Salicylate metal-binding isosteres as fragments for metalloenzyme inhibition

Metalloenzyme inhibitors typically share a common need to possess a metal-binding pharmacophore (MBP) for binding the active site metal ions. However, MBPs can suffer from physicochemical liabilities, impeding the pharmacological properties and drug-likeliness of inhibitors. To circumvent this, problematic features of the MBP can be identified and exchanged with isosteric replacements. Herein, the carboxylic and hydroxyl group of the salicylic acid MBP were replaced and a total of 27 salicylate metal-binding isosteres (MBIs) synthesized. Of these 27 MBIs, at least 12 represent previously unreported compounds, and the metal-binding abilities of >20 of the MBIs have not been previously reported. These salicylate MBIs were examined for their metal-binding features in model complexes, physicochemical properties, and biological activity. It was observed that salicylate MBIs can demonstrate a range of attractive physicochemical properties and bind to the metal in a variety of expected and unexpected binding modes. The biological activity of these novel MBIs was evaluated by measuring inhibition against two Zn2+-dependent metalloenzymes, human glyoxalase 1 (GLO1) and matrix metalloproteinase 3 (MMP-3), as well as a dinuclear Mn2+-dependent metalloenzyme, influenza H1N1 N-terminal endonuclease (PAN). It was observed that salicylate MBIs could maintain or improve enzyme inhibition and selectivity. To probe salicylate MBIs as fragments for fragment-based drug discovery (FBDD), an MBI that showed good inhibitory activity against GLO1 was derivatized and a rudimentary structure–activity relationship was developed. The resulting elaborated fragments showed GLO1 inhibition with low micromolar activity.

at 45 °C. Water (10 mL) was added and the mixture was extracted with EtOAc (3x20 mL).

Methyl 2-(methylsulfonamido)benzoate (12)
Compound 2 (150 mg, 0.70 mmol, 1.00 equiv) was dissolved in anhydrous MeOH (2 mL) and acetyl chloride (0.15 mL, 2.10 mmol, 3.00 equiv) was added dropwise. The mixture was stirred at 70 °C for 14 h. All volatiles were removed under reduced pressure and sat. NaHCO3 was added until a neutral pH was reached. The aqueous phase was extracted with EtOAc (3x10 mL). The combined organics were washed with brine (10 mL), dried over MgSO4 and concentrated under reduced pressure.

General conditions for Suzuki cross-coupling:
Compound s6 (50.0 mg, 0.16 mmol, 1.00 equiv) and boronic acid (0.19 mmol, 1.20 equiv) were mixed in dioxane (0.5 mL) and aq. Na2CO3 (2 M, 0.5 mL) was added. The resulting mixture was degassed (Argon bubbling) for 15 min and stirred at 90 °C for 14 h. The mixture was allowed to cool to room temperature, H2O (5 mL) and EtOAc (15 mL) were added, the phases were separated and the aqueous phase was extracted with EtOAc (3x10 mL). The combined organic phases were washed with brine, dried over MgSO4 and concentrated under reduced pressure. Purification by column chromatography.

General conditions for demethylation A:
Phenol ether (1.00 equiv) was dissolved in CH2Cl2 (1.0 mL) and BBr3 (1 M in heptane, 4.00 equiv) was added dropwise at 0 °C. The resulting mixture was stirred for 14 h at room temperature. Sat. aq. Na2CO3 (5 mL) and EtOAc (15 mL) were added, the phases were separated and the aqueous phase was extracted with EtOAc (3x10 mL). The combined organic phases were washed with brine, dried over MgSO4 and concentrated under reduced pressure. Purification by column chromatography.

Computational docking
The geometry of the molecule was determined using density-functional theory (   shown; with atom labeling (bottom) (see Table S4).

Dose-response curves against GLO1
The raw data from the GLO1 assay was background corrected and normalized from 0 to 100 with 100 being the largest mean of each dataset. A variable slope log-logistic model was chosen for curve fitting. Some compounds did not dissolve well at higher concentrations (>50 µM). The corresponding datapoints are shown as empty bullets and were not included in the curve-fitting. The resulting restricted datasets were fitted according to a literature procedure setting the lower limit of the dose-response curve to zero. This compensates for the loss of information caused by trimming of the dataset. 5,6 Figure S7. Dose-response curve of 11 against GLO1.

X-Ray crystallography data
Single Crystal X-ray Diffraction was solved with the ShelXT 7 structure solution program using direct methods and refined with the XL 8 refinement package using least squares minimization using Olex2. 9 The crystal data files were deposited into the Cambridge Crystallographic Data Centre (CCDC).