Fig 1.
Ribbon model of the NS3/4A protease domain based on the crystal structure with PDB ID: 4JMY [36] (A). The β-sheet of the NS4A cofactor is in yellow and NS3 membrane contact sites are in magenta. The zinc coordinating residues (B) and the catalytic triad (C) are also shown.
Fig 2.
The all-atom model of PEG (A) and Ficoll (B) crowders used in MD simulations.
Fig 3.
Snapshots of the systems simulated in this study.
The NS3/4A enzyme is shown in brown. Substrates are shown in purple. Crowder molecules are shown in grey, but selected molecules are colored to indicate their size (PEG: green, Ficoll: blue). For clarity of the figure, water molecules and ions are not shown.
Table 1.
Translational diffusion of NS3/4A, crowders, and substrates.
Table 2.
Rotational diffusion of NS3/4A, crowders, and substrates.
Fig 4.
Radial distributions of crowder and substrate atoms around NS3 from simulations with PEG (green) or Ficoll (blue) crowders or in the absence of crowders (red).
Distributions are shown for crowder heavy atoms (A), crowder oxygen atoms (B), and for substrate heavy atoms (C). Dashed lines in A and B indicate crowder distributions in the presence of substrates. Distances from the crowder or substrate atoms to the closest NS3 atom were counted and normalized by the total available volume at a given distance from the NS3 surface. For the simulations with only substrates (C), the distribution functions were scaled by a factor of 2 to account for the difference in substrate molarity between the systems without and with crowders (see S1 Table). The dotted line shows the results obtained without scaling. Distribution functions were averaged over all trajectories for the systems with the crowders but without substrates. The shaded areas indicate standard errors.
Fig 5.
Relative distribution of contacts between PEG and Ficoll crowders and NS3 according to the NS3 amino acids that crowders are in contact with.
A contact is defined based on a minimum distance between a crowder oxygen and the closest NS3 heavy atom of less than 5 Å.
Fig 6.
Surface representation of NS3 colored by frequency of crowder and substrate contacts per residue.
Different surface colors indicate interactions with PEG (green), Ficoll (blue), and substrates (purple). Different views were generated by rotating the NS3 structure as indicated in the figure. A cartoon representation of NS3 in the same orientations as the surfaces, colored according to secondary structure elements and with van der Waals sphere representation of the catalytic residues, is shown for reference. For clarity, higher number of contacts up to a maximum value of 0.5 contacts per residue are shown at the fully saturated color levels for PEG and Ficoll contacts.
Table 3.
Average number of contacts and hydrogen bonds between NS3, substrates, and crowders.
Fig 7.
(A) Radial distribution of PEG (green) or Ficoll (blue) crowder heavy atoms around substrates. Distances from the crowder atoms to the closest substrate atom were counted and normalized by the total available volume at a given distance from the NS3 surface. (B) Contacts with PEG (green) or Ficoll (blue) per substrate residue along the substrate sequence. Trajectory averages are shown as solid lines with the shaded areas indicating standard errors.
Fig 8.
Radial distribution of PEG (green) or Ficoll (blue) crowder heavy atoms (top row) or substrate heavy atoms (bottom row) in simulations with PEG (green), Ficoll (blue) or without crowders (red) around NS3 active site residues. Distances from the crowder atoms to the closest substrate atom were counted and normalized by the total available volume at a given distance from the given NS3 residues. The distribution function of substrates in the simulations without crowders are not scaled; note that substrates are present at twice the overall concentration in those systems. Dashed lines in the top row panels indicate crowder distributions in simulations with substrate. Trajectory averages are shown as lines with the shaded areas indicating standard errors.
Fig 9.
Relative probabilities of sampling different values of root mean-square deviations (RMSD) of heavy atoms of active site residues (57, 81, 139).
RMSD values were calculated with respect to experimental reference structure (PDB ID: 4JMY) after optimal superposition of the respective residues. Relative probabilities were shifted so that the most likely distance corresponds to a log(p) value of 0. Results are shown for simulations in water (red), with only substrates (purple), and in the presence of PEG (green) or Ficoll (blue) crowders. The top panel (A) compares distributions in water with distributions when crowders are in contact (<5 Å). Solid lines show results without substrate, dashed lines show results when substrates are present along with the crowders. In the bottom panel (B), distributions are shown for cases where the substrate is in contact (<3 Å). For comparison, the dashed line shows the distribution in water. Trajectory averages are shown as lines with the shaded areas indicating standard errors.
Fig 10.
A cartoon representation of NS3/4A is shown for reference with labels identifying key residues in the active site and those involved in substrate binding (A). Substrate binding near NS3 active site from simulations is indicated without crowders (B), with Ficoll crowders (C) and with PEG crowders (D). Substrate conformations shown are representative structures closest to cluster centers for the largest clusters covering the top 25% (dark magenta) and the next 25% (light magenta) of all substrate conformations extracted for a given system. Surface representations of NS3 are colored based on crowder contact frequencies, with Ficoll (C, blue) and with PEG (D, green). The bottom row of figures (E-H) shows the representations in A-D after rotation to the left by about 45°.