Figure 1.
Predicted membrane topology of the Dengue virus polyprotein and its cleavage sites.
The polyprotein is composed of three structural subunits: capsid (C), precursor of membrane protein (PrM) and envelope (E), as well as seven nonstructural (NS) subunits: NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5. The NS3 is a multifunctional protein composed of a helicase domain (NS3HEL, see detailed box) and a protease (NS3PRO, also observable in the detailed box) domain, which in turn needs the hydrophilic loop of NS2B (NS2BCF, marked in green) as a cofactor to be fully active. Red arrows indicate sites processed by the viral NS2B/NS3 protease. Based on Figure 1 from Umareddy et al., 2007 [12], for illustrative purposes only.
Figure 2.
Dengue and West Nile virus NS2BCF-Gly-NS3PRO multiple sequence alignment extracted from the protein databank files.
This alignment allowed us to verify the completeness of each deposed structure, as well as to fill structural gaps, direct N- (blue box) and C-terminus (red box) regions and the cofactor itself. Green triangles indicate the catalytic triad (HIS51, ASP75 and SER135); sequences are numbered accordingly to the full constructs and shaded based on sequence similarities (black for identical, dark gray for strongly similar and soft gray for weakly similar residues).
Figure 3.
Structural alignments between (A) core region from all templates used in homology modeling, and (B) obtained model and recently available ligand bound DENV NS3PRO structure (PDB id: 3U1I)
[36]. In (A), residues were colored based on their Qres factor, obtained after a STAMP structure alignment performed with the VMD Multiseq plugin [68]. Color ranges from dark blue (highly conserved positions) to red (not conserved at all). In (B), crystallographic structure was colored with gray (NS3PRO) and red (NS2BCF); homology model was colored in lighter shades – white (NS3PRO) and pink (NS2BCF). In addition to the highly conserved active site residues position, the oxyanion hole was also preserved in this model (boxed detail).
Figure 4.
Secondary structure timeline analysis as computed by the timeline plugin contained in VMD.
In the graphic, the β-sheet components turn (T) and extended conformation (E) are represented in teal and yellow respectively; isolated bridges are in dark yellow; degrees of helix are in pink (α-helix), blue (3-10 helix) and red (π-helix); random coils are in white.
Figure 5.
Single- and two-dimensional RMSD plots from the MD simulation trajectory, and their representative conformers.
In (A), RMSD values from the whole construct (NS3PRO F, black line) were compared with RMSD values from the protein core only (NS3PRO C, gray line). In (B), a two-dimensional RMSD plot was utilized to provide information about conformational families during trajectory. Color ranges from black (as out of the cut-out distance) to white (as in identical structures), with intermediate values depicted in the color scale. White stars mark the representative frames. (C) Snapshots of representative conformations of each family extracted from the clustered (two-dimensional) RMSD analysis (NS2BCF and linker in red, NS3PRO in gray, active site residues in green and NDL in orange).
Figure 6.
Individual residue mobility in principal component analysis.
(A) The greatest eigenvector represents NS3PRO C-terminal domain shifting, and the second eigenvector represents NS2BCF C-terminal and glycine linker shifts. (B) Normal mode visualization from the first eigenvector, also revealing small motions that may be important to binding site plasticity.
Figure 7.
Pocket detection and evolution during MD simulation trajectory.
Active site pocket (A) was identified by METAPOCKET and later monitored for changes in volume (B) by MD Pocket. In (A), DENV NS3PRO is represented in gray; NS2BCF and the glycine linker are in red. Active site residues are represented by green sticks, and the detected pocket is shown in a cyan surface. The pocket opposed to the active site, usually found in ligand bound structures (35) is depicted in a brown surface, merely for illustration. In (B), the red line is a smoothed curve of the black line, intended to clarify the breathing behavior of the active site.
Figure 8.
Binding mode fluctuations during the MD simulation.
Both ligand and binding site plasticity contributes to changes in the binding mode between the ligand and protein. The solvent accessible surface of an interacting residue or atom is represented by a blue halo around it. The diameter of the circle is proportional to the solvent accessible surface. Pink circles indicate residues involved in hydrogen-bond, polar or charged interactions; green circles indicates residues involved in van der Waals interactions; π-interactions are represented by orange lines; green arrows indicate hydrogen-bonds interactions with amino acid main chains; blue arrows indicate side chain hydrogen-bond interactions. Arrowhead directs towards the electron donor.
Figure 9.
Total pair interaction energies (potential plus kinetic).
(A) Pair interactions from the ligand NDL versus protein (black line), and NDL versus water (gray line). (B) Average total energy of pair interactions from the ligand and each protein residue, allowing for contact points mapping.