS (positive): m/z calcd for C.sub.27H.sub.30O.sub.6+H.sup.+: 451.211500 [M+H].sup.+(monoisotopic mass). found: 451.211495.

(89) The .sup.1H-NMR data of this product were identical to an authentic sample of Glabrescione B.

Example n. 8: Chemical Synthesis of Compounds NT8, NT9, NT10 and NT11

(90) ##STR00018##

(91) A mixture of 3,5-dimethoxyphenol (1.3 mmol, 463 mg) (SIGMA-ALDRICH 132632), 3,4-dihydroxyphenylacetic acid (2, 3 mmol, 504.5 mg) (SIGMA-ALDRICH 850217) and BF.sub.3.Et.sub.2O (15.3 mmol, 1.94 ml) was stirred at 90 C. for 90 min under Argon. The reaction mixture was poured into 10% aqueous NaOAc solution (100 ml) and allowed to stand 4 h. The solution was extracted with EtOAc. The combined organic layers were washed with saturated solution of NaHCO.sub.3, dried over Na.sub.2SO.sub.4 and finally concentrated under reduced pressure. The residue was purified by column chromatography using hexane-EtOAc mixture as eluent to obtain 1-(2-hydroxy-4,4-dimethoxy-phenyl)-2-(3-hydroxy-4-methoxy-phenyl)-ethanone 7. A mixture of 7 (3 mmol) and BF.sub.3.Et.sub.2O (9 mmol, 1.2 mL) was cooled to 10 C. and DMF (4.6 ml) was added drop wise. In another flask, DMF (8 mL) was cooled to 10 C. and PCl.sub.5 (4.5 mmol) was added. The mixture was then allowed to stand to 55 C. for 20 min. The light yellow colored solution containing N,N-dimethyl(chloromethylene)ammonium chloride was then added to the above reaction mixture at 20-25 C. The mixture was stirred at r.t. for 2 h then poured into methanolic HCl (0.1N) and allowed to stand at 700 for 2 h. After removing the solvent the solution was extracted with EtOAc (3100 mL) and the combined organic layer was washed with brine, dried over Na.sub.2SO.sub.4, and finally concentrated under reduced pressure. The residue was purified by column chromatography using hexane-EtOAc as eluent to give 3-(3-hydroxy-4-methoxy-phenyl)-5,7-dimethoxy-chromen-4-one 8.

(92) To a solution of 8 (0.18 mmol, 60 mg) in acetone (5 ml) at 45 C. was added solid K.sub.2CO.sub.3 (8 eq). R.sub.5Br (1.5 eq) was added drop wise to the mixture and stirred at 45 C. for 3 h. The progress of the reaction was monitored by thin layer chromatography (SIO.sub.2 gel, developing solvent V exane:V EtOAc=3:7). When only one spot was on the TLC the reaction was quenched, extracted with EtOAc, the combined organic layer was dried over Na.sub.2SO.sub.4 and evaporated under reduced pressure.

3-[3-(3-methyl-but-2-enyloxy)-4-methoxyphenyl]-5,7-dimethoxy-4H-chromen-4-one (NT8)

(93) ##STR00019##

(94) Yellow oil, 47% yield. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.76 (s, 1H, H-2), 7.22 (d, 1H, J=1.6 Hz, H-2), 7.01 (dd, 1H, J=8 Hz and J=1.6 Hz, H-6), 6.88 (d, 1H, J=8.4 Hz, H-5), 6.44 (d, 1H, J=2 Hz, H-8), 6.37 (d, 1H, J=2 Hz, H-6), 5.55 (t, 1H, J=6.4 Hz, CH), 4.59 (d, 2H, J=6.8 Hz, CH.sub.2), 3.94 (s, 3H, OCH.sub.3), 3.88 (s, 6H, 2OCH.sub.3), 1.76 (s, 3H, CH.sub.3), 1.71 (s, 3H, CH.sub.3).

