Development and in vitro Profiling of Dual FXR/LTA4H Modulators

Abstract Designed polypharmacology presents as an attractive strategy to increase therapeutic efficacy in multi‐factorial diseases by a directed modulation of multiple involved targets with a single molecule. Such an approach appears particularly suitable in non‐alcoholic steatohepatitis (NASH) which involves hepatic steatosis, inflammation and fibrosis as pathological hallmarks. Among various potential pharmacodynamic mechanisms, activation of the farnesoid X receptor (FXRa) and inhibition of leukotriene A4 hydrolase (LTA4Hi) hold promise to counteract NASH according to preclinical and clinical observations. We have developed dual FXR/LTA4H modulators as pharmacological tools, enabling evaluation of this polypharmacology concept to treat NASH and related pathologies. The optimized FXRa/LTA4Hi exhibits well‐balanced dual activity on the intended targets with sub‐micromolar potency and is highly selective over related nuclear receptors and enzymes rendering it suitable as tool to probe synergies of dual FXR/LTA4H targeting.


In vitro biological evaluation
Aqueous solubility. Aqueous solubility of compounds 3 and 19 were determined using Whatman Uniprep filters (Whatman plc, Maidstone, UK). 5 mg of each compound and 3 mL H2O dest. were inserted into the Uniprep vessel and the mixture was shaken at 37 °C for 24 h. The mixture was then pressed through the Uniprep filter and the concentration of dissolved compound in filtrate was quantified by HPLC (Varian ProStar, SpectraLab Scientific Inc. equipped with a MultoHigh100 Phenyl 5 µ 240+4 mm column, CS-Chromatographie Service GmbH) using external calibration.
WST-1 assay. WST-1 assay (Roche Diagnostics International AG, Rotkreuz, Schweiz) was performed according to manufacturer's protocol. In brief, HepG2 cells were seeded in DMEM high glucose, supplemented with sodium pyruvate (1 mM), penicillin (100 U/mL), streptomycin (100 µg/mL) and 10% FCS in 96-well plates (3•10 4 cells/well). After 24 h, medium was changed to DMEM high glucose, supplemented with penicillin (100 U/mL), streptomycin (100 µg/mL) and 1% charcoal stripped FCS and cells were incubated with 3 or 19 (final concentrations 0.1 µM, 1 µM, 10 µM, 30 µM and 100 µM), Revlotron as positive control, and DMEM/1% DMSO as negative control. After 48 h, WST-1 reagent (Roche Diagnostics International AG) was added to each well according to manufacturer's instructions. After 45 min incubation, absorption (450 nm/ reference: 620 nm) was determined with a Tecan Infinite M200 (Tecan Deutschland GmbH, Germany). Each experiment was repeated at least three times in duplicates. Results (expressed as mean percent of untreated control±SEM; n=4; DMSO=100%). sEH activity assay. Human sEH full length protein was purified as described by Lukin et al. 4 . In brief, the protein was expressed in E.coli BL21 (DE3) and purified on an ÄKTA purifier (GE Healthcare) with a 5 ml HisTrap HP (GE Healthcare) followed by a dialysis with a volume ratio of ~ 1:100 for at least 10 h at 4°C. Pure protein in buffer (50 mM Tris, 500 mM NaCl, 5% glycerol, and pH 8 supplemented with an additional 20% glycerol) was frozen in liquid nitrogen and stored at -80°C. The sEH-hydrolase activity assay procedure was adapted from Hahn et al. and Lukin et al. 4 as described by Brunst et al. 5 . Final compound concentrations were measured between 0 -300 µM.
Isothermal titration calorimetry. Recombinant FXR LBD for ITC experiments was obtained as described previously 2,5 .In brief, the protein was expressed in E.coli BL21 (DE3) and purified on an ÄKTA purifier (GE Healthcare) with a 5 ml HisTrap HP (GE Healthcare). Cleavage of the His6-tag was achieved by adding self-produced TEV protease in a molecular ratio of 1:10. This was followed by another purification on an ÄKTA purifier with a 5 ml HisTrap HP (GE Healthcare). This time the flow trough was collected and a size-exclusion chromatography (HiLoad 16/600 Superdex 200 pg TM , GE Healthcare) was performed. Pure protein in buffer (10 mM Tris, 100 mM NaCl, 5 mM DTT, pH=8.3) was frozen in liquid nitrogen and stored at -80°C. LTA4H protein was obtained as described by Hiesinger et al. 6 . In brief, the protein was expressed in E.coli BL21 (DE3) and purified on an ÄKTA purifier (GE Healthcare) with a 5 ml HisTrap HP (GE Healthcare) followed by size-exclusion chromatography (HiLoad 16/600 Superdex 200 pg TM , GE Healthcare). Pure protein in buffer (50 mM Tris, 50 mM NaCl, pH=8) was frozen in liquid nitrogen and stored at -80°C.
For ITC measurements, the respective protein (FXR or LTA4H) was dialyzed for 4 to 12 hours with the respective buffer in a 1:250 volume ratio, centrifuged at 4°C with 21130 xg for 10 min and diluted to the desired concentration using the dialysis buffer and DMSO to a final concentration of 1% DMSO. Test compounds were prepared by diluting stock solution in dialysis buffer and adding the respective amount of DMSO for a final concentration of 1%. The respective protein (150 µM FXR LBD or 40 µM LTA4H) was titrated to compound (10-60 µM 8 or 15-60 µM 19) with varying test compound concentrations. ITC measurements were performed on an Affinity ITC (TA-Instruments) at a temperature of 25 °C and a stirring rate of 75 rpm. Control experiments were performed by titrating protein to buffer or buffer to compound with otherwise identical conditions. Each protein-ligand interaction was tested twice with varying ligand concentration. Buffers were the following. FXR LBD: 10 mM Tris, 100 mM NaCl, 5 mM DTT, pH=8.3. LTA4H: 50 mM Tris, 50 mM NaCl, pH=8.0.

Computational Methods
Molecular docking. For molecular docking, MOE2019.0102 (Chemical Computing Group, Montreal, Canada) was used. X-ray structures in complex with compounds displaying structural similarity to compound 19 (FXR: 4QE8 2 ; LTA4H: 3FHE 1 ) were downloaded from the Protein Data Bank and prepared for docking using the QuickPrep routine with default settings. Ligand 19 was loaded in an MOE database and prepared for docking using Wash routine (default settings, adjusting dominant protonation state at pH 7.4, rebuilding 3D coordinates). Molecular docking was conducted using Similarity placement and induced fit method. Rescoring was performed using GBVI/WSA dG scoring function for generating 5 diverse conformations, followed by energy minimization of the binding site (ligand and residues within 4.5A). The most probable binding mode was selected manually, with special attention to interactions of polar groups.