Development and Characterization of Potent Succinate Receptor Fluorescent Tracers

Abstract

The succinate receptor (SUCNR1) has emerged as a potential target for the treatment of various metabolic and inflammatory diseases, including hypertension, inflammatory bowel disease, and rheumatoid arthritis. While several ligands for this receptor have been reported, species differences in pharmacology between human and rodent orthologs have limited the validation of SUCNR1's therapeutic potential. Here, we describe the development of the first potent fluorescent tool compounds for SUCNR1 and use these to define key differences in ligand binding to human and mouse SUCNR1. Starting from known agonist scaffolds, we developed a potent agonist tracer, TUG-2384 (22), with affinity for both human and mouse SUCNR1. In addition, we developed a novel antagonist tracer, TUG-2465 (46), which displayed high affinity for human SUCNR1. Using 46 we demonstrate that three humanizing mutations on mouse SUCNR1, N18^1.31E, K269^7.32N, and G84^EL1W, are sufficient to restore high-affinity binding of SUCNR1 antagonists to the mouse receptor ortholog.

Figure 1

Design strategy of antagonist and agonist tracers. The flexible sites identified either by the published X-ray complex of the antagonist 1 or by SAR investigations of the agonists 2 and 3 were explored using a small linker library to identify the optimal linker for attachment of the NBD-fluorophore.

Scheme 1. Synthesis of Agonist Tracer Precursors and the First Agonist Tracer 7

Reagents and conditions: (a) BocNH(CH₂CH₂O)₂Ts or BocNH(CH₂)_2–6X or -OTs, K₂CO₃, MeCN, 80 °C, 21–48 h, 23–79%; (b) 4 M HCl in 1,4-dioxane, DCM, rt, 20–48 h, 15–94%; (c) 0.6 M LiOH (aq), THF, rt, 12–39 h, 53–100%; (d) NBD-Cl, NEt₃, MeOH, rt, 18 h, 40%; (e) BTFFH, DIPEA, DCM, 80 °C, 12 h, 88%; (f) 4-hydroxyphenylboronic acid, XPhos-Pd-G4, 0.5 M K₃PO₄ (aq), THF, rt, 22 h, 62%.

Scheme 2. Synthesis of Optimized Agonist Tracer Precursors and Agonist Tracer 22

Reagents and conditions: (a) Br(CH₂)₆Br, K₂CO₃, KI, MeCN, 80 °C, 3.5 h, 71%; (b) BocNH(CH₂)₂NH₂, K₂CO₃, KI, MeCN, 80 °C, 52 h, 41%; (c) 0.6 M LiOH (aq), THF, rt, 2–53 h, 43–96%; (d) Br(CH₂)₆OH, K₂CO₃, KI, MeCN, 80 °C, 22 h, 43%; (e) DMP, NaHCO₃, DCM, 0 °C, 2 h, 82%; (f) CH₃NH₂·HCl or (CH₃)₂NH·HCl, NaBH(OAc)₃, NEt₃, DCE, rt, 2–3 h, 11–17%; (g) NBD-NH(CH₂)₂NH₂, NaBH(OAc)₃, THF, rt, 23 h, 8%.

Scheme 3. Synthesis of Antagonist Tracer Precursors and the Antagonist Tracer 46

Reagents and conditions: (a) HATU, DIPEA, DMF, rt, 3–41 h, 40–97%; (b) 4-chlorophenylboronic acid or (4-chloro-2-methylphenyl)boronic acid, PdCl₂(PPh₃)₂, 1 M Na₂CO₃ (aq), EtOH or MeOH, toluene, 70–80 °C, 17–20 h, 27–70%; (c) 0.6 M LiOH (aq), THF, rt, 2–14 h, 45–100%; (d) 4-hydroxyphenylboronic acid or (4-hydroxy-2-methylphenyl)boronic acid, Pd-XPhos-G4, 0.5 M K₃PO₄ (aq), THF, rt–50 °C, 19–22 h, 70–75%; (e) alkyl halide/tosylate, K₂CO₃, (KI), MeCN or DMF, 50–90 °C, 22–43 h, 43–57%; (f) 4 M HCl (aq), THF, 55 °C, 15 h, 67% or 4 M HCl in dioxane, rt, 17–23 h, 71–100%; (g) 1) BBA, XPhos-Pd-G2, XPhos, KOAc, EtOH, 80 °C, 1–2.5 h; 2) aryl halide, 1.8 M K₂CO₃ (aq), 80 °C, 18–22 h, 24–75%; (h) TFA, DCM, rt, 3 h, 100%; (i) NBD-Cl, MeOH, Et₃N, rt, 29 h, 30%.

