Biomedical Engineering Reference
In-Depth Information
cross-peak. To use the NOE as a distance-restraint the cross-peak has to be
assigned to a specific pair of atoms using their respective chemical shifts.
The peak-volume is proportional to r
26
, where r is the distance of the
involved protons. These distances are not very precise, however, since the
effects of spin-diffusion and motion are difficult to quantify accurately.
More precise distances can be obtained from NOE build-up curves
52
but the
experimental effort is larger.
For larger proteins NOESY cross-peak assignment cannot be resolved
unambiguously due to similar chemical shifts for different atoms and due to
overlap of multiple signals. The ambiguity and the overlap can be reduced by
introducing more dimensions to the NOESY spectrum. 3D and 4D NOESY
experiments are common, in which one or both of the proton frequencies are
labelled by the frequency of their covalently bound N or C atoms.
53,54
In this
way, structures up to ca. 15 kDa molecular weight are routinely solved with
good accuracy and a reasonable experimental effort.
4.3.1.1 Advantages
NOE experiments potentially yield many long-range (tertiary) contacts that are
readily interpretable as distance restraints. Even only a handful of
unambiguously assigned NOE distance restraints can be of high value for
structure calculation.
4.3.1.2 Disadvantages
NOE cross-peak assignment is ambiguous and thus error-prone. Even a single
mis-assigned long-range NOE can lead to significant structural distortions if
used at face value. No reasonable error distribution can model this mis-
assignment without losing all resolution of the restraint. Structure calculations
usually assume error-free assignments but might model ambiguity in the
restraints explicitly. If dynamics are present, shorter distances dominate the
NOE due to the r
26
averaging, which can lead to distorted and overly
restrained structures.
For high-molecular-weight structures (.20 kDa), NOEs are increasingly
difficult to obtain. The slow molecular tumbling time of larger molecules leads
to line broadening, whereas the larger number of residues exacerbates spectral
overlap. Perdeuteration of proteins greatly reduces overlap and improves the
signal-to-noise ratio by rendering the spin diffusion network sparse. However,
the removal of protons also results in a more sparse set of NOE-derived
distance restraints.
4.3.1.3 The Bottom Line
Despite the error-prone assignment and the low signal-to-noise ratio in high
molecular weight structures, NOEs remain the workhorse of de novo NMR
