BI 1002494, a Novel Potent and Selective Oral Spleen Tyrosine Kinase Inhibitor, Displays Differential Potency in Human Basophils and B Cells S

Home , Boehringer Ingelheim, CAMK2D, Cerdulatinib, PIM2 (gene), PTK2, PTK2B

Supplemental material to this article can be found at: http://jpet.aspetjournals.org/content/suppl/2016/04/05/jpet.116.233155.DC1

1521-0103/357/3/554–561$25.00 http://dx.doi.org/10.1124/jpet.116.233155 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 357:554–561, June 2016 Copyright ª 2016 by The American Society for Pharmacology and Experimental Therapeutics

BI 1002494, a Novel Potent and Selective Oral Spleen Tyrosine Kinase Inhibitor, Displays Differential Potency in Human Basophils and B Cells s

David J. Lamb, Stefan Lutz Wollin, Andreas Schnapp, Daniel Bischoff, Klaus J. Erb, Thierry Bouyssou, Bernd Guilliard, Christine Strasser, Eva Wex, Sylvia Blum, Eva Thaler, Helga Nickel, Oliver Radmacher, Hannah Haas, Jennifer L. Swantek, Don Souza, Melissa Canfield, Della White, Mark Panzenbeck, Mohammed A. Kashem, Mary Sanville-

Ross, Takeshi Kono, Katherina Sewald, Armin Braun, Helena Obernolte, Olga Danov, Downloaded from Gerhard Schaenzle, Georg Rast, Gerd-Michael Maier, and Matthias Hoffmann Immunology and Respiratory Research (D.J.L., S.L.W., A.S., K.J.E., T.B., B.G., C.S., E.W., S.B., E.T., H.N., O.R., H.H.), Discovery Drug Support (D.B., G.S., G.R., G.-M.M.), and Medicinal Chemistry (M.H.), Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach an der Riß, Germany; Immunology and Respiratory Research (J.L.S., D.S., M.C., D.W., M.P.) and Small Molecule Discovery Research (M.A.K., M.S.-R.), Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut; Kobe Pharma Research Institute, Nippon Boehringer Ingelheim Co., Chuo-ku, Kobe City, Japan (T.K.); and Fraunhofer Institute for Toxicology jpet.aspetjournals.org and Experimental Medicine, Hannover, Germany (K.S., A.B., H.O., O.D.) Received February 25, 2016; accepted March 31, 2016

ABSTRACT BI 1002494 [(R)-4-{(R)-1-[7-(3,4,5-trimethoxy-phenyl)-[1,6] a similar species potency shift was not observed in B cells. The napthyridin-5-yloxy]-ethyl}pyrrolidin-2-one] is a novel, po- lower potency in rat basophils was confirmed in both ex vivo at ASPET Journals on September 30, 2021 tent, and selective spleen tyrosine kinase (SYK) inhibitor with inhibition of bronchoconstriction in precision-cut rat lung slices and sustained plasma exposure after oral administration in rats, in reversal of anaphylaxis-driven airway resistance in rats. The which qualifies this molecule as a good in vitro and in vivo tool different cellular potencies translated into different in vivo efficacy; compound. BI 1002494 exhibits higher potency in inhibiting high- full efficacy in a rat ovalbumin model (that contains an element of affinity IgE receptor–mediated mast cell and basophil degranula- mast cell dependence) was achieved with a trough plasma tion (IC50 5 115 nM) compared with B-cell receptor–mediated concentration of 340 nM, whereas full efficacy in a rat collagen- activation of B cells (IC50 5 810 nM). This may be explained by induced arthritis model (that contains an element of B-cell de- lower kinase potency when the physiologic ligand B-cell linker was pendence) was achieved with a trough plasma concentration of used, suggesting that SYK inhibitors may exhibit differential 1400 nM. Taken together, these data provide a platform from potency depending on the cell type and the respective signal which different estimates of human efficacious exposures can be transduction ligand. A 3-fold decrease in potency was observed in made according to the relevant cell type for the indication intended rat basophils (IC50 5 323 nM) compared with human basophils, but to be treated.

Introduction 2010), primarily expressed in hematopoietic tissues but also in a variety of different tissues (Yanagi et al., 2001). SYK Spleen tyrosine kinase (SYK) is a member of the Syk family propagates signal transduction for a number of immunore- of nonreceptor cytoplasmic tyrosine kinases (Riccaboni et al., ceptor tyrosine-based activation motif–dependent proinflammatory pathways, including Fc receptor, B-cell receptor This research was supported by Boehringer Ingelheim Pharma GmbH & Co. (BCR), and integrin signaling (Riccaboni et al., 2010). The KG and Boehringer Ingelheim Pharmaceuticals, Inc. dx.doi.org/10.1124/jpet.116.233155. proximal location of SYK in these multiple pathways confers a s This article has supplemental material available at jpet.aspetjournals.org. broad range of anti-inflammatory activity and makes SYK an

ABBREVIATIONS: BAY 61-3606, 2-[7-(3,4-dimethoxyphenyl)-imidazo[1,2-c]pyrimidin-5-ylamino]nicotinamide dihydrochloride; BCR, B-cell receptor; BI 1002494, (R)-4-{(R)-1-[7-(3,4,5-trimethoxy-phenyl)-[1,6]napthyridin-5-yloxy]-ethyl}pyrrolidin-2-one; BLNK, B-cell linker; BSA, bovine serum albumin; CIA, collagen-induced arthritis; DMSO, dimethylsulfoxide; DNP, dinitroprusside; Fc«R1, high-affinity IgE receptor; GS-9973, 2-[[(1R,2S)-2-aminocyclohexyl] amino]-4-[3-(triazol-2-yl)anilino]pyrimidine-5-carboxamide; hERG, human ether-a-go-go–related gene potassium channel; JAK, Janus kinase; OVA, ovalbumin; PAK, p21-activated kinase; PBS, phosphate-buffered saline; PCLS, precision-cut lung slice; PK, pharmacokinetics; PRT062070, 4-(cyclopropylamino)-2-[4-(4-ethylsulfonylpiperazin-1-yl)anilino]pyrimidine-5-carboxamide; PRT062607, 2-[[(1R,2S)-2-aminocyclohexyl]amino]-4-[3-(triazol- 2-yl)anilino]pyrimidine-5-carboxamide; R343, 5,7-dibromoquinolin-8-ol; R406, 6-[[5-fluoro-2-(3,4,5-trimethoxyanilino)pyrimidin-4-yl]amino]-2,2-dimethyl- 4H-pyrido[3,2-b][1,4]oxazin-3-one; R788, 6-[[5-fluoro-2-(3,4,5-trimethoxyanilino)pyrimidin-4-yl]amino]-2,2-dimethyl-3-oxopyrido[3,2-b][1,4]oxazin-4-yl] methyl dihydrogen phosphate; SYK, spleen tyrosine kinase; TAK-659, diethyl 2-dimethoxyphosphinothioylsulfanylbutanedioate.

