Posaconazole, a Second-Generation Triazole Antifungal Drug, Inhibits the Hedgehog Signaling Pathway and Progression of Basal Cell Carcinoma

Abstract

Deregulation of Hedgehog (Hh) pathway signaling has been associated with the pathogenesis of various malignancies, including basal cell carcinomas (BCC). Inhibitors of the Hh pathway currently available or under clinical investigation all bind and antagonize Smoothened (SMO), inducing a marked but transient clinical response. Tumor regrowth and therapy failure were attributed to mutations in the binding site of these small-molecule SMO antagonists. The antifungal itraconazole was demonstrated to be a potent SMO antagonist with a distinct mechanism of action from that of current SMO inhibitors. However, itraconazole represents a suboptimal therapeutic option due to its numerous drug-drug interactions. Here, we show that posaconazole, a second-generation triazole antifungal with minimal drug-drug interactions and a favorable side-effect profile, is also a potent inhibitor of the Hh pathway that functions at the level of SMO. We demonstrate that posaconazole inhibits the Hh pathway by a mechanism distinct from that of cyclopamine and other cyclopamine-competitive SMO antagonists but, similar to itraconazole, has robust activity against drug-resistant SMO mutants and inhibits the growth of Hh-dependent BCC in vivo Our results suggest that posaconazole, alone or in combination with other Hh pathway antagonists, may be readily tested in clinical studies for the treatment of Hh-dependent cancers. Mol Cancer Ther; 15(5); 866-76. ©2016 AACR.

Figure 1. Posaconazole Inhibits The Hedgehog Pathway

8x-Gli binding site luciferase reporter SHH-Light2 cells stimulated with SHHN conditioned medium were used. (A) Posaconazole inhibited Hh pathway activation whereas voriconazole, another triazole antifungal drug, did not. (B) Inhibition of Hh pathway activation cannot be reversed with the addition of cholesterol, lathosterol, and desmosterol – components of cholesterol biosynthesis downstream of 14α-lanosterol demethylase, the target of posaconazole’s antifungal activity. (C) Addition of cholesterol, lathosterol, and desmosterol in the absence of SHHN conditioned medium did not induce Hh pathway activity. Data represents mean of triplicates +/- S.D.

Figure 2. Posaconazole Inhibits the Hh Pathway Downstream Of PTCH

(A) In constitutively active Ptch^-/- murine fibroblasts transfected with an 8x-Gli luciferase reporter, posaconazole inhibited Hh pathway activity. (B) Posaconazole was unable to inhibit the pathway activity of SHH-Light SMOA1 cells stably expressing constitutively active mutant SMOA1. (C-E) Histograms of BODIPY-cyclopamine (BD) fluorescence (x-axis), as measured by FACS, using tetracycline-inducible SMO-expressing HEK-293 cells. Graphs are from same experiment separated for clarity. When used, BODIPY-cyclopamine is 5 nM in all panels. (C) Black tracing represents uninduced cells (no SMO) with no BODIPY-cyclopamine. Blue trace represents uninduced cells with BODIPY-cyclopamine. Green trace represents maximal fluorescence in cells with induced SMO-overexpression and BODIPY-cyclopamine. (D) Addition of KAAD-cyclopamine 200 nM (orange - filled trace; 10x IC₅₀) and 400 nM (red - filled trace; 20x IC₅₀) reduced fluorescence back to baseline levels. (E) Treatment with increasing concentrations of posaconazole up to 10 μM (~11x IC₅₀) did not significantly decrease fluorescence of BODIPY-cyclopamine bound to SMO.

Figure 3. Posaconazole Inhibits The Hh Pathway At Distinct Sites From Other SMO Modulators

