The Drug development process - Pharmaceutical Biotechnology: Concepts and Aapplications - page 67

in practice the fi eld focuses upon protein structure. The basic approach to structural genomics
entails the cloning and recombinant expression of cellular proteins, followed by their purifi cation
and three-dimensional structural analysis. High-resolution determination of a protein's structure
is amongst the most challenging of molecular investigations. By the year 2000, protein structure
databanks housed in the region of 12 000 entries. However, such databanks are highly redundant,
often containing multiple entries describing variants of the same molecule. For example, in excess
of 50 different structures of 'insulin' have been deposited (e.g. both native and mutated/engineered
forms from various species, as well as insulins in various polymeric forms and in the presence of
various stabilizers and other chemicals). In reality, by the year 2000, the three-dimensional struc-
ture of approximately 2000 truly different proteins had been resolved.
Until quite recently, X-ray crystallography was the technique used almost exclusively to resolve
the three-dimensional structure of proteins. As well as itself being technically challenging, a ma-
jor limitation of X-ray crystallography is the requirement for the target protein to be in crystalline
form. It has thus far proven diffi cult/impossible to induce the majority of proteins to crystallize.
NMR is an analytical technique that can also be used to determine the three-dimensional struc-
ture of a molecule, and without the necessity for crystallization. For many years, even the most
powerful NMR machines could resolve the three-dimensional structure of only relatively small
proteins (less than 20-25 kDa). However, recent analytical advances now render it possible to
analyse much larger proteins by this technique successfully.
The ultimate goal of structural genomics is to provide a complete three-dimensional descrip-
tion of any gene product. Also, as the structures of more and more proteins of known function are
elucidated, it should become increasingly possible to link specifi c functional attributes to specifi c
structural attributes. As such, it may prove ultimately feasible to predict protein function if its
structure is known, and vice versa.
4.7 Pharmacogenetics
Pharmacogenetics relates to the emerging discipline of correlating specifi c gene DNA sequence
information (specifi cally sequence variations) to drug response. As such, the pursuit will ulti-
mately impinge directly upon the drug development process and should allow doctors to make
better-informed decisions regarding what exact drug to prescribe to individual patients.
Different people respond differently to any given drug, even if they present with essentially
identical disease symptoms. Optimum dose requirements, for example, can vary signifi cantly.
Furthermore, not all patients respond positively to a specifi c drug (e.g. IFN-β is of clinical benefi t
to only one in three multiple sclerosis patients; see Chapter 8). The range and severity of adverse
effects induced by a drug can also vary signifi cantly within a patient population base.
While the basis of such differential responses can sometimes be non-genetic (e.g. general
state of health, etc.), genetic variation amongst individuals remains the predominant factor.
Although all humans display almost identical genome sequences, some differences are evi-
dent. The most prominent widespread-type variations amongst individuals are known as single
nucleotide polymorphisms (SNPs, sometimes pronounced 'snips'). SNPs occur in the general
population at an average incidence of 1 in every 1000 nucleotide bases; hence, the entire human
genome harbours 3 million or so. SNPs are not mutations; the latter arise more infrequently, are
more diverse and are generally caused by spontaneous/mutagen-induced mistakes in DNA re-
pair/replication. SNPs occurring in structural genes/gene regulatory sequences can alter amino