Chemistry Reference
In-Depth Information
reducing the amount of experimental effort. However, a virtual screen does have
significant limitations that prevent it from completely replacing traditional HTS
[
50
-
52
]. These limitations include inaccurate scoring functions, use of rigid
proteins, and simplified solvation models. In essence, a virtual screen only increases
the likelihood that a predicted ligand actually binds the protein target, experi-
mental verification is essential. Despite the individual drawbacks, NMR ligand
affinity screens and molecular docking are complementary techniques. This review
will highlight the use of NMR ligand affinity screens and molecular docking in
drug discovery and describe recent examples where the combination of the two
techniques provides a powerful approach to identify new and effective therapeutic
drugs.
2 NMR Ligand Affinity Screens
NMR ligand affinity screening is a versatile technique that is useful for multiple
stages of the drug discovery process [
15
,
17
,
22
,
53
]. This versatility arises from the
ability of NMR to directly detect protein-ligand binding based on changes in
several NMR parameters. A binding event is detected by the relative differences
between the protein or ligand NMR spectrum in the bound and unbound states.
However, the specific type of information obtained about the binding process
depends on whether a ligand-based or target-based NMR experiment is used.
2.1 Ligand-Based NMR Screens
Ligand-based NMR screens typically monitor the NMR spectrum of a ligand
under free and bound conditions. Distinguishing between a free ligand and
a protein-ligand complex is generally based on the large molecular weight differ-
ence that affects several NMR parameters. Small molecular weight molecules have
slow relaxation rates (R
2
), negative NOE cross-peaks, and large translational
diffusion coefficients (D
t
). If a protein-ligand binding event occurs, the ligand
adopts the properties of the larger molecular-weight protein, increasing R
2
, produc-
ing positive NOE cross-peaks, and decreasing D
t
, all of which can be observed by
NMR [
54
]. Most ligand-based NMR screens use one-dimensional (1D)
1
H-NMR
experiments to monitor these changes, which provide significant benefits for a high-
throughput screen. 1D NMR experiments are typically fast (2-5 min) and routinely
use mixtures without the need to deconvolute [
55
]. The deconvolution of mixtures
is avoided by ensuring that NMR ligand peaks do not overlap in the NMR spectrum
(Fig.
1
). The application of mixtures allows for hundreds to thousands of
compounds to be screened in a single day. Another advantage of ligand-based
NMR methods is the minimal amount of protein required (
M) for each
experiment. Additionally, isotopically labeled proteins are not needed for the
<
10
m
