Ng, Jia Tsing, Dekker, Carien, Reardon, Paul and von Delft, Frank (2016) Lessons from 10 Years of Crystallization Experiments at the SGC. Acta Crystallographica Section D, 72 (2). pp. 224-235. ISSN 2059-7983

Abstract

Abstract Though protein crystallization is generally considered more art than science and remains significantly trial-and-error, large-scale datasets hold the promise of general learnings. Here we present observations from retrospective analyses of the strategies actively deployed for the extensive crystallization experiments at the Oxford site of the Structural Genomics Consortium (SGC) during its first 10 years of operation. We report on the importance in the redundancy of crystallizing conditions, specifically using several mixing ratio of protein sample and precipitant as well as different incubation temperatures. From a comparison of the four most-widely used sparse-matrix (“coarse”) screens that yielded crystals eventually resulting in deposited structures, we find no meaningful difference in performance. We therefore conclude that statements regarding differential or superior behaviours of any sparse-matrix screen will always be meaningless without extensive cross-testing. Our analyses do not disparage the redundancy in crystallization conditions, but de-emphasize on the importance of the contents of sparse-matrix screen. Specifically, where ~200µl of concentrated protein sample is available, it is effective and meaningful to set up four sparse-matrix screens at two incubation temperatures, with each condition in three 150nl droplets at varying protein to precipitant mixing ratios of 2:1, 1:1 and 1:2. In contrast, where protein sample is limiting, the number of conditions should be prioritised ahead of redundancy, whether in mixing ratio or temperature, in order to maximise likelihood of identifying hits. Finally, concerning the frequently asked logistical question of how long experiments should be stored, we observe that only 2% of all crystals that led to deposited structures appeared after 30 days. Overall, our data serve as practical guidelines for the design of initial screening experiments for new crystallization targets.

Item Type:	Article
Date Deposited:	27 Apr 2016 23:45
Last Modified:	27 Apr 2016 23:45
URI:	https://oak.novartis.com/id/eprint/27128