124 One Step Protein Purification via the Type 3 Secretion System

Faculty Mentor: Fabienne Chevance (School of Biological Sciences, University of Utah)

Protein production via bacteria requires an intensive purification process. The cells need to be burst open, then proteins of interest need to be separated from among thousands of others. Killing the cells requires a new batch to be grown for every harvest, which takes time. Separation requires additional equipment and can be expensive on an industrial scale. Secreting proteins into the growth medium can circumvent both of these issues, and all one needs to do is spin down the culture to separate the cells from the secreted proteins.

Salmonella have such a system already built-in. The Type 3 Secretion System is used to construct flagella on the surface of the cell. It contains many moving parts and is an amazing example of a molecular machine in its own right. Of note, it has a genetic switch activated by a molecular ruler protein which measures the distance from the secretion apparatus at the base of the flagellar structure to the end of a flexible hook structure, which is a universal joint controlling the direction of the flagellar filament. Once the hook has grown to the appropriate length, the ruler interacts with the secretion apparatus to flip the switch and turn swap from expressing class 2 to class 3 genes (early secretion proteins to late).

My research has focused on co-opting this machinery to secrete proteins of interest that we have engineered into the cell. We are investigating multiple techniques, utilizing different stages of construction to optimize levels of secretion. One, the early method, involves preventing the switch from flipping, leaving the path through the cell wall open for secretion and utilizing the secretion signals of early proteins. Another flips the switch early by use of a gene fusion of the ruler and the hook cap and utilizes the late genes that are turned on to shuttle our proteins out of the cell, while also significantly reducing the length of the channel that these proteins are shuttled through.

We have shown both methods to be successful and are currently optimizing them. We hypothesize that secretion can be improved by adjusting the timing of the expression of our genes by use of different promoters; if they’re being built too fast, they may be clogging the secretion channel, for example. There may also be partial folding of our ruler fusion, electrostatic interactions with the secretion machinery, or even unknown interactions between components of said machinery which could reduce switch efficiency. These hurdles offer potential ways to increase secretion even further, and the project as a whole provides a potential technique to produce proteins cheaply and easily.