Chemistry Reference
In-Depth Information
In summary, the fragment screening campaign very efficiently generated a series of three
compounds for A-FABP which fulfill the hit criteria that are normally set in an HTS, i.e.
at least 10 M potency and reasonable selectivity. In addition, these compounds are small,
have high ligand efficiencies and are highly soluble. Hence they allow for further expansion,
in order to generate molecules with improved affinity and selectivity, without the imme-
diate risk of obtaining drug candidates too large and lipophilic to become optimal drugs.
The generation of highly soluble hits makes it possible to obtain structures of ligand-target
complexes very early in the drug discovery process (pre-HTS in this case) and learn more
about key interactions in the binding pocket, which promises to streamline the upcoming
optimization efforts of compounds originating both from fragment screens and HTS. Fur-
thermore, the compounds generated from fragment screening could constitute very different
classes of compounds to what is obtained from screening of traditional HTS libraries. In
this context, it is interesting to note that Vertex also has published results of an NMR screen
using the SHAPES library of A-FABP.
[
6
]
The primary screen yielded 13 hits with affinities
ranging from 0.3 to 800 M and two crystal structures of ligand-target complexes were
obtained. Based on the structural information, 134 commercially available analogues were
purchased and tested in a calorimetric screen. Nine compounds with low micromolar to
nanomolar potencies were identified and they were different from those presented here.
As was noted by Lepre
et al
.,
[
6
]
although both groups applied very similar approaches to
exactly the same target protein, the resulting lead series were very different. This illustrates
that fragment-based methods are particularly effective for discovering
novel
leads.
4.6.2 Expansion of Primary Fragment Hit - Ser/Thr Kinase
In general, it is difficult to develop binding fragments into high potency lead molecules
without structural information such as crystal structures of fragment-target complexes.
This example, however, serves to illustrate that in favorable cases it is possible to obtain
promising hit and lead series by fragment-based screening despite a lack of both structural
information and synthetic chemistry resources.
Protein kinases have become the second most important class of drug targets, after
G-protein-coupled receptors.
[
127
]
The main therapeutic areas include oncology, inflamma-
tion and metabolic diseases. As most protein kinase inhibitors target the ATP binding site,
not only specificity but also intellectual property is an important issue. There are several
compound libraries targeted at the ATP binding site of protein kinases on the market, but
there are also many pharmaceutical companies using these libraries, leading to a crowded
patent situation. With this in mind, fragment-based screening should have an important
role to play in kinase inhibitor discovery, especially when it comes to finding novel and
patentable scaffolds.
The Ser/Thr kinase target was a novel target: no small molecule inhibitors had been pub-
lished and no crystal structure was available. The fragment screening by NMR took place
before the HTS campaign but there was a low to medium throughput kinase activity assay
(a nonhomogeneous scintillation assay usingATP--
33
P) available that was not compatible
with HTS. Using this assay, it was shown that the nonselective protein kinase inhibitor
staurosporine is a highly potent inhibitor of this Ser/Thr kinase and could therefore be used
for competition experiments. The NMR screen using STD identified 32 binding fragments
which were tested in the kinase activity assay at two concentrations, 250 and 500 M. Six
