Antimicrobial activity of compounds identified by artificial intelligence discovery engine targeting enzymes involved in Neisseria gonorrhoeae peptidoglycan metabolism

Abstract

Background: Neisseria gonorrhoeae (Ng) causes the sexually transmitted disease gonorrhoea. There are no vaccines and infections are treated principally with antibiotics. However, gonococci rapidly develop resistance to every antibiotic class used and there is a need for developing new antimicrobial treatments. In this study we focused on two gonococcal enzymes as potential antimicrobial targets, namely the serine protease L,D-carboxypeptidase LdcA (NgO1274/NEIS1546) and the lytic transglycosylase LtgD (NgO0626/NEIS1212). To identify compounds that could interact with these enzymes as potential antimicrobials, we used the AtomNet virtual high-throughput screening technology. We then did a computational modelling study to examine the interactions of the most bioactive compounds with their target enzymes. The identified compounds were tested against gonococci to determine minimum inhibitory and bactericidal concentrations (MIC/MBC), specificity, and compound toxicity in vitro.

Results: AtomNet identified 74 compounds that could potentially interact with Ng-LdcA and 84 compounds that could potentially interact with Ng-LtgD. Through MIC and MBC assays, we selected the three best performing compounds for both enzymes. Compound 16 was the most active against Ng-LdcA, with a MIC50 value < 1.56 µM and MBC50/90 values between 0.195 and 0.39 µM. In general, the Ng-LdcA compounds showed higher activity than the compounds directed against Ng-LtgD, of which compound 45 had MIC50 values of 1.56-3.125 µM and MBC50/90 values between 3.125 and 6.25 µM. The compounds were specific for gonococci and did not kill other bacteria. They were also non-toxic for human conjunctival epithelial cells as judged by a resazurin assay. To support our biological data, in-depth computational modelling study detailed the interactions of the compounds with their target enzymes. Protein models were generated in silico and validated, the active binding sites and amino acids involved elucidated, and the interactions of the compounds interacting with the enzymes visualised through molecular docking and Molecular Dynamics Simulations for 50 ns and Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA).

Conclusions: We have identified bioactive compounds that appear to target the N. gonorrhoeae LdcA and LtgD enzymes. By using a reductionist approach involving biological and computational data, we propose that compound Ng-LdcA-16 and Ng-LtgD-45 are promising anti-gonococcal compounds for further development.

Keywords: Neisseria gonorrhoeae; Artificial intelligence; Bactericidal; Computational modelling; Peptidoglycan.

Fig. 1

Screening of compounds in a standard Minimum Inhibitory Concentration (MIC) assay. Ng-LdcA (n = 74) and Ng-LtgD (n = 84) compounds were diluted and tested at a single concentration of 50 µM in wells containing ~ 10⁵ CFU of N. gonorrhoeae strain P9-17. Controls were bacteria alone, GC broth alone (control in the graphs), GC broth with DMSO (10µL/well) alone, and positive control was bacteria treated with ceftriaxone. DMSO alone has no effect on bacterial growth. Optical density was measured after 24 h incubation. The data are shown as the % reduction in optical density compared to the control bacteria alone. Data are from a representative experiment done twice. The red lines denote the 50% cut offline for selecting compounds for further analyses

Fig. 2

Homology-based models of Ng-LdcA and Ng-LtgD. The E. coli 5Z01 template-based model of Ng-LdcA and the E. coli 1D0K template-based model of Ng-LtgD were generated by AlphaFold. The Ng-LdcA homo-dimeric protein consists of two identical chains (Chain A in green and Chain B in red), whereas Ng-LtgD is a monomer (in green)

Fig. 3

Identification of binding pockets with DeepFold. The binding pocket (active site) predicted by DeepFold is shown for Ng-LdcA and Ng-LtgD. The active site regions show the conserved amino acid residues predicted to be involved in protein–ligand binding and conserving the architecture of the binding cavity. See text for description of the amino acids involved

Fig. 4

Molecular docking studies for Ng-LdcA. The left-hand exploded diagram shows the binding site at the surface of modelled protein for Ng-LdcA. On the right-hand side are images of compounds Ng-LdcA-16, -37 and -69 docked in the binding cavity. All the docked ligands exhibited good docking scores and retained all the conserved residues in the protein–ligand binding interaction

Fig. 5

Molecular docking studies for Ng-LtgD. The left-hand exploded diagram shows the binding site at the surface of modelled protein for Ng-LtgD. On the right-hand side are images of compounds Ng-LtgD-45, -52 and -69 docked in the binding cavity. All the docked ligands exhibit good docking scores and retained all the conserved residues in the protein–ligand binding interaction

Fig. 6

LigPlot⁺ 2D interaction diagrams for Ng-LdcA. The 2D interaction diagrams of the compounds docked with modelled Ng-LdcA proteins, plotted using LigPlot⁺. The interactions of all the molecules obtained by molecular docking are shown for compounds Ng-LdcA-16, -37 and -69

Fig. 7

LigPlot⁺ 2D interaction diagrams for Ng-LtgD. The 2D interaction diagrams of the compounds docked with modelled Ng-LtgD proteins, plotted using LigPlot⁺. The interactions of all the molecules obtained by molecular docking are shown for compounds Ng-LtgD-45, -52 and -69

Fig. 8

Plots to investigate the energy deviation, conformation stability and surface area accessible during simulation for Ng-LdcA and Ng-LtgD proteins and ligands in bound state with the protein. A represents RMSD, B represents radius of gyration, C represents Solvent Accessible Surface Area (SASA), and D represents number of hydrogen bonds. Black colour shows control, whereas red, green and blue colour shows 1st, 2nd and 3rd ligands respectively. Ng-LdcA 1st ligand = compound -16, 2nd = compound -37, 3rd = compound -69. Ng-LtgD 1st ligand = compound -45, 2nd = compound -52, 3rd = compound -69

Fig. 9

Binding free energy calculations using MM-PBSA tool for the potential small molecule inhibitor compounds along with the control. Colour coding is represented in the figure. Ng-LdcA 1st ligand = compound -16, 2nd = compound -37, 3rd = compound -69. Ng-LtgD 1st ligand = compound -45, 2nd = compound -52, 3rd = compound -69

Fig. 10

A Cytotoxicity of compounds. Human Chang conjunctival cells were treated with 50 µM (final concentration) of Ng-LcdA-16, -37 and -69 and Ng-LtgD-45, -52 and -69 compounds and cytotoxicity was measured using a standard resazurin assay. Controls included untreated cells, DMSO alone, medium alone, cell lysis (i.e. induced death) and ceftriaxone. The columns represent the means, and the error bars the standard error of the means of three independent experiments. B Determination of time to kill for compound Ng-LdcA-16. Bacteria (10⁵ CFU/well, n = 3) were treated with 50 µM (final concentration) of Ng-LdcA-16 and viable counts were made over time. Controls were bacteria alone, and bacteria treated with ceftriaxone (50 µM final concentration). Data are from one representative experiment of experiments done at least twice

Fig. 11

Activity of Ng-LdcA-16 and Ng-LtgD-45 compounds against other Neisseria spp. A Meningococci and B N. lactamica (10⁵ CFU/well, n = 3) were treated with various concentrations of compound Ng-LdcA-16 in the standard MIC and MBC assay, over 24 h. C Meningococci and D N. lactamica (10⁵ CFU/well, n = 3) were treated with various concentrations of compound Ng-LtgD-45 in the MBC assay in PBSB for 1 h with viable counting. Symbols represent the mean and any error bars the standard error of the means from three independent experiments