3-(3-Benzyloxy-4-methoxyphenyl)-5,7-dimethoxy-4H-chromen-4-one (NT9)

(95) ##STR00020##

(96) White powder, 44% yield. MP: 126 C. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.71 (s, 1H, H-2), 7.46 (dd, 2H, J=7.6 and J=1.6 Hz, ArH), 7.36 (t, 2H, J=7.6 Hz, ArH) 7.31 (dd, 1H, J=7.2 and J=2 Hz, ArH), 7.29 (d, 1H, J=2 Hz, H-2) 7.07 (dd, 1H, J=8.4 and J=2 Hz, H-6), 6.92 (d, 1H, J=8.4 Hz, H-5), 6.44 (d, 1H, J=2 Hz, H-8), 6.37 (d, 1H, J=2 Hz, H-6), 5.17 (s, 2H, OCH.sub.2), 3.95 (s, 3H, OCH.sub.3), 3.90 (s, 3H, OCH.sub.3), 3.89 (s, 3H, OCH.sub.3).

3-[3-(3,7-Dimethyl-octa-2,6-dienyloxy)-4-methoxy-phenyl]-5,7-dimethoxy-chromen-4-one (NT10)

(97) ##STR00021##

(98) White powder, 44% yield .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.76 (s, 1H, H-2), 7.22 (d, 1H, J=1.6 Hz, H-2), 7.01 (dd, 1H, J=8 Hz and J=1.6 Hz, H-6), 6.88 (d, 1H, J=8. Hz, H-5), 6.44 (d, 1H, J=2 Hz, H-8), 6.37 (d, 1H, J=2 Hz, H-6), 5.55 (t, 1H, J=6.4 Hz, CH), 5.39 (t, 1H, J=7 Hz, CH), 4.59 (d, 2H, J=6.4 Hz, OCH.sub.2), 3.94 (s, 3H, OCH.sub.3), 3.88 (s, 6H, 2OCH.sub.3), 2.14 (t, 2H, J=6.9 Hz CH.sub.2), 2.00 (q, 2H, J=7 Hz, CH.sub.2CH.sub.2CH), 1.76 (s, 3H, CH.sub.3), 1.71 (s, 6H, 2CH.sub.3).

5,7-Dimethoxy-3-[4-methoxy-3-(4-trifluoromethyl-benzyloxy)-phenyl]-chromen-4-one (NT11)

(99) ##STR00022##

(100) White powder, 78% yield. MP: 128.6-129.1 C. .sup.1H NMR (400 MHz, acetone-d6): (ppm) 8.01 (s, 1H, H-2), 7.78 (m, 4H, ArH), 7.36 (d, 1H, J=2.1 Hz, H-2), 7.15 (dd, 1H, J=8.3 Hz and J=2.1 Hz, H-6), 7.02 (d, 1H, J=8.4, H-5), 6.57 (d, 1H, J=2.3 Hz, H-8), 6.49 (d, 1H, J=2.3 Hz, H-6), 5.27 (s, 2H, OCH.sub.2), 3.93 (s, 3H, OCH.sub.3), 3.89 (s, 3H, OCH.sub.3), 3.88 (s, 3H, OCH.sub.3).

Example n. 9: Chemical Synthesis of Compounds NT12, NT13, NT14 and NT15

(101) ##STR00023##

(102) A mixture of 3,5-dimethoxyphenol (1.3 mmol, 463 mg), 3,4-dihydroxyphenylacetic acid (2, 3 mmol, 504.5 mg) and BF.sub.3.Et.sub.2O (15.3 mmol, 1.94 ml) was stirred at 90 C. for 90 min under Argon. The reaction mixture was poured into 10% aqueous NaOAc solution (100 ml) and allowed to stand 4 h. The solution was extracted with EtOAc. The combined organic layers were washed with saturated solution of NaHCO.sub.3, dried over Na.sub.2SO.sub.4 and finally concentrated under reduced pressure. The residue was purified by column chromatography using hexane-EtOAc mixture as eluent to obtain 9.

