Feedback-System-Control Integrated Microfluidic System for Fast Screening of Protein Crystallization Conditions

Proteins are the essential components of all living creatures that link genes and biofunctionalities, and their crystallizations have attracted much interest; the study of the 3D structures through X-ray crystallography would provide critical information for novel drug design and disease treatment. To reduce searching efforts, to minimize the amount of protein solutions, and to increase the final accuracy, here we demonstrate protein crystallization optimization using the in-house fabricated hydrogel-digitized microfluidic system (HD-MFS). This technique is incorporated with a two-stage feedback-system-control (2s-FSC) algorithm for the fast and easy screening of chemical conditions required for obtaining high quality and large protein crystals. We have demonstrated the efficacy of our technique by crystallizing three different types of proteins, including lysozyme, proteinase-K, and KDM-4A. Large and high quality protein crystals could be obtained by conducting many fewer crystallization trials (tens experiments) in parallel from a pool of thousands of possible trials; thus, our methodology allows for thorough screening of crystallization conditions without the need for elaborate and expensive robotic solutions. In addition, the expanded two-stage approach for the feedback-system-control algorithm (FSC) can not only rapidly identify the possible locations in a vast parameter space but also refine the final searching results in a more confined and accurate region, where the desired chemical conditions can be identified quickly and accurately to obtain large and top quality protein crystals. The performance comparison with other published literature has been carried out, and the results indicate that the current design can perform protein crystallization more efficiently by the integration of the HD-MFS and FSC algorithm. We strongly believe that the hydrogel-digitized microfluidic system with the 2s-FSC algorithm could be one of the easiest ways to obtain protein crystals of unsolved novel protein macromolecules.