Fig 1.
Structure of the HPV-16 E6 protein homology model in interaction with LxxLL helical motif (green).
Residues belonging to the E6 pocket are shown in stick representation (red).
Fig 2.
Root mean square deviation (RMSD) and pocket volume analysis of apo-E6 and E6-hx systems.
a) RMSD analysis was carried out for the molecular dynamics simulations of each system considering both the Cα atoms of the whole protein (whole) and only those belonging to the pocket (Pocket). Pearson correlation results between whole and Pocket RMDS values were as follows. apo-E6: A1 = 0.98, A2 = 0.96, A3 = 0.88; E6-hx: A1 = 0.54, A2 = 0.56, A3 = 0.66. b) E6 pocket volumes throughout MD trajectories.
Fig 3.
Principal Component Analysis of the three apo-E6 trajectories.
a) Projection of the conformational distribution onto the subspace defined by the first two PCs of each assay (red: 0 ns to blue: 100 ns). b) Superimposed structures representing a trajectory obtained from the variance contribution of each residue in the first CP of each assay. The opening of the pocket is exhibited by the flexibility of the linker-helix joining the two Zn2+ binding domains. c) Conformational clusters of each apo-E6 trajectory displayed on the PCA scatter plot (left) and the RMSD plot (right) of each assay. Each dot represents an E6 conformer, and the yellow markers indicate the representative conformations (medoids) of each cluster.
Fig 4.
LxxLL binding pocket of the four E6 conformations selected for the EBD with Vina.
The residues structuring the pocket of each E6 conformer and the surface representation of their side chains are shown, highlighting the position of arginine residues. The volume and surface area values are also presented. Molecular graphics were produced by using UCSF Chimera software [62].
Fig 5.
Structure-Based virtual screening.
a) Results of the Ensemble-Based Docking using Vina. The 834 molecules with the most favorable binding values (red) were selected to be evaluated with AD4. For visualization purposes, only the results of three of the conformations (A1c, A2a, and A3a) are shown. Full results can be consulted in S9 Fig. b) Results of the evaluation of the 834 molecules docked to the A3a E6 conformation using AD4. The 100 ligands (green) with the lowest binding values were selected. The reference compounds were also analyzed. c) Rescoring of the AD4 top-ranked ligands using MM/PB(GB)SA. The Pareto front is defined by the red line connecting the three optimal solutions, corresponding to the three final candidate compounds: Lig1 (ZINC111606147), Lig2 (ZINC362643639), and Lig3 (ZINC96096545).
Table 1.
Screened candidate compounds and their pharmacokinetic properties.
Table 2.
Trajectory analysis for each of the E6-lig systems along 50 ns of MD.
Fig 6.
Ligand interaction diagrams of luteolin (a) and the three candidate compounds (b, c, d) with the E6 pocket residues.
For each ligand, the involving hydrogen bonds and its occupancy are highlighted. Additionally, the interaction structures at the beginning (0 ns) and the end (50 ns) of the MD production are presented. As shown in 2D diagrams, C58 was predicted as a deprotonated cysteine by PROPKA.
Table 3.
Trajectory analysis results for each of [E6+lig]-hx systems during 50 ns of MD.
Fig 7.
Efect of the docked compounds to E6 pocket over the E6-LxxLL binding affinity.
a) MM/GBSA ΔHbind decomposition per E6 pocket residue of each of the four [E6+lig]-hx systems and [E6]-hx system. The ΔHsubtotal (kcal/mol) values of each ligand are also presented (right). Full results can be consulted in S14 Fig. b) H-bonds occupancy between E6 pocket residues and LxxLL motif over 50 ns of MD. c) Sequence logo of E6 pocket of the most prevalent HR HPV types, depicting the degree of amino acid conservation at each position. Red letters indicate the corresponding residue of HPV-16 E6 protein.