(103) A mixture of 9 (3 mmol) and BF.sub.3.Et.sub.2O (9 mmol, 1.2 mL) was cooled to 10 C. and DMF (4.6 ml) was added drop wise. In another flask, DMF (8 mL) was cooled to 10 C. and PCl.sub.5 (4.5 mmol) was added. The mixture was then allowed to stand to 55 C. for 20 min. The light yellow colored solution containing N,N-dimethyl(chloromethylene)ammonium chloride was then added to the above reaction mixture at 20-25 C. The mixture was stirred at r.t. for 2 h then poured into methanolic HCl (0.1N) and allowed to stand at 700 for 2 h. After removing the solvent the solution was extracted with EtOAc (3100 mL) and the combined organic layer was washed with brine, dried over Na.sub.2SO.sub.4, and finally concentrated under reduced pressure. The residue was purified by column chromatography using hexane-EtOAc as eluent to give 10.

(104) To a solution of 10 (0.18 mmol, 60 mg) in acetone (5 ml) at 45 C. was added solid K.sub.2CO.sub.3 (8 eq). R.sub.5Br (1.5 eq) was added drop wise to the mixture and stirred at 45 C. for 3 h. The progress of the reaction was monitored by thin layer chromatography (SIO.sub.2 gel, developing solvent V exane:V EtOAc=3:7). When only one spot was on the TLC the reaction was quenched, extracted with EtOAc, the combined organic layer was dried over Na.sub.2SO.sub.4 and evaporated under reduced pressure.

5,7-Dimethoxy-3-[3-methoxy-4-(3-methyl-but-2-enyloxy)-phenyl]-chromen-4-one (NT12)

(105) ##STR00024##

(106) Yellow powder, 45% yield. MP: 130.7-131.8 C. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.78 (s, 1H, H-2), 7.23 (d, 1H, J=2.0 Hz, H-2), 6.96 (dd, 1H, J=8.0 and J=2.0 Hz, H-6), 6.88 (d, 1H, J=8.0 Hz, H-5), 6.44 (d, 1H, J=2.4 Hz, H-8), 6.37 (d, 1H, J=2.4 Hz, H-6), 5.53 (m, 1H, CH), 4.60 (d, 2H, J=6.8 Hz, OCH.sub.2), 3.94 (s, 3H, OCH.sub.3), 3.89 (s, 6H, 2OCH.sub.3), 1.77 (s, 3H, CH.sub.3), 1.73 (s, 3H, CH.sub.3).

3-(4-Benzyloxy-3-methoxy-phenyl)-5,7-dimethoxy-chromen-4-one (6b) (NT13)

(107) ##STR00025##

(108) White powder, 56% yield. MP: 142.8-143.7 C. .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.77 (s, 1H, H-2), 7.45 (dd, 2H, J=8.0 Hz and J=2.0 Hz, ArH), 7.37 (t, 2H, J=8.0 Hz, ArH), 7.31 (dd, 1H, J=8 Hz and J=1.6 Hz, ArH), 7.30 (d, 1H, J=2.0 Hz, H-2), 6.89 (m, 2H, H-6 and H-5), 6.45 (d, 1H, J=2 Hz, H-8), 6.38 (d, 1H, J=2 Hz, H-6), 5.19 (s, 1H, OCH.sub.2), 3.94 (s, 3H, OCH.sub.3), 3.93 (s, 3H, OCH.sub.3), 3.90 (s, 3H, OCH.sub.3)

3-[4-(3,7-Dimethyl-octa-2,6-dienyloxy)-3-methoxy-phenyl]-5,7-dimethoxy-chromen-4-one (NT14)

(109) ##STR00026##

(110) White powder, 44% yield .sup.1H NMR (400 MHz, CDCl.sub.3): (ppm) 7.76 (s, 1H, H-2), 7.22 (d, 1H, J=1.6 Hz, H-2), 7.01 (dd, 1H, J=8 Hz and J=1.6 Hz, H-6), 6.88 (d, 1H, J=8 Hz, H-5), 6.44 (d, 1H, J=2 Hz, H-8), 6.37 (d, 1H, J=2 Hz, H-6), 5.42 (t, 1H, J=6.4 Hz, CH), 5.12 (t, 1H, J=7 Hz, CH), 4.41 (d, 2H, J=6.4 Hz, OCH.sub.2), 3.94 (s, 3H, OCH.sub.3), 3.90 (s, 3H, OCH.sub.3), 3.88 (s, 3H, OCH.sub.3), 2.14 (t, 2H, J=6.9 Hz CH.sub.2), 2.00 (q, 2H, J=7 Hz, CH.sub.2CH.sub.2CH), 1.76 (s, 3H, CH.sub.3), 1.71 (s, 6H, 2CH.sub.3).