554 SYK Inhibitor with Differential Cellular Pharmacology 555 attractive therapeutic target for a range of inflammatory of 10 nM on human basophils (Yamamoto et al., 2003). diseases. For example, the early and late allergic responses However, the poor rodent pharmacokinetics (PK) profile observed in asthma and other atopic diseases can be directly required very high doses to achieve an inhibitory effect in linked to the degranulation of mast cells that have been vivo (Yamamoto et al., 2003; Bhagwat, 2009). More recently, sensitized with allergen-specific IgE binding to the high- Merck published a very potent advancement on the more affinity IgE receptor (Fc«R1) (Allan et al., 2012). SYK is selective Portola compound structure (SYK kinase IC50 5 immediately downstream of Fc«R1, and inhibition of SYK has 60 pM, whole blood basophil IC50 5 58 nM) with which only been shown to abrogate these responses (Wex et al., 2011). zeta chain–associated protein kinase 70 in the 265-kinase B-cell activity has been linked to other inflammatory diseases, panel was inhibited within 100-fold of the SYK IC50 (Ellis such as rheumatoid arthritis, where they may present antigen et al., 2015). to T cells or secrete proinflammatory cytokines and rheuma- For respiratory diseases, the SYK inhibitor R343 (5,7- toid factor (Panayi, 2005), and in which the B-cell–depleting dibromoquinolin-8-ol) from Rigel Pharmaceuticals has been antibody rituximab has been shown to be clinically effective evaluated using the inhalation route in clinical trials in (Cohen et al., 2006). SYK is also downstream of BCR, and codevelopment with Pfizer, showing complete inhibition of inhibition of SYK has been shown to inhibit BCR-mediated both the early and late atopic phases after allergen challenge activation of B cells (Braselmann et al., 2006). in individuals with mild asthma (Allan et al., 2012). However, Because of this broad mechanism of action, several SYK in a small 42-kinase panel, 12 kinases (29% of the kinase inhibitors are currently in clinical development but all seem panel) were inhibited with a higher potency than SYK and 13 Downloaded from to be hampered by some liabilities. Fostamatinib [R406 (6- additional kinases (32% of the kinase panel) within 20% of [[5-fluoro-2-(3,4,5-trimethoxyanilino)pyrimidin-4-yl]amino]- SYK activity (Ramis et al., 2015), again suggesting that a large 2,2-dimethyl-4H-pyrido[3,2-b][1,4]oxazin-3-one); Rigel proportion of the kinome may be inhibited. Recently, Almirall Pharmaceuticals, South San Francisco, CA) has been evalu- reported that another inhaled SYK inhibitor (LAS189386, ated in clinical trials for rheumatoid arthritis in codevelopment 1-{2-[(1S,4S)-2,5-diazabicyclo[2.2.1]hept-2-yl]pyridin-4-yl}- with AstraZeneca; however, fostamatinib failed in one of the N-pyrazin-2-yl-1H-indazol-3-amine) shows preclinical activity jpet.aspetjournals.org two coprimary phase III efficacy endpoints (Genovese et al., in rodent models of ovalbumin (OVA)–induced early asthmatic 2014; Weinblatt et al., 2014) and is currently in development for response (Ramis et al., 2015), although this compound exhibits immune thrombocytopenia and IgA-mediated nephropathy. only limited selectivity for SYK, with 7 other kinases (17% of the Recent analysis suggests that the soluble prodrug of R406, kinase panel) being inhibited within 20% of SYK activity in a R788 ([6-[[5-fluoro-2-(3,4,5-trimethoxyanilino)pyrimidin-4-yl] short 42-kinase selectivity panel. amino]-2,2-dimethyl-3-oxopyrido[3,2-b][1,4]oxazin-4-yl]methyl We report here a novel potent and selective SYK inhibitor, dihydrogen phosphate)] (Braselmann et al., 2006), has a Kd BI 1002494 [(R)-4-{(R)-1-[7-(3,4,5-trimethoxy-phenyl)-[1,6] at ASPET Journals on September 30, 2021 within 10-fold of the Kd for SYK for 110 of 440 kinases napthyridin-5-yloxy]-ethyl}pyrrolidin-2-one], with sustained (including FMS-related tyrosine kinase 3, lymphocyte cell- plasma exposure in rodents after oral administration, which specific protein tyrosine kinase, Janus kinases JAK1 and exhibits enhanced potency in basophils/mast cells versus JAK3, and c-kit), suggesting that more than 20% of the kinome B cells and displays an excellent in vitro–ex vivo–in vivo may be inhibited by R406 within 10-fold of the concentration potency translation. required for SYK inhibition and may thus challenge its selectivity (Davis et al., 2011). The dual SYK/JAK inhibitor cerdulatinib (PRT062070 [4- Materials and Methods (cyclopropylamino)-2-[4-(4-ethylsulfonylpiperazin-1-yl)anilino] Animals. Adult, test-naïve, male brown Norway rats (48 BN/Crl, pyrimidine-5-carboxamide]; Portola Pharmaceuticals, South aged 15 to 16 weeks, weighing 190–250 g) were purchased from San Francisco, CA) is in a phase II trial for leukemia and Charles River Laboratories (Sulzfeld, Germany). Adult, test-naïve, – lymphoma but also inhibits 19 other tested kinases, with IC50 female Lewis rats (nulliparous and nonpregnant, aged 8 10 weeks, weighing 18–24 g) were purchased from Charles River Laboratories values less than 200 nM and within 100-fold of the SYK IC50 (Coffey et al., 2014). The more selective SYK inhibitor from (Willimantic, CT). Rat OVA and passive anaphylaxis animal experi- Portola, PRT062607 (2-[[(1R,2S)-2-aminocyclohexyl]amino]-4- mentation was conducted in accordance with German national guidelines and legal regulations and was approved by the Regierung- [3-(triazol-2-yl)anilino]pyrimidine-5-carboxamide), has also spräsidium Tübingen Ethical Committee in Germany (permit number been evaluated in a phase I trial, but no further develop- 13-025). ment has since been reported (Coffey et al., 2012). Despite Animals were housed under conventional and certified laboratory the increased selectivity over PRT062070, PRT062607 in- conditions with a regular 12-hour dark/light cycle at ambient temper- hibits FGR and MLK1 within 100-fold of the SYK IC50 (Coffey ature of 22°C 6 2°C and a relative air humidity of 60% 6 15%, and they et al., 2012). Recently, investigators started phase II trials of were kept in rooms maintained at constant temperature (22°C 6 2°C) entospletinib [GS-9973 (2-[[(1R,2S)-2-aminocyclohexyl]amino]-4- and humidity (60% 6 15%) under a 12-hour light/dark cycle. The [3-(triazol-2-yl)anilino]pyrimidine-5-carboxamide); Gilead Sci- animals were housed in groups of two in isolated ventilated cages and ences, Foster City, CA] for refractory hematologic conditions were allowed free access to water and standard food. All efforts were and refractory lymphoma. Only nonreceptor tyrosine-protein made to minimize suffering. Animals were acclimated for at least 2 weeks before use. Animals were constantly monitored during in- kinase 1 was inhibited within 10-fold of the SYK IC (Currie 50 duction and maintenance of anesthesia, and the level of narcotic was et al., 2014); however, 34 kinases were reported to be inhibited adjusted to maintain optimal depth of anesthesia. within 100-fold of the SYK IC50. Bayer published the de- Compound Preparation. BI 1002494 was synthesized as de- scription of an ATP-competitive selective SYK inhibitor, BAY scribed in patent WO2011092128 as example 35. For in vitro assays, 61-3606 (2-[7-(3,4-dimethoxyphenyl)-imidazo[1,2-c]pyrimidin- BI 1002494 was dissolved in dimethylsulfoxide (DMSO) and serial 5-ylamino]nicotinamide dihydrochloride), with an IC50 value dilutions were prepared in DMSO prior to final dilution in assay 556 Lamb et al. buffer. The final DMSO concentration was 1% in the kinase assays and closely performed as described previously (Wohlsen et al., 2001). 0.1% in the cellular assays. For animal experiments, BI 1002494 was PCLSs were prepared from rat lungs. Animals were euthanized with dissolved in 0.5% hydroxyethylcellulose containing 0.01% Tween 20 an intraperitoneal overdose of pentobarbital-Na (Merial GmbH, and was applied by oral gavage. Hallbergmoos, Germany). Extraction of lung tissue was performed SYK Enzyme Kinase Domain ATP Detection Assay. The directly postmortem to conserve vitality of the tissue. Lung lobes were catalytic activity of SYK was measured utilizing Kinase-Glo ATP cannulated and filled with prewarmed 1.5% low-melting agarose/ detection reagent (Promega, Madison, WI), which quantifies residual medium solution and cooled on ice. Tissue cores of 8-mm caliber were ATP remaining in solution after the kinase reaction. The assay was prepared and cut into approximately 300-mm–thick slices using a performed in a 384-well OptiPlate (PerkinElmer, Waltham, MA) microtome (Krumdieck tissue slicer; Alabama Research and Develop- containing 4 nM GST (Glutathione S-transferase)-Syk kinase domain, ment, Munford, AL; or Vibratome OTS-5000; Science Services GmbH, 0.15 mM ATP, and 100 mg/ml poly(glutamic acid/tyrosine 4:1) Munich, Germany) in Earle’s balanced salt solution. Sections were [poly(EY) 4:1] in assay buffer consisting of 25 mM HEPES, pH 7.5, incubated overnight with 1% serum of OVA-sensitized rats. Reference

