- 1:10 PM

Microfluidic Platforms for (Membrane) Protein Crystallization

Paul J. A. Kenis1, Sarah L. Perry1, Griffin W. Roberts1, Sameer Talreja1, Joshua D. Tice1, Charles F. Zukoski1, and Robert B. Gennis2. (1) Chemical & Biomolecular Engineering, University of Illinois at Urbana-Champaign, 600 South Mathews Avenue, Urbana, IL 61801, (2) Biochemistry, University of Illinois at Urbana-Champaign, 600 South Mathews Avenue, Urbana, IL 61820

Despite their critical role in many biological processes, the precise structure of a disproportionate number of membrane proteins is still unknown. Their limited availability and amphiphilic nature have seriously hampered both the identification of suitable crystallization conditions as well as the subsequent growth of X-ray quality crystals. We exploit microfluidics to facilitate the identification of suitable crystallization conditions on-chip using minute quantities of proteins on both the in-surfo and in-meso approach.

In Surfo: Determining Phase Diagrams and Improving Crystal Quality.

We have shown that kinetic control of crystallization by gradual, direct evaporation of the solvent is an effective method to identify and optimize crystal-producing conditions for lysozyme, thaumatin, and RNAse. We created a microfluidic platform comprised of 144 wells for ~20 nL-sized droplets, capable of screening 48 precipitants at 3 different protein/precipitant ratios. We have also added a reverse vapor diffusion component to the protocol to decouple crystal nucleation and growth, thereby reducing the number of nuclei and increasing crystal quality. Using a similar protocol, we can determine the solubility diagram of a protein using less than 20 ÁL of protein solution. We have done this for example for the membrane protein bacteriorhodopsin.

In Meso: Membrane Protein Crystallization from Lipidic Phases.

Landau & Rosenbush, and later Caffrey and coworkers have developed the crystallization of membrane proteins from a lipidic mesophase. An aqueous protein solution and a viscous lipid are mixed to form a lipidic mesophase with the protein embedded in the bilayers that form a bicontinuous cubic structure. Upon the addition of salt, a lamellar phase forms in which the membrane proteins align into stacks of 2D sheets, thus facilitating crystallization. We have created a microfluidic chip capable of mixing viscous lipids (e.g. monoolein) with a 30x less viscous, protein containing aqueous phase. The required cubic mesophase is formed by pushing the solutions back and forth between a central lipid containing compartment and two adjacent protein solution containing compartments. Upon the addition of salt the solution transitions to a lamellar phase. We have successfully grown crystals of bacteriorhodopsin on chip. Presently we use IR microscopy to confirm that the crystals are indeed protein and not salt. The on-chip, in meso crystallization of several other membrane proteins is in progress as well as the development of chips that allow for on-chip X-ray evaluation of crystal quality.