5,7-Dimethoxy-3-[3-methoxy-4-(4-trifluoromethyl-benzyloxy)-phenyl]-chromen-4-one (NT15)

(111) ##STR00027##

(112) White powder, 25% yield. MP: 128.4-130.2 C. .sup.1H NMR (400 MHz, acetone-d6): (ppm) 8.04 (s, 1H, H-2), 7.76 (m, 4H, ArH), 7.29 (d, 1H, J=2.1 Hz, H-2), 7.06 (m, 2H, H-5 and H-6), 6.58 (d, 1H, J=2 Hz, H-8), 6.50 (d, 1H, J=2 Hz, H-6), 5.28 (s, 2H, OCH.sub.2), 3.94 (s, 3H, OCH.sub.3), 3.89 (s, 6H, 2OCH.sub.3).

Example n.10: Preparation of the Virtual Library

(113) The in house unique library was composed of 816 different natural products. Single molecular entries were generated in SMILES format and then transformed in the 3D SDF format. The LigPrep application of the Schrodinger Maestro suite was used for ionizing compounds at pH=7.51, for generating tautomers and for energy minimization with the OPLS2005 force field (Jorgensen, Maxwell et al. 1996). Ionization and tautomerization states endowed with a normalized probability higher than 0.6 were retained in the library. After this step the library was composed of 1111 individual entries. The Qik-Prop application of the Maestro suite was used to predict chemical and chemico-physical features of all compounds. The conformational analysis was carried out by means of the Build 3D Database protocol of Discovery Studio 2.5 using the CAESAR conformation method with default options and keeping up to 600 conformers for each ligand.

Example n.11: Generation of Pharmacophore Models

(114) A training set of 9 potent SMO antagonists was used to generate pharmacophore models according with a ligand-based procedure. The Common Feature Pharmacophore Generation protocol implemented in Discovery Studio 2.5 from Accelrys was used. Up to 20 pharmacophores were generated, composed by a maximum of 10 features. The minimum inter-feature distance was set at 2.0 while the number of leads that may miss was kept at the default value (1). Conformational analysis of the training set was carried out with the CAESAR method. The maximum number of omitted features during the alignment of the training set to pharmacophores was set at 1. This procedure generated twenty pharmacophores (maximum allowed value, according to custom settings). The six top-ranking pharmacophores are divided into two groups, based on the pharmacophoric feature composition. Type1 pharmacophores have three hydrogen bond acceptor (HBA) and three hydrophobic (HYD) features. Type2 pharmacophores have three HBA, two HYD and a hydrogen bond donor (HBD) features (FIG. 1). Coordinates of pharmacophores and inter-feature distances are given in Tables 1 and 2 for the representative Type1 pharmacophore model (3 HYD; 3 HBA) and in Tables 3 and 4 for the representative Type2 pharmacophore (2 HYD; 1 HBD; 3 HBA).

(115) TABLE-US-00001 TABLE 1 Coordinates of the representative pharmacophore Type 1. Feature x y z Radius () HYD1 5.940 2.940 0.780 1.7 HYD2 4.000 2.360 1.260 1.7 HYD3 5.420 2.960 0.980 1.7 HBA1-tail 1.998 3.801 0.538 1.7 HBA1-head 0.520 6.360 1.180 2.3 HBA2-tail 6.760 5.444 1.054 1.7 HBA2-head 9.760 5.440 1.140 2.3 HBA3-tail 2.885 1.548 0.149 1.7 HBA3-head 1.000 1.980 2.460 2.3 HYD: Hydrophoic feature; HBA: H-Bond Acceptor feature. Coordinates are in .

