Chemistry Reference
In-Depth Information
The synthetic advantages and target-informed product formation with
in situ
DCC and
Click chemistry are of little consequence unless we have the ability to identify and charac-
terize readily the linked fragments of interest. Accomplishing the rapid detection of these
enriched product(s) is clearly a make-or-break factor to
in situ
DCC and Click chemistry
taking a prominent place in fragment-based optimization for drug discovery.
[
45
]
The stand-
ard DCL screening approach is indirect and must carry out the screening of identical DCLs
twice, with and without the target. The equilibrium concentration profiles of all library
ligands are then compared, typically following disruption of the ligand-target noncova-
lent complexes of the target protein. The detection of ligand enrichment in the targeted
DCL is the basis of identifying the 'best binders'. The complexity of screening with this
indirect method increases with the size of the DCL owing to chromatographic (e.g. HPLC)
and/or spectral (e.g. NMR) overlap so that either the need for the synthesis of individual
library components to validate assignments and/or the preparation and screening of DCL
sublibraries becomes necessary for deconvolution of larger DCLs; this is both labour and
time intensive and has the effect of undermining the promoted advantages of DCC. For
DCC to have a significant impact on drug discovery necessitates the development of a
direct screening protocol(s) for rapid identification and characterization of those ligands
with affinity for the target protein, against the background of inactive DCL constituents. It
would be a tremendous advantage to drug discovery applications of DCC if DCL screening
methodologies could proceed without the need for chromatography, conversion to a static
library, preparation of sublibraries, prior disruption of the protein-ligand complexes, syn-
thesis of individual library ligands to validate spectral assignments or for the preparation
and duplicate screening of identical DCLs (with and without the target). In pursuing this,
we must relinquish our 'need to know everything' philosophy of medicinal chemistry and
instead settle for what we 'really need to know', that being the molecular structure of the
linked fragments that are selected and amplified in the system under investigation.
The strong amplification of selected DCL constituents has been observed in many
examples, particularly from smaller DCLs. For drug discovery we may sometimes also
want to work with larger libraries, so what can we expect of amplification levels when we
have an increased number of initial library fragments? Fortunately, the question of what the
limit to the size of effective DCLs is has already received serious consideration, with vari-
ous groups having modelled DCL product distributions under a variety of library scenarios.
This material has been reviewed recently and a complete discussion is beyond the scope
of this chapter.
[
46, 47
]
The results from one of the most relevant models to drug discovery
applications are summarized in Figure 7.4. This model confirms the intuitive reduction in
the observed concentration of the best binders that occurs with increasing number of DCL
fragments (and possible products) that is a consequence of spreading the fragments out
over an increasing number of possible library constituents.
[
46
]
However, this decrease in
concentration is not as sharp as expected; for example, in a 1000 compound library the
best binder will become 13 % (on average) of the library composition and in a 10 000
compound library the level is 8 % (on average) (Figure 7.5). Several good binders are a
probable outcome from
in situ
DCC; with this scenario, amplification levels for the good
binders are likely to remain within the detection capability of modern analytical equipment.
Like DCC, applying
in situ
Click chemistry against drug discovery targets necessitates
the identification of the best binding ligands. The experimental reaction mixture for
in
situ
Click chemistry shares many component attributes in common with DCC, namely
