Many Gram-negative bacteria utilize contact-dependent growth inhibition (CDI) systems to deliver toxins to nearby cells using a large (~350kD) filamentous protein, CdiA, expressed on their cell surface. Recently, we discovered that biogenesis of CdiA is tightly regulated such that the N-terminal half of CdiA is first exported to form a 33nm stick on the cell surface, while the C-terminal half remains in the periplasm. When CdiA recognizes a receptor on a neighboring cell, the periplasmic half resumes secretion and penetrates the neighboring cell’s outer membrane to deliver the C-terminal toxin domain. This mode of programmed secretion is crucial for CdiA function because prematurely secreting its C-terminal half releases CdiA off the cell surface and renders it ineffective. Since there is no obvious source of energy or potential difference across the outer membrane, it is intriguing how secretion of a large protein like CdiA is arrested at a precise configuration only to be resumed by a receptor-binding event.

This thesis will highlight the recent findings about this mechanism of programmed secretion of CdiA. Deletions and sequence-replacement mutants have identified a 50-residue long, Tyr-Pro enriched, region in the periplasmic half of the protein responsible for secretion arrest. The CdiA receptor-binding event specifically relays a signal to release this arrest and resume secretion. Studies with a chimeric CdiA-DHFR have shown that the periplasmic CdiA domain must first unfold in order to resume export. Substrate-induced folding of the periplasmic DHFR fusion disrupts secretion restart. Interestingly, enzymatic fusions to CdiA demonstrate that the system functions as a nanoscale switch to direct the enzymatic activity to the extracellular space only when CdiA binds a receptor. Lastly, the filamentous N-terminal CdiA is a robust toxin delivery machine which tolerates shortening and severing at its base.