Panels A-F represent signaling assays using SHH-Light2 cells stimulated with SAG, a small molecule SMO agonist, or SHHN CM. (A) Addition of KAAD-cyclopamine 20 nM (1x IC₅₀) and 100 nM (5x IC₅₀) increased the EC₅₀ of SAG ~3.5- and ~7-fold, respectively, while maintaining maximal pathway activation. (B) Treatment with posaconazole 1 μM (~1.1 × IC₅₀) and 7.5 μM (~5.5x IC₅₀) decreased the maximal activity of SAG but did not significantly affect its EC₅₀. (C) Increasing concentrations of posaconazole while titrating 20(S)-hydroxycholesterol, an agonist that binds SMO at a distinct site from SAG, decreased maximal pathway activation of 20(S)-hydroxycholesterol but did not alter its EC₅₀. (D) Increasing concentrations of 20S-hydroxycholesterol, up to 10 μM, did not significantly alter the IC₅₀ of posaconazole. (E) The addition of posaconazole caused a reduction in the IC₅₀ of KAAD-cyclopamine. (F) Treatment with posaconazole 1 μM (~1.1 × IC₅₀), KAAD-cyclopamine 10 nM (~0.5 × IC₅₀) and GDC-0449 20 nM (~1.5 × IC₅₀) partially inhibited Hh pathway activation by SHHN CM. Combination of either KAAD-cyclopamine or GDC-0449 with posaconazole at the same doses further suppressed pathway activation. (G) Treatment of ASZ and BSZ, constitutively active murine BCC cells, with posaconazole 1 μM (~1.1 × IC₅₀), KAAD-cyclopamine 40 nM (~2.0 × IC₅₀) and GDC-0449 40 nM (~3.1 × IC₅₀) significantly inhibited the pathway as measured by Gli1 mRNA. Combination of posaconazole with either KAAD-cyclopamine or GDC-0449 completely inhibited pathway activity. For panels A-E, respective IC₅₀ or EC₅₀ of the titrated compound under varying conditions are listed in the figure. All data represents mean of triplicates +/- S.D.

Figure 4. Posaconazole Inhibits SMO Accumulation In Primary Cilia

(A) Representative immunofluorescent images of NIH-3T3 cells. Stimulation of Hh pathway with SHHN led to accumulation of SMO in primary cilia. Treatment of posaconazole 7.5 μM with SHHN inhibited SMO accumulation in primary cilia. Insets show enlarged views of boxed primary cilia with acetylated tubulin (red) and SMO (green) channels offset for clarity. (B) Tabulation of SMO accumulation in primary cilia. Data represent mean of at least 7 fields +/- S.D. Scale bar represents 10 μm.

Figure 5. In Vivo Treatment With Posaconazole

(A) Plasma Pharmacokinetics of Posaconazole. Balb-c mice were treated with one dose of posaconazole and plasma levels of posaconazole were measured at various time points after treatment. Each data point represents mean of 3 mice +/- S.D. (B) Posaconazole treatment of nude mice bearing subcutaneous allografts of murine Ptch^+/-;p53^-/- basal cell carcinomas suppressed tumor growth compared to vehicle control. ‘**’ p < 0.005. (C) Posaconazole treatment decreased Hh pathway activation in allografted basal cell carcinomas compared to vehicle control treated tumors, as measured by Gli1 mRNA levels. Gli1 mRNA levels of posaconazole treated tumors were normalized to the mean of vehicle treated tumors. Treatments were given twice per day by oral gavage for 5 days. Line represents mean of data sets.‘*’ p = 0.0224, ‘**’ p = 0.0033 (D and E) BCC tumors treated with posaconazole showed increased tumor necrosis compared to control vehicle. (D) Tabulation of necrosis scores of tumors (see Methods). Line represents mean of data set. ‘**’ p = 0.0028. (E) Representative images of BCC treated with vehicle control and posaconazole 30 mg/kg. Vehicle control treated tumors showed structured and ordered nests of tumor cells surrounded by stroma (black arrowhead). In contrast, posaconazole treated tumors showed destruction of organized tumor epithelial architecture (arrow) with increased stroma (white arrowhead). Scale bar represents 200 μm.

Figure 6. Posaconazole and ATO Inhibit the Activity of Drug-resistant SMO Mutants

Signaling assays of Smo^-/- fibroblasts transfected with 8x-Gli luciferase reporter and either wild-type or mutant Smo constructs resistant to vismodegib (GDC-0449) or sonidegib (LDE225) and stimulated with SHHN CM. Posaconazole, ATO or combination inhibited the drug-resistant mutant SMO activity. SMO D477G and E522K are vismodegib-resistant. All other SMO mutants are resistant to sonidegib. “WT” = Wild Type. Data represents mean of triplicates +/- S.D.