Chemistry Reference
In-Depth Information
2.2 Phase I Activities
2.2.1 Target Assessment
It has been estimated that
10% of the entire human genome is involved in disease onset
or progression,
[
13
]
resulting in several thousand potential targets suitable for therape
u
tic
intervention. Most drug discovery targets are proteins, but this is not always the case.
[
14 17
]
We will focus on protein targets only for the rest of this chapter, as the concepts for nonpro-
tein targets are the same. The first stage in the drug discovery process is target identification
and validation. If a target is not validated with the disease, resources can be wasted in a
fruitless search for a drug. Sometimes the result of FBDD (or LLDD) is de-validation of
a target, which can be just as important as finding a lead compound against a target, just
not nearly as glorified. Some people would argue for 'druggability' as a relevant part of
target assessment but, as discussed below, we do not. An estimated 60% of small molecule
drug discovery projects fail because the biological target is found to be not 'druggable'.
[
18
]
The easy 'druggable' targets have already been the focus of intense drug discovery efforts;
the future of drug discovery lies in drugging 'undruggable' or novel targets. The current
lead-like drug disco
v
ery paradigm consists of the creation of libraries around previous
work for a target;
[
1, 19 21
]
therefore, 'undruggable' and novel targets will be a difficult task.
This also creates inherent issues with intellectual property; issues that are not inherent
in fragments.
Once a validated target has been chosen, it is important to develop a detailed dossier on
the protein. This information will affect both assay assessment (Figure 2.1) and compound
assessment (Figure 2.1).
[
22
]
This obviously starts with the classification of the target into a
given class: nuclear hormone receptor, membrane protein, kinase, protease, carrier protein,
chaperone, metalloenzyme, etc. Some target classes are easier to find hits against than
others, most notably enzyme targets.
[
23, 24
]
Soluble, single-domain proteins (or those that
have isolatable enzymatic domains) are much easier to work with, in both fragment-based
and lead-like drug discovery, e.g. kinases and proteases. Membrane proteins, which are the
most difficult to work with, especially from a biophysical standpoint, make up 50% of the
pharmaceutically relevant targets.
[
25
]
It is at this point that the target validation status of the target is determined. It is beyond
the scope of this chapter to discuss the criteria used to determine this, but there are many
excellent papers that describe some relevant aspects.
[
26, 27
]
One proposed method is to
prioritize targets based upon their druggability. Figure 2.2 shows the calculated druggability
of targets of high pharmaceutical interest.
[
28
]
Of particular note is the very high calculated
druggability (10 M) of the 'undruggable' target HIV integrase (a typical 'undruggable'
case scenario). In 2007, Merck launched Isentress, which inhibits this target, showing that
'undruggable' in this context does not mean what one would think it means. We would
argue that target druggability is irrelevant as long as the target is suitably well validated.
Where there is a validated target and a will, there is a (drug discovery) way.
We feel that target validation is the key comp
o
nent of the target assessment step and
many interesting approaches have been detailed.
[
29 31
]
An interesting approach has recently
emerged for target assessment involving chemical genomics. Both forward and reverse
chemical genomics can play a role in target validation.
[
32
]
These paradigms are depicted in
Figure 2.3. Forward chemical genomics explores phenotypes by screening libraries to obtain
∼
