Interpreting protein structural dynamics from NMR chemical shifts

Abstract

In this investigation, semiempirical NMR chemical shift prediction methods are used to evaluate the dynamically averaged values of backbone chemical shifts obtained from unbiased molecular dynamics (MD) simulations of proteins. MD-averaged chemical shift predictions generally improve agreement with experimental values when compared to predictions made from static X-ray structures. Improved chemical shift predictions result from population-weighted sampling of multiple conformational states and from sampling smaller fluctuations within conformational basins. Improved chemical shift predictions also result from discrete changes to conformations observed in X-ray structures, which may result from crystal contacts, and are not always reflective of conformational dynamics in solution. Chemical shifts are sensitive reporters of fluctuations in backbone and side chain torsional angles, and averaged (1)H chemical shifts are particularly sensitive reporters of fluctuations in aromatic ring positions and geometries of hydrogen bonds. In addition, poor predictions of MD-averaged chemical shifts can identify spurious conformations and motions observed in MD simulations that may result from force field deficiencies or insufficient sampling and can also suggest subsets of conformational space that are more consistent with experimental data. These results suggest that the analysis of dynamically averaged NMR chemical shifts from MD simulations can serve as a powerful approach for characterizing protein motions in atomistic detail.

Figure 1

Comparison of Sparta+ chemical shift predictions for ttRNH obtained from an X-ray structure and a 100 ns MD simulation in the amber99SB force field. For each residue for which an experimentally measured chemical shift (δExp) was available, the magnitude of the deviation between the X-ray predicted value (δX-ray) and the experimental value, | δX-ray - δExp |, and the magnitude of the deviation between the MD averaged prediction (δMD) and the experimental value, | δMD - δExp |, are compared. |δX-ray - δExp| - | δMD - δExp | is shown for Cα atoms in A and HN atoms in B. A positive value of |δX-ray - δExp| - |δMD - δExp| indicates that δMD is in better agreement with experiment, while a negative value indicates that δX-ray is in better agreement. Residues are colored according to their secondary structure in the X-ray structure with green, red, and yellow corresponding to coil, helix, and sheet respectively. Residues with aromatic side chains are displayed as cyan squares in B. The reported standard deviation of Sparta+ predictions obtained from a benchmark database of X-ray structures for Cα and HN atoms are displayed as dotted lines on for comparison.

Figure 2

Chemical shift predictions of Ala 145 Cα in a 100 ns MD simulation of ttRNH in the amber99SB force field. A) The value of the Sparta+ predicted chemical shift of Ala 145 Cα for snapshots saved every 4.5 ps of the MD trajectory. B) The normalized distribution of the Sparta+ predicted shifts of Ala 145 Cα from the MD trajectory. The Sparta+ prediction obtained from the X-ray structure (pdb code 1RIL) is shown as a red square, the average value of the Sparta+ predictions over the entire MD trajectory is shown as a blue square, and the experimentally measured value is shown as a black diamond. C) The conformation of Ala 145 observed in the X-ray structure (red) and from two representative MD snapshots (blue), along with the corresponding Sparta+ shift prediction of Ala 145 Cα for those conformations. Ala 145 Cα is colored cyan in the X-ray structure and magenta in the MD snapshots.

Figure 3

Chemical shift predictions of Glu 4 HN in a 1 μs MD simulation of ecRNH in the amber99SB force field. A) The value of the Sparta+ predicted chemical shift of Glu 4 HN for snapshots saved every 4.5ps of the MD trajectory. B) The value of the ψ dihedral angle of Val 3 (ψ₃) for snapshots saved every 4.5ps of the MD trajectory C) The normalized distribution of the Sparta+ predicted shifts of Glu 4 HN from the trajectory. The Sparta+ prediction obtained from the X-ray structure (pdb code 2RN2) is shown as a red square, the average value of the Sparta+ predictions over the entire MD trajectory is shown as a blue square, and the experimentally measured value is shown as a black diamond. D) The conformation of Glu 4 observed in the X-ray structure (red) and from two representative MD snapshots (blue), along with the corresponding Sparta+ shift prediction of Glu 4 HN and the ψ value of Val 3 for those conformations. Glu 4 HN is colored cyan in the X-ray structure and magenta in the MD snapshots.

