Biomedical Engineering Reference
In-Depth Information
products available on the market are “unique drugs” (of which about 90% are small
organic molecules and 10% are biopharmaceutical, i.e., proteins and nucleic acid
derivatives) [7]. Barely 324 different drug-target protein pathways (of which 266
identified in the human genome and 58 in the pathogen genome) and 130 privileged
“druggable” protein domains (of the 16,000 protein families and 10,000 different
folds) account for all the drug targets used in pharmaceutical developments so far [7].
Paradoxically, the gap between the number of drug molecules and drug targets
available is
not
due to a lack of chemical compounds; rather it reflects a lack of ade-
quate drug discovery tools for the efficient identification of optimal binding molecules
among a vast, fast-growing chemical space [10].
In the era of molecular medicine, there is an inevitable need to develop more
efficient and versatile drug discovery techniques, capable of rapidly generating and
interrogating massive collection of chemical compounds in order to improve drasti-
cally the ability to tackle new target classes, difficult to treat with conventional drug
discovery approaches.
11.1.2 Selecting Chemicals
Every pharmaceutical agent is, to begin with, a molecule that binds selectively to
one or more protein targets. Thus, the identification of molecules capable of specific
protein recognition is at the heart of modern drug discovery technologies.
Low-molecular-weight compounds (small organic molecules) are extremely pow-
erful tools; such compounds are generally able to quickly penetrate cell membranes
and initiate biomolecular modifications, eventually leading to the desired therapeutic
effects. However, the identification of small-molecular probes capable of selectively
binding to any target protein of choice is frequently a massively cumbersome task
[11,12].
Conventional approaches for the identification of small molecular biological
agents often require the discrete screening [high-throughput screening (HTS)] of
very large chemical collections (compound libraries), comprising up to a few million
compounds. Frequently, to expand the chemical space and explore underrepresented
classes of chemical structures, rational design approaches [13,14], combinatorial
chemistry [15,16], a fragment-based approach [11,17,18], and diverse-oriented syn-
thesis [19-21] are employed.
Virtual screening [22,23] and structure-activity relationships (SARs) [24] may
sensibly reduce the number of compounds that need to be evaluated. Nonetheless,
high-throughput-screening campaigns routinely require the repetitive screening of
hundreds of thousands to millions of chemical compounds against a single drug
target in order to identify a handful of candidates with adequate biological properties
(Figure 11.1a) [12].
Despite the practical burdens (tremendously demanding in terms of library syn-
thesis, compound management, robotics, and logistics) and the intrinsic inefficiency,
high-throughput screening is still the drug discovery technology of choice for the
identification of small-molecule hits against a given biological target [25].
