The Orally Available Spleen Tyrosine Kinase Inhibitor 2-[7-(3,4-Dimethoxyphenyl)-imidazo[1,2-c]pyrimidin-5-ylamino]nicotinamide Dihydrochloride (BAY 61-3606) Blocks Antigen-Induced Airway Inflammation in Rodents

Spleen tyrosine kinase (Syk) is a cytosolic 72-kDa protein tyrosine kinase that plays an essential role in high-affinity IgE receptor (FcϵRI)-mediated signaling in mast cells and basophils (Beaven and Baumgartner, 1996). Mast cells developed from Syk-deficient mice conclusively demonstrated the essential role of Syk in FcϵRI signaling not only for degranulation but also for lipid mediator synthesis and cytokine production (Costello et al., 1996). Mast cells and basophils produce cytokines important for the late phase allergic reaction (Costello et al., 1996; Shichijo et al., 1999). Mast cell-deficient mice did not exhibit airway inflammation (Kung et al., 1995) or hyperresponsiveness (Kobayashi et al., 2000). These experimental results suggest that mast cells play important roles not only in early but also in late phase allergic reactions and that Syk inhibitors would prevent both phases.

In addition to the critical role of Syk in FcϵRI signaling, it has been reported that Syk is essential in signaling from receptors for IgG (FcγR). Syk-deficient macrophages and neutrophils failed to phagocytose IgG-coated antigen through FcγR (Crowley et al., 1997; Kiefer et al., 1998). Antigen presentation mediated by antibody and FcR (FcγR and FcϵRI) was demonstrated to be around 100-fold more efficient than that in the absence of the antibody (Sallusto and Lanzavecchia, 1994; Maurer et al., 1996). In fact, an antigen/IgE immune complex more efficiently induced airway inflammation than the antigen alone (Zuberi et al., 2000). The bronchoalveolar lavage (BAL) fluid from the ovalbumin (OVA)-challenged mice contained significant amounts of antigen-specific IgE and IgE-OVA immune complexes. These observations suggest the importance of FcR-mediated phagocytosis/antigen presentation for the deterioration of inflammation and thereby the signaling through Syk to facilitate phagocytosis in maintaining chronic inflammation by repeated and effective antigen presentation.

Furthermore, recent in vivo experiments, in which syk antisense oligodeoxynucleotide treatment inhibited airway inflammation in rats, directly suggest an important role of Syk in pulmonary inflammation (Stenton et al., 2000).

These literature data strongly suggest that Syk is an important enzyme in various inflammation pathways relevant to respiratory diseases and therefore a key target for a novel antiasthmatic therapy. We have recently identified an orally available Syk kinase inhibitor, BAY 61-3606, and in this study, we have characterized the pharmacological profiles of BAY 61-3606 both in vitro and in vivo.

Materials and Methods

Chemicals, Antibodies, and Kits. 2-[7-(3,4-Dimethoxyphenyl)-imidazo[1,2-c]pyrimidin-5-ylamino]-nicotinamide dihydrochloride (BAY 61-3606) was synthesized by the Department of Chemistry (Bayer Yakuhin, Ltd., Kyoto, Japan). The structure of BAY 61-3606 is shown in Fig. 1. Peptide substrates for Syk (biotin-KISDFGLSKALRADENYYKAQTHGKWPVK W) and Lyn (biotin-Ahx-KVEKIGEGTYGVVYK-NH₂), and Boc-Ala-Gly-Pro-Arg-MCA were obtained from Peptide Institute (Osaka, Japan). Europium-labeling kit, LTC₄/D₄/E₄ enzyme immunoassay kit, leukotriene (LT) C₄/D₄/E₄, granulocyte-macrophage colony-stimulating factor (GM-CSF) ELISA kit, and Percoll were purchased from Amersham Biosciences UK, Ltd. (Little Chalfont, Buckinghamshire, UK). Src was purchased from Upstate Biotechnology (Lake Placid, NY). Recombinant human stem cell factor and IL-6 were from Peprotech (Rocky Hill, NJ). Histamine ELISA kit was from Immunotech (Marseille, France). PGD₂-methoxylamine enzyme immunoassay kit was from Cayman Chemical (Ann Arbor, MI). Anti-human IgG F(ab′)₂ were from ICN Biomedicals Inc. (Costa Mesa, CA). All culture mediums and culture supplements were from Invitrogen (Carlsbad, CA). Anti-dinitrophenol (DNP) monoclonal mouse IgE (SPE-7), DNP-conjugated bovine serum albumin (DNP-BSA), and other reagents were obtained from Sigma-Aldrich (St. Louis, MO).