25 mM MgCl2,5mMMnCl2, 50 mM KCl, 0.01% 3-[(3-cholamidopropyl) tissue was incubated with 1% serum of nonsensitized rats. Single dimethylammonio]-1-propanesulfonic acid, 100 mMNa3VO4, and 1% tissue sections were placed in cavities of petri dishes and fixed by a DMSO. The buffer additionally contained either 0.2% or 1% human nylon thread attached by a wire to avoid movement during measure- serum albumin. Test compound was preincubated with kinase for ment. Airways of PCLSs were imaged and digitized using an inverted 15 minutes at room temperature before starting the reaction with the microscope and a digital video camera. Camera control and image substrates. Positive control wells contained no test compound, and analysis were achieved by Axiovision software (Carl Zeiss, Jena,

negative control wells contained no kinase or test compound. The Germany). The airway area before the addition of allergen was defined Downloaded from kinase reaction mixture (15 ml total volume) was incubated for as 100%. Bronchoconstriction was expressed as the percentage of the 60 minutes at room temperature, followed by the addition of 10 ml initial airway area. Only slices with airways free of agarose, with Kinase-Glo reagent. After 15 minutes of incubation at room temper- beating cilia, and with an intact smooth muscle layer were used. Acute ature, the relative light unit signal was measured on an Envision allergic airway constriction was induced by the addition of 0.1% OVA. reader (Agilent Technologies, Santa Clara, CA). Images of bronchoconstriction were taken by video microscopy. Details of the full-length SYK assays can be found in the Supple- Images were recorded every 15 seconds for a 10-minute period. The

mental Methods. experiment was conducted after preincubation with BI 1002494 for jpet.aspetjournals.org Human and Rat Basophil Assays. Human whole blood collected 30 minutes at 37°C. Preincubation with serum of nonallergic rats was into heparin blood collection tubes was obtained from healthy used as a reference curve. Each measurement was performed with volunteers who had previously given their informed consent. All lung tissue from three rats. samples were anonymized prior to use. Two-hundred microliters of Rat Pulmonary Passive Anaphylaxis Model. Brown Norway heparinized whole blood was aliquoted in a 48-well plate with 10 mlBI rats were briefly anesthetized with isoflurane [3% to 4% (v/v)] and 1002494 and incubated at 37°C for 30 minutes. Forty microliters of 200 ml monoclonal anti–DNP IgE antibody [50 mg/ml in phosphate- hJL3 (stimulation buffer) was added, mixed, and incubated at 37°C in buffered saline (PBS)] was administered intratracheally. Each animal a water bath for 10 minutes. Two-hundred microliters of 60 ng/ml received an additional 10 ml antibody (1 mg/ml in PBS) intradermally at ASPET Journals on September 30, 2021 dinitroprusside (DNP)/bovine serum albumin (BSA) was added, mixed, into the right ear. Twenty-four hours later, animals were briefly and incubated at 37°C for 20 minutes. The reaction was stopped with anesthetized and 1 mg DNP-BSA diluted in 0.9% saline with 1% the addition of 2 ml phospho-Fix solution (prediluted 1:5 in water and Evans blue was injected into the tail vein. BI 1002494 was dosed prewarmed to 37°C) and incubated at 37°C in a water bath for orally 1 hour before challenge and the rats were euthanized by 10 minutes. Cells were immunostained with mouse anti–human IgE-PE an intraperitoneal injection of pentobarbital (160 mg/kg narcoren), antibody and mouse anti–human CD63- fluorescein isothiocyanate. 30 minutes after the DNP-BSA challenge. Airway compliance was IgE/CD631 cells were identified as basophils. measured using the SCIREQ flexiVent (Emka Technologies, Paris, Twenty-five microliters of anti–DNP IgE monoclonal antibody Spe7 France). Under terminal anesthesia, blood samples were drawn from was added to 8 Â 105 cells/ml rat basophil cell line RBL-2HS and the retro-orbital plexus into EDTA for analysis of BI 1002494 incubated for 30 minutes. Cells were washed once with Tyrode’s buffer concentrations. Animals were euthanized and the respiratory tract, and 10 ml BI 1002494 was added and incubated for an additional including the larynx, trachea, and lung, was removed. Ear tissue 30 minutes in the presence of 25% rat serum. Five microliters of biopsies were taken from both ears using a tissue punch (8-mm DNP-BSA (100 ng/ml final concentration) was added and incubated diameter). Evans blue was extracted by incubating the trachea for 60 minutes. Twenty-five microliters of substrate (2.4 mM together with the bronchial tissue in 2.5 ml formamide and the ear 4-metylumbelliferyl-2-acetamide-2-deoxy-b-D-glucopyranoside) was tissue in 0.5 ml formamide at 65°C overnight with gentle agitation added and incubated for 45 minutes. Fluorescence intensity of the (450 rpm). Twenty-four hours later, the absorbance in the supernatant hydrolyzed substrate was measured in a SpectraMax M5 microplate was measured at 620 nm using a spectrophotometer. reader (Molecular Devices, Sunnyvale, CA) using 350-nm and 460-nm Rat OVA Model. Fifteen milligrams of OVA was dissolved in excitation and emission filters, respectively. 10 ml PBS prior to being mixed with 5 ml alum. On days 0, 1, and 2, brown B-Cell CD69 Assay. Human whole blood collected into heparin Norway rats were immunized intraperitoneally with approximately blood collection tubes was obtained from healthy volunteers who had 1 ml/kg with PBS/alum (negative controls) or OVA/alum (positive previously given their informed consent. All samples were anonymized controls and treatment groups). Animals were returned to their home prior to use. Heparinized whole blood was mixed gently and 45 mlwas cage to recover. On days 5 and 6, rats were challenged with inhaled aliquoted on the bottom of a 96-well plate. Five microliters per well of BI saline [negative controls] or OVA [1% (w/v); positive controls and 1002494 was added and preincubated for 30 minutes at 37°C in a water treatment groups] at a flow rate of 10 l/min for 20 minutes. Animals bath. Then 12.5 mg/ml goat anti–human IgD was added and the plate remained in the inhalation chambers for an additional 1 minute, after was incubated for 4 hours in 37°C in a water bath. Cells were stained which the air was evacuated (60 l/min) and the animals were returned with mouse anti–human CD19-fluorescein isothiocyanate and mouse to their home cage. BI001002494 was dosed orally 1 hour before and anti–human CD69-phycoerythrin. CD191 B cells were identified by flow 8 hours after OVA challenge on days 5 and 6. On day 7, rats were cytometry and the percentage of surface expression of CD69 was measured. euthanized and the lungs were lavaged with Hanks’ balanced salt solution, Rat Precision-Cut Lung Slices. To understand whether the 0.6 mM EDTA, and 1% BSA. Total cell counts in bronchioalveolar lavage cellular potencies of BI 1002494 translated into more functional samples were performed in a Sysmex XT-1800i analyzer (Sysmex, readouts, the ex vivo PCLS platform was used to measure broncho- Lincolnshire, IL). Cytospins were prepared and stained with May- constriction. Preparation of PCLSs and subsequent experiments were Grünwald-Giemsa stain for differential cell counts. SYK Inhibitor with Differential Cellular Pharmacology 557

Rat Collagen-Induced Arthritis Model. In the rat collagen- TABLE 1 induced arthritis (CIA) model, Lewis rats received two divided Potency of BI 1002494 in SYK enzyme assays intradermal injections totaling 400 ml collagen emulsion (containing 400 mg type II collagen emulsified with Freund’s incomplete adjuvant) SYK Kinase Domain Full-Length SYK at the base of the tail. Seven days later, rats received a booster Assay Binding TK Substrate BLNK injection of 100 ml collagen emulsion (containing 100 mg type II 0.2% HSA 1% HSA Substrate (0.01% BSA) (0.05% BSA) (0.05% BSA) collagen emulsified with Freund’s incomplete adjuvant) at the base of the tail. Three days after the booster injection, rats with visible IC50,nM 1 1 34 36 220 arthritis symptoms were randomly enrolled into treatment groups. HSA, human serum albumin. Hind paw volume was assessed daily.

