High-throughput protein crystallography and drug discovery

Abstract

Single crystal X-ray diffraction is the technique of choice for studying the interactions of small organic molecules with proteins by determining their three-dimensional structures; however the requirement for highly purified protein and lack of process automation have traditionally limited its use in this field. Despite these shortcomings, the use of crystal structures of therapeutically relevant drug targets in pharmaceutical research has increased significantly over the last decade. The application of structure-based drug design has resulted in several marketed drugs and is now an established discipline in most pharmaceutical companies. Furthermore, the recently published full genome sequences of Homo sapiens and a number of micro-organisms have provided a plethora of new potential drug targets that could be utilised in structure-based drug design programs. In order to take maximum advantage of this explosion of information, techniques have been developed to automate and speed up the various procedures required to obtain protein crystals of suitable quality, to collect and process the raw X-ray diffraction data into usable structural information, and to use three-dimensional protein structure as a basis for drug discovery and lead optimisation. This tutorial review covers the various technologies involved in the process pipeline for high-throughput protein crystallography as it is currently being applied to drug discovery. It is aimed at synthetic and computational chemists, as well as structural biologists, in both academia and industry, who are interested in structure-based drug design.

Soft matter perspective on protein crystal assembly

“…Although some of these processes can be characterized at a relatively coarse scale, more often than not a microscopic understanding of the structure and dynamics of the involved biomolecules is also essential [1]. Proteins, for instance, interact with each other and their environment through fine-tuned structural features so as to perform their biological functions [2][3][4][5][6]; reciprocally, protein malfunction is often due to structural defects, and may result in diseases [7]. Proteins also represent some of the most sophisticated nano-machines known, having been shaped by natural selection.…”

Ligand identification using electron-density map correlations

“…The identification of the ligand in these cases can be an important step in characterizing the macromolecule, as it may give clues as to the natural function of the macromolecule. A related situation occurs in increasingly many drug-discovery and ligand-discovery projects in which a mixture of ligands is included in crystallization or after crystallization, a structure is determined and the identify of the bound ligand is determined from the density (see, for example, Tickle et al, 2004).…”

Automated Protein–Ligand Crystallography for Structure‐Based Drug Design

“…Moreover, there have been significant advances in methodologies and automation for protein production and crystallisation. [2] As a result, obtaining crystals is no longer necessarily a bottleneck for the structure determination of protein-ligand complexes. Also data collection has often been a relatively slow process but significant advances have been made.…”

“…Also data collection has often been a relatively slow process but significant advances have been made. [2] For example, robotic systems that are able to store and automatically manipulate protein crystals [3,4] have removed the need for manual intervention during X-ray data collection.…”