Biomedical Engineering Reference
In-Depth Information
reversible interaction with the target but rather non-specific effects such as
interference with the assay itself or (commonly) aggregation.
45
Thus, there is
an emerging need to triage the list of HTS hits. At Pfizer, Miller and colleagues
have conducted hit-prioritisation campaigns using biochemical and biophysi-
cal techniques.
46
The first validation step consists of an orthogonal enzymatic
assay to validate the results of the HTS screen. Then, to ensure integrity, a re-
synthesis step of the compound is carried out followed by SAR studies.
Finally, STD and isothermal titration calorimetry (ITC) are used in parallel as
a last step of validation. This extensive triage of compounds allows
classification in distinct series and the majority of unwanted compounds are
eliminated in the early stages. One caveat is that in our experience, many HTS
hits have very limited aqueous solubility rendering ligand-observed NMR
challenging.
As hits from a fragment or HTS screen are progressed, the affinity typically
drops into the nM range. In this range k
off
becomes slow on the NMR
timescale and the K
D
is lower than the protein concentration typically required
for either ligand- or protein-observed NMR. In such cases, determining the
affinity using NMR is problematic. However, in a series of compounds often
the relative affinities are more important than the absolute affinities. Jahnke
and colleagues have developed a method to acquire the relative affinity of two
compounds whose absolute affinities may be close or widely separated.
47
The
method can be implemented using either protein- or ligand-observed spectro-
scopy and has proven useful in the elaboration of fragments. Although the
protein-observed method is the most robust, it is more restrictive than the
ligand-observed method.
11.4 Protein-Ligand Structures
Structure-based drug design has become the de facto standard method for
targets in which high-throughput crystallography is enabled. An analysisof
FBDD projects suggested that the success rate (as measured by achieving
inhibition (IC
50
) better than 100 nM) was three-fold higher when three-
dimensional (3D) structures of protein-ligand complexes were available.
48
NMR and X-ray crystallography remain the only methods for high-resolution
protein structure determination while, for commercial drug discovery, itis
imperative that this information is rapidly available. There have been a
number of exciting developments that are beginning to enable NMR to
provide structural information in the timeframe demanded. Below, we point
out a few that are, in our opinion, the most exciting.
While ultimately it is ideal to have high-resolution 3D structures of the
protein-ligand complex, information about the conformation of the ligand
itself can be valuable. Transferred NOE (trNOE) is a simple technique that can
be applied to a variety of targets. A NOESY spectrum of a ligand in rapid
exchange with the target is acquired under conditions of ligand excess. NOEs
between
1
Hs of the ligand only build up when bound to the larger protein. The
