An Unusual Binding Model of the Methyl 9-Anilinothiazolo[5,4-f] quinazoline-2-carbimidates (EHT 1610 and EHT 5372) Confers High Selectivity for Dual-Specificity Tyrosine Phosphorylation-Regulated Kinases

Supporting Information An Unusual Binding Mode of the Methyl 9-Anilinothiazolo[5,4f]quinazoline-2-carbimidates (EHT 1610 and EHT 5372) Confers High Selectivity for Dual-specificity Tyrosine Phosphorylation-Regulated Kinases. Apirat Chaikuad,† Julien Diharce,‡ Martin Schröder, ¥ Alicia Foucourt,§ Bertrand Leblond,ǁ Anne-Sophie Casagrande,ǁ Laurent Désiré,ǁ Pascal Bonnet,‡ Stefan Knapp,*,†,¥ Thierry Besson*,§ † Target Discovery Institute (TDI), and Structural Genomics Consortium (SGC), University of Oxford, Old Road Campus Research Building, Oxford OX3 7DQ, U.K. ‡ Institut de Chimie Organique et Analytique, UMR CNRS-Université d’Orléans 7311, Université d’Orléans BP 6759, Orléans 45067 Cedex 2, France § Normandie Univ, UNIROUEN, INSA Rouen, CNRS, COBRA UMR 6014, 76000 Rouen, France ǁ Diaxonhit, 63-65 Boulevard Masséna, 75013 Paris, France ¥ Institute of Pharmaceutical Chemistry and Buchman institute for life sciences, Goethe-University, Max-vonLaue-Str. 9, 60438 Frankfurt am Main, Germany Table of contents 1. Chemistry S2 1.1. Schematic synthesis of the target compounds 1 (EHT 5372) and 2 (EHT 1610). S2 1.2. 1H NMR of compounds 1 (EHT 5372) and 2 (EHT 1610). S3 1.3. Purity of compounds 1 (EHT 5372) and 2 (EHT 1610). S4 2. Protein expression and purification S6 3. Crystallization, data collection and structure determination S6 4. Thermal stability (Tm shift) assays S6 5. Structural modelling of DYRK1B and docking of 1 and 2 in DYRK1A and DYRK1B S6 5.1. Docking Experiment S6 5.2. Isothermal titration calorimetry S7 6. References S7 7. PDB accession codes S8 S1 1. Chemistry. 1.1. Schematic synthesis of the target compounds 1 (EHT 5372) and 2 (EHT 1610). Compounds 1 (EHT 5372) and 2 (EHT 1610) were prepared in three steps from the key 6aminobenzo[d]thiazole-2,7-dicarbonitrile itself obtained in six steps from 5-nitroanthranilinitrile. All details covering the various synthetic routes of compounds 1 (EHT 5372) and 2 (EHT 1610) are given in papers cited in ref 24, 25 and 26 in the manuscript. The scheme below is describing the experimental conditions and the yields obtained. Scheme S1. Multistep microwave-assisted (w) synthesis of methyl 9-anilinothiazolo[5,4f]quinazoline-2-carbimidates 1 (EHT 5372) and 2 (EHT1610) used in this study. Reagents and conditions: (a) Boc2O, DMAP, Et3N, CH2Cl2, rt, 4 h; (b), HCO2NH4, Pd/C, EtOH, 85 °C (w), 0.5 h; (c) Br2, AcOH, CH2Cl2, rt, 2.5 h; (d) Appel salt, Py. (2 equiv), CH2Cl2, rt, 4 h; (e) AcOH, 118 °C (w), 2 h; (f) CuI, Py, 130 °C (w), 20 min; (g) DMFDMA, DMF, 70 °C (w), 2 min, 86%; (h) 2,4-dichloroaniline for 1 and 2-fluoro-4-methoxyaniline for 2 (1.5 equiv), AcOH, 118°C (w), 50 min; 35% for 1 and 5 min; 85% for 2; (i) NaOMe (0.5 M in MeOH), MeOH, 65 °C (w), 0.5 h, 81% (1) and 82% (2). S2 1.2. 1NMR of compound 1 (EHT 1610) and 2 (EHT 5372) Methyl 9-(2,4-dichlorophenylamino)thiazolo[5,4-f]quinazoline-2-carbimidate (1). Methyl 9-(2-fluoro-4-methoxyphenylamino)thiazolo[5,4-f]quinazoline-2-carbimidate (2). S3 1.3. Purity of compounds 1 (EHT 5372) and 2 (EHT 1610) : Chromatograms and results. Injection Details Injection Name: Vial Number: Injection Type: Calibration Level: Instrument Method: Processing Method: Injection Date/Time: EHT5372 GB3 Unknown Run Time (min): Injection Volume: Channel: Wavelength: Bandwidth: Dilution Factor: Sample Weight: 60C-40D-15min methode traitement 04/nov/11 10:56 Integration Results No. Peak Name Retention Time Area min mAU*min 1 2,690 0,044 2 5,113 0,044 3 9,407 0,026 4 11,820 3,132 Total: 3,246 Height mAU 0,059 0,055 0,039 9,966 10,119 S4 15,00 20,00 UV_VIS_2 289,0 2 1,0000 1,0000 Relative Area Relative Height Amount % % n.a. 1,36 0,58 n.a. 1,35 0,54 n.a. 0,80 0,39 n.a. 96,49 98,49 n.a. 100,00 100,00 