Discovery and Characterization of a Novel Series of Chloropyrimidines as Covalent Inhibitors of the Kinase MSK1

We describe the identification and characterization of a series of covalent inhibitors of the C-terminal kinase domain (CTKD) of MSK1. The initial hit was identified via a high-throughput screening and represents a rare example of a covalent inhibitor which acts via an SNAr reaction of a 2,5-dichloropyrimidine with a cysteine residue (Cys440). The covalent mechanism of action was supported by in vitro biochemical experiments and was confirmed by mass spectrometry. Ultimately, the displacement of the 2-chloro moiety was confirmed by crystallization of an inhibitor with the CTKD. We also disclose the crystal structures of three compounds from this series bound to the CTKD of MSK1, in addition to the crystal structures of two unrelated RSK2 covalent inhibitors bound to the CTKD of MSK1.

Starting materials were either commercially available or can be prepared according to reported literature procedures. All reactions involving air-or moisture-sensitive reagents were performed under a nitrogen or argon atmosphere using anhydrous solvents and dried glassware. Commercial solvents and reagents were generally used without further purification, including anhydrous solvents when appropriate (usually Sure-Seal TM products from Aldrich Chemical Company or AcroSeal™ from ACROS Organics). In general reactions were followed by LC-MS analysis. Products were generally dried under vacuum before final analyses and submission to biological testing. LC/MS analyses are performed as follows. No unexpected or unusually high safety hazards were encountered.
Method A: Acidic Shimadzu 2010EV single quadrupole mass spectrometer was used for LC/MS analysis. This spectrometer was equipped with an ESI source and LC-20AD binary gradient pump, SPD-M20A photodiode array detector (210-400 nm). Data was acquired in a full MS scan from m/z 70 to 1200 in positive and negative mode. The reverse phase analysis was carried out by using Waters XBridge C 18 (30 X 2.1)mm 2.5 µ column. Gradient elution was done with 5 mM ammonium formate in water +0.1% formic acid (Phase A) and Acetonitrile +5% solvent A +0.1% formic acid (Phase B), with gradient 5-95%B in 4.0 min hold till 5.0 min, 5% B at 5.1 min hold till 6.5 min. HPLC flow rate: 1.0 ml/min, injection volume: 5 µL. Retention times (T R ) given in minutes.
MS parameters: Detector voltage 1.5 kV. Source block temperature 200°C. Desolvation temperature 240°C. nebulizing gas flow 1.2 L/min (Nitrogen). Data was acquired in a full MS scan from m/z 70 to 1200 in positive and negative mode.

Method B: Basic
Shimadzu 2010EV single quadrupole mass spectrometer was used for LC-MS analysis. This spectrometer was equipped with an ESI source and LC-20AD binary gradient pump, SPD-M20A photodiode array detector (210-400 nm). Data was acquired in a full MS scan from m/z 70 to 1200 in positive and negative mode. The reverse phase analysis was carried out by using Waters XBridge C 18 (30 X 2.1)mm 2.5 µ column Gradient elution was done with 5 mM ammonium formate in water +0.1% Ammonia (solvent A),or Acetonitrile +5% solvent A+0.1% Ammonia (solvent B), with gradient 5-95% B in 4.0 min hold till S3 5.0 min, 5%B at 5.1 min hold till 6.5 min. HPLC flow rate: 1.0 ml/min, injection volume: 5 µL. Retention times (T R ) given in minutes.
MS parameters: Detector voltage 1.5 kV. Source block temperature 200°C. Desolvation temperature 240°C. Nebulising gas flow 1.2 L/min (Nitrogen). Data was acquired in a full MS scan from m/z 70 to 1200 in positive and negative mode. NMR NMR spectra are recorded on a Varian VNMR 400 MHz NMR fitted with Linux 3.2 software with operating system Redhat enterprise Linux 6.3 and 5 mm inverse 1 H/ 13 C/ 19 F triple probe head. The compounds are studied in deuterated solvents as indicated at a probe temperature of 300 K and at a concentration around 4-5 mg/mL. The instrument was locked on the deuterium signal of the deuterated solvent used. Chemical shifts are given in ppm downfield from TMS (tetramethylsilane) taken as internal standard. Splitting patterns are designated as s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad.

