High-throughput screening identifies novel inhibitors of the acetyltransferase activity of Escherichia coli GlmU

Abstract

The bifunctional GlmU protein catalyzes the formation of UDP-N-acetylglucosamine in a two-step reaction using the substrates glucosamine-1-phosphate, acetyl coenzyme A, and UTP. This metabolite is a common precursor to the synthesis of bacterial cell surface carbohydrate polymers, such as peptidoglycan, lipopolysaccharide, and wall teichoic acid that are involved in the maintenance of cell shape, permeability, and virulence. The C-terminal acetyltransferase domain of GlmU exhibits structural and mechanistic features unique to bacterial UDP-N-acetylglucosamine synthases, making it an excellent target for antibacterial design. In the work described here, we have developed an absorbance-based assay to screen diverse chemical libraries in high throughput for inhibitors to the acetyltransferase reaction of Escherichia coli GlmU. The primary screen of 50,000 drug-like small molecules identified 63 hits, 37 of which were specific to acetyltransferase activity of GlmU. Secondary screening and mode-of-inhibition studies identified potent inhibitors where compound binding within the acetyltransferase active site was requisite on the presence of glucosamine-1-phosphate and were competitive with the substrate acetyl coenzyme A. These molecules may represent novel chemical scaffolds for future antimicrobial drug discovery. In addition, this work outlines the utility of catalytic variants in targeting specific activities of bifunctional enzymes in high-throughput screens.

FIG. 1.

(A) Scheme of the developed coupled acetyltransferase assay for E. coli GlmU. The addition of excess GlmU_H363N and inorganic pyrophosphatase (PPase) have rendered the acetyltransferase activity rate limiting. The 2 units of inorganic phosphate produced by each acetyltransferase turnover are detected by absorbance of 660 nm after the addition of malachite green assay reagents. (B) Evaluation of GlmU acetyltransferase screen. The high controls (100% residual activity; circles) and the low controls (0% residual activity; squares) are plotted. The assay shows a large hit window between the high and low controls with a calculated Z′ factor of 0.71.

FIG. 2.

Replicate plot of the screen of 50,000 small molecules against E. coli GlmU acetyltransferase reaction. The percent residual activity (%RA) of duplicates is plotted on opposite axes. Data points that fall on a slope of 1 are considered to be in agreement. Molecules within the outlined area below 60% residual activity for each replicate reproducibly showed inhibition below the determined cutoff value and were selected for further study.

FIG. 3.

Inhibition studies with MAC0021939. Initial rate data of different amounts of inhibitor and either AcCoA (A) or GlcN-1-P (B). The fixed substrate was maintained at 100 μM. MAC0021939 was used at concentrations of 0 nM (•), 50 nM (○), 100 nM (▾), 200 nM (▿), and 300 nM (▪). (A) The inhibitor is competitive with regard to AcCoA with a K_ic of 17.2 ± 2.30 nM (B) and uncompetitive with GlcN-1P with a K_iu of 189 ± 17.7 nM.

FIG. 4.

Inhibition studies with MAC0008028. Initial rate data of different amounts of inhibitor and either AcCoA (A) or GlcN-1-P (B). The fixed substrate was maintained at 100 μM. MAC0008028 was used at concentrations of 0 nM (•), 1,000 nM (○), 2,000 nM (▾), and 3,000 nM (▪). (A) The inhibitor is competitive with regard to AcCoA with a K_ic of 404 ± 59.1 nM (B) and uncompetitive with GlcN-1P with a K_iu of 1.11 ± 0.08 × 10³ nM.

FIG. 5.

Inhibition studies with MAC0029665. Initial rate data of different amounts of inhibitor and either AcCoA (A) or GlcN-1-P (B). The fixed substrate was maintained at 100 μM. MAC0029665 was used at concentrations of 0 nM (•), 20 nM (○), 40 nM (▾), 60 nM (▿), and 80 nM (▪). (A) The inhibitor shows mixed inhibition with regard to AcCoA with a K_ic of 22.8 ± 2.70 nM and K_iu of100.9 ± 28.4 nM. (B) Uncompetitive inhibition with GlcN-1P is observed with a K_iu of 28.2 ± 1.83 nM.