Syk Kinase Assay. Assay was implemented as described previously (Yamamoto et al., 2003).

Tyrosine Kinase Selectivity Assays. Fyn, glutathione S-transferase-tagged Lyn, and histidine₆ (His)-tagged Itk and Btk were expressed in insect cells using the baculovirus expression system from BD PharMingen (San Diego, CA) according to standard protocol. T cell receptor ζ-chain tagged with His (His-zeta) was expressed in Escherichia coli and purified by Ni²⁺-resin. His-zeta was used as a substrate for all these kinases except Lyn. The concentration of ATP and apparent K_m, respectively, were 30, 35 (Lyn), 10, 10 (Fyn), 10, 10 (Src), 40, 36 (Itk), and 30, 29 μM (Btk).

RBL-2H3 Degranulation Assay. Degranulation assays were performed as previously reported with slight modification (Taylor et al., 1995). Briefly, RBL-2H3 cells were maintained in minimum essential medium supplemented with 15% fetal calf serum and antibiotics-antimycotics. Cells were seeded in plates and cultured for more than 24 h with 0.3 μg/ml of SPE-7. After washing with PIPES buffer (25 mM PIPES, 125 mM NaCl, 2.7 mM KCl, 5.6 mM glucose, 1 mM CaCl₂, and 0.1% BSA, pH 7.4), cells were treated with a test compound for 15 min at 37°C. Cells were then stimulated with 0.1 μg/ml DNP-BSA for 45 min. The activity of hexosaminidase in the supernatant was measured by an enzyme assay with p-nitrophenyl-β-d-glucosaminide as a substrate.

Human Cultured Mast Cell (HCMC) Assays. HCMCs were developed from cord blood stem cells by culturing with 80 ng/ml stem cell factor and 50 ng/ml IL-6 for more than 12 weeks according to the method of Saito et al. (1995). HCMCs were sensitized with 1 μg/ml human IgE overnight. After washing cells with Hanks' balanced salt solution with 0.1% BSA, HCMCs were treated with a test compound for 15 min and then challenged with 4 μg/ml anti-human IgE for 30 min (for histamine and PGD₂). For cytokine measurements, cells were resuspended with culture medium and stimulated for 6 h. Supernatants were recovered after centrifugation and kept at -20°C until ELISA assays for histamine, PGD₂, LTC₄/D₄/E₄, and GM-CSF. Released tryptase was measured by the enzyme assay with Boc-Ala-Gly-Pro-Arg-MCA as a substrate.

Human Basophil Degranulation Assay. Peripheral blood samples were obtained from atopic and nonatopic volunteers. Leukocytes were purified by dextran sedimentation. Leukocytes were resuspended in Hanks' balanced salt solution with 0.1% BSA and seeded into 96-well plates (5 × 10⁵ cells/well). After incubation with a test compound for 15 min, cells were stimulated with 4 μg/ml anti-human IgE antibody for 30 min at 37°C. Supernatants were recovered after centrifugation and stored at -20°C until use. Released histamine was measured by ELISA.

Mouse Eosinophil Superoxide Production Assay. Spleens were obtained from IL-5 transgenic mice (kind gift from Dr. Tominaga; Department of Medical Biology, Kochi Medical School, Japan) (Tominaga et al., 1991). Mouse eosinophils were purified from the splenocytes by negative selection with magnetic beads-labeled with anti-Thy1.2 and anti-B220 monoclonal antibodies by using MACS (Miltenyi Biotec Inc., Auburn, CA). After purification, cells were cultured overnight in RPMI 1640 medium containing 10% fetal calf serum. Cells were suspended with Hanks' balanced salt solution in 96-well plates (10⁵ cells/well) and treated with the indicated concentration of a test compound for 15 min at 37°C. The cells were then transferred to plates coated with 3 μg/ml human IgG and preblocked with 1% BSA. After incubation for 20 min at 37°C, superoxide production was measured by TopCount by monitoring the luminescence from luminol included in the assay buffer.