Injection Details Injection Name: Vial Number: Injection Type: Calibration Level: Instrument Method: Processing Method: Injection Date/Time: EHT1610 GB1 Unknown Run Time (min): Injection Volume: Channel: Wavelength: Bandwidth: Dilution Factor: Sample Weight: 60C-40D-15min methode traitement 11/avr/11 08:37 15,00 20,00 UV_VIS_2 289,0 2 1,0000 1,0000 Integration Results No. 1 2 3 4 Total: Peak Name Retention Time Area min mAU*min 1,807 0,027 3,530 0,013 4,397 0,008 6,027 12,357 12,404 S5 Height mAU 0,172 0,108 0,068 73,148 73,497 Relative Area Relative Height Amount % % n.a. 0,21 0,23 n.a. 0,10 0,15 n.a. 0,06 0,09 n.a. 99,62 99,53 n.a. 100,00 100,00 2. Protein expression and purification Recombinant DYRK2 kinase domain was expressed as previously described.1 The recombinant protein was initially purified using affinity chromatography, and the His6-tag was subsequently cleaved using TEV protease. The cleaved protein was further purified using reverse affinity chromatography and size-exclusion chromatography in buffer containing 50 mM HEPES, ph 7.5, 250 mM NaCl and 0.5 mM TCEP. The pure protein was concentrated to 10.3 mg/mL. 3. Crystallization, data collection and structure determination DYRK2 kinase domain was pre-incubated with the inhibitors at 1 mM. The crystals of the complexes were obtained using sitting drop vapour diffusion method at 4 °C. Viable DYRK21 crystals grew in 25% PEG 3350, 0.2 M NaCl and 0.1 M bis-tris, pH 5.5, while the crystals of the DYRK2-2 complex was obtained using the condition containing 1.5 M Li2SO4 and 0.1 M HEPES, pH 7.5. Crystals of both complexes were cryo-protected in the reservoir solution supplemented with 22% ethylene glycol, and flash-cooled in liquid nitrogen. Diffraction data were collected at Diamond Light Source, beamline I04-1, and were processed and scaled with Mosflm2 and Scala,3 respectively. Structures were solved by molecular replacement method using Phaser4 and the published coordinates of DYRK2.1 Iterative cycles of manual model building alternated with structure refinement were performed in COOT5 and REFMAC.6 The final models were verified for their geometric correctness using MOLPROBITY.7 4. Thermal stability (Tm shift) assays The proteins at 2 µM in 10 mM HEPES, pH 7.5 and 500 mM NaCl were incubated with the compounds at 10 µM, and the complexes were mixed with SyproOrange. For CLK1 and CLK3, 50 mM arginine-glutamate mix was also supplemented into the buffer. The assays and data analyses were performed using a Real-Time PCR Mx3005p machine according to the protocol previously described.8 5. Structural modelling of DYRK1B and docking of 1 and 2 in DYRK1A and DYRK1B Structure preparation DYRK1A coordinates have been taken from crystallographic structure (PDB ID: 4YLJ). For DYRK1B, since no crystal structure has been reported, an homology model was constructed. Sequence alignment, using the MAFFTT module of Jalview,9 reveals a very high identity on the kinase domain between DYRK1A and DYRK1B (85% and 93% for identity and similarity respectively). MODELLER10 software was used to build the DYRK1B model, with the DYRK1A structure as template. Structural validation has been made with the use of PROCHECK.11 Preparation of both structures for docking calculation was performed using MOE2014 (Molecular Operating Environment) software.12 Physiological pH is considered for residue protonation. Solvent-accessible surface area (SASA) calculations have been made with MOE. 5.1. Docking Experiment Prediction of binding mode of inhibitors was carried out by docking experiments with rDock software13 using default genetic algorithm parameters. Grid cavity was generated with rDock and centered on the DYRK1A co-crystal ligand. For DYRK1B, the grid was build and centered on the same co-crystal ligand after superimposition of DYRK1A crystal structure on DYRK1B homology model. S6 5.2. Isothermal titration calorimetry The ITC measurement was performed on a NanoITC (TA Instruments) at 30 °C in buffer containing 50 mM HEPES pH 7.5, 500 mM NaCl, 0.5 mM TCEP and 5% Glycerol. DYRK1A at 45 µM was injected into the cell, containing compound 2 at 3 µM. The integrated heat of titration was calculated and fitted to a single, independent binding model using the software provided by the manufacture. The thermodynamic parameters (ΔH and TΔS), equilibrium association and dissociation constants (Ka and KD), and stoichiometry (n) were calculated. 