Characterization of the binding of MRTX1133 as an avenue for the discovery of potential KRASG12D inhibitors for cancer therapy

Abstract

The Kirsten rat sarcoma (KRAS) oncoprotein has been on drug hunters list for decades now. Initially considered undruggable, recent advances have successfully broken the jinx through covalent inhibition that exploits the mutated cys12 in the switch II binding pocket (KRAS^G12C). Though this approach has achieved some level of success, patients with mutations other than cys12 are still uncatered for. KRAS^G12D is the most frequent KRAS mutated oncoprotein. It is only until recently, MRTX1133 has been discovered as a potential inhibitor of KRAS^G12D. This study seeks to unravel the structural binding mechanism of MRTX1133 as well as identify potential drug leads of KRAS^G12D based on structural binding characteristics of MRTX1133. It was revealed that MRTX1133 binding stabilizes the binding site by increasing the hydrophobicity which resultantly induced positive correlated movements of switches I and II which could disrupt their interaction with effector and regulatory proteins. Furthermore, MRTX1133 interacted with critical residues; Asp69 (- 4.54 kcal/mol), His95 (- 3.65 kcal/mol), Met72 (- 2.27 kcal/mol), Thr58 (- 2.23 kcal/mol), Gln99 (- 2.03 kcal/mol), Arg68 (- 1.67 kcal/mol), Tyr96 (- 1.59 kcal/mol), Tyr64 (- 1.34 kcal/mol), Gly60 (- 1.25 kcal/mol), Asp12 (- 1.04 kcal/mol), and Val9 (- 1.03 kcal/mol) that contributed significantly to the total free binding energy of - 73.23 kcal/mol. Pharmacophore-based virtual screening based on the structural binding mechanisms of MRTX1133 identified ZINC78453217, ZINC70875226 and ZINC64890902 as potential KRAS^G12D inhibitors. Further, structural optimisations and biochemical testing of these compounds would assist in the discovery of effective KRAS^G12D inhibitors.

Figure 1

Schematic diagram of the approach applied to this study. First four sections indicate the characterization approach applied to MRTX1133, while rest of the sections depict the process of hit compounds identification.

Figure 2

Comparative stability and solvent accessibility surface area of the binding site of KRAS^G12D upon the binding of MRTX1133 and the holo protein. (A) Shows the RMSD plots of the C-a atoms of the binding site residues indicating MRTX1133 binding stabilized the residues (red) relative to the holo (black). (B) Shows the SASA plots of the binding site residues indicating a reduction in surface area availability upon MRTX1133 binding (red) relative to the holo protein (black). (C) Shows the hydrophobicity of the binding site. (D) Shows the solvent accessible of area of the binding site.

Figure 3

Cross-correlation matrices of the fluctuations of coordinates of C–a atoms of the switches around their mean positions during the 450 ns simulation. DCCM plots of switch I MRTX1133 complex (A), Switch I holo (A1), Switch II MRTX1133 complex (B), and (B1) Switch II holo protein. This indicates MRTX1133 complexing induced a more corelated movement. (C) KRAS structure showing switch I (black) and switch II (red).

Figure 4

2D interactions of MRTX1133 with KRAS^G12D binding site residues. Snapshots depicts interactions at 100 ns, 200 ns, 300 ns and 450 ns. MRTX1133 engages in varying interactions.

Figure 5

Per-residue decomposition analysis showing binding site residue energy contributions towards the complexing of MRTX1133. Electrostatic energy featured prominently towards the total free binding energy. (A) Shows the energy plots. (B) Shows the interacting residues.

Figure 6

A pharmacophore model of MRTX1133. (A) 2D illustration of MRTX1133 interacting with binding site residues after 450 ns MD simulation. (B) The selected pharmacophoric moieties in MRTX1133, which interacted with the most negative energy contributing binding site residues. (C) PRED plot highlighting only the total binding energy contributed by the selected active site residues.

Figure 7

2D molecular interactions of MRTX1133 and selected compounds within the binding site of KRAS allosteric binding site. The selected compounds and MRTX1133 show similar interactions with the binding site residues suggesting the selected compounds have the potential to elicit similar therapeutic effects.

Figure 8

(A) Superimposition of KRAS^G12D complexed with ZINC78453217, ZINC70875226, ZINC64890902 and MRTX1133. Insert shows the binding orientation of each inhibitor at 100 ns, 300 ns, and 450 ns. (B) Comparative inhibitor stability plots using RMSD metrics for ZINC78453217, ZINC70875226, ZINC64890902 and MRTX1133. (C) Comparative inhibitor SASA plots of ZINC78453217, ZINC70875226, ZINC64890902 and MRTX1133.

Figure 9

Principal component analysis of the motions of the selected compounds at the allosteric binding site of KRAS^G12D.

Figure 10

2D interactions of hit compounds with KRAS^G12D binding site residues. Snapshots depicts interactions at 100 ns, 200 ns, 300 ns and 450 ns. (A) Shows ZINC78453217 interactions with the binding site residues. (A1) Shows ZINC70875226 interaction with the binding site residues. (A2) Shows ZINC64890902 interactions with the binding site interactions.

Figure 11

Comparative RMSD, RMSF and RoG plots of the hit compounds, MRTX1133 and unbound (holo) KRAS^G12D protein. (A) Shows the binding site RMSD elicited by the binding of the hit compounds. (B) Depicts the global stability of KRAS^G12D upon compounds binding. (C) Shows the residual fluctuations exhibited by the global protein residues during the simulation period. Highlights (deep yellow) show the switches regions. (D) Shows the RoG exhibited by the global protein. The bound conformations are contrasted to the unbound (Holo).

Figure 12

Most promising inhibitors of KRAS^G12D.