B cells was 7-fold lower than in basophils/mast cells, with an Results IC50 value of 810 nM (Fig. 2). An analysis of 171 additional compounds that included examples from a diverse set of Kinase and Cellular Potencies of BI 1002494. The compounds revealed only 21 compounds (12%) with better structure of BI 1002494 is shown in Fig. 1. To understand the potency on CD69 versus CD63 expression (Fig. 3A); across the potency of BI 1002494 against SYK, a number of kinase compound panel, the average difference between CD63 and assays were employed (or developed) using either the kinase CD69 IC50 was 6.9-fold (Fig. 3B), suggesting that the finding domain only (i.e., constitutively active) or the full-length Downloaded from with BI 1002494 indeed represents a biologic difference of kinase (i.e., not constitutively fully activated), together with SYK inhibition in mast cells and B cells. a number of different generic and/or physiologically relevant Unlike the cross-species potency differences we observed in substrates. The potency of BI 1002494 in the SYK kinase the basophils/mast cells, the mouse B-cell CD69 IC was domain activity assay in ATP detection format with poly(Glu- 50 933 nM, which agreed well with the human B-cell potency Tyr)(4:1) as a substrate was 1 nM in the presence of both 0.2% (Fig. 2; Table 2). Because of technical limitations, it was not and 1% human serum albumin (Table 1). In the full-length possible to generate a B-cell CD69 potency in rat B cells. jpet.aspetjournals.org SYK competition binding assay and full-length SYK catalytic Physicochemical Properties and Selectivity of BI activity assay using the KinEASE-TK peptide as a substrate, 1002494. Plasma protein binding varied between species, the potency was approximately 30 nM. This potency was with values ranging from 93.2% to 95.6% (Table 3). Hepatocyte 220 nM in the full-length SYK catalytic activity assay when stability was estimated between 36% and 68% of human liver B-cell linker (BLNK) protein was used as a substrate. blood flow, depending on the species, resulting in a sustained The potency of BI 1002494 was tested in a number of plasma exposure above the in vitro IC values from the twice- different cellular assays to understand the effect and potency 50 daily dosing profile in rat (Fig. 2). Cytotoxicity, human ether- at ASPET Journals on September 30, 2021 of the SYK inhibitor across different cell types that are a-go-go–related gene potassium channel (hERG) inhibition, implicated in different inflammatory/autoimmune diseases. good cell permeability, and aqueous solubility were all within Where possible, assays were performed in whole blood or in the acceptable range for an in vivo tool compound (Table 3). the presence of protein for better comparison of intracellular In the LanthaScreen, Z9-Lyte, and Adapta kinase profiling potencies. The IC of BI 1002494 for Fc«R1-mediated hista- 50 screens (Invitrogen/Thermo Fisher Scientific, Carlsbad, CA), mine release in human monocyte-derived mast cells was 7 nM BI 1002494 inhibited only 23 of the 283 available kinases in the presence of 0.1% albumin (Table 2), which translates greater than 50% at 1 mM test concentration (Supplemental into an IC value of 159 nM when adjusted for the proportion 50 Table 1). Furthermore, only three receptors were bound more of compound bound to protein in the assay (93.2%). This than 50% at 10 mM in the ExpresSProfile screen (Eurofins agreed well with the potency of 115 nM for Fc«R1-mediated Cerep SA, Celle L’Evescault, France) (Supplemental Table 2). CD63 surface expression on peripheral basophils in human Potency Translation of BI 1002494 to Ex Vivo and In whole blood. The IC in the rat basophil cell line RBL-2H3 50 Vivo Models. Finally, we sought to determine whether the was 323 nM in the presence of 25% rat serum, approximately different potencies generated in the kinase and cellular assays 3-fold less potent than that for human basophils. translated into differences in efficacy in the ex vivo PCLS BI 1002494 also inhibited the activation of B cells (Table 2). platform and across a range of in vivo models. However, the potency of BI 1002494 in blocking CD69 Bronchoconstriction of passively sensitized and challenged expression after anti-IgD (BCR) stimulation of peripheral rat PCLSs was inhibited by BI 1002494 with an IC50 value of 70 nM (Fig. 4). These findings were obtained in the presence of 1% serum. The shift in potency in the RBL-2H3 assay from 1% serum to 25% serum was approximately 6-fold (data not shown). Taking into account this serum shift, this IC50 value agrees well with the rat basophil IC50 of 323 nM. Airway resistance (Fig. 4) was also measured in a pulmonary passive sensitization and challenge rat model with IC50 values of 735 nM, which also agrees well with both basophil and PCLS potencies. BI 1002494 dose-dependently inhibited OVA-induced bronchioalveolar lavage eosinophils in brown Norway rats, with twice-daily 30 mg/kg providing 91% inhibition (Fig. 5A). An exposure simulation of twice-daily 30 mg/kg in brown Fig. 1. Structure of BI 1002494. Norway rats resulted in a trough plasma concentration of 558 Lamb et al.

TABLE 2 Cellular potencies of BI 1002494 Values are given as means 6 S.E.M. unless specified otherwise.

Assay Human Mast Cell Human Basophil Rat RBL-2H3 Human B-Cell Mouse B-Cell Format 0.1% HSA Whole blood 25% serum Whole blood Whole blood Number 3 263 8 35 4 IC50,nM 76 31156 18 323 6 146 810 6 146 933 6 44 PPB-adjusted IC50, nM 159 n/a n/a n/a n/a

HSA, human serum albumin; n/a, not applicable; PPB, plasma protein binding.

340 nM (Fig. 6), which provided 24-hour coverage of the rat than fostamatinib (Braselmann et al., 2006), cerdulatinib basophil IC50 adjusted for plasma protein binding (323 nM). (Coffey et al., 2014), R343 (Singh et al., 2012), and LAS189386 BI 1002494 was tested in the Lewis CIA rat (Fig. 5B). A dose (Ramis et al., 2015) and has better rodent PK compared with of 10 mg/kg provided full efficacy; however, despite the rela- BAY 61-3606 (Bhagwat, 2009). The selectivity profile of BI tively lower dose, an exposure simulation of this dose in Lewis 1002494 is broadly similar to that of PRT062607 (Coffey et al., rats resulted in a higher trough plasma concentration of 2012); however, unlike PRT062607, BI 1002494 demonstrates 1400 nM (Fig. 6), a 40% increase in the plasma concentration 10-fold greater selectivity over p21-activated kinase PAK4 Downloaded from area under the curve (data not shown), and 24-hour coverage (PRT062607 PAK4 IC50 5 24 nM; BI 1002494 PAK4 IC50 5 of the human/mouse B-cell IC50 adjusted for plasma protein 189 nM), which may inhibit cell proliferation in tissues with binding (810/938 nM respectively), which was significantly high cellular turnover (e.g., bone marrow). The recently higher than the rat basophil IC50. Unfortunately, it was not published data on entospletinib (GS-9973) suggest that the technically possible to generate a rat B-cell IC50; nevertheless, selectivity profile is similar to that of BI 1002494 but is 3- to jpet.aspetjournals.org there was good agreement between potency values in both 10-fold less potent on SYK (IC50 5 7.7 nM versus 0.8 nM) and humans and mice (Table 2). in the basophil CD63 assay (IC50 5 387 nM versus 115 nM) (Currie et al., 2014). We observed that BI 1002494 was more potent in human Discussion basophils and mast cells, as measured by CD63 surface expression and histamine release, respectively, compared We describe here a novel, potent, and selective SYK in- with B-cell activation, as measured by CD69 expression. We hibitor with a good rodent PK profile, low cytotoxicity, hERG initially hypothesized that this finding was an artifact of at ASPET Journals on September 30, 2021 inhibition, and good cell permeability. When designing, variability in the assays. However, an analysis of a panel of interpreting, and reporting preclinical studies, it is important 171 SYK inhibitors (which included examples from a diverse to fully understand the cellular potencies, selectivity, and PK set of compounds) revealed 147 compounds (86%) with better of the tool compound selected, as recently stated by a cross- potency in mast cells (basophils) versus B cells. Furthermore, academic/pharmaceutical working group (Arrowsmith et al., the average difference between the potencies was 6.9-fold, 2015). BI 1002494 fulfills these criteria; it is more selective which agrees very well with the CD63/CD69 ratio of 7.0 for BI 1002494, suggesting that this discrepancy between basophil and B-cell potencies is real. The number of repetitions for CD63 (n 5 317) and CD69 (n 5 21) also suggests that these differences are not a result of assay variability. Both fostamatinib (Braselmann et al., 2006) and LAS189386 (Ramis et al., 2015) are reported to have potency values for B cells equal to or even better than those for basophils, which may be a consequence of their lower selectivity profile. The potency of BI 1002494 was initially determined using a kinase domain–only fragment of SYK and the poly(EY) 4:1 peptide substrate, a universal substrate for protein tyrosine kinases. However, this potency was lowered when the kinase domain was replaced with the full-length kinase and the potency determined by competitive binding with an ATP- competitive kinase inhibitor scaffold (kinase tracer 236). The recently published crystal structure of the full-length SYK Fig. 2. Potency of BI 1002494 in basophils and B cells. In all experiments, kinase, combined with some elegant mutation and binding cells or whole blood were preincubated with BI 1002494 for 30 minutes. studies, reveals that contrary to the kinase domain (in which For human basophils (solid black circles), whole blood was sensitized with the ATP binding site is active), the full-length kinase ATP anti–DNP IgE and challenged with DNP. Basophil degranulation was determined by the surface appearance of CD63 (n = 263). For rat basophils binding site exists in an autoinhibited conformation (Grädler (solid gray squares), RBL-H2S cells in the presence of 25% rat serum were et al., 2013). An activation loop that enables the phosphorylated sensitized with anti–DNP IgE and challenged with DNP. Basophil de- immunoreceptor tyrosine-based activation motif domain to b granulation was determined by the release of -hexosamidase into the bind to the SH2 domain facilitates activation of the site. This medium (n = 8). For human B cells (open black circles; n = 35) and mouse B cells (open gray triangles; n = 4), whole blood was stimulated with anti-IgD may explain the decrease in potency of BI 1002494 from an and B-cell activation was determined by the surface appearance of CD69. assay containing the “active” ATP binding site kinase domain to SYK Inhibitor with Differential Cellular Pharmacology 559