U937 Superoxide Production Assay. Assay was implemented as described previously (Yamamoto et al., 2002).

Rats for in Vivo Studies. Male Wistar rats for acute models and male Brown Norway rats for a chronic model (both 6 week old or older) (Charles River Japan, Yokohama, Japan) were used.

Passive Cutaneous Anaphylaxis (PCA) Reactions. Rats were passively sensitized by s.c. injection of 5 ng of anti-DNP IgE in dorsal skin. One day later, a test compound in saline containing 10% cremophor was administered 5 min (i.v.) or 60 min (p.o.) before DNP-BSA administration (1 mg in saline containing 0.5% Evans blue, i.v.). Thirty minutes later, rats were sacrificed, and Evans blue in the sites of sensitization was extracted by formamide overnight at 65°C. Absorbency at 620 nm was measured to determine the amount of Evans blue.

Bronchoconstriction and Bronchial Edema Models in Rats.Rats were passively sensitized by i.v. injection of 10 μg of SPE-7 1 day before experiments. After urethane anesthesia, main bronchi were exposed and cannulated to measure the change in pulmonary pressure. BAY 61-3606 was administered (p.o.) 60 min before injection of the antigen (1.5 μg of DNP-BSA in saline containing 0.5% Evans blue, i.v.). Change in pulmonary pressure was monitored for 10 min after antigen exposure. Thirty minutes after the challenge, rats were sacrificed and lungs were perfused with 20 ml of phosphate-buffered saline. Evans blue in main bronchus was extracted by formamide and measured as described above.

Airway Inflammation Model in Rats. Rats were immunized by i.p. injection of OVA in Al(OH)₃ suspension on days 0 and 14. On days 20 and 21, an aerosol of 1% OVA in saline was administered by inhalation. BAL fluid was collected, and cell number and differential counts were determined. BAY 61-3606 was administered (p.o.) from days 0 to 21 (b.i.d.). Dexamethasone was administered (p.o.) from days 0 to 9 and days 18 to 21 (b.i.d.).

Results

Biochemical Characterization. We have identified an orally available Syk kinase inhibitor, BAY 61-3606 (Fig. 1), from a series of imidazopyrimidine analogs. BAY 61-3606 inhibited kinase activity of Syk in a concentration-dependent manner with an IC₅₀ value of 10 nM (Fig. 2a). Lineweaver-Burk analysis confirmed competitive inhibition against ATP (Fig. 2b), and the K_i value was determined as 7.5 nM.

BAY 61-3606 was a highly selective inhibitor of Syk kinase. Other selected tyrosine kinases, Lyn, Fyn, Src, Itk, and Btk, were not inhibited by BAY 61-3606 in concentrations up to 4.7 μM (Table 1).

Inhibition of Cellular Function. In functional assays to measure FcϵRI-mediated degranulation in mast cells, BAY 61-3606 inhibited the release of various inflammatory mediators in a concentration-dependent manner. The IC₅₀ values for the FcϵRI-mediated hexosaminidase release from a rat basophilic leukemia cell line, RBL-2H3 (Fig. 3a), and serotonin release from rat peritoneal mast cells (Table 2) were found to be 46 and 17 nM, respectively. In RBL-2H3 cells, phosphorylation of Syk was also attenuated (data not shown). None of the reference compounds, dexamethasone (glucorticoid), disodium cromoglycate (DSCG, mast cell stabilizer), and montelukast (LT antagonist), were found to inhibit hexosaminidase release from RBL-2H3 cells at the concentrations evaluated in this study (Fig. 3a).

