Chemistry Reference
In-Depth Information
dipole-dipole interaction between an unpaired electron and a proton will dominate over all
other relaxationmechanisms and is effective over much longer through-space distances than
is typical for
1
H-
1
H relaxation (up to 15-20 Å). Further, the strong transverse relaxation
effect makes it possible to use low protein concentrations, on the order of a factor 10 lower
than in a normal transverse relaxation filter experiment.
[
35
]
The main disadvantage of the
method is the need to introduce the spin label at a suitable amino acid side-chain near the
binding site of interest. There is a risk that the introduction of the spin label might alter
the structure and/or the binding properties of the target protein. An attractive variant of the
method is the use of second-site screening. A spin-label is then attached to a known ligand
and used to detect fragments binding to a site adjacent to the binding site of the spin-labeled
ligand. The quenching effect of the spin label on compounds is observed if, and only if,
both the spin-labeled compound and the other compound bind at the same time and in the
vicinity of each other. Here, it is important that the attachment of the spin label does not
alter either the first or the second binding site.
Fluorine relaxation filter.
The
19
F nucleus exhibits a number of features that make it attract-
ive to use for detection in NMR binding experiments:
[
88
]
(i) the
19
F transverse relaxation
rate and chemical shift are very sensitive to changes in the microenvironment (e.g. upon
ligand-target complex formation), (ii)
19
F occurs at 100% natural abundance and has a
gyromagnetic ratio nearly as high as that for
1
H and (iii) the absence of
19
F in biological
macromolecules, most organic compounds and buffer components results in very clean
19
F
spectra. However, direct observation of
19
F in a primary fragment screen would require
that all fragments contain at least one fluorine atom. In practice, therefore,
19
F detection
in screening campaigns is applied in competition experiments where a fluorine containing
'spy' molecule is employed.
[
86, 87
]
The observed parameter of the spy molecule is typically
the
19
F signal intensity after a spin-echo filter.
4.5.2 Protein-detected Techniques
Protein-detected techniques rely on detecting changes in the NMR observables of a pro-
tein upon exposure to a ligand. The best known approach is the pioneering SAR-by-NMR
fragment-linking scheme introduced by Fesik and co-workers in 1996.
[
32
]
In the first step
of the scheme, ligands are identified by monitoring alterations of target signals in a 2D
1
H-
15
N correlation spectrum. The spectrum can be regarded as a residue-resolved map
over the entire protein backbone (except prolines). The amide resonances in the binding
site of the ligand usually show the largest changes as compared with a spectrum of the
apo
-protein. If a cocktail of fragments is tested, then deconvolution of the cocktail will
be necessary to identify the active fragment. Protein-detected techniques require isotop-
ically enriched samples, but have the benefit (provided that sequence-specific resonance
assignments have been obtained) of being able to identify directly the binding epitope on
the target, which also makes it possible to distinguish between specific and nonspecific
binding events. It is also possible to assess whether any significant conformational changes
of the target protein occur upon binding. Further, protein-detected methods do not rely on
fast exchange to retrieve information from the bound state, making it possible to detect
both low- and high-affinity hits. The major drawbacks of protein-detected methods are the