Fig. 3. Differential potency of a panel of SYK inhibitors in basophils and B cells. (A) Individual CD63 and CD69 IC50 values from a panel of 171 SYK inhibitors. (B) Mean IC50 values 6 S.E.M. for BI 1002494 in the basophil CD63 and B-cell CD69 assays. Downloaded from

an assay containing the full-length “autoinhibited” ATP binding exchanged for the physiologic substrate for SYK in B cells,

site. The activation loop contains at least two phosphorylation BLNK (Ishiai et al., 1999). Although it is difficult to draw firm jpet.aspetjournals.org sites: Tyr348 and Tyr352. It is not known in a cellular setting conclusions without the comparison with the physiologic whether the full-length SYK exists in the autoinhibited, active, substrate for SYK in basophils/mast cells, LAT (Linker of or partially active conformation. We can assume that the potency activated T-cells), it provides a plausible explanation for the is probably lower in cells compared with purified kinase protein differences between the basophil/mast cell and B-cell potency. because of the requirement for the compound to partition across Consistent with these findings, Gilead also reported that GS- the plasma membrane. Given that the human mast cell IC50 9973 displays an offset in potency between SYK (IC50 5 5 for BI 1002494 in the presence of only low levels of albumin was 7.7 nM) and SYK with BLNK as the substrate (IC50 26 nM) at ASPET Journals on September 30, 2021 7 nM, which is closer to the potency in the kinase domain (Currie et al., 2014). (activated kinase form) assay (IC50 5 1 nM) compared with that of the full-length kinase assay (IC50 5 34 nM), this suggests that the full-length kinase may be fully or at least partially activated in the cell. Next we asked what mechanism(s) could account for the discrepancy between basophil/mast cell and B-cell potency of BI 1002494. In the full-length kinase assay, the potency of BI 1002494 was similar in assays measuring displacement of a tracer (IC50 5 34 nM) and phosphorylation of a tyrosine kinase peptide substrate (homogeneous time resolved fluorescence KinEASE-TK substrate, 36 nM). However, there was a marked decrease (IC50 5 220 nM) when this substrate was

TABLE 3 Summary of BI 1002494 properties

Property Human Rat Mouse Value Fig. 4. Efficacy of BI 1002494 in rat basophils, rat PCLSs, and rat Plasma protein binding, % 95.6 93.2 95.1 pulmonary bronchoconstriction. Translation of BI 1002494 potency Microsomal stability, %QH 61.8 55.8 67.5 between rat basophils (solid black squares), bronchoconstriction in rat Hepatocyte stability, %QH 55 40 n/d PCLSs (open black circles), and airway resistance in a rat model of Bioavailability, % 73 pulmonary passive sensitization and challenge (solid gray circles). Rat hERG blockade, mM .10 RBL-2HS cells were preincubated with BI 1002494 in the absence of Cytotoxicity (U937 cells), mM25serum, and the Fc«R1 was loaded with anti–DNP IgE and stimulated with P-glycoprotein inhibition, mM6DNP. Degranulation was measured by release of b-hexosamidase into the Aqueous solubility (HCl salt), mg/ml medium (n = 4). PCLSs were preincubated with BI 1002494 and passively pH 1 0.9 sensitized with serum from OVA-sensitized rats. Sections were stimulated pH 3 0.9 with OVA and bronchoconstriction was measured by video microscopy. pH 5 0.9 Experiments were performed in duplicate with material from three rats. pH 7.4 0.5 Finally, pulmonary mast cells in rats were passively sensitized by pKa3.2intratracheal administration of anti–DNP IgE. Twenty-four hours later, logP 2.5 rats were challenged with DNP and airway resistance was measured via a flexiVent. Each data point represents four rats. Data are adjusted for the %QH, percentage liver blood flow. protein unbound fraction of compound. 560 Lamb et al.

Fig. 6. Simulated plasma exposure of BI 1002494 in brown Norway and Lewis rats after twice-daily oral dosing. Plasma exposures for twice-daily dosing (at 0 hours and 8 hours) were simulated for brown Norway rats (black squares) and Lewis rats (gray circles) from 24-hour plasma exposure profiles after a single 10-mg/kg oral administration in three Downloaded from rats. Trough plasma levels of 1400 nM (Lewis rat, 10 mg/kg,) and 340 nM (brown Norway rat, 30 mg/kg) were calculated.

efficacious exposures is somewhat easier when there is no observable species potency shift between human and mouse B cells. Similarly, there was no difference between mouse jpet.aspetjournals.org basophil and mouse B-cell potencies for BI 1002494 (data not shown). However, if the basophil/mast cell axis is targeted in diseases such as asthma (R-343, LAS189386), then a lower dose that provides exposure covering the potency in these cell types may be sufficient. Whether coverage of B-cell activation

is required for better efficacy in asthma is an open question. at ASPET Journals on September 30, 2021 However, there is no direct evidence that B-cell depletion and/ or loss of function would be beneficial in asthma, and loss of regulatory B cells exacerbates allergic airway inflammation in mice (Amu et al., 2010). Furthermore, compromising B-cell function may even be deleterious in patients with asthma who Fig. 5. Efficacy of BI 1002494 in the brown Norway rat OVA model and rely on vaccine protection from respiratory viruses (e.g., Lewis rat CIA model. (A) Brown Norway rats were immunized in- influenza) to prevent exacerbation (Wells et al., 2015), and traperitoneally with OVA plus alum on days 1 and 2. On days 5 and 6, rats there is evidence that B-cell depletion impairs response to were challenged with inhaled OVA [1% (w/v)]. BI 1002494 was dosed the influenza vaccine (Bedognetti et al., 2011). However, orally 1 hour before and 8 hours after OVA challenge on days 5 and 6. On day 7, rats were euthanized, the lungs were lavaged, and the number of fostamatinib demonstrated no effect on IgG or IgM responses eosinophils was counted. Each data point is the mean 6 S.E.M. from eight after keyhole limpet hemocyanin challenge, nor did it affect rats. (B) Lewis rats received two intradermal injections of collagen bacterial or viral clearance in listeria, streptococcal, or influenza emulsion in Freund’s incomplete adjuvant at the base of the tail. Seven days later, rats received a booster injection of collagen in Freund’s host-resistance mouse models (Zhu et al., 2007), suggesting that incomplete adjuvant. Three days after the booster injection, rats with B-cell responses may not be impaired, even at higher exposures visible arthritis symptoms were randomly enrolled into treatment groups that cover the CD69 activation assay. The choice of indication- of vehicle (solid black circles), 3 mg/kg BI 1002494 (open black circles), 10 mg/kg BI 1002494 (solid black squares), or 30 mg/kg BI 1002494 (open relevant cellular assay therefore has an important influence on black squares) and treated twice daily for 14 days. Hind paw volume was the estimated human efficacious exposure. assessed daily. Each data point is the mean 6 S.E.M. from eight rats. Acute asthmatic attacks in patients, in the in vivo rat , , ***P 0.005; ****P 0.001. anaphylaxis model, and, to a lesser extent, in the OVA model rely on mast cell degranulation. However, assessment of The apparent lower potency in human B cells may have a target engagement by a pharmacological intervention in consequence for selection of a human dose, depending on the tissue cells is difficult. It is therefore common to use a indication in which the SYK inhibitor is positioned, which was peripheral cell as a surrogate; in this case (and in the supported by the higher exposures required in the rat CIA literature; Braselmann et al., 2006), the basophil was selected. model for full efficacy. If the B cell is the desired target Our work increases the confidence that the potency of an SYK in diseases such as rheumatoid arthritis (fostamatinib, inhibitor in the peripheral basophil is a good marker for PRT062607, HMPL-523, GS-9876) or B-cell lymphoma potency in the tissue mast cell. We compared the potency of (PRT062070, GS-9973, TAK-659 [diethyl 2-dimethoxyphos- BI 1002494 in the in vitro rat basophil degranulation assay phinothioylsulfanylbutanedioate]), then a dose resulting in a (IC50 5 323 nM), ex vivo bronchoconstriction in rat PCLSs higher exposure that covers the potency in the B-cell CD69 (IC50 5 70 nM, which was generated in the presence of 1% assay is likely required. In this scenario, estimating human serum and would result in an IC50 value of approximately SYK Inhibitor with Differential Cellular Pharmacology 561

