| A major hurdle facing in vitro protein characterization is obtaining soluble protein from targets that tend to aggregate and form insoluble inclusion bodies. Soluble protein is essential for any biophysical data collection and new methods are needed to approach this significant problem. Directed evolution can be used to discover mutations which lead to improved solubility using an appropriate screening method. Green fluorescent protein (GFP) has been shown to be an effective solubility reporter which can be used to screen for soluble protein variants. We have chosen three diverse enzymes as targets for improving protein solubility using this technique: arachidonate 5-lipoxygenase---an enzyme which converts fatty acids into leukotrienes, PhnG---an enzyme belonging to the bacterial carbon-phosphorus lyase pathway, and RebG---a glycosyltransferase. Error-prone PCR and DNA shuffling were used to generate libraries of mutants which were subsequently cloned into a GFP-fusion screening vector. From the evolution of 5LO and RebG, much was learned about the optimization of the protocols involved in this methodology, including valuable information about how to avoid common "false-positive" results in which fluorescent colonies arise while screening but do not represent an improvement of the target. Evolution of these two targets did not result in an improvement of solubility, however truncation strategies may still prove to be effective, and more work needs to be done in this area. Evolution of PhnG successfully produced one variant, named clone B6, which showed both an improvement in expression and folding over wild type PhnG. It was also discovered that GFPuv can act as an effective solubility enhancing fusion tag for PhnG. Prior to the current studies PhnG had not been effectively expressed and purified in E. coli, however purification and refolding of resolubilized inclusion bodies of the clone B6 PhnG-GFP fusion construct was shown to yield enough soluble protein for future crystallographic studies. |
