Fig 1.
Comparative analysis of thermal stability plots of recombinant BmMetRS with ligands.
Thermal shift assays Tm curves for apo form (control) of the BmMetRS (red plot) and inhibitors bound with 1312 (amber), 1433 (blue) and 1415 (green) plots. A significant shift in melting temperature (ΔTm) of 7.8°C was observed when compound 1312 was complexed with BmMetRS compared to the Apo BmMetRS (unbound). Two other compounds, 1415 and 1433, exhibited ΔTms of 3.1°C and 2.5°C respectively.
Fig 2.
The lowest effective concentration of a methionyl-tRNA synthetase inhibitor required to inhibit the growth of B. melitensis strain 16M (MIC) was assessed from values observed over concentrations ranging from 0.716–71.6 μg/mL for compound 1433, 0.696–69.6 μg/mL for 1415 while 1312 was determined at inhibitor concentration 0.865–37.6 μg/mL. Average growth (OD450 nm) at 28 hours (late log phase), 37°C, in Brucella minimal medium was measured. Inhibitor 1415 MIC = 53.93μg/mL; Inhibitor 1433 MIC = 38.67 μg/mL; Inhibitor 1312, MIC = 28.96 μg/mL. Control = gentamicin, no growth, 0.7–50 μg/mL; DMSO control has no inhibition of bacterial growth. Data table shows MICs from B. melitensis growth inhibition and IC50s of inhibitors (Kinase-Glo®, aminoacylation assays). The characteristics of each inhibitor experimentally measured against human mitochondria MetRS, EcMetRS, human liver hepatocellular (Hep G2) and lymphocytic (CRL-8155) cell lines are also shown.
Table 1.
OD450 counts and growth inhibitory effects on B. melitensis exposed to MetRS inhibitors.
Table 2.
Data collection and model refinement statistics.
Fig 3.
A. MetRS bound to SeMet. Structural features of BmMetRS include a catalytic domain formed by a Rossmann fold (red), inserted connective peptide (CP) zinc finger domain (blue), a stem-contact fold (SCF) domain (magenta), and an anti-codon binding α-helix bundle (green). B. Overlay of MetRS bound to SeMet and MetRS bound to 1312. MetRS bound to 1312 is colored in yellow. Domains in MetRS bound to SeMet are colored identically to Fig 3a above. Movement of the CP domain is highlighted by the arrow.
Fig 4.
Crystal structure of BmMetRS in-complex with selenomethione and three different inhibitors to define structural pockets.
The binding of each ligand; selenomethionine (Fig 4a), 1312 (Fig 4b), 1415 (Fig 4c) and 1433 (Fig 4d) in the crystal structure at the active site causes similar conformational changes to the residues surrounding the methionine-binding site.
Fig 5.
Amino acid sequence alignments within the benzyl and quinolone pocket of BmMetRS, TbMetRS and LmMetRS.
Inhibitor binding residues in BmMetRS compared to TbMetRS and LmMetRS are highlighted in red boxes. The most significant differences can be found in residues 211–213 (TTF) of the BmMetRS which are analogous to a larger run of residues (TbMetRS: KRETLH LmMetRS: KRESVM) in the trypanosome structures. Inhibitor interaction with Phe213 within the BmMetRS complex led to different protein geometry relative to the TbMetRS complex. Table shows list of residues within the benzyl and quinolone pockets interacting with inhibitors relative to other MetRS. Sequence numbers refer to the BmMetRS sequence. # LR = linker region, BP = benzyl pocket, QP = quinolone pocket.
Fig 6.
Compound 1312 additional hydrogen bonding of its ketone group with ASP51.
The crucial hydrogen bonds in all 3 inhibitors with Asp51 are conserved in all BmMetRS inhibitor complexes. In addition, the 4-ketone group in the aminoquinolone compound 1312 makes water-mediated interactions (water molecule is shown as red sphere) with the phenol hydroxyl of Tyr228.
Fig 7.
A superposition of BmMetRS (PDB ID 4PLY2 Chain B) and TbMetRS (PDB ID 4MVW Chain) bound to compound 1433.
A key difference is the interaction of BmMetRS Phe213 which is functionally equivalent to Leu456 in TbMetRS but led to different protein geometry. The TbMetRS structure is shown in blue and the 2 different residues in orange.
Fig 8.
Homology model of human mitochondrial MetRS overlaid with the crystal structure of BmMetRS in complex with 1433.
Very few differences could be observed between amino acid residues within the inhibitor binding site and interaction of the compound in the 2 enzymes. This fits well with experimental data showing similar level of inhibition between BmMetRS and human mitochondrial MetRS enzymes.