Skip to main content

Novel tools for crystallisation of macromolecules

Final Report Summary - TOPCRYST (Novel tools for crystallisation of macromolecules)

Genes in our deoxyribonucleic acid (DNA) code for the production of polypeptides are folded to make proteins, which are the building blocks of life. Protein function is crucial for many diseases and so the ability to understand and then inhibit or enhance their action is essential for rational drug design and medical research in general. It has therefore become a priority to determine the three-dimensional structures and hence understand the function of these molecules in atomic detail.

The prime method for determining such structures is X-ray crystallography. However, this technique requires that the protein molecules in solution be converted into an ordered crystal. The 'recipe' for doing that varies for each protein. Originally, the general consensus was that by subjecting each protein to numerous different chemical conditions, it would be possible to find ones suitable for crystallising them. Unfortunately, a bottleneck was reached, with only around 20 % of the proteins tried being successfully crystallised. Most of the proteins we were hoping to crystallise were thus stuck in a cul-de-sac.

The TOPCRYST project and its objectives

TOPCRYST, an industry-academia pathways and partnerships European Commission (EC) project coordinated by Dr Emmanuel Saridakis in collaboration with Dr Irene Mavridis at the National Centre for Scientific Research 'Demokritos' in Athens, aimed to develop a method for helping to tackle the problems of discovering and assessing crystallisation conditions.

The project was carried out in partnership with Dr Attia Boudjemline, Marcus Swann and Geoff Platt of Farfield Group Ltd., a British small and medium-sized enterprise (SME) specialising in the design and manufacturing of optics-based scientific instrumentation, and with Professor Naomi Chayen, Head of the Protein Crystallisation Group at the Biomolecular Medicine Section, Faculty of Medicine, Imperial College London.

Farfield has pioneered the use of a method known as dual polarisation interferometry (DPI) for a wide range of applications. DPI is a surface analytical technique used for the characterisation of structure and interactions of molecular layers on an optical waveguide surface for a wide range of applications, including protein-protein interactions and conformational changes. It is based on a waveguide structure which supports light confined within a core. Information is obtained by studying the change in the propagation velocity of light travelling through the waveguide structure which manifests itself as a change in the position of interference fringes, i.e. a phase shift.

This instrument was now shown to be effective in detecting macromolecular crystal nucleation at its very start. When crystals begin to form, light loss from the top of the waveguide increases dramatically, rapidly leading to the loss of the interference fringes. The main objective of this project was therefore to explore the possible use of DPI technology as a new tool for crystallisation diagnostics, via detection of crystallisation events at the earliest possible stage.

Main innovations, results and conclusions

Preliminary observations of crystallisation by DPI had been undertaken in a batch mode, which involved mixing the protein and precipitating agents before proceeding to the measurements. This had a number of disadvantages, rendering the method impractical. We modified the apparatus for implementing DPI in a dialysis setup, by splitting the measurement chamber in the horizontal direction via interposition of a dialysis membrane. This allowed the scanning of a wide range of crystallisation conditions at once, along a continuous or step-wise gradient of concentration or pH. Further modifications led to minimisation of the protein volume loaded in the chamber to less than 10 µL, which is crucial for the usually scarce target proteins.

The new method was first applied to lysozyme, catalase, and thaumatin, under crystallisation and non-crystallisation conditions. These are model proteins, for which crystallisation conditions are already known and which are often used to develop, test and validate new crystallisation methods.

By comparing the phase shifts and, more importantly, contrast loss of the interference fringes at conditions leading to:

(i) crystallisation,
(ii) clear solutions and
(iii) amorphous precipitation,

we found an unequivocal, 'signature' signal pattern that only appeared upon crystallisation. The method also allowed us to unexpectedly find an unusual crystallisation condition for the thaumatin.

Rat dynamin and xylanase are two important target proteins for which finding crystallisation conditions had proved to be problematic. Rat dynamin did not yield a DPI indication of crystallisability. The xylanase displayed the characteristic crystallisation 'signature' when subjected to conditions that had given crystals in unconventional setups but not with more standard techniques. These conditions would therefore normally have been missed. This result was again an indication of the effectiveness of DPI.

In conclusion, real-time DPI monitoring of trials can indicate whether a particular precipitant or precipitant / buffer combination is worth pursuing and optimising for crystal production. DPI was thus shown to be effective as a guide to selecting conditions that will lead to the crystallisation of macromolecules, drastically minimising the number of preliminary initial conditions to be searched.

Societal impact

The results of the project have obvious application in the field of drug development and healthcare.

Another important element of the project was knowledge transfer on state-of-the-art crystallisation methodology and on DPI between the academic and industrial partners and to the wider structural biology community. This was achieved through the secondments and a very well-attended workshop with distinguished speakers in the field that took place in Athens towards the end of the project.

Project website: