27.2 Protein–Protein Interactions

367

Table 27.2 Stages of drug discovery and development

Stage

Desired outcome

Technologies involved

1. Target selection

A gene

(Functional) genomics;

genotyping

2. Protein expression

A three-dimensional protein

structure

Protein chemistry

3. Screening

A drug which binds

Binding studies

4. ADME

A usable drug

Interaction studies

5. Trials

An efficacious drug

Clinical trials

27.1 Routes to Discovery

Once a target has been identified, the next stage is to find a small molecule that binds

to it. It is assumed that the structure is known. ² If not, advanced computational tools

should be able to predict the structure corresponding to a gene sequence. ³

The first approach to finding a small molecule that can bind to the relevant part of

the protein target is based on the venerable “lock and key” concept. Virtual reality

(VR) has enormously enhanced the efficiency of molecule finding: a researcher can

play around with candidate molecules and the target to get a feel for what might

bind. There is of course an enormous and continually growing corpus of knowledge

about what molecules bind to which motifs to guide the researcher.

The primitive “lock and key” concept is unlikely to be successful as a general

strategy, however, because it neglects the exceedingly important phenomenon of

induced fit, which describes how the shape of an enzyme changes when it binds to

its substrate. This enormously increases the dimensionality of the parameter space

and typically the problem then far exceeds the ability of a human being to optimize a

drug–protein interaction by visually playing around, even with the assistance of VR.

Artificial intelligence is one way of automatically steering design in the presence of

drug-induced protein conformational changes. ⁴

27.2 Protein–Protein Interactions

Proteins in cells do not exist in isolation. They bind to other proteins to form multi-

protein structures that inter alia are the elements of pathways that control functions

2 in 2023 there are about 180,000 structures in the Protein Data Bank (see https://www.rcsb.org/

stats/growth/growth-protein), but only about 5% of these are of human proteins.

3 Fernández (2020).

4 Fernández (2021).