6. References 1) Soundararajan, M.; Roos, A.K.; Savitsky, P.; Filippakopoulos, P.; Kettenbach, A.N.; Olsen, J.V.; Gerber, S.A.; Eswaran, J.; Knapp, S.; Elkins, J.M. Structures of Down syndrome kinases, DYRKs, reveal mechanisms of kinase activation and substrate recognition. Structure. 2013, 22, 986-996 2) Powell, H.R.; Johnson, O.; Leslie, A.G. Autoindexing diffraction images with iMosflm. Acta Cryst. D. 2013, 69, 1195-1203 3) Evans, P.R. An introduction to data reduction: space-group determination, scaling and intensity statistics. Acta Cryst. D. 2011, 67, 282-292 4) McCoy, A.J.; Grosse-Kunstleve, R.W.; Adams, P.D.; Winn, M.D.; Storoni, L.C.; Read, R.J. Phaser crystallographic software. J. Appl. Cryst. 2007, 40, 658-674 5) Debreczeni, J.É.; Emsley, P. Handling ligands with Coot. Acta Cryst. D. 2012, 68, 425430 6) Rinaldelli, M.; Ravera, E.; Calderone, V.; Parigi, G.; Murshudov, G.N.; Luchinat, C. Simultaneous use of solution NMR and X-ray data in REFMAC5 for joint refinement/detection of structural differences. Acta Cryst. D. 2014, 70, 958-967 7) Chen, V.B.; Arendall, W.B. 3rd; Headd, J.J.; Keedy, D.A.; Immormino, R.M.; Kapral, G.J.; Murray, L.W.; Richardson, J.S.; Richardson, D.C. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Cryst. D. 2010, 66, 12-21 8) Fedorov, O.; Niesen, F.H.; Knapp, S. Kinase inhibitor selectivity profiling using differential scanning fluorimetry. Methods Mol Biol. 2012, 795, 109-18 9) Waterhouse, A.M., Procter, J.B., Martin, D.M.A., Clamp, M., Barton, G.J. Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics, 2009, 25(9), 1189-1191 10) B. Webb, A. Sali. Comparative Protein Structure Modeling Using Modeller. Current Protocols in Bioinformatics, John Wiley & Sons, Inc., 5.6.1-5.6.32, 2014. 11) Laskowski, R.A., MacArthur, M.W., Moss, D.S., Thornton, J.M. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr, 1993, 26(2), 283-291. 12) Molecular Operating Environment (MOE), 2014.09; Chemical Computing Group Inc., 1010 Sherbooke ST. West, Suite#910, Montreal, QC, Canada, H3A 2R7, 2014. 13) Ruiz-Carmona S, Alvarez-Garcia D, Foloppe N, Garmendia-Doval AB, Juhos S, et al. rDock: A Fast, Versatile and Open Source Program for Docking Ligands to Proteins and Nucleic Acids. PLoS Comput. Biol., 2014, 10(4): e1003571. S7 Supplementary Table S1: Data collection and refinement statistics Complex PDB accession code Data Collection Beamline Wavelength (Å) Resolutiona (Å) Spacegroup Cell dimensions No. unique reflectionsa Completenessa (%) I/σIa Rmergea Redundancya Refinement ligands No. atoms in refinement (P/L/O)b Rfact (%) Rfree (%) Bf (P/L/O)b (Å2) rms deviation bondc (Å) rms deviation anglec (°) Molprobity Ramachandran favour Ramachandran allowed DYRK2-1 5LXC DYRK2-2 5LXD Diamond, beamline I04-1 Diamond, beamline I04-1 0.9200 0.9200 49.43-2.15 (2.27-2.15) 47.04-2.58 (2.72-2.58) C2 C2 a = 130.2, b = 61.0, c = 148.8 a = 66.2, b = 130.0, c = 136.3 Å Å α = γ = 90.0, β = 105.0° α = γ = 90.0, β = 90.4° 60,094 (8,856) 35,477 (5,224) 97.6 (98.7) 97.8 (99.4) 8.2 (2.1) 9.1 (2.0) 0.107 (0.578) 0.091 (0.649) 4.0 (4.0) 3.4 (3.5) 1 6,269/ 52/ 440 2 6,257/ 54/ 233 20.6 25.4 50/ 40/ 45 0.015 1.4 19,7 25.3 68/ 54/ 51 0.011 1.2 94.46 99.74 94.96 99.74 a Values in brackets show the statistics for the highest resolution shells. P/L/O indicate protein, ligand molecule, and other (water and solvent molecules), respectively. c rms indicates root-mean-square. b S8