Fig 1.
In silico screening pipeline for SMO binders.
(a) Schematic representation of the procedure (b) Ligand-APF Models used for screening of the DrugBank Database. Aromatic and aliphatic features are represented by white and yellow, respectively; hydrogen bond donors and acceptors are blue and red, respectively; positive and negative charges are gray and pink, respectively. (c) Pocket Docking Models used for the docking procedure. Carbon, oxygen, nitrogen, and sulfur atoms in the pockets are colored white, red, blue, and yellow, respectively.
Table 1.
Pocket docking scores, and hit list percentile ranks for known SMO modulators and the selected screening candidates.
For the known SMO modulators, literature-reported SMO inhibitory potency is given.
Fig 2.
Crystallographic or predicted binding poses of selected SMO ligands (Cyclopamine, Vismodegib, SANT-1, Nilotinib, Imatinib and Mebendazole).
The binding cavity of SMO can be tentatively separated into two subpockets–upper binding pocket and lower binding pocket. Cyclopamine (a), Vismodegib (b), and SAG (not shown) bind in the upper binding pocket, while SANT-1 (c) binds in the lower binding pocket. In their predicted binding poses, Nilotinib (d) and Imatinib (e) span both upper and lower binding sub-pockets and form hydrogen bonds with Y394, R400 / E518 and H470. By contrast, Mebendazole (f) is more similar to SANT-1 because it occupies the lower sub-pocket only. Receptor residues making contacts with the ligands are shown as sticks. Selected residues making the strongest contacts[82] are labeled and colored pink (upper binding sub-pocket) and purple (lower binding sub-pocket). Residues making hydrogen bonds with the ligand are labeled in bold print.
Fig 3.
Similarity between SMO and ABL1 binding cavities.
(a) Ribbon diagram of structure of the SMO bound to ANTA XV antagonist (PDB 4QIM) (b) Ribbon diagram of structure of ABL1 bound to Nilotinib (PDB 3CS9) (c) Molecular surface of the TM binding pocket of SMO with ANTA XV (d) Molecular surface of the binding pocket of ABL1 with Nilotinib.
Fig 4.
Hedgehog pathway inhibition by Nilotinib, Imatinib and Mebendazole.
The drugs were applied at different concentrations to the NIH 3T3 Gli-RE cells followed by stimulation of the cells with ShhN-conditioned media (n = 3). The data is presented as percent of the maximum luminescence in the absence of antagonists. Cyclopamine and Vismodegib were included as control antagonists. Nilotinib, Imatinib and Mebendazole, but not Nimesulide or Thalidomide, demonstrated dose-dependent inhibition of ShhN-induced Hh pathway activation. Reported clinically achievable therapeutic concentrations of Nilotinib are marked with a pink rectangle. Data represent the mean and standard deviation for at least three independent experiments.
Table 2.
IC50 values of control and test compounds observed in various assays.
Fig 5.
Nilotinib binds to SMO and competes with Cyclopamine and SAG.
(a) Representative distributions of cell fluorescence intensities measured by flow cytometry and demonstrating competitive displacement of BODIPY-Cyclopamine by Vismodegib, Nilotinib and Imatinib at 10 μM concentration. (b) Concentration response curve for flow cytometry based competitive binding assay in HEK293T cells transiently transfected with mSMO: positive controls (Cyclopamine and Vismodegib) and test compounds—Nilotinib and Imatinib displaced BODIPY-Cyclopamine from SMO in a dose-dependent manner. Mebendazole did not demonstrate any appreciable displacement of BODIPY-Cyclopamine in this assay. Data represent the mean and standard deviation for two independent experiments, max GMFI: Maximum Geometric Mean Fluorescence Intensity; UT: Untransfected HEK293T cells (c) Concentration response curve of SAG in the Gli-Luciferase functional assay in the absence or presence of inhibitors: Nilotinib or Vismodegib shifted EC50 of SAG to the right without decrease in maximal response, suggesting a competitive relationship. Imatinib did not shift EC50 of SAG possibly due to its low potency as observed in the functional assay. Data represent the mean and standard deviation for three independent experiments.
Fig 6.
Nilotinib inhibits cell viability and neurosphere formation in medulloblastoma cells (MB-PDX and DAOY).
(a) Representative bright field images of MB-PDX cells cultured in Neurosphere (NS) and Adherent (Adh) conditions with vehicle and Nilotinib (5 and 20 μM) (magnification 10×). Concentration response curve for cell viability in low serum (NS-formation) conditions (n = 3) in MB-PDX cells (b) and (c) Concentration response curve for cell viability in low serum (NS-formation) conditions (MB-PDX cells–b and DAOY cells–c); n = 3. The IC50 values for cell viability assay observed in MB-PDX and DAOY cells are also provided next to panel (c) under the heading “IC50 in NS Culture Conditions”. (d) Concentration response curves for inhibition of cell viability in adherent conditions (n = 3). Drugs did not affect viability of cells grown in adherent conditions. (e) Neurosphere formation assay performed three times in presence of 10 μM concentration of drugs in MB-PDX and DAOY cells. Nilotinib was observed to be most potent in inhibiting the formation of NS from single cells. (f) A representative Western blot for quantitation of nuclear Gli-1 in MB-PDX and DAOY cells shows a decrease in Gli-1 protein after treatment with 5 μM Nilotinib for 24 hours. Tata Binding Protein (TBP) was used as loading control for nuclear fractions. For positive controls (mSMO-transfected HEK293T cell lysate), a reduced amount of cell lysate was loaded to avoid overstaining of the blot: this explains disproportional loading controls one sample in each blot. (g) Quantification of nuclear Gli-1 in MB-PDX and DAOY cells treated or not with drugs. Fig. 6 b-e,f: Image represents the mean and standard deviation of at least three experiments performed on different days. Asterisks represent p-values: *, p < 0.05; **, p < 0.005; ***, p < 0.0005.
Fig 7.
Nilotinib reduces tumor growth and Gli-1 mRNA expression in-vivo.
(a) Nilotinib treatment (40 mg/kg/day) led to significant reduction in tumor volume as compared to vehicle treated group in the MB-PDX xenograft model (n = 2). (b) Gli-1 mRNA expression is reduced in Nilotinib (10 μM) treated tumors as compared to vehicle treated tumors of MB-PDX xenograft model (n = 3). (c) Gli-1 mRNA expression is decreased in the MB (Math1-Cre; ptch1-flox/flox) cells propagated in mice and treated ex-vivo with Nilotinib (10 μM) in ultra-low attachment plates for 24 hours (n = 3). Data represent the mean and standard deviation for repeat experiments, and asterisks represent p-values as follows: *, p < 0.05; **, p < 0.005.
Fig 8.
Multi-pathway pharmacology of Nilotinib vs SMO-specific drugs for Hh-dependent cancers.
Targets of Nilotinib include the Hh-pathway (identified in this study) and several other pathways that are either already known to be dysregulated in Hh-dependent cancers or may serve as escape pathways and are implicated in other cancers. The ability of Nilotinib to inhibit multiple targets simultaneously in Hh-dependent cancers makes it suitable candidate for therapeutic candidate for personalized medicine as compared to specific SMO inhibitors like Vismodegib. Target type is shown by color. RTK stands for receptor tyrosine kinase and nRTK stands for non-receptor tyrosine kinase.