Biomedical Engineering Reference
In-Depth Information
The results identified an alternating pattern for the amplitude of the c-and
b-motions in the b-sheet. The motional amplitudes increase from one edge of
the sheet to the other, providing maximal dynamic sampling at the interaction
site, suggesting that conformational selection plays a role in the interaction
with the physiological partner F
ab
. Extensive cross-validation against data that
were not used in the analysis corroborated these findings, and correlations
between motions of peptides planes involved in hydrogen bonds across the b-
sheet were found from
3
J
N9C
trans-hydrogen-bond scalar couplings. These
results demonstrate how proteins can transfer information over long distances
via correlated motions, and have implications for our understanding of
allosteric regulation. The dependence on the structural model used to represent
the average coordinates in solution was shown to be minimal, by repeating the
procedure using an NMR structure refined using many of the RDCs used in
the dynamic analysis.
63
It is apparent from these results, and from the nature of the expression
shown in eqn (8.2), that it is in principle possible to simultaneously determine
of the average structure and the associated dynamic modes and amplitudes
describing motions about this mean. This was demonstrated using a combined
approach, called Dynamic Meccano, that determines all tensor parameters, the
average structure, and the principle motions about this mean, using only
RDCs.
64,73,74
Application of a 3DGAF analysis of the same backbone data
from GB3 led to very similar results to the previous analysis determined using
the high-resolution crystal structure. The backbone coordinates were
remarkably similar (backbone rmsd ,0.5
˚
) to high resolution X-ray
crystallographic and NMR structures, suggesting that main-chain dynamics
in this protein occur in a pseudo simple-harmonic potential. This approach
was cross-validated using RDCs that were not used in the analysis (for
example couplings between a-carbons and protons), demonstrating that the
dynamic description better reproduced independent data than an optimally
applied static approach.
64
As described above, the power of RDCs to simultaneously define high-
resolution structure and long-timescale dynamics is now established. However
the quantitative determination of these dynamics remains challenging. This
issue is important, because little is known about the absolute level of slow
dynamic transitions in proteins, and because RDCs are sensitive to errors of
absolute dynamic amplitude, due to the potential for a component of the
dynamic averaging of RDCs to be absorbed into the estimated magnitudes of
the alignment tensors. When considering how well one can determine the
absolute level of dynamic fluctuations on timescales up to the millisecond, a
number of potential sources of artefacts should be considered. These include
the accurate estimation of the alignment of the protein, the influence of the
alignment medium on protein structure and dynamics, the influence of noise,
and the dependence on the coordinates of the average structure used in the
analysis. These questions were addressed in a subsequent study of ubiquitin,
75
using the coherent data sets that had been identified from the SECONDA