In a manner similar to its effect on the degranulation of RBL-2H3 cells and rat peritoneal mast cells, BAY 61-3606 inhibited FcϵRI-mediated histamine and tryptase release from HCMCs with IC₅₀ values of 5.1 and 5.5 nM, respectively (Fig. 3, b and c). In addition to the effects on the degranulation, BAY 61-3606 inhibited FcϵRI-mediated lipid mediator release (PGD₂ and LTC₄/D₄/E₄) and de novo synthesis of the cytokine GM-CSF in HCMCs (IC₅₀ = 5.8, 3.3, and 200 nM, respectively; Fig. 3, d-f). Montelukast (IC₅₀ = 6.8 μM for histamine and 7.0 μM for tryptase) and DSCG (IC₅₀ = 860 μM for histamine) inhibited degranulation from HCMCs only very weakly, and dexamethasone showed no effect up to 30 μM (Fig. 3, b and c). The potency of reference compounds was weak for lipid mediator synthesis (montelukast, IC₅₀ = 6.8 μM for LT; Fig. 3, d and e). Dexamethasone showed higher potency (IC₅₀ = 6 nM) than BAY 61-3606 for inhibition of cytokine production (Fig. 3f).

BAY 61-3606 was also found to inhibit the degranulation of human freshly isolated basophils. Leukocyte fractions including basophils were isolated from peripheral blood of both high-(>280 U/ml) and low (<280 U/ml)-serum IgE donors. The expression level of FcϵRI in leukocytes from high-serum IgE donors was higher than that from low-serum IgE donors as demonstrated by flow cytometry (data not shown). The leukocytes from both groups responded to anti-IgE stimulation by releasing histamine. BAY 61-3606 inhibited histamine release from leukocytes in high and low IgE groups equipotently, giving IC₅₀ values of 8.1 and 10 nM, respectively (Fig. 4).

BAY 61-3606 was also found to inhibit B cell receptor (BCR)-mediated signaling. The IC₅₀ values for BCR-stimulated increases in intracellular calcium concentration in the Ramos human B cell line and for BCR-induced mouse splenic B cell proliferation were 81 and 58 nM, respectively (Table 2). Phosphorylation of Syk in Ramos cells was concentration dependently reduced (data not shown). Dexamethasone showed a similar potent inhibition of B cell growth (IC₅₀ = 30 nM) as BAY 61-3606, whereas DSCG and montelukast showed no effect up to 10 μM (data not shown).

Furthermore, BAY 61-3606 was found to block FcγR-mediated activation of monocytes effectively. BAY 61-3606 inhibited FcγR-mediated superoxide production from a human monocytic cell line, U937 (Fig. 5b) and human monocytes freshly isolated from peripheral blood (Table 2) (IC₅₀ = 52 and 12 nM, respectively). The reference compounds were without effect up to 10 μM on the respiratory burst from U937 triggered by FcγRI-aggregation (Fig. 5b). We also examined effects of BAY 61-3606 on the respiratory burst in eosinophils by FcγR stimulation. In mouse, immobilized IgG elicited superoxide production, and it was suppressed by the pretreatment of cells with anti-FcγRIII/II monoclonal antibody 2.4G2 (Fig. 5a). BAY 61-3606 inhibited respiratory burst in a concentration-dependent manner with an IC₅₀ value of 35 nM (Table 2), and it completely suppressed superoxide production at 1 μM, whereas the efficacy of other reference compounds was none or weak at the concentrations used in this study (Fig. 5a).

In Vivo Characterization. PCA assay was carried out to investigate the effect of BAY 61-3606 on mast cell-mediated type-I allergic reactions in rats. Oral administration of BAY 61-3606 dose dependently inhibited the PCA reaction with an ED₅₀ value of 8 mg/kg (statistical inhibition at 3 mg/kg; p < 0.05) (Fig. 6a). Although DSCG (30 mg/kg i.v.) showed 60% inhibition of dye leakage by the PCA reaction (Fig. 6b), this compound was not active with oral administration (data not shown). Neither montelukast (10 mg/kg p.o.) nor dexamethasone (0.3 mg/kg p.o.) showed any inhibitory activity (Fig. 6, c and d).

