Discovery of BAY-405: An Azaindole-Based MAP4K1 Inhibitor for the Enhancement of T-Cell Immunity against Cancer

Abstract

Mitogen-activated protein kinase kinase kinase kinase 1 (MAP4K1) is a serine/threonine kinase that acts as an immune checkpoint downstream of T-cell receptor stimulation. MAP4K1 activity is enhanced by prostaglandin E2 (PGE2) and transforming growth factor beta (TGFβ), immune modulators commonly present in the tumor microenvironment. Therefore, its pharmacological inhibition is an attractive immuno-oncology concept for inducing therapeutic T-cell responses in cancer patients. Here, we describe the systematic optimization of azaindole-based lead compound 1, resulting in the discovery of potent and selective MAP4K1 inhibitor 38 (BAY-405) that displays nanomolar potency in biochemical and cellular assays as well as in vivo exposure after oral dosing. BAY-405 enhances T-cell immunity and overcomes the suppressive effect of PGE2 and TGFβ. Treatment of tumor-bearing mice shows T-cell-dependent antitumor efficacy. MAP4K1 inhibition in conjunction with PD-L1 blockade results in a superior antitumor impact, illustrating the complementarity of the single agent treatments.

Figure 1

(A) Schematic representation of the high throughput biochemical kinase assay that is based on the phosphorylation of a synthetic peptide substrate (40 nM), representing IKKRKLTRRKSLKG, by recombinant human MAP4K1 protein in the presence of 10 μM ATP and 10 μM compound, where the quantity of phosphorylated substrate is measured by means of HTRF (Homogeneous Time Resolved Fluorescence). (B) Determination of assay linearity with respect to the concentration of human MAP4K1 protein and incubation time, based on which 77pM MAP4K1 and an incubation period of 60 min were selected. (C) Efficacy distribution of the 60,160 primary hits based on % inhibition in the primary screen. (D) High-throughput screen hit triaging resulted in the identification of one priority cluster represented by screening hit 1 containing a pyrrolo-pyridine as hinge binding motif. This binding mode features two hydrogen bonds and has been experimentally characterized by more than 150 different Protein Data Bank (PDB) entries in KLIFS.

Figure 2

Kinome tree of 1 (BAY-755) and 38 (BAY-405) representing IC₅₀ values recorded for all kinases with % inhibition >80% at 1 μM. MAP4K1 is represented by the blue dot. Total number of kinases tested for 1 (n = 318) and 38 (n = 373). Image generated using TREEspot Software Tool and reprinted with permission from KINOMEscan (DiscoveRx Corp., Fremont, USA).

Figure 3

Co-crystal structure of human MAP4K1 in complex with 38 (PDB accession code 8PAR). (A) Overview of BAY-405 (green) bound to the ATP site of human MAP4K1. The electron density of the bound inhibitor is shown at a sigma of 1.0 (blue mesh). Protein secondary structures are designated in red (alpha-helices) and blue (beta-strands). (B) Hinge-binding interactions of the aza-indole of BAY-405. (C) Hydrogen bonding interactions of the oxazine of BAY-405 to Asp101.

Figure 4

In vitro Pharmacology of BAY-405. (A) Inhibition of MAP4K1 activity in Jurkat T-cells by BAY-405 as measured on the basis of the intracellular levels of phospho-SER376-SLP76 (pSLP76), using MAP4K1 knockdown Jurkat T-cells as controls (Supporting Information Figure S2A). Jurkat T-cells were stimulated with plate-bound anti-CD3 Abs (1 mg/mL) for 30 min after which the pSLP76 levels were determined by a HTRF-based method. (B) Dose-dependent enhancement of T-cell reactivity by BAY-405 and its predecessor BAY-755 in human PBMC cultures stimulated with 30 ng/mL anti-CD3 Ab in the presence of 1 μM PGE2. Secreted IFNγ was analyzed after 24 h by means of ELISA. 2500 nM BAY-405 vs vehicle p = 0.0024; 156 nM BAY-755 vs vehicle p = 0.0059 (student t-test). (C) T-cell reactivity assay as in (B), but performed in the presence of 20 ng/mL human TGFβ. 2500 nM BAY-405 vs vehicle p = 0.0012; 156 nM BAY-755 vs vehicle p = 0.0028 (student t-test). (D,E) T-cell assays as in (A,B) showing the impact of 1 μM BAY-405 on T-cell activation in the presence of different concentrations of anti-CD3 Ab as well as in the presence and absence of 1 μM PGE2 or 20 ng/mL human TGFβ. PGE2 + BAY-405 vs PGE2 + vehicle p < 0.0001. BAY-405 vs vehicle p = 0.7175; TGFβ + BAY-405 vs TGFβ + vehicle p < 0.0001. BAY-405 vs vehicle p = 0.2195 (ANOVA). (F) T-cell reactivity assay with mouse splenocytes from wild-type and MAP4K1 kinase dead knock in mice (Supporting Information Figure S2B,C) in the presence of 300 ng/mL anti-CD3 Ab and 1 μM PGE2 as well as 500 nM BAY-405 or vehicle. WT vehicle vs KI vehicle p = 0.002; WT vehicle vs WT BAY-405 p = 0.0108; KI vehicle vs KI BAY-405 p = 0.4899 (Student’s t-test).

