Chemistry Reference
In-Depth Information
extended periods are not especially amenable to many biophysical screens. If the target
was the focus of previous efforts that utilized primarily biochemical screens and resulted
in no hits suitable for lead optimization, then the next iteration on this target may warrant a
biophysical screen, similar to reverse chemical genomics. If there is structural data of the
target or suitable surrogate, then any of the multitude of structure-based drug discovery
(SBDD) or
in silico
design paradigms may be utilized.
[
46, 47
]
The breadth of these is far
beyond th
e
scope of this chapter and this topic has recently been the focus of many recent
reviews.
[
48 50
]
The primary choice in this assessment is whether to utilize a biochemical or biophysical
screen as the primary filter. Although it seems like an either/or choice, this is a false dicho-
tomy. The most successful fragment screens obtain orthogonal data, i.e. both biochemical
and biophysical data in parallel or in quick succession.With orthogonal data, the probability
of false positives (or negatives) is reduced. Most commonly biochemical and biophysical
data are obtained. However, all the different biochemical and biophysical screens can be
considered orthogonal. We would recommend that if two biophysical methods are to be
used at least one should be a direct method (discussed below).As is noted many times in this
topic, rapid iterations among the various data sources are the key to a successful process.
Biochemical versus biophysical screens.
As shown in Figure 2.1, the first three steps of
FBDD are interdependent: the choice made in one assessment impacts the choices that
are/can be made in the others. For example, a fluorescent biochemical screen requires
compounds which do not quench the assay or give false positives. However, fluorescence
quenching compounds can be easily run in a biophysical screen. On the other hand, a
fluorescence biochemical screening looking for changes in fluorescence polarization aniso-
tropy
[
51, 52
]
requires the exact opposite set of characteristics in a molecule. A biophysical
screen using mass spectrometric (MS) detection requires moderately soluble compounds
which will ionize,
[
53
]
whereas an NMR-based scr
e
en requires compounds with at least
one-nonexchangeable proton and high solubility.
[
54 57
]
Most of the time the choice of bio-
chemical versus biophysical screen is made out of comfort and available expertise. Cost
is, of course, a concern at this point as many biophysical screens require equipment that
have expensive upfront costs. However, it would be unusual to consider a screen unless the
equipment was already available or a suitable collaboration partner could be found. The
nature of FBDD requires such close commitment of primary and secondary screens. For
FBDD these terms are really inadequate, but for now will have to suffice. The assumption
that new biochemical assays need to be developed for fragments is generally unwarran-
ted.. Fragments are typically screened at higher concentrations than lead-like compounds
(
≥
1 mM versus 10-25 M) and therefore the highest DMSO concentration (fragments are
typically solubilized at
100 mM in DMSO) in the biochemical screen is 1%. As long
as the assay's DMSO tolerance is known, it can be used for FBDD. A brief summary of
each of the main types of screening approaches used in FBDD is given in Table 2.1 and
discussed in Section 2.3.1.
≥
2.2.3 Compound Assessment
Compound assessment and library design are part and parcel. However, one trap that can
be tempting, but in the long run inefficient, is the belief that a universal fragment library
