View All Posters

P13030
Rapid Improvement of Protein Stability by Thermal Profiling
Presenter William Chiou, Abbott Laboratories, USA
Additional Authors: Kenneth M. Comess, Linda Traphagen, Usha Warrior, Scott C. Galasinski, David J. Burns

Buffer conditions are critical to protein folding, stability, and the ability to concentrate and crystallize. True buffer optimization involves the testing of a variety of conditions involving pH, buffer type, ionic strength, and stabilizing additives.  This can be a very tedious process, and often consumes large amounts of protein. We describe here a two-step strategy that can be used to screen buffers and additives in a high-throughput fashion using 384 well plates. Protein stability is assessed by the midpoint of thermal denaturation (Tm), and stabilizing or destabilizing conditions can be detected by deviations that increase or decrease the Tm respectively. The Tm is determined by gradually heating the protein, and as the denatured protein unfolds, it exposes hydrophobic regions that can be bound by an environmentally sensitive dye.  Dye binding results in enhanced fluorescence, which can be detected in a fluorescence plate reader at EX/EM=465nm/590nm. In the first step, conditions consisting of a buffer salt (10-50 mM, pH of 5.5-8.5), a salt (2-98 mM), an additive, and in some cases, a detergent were assembled to form 384 buffer conditions using incomplete factorial mixing algorithm In the second step, 96 additives commonly used for protein crystallization were screened against the four best buffer conditions identified in the first step. Two E. Coli-expressed human Aurora-B constructs (6His-(Thr)-[Aur1 (1-344)] and [Aur1(h) (44-344) C170R] + Incenp) were used to test this screening strategy. The results demonstrate that the thermal stability for Aurora-B (1-344) was improved by 11 oC when compared to a commonly used buffer found in the literature for kinases. The Tm for Aurora-B (44-344) was also significantly improved by 18 oC after the two-step screen. The pH of the buffer correlated as the dominant factor determining stability.  The cofactor screen revealed expected stabilizing interactions, e.g., ATP, as well as stabilization by a number of multivalent metal ions.  The inclusion of an additive contributed to an additional 10 degrees of thermal stability.   The protein consumption in this study was 10 ul at 2.5 uM for each buffer condition.  In conclusion, we report here a two-step screening strategy that could rapidly and significantly improve the protein stability, consuming less than 1 mg of protein.