Figure 4

Chemical shift predictions of His 28 HN in a 100 ns MD simulation of ttRNH in the amber99SB force field. A) The value of the Sparta+ predicted chemical shift of His 28 HN for snapshots saved every 4.5 ps of the MD trajectory. B) The value of the contribution of the ring currents to the chemical shift of His 28, δ_RC, as calculated by Sparta+ for snapshots saved every 4.5 ps of the MD trajectory C) The normalized distribution of the Sparta+ predicted shifts of His 28 HN from the MD trajectory. The Sparta+ prediction obtained from the X-ray structure (pdb code 1RIL) is shown as a red square, the average value of the Sparta+ predictions over the entire MD trajectory is shown as a blue square, and the experimentally measured value is shown as a black diamond. D) The conformation of His 28 observed in the X-ray structure (red) and from two representative MD snapshots (blue), along with the corresponding Sparta+ shift prediction of His 28 HN and the δ_RC value for those conformations. His 28 HN is colored cyan in the X-ray structure and magenta in the MD snapshots.

Figure 5

Chemical shift predictions of Val 5 N in a 450 ns MD simulation of ecRNH in the amber99SB-ILDN force field. A) The value of the Sparta+ predicted chemical shift of Val 5 N for snapshots saved every 4.5 ps of the MD trajectory. B) The value of the χ₁ dihedral angle of Val 5 for snapshots saved every 4.5 ps of the MD trajectory. C) The normalized distribution of the Sparta+ predicted shifts of Val 5 N from the MD trajectory. The Sparta+ prediction obtained from the X-ray structure (pdb code 2RN2) is shown as a red square, the average value of the Sparta+ predictions over the entire MD trajectory is shown as a blue square, and the experimentally measured value is shown as a black diamond. D) The conformation of Val 5 observed in the X-ray structure (red) and from two representative MD snapshots (blue), along with the corresponding SPARTA+ shift prediction of Val 5 N and Val 5 χ₁ value for those conformations. Val 5 N is colored cyan in the X-ray structure and magenta in the MD snapshots.

Figure 6

Chemical shift predictions of Phe 27 C′ in a 100 ns MD simulation of ttRNH in the amber99SB force field. A) The value of the Sparta+ predicted chemical shift of Phe 27 C′ for snapshots saved every 4.5 ps of the MD trajectory. B) The value of the φ dihedral angle of Phe 27 (φ₂₇) for snapshots saved every 4.5 ps of the MD trajectory. C) The normalized distribution of the Sparta+ predicted shifts of Phe 27 C′ from the MD trajectory. The Sparta+ prediction obtained from the X-ray structure (pdb code 1RIL) is shown as a red square, the average value of the Sparta+ predictions over the entire MD trajectory is shown as a blue square, and the experimentally measured value is shown as a black diamond. D) The conformation of Phe 27 observed in the X-ray structure (red) and from two representative MD snapshots (blue), along with the corresponding Sparta+ shift prediction of Phe 27 C′ and the φ value of Phe 27 (φ₂₇) for those conformations. Phe 27 C′ is colored cyan in the X-ray structure and magenta in the MD snapshots.

Figure 7

Chemical shift predictions of Leu 73 HN in a 100 ns MD simulation of ttRNH in the amber99SB force field. A) The value of the Sparta+ predicted chemical shift of Leu 73 HN for snapshots saved every 4.5 ps of the MD trajectory. B) The distance, in Å, between Leu 73 HN and the Sidechain hydroxyl O of Ser 70 for snapshots saved every 4.5 ps of the MD trajectory. C) The normalized distribution of the SPARTA+ predicted shifts of Leu 73 HN from the MD trajectory. The Sparta+ prediction obtained from the X-ray structure (pdb code 1RIL) is shown as a red square, the average value of the Sparta+ predictions over the entire MD trajectory is shown as a blue square, and the experimentally measured value is shown as a black diamond. D) The conformation of Leu 73 HN observed in the X-ray structure (red) and from two representative MD snapshots (blue), along with the corresponding Sparta+ shift prediction of Leu 73 HN for those conformations. Leu 73 HN is colored cyan in the X-ray structure and magenta in the MD snapshots.