(116) TABLE-US-00002 TABLE 2 Distance matrix of the representative pharmacophore Type1. For HBA features, distances have been calculated referring to the tail. Distances are expressed in . Feature HYD1 HYD2 HYD3 HBA1 HBA2 HBA3 HYD1 10.164 12.921 4.042 2.649 9.921 HYD2 5.513 6.425 11.430 4.301 HYD3 10.151 14.937 3.114 HBA1 5.064 7.253 HBA2 11.947 HBA3

(117) TABLE-US-00003 TABLE 3 Coordinates of the representative pharmacophore Type 2. Feature x y z Radius () HYD1 5.620 2.740 0.900 1.7 HYD2 4.740 2.800 0.280 1.7 HBD-tail 3.007 3.615 0.875 1.7 HBD-head 2.040 5.900 2.640 2.3 HBA1-tail 3.058 1.953 0.696 1.7 HBA1-head 2.220 0.080 2.900 2.3 HBA2-tail 2.512 4.671 0.244 1.7 HBA2-head 0.440 4.120 0.520 2.3 HBA3-tail 4.167 1.126 0.138 1.7 HBA3-head 6.120 3.440 0.340 2.3 HYD: Hydrophoic feature; HBA: H-Bond Acceptor feature. Coordinates are in .

(118) TABLE-US-00004 TABLE 4 Distance matrix of the representative pharmacophore Type2. For HBA and HBD features, distances have been calculated referring to the tail. Distances are expressed in . Feature HYD1 HYD2 HYD3 HBA1 HBA2 HBA3 HYD1 11.765 2.756 3.119 11.022 9.973 HYD2 10.076 9.184 2.910 3.990 HBD 2.288 9.976 7.661 HBA1 8.706 7.294 HBA2 6.041 HBA3

Example n.12: Pharmacophoric Screening

(119) The in house unique library of natural products, generated as described in example n.10 was screened through the six pharmacophores. First, the Search 3D Database protocol implemented in Discovery Studio 2.5 was used to select molecules able to map the pharmacophores. The Best search method was used, which perform a flexible fit of the ligand conformations against the pharmacophore. The Ligand Pharmacophore Mapping protocol was subsequently used to fit the selected conformations to the pharmacophores and to calculate the FitValue. No omitted features were allowed during ligand fitting to pharmacophores. The flexible fitting method was used, which permit a slight modification of each ligand conformation to better fit the pharmacophore.

(120) By pharmacophore screening, three groups of molecules were selected from the initial library: 1) ligands that map all pharmacophores; 2) ligands that map only type1 pharmacophores and 3) ligands that map only type2 pharmacophores. With the aim of prioritizing small molecular compounds, the Ligand Efficiency (LE) was further calculated for each compound as the ratio between its FitValue and the number of heavy atoms (LE=FitValue/no. heavy atoms). 16 natural compounds endowed with the highest LE were selected for in vitro studies (see above).

Example n.13: Structure-Based Virtual Screening

(121) Initial coordinates of the Gli1-ZF/DNA complex were retrieved from the Protein Data Bank under the PDB accession code 2GLI. Coordinates of crystal water molecules were manually removed from the complex, while cobalt ions were manually replaced with zinc ions within the coordination system of each zinc finger. The Amber11 program (Case, Cheatham et al. 2011) was used for generating MD trajectories and performing energy calculations. The AmberToolsl.5 software was used for preparing input coordinates and topology files, and for performing preliminary analysis on MD trajectories by the ptraj and cpptraj modules. The ff99bsc0 and General Amber (GAFF) force fields were used for parameterizing protein and ligands, respectively. Zinc ions were treated following a bound approach; parameters for the zinc ion and zinc-coordinating residues were adapted from a previous QM calculation (Mori, Dietrich et al. 2010). The Gli1-ZF/DNA complex was solvated in a rectilinear box of explicit water molecules, buffering 8 from the macromolecular system. The TIP3P water model was used. The total charge of the system was neutralized by the addition of sodium counterions (Na+). The solvated macromolecular system was first energy minimized as follows: the water solvent and counterions were first minimized for 250 steps by using a steepest descent algorithm (SD) and for 750 steps by using a conjugate gradient algorithm (CG), while keeping the Gli1-ZF/DNA coordinates as frozen; then, the solvated system was energy minimized for 1000 steps SD and further 4000 steps CG without positional restraints. Before the final production of trajectories, the energy minimized system was gradually heated from 0 to 300 K for 50 s using the Langevin control of the temperature at constant pressure and constraining the Gli1-ZF backbone and the DNA phosphate backbone with a harmonic force constant of 5.0 kcal.Math.mol.sup.1.Math..sup.2. Then, the density of the system was equilibrated for 50 s, by applying the same constraints mask used in the heating step. Restrained MD trajectory were produced for 3 ns while the force constant applied to Glil-ZF and DNA backbone was gradually decreasing from 5 to 2 to 1 kcal.Math.mol.sup.1.Math..sup.2 every 1 ns. After this step, unrestrained MD trajectories were generated for 20 ns by using SANDER. During all MD simulations, a time step of 0.001 s was used. A representative Gli1-ZF structure was extracted from MD trajectories and used as rigid receptor for docking simulations. The GOLD docking program (version 5.0.1) was used to dock the in house library towards the binding site centered on the side chain of Thr374 and having a radius of 18 . The GoldScore function was used whit Genetic Algorithm (GA) efficacy set at 150% and generating 50 runs for ligands, while other parameters were kept at their default values. Docking poses were rescored by means of the MM-GB SA method implemented in Amber11.

