Page 91 - TD-4-1

Tumor Discovery Identification of a potential KRAS(G12C) inhibitor

A B

Figure 1. (A) The binding modes of the X-ray-bound Sotorasib and the re-docked Sotorasib were superimposed. (B) Amino acid interactions of Sotorasib
with Kirsten rat sarcoma viral oncogene homolog(G12C) and compound prioritization based on molecular dynamics simulations and binding free energy
calculations.

Table 2. The binding affinities and physicochemical properties of KRAS-C01, KRAS-C02a, KRAS(G12C)-C02b, and KRAS-C03
resembling the chemical scaffold of Sotorasib. TPSA: vdW polar surface area, LogP: Predicted octanol/water partition coefficient
Compound number Compound name Similarity Glide score (kcal/mol) Prime energy Molecular weight TPSA LogP
Sotorasib 1.00 −8.157 −19.6 560.6 102 4
C01 Shen_92298609_93443091_95741143 0.917 −7.801 −78.3 605.770 95.660 7.178
C02a Shen_92298609_44473994_95741143 0.916 −7.596 −16.7 600.731 108.550 6.512
C02b Shen_92298609_36250864_95741143 0.916 −7.347 −22.2 600.731 108.550 6.512
C03 Shen_92298609_50382533_8869305 0.916 −7.074 −5.6 611.675 130.480 4.944
Abbreviations: TPSA: Topological polar surface area.

candidates with the clinically approved drug Sotorasib value ranging from 0.2 to 0.4 nm throughout the entire
using metrics such as RMSD and MM/GBSA to assess their simulation. Thus, based on the RMSD of the ligand after
binding strength. The 100 ns MD simulations revealed least square fitting to the protein backbone, KRAS(G12C)-
that, apart from KRAS-C02a and KRAS-C03, all complexes C02b exhibits a therapeutic candidate attribute comparable
maintained an average RMSD value ranging from 0.2 to to the reference complex, KRAS(G12C)-Sotorasib, in that
0.4 nm, indicating a relatively stable simulation completion it maintains binding stability and shows no evidence of
that was closely matched with their initial conformations dissociation from the binding site.
(Figure 3). Furthermore, KRAS-C01 was unable to To understand the fluctuations at the per residue
maintain its stability in subsequent 300 ns MD simulations level upon ligand binding, root mean-square fluctuation
(Figure 3). Complexes such as KRAS(G12C)-C02b and (RMSF) calculations were performed. Consistent with
KRAS(G12C)-Sotorasib, on the other hand, successfully previously published findings, 26,31,46 it was observed
completed the simulation with RMSD values ranging that, in contrast to the entire protein, fluctuations were
from 0.2 to 0.4 nm (Figure 3). Continuing the 500 ns MD predominantly concentrated around switch-I (residues
simulation, KRAS(G12C)-C02b maintained its stability 30 – 38) and switch-II (residues 60 – 76). In Figure 4A,
until around 450 ns when its RMSD ranged from 0.3 to high peaks were noted in the RMSF graph within the
0.6 nm near the end of the simulation at 500 ns (Figure 3). region encompassing switch-I and switch-II loops in both
Of note, KRAS(G12C)-Sotorasib maintained an RMSD complexes. Furthermore, hydrogen bond occupancy,

Volume 4 Issue 1 (2025) 83 doi: 10.36922/td.5163