The effects of BAY 61-3606 on antigen-induced asthmatic models were investigated. In a bronchoconstriction model, BAY 61-3606 dose dependently inhibited the DNP-BSA-induced increase in pulmonary pressure, and the dose of 3 mg/kg showed statistically significant suppression (Fig. 7). As shown in Fig. 8, BAY 61-3606 also significantly attenuated DNP-BSA-induced bronchial edema at and over dosages of 3 mg/kg. Furthermore, BAY 61-3606 significantly inhibited eosinophil accumulation in the BAL fluid at dose of 30 mg/kg to the same level as dexamethasone at a dose of 0.3 mg/kg (Fig. 9). The inhibitory effect of BAY 61-3606 on the total cell number in BAL fluid was also observed at 30 mg/kg, but it was not statistically significant (data not shown). Body weight of dexamethasone-treated rats decreased gradually during the course of experiment (starting from 140 g, -20 g at the end of the experiment); however, those of BAY 61-3606-treated groups at all dosages were almost same as the vehicle group (∼40 to 60 g at the end of the experiment) (data not shown).

Discussion

BAY 61-3606 potently inhibited recombinant Syk kinase activity in an ATP-competitive manner (K_i = 7.5 nM; Fig. 2). Furthermore, more than 626-fold selectivity against several other tyrosine kinases, such as Lyn, Fyn, Src, Itk, and Btk, was demonstrated (Table 1).

The efficacy of BAY 61-3606 on antigen-induced degranulation was confirmed both in RBL-2H3 cells (IC₅₀ = 46 nM; Fig. 3a) and freshly isolated rat mast cells (IC₅₀ = 17 nM; Table 2). Also in vivo, BAY 61-3606 was effective in suppressing PCA reactions in skin (Fig. 6). Oral dosing over 3 mg/kg was significantly effective and inhibition was dose-dependent. Furthermore, effects of BAY 61-3606 on lung mast cells were confirmed in two acute asthmatic models; DNP-BSA-induced bronchoconstriction (Fig. 7) and bronchial edema (Fig. 8). Statistic significance was obtained over 3 mg/kg p.o. in both assays. Also in mice, oral administration of BAY 61-3606 suppressed PCA reaction dose dependently (data not shown). This in vivo evidence suggests that BAY 61-3606 may be an effective orally available antiallergy medicine.

Although RBL-2H3 and rat peritoneal mast cells have been used frequently to study the effect of compounds on mast cells, compounds effective on these cells were often less active in human mast cells (Pearce et al., 1982). We thus studied the efficacy of BAY 61-3606 on cord blood stem cell-derived HCMCs. HCMCs have been recognized to show similar pharmacological characteristics to human lung mast cell (Shichijo et al., 1998), which is one of the primary target cells for asthma therapy. Similarly to rat cells, BAY 61-3606 blocked activation of HCMCs by FcϵRI-aggregation (Fig. 3, b-f). The efficacy was not only on degranulation but also on lipid mediators and cytokine production. These results are consistent with the phenotypes observed in Syk-deficient mast cells (Costello et al., 1996). Relatively smaller IC₅₀ values in rapid mediator release/synthesis in HCMCs compared with those in rat mast cells would indicate heterogeneity of mast cells. The relatively high IC₅₀ value of BAY 61-3606 in GM-CSF production might be due to longer incubation time in this assay or to difference in sensitivity to Syk inhibition between readouts. Tryptase has been considered to cause the remodeling in the airway (Sommerhoff, 2001). Reticular basement membrane thickness occurs early in the asthma process even in childhood (Jeffery, 2001). Therefore, effective inhibition of tryptase release by BAY 61-3606 might be effective for the airway remodeling as a long-term efficacy.

We extended the efficacy study to freshly isolated human cells, including basophils, which express FcϵRI. BAY 61-3606 showed almost similar efficacy on the degranulation of cells from high-serum IgE (IC₅₀ = 8.1 nM) and low-serum IgE (IC₅₀ = 10.2 nM) donors (Fig. 4), indicating its efficacy in humans, including atopic patients. Based on these mast cell and basophil data, BAY 61-3606 is a potent inhibitor of human mast cell/basophil activation by antigen.

BAY 61-3606 also suppressed BCR signaling. The inhibition of BCR engagement-induced calcium mobilization in a human B cell line, Ramos, by BAY 61-3606 (IC₅₀ = 81 nM; Table 2) is consistent with the phenotype of Syk-deficient DT40 cells (Takata et al., 1994). Purified splenic B cells responded to anti-IgM antibody to show proliferation. BAY 61-3606 attenuated this cell growth (IC₅₀ = 58 nM; Table 2). This is the first pharmacological demonstration, as far as we know, that inhibition of Syk results in the prevention of a B cell function. These B cell data imply that clonal expansion after BCR-engagement by antigen and further maturation could be attenuated by Syk kinase inhibitors.

