Figure 1.
Graphical representation of the four distances for the Arg-Asp residue pair.
Two representative atoms were chosen for each residue: NH1, NH2 and OD1, OD2 for Arg and Asp, respectively. Four Euclidian distances are measured: d11 for the NH1–OD1 pair; d12 for the NH1–OD2 pair; d21 for the NH2–OD1 pair; d22 for the NH2–OD2 pair.
Figure 2.
Six projections of the four distances connecting Arg and Asp.
The four distances are: NH1–OD1 (d11), NH1–OD2 (d12), NH2–OD1 (d21), and NH2–OD2 (d22). Six projection are needed to plot the four dimensional data. Only cases with correct IUPAC atom name assignment protocol were considered.
Figure 3.
Only cases with correct IUPAC atom name assignment protocol were considered. (A) The CG2–CG1 distance (d21) has a higher occupancy at 4 Å than the CG1–CG1 distance (d11). (B) Val atom notation in the PDB. The CB atom is a prochiral atom, putting the CB hydrogen at the back, and starting at clockwise direction from the CA atom the following atoms are always CG1 then CG2. (B) The V shape objects represent the covalent bond between CG1 CG2 and CB in the Val side chain. The possible Val conformation with CG1–CG1 (d11) and CG2–CG1 (d21) distances of either at 4 Å or at 6 Å (C,D). Two histograms depicting the distance between CG1 (C) and CG2 (D) from the backbone nitrogen and oxygen. (C) CG1 atom is generally closer to the backbone oxygen. (D) CG2 atom is generally closer to backbone nitrogen.
Figure 4.
Asymmetry in 4-D distribution of His, Ile, Phe, Glu, Leu, and Pro residues.
Only cases with correct IUPAC atom name assignment protocol were considered. (A) The NE2–NE2 distance of the His-His pair has a higher occupancy around 3.5 Å compared to the ND1–ND1 distance. (B) The CD1–CG2 distance of the Ile-Ile pair has a higher occupancy around 4 Å compared to the CG2–CG2 distance. (C) The CE1–CE2 distance of the Phe-Phe pair has a higher occupancy at around 4 Å compared to the CE1–CE1 distance. The CD atom that makes the smaller torsion angle is named CD1 (see inset). (D) The OE1–CD distance between Glu and Pro has a higher occupancy at the 3–4 Å interval compared to the OE2–CD distance. (E) The CD1–CD1 distance of Leu-Leu has a higher occupancy at around 4 Å compared to the CD2–CD2 distance. (F) The CD1–CD1 distance of the Pro-Pro pair has a higher occupancy at around 4 Å compared to the CD2–CD2 distance.
Figure 5.
4-D histograms of the Arg–Asp pair (d12 versus d21) from different data sources.
(A) The four distances for the Arg–Asp residue pair. (B) High-resolution X-ray data (<2 Å). (C) High-resolution X-ray data (<2 Å) collected at temperature between 275 to 300 Kelvin. (D) Low resolution X-ray data (2.5–3.0 Å). (E) NMR data. (F) Same data set as in (A) though all rotamers were randomly shuffled.
Figure 6.
Building a random model for the KBP.
(A) Data collected using a 5 Å cutoff (i.e. at least one distance <5 Å) for real data (blue) and random data (red). (B) Data collected using a 30 Å cutoff. (C) and (D) Log ratio of real and random distribution for (A) and (B) respectively.
Figure 7.
Bin counts for the real and random distributions of the Leu–Val pair. The y-axis is the normalized amount of data. The amount of data is the product of the number of measurements in a certain bin times the number of bins with the same number of measurements in this histogram.
Figure 8.
Optimal arrangement of Arg and Asp side chains simulated from the derived 4-D histograms.
In the preferred orientation the carboxyl group of Asp is closer to NH2 of Arg than to NH1. To generate the depicted orientation, two residues in arbitrary conformation were placed randomly relative to each other, multiple times. The top scoring orientation, according to Equation 2, was identified and drawn in the figure. This simulation was repeated several times resulting in similar asymmetric placement of Arg and Asp side chains.