Example n.14

(122) NMR experiments were carried out to probe the direct interaction of Glabrescione B to Gli1-ZF and to support computational and biological data.

(123) A sample of Glabrescione B 0.412 mM in DMSO-d.sub.6 was prepared and added to 20 L of a solution of Gli1-GST (Glutathione S-transferase) fusion protein at 5 g/L, providing a molar ratio of Glabrescione B/Gli1-GST of 150:1. Relaxation speeds of Glabrescione B protons were monitored via NMR (600 MHz), results are showed in Table 5. In experimental conditions, Glabrescione B protons experiencing the most significant perturbation of the relaxation speed, due to the presence of Gli1-GST, belong to the ring A of the isoflavone nucleus, namely aromatic protons H.sub.1, H.sub.3 and those belonging to methoxyl groups 2 and 4. In addition, protons H.sub.11 and H.sub.15 belonging to the ring C showed a significant perturbation of the relaxation speed in presence of Gli1, while H.sub.8 showed only a small variation.

(124) It is worth noting that the same experiments were conducted also in presence of GST alone to monitor the influence of GST in the fusion protein Gli1-GST to the binding of Glabrescione B. Results showed that the normalized perturbation of the relaxation speed is significantly lower than that observed in presence of Gli1-GST (Table 5) and suggested that Glabrescione B interacts most strongly with Gli1 than GST.

(125) TABLE-US-00005 TABELLA 5 Normalized relaxation speed (R/R.sub.f) of Glabrescione B protons (0.412 mM) in presence of Gli1-GST and GST alone. GlaB:Gli1-GST GlaB:GST GlaB 1:150 1:150 proton R.sub.f.sup.ms (S.sup.1) R/R.sub.f R/R.sub.f 1 0.41 0.341 0.024 2 1.20 0.267 0.058 3 0.63 0.206 0.100 4 1.39 0.482 0 8 0.36 0.028 0 11 0.63 0.286 n.d. 12 0.89 0 0.034 15 0.76 0.329 0.138 19 0.85 0 0

(126) Moreover, Glabrescione B protons involved in binding to GST are significantly different from those involved in binding to Gli1.

(127) Finally, since Gli1 is a zinc binding protein, it was monitored via NMR the direct interaction between Glabrescione B and Zn.sup.2+, showing that the molecule is not capable of binding to the metal ions.

Example n.15

(128) Since Glabrescione B was the most potent Hh inhibitor identified by the screening, a number of Glabrescione B analogues (namely NT8, NT9, NT10, NT11, NT12, NT13, NT14 and NT15) were synthesized and tested in vitro to improve inhibitory potency against the Hh pathway and to afford Structure-Activity Relationships (SAR) for the congeneric series (see examples 8 and 9). Evaluation of the Hh inhibitory activity of these molecules at 5 M was preliminarily conducted in vitro on Shh Light II cells (see example 2). Preliminary results showed that some of these molecules were at least as active as Glabrescione B, with NT8 and NT9 being more potent than Glabrescione B (see FIG. 19).

