ETH Zurich, Switzerland
Title: Microrobotics enables non-contact, fully automated protein crystal harvesting
David Sargent obtained his PhD in biophysics from the University of Western Ontario, Canada, followed by postdoctoral studies at the ETH Zurich and the University of Sydney (Australia). He has had extensive experience in macromolecular crystallography at the ETH Zurich, and recently has also been associated with the Multiscale Robotics Laboratory (ETH Zurich) of Bradley J. Nelson. David is one of the founders of MagnebotiX, a spinoff of the ETH, which provides tools for magnetic propulsion and guidance at the microscopic scale. The work reported here uses this technology to streamline and accelerate the process of macromolecular crystal structure determination.
Statement of the Problem: Most aspects of macromolecular structure determination, from synthesis and purification of materials, through crystallization, data collection and model building, are highly automated, but the recognition, harvesting and cryocooling of crystals remains a predominantly manual task. Several concepts, including in situ crystallography, are being developed to overcome these difficulties, but frequently impose other restrictions, such as on data collection strategies. We are developing hardware and software to support crystal harvesting using standard crystallization procedures, thus avoiding such limitations. Methodology & Theoretical Orientation: We use a magnetically driven, mobile, rolling microrobot, the “RodBot”, to locally move the liquid surrounding a crystal, and the crystal then passively follows the flow. Crystal position is monitored using low level uv-light. Transport is controlled using flexible algorithms that allow for error-recovery following stochastic disturbances. Findings: We demonstrate the effectiveness of the technique using crystals of different geometries and densities in a variety of buffers and cryoprotectants. Even at this developmental stage average harvesting time is reduced compared to manual operations. Conclusion & Significance: This non-destructive, non-contact method allows crystals to be extracted reliably from the growth droplet in a completely automated process. Harvesting can take place remotely in climate-controlled chambers, ensuring optimal conditions throughout the process with respect to temperature, humidity and composition of the environment. Damage to valuable crystals due to operator jitter or fatigue is eliminated. Incorporation into existing robotics setups for sample handling will also allow increased reproducibility of flash-cooling. Fully automated structure determination pipelines using well-established techniques are now possible and will yield improved data quality at reduced cost.