Discovery of Highly Potent and BMPR2-Selective Kinase Inhibitors Using DNA-Encoded Chemical Library Screening

Abstract

The discovery of monokinase-selective inhibitors for patients is challenging because the 500+ kinases encoded by the human genome share highly conserved catalytic domains. Until now, no selective inhibitors unique for a single transforming growth factor β (TGFβ) family transmembrane receptor kinase, including bone morphogenetic protein receptor type 2 (BMPR2), have been reported. This dearth of receptor-specific kinase inhibitors hinders therapeutic options for skeletal defects and cancer as a result of an overactivated BMP signaling pathway. By screening 4.17 billion "unbiased" and "kinase-biased" DNA-encoded chemical library molecules, we identified hits CDD-1115 and CDD-1431, respectively, that were low-nanomolar selective kinase inhibitors of BMPR2. Structure-activity relationship studies addressed metabolic lability and high-molecular-weight issues, resulting in potent and BMPR2-selective inhibitor analogs CDD-1281 (IC₅₀ = 1.2 nM) and CDD-1653 (IC₅₀ = 2.8 nM), respectively. Our work demonstrates that DNA-encoded chemistry technology (DEC-Tec) is reliable for identifying novel first-in-class, highly potent, and selective kinase inhibitors.

Figure 1

Enrichment profile of BMPR2 kinase domain binders using Baylor College of Medicine DECLs. Enrichment profile of BMPR2 at 0.2 μM from two DECLs: (A) qDOS18_2 and (B) qDOS27. For each library, the selection data show the enrichment of the same building block 3 (BB3; in black) and with various BB1 (in blue) and BB2 (in red). The enrichment of each hit series was shown as count/z-score in the box and graphed as z-score in the BMPR2 0.2 μM selection on the x axis versus z-score for the selection against no-target control on the y axis for (A) or versus z-score for the selection against BMPR2 0.2 μM plus CDD-1115 on the y axis for (B). CDD-1115 and CDD-1431 correspond to the on-DNA hits with the highest affinity in the qDOS18_2 and qDOS27 selections, respectively. Building blocks closest to farthest from the attached DNA are sequentially shown in blue, red, and black.

Figure 2

BMPR2-specific compounds from DECLs. (A) Structures of compounds CDD-1115 (compound 4a) and CDD-1431 (compound 7a). For CDD-1115 and CDD-1431, cycle 1 building block is in blue, cycle 2 building block is in red, and cycle 3 building block is in black. (B) Thermal shift assay to determine the binding of CDD-1115 and CDD-1431 compounds to BMPR2 kinase domain protein. CDD-1115 and CDD-1431 compounds produced large positive thermal shifts in BMPR2 kinase domain protein.(C) Inhibition K_iapp value determination for BMPR2. Concentration-dependent inhibition curves of CDD-1115 and CDD-1431 and K_iapp values were obtained as described in the Experimental Section.

Scheme 1. General Synthetic Routes for the Synthesis of DEC-Tec Selection Hits for (A) CDD-1115 Series, (B) CDD-1431 Series, and SAR Analogs

Reagents and conditions for each synthesis step are as follows: (i) different substituted benzyl amines R¹NH₂, O-(7-azabenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HATU), N,N diisopropylethylamine (DIEA), DMF, rt, 16 h; (ii) different substituted aliphatic amines R²NH₂, TEA, DMF, 85 °C, 2 h or microwave 120 °C, 10 min (on-DNA condition: pH 9.5 borate buffer, DMSO, 80 °C, 16 h); (iii) for R² containing methyl ester: LiOH·H₂O, THF, H₂O, rt, 1 h, then CH₃NH₂·HCl, HATU, DIEA, DMF, rt, 16 h; and (iv) different substituted carbaldehydes R³CHO, Na₂S₂O₄, DMSO/H₂O (v/v = 4:1), 85 °C, 30 min (on-DNA nitroreduction: Na₂S₂O₄, methyl viologen dichloride, 80 °C, 15 min; on-DNA benzimidazole synthesis: pH 5.8 MES buffer, rt, 16 h); (v) different substituted 2,4-dichloropyrimidine, various substituted phenylpiperazines or phenyl azitidines, DIEA, 40–80 °C, 12 h; and (vi) substituted aminobenzenesulfonamide, BrettPhos G2, dioxane, 110 °C, 1 h in a microwave (on-DNA Buchwald coupling reaction: Pd-PEPPSI, CsOH, Na ascorbate, DMA, 95 °C, 15 min); for R² containing Boc-piperazine, 4 M HCl/dioxane.

Figure 3

Kinase selectivity profile of compounds CDD-1115, CDD-1496, and CDD-1653. Compounds were assayed at 1 μM against 403 kinases using the DiscoveRx KINOMEscan screen. Compound selectivity is represented in a TREEspot kinase dendrogram view of the human kinome phylogenetic tree. (A) 0.25% percent control for CDD-1115, (B) 0.1% percent control for CDD-1496, and(C) 0% percent control for CDD-1653. % inhibition = 1 – % control; the lower percent control and larger red circles indicate stronger inhibition against the corresponding kinases; all other kinases tested were inactive as indicated by the green circles.

Figure 4

Inhibition of lead compounds on BMP-mediated gene expression in mammalian cell cultures. (A) Inhibition of BMP-stimulated luciferase transactivation in HEK293T BMP reporter (293T BRE-R) cells treated with 5 ng/mL BMP2 for 6 h in the presence or absence of various concentrations of the small molecule inhibitors CDD-1653, CDD-1431, CDD-1496, CDD-1115, CDD-1280, and CDD-1281 or the ALK2/3/6 inhibitor LDN-193189. Dose–response data represent a nonlinear fit analysis model (inhibitor versus response with variable slope, eq Y = bottom + (top – bottom)/(1 + (IC50/X)^HillSlope)). (B) Calculated IC₅₀ values for the dose–response curves in (A). (C) HEK293T cells were pretreated with small molecule inhibitors (25 μM CDD-1653, 25 μM CDD-1496, or 25 μM CDD-1281) or the ALK2/3/6 inhibitor (1 μM LDN-193189) for 30 min followed by stimulation with 5 ng/mL BMP2 for 15 min. Western blot was used to detect phosphorylated SMAD1/5 (pSMAD1/5), total SMAD1, SMAD5, and GAPDH. (D) Densitometric analysis of pSMAD1/5 in HEK293T cells (C) treated with the various small molecule inhibitors (n = 3 biological replicates, one-way ANOVA with Tukey’s post hoc analysis). (E) Human umbilical vein endothelial cells (HUVECs) were pretreated with small molecule inhibitors for 30 min (at the same concentrations as 293T cells) followed by stimulation with 0.5 ng/mL BMP9 for 15 min. Western blot detection of phosphorylated SMAD1/5 (pSMAD1/5), total SMAD1, SMAD5, and GAPDH. (F) Densitometric analysis of pSMAD1/5 in HUVECs (shown in E) treated with the various small molecule inhibitors (n = 3 biological replicates, one-way ANOVA with Tukey’s post hoc analysis).