BAY 61-3606 concentration dependently inhibited FcγR-mediated respiratory burst not only in mouse eosinophils (IC₅₀ = 35 nM; Table 2) but also in a human monocytic cell line, U937 (IC₅₀ = 52 nM; Fig. 5b) and freshly isolated human monocytes (IC₅₀ = 12 nM; Table 2). These data are consistent with the previous publications using syk antisense or cells derived from knockout mice (Matsuda et al., 1996; Crowley et al., 1997; Kiefer et al., 1998; Lach-Trifilieff et al., 2000) and indicate one aspect of anti-inflammatory profiles of BAY 61-3606.

To confirm an outcome of inhibitory actions of BAY 61-3606 in various types of inflammatory cells, we examined the efficacy of BAY 61-3606 in a rat OVA-induced airway inflammation model. BAY 61-3606 at 30 mg/kg p.o., b.i.d., greatly suppressed accumulation of eosinophils in BAL fluid (Fig. 9). The inhibition was 70% and comparable with that of dexamethasone (0.3 mg/kg p.o., b.i.d.). Thus, not only mast cell-stabilizing activity but also anti-inflammatory activity of BAY 61-3606 was confirmed in vivo. The requirement of higher dose in the chronic model might be related to the pharmacokinetic profile of BAY 61-3606. When rats were treated with a single dose of 10 mg/kg p.o. BAY 61-3606, a maximal concentration of 0.167 mg/l (360 nM) was reached 2 h after administration. Due to relatively fast elimination (t_1/2 = 1.78 h), we administered the compound twice a day. It should be noted that dexamethasone reduced the increase in body weight of rats, but BAY 61-3606 had no effect during the course of this experiment (data not shown).

The potency of BAY 61-3606 in mast cell and basophil assays was superior to that of DSCG, a widely used mast cell stabilizer as an inhalant (Figs. 3 and 6). Moreover, dexamethasone and montelukast showed little or no effect in many assays selected for this study (Figs. 3, 5, and 6). This difference in efficacy profiles between BAY 61-3606 and other widely used drugs make it interesting to try BAY 61-3606 as an alternative anti-asthma/allergic medicine with a novel mechanism of action.

In this study, we focused on FcR- and BCR-mediated signals to examine the efficacy of BAY 61-3606. In addition to these signaling cascades, much evidence has been accumulating that Syk also play a critical role in some parts of integrin signaling in neutrophils and macrophages (Vines et al., 2001; Mocsai et al., 2002). Furthermore, an essential role of Syk in IL-1-induced RANTES (regulated on activation normal T cell expressed and secreted) production in human nasal fibroblasts was suggested by an antisense experiment (Yamada et al., 2001). These data increase the value of Syk kinase inhibitor as a broad anti-inflammatory agent.

Contrary to the positive data for drug development described above, evidence exists that indicates risks and potential side effects of Syk kinase inhibitors. Syk-deficient mice showed perinatal lethality due to severe systemic hemorrhaging (Cheng et al., 1995; Turner et al., 1995). The mechanism might be the impaired proliferation and migration of endothelial cells (Inatome et al., 2001). In platelets, Syk was shown to be essential in collagen-induced activation, which is important for the clotting reaction (Watson et al., 2000). In breast cancer, Syk has been implicated as an important inhibitor of cancer cell growth and metastasis (Coopman et al., 2000). In natural killer cells, expression of dominant-negative Syk attenuated natural cytotoxicity (Brumbaugh et al., 1997). In addition, the expression of Syk was reported in several other nonhematopoietic cells (Yanagi et al., 2001). All of these concerns should be clarified in detail in future safety toxicological studies.

In conclusion, BAY 61-3606 is an orally available Syk-selective kinase inhibitor, which exhibits a variety of actions on mast cells, basophils, B cells, eosinophils, and antigen-presenting cells. BAY 61-3606 would have benefits in the treatment of asthma by preventing antigen-induced bronchoconstriction and airway inflammation.