Coupling Crystallisation With X-ray Data Collection

Protein crystallisation is still, after the expression of soluble proteins, the major bottleneck in protein structure determination. Considerable efforts are underway worldwide to enhance the basic understanding of protein crystallisation and to target it towards growing crystalline samples that are sufficiently good to produce diffraction patterns suitable for subsequent X-ray structure determination. One approach is to set up crystallisation experiments in a massively parallel way automatically by robotics, while at the same time minimising protein consumption and enhancing reproducibility. The sheer number of experiments requires that the results and their time dependence are collected automatically and stored in a database. With this enormous amount of data at hand, the outcome of the crystallisation experiments can be linked to the physicochemical properties of the protein and more can be learned about why and when and under what conditions it crystallises. The role of the BIOXHIT project in this was to link the existing and future facilities together and to coordinate them with the needs at the synchrotron beamlines. This should provide the possibility of crystallisation at the synchrotron site, thus allowing direct feedback from the diffraction experiment to the crystallisation. This should also allow crystal improvement studies to be carried out at the synchrotron site in challenging projects, where it has proved difficult to obtain good quality crystals of sufficient size for testing in the home laboratory.

As part of the development of xtalPIMS, Figure 3.1 a, a certain configuration of computing resources to allow access to images has been suggested. This entails one computer to serve images, another to serve the database and a third computer to serve the web site. For smaller labs, it is likely that the database and the imager service functionality will be co-located on the same machine, although this is likely to be dependent on other factors, e.g. the use of PIMS within the lab. Progress has also been made on the optimisation of queries to the PIMS database for accessing image and crystallisation-related data.

Figure 3.1 a. The Protein Information management System

Figure 3.1 b. Crystals grown in a microfluidics

Figure 3.1 c. Microfluidics crystalliser

Partner 1A (EMBL-HH) has developed and successfully tested a system for the setup of free-interface diffusion experiments. In order to investigate the possibilities for commercialisation, Partner 1A has initiated contacts with relevant SMEs. In the system, crystals can be grown from volumes of less than 5 nL in a multitude of reaction chambers, Figure 3.1 b. The system allows the controlled and automatic addition of sample and the controlled delivery of both sample and crystallisation solution on a CD-like crystallisation chip, Figure 3.1 c. The system was successful in 4 test cases with proteins. In contrast to the commercial system from Fluidigm (Topaz chip system), the BIOXHIT system works without the need of pumps or pressure driven valves. The production costs of the chips are also expected to be a fraction of the typical costs of a Topaz chip. The ability to investigate the intrinsic diffraction quality of crystals in-situ was also demonstrated successfully.