400 nM if adjusted for the approximate 6-fold serum shift Braselmann S, Taylor V, Zhao H, Wang S, Sylvain C, Baluom M, Qu K, Herlaar E, Lau A, and Young C, et al. (2006) R406, an orally available spleen tyrosine kinase observed in the RBL-2H3 assay), and in vivo reversal of airway inhibitor blocks Fc receptor signaling and reduces immune complex-mediated in- – resistance in a rat pulmonary anaphylaxis model (IC50 5 flammation. J Pharmacol Exp Ther 319:998 1008. Coffey G, Betz A, DeGuzman F, Pak Y, Inagaki M, Baker DC, Hollenbach SJ, Pandey 735 nM). The good agreement between potencies generated in A, and Sinha U (2014) The novel kinase inhibitor PRT062070 (cerdulatinib) the cellular assay (as a surrogate for the peripheral basophil) demonstrates efficacy in models of autoimmunity and B-cell cancer. J Pharmacol Exp Ther 351:538–548. and the tissue mast cells both in tissue preparations and in Coffey G, DeGuzman F, Inagaki M, Pak Y, Delaney SM, Ives D, Betz A, Jia ZJ, an in vivo model suggest that in this case, the basophil is a Pandey A, and Baker D, et al. (2012) Specific inhibition of spleen tyrosine kinase suppresses leukocyte immune function and inflammation in animal models of reasonable surrogate for estimation of compound potency in rheumatoid arthritis. J Pharmacol Exp Ther 340:350–359. tissue mast cells. Cohen SB, Emery P, Greenwald MW, Dougados M, Furie RA, Genovese MC, Key- Although the fully efficacious dose in the rat CIA model was stone EC, Loveless JE, Burmester GR, and Cravets MW, et al.; REFLEX Trial Group (2006) Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis lower than that required for the rat OVA model, the plasma factor therapy: results of a multicenter, randomized, double-blind, placebo- exposures differed considerably between rat strains. Conse- controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks. Arthritis Rheum 54:2793–2806. quently, higher exposures were required for efficacy in the CIA Currie KS, Kropf JE, Lee T, Blomgren P, Xu J, Zhao Z, Gallion S, Whitney JA, Maclin model compared with the OVA model. It is therefore impor- D, and Lansdon EB, et al. (2014) Discovery of GS-9973, a selective and orally efficacious inhibitor of spleen tyrosine kinase. J Med Chem 57:3856–3873. tant to consider both dose and exposure when assessing the Davis MI, Hunt JP, Herrgard S, Ciceri P, Wodicka LM, Pallares G, Hocker M, Treiber potency of a tool compound in vivo. DK, and Zarrinkar PP (2011) Comprehensive analysis of kinase inhibitor selectivity. Nat Biotechnol 29:1046–1051. In summary, BI 1002494 is a novel, potent, and selective Ellis JM, Altman MD, Bass A, Butcher JW, Byford AJ, Donofrio A, Galloway S, SYK inhibitor with a good rodent PK profile, low cytotoxicity, Haidle AM, Jewell J, and Kelly N, et al. (2015) Overcoming mutagenicity and ion Downloaded from channel activity: optimization of selective spleen tyrosine kinase inhibitors. J Med and hERG and P-glycoprotein inhibition. Potency in human Chem 58:1929–1939. mast cells and basophils was consistently better than in Genovese MC, van der Heijde DM, Keystone EC, Spindler AJ, Benhamou C, Kavanaugh A, Fudman E, Lampl K, O’Brien C, and Duffield EL, et al. (2014) A human B cells, which is perhaps explained by a lower potency phase III, multicenter, randomized, double-blind, placebo-controlled, parallel- when the physiologic B-cell substrate BLNK was substituted group study of 2 dosing regimens of fostamatinib in patients with rheumatoid arthritis with an inadequate response to a tumor necrosis factor-a antagonist. in the kinase assay. A potency shift was observed between JRheumatol41:2120–2128. humans and rats, but these potencies translated well between Grädler U, Schwarz D, Dresing V, Musil D, Bomke J, Frech M, Greiner H, Jäkel S, jpet.aspetjournals.org in vitro and ex vivo test systems and into efficacious exposures Rysiok T, and Müller-Pompalla D, et al. (2013) Structural and biophysical char- acterization of the Syk activation switch. J Mol Biol 425:309–333. in the respective species in vivo models. Taken together, these Ishiai M, Kurosaki M, Pappu R, Okawa K, Ronko I, Fu C, Shibata M, Iwamatsu A, data provide a robust platform from which to estimate human Chan AC, and Kurosaki T (1999) BLNK required for coupling Syk to PLC gamma 2 and Rac1-JNK in B cells. Immunity 10:117–125. efficacious exposures according to the relevant cell type for the Panayi GS (2005) B cells: a fundamental role in the pathogenesis of rheumatoid indication intended to be treated. arthritis? Rheumatology (Oxford) 44 (Suppl 2):ii3–ii7. Ramis I, Otal R, Carreño C, Domènech A, Eichhorn P, Orellana A, Maldonado M, De Alba J, Prats N, and Fernández JC, et al. (2015) A novel inhaled Syk inhibitor

Authorship Contributions blocks mast cell degranulation and early asthmatic response. Pharmacol Res 99: at ASPET Journals on September 30, 2021 Participated in research design: Lamb, Wollin, Schnapp, Bischoff, 116–124. Riccaboni M, Bianchi I, and Petrillo P (2010) Spleen tyrosine kinases: biology, Erb, Blum, Thaler, Radmacher, Haas, Swantek, Souza, Kashem, therapeutic targets and drugs. Drug Discov Today 15:517–530. Kono, Braun, Schaenzle, Rast, Maier. Singh R, Masuda ES, and Payan DG (2012) Discovery and development of spleen Conducted experiments: Guilliard, Strasser, Wex, Blum, Thaler, tyrosine kinase (SYK) inhibitors. J Med Chem 55:3614–3643. Weinblatt ME, Genovese MC, Ho M, Hollis S, Rosiak-Jedrychowicz K, Kavanaugh A, Nickel, Radmacher, Haas, Souza, Canfield, White, Panzenbeck, Millson DS, Leon G, Wang M, and van der Heijde D (2014) Effects of fostamatinib, Sanville-Ross, Kono, Obernolte, Danov. an oral spleen tyrosine kinase inhibitor, in rheumatoid arthritis patients with an Contributed new reagents or analytic tools: Hoffmann. inadequate response to methotrexate: results from a phase III, multicenter, randomized, double-blind, placebo-controlled, parallel-group study. Arthritis Rheu- Performed data analysis: Lamb, Wollin, Bouyssou, Guilliard, matol 66:3255–3264. Strasser, Wex, Blum, Thaler, Radmacher, Haas, Swantek, White, Wells RE, Garb J, Fitzgerald J, Kleppel R, and Rothberg MB (2015) Factors associ- Panzenbeck, Sanville-Ross, Schaenzle, Rast, Maier. ated with emergency department visits in asthma exacerbation. South Med J 108: 276–280. Wrote or contributed to the writing of the manuscript: Lamb, Wex E, Bouyssou T, Duechs MJ, Erb KJ, Gantner F, Sanderson MP, Schnapp A, Wollin, Schnapp, Bischoff, Bouyssou, Haas, Swantek, Kashem, Stierstorfer BE, and Wollin L (2011) Induced Syk deletion leads to suppressed Sewald, Braun, Hoffmann. allergic responses but has no effect on neutrophil or monocyte migration in vivo. Eur J Immunol 41:3208–3218. References Wohlsen A, Uhlig S, and Martin C (2001) Immediate allergic response in small airways. Am J Respir Crit Care Med 163:1462–1469. Allan R, Jones I, Bartley M, Willbraham D, and Singh D (2012) A randomised, Yamamoto N, Takeshita K, Shichijo M, Kokubo T, Sato M, Nakashima K, Ishimori M, placebo controlled trial of the effect of an inhaled SYK inhibitor on allergen induced Nagai H, Li YF, and Yura T, et al. (2003) The orally available spleen tyrosine airway response in mild asthma (Abstract). Am J Respir Crit Care Med 185:A2774. kinase inhibitor 2-[7-(3,4-dimethoxyphenyl)-imidazo[1,2-c]pyrimidin-5-ylamino] Amu S, Saunders SP, Kronenberg M, Mangan NE, Atzberger A, and Fallon PG (2010) nicotinamide dihydrochloride (BAY 61-3606) blocks antigen-induced airway in- Regulatory B cells prevent and reverse allergic airway inflammation via FoxP3- flammation in rodents. J Pharmacol Exp Ther 306:1174–1181. positive T regulatory cells in a murine model. J Allergy Clin Immunol 125: Yanagi S, Inatome R, Takano T, and Yamamura H (2001) Syk expression and novel – 1114 1124.e8. function in a wide variety of tissues. Biochem Biophys Res Commun 288:495–498. Arrowsmith CH, Audia JE, Austin C, Baell J, Bennett J, Blagg J, Bountra C, Zhu Y, Herlaar E, Masuda ES, Burleson GR, Nelson AJ, Grossbard EB, and Clemens Brennan PE, Brown PJ, and Bunnage ME, et al. (2015) The promise and peril of GR (2007) Immunotoxicity assessment for the novel Spleen tyrosine kinase in- – chemical probes. Nat Chem Biol 11:536 541. hibitor R406. Toxicol Appl Pharmacol 221:268–277. Bedognetti D, Zoppoli G, Massucco C, Zanardi E, Zupo S, Bruzzone A, Sertoli MR, Balleari E, Racchi O, and Messina M, et al. (2011) Impaired response to influenza vaccine associated with persistent memory B cell depletion in non-Hodgkin’s Address correspondence to: David J. Lamb, Immunology and Respiratory lymphoma patients treated with rituximab-containing regimens. J Immunol 186: Research, Boehringer Ingelheim Pharma GmbH & Co. KG, Birkendorfer 6044–6055. Straße 65, 88397 Biberach an der Riß, Germany. E-mail: david.lamb@ Bhagwat SS (2009) Kinase inhibitors for the treatment of inflammatory and auto- boehringer-ingelheim.com immune disorders. Purinergic Signal 5:107–115. JPET #233155