Figure 2

7 is a modest affinity agonist tracer for hSUCNR1 and mSUCNR1. BRET saturation binding for 7 using N-terminally Nluc-tagged hSUCNR1 (A) and mSUCNR1 (B). Non-specific (NS) binding was obtained by treating cells with 100 μM 17. Data are shown as mean ± SEM from three independent experiments carried out in duplicate. Data were globally fit to a total and non-specific binding equation and yielded K_d values of 1.58 μM (95% CI = 0.47–4.98 μM) for hSUCNR1 and 1.16 μM (95% CI = 0.47–4.97 μM) for mSUCNR1.

Figure 3

9d in complex with h- (salmon) and overlaid with mSUCNR1 (petrol) receptor model. Helixes and conserved residues are in gray.

Figure 4

22 is a fluorescent agonist tracer for both human and mouse SUCNR1. BRET saturation binding for 22 using N-terminally Nluc-tagged hSUCNR1 (A) and mSUCNR1 (B). Non-specific (NS) binding was obtained by treating cells with 100 μM 17. Saturation binding data are shown as mean ± SEM from three independent experiments carried out in duplicate. Saturation binding data were globally fit to a total and non-specific binding equation, yielding a K_d value of 0.99 μM (95% CI = 0.45–2.15 μM) for hSUCNR1 and 1.17 μM (95% CI = 0.90–1.51 μM) for mSUCNR1. Kinetic BRET binding experiments are shown for Nluc-hSUCNR1 (C) and Nluc-mSUCNR1 (D). The kinetic data are shown as specific BRET, subtracting signal obtained when treating with equivalent concentrations of 22 in the presence of 100 μM 17. Kinetic binding data are shown as mean ± SEM from three independent experiments carried out in triplicate and were globally fit to an equation for binding of multiple concentrations of labeled ligand.

Figure 5

46 is a fluorescent antagonist tracer for human but not mouse SUCNR1. 46 retains high-potency antagonism in the cAMP assay for the Nluc-hSUCNR1 construct (A). BRET saturation binding for 46 using N-terminally Nluc-tagged hSUCNR1 (B) and mSUCNR1 (C). Non-specific (NS) binding was obtained by treating cells with 100 μM 17. Saturation binding data are shown as mean ± SEM from three independent experiments in duplicate and globally fit to a total and non-specific binding equation, yielding a K_d value of 300 nM (95% CI = 178–528 nM) at hSUCNR1. Kinetic BRET binding experiments are shown for Nluc-hSUCNR1 (D). Kinetic data are shown as specific BRET after subtracting the signal obtained when treating with the equivalent concentrations of 46 in the presence of 100 μM 17. Kinetic binding data are mean ± SEM from three independent experiments in triplicate and were globally fit to an equation for binding of multiple concentrations of labeled ligand.

Figure 6

Humanizing mutations N18^1.31E, K269^7.32N, and G84^EL1W allow for antagonist binding to mSUCNR1. 1 in complex hSUCNR1 (salmon) and overlay with mSUCNR1 (petrol) highlighting the three central amino acids for species selectivity (A). Recovery of antagonist function at hmSUCNR1 was confirmed using a cAMP assay measuring inhibition of an EC₈₀ concentration of succinate with 1 at hSUCNR1, mSUCNR1, and hmSUCNR1 (N18^1.31E, K269^7.32N, and G84^EL1W) (B). Data in cAMP experiments are mean ± SEM of three independent experiments in triplicate. Saturation binding results are shown for hmSUCNR1 using 46 (C) or 22 (D). Non-specific binding (NS) was measured using 100 μM 17. Saturation binding data are shown as mean ± SEM from three independent experiments completed in triplicate. Saturation binding data were globally fit to a total and non-specific binding equation yielding a K_d value for 22 of 490 nM (95% CI = 370–640 nM) and for 46 of 390 nM (95% CI = 230–650 nM).

Figure 7

Humanizing mutations N18^1.31E, K269^7.32N, and G84^EL1W transform mSUCNR1 pharmacology to match that of hSUCNR1. Competition binding experiments were conducted for various SUCNR1 agonists using 22 as the tracer ligand at hSUCNR1 (3 μM 22) (A), mSUCNR1 (1 μM 22) (B), and hmSUCNR1 (1 μM 22) (C). Comparable experiments were conducted using 22 as the tracer for various SUCNR1 antagonists at the same three receptor constructs with their succinate competition curve shown for reference (D–F). All competition binding data are shown as mean ± SEM from three independent experiments conducted in duplicate. The pK_i affinity values obtained at hSUCNR1 (G) or hmSUCNR1 (H) in competition binding experiments using 22 as the tracer ligand are correlated with the values obtained using 46 as the tracer for the same set of competing ligands. A correlation for pK_i values obtained with 22 at hSUCNR1 against the values obtained with the same ligands at either mSUCNR1 or hmSUCNR1 (I).