BIBLIOGRAPHY

(129) Ahn, S. and A. L. Joyner (2005). Nature 437(7060): 894-897. Aszterbaum, M., et al., (1999). Nat Med 5: 1285-1291. Atwood, S. X. Lynn, A., Chang, S. and Oro, A. E. (2012). J. Cell Biol. 199 (2):193-197. Barnum, D., J. et al. (1996). J Chem Inf Comput Sci 36(3): 563-571. Biegasiewicz, K. F., et al., (2010). Tetrahedron Letters 51(33): 4408-4410. Buonamici, S., et al., (2010). Sci Transl Med 2(51): 51ra70. Case D. A., et al., AMBER 11, University of California, San Francisco. (2011). Clement, V., et al. (2007). Curr Biol 17(2): 165-172. Chen, J. K., et al. (2002). PNAS, 99(22): 14071-14076. Dahlen A, et al. (2004) Am J Pathol 164:1645-1653. Dahmane, N. and A. Ruiz i Altaba (1999). Development 126(14): 3089-3100. De Smaele, E., E. Ferretti and A. Gulino (2010). Curr Opin Investig Drugs 11(6): 707-718. Delle Monache, F.; et al. (1977). Gazzetta Chimica Italiana 107(7-8): 403-407. Di Marcotullio, L., et al. (2006). Mol Neurobiol 34(3): 193-204. Di Marcotullio L, et al. (2011). Oncogene 30, 65-76. Dijkgraaf, G. J., et al. (2011). Cancer Res 71(2): 435-444. Eberl, M., et al. (2012). EMBO Mol Med 4:218-233. Ellison, D. (2002). Neuropathol Appl Neurobiol 28(4): 257-282. Epstein, E. H. (2008). Nat. Rev. Cancer. 8, 743-754. Everett, L., Crabb, D. W. (1999). J. Steroid Biochem Mol Biol, 70(4-6):197-201. Goodrich, L. V., et al. (1997). Science 277(5329): 1109-1113. Hallahan, A. R., et al. (2004). Cancer Res 64(21): 7794-7800. Ingham, P. W. and M. Placzek (2006). Nat Rev Genet 7(11): 841-850. Jorgensen, W. L., et al. (1996). Journal of the American Chemical Society 118(45): 11225-11236. Kinzler K W, et al. (1987) Science 236:70-73. Kogerman P, et al. (1999). Nat Cell Biol 1:312-319. Lai, K., B. K. Kaspar, F. H. Gage and D. V. Schaffer (2003). Nat Neurosci 6(1): 21-27. Lauth, M., Bergstrom, A., Shimokawa, T., and Toftgard R (2007). PNAS 104(20): 8455-8460 Machold, R., et al. (2003). Neuron 39(6): 937-950. Manetti, F., et al. (2010). Mol Pharmacol 78(4): 658-665. Mori, M., et al. J Chem Inf Model 50, 638-50 (2010). Palma, V., et al. (2005). Development 132(2): 335-344. Palma, V. and A. Ruiz i Altaba (2004). Development 131(2): 337-345. Pavletich, N. P. & Pabo, C. O. (1993). Science 261, 1701-7. Ruiz i Altaba, A., P. Sanchez and N. Dahmane (2002). Nat Rev Cancer 2(5): 361-372. Stecca, B. and A. Ruiz i Altaba (2009). EMBO J 28(6): 663-676. Sheng, T., et al. (2006). J Biol Chem 281, 9-12. Svrd, J., et al. (2006). Dev. Cell 10, 187-197. Taipale, J., et al. (2000). Nature 406(6799): 1005-1009. Taylor, M. D., et al. (2002). Nat Genet 31(3): 306-310. Teglund, S. and R. Toftgard (2010). Biochim Biophys Acta 1805(2): 181-208. Tremblay, M. R., et al. (2009). Expert Opin Ther Pat 19(8): 1039-1056. Verdonk, M. L., et al. (2003). Proteins 52, 609-23. Wallace, V. A. (1999). Curr Biol 9(8): 445-448. Wechsler-Reya, R. J. and M. P. Scott (1999). Neuron 22(1): 103-114. Yauch, R. L., et al. (2009). Science 326(5952): 572-574. Ye, Y. Q., et al. (2009). Organic Letters 11(21): 5074-5077.