Figure 5

Enhancement of in vitro and in vivo antitumor T-cell reactivity. (A) xCelligence real time analysis of the cytotoxicity of MART-1-specific human T-cells toward the HLA-A*0201, MART-1 positive human melanoma cell line COLO800 in the presence of indicated concentrations of BAY-405. P-value represents the statistical differences at the 100 h time point (Student’s t-test). (B) Systemic exposure of BAY-405 (unbound plasma concentration) after a single oral dose in a mouse as indicated. Indicated in the graph are the BAY-405 IC₅₀ levels determined in the relevant biochemical and cellular assays. See Supporting Information Table S9 for underlying data. (C) Impact of indicated b.i.d. doses of BAY-405 on tumor outgrowth in lungs after i.v. injection of 2 × 10⁴ B16-OVA cells. The numbers of lung nodes were analyzed at day 14 after tumor cell injection. Nine mice were used under each condition. Vehicle vs 30 mg/kg p = 0.0219. Vehicle vs 60 mg/kg p = 0.0001 (Student’s t-test). The experiment was repeated twice with similar outcomes (See Figure 5G and Supporting Information Figure S4B). (D) Suppression of tumor outgrowth by indicated doses of BAY-405 in mice subcutaneously challenged with 2 × 10 × 10⁵ of B16-OVA cells. Four million OT-I were injected iv at day 7. Treatment with BAY-405 started at day 8 and lasted until day 18. Nine mice were used for each condition. Vehicle vs 60 mg/kg b.i.d. p = 0.0393 (ANOVA). (E) Suppression of tumor outgrowth by indicated doses of BAY-405 in mice subcutaneously challenged with 5 × 10⁵ of EMT6 tumor cells. Ten mice were used for each condition. Vehicle vs 60 mg/kg b.i.d. p = 0.0026; vehicle vs 100 mg/kg b.i.d. p = 0.0002 (ANOVA). (F) Suppression pf pSLP76 levels in splenocytes of mice treated with BAY-405 at 60 mg/kg, 100 mg/kg or vehicle b.i.d., as detected by Immunoblotting using a proprietary antimouse pSer376-SLP76 antibody. From 8 mice of each treatment group, spleen samples were taken 1 h after the last compound treatment reflecting Cmax. P-value: vehicle vs BAY-405 60 mg/kg = 0.0006; vehicle vs BAY-405 100 mg/kg > 0.0001. (G) Impact of 60 mg/kg b.i.d. BAY-405, conc. PD-L1 blocking Ab (10 mg/kg twice weekly) and/or a combination thereof on B16-OVA melanoma tumor outgrowth in lungs after i.v. injection of 20,000 tumor cells. The numbers of lung tumor nodes were analyzed at day 14 after tumor grafting. Nine mice were used for each condition. Vehicle vs BAY-405 p = 0.0002; vehicle vs PD-L1 Ab p = 0.0472; vehicle vs BAY-405/PD-L1 Ab p < 0.0001; PD-L1 Ab vs BAY-405/PD-L1 Ab p < 0.0001 (Student’s t-test). (H) Suppression of B16 melanoma outgrowth by 60 mg/kg b.i.d. BAY-405, 10 mg/kg PD-L1 blocking Ab (twice weekly) and/or a combination thereof in mice subcutaneously challenged with 200,000 B16-OVA melanoma cells; 5 × 10⁶ splenocytes from OT-I mice were injected i.v. at day 6 and 11. Treatment with BAY-405 started at day 7 and lasted until day 17. Ten mice were used for each condition. Vehicle vs BAY-405 p = 0.0064; vehicle vs PD-L1 Ab p = 0.0045; PD-L1 Ab vs BAY-405/PD-L1 Ab p < 0.0146 (ANOVA). Experiment was repeated twice.

Scheme 1. Synthesis of BAY-405 (38)

Scheme 2. Synthesis of Oxazines via Acid-Induced Rearrangement of Oxetane Containing Ureas