Figure 1.
The time evolution of the RMSD values in the apo, ADP and ATP simulations.
Only Cα atoms are taken into account. The model linker between NTD and M-domain, as well as the ATP lid, are not included in the calculation.
Figure 2.
The time evolution of the RMSD values for individual domains of the dimer.
Only Cα atoms are taken into account. From left to right: NTD, M-domain and CTD; from top to bottom, apo, ADP and ATP simulation.
Figure 3.
The average fluctuation per residue calculated over the 70 ns trajectory.
For each simulation, the two protomers forming the dimer are plotted on the same graph, with each of them spanning residues 1–677. From top to bottom: apo, ADP and ATP simulation.
Figure 4.
Cross-Correlation matrices calculated considering the motion of Cα atoms around the average position.
A) Cross-Correlation Matrix for the Hsp90-ATP complex; B) Cross-Correlation Matrix for the Hsp90-ADP complex; C) Cross-Correlation Matrix for apo Hsp90. A correlation close to 1 (color code: yellow) corresponds to highly coordinated motion of the atom pair along the same direction, whereas a negative correlation (color code: blue) indicates motion in opposite directions. The 3D structure of the full-length Hsp90 is reported for reference, with the different domains highlighted in different colours: blue, N-terminal domain; green, M-Domain; orange, C-Domain.
Figure 5.
Time evolution of the N terminal interface distance.
The interface distance is defined as the distance between the center of mass of the two residue groups from each protomer involved in the interface.
Figure 6.
Structural representation of the N-terminal interface.
The structures represent the most populated conformations obtained from the cluster analysis of the liganded dimers. The interface residues involved in hydrophobic packing are highlighted with a sphere representation (residues 22, 23, 24, 376, 378, 380). In the presence of ATP the interface becomes tighter during the simulation (A). In the ADP simulation the hydrophobic packing is disrupted (B).
Figure 7.
Structural detail of the interface showing the difference between the ATP and the ADP simulation.
(A) In the presence of ATP the loop containing Y24 protrudes from one protomer to the other and the two tyrosines are pointing to each other. Helices 1, depicted in red in both protomers, are nearly parallel. (B) In the ADP simulation, Y24 of one protomer rotates away from the interface towards its nucleotide binding site, which results in an increased tilting of helices 1(in red) and a relative displacement of the other protomer.
Figure 8.
Time evolution of the tilt angle between the two N-Terminal helices 1.
The figure represents the time evolution of the cosine of the dihedral angle between the axis helix 1 from one protomer and the axis of the second helix 1 from the other protomer. In the ADP simulation a clear variation of the cosine of the angle can be observed, identifying the increase of the tilt and the unwrapping of the helices described in the text.
Table 1.
Flexibility parameters for different domains of full-length Hsp90 bound to different ligands.
Figure 9.
The average residue fluctuations projected on the main ED eigenvector.
This quantity is calculated by projecting the trajectory over the first eigenvector according to the ED analysis. For each simulation, the two protomers (each spanning residues 1–677) forming the dimer are plotted on the same graph. From top to bottom: apo, ADP and ATP simulation.
Figure 10.
Structural analysis of the ATP-bound Hsp90.
Different consecutive snapshots from the MD trajectory in the ATP simulation are projected along the main ED eigenvector and their structures superimposed. Regions undergoing relevant correlated motions are highlighted in colors; they involve the ATP lid of both protomers and helices 3 at the CTDs. The color code from green to blue reflects the time evolution of the trajectory.
Figure 11.
Structural analysis of the ADP-bound Hsp90.
Different consecutive snapshots from the MD trajectory in the ADP simulation are projected along the main ED eigenvector and their structures superimposed. Regions undergoing relevant correlated motions are highlighted in colors. With respect to the ATP case (Figure 10), the motions are overall less pronounced and involve mainly the M-domain and the helices at the CTD interface. The color code from green to blue reflects the time evolution of the trajectory.
Figure 12.
The histogram analysis of the communication efficiency.
(A) The histogram for the Hsp90-ATP complex; (B) The histogram for the Hsp90-ADP complex; (C) The histogram for apo Hsp90. Each bin refers to a residue and shows the fraction of residues of the whole protein that are highly prone to communicate with it, i.e. such that their pair CP value is below 0.025. In each histogram only communications at distances greater than a given threshold are considered, namely over 40, over 60 and over 80 Angstroms.
Figure 13.
Structural analysis of the communicating residues in the Hsp90 dimer.
Residues active in long range signalling (i.e. active in communication with other residues that are more than 40 Angstroms apart) are coloured according to the following scheme: yellow: Residues active both in ATP and ADP system. Red: Residues active in the ATP system. Blue: Residues active in the ADP system. Mutation sites leading to different phenotypes [49] are highlighted with spheres and labelled.
Figure 14.
Representation on the 3D structure of residues communicating with high efficiency.
Two views of Hsp90 dimer, where residues active in very long range signalling (i.e. active in communication with other residues that are more than 80 Angstroms apart) are coloured according to the following scheme: Yellow: Residues active both in ATP and ADP system. Red: Residues active in the ATP system. Blue: Residues active in the ADP system. The mutation site A577 leading to loss of the ATPase activity is highlighted with spheres and labelled.