BI 1002494, a novel potent and selective oral SYK inhibitor displays differential potency in human basophils and B-cells

Lamb DJ, Wollin SL, Schnapp A, Bischoff D, Erb KJ, Bouyssou T, Wex E, Blum S, Thaler E, Nickel H, Radmacher O, Haas H, Swantek JL, Souza D, Canfield M, White D, Panzenbeck M, Kashem MA, Sanville-Ross M, Kono T, Sewald K, Braun A, Obernolte H, Danov O, Schaenzle G, Rast G, Maier GM, Hoffmann M.

Journal of Pharmacology and Experimental Therapeutics (JPET)

METHODS

SYK enzyme - full length LanthaScreen competition binding assay

LanthaScreen competition binding assay (Life Technologies) of GST-Syk was based on the binding and displacement of Alexa Fluor647-labeled kinase tracer to the ATP-binding site of the kinase with TR- FRET detection using a europium-labeled anti-GST antibody. The assay was performed in 384-well, low volume NBS black plates (Corning) containing 2 nM GST-Syk full-length (Life Technologies), 1 nM kinase Tracer236 and 2 nM Eu3+-anti-GST antibody in assay buffer consisting of 50 mM HEPES, pH

7.5, 10 mM MgCl2, and 0.01% Brij 35, 0.01% BSA, 1 mM DTT, and 1% DMSO. Positive control wells contained no test compound, and negative control wells contained no kinase or test compound. The kinase reaction mixture (15 µL total volume) was incubated for 60 min at room temperature, and then the TR-FRET signal was read on a VIEWLUX using 340 nm excitation filter and 671 nm emission filter.

SYK enzyme full length HTRF assay usingTK peptide substrate

The catalytic activity of full-length Syk was measured using HTRF KinEASE-TK assay (CisBio) which uses a unique biotinylated peptide substrate (TK-peptide) containing a single tyrosine phosphorylation site, which when phosphorylated is recognized by a monoclonal anti-phospho- tyrosine antibody labeled with Eu3+cryptate. The assay was performed in 384-well Optiplate containing 3 nM GST-Syk full-length (Life Technologies), 500 nM biotin-TK pepride and 5 µM ATP in assay buffer consisting of 250 mM HEPES, pH 7.0, 0.05% BSA, 0.1% NaN3, 10 mM MgCl2, 100 µM

Na3VO4, 1 mM DTT and 1% DMSO. Test compound was pre-incubated with kinase for 15 min at room temperature before starting the reaction with the substrates. Positive control wells contained no test compound, and negative control wells contained no kinase or test compound. The kinase reaction mixture (10 µL total volume) was incubated for 60 min at 37C followed by addition of 10 µL detection mix containing a 1:1 mixture of TK antibody Cryptate and 250 nM Streptavidin XL-665. After additional 60 minutes of incubation at room temperature, TR-FRET signal was read on an EnVision (Excitation: 340 nm; Emissions:615 and 665 nm).

SYK enzyme full length HTRF assay using BLNK protein substrate

The catalytic activity of full-length SYK was measured utilizing His-tagged BLNK protein substrate (GenWay) with TR-FRET detection of phosphorylated BLNK with Eu3+-labeled anti-His antibody (Life Technologies) and custom d2-labeled anti-phospho BLNK (Try84) antibody (CisBio). The assay was performed in 384-well Optiplate containing 3 nM GST-Syk full-length (Life Technologies), 25 nM His- JPET #233155

BLNK protein substrate, and 5 µM ATP in assay buffer consisting of 250 mM HEPES, pH 7.0, 0.05%

BSA, 0.1% NaN3, 10 mM MgCl2, 100 µM Na3VO4, 1 mM DTT and 1% DMSO. Test compound was preincubated with kinase for 15 min at room temperature before starting the reaction with the substrates. Positive control wells contained no test compound, and negative control wells contained no kinase or test compound. The kinase reaction mixture (10 µL total volume) was incubated for 60 min at 37C followed by addition of 10 µL detection mix containing a 1:1 mixture of 20 ng/well d2- labeled anti-phospho BLNK (Try84) and 4 nM Eu3+-labeled anti-His. After additional 60 minutes of incubation at room temperature, TR-FRET signal was read on an Envision (Excitation: 340 nm; Emissions:615 and 665 nm).

Histamine release from human peripheral CD34+ progenitor cell-derived

Human peripheral blood derived CD34+ cells (2G-101C) were obtained from LONZA. Into each well of a 24-well plate are placed 5 x 104 cells in METHOCULT SFBIT H4236 medium containing 50 ng/mL SCF ) 50ng/mL IL-6 and 25 ng/mL IL-3 and cultured for 28 days in a CO2 incubator at 37°C / 5%CO2. On day 28, 1 mL of serum free mast cell medium supplemented with cytokines (SCF (100 ng/mL), IL-6 (50 ng/mL), Albumin solution (0.1%), Insulin-Transferrin-Selenium supplement (1% v/v), Penicillin- Streptomycin solution (1% v/v) and 2-Mercaptoethanol (55 μM) was added to each well. After day 84 mature mast cells were identified by FACS analysis (cKit-High, Fc epsilon RI alpha-High) and by Western blotting (tryptase-positive). 100,000 cells/mL are sensitized with 1 μg/mL of human IgE, overnight. To stimulate the cells 3 μg/mL of anti-IgE antibody was added to each well and incubated for 30 minutes at 37°C in a water-bath. To stop the mast cell degranulation the plates were quickly cooled down in ice water and centrifuged for 5 min at 1000 rpm and 4°C. The supernatants were collected and the histamine concentration of each sample is measured by a histamine EIA kit (Bio- Rad), according to the manufacturer's protocol.

