UC Berkeley

Secretion is emerging as a useful strategy for producing and purifying proteins of biotechnological interest from bacteria. The T3SS in Salmonella enterica is an ideal path to protein export because (a) the organism is genetically tractable; (b) the T3SS is non-essential for bacterial metabolism; and (c) the T3SS allows for target proteins to cross both bacterial membranes in one step. In this dissertation, we describe our work in characterizing the mechanisms of action that govern type III secretion, and adapting the system for new, synthetic functions.

First, we investigated the roles of two biophysical factors in protein secretion by the T3SS: free energy of unfolding (∆G) and refolding kinetics of target proteins, since the secretion pathway includes a protein-unfolding step. These experiments demonstrated that the T3SS preferentially exports proteins whose G values and folding rates fall within a specific range. This study provides insight for predicting which proteins are suitable for secretion by the T3SS, and how the system specificity may be engineered.

We also engineered a hyper-secreting strain of S. enterica for the high-titer production (100 mg/L) of a variety of heterologous proteins. To design the hyper-secreting strain, we investigated a single T3SS protein, SipD. We found that SipD has two distinct roles on different sides of the cell envelope, which have opposite effects on protein secretion. We have proposed a physical and chemical mechanism for this phenomenon, and we are currently using our hyper-secreting S. enterica strain to secrete enzymes, cytochromes, antimicrobial peptides and biopolymer proteins.

In secreting biopolymer proteins, we observed a 35-60% enhancement in the homogeneity of full-length forms of secreted pro-resilin, tropo-elastin, and silk proteins, as compared with proteins purified from the cytosol. We introduced several modifications to resilin-based peptides to make materials of sufficiently high quantity and yield for the production of antimicrobial hydrogels with reproducible rheological properties. The hydrogels permitted the growth of mammalian cells but inhibited growth of bacterial cells. The ease of this process lends support for using bacterial secretion to produce and purify other proteins that are traditionally difficult to express and isolate, and which may find use in multifunctional protein-based materials.

Finally, we modified the self-assembling T3SS filament protein, PrgI, as a scaffold for inorganic nanostructure synthesis. In T3SS assembly, about 120 copies of PrgI spontaneously assemble into a growing filament structure that extends outwards from the cell membrane. We engineered the flexible N-terminus of PrgI to specifically bind metal ions and particles, and used the resulting assemblies as a template to create metal-plated nanowires of tunable length and composition in vitro and on live cells. These cell-tethered metallic antennae could be used to controllably drive current from the cell cytosol to external electrodes via a non-native protein electron transfer chain localized in the cell membrane.