Chemistry Reference
In-Depth Information
of highly potent inhibitors. By tracing back the end compounds through to constituent frag-
ments and analysing the change in physicochemical properties with potency, he formulated
a measure of the likelihood of developing a drug-like molecule from a particular hit frag-
ment based on the size and potency of that molecule. Essentially, potency was observed
to increase proportionately with mass along an ideal optimisation path, suggesting that
ligand efficiency, discussed further below, should be used in the process of both selecting
the most desirable hit fragments and also evaluating the effectiveness of each modific-
ation in the development phase. Hadjuk's analysis looks at requirements for individual
fragment hits; Makara
[
27
]
investigated another aspect of fragment screening, library size,
in particular in relation to sampling available fragment space. Key conclusions are that a
relatively small number of fragments yield hits against many targets with libraries of 10
3
compounds providing more than sufficient chemical matter to follow up. Diversity, in this
instance across the reagents used to construct the library, is vital in formulating a fragment
screening deck.
What, then, is a fragment? As for drug-like classification, there is no single, unifying
definition that can categorically distinguish between fragments and non-fragments, but
there are some general rules of thumb:
•
Fragments are smaller than drug-like molecules, whether that is defined by MW or the
number of heavy atoms.
•
Fragments, in addition to being smaller, are less complex than drug-like molecules.
•
Fragments, due to the manner in which they are used, should be highly soluble.
•
Fragments should be devoid of undesirable chemical functionality while at the same time
facilitating rapid development to more potent compounds.
1.3 Why Use Fragments?
There are a multitude of reasons for the current popularity of using fragments in drug dis-
covery stemming from both the conceptual advantages of using fragments and the observed
failures of alternative methodologies.As stated above, the concept of using fragments arose
in the early 1980s when it was theorized that the total affinity of a molecule could be taken
as a function of the affinity of constituent components. However, it was only in the mid-
1990s, with technological advances in the ability to detect weakly binding fragments, that
theory was put into practice with what is often cited as the first demonstration of fragment-
based drug discovery.
[
28
]
In the meantime, drug discovery became a game of numbers
as it was perceived that quantity could compensate for a lack of understanding with the
development of HTS and combinatorial chemistry. HTS has had successes and is still the
predominant means of hit identification, but it was quickly acknowledged that screening
large compound libraries hasmany shortcomings and if anything led to higher attrition rates.
Subsequently, screening libraries were overhauled based on drug/lead-like characteristics,
high-content screening was introduced and ultimately companies looked to alternative
means of initiating discovery programmes including ever more sophisticated fragment
based approaches.
One of the postulates of fragment-based drug design is that considerable gains in potency
can be achieved through linking together fragments binding in different regions of the