JPET #233155

DATA

Table 1: % inhibition of kinases in Invitrogen panel of BI 1002494 at 1µM (23 out of 239 non- mutant kinases)

% % % Kinase Kinase Kinase inhib. inhib. inhib. SYK 107 JAK2 11 CDC42 BPA (MRCKA) 0 MAPK14 (p38 alpha) MAP3K9 (MLK1) 103 BLK 10 0 Direct STK22D (TSSK1) 99 ABL1 10 STK3 (MST2) 0 FER 98 CDK5/p25 10 IKBKB (IKK beta) 0 LRRK2 97 BMX 10 TEK (Tie2) 0 FES (FPS) 95 MAP3K8 (COT) 10 SRMS (Srm) 0 MAP4K2 (GCK) 90 KDR (VEGFR2) 10 AKT3 (PKB gamma) -1 RPS6KA2 (RSK3) 88 ABL1 T315I 9 CLK3 -1 PTK6 (Brk) 87 CDK5/p35 9 CDK9/cyclin T1 -1 STK22B (TSSK2) 86 FLT1 (VEGFR1) 9 PLK2 -1 CAMK2D (CaMKII RPS6KA6 (RSK4) 85 MELK 8 -1 delta) RPS6KA3 (RSK2) 85 LYN B 8 MARK1 (MARK) -1 CSF1R (FMS) 85 MET (cMet) 8 FGFR1 -2 CSNK1G3 (CK1 gamma RET 85 PRKCA (PKC alpha) 8 -2 3) CAMK2A (CaMKII CHEK2 (CHK2) 83 CHUK (IKK alpha) 8 -2 alpha) ALK 82 PRKD1 (PKC mu) 8 ERBB2 (HER2) -2 RPS6KA1 (RSK1) 81 PHKG1 7 PIM1 -2 ROS1 80 CSNK1G2 (CK1 gamma 2) 7 MAPK10 (JNK3) -2 PAK4 75 NTRK3 (TRKC) 7 ADRBK2 (GRK3) -2 PAK7 (KIAA1264) 74 YES1 6 MARK3 -3 LTK (TYK1) 68 CDK1/cyclin B 6 SPHK1 -3 AURKA (Aurora A) 66 EPHB4 6 NEK6 -3 PIK3C2B (PI3K-C2 beta) 54 AURKB (Aurora B) 6 BRSK1 (SAD1) -3 CSNK1A1 (CK1 alpha PRKCG (PKC gamma) 53 PKN1 (PRK1) 6 -3 1) FLT4 (VEGFR3) 48 MST1R (RON) 6 AMPK A1/B1/G1 -3 PTK2 (FAK) 45 FGR 6 DYRK3 -3 PIK3CD/PIK3R1 (p110 IRAK4 43 6 PASK -4 delta/p85 alpha) PI4KB (PI4K beta) 43 GRK5 6 AMPK A2/B1/G1 -4 INSR 42 SRPK1 6 EEF2K -4 FLT3 42 PRKACA (PKA) 5 MAPK11 (p38 beta) -4 PDGFRA (PDGFR alpha) 36 TYRO3 (RSE) 5 FGFR2 -4 BTK 36 CDK2/cyclin A 5 FGFR4 -4 CLK2 36 RPS6KA5 (MSK1) 5 EPHA2 -4 IGF1R 34 DYRK4 5 PRKG2 (PKG2) -4 RPS6KB1 (p70S6K) 32 MAPKAPK5 (PRAK) 5 ADRBK1 (GRK2) -5 JPET #233155

PIK3CA/PIK3R1 (p110 30 SGKL (SGK3) 5 MARK4 -5 alpha/p85 alpha) SNF1LK2 26 CAMK2B (CaMKII beta) 5 NTRK1 (TRKA) -6 JAK1 26 PRKCB1 (PKC beta I) 5 NTRK2 (TRKB) -6 BRAF 25 CSNK1G1 (CK1 gamma 1) 5 DAPK3 (ZIPK) -6 PAK3 25 HIPK2 5 TAOK2 (TAO1) -6 ERBB4 (HER4) 22 CSNK1E (CK1 epsilon) 5 PAK2 (PAK65) -6 EPHB2 22 PRKG1 5 DYRK1A -6 PIK3C2A (PI3K-C2 alpha) 21 JAK2 JH1 JH2 5 GSK3B (GSK3 beta) -6 EPHA1 21 NEK7 4 PLK3 -6 PTK2B (FAK2) 19 KIT 4 DAPK1 -6 EPHA5 19 CHEK1 (CHK1) 4 MAPK12 (p38 gamma) -6 GRK4 19 EPHA8 4 MYLK2 (skMLCK) -6 CLK1 19 GSK3A (GSK3 alpha) 4 PHKG2 -6 IKBKE (IKK epsilon) 19 MAPK3 (ERK1) 4 MAP2K2 (MEK2) -6 AURKC (Aurora C) 19 ACVR1B (ALK4) 4 SGK (SGK1) -7 LCK 19 HCK 4 PDK1 Direct -8 MAPK8 (JNK1) 18 PI4KA (PI4K alpha) 4 NEK9 -8 EPHA3 18 CSNK2A1 (CK2 alpha 1) 3 MAP4K4 (HGK) -9 EPHB1 18 MST4 3 STK4 (MST1) -10 STK23 (MSSK1) 18 ABL2 (Arg) 3 GSG2 (Haspin) -10 ZAP70 17 MAPK1 (ERK2) 3 CDC42 BPB (MRCKB) -10 FRK (PTK5) 16 MARK2 3 NEK1 -10 CAMK1D (CaMKI GRK7 15 MAPKAPK3 3 -10 delta) MAPK9 (JNK2) 15 FYN 3 PRKCE (PKC epsilon) -11 AKT2 (PKB beta) 15 CSNK1D (CK1 delta) 3 ITK -11 EPHA4 15 CSNK2A2 (CK2 alpha 2) 3 PIM2 -11 MAP2K1 (MEK1) 13 AKT1 (PKB alpha) 2 MATK (HYL) -12 CSK 13 PRKCZ (PKC zeta) 2 ROCK1 -12 MERTK (cMER) 13 RPS6KA4 (MSK2) 2 SRPK2 -12 MAP2K6 (MKK6) 13 CAMK4 (CaMKIV) 2 PRKCQ (PKC theta) -12 ROCK2 13 CDK7/cyclin H/MNAT1 2 STK25 (YSK1) -13 PRKX 13 GRK6 2 HIPK4 -14 PIK3CG (p110 gamma) 13 NEK4 2 PRKCB2 (PKC beta II) -14 MUSK 12 MAPK13 (p38 delta) 2 FGFR3 -15 EGFR (ErbB1) 12 PLK1 1 NEK2 -16 SRC N1 12 CAMK1 (CaMK1) 1 MAPK14 (p38 alpha) -16 TBK1 12 PDGFRB (PDGFR beta) 1 STK24 (MST3) -17 LYN A 12 EPHB3 1 PRKCD (PKC delta) -19 TYK2 11 SRC 1 PRKCH (PKC eta) -19 PRKCN (PKD3) 11 AXL 1 PRKCI (PKC iota) -20 JAK3 11 PDK1 1 PIK3C3 (hVPS34) -20 PAK6 11 DYRK1B 1 MINK1 -20 INSRR (IRR) 11 MAPKAPK2 1 FRAP1 (mTOR) -21 DCAMKL2 (DCK2) 11 HIPK1 (Myak) 0 MAP4K5 (KHS1) -26 JPET #233155

PRKD2 (PKD2) 11 SGK2 0 SPHK2 -39

Table 2: % inhibition of receptors in CEREP ExpresSProfile panel of BI 1002494 at 1µM

Assay % Inhibition

M1 (h) 70

A1 (h) 63

A2A (h) 59

A3 (h) 43

M2 (h) 32

5-HT2B (h) 26

5-HT2A (h) 20

d2 (DOP) (h) 19 Na+ channel (site 2) 18

norepinephrine transporter (h) 14 m (MOP) (h) 11 k (KOP) 10

MC4 (h) 7

5-HT1A (h) 6

MT1 (ML1A) (h) 5

5-HT3 (h) 4

D1 (h) 4

a1 (non-selective) 3

H1 (h) 3

5-HT6 (h) 2

a2 (non-selective) 2 CXCR2 (IL-8B) (h) 2

dopamine transporter (h) 2 NOP (ORL1) (h) 1 5-HT transporter (h) 0

B2 (h) 0 Cl- channel (GABA-gated) 0

GAL2 (h) 0

TP (h) (TXA2/PGH2) 0

5-HT1B -1

5-HT5a (h) -1

b2 (h) -1

D2S (h) -1 GABA (non-selective) -1

V1a (h) -1

5-HT7 (h) -2 Ca2+ channel (L, verapamil site) (phenylalkylamine) -2

H2 (h) -2

NK2 (h) -2

AT1 (h) -3 CCR1 (h) -3 JPET #233155

CB1 (h) -4

M3 (h) -4

KV channel -5

VPAC1 (VIP1) (h) -5 BZD (central) -6

NK3 (h) -6

b1 (h) -7

SKCa channel -7

ETA (h) -8

Y2 (h) -8

NTS1 (NT1) (h) -10 sst (non-selective) -12

Y1 (h) -12