Preparative Purification
Preparative purification was performed in an in-line mass-directed reverse phase semi-preparative HPLC.A SQD Waters single quadrupole mass spectrometer was used for LC-MS purification analysis. This spectrometer was equipped with an ESI source, General Procedure for the addition of 4-azaindole to chloropyrimidines. A solution of 4-azaindole (100 mg, 0.8126 mmol) in anhydrous DMF (4 mL) was treated with sodium hydride (36 mg, 0.9 mmol, 60 mass% in mineral oil) and the resulting mixture was stirred at room temperature for 10 minutes before addition of 2,4,5-trichloropyrimidine (183 mg, 0.9778 mmol). After stirring at room temperature for 15 minutes, the reaction mixture was diluted with EtOAc and water. The layers were separated and the aqueous layer was extracted with EtOAc (2 x). The combined organic layers were washed with water (3 x), saturated brine, dried over Na 2 SO 4 , filtered and concentrated in vacuo to give a yellow residue. The product was purified by silica gel column chromatography (regular silica 30 μ, 10 g SNAP Ultra, liquid loading in CH 2 Cl 2 ) eluting with a gradient of CH 2 Cl 2 /MeOH 1:0 to 9:1) followed by silica gel column chromatography (regular silica 30 μ, 25 g SNAP Ultra, solid loading on Celite) eluting with a gradient of heptane/EtOAc 4:1 to 7:3 to provide 1-(2,5dichloropyrimidin-4-yl)pyrrolo [3,2-b]   To a solution of 1-(2,5-dichloropyrimidin-4-yl)pyrrolo [3,2-b]pyridine (1, 77 mg, 0.29 mmol) in acetic acid (2 ml) was added hydrobromic acid (33 mass% in acetic acid) (225.6 mg, 0.93 mmol). The resulting mixture was stirred at 80°C for 5 days. The reaction mixture was then taken up with dichloromethane and washed twice with a saturated aqueous solution of sodium hydrogen carbonate. The organic layer was dried over magnesium sulfate, filtered and concentrated in vacuo. The residue was purified by reverse phase chromatography in basic mode to provide 1-(2-bromo-5-chloro-pyrimidin-4-yl)pyrrolo [3,2-  HPLC Purity: 97%.

S13
A solution of 5-(2,5-dichloropyrimidin-4-yl)-7-(p-tolylsulfonyl)pyrrolo [2,3-d]pyrimidine (21a, 44.5 mg, 0.106 mmol) in DCM (2 mL) was treated with TBAF (0.1 mL, 1 mmol, 1 M in THF) and the resulting orange solution was stirred at room temperature for 30 minutes. The reaction was treated with TBAF (0.1 mL, 1 mmol, 1 M in THF) and the resulting orange solution was stirred at room temperature for 30 minutes. The reaction was quenched with a saturated solution of sodium bicarbonate and the aqueous layer was extracted with DCM.

Recombinant expression and purification
Bacterial MSK CTKD constructs were transformed into BL21(DE3) cells and expressed overnight at 25°C in TB media with kanamycin after induction with 1 mM IPTG. For expression in insect cells, each MSK DNA was first co-transfected into Sf9 cells with linearized baculovirus DNA to create viral stocks. Tni cells were then infected with each baculovirus at an MOI of 2 and cells were harvested after 2-3 days of expression at 27°C. For mammalian expression, MSK1 DNA was transfected into HEK293 cells using PEI and cells were harvested after expression for 2 days at 37°C. Cell pastes for all MSK variants were lysed and purified by Ni-IMAC. N-terminal tags were cleaved by in-house proteases (TEV or Ulp-1) and cleaned up by a second Ni-IMAC step and polished with size exclusion chromatography. Avi-tagged MSK was in vitro biotinylated with BirA (Avidity). Lambda phosphatase treatment was used to dephosphorylate MSK. For crystallography, the final protein was concentrated to 10 mg/mL in 25mM HEPES pH 7.5, 150mM NaCl, 5% Glycerol, and 5mM BME, aliquoted and flash frozen.

Recombinant Enzyme assays
ERK2-MSK1 cascade assay: 0.125 nM inactive MSK1 enzyme was pre-incubated for 1 hour with compounds, before addition of 0.006 nM ERK2 enzyme (Thermo PR5257A), 0.5 µM peptide substrate (FAM-PSKPAATRKRRWSAPESR-NH2) and cofactors 25 µM ATP (low ATP) or 1 mM (high ATP) and 5 mM MgCl 2 in a well of a 384 microtiter plate followed by incubation for additional 2 hours at 25ºC. At the end of the incubation, the reaction is quenched by the addition of a 20 mM EDTA-containing buffer. Substrate and phosphorylated product are separated electrophoretically using the microfluidic-based and 5 mM MgCl 2 in a well of 384 microtiter plate, followed by incubation for additional 70 min at 37°C. ATP to ADP conversion is measured using ADP-Glo kinase assay kit (V9102 Promega).
All assays in routine screening mode had 1 hour of pre-incubation with compounds before adding substrate and co-factors, unless specific mode of action studies were conducted to look at time-dependent potency shifts or jump dilution experiments.

Mass Spectrometry
MSK1 CTKD (residues 414-738), 10 M, was incubated (at various times, detailed in the table below) with compounds (100 M 2% DMSO) and assessed by native mass spectrometry (binding) and LCMS (covalent binding) at the defined time points.
For native mass spectrometry, the protein stock solution was passed through two pre-washed Zeba columns to exchange into Ammonium Acetate (50mM, pH 6.9). Protein concentration was measured using a Nanodrop Spectrophotometer and diluted to a final concentration of 20 M). For both the native and LCMS studies, protein and compound were mixed in 1:1 ratio (e.g 5ml:5ml) to give a sample with final concentrations of protein (10 M): compound (100 mM), 1% DMSO in ammonium acetate (50 mM). This sample was analysed (using the Nanomate-Exactive EMR Plus MS) immediately, or at the specified time points.
For LCMS (covalent binding), after the defined incubation period, samples were immediately analyzed (using the Acquity- reconstituted in 20 mL 0.1% Rapigest, vortexed three times and left at room temperature for 1 hour. 10 mL of 0.1 mg/mL trypsin was added and incubated at 37 o C overnight. Following digestion, a 5 mL sample was analyzed using a Acquity-Xevo G2 MS and a C4 reverse phase UPLC column and eluted over 30 minute run time. The peptide containing Cys440 was only found in the control, not the analyte, suggesting in the analyte that the peptide is completely bound to compound 1.

GSH incorporation experiments
Compounds (10 M) in phosphate buffer (pH 7.4) were incubated in the absence or presence of human liver microsomes (1mg protein/mL), NADPH (2.5 mM), MgCl 2 (5 mM) and GSH (5 mM), with stirring, at 37 o C for 30 minutes. The reaction was terminated with 1 volume of cold acetonitrile containing an internal standard (Midazolam) at 5 M. The plate was centrifuged at 1850 t/min for 10 minutes at 4 o C. 150 L of the supernatant was taken and diluted with 1 volume of water and analyzed by UPLC-QTOF.

S18
Kinase selectivity data The 11 kinases with Cys in the same position (according to reference 5) are highlighted, namely MAP3K1 (MEKK1), PLK1, PLK2, PLK3, RIPK5 (SgK496), RSK1 CTKD, RSK2 CTKD, RSK3 CTKD, RSK4 CTKD, MSK1 CTKD &MSK2 CTKD are highlighted in orange. Data listed as percentage target captured by immobilized ligand compared to control, ie the higher the percentage the lower the binding of the test ligand to the kinase.   Figure S1. TREEspot TM interaction map for compound 21 @ 10 M.
Image generated using TREEspot™ Software Tool and reprinted with permission from KINOMEscan®, a division of DiscoveRx Corporation, © DISCOVERX CORPORATION 2010. Figure S2. TREEspot TM interaction map for compound 22 @ 10 M.

S43
Image generated using TREEspot™ Software Tool and reprinted with permission from KINOMEscan®, a division of DiscoveRx Corporation, © DISCOVERX CORPORATION 2010

Discussion on the reactivity of Cys440 of the CTKD of MSK1
Our data show that the chloro-pyrimidines and -pyridines described herein irreversibly covalently label Cys440 in the CTKD of MSK1, the same residue labelled by the reversible covalent inhibitors 25 and 26, which have been described as reversible covalent inhibitors of RSK2 (at the equivalent Cys, ie Cys436). Interestingly, Cys440 of MSK1 was not identified as a reactive cysteine in a recent quantitative cysteine proteome profiling study. 1 This study was able to identify catalytic and non-catalytic cysteines, and did identify Cys579 (not Cys436, although this could be due to differences in numbering) as a reactive, non-catalytic cysteine in RSK2 (RPS6KA3, same family as MSK1) and Cys90 of p70S6K (RPS6KB1). There are multiple possible reasons why Cys440 in the CTKD of MSK1 may not have been identified. Firstly the study utilized human MCF7, MDA-MB-231 and Jurkat cells, which may not express MSK1, or express MSK1 at levels below detection, or the kinase may be in the inactive form which may preclude binding. Furthermore, the iodoacetamide probe used in this cysteinome study needs to be able to bind to the target protein with sufficient affinity and residence time to enable the reaction to take place, which requires both electronic and steric complementaraity. If the probe is able to bind to a protein, sufficient nucleophilicity of the nearby cysteine is also required. Cysteine reactivity can me predicted by various methods, including by calculated pKa, for which there are also various methods. Recent estimates of the pKa of Cys440 of MSK1, circa 15, place it a similar range to Cys432 of RSK1 and Cys436 of RSK2, circa 12-13, and other kinases which have been covalently targeted, such as EGFR and BTK etc. 2

Literature data on the half-life of MSK1
Several studies have estimated the half-life of MSK1 and related kinases, summarized in Table S1 below. 3,4