A deadly hug : contact-dependent killing by Caulobacter crescentus, via cell surface-associated glycine-zipper proteins

Abstract

In the battle for resources within microbial communities, antagonistic interactions between bacterial species are often mediated by diffusible inhibitory compounds, which can be diffusible or delivered in a contact-dependent manner. Bacteriocins are one ubiquitous type of such antimicrobials, and collectively constitute a very diverse group of ribosomally-synthesized diffusible proteinaceous toxins. The majority of well-characterized bacteriocin systems belong to a limited group of bacterial clades and environments. In contrast, interbacterial interactions are poorly characterized in the nutrient-poor aquatic environments where the a-proteobacterium Caulobacter crescentus thrives. Here, I describe the discovery and characterization of a new type of bacteriocin in C. crescentus. The Cdz toxin is composed of two small hydrophobic proteins, each harboring an extended glycine-zipper motif often found in amyloids. These proteins are retained on the surface of producer cells where they form large, insoluble aggregates. I show that Cdz mediates cell contact-dependent killing of closely related species. The Cdz bacteriocin uses a type I secretion system and is unrelated to previously described contact-dependent inhibition systems. Delivering the bacteriocin directly to the recipient, rather than secreting it into the extracellular milieu, likely enables C. crescentus to avoid the dilemma of producing an expensive common good that would rapidly diffuse away in its aqueous environment. Furthermore, Cdz-like systems are found in many clades of bacteria, including pathogens such as Klebsiella pneumoniae and Pseudomonas aeruginosa, suggesting that this form of contact-dependent inhibition is widespread. Using a cationic membrane stain, I showed that the Cdz bacteriocin causes inner membrane depolarization in the target cells. To further characterize the mechanism of delivery and toxicity, I conducted a suppressor screen to identify mutations that confer Cdz resistance to a target strain. I identified the putative surface receptor for the Cdz toxins, PerA, a protein harboring several pentapeptide-repeat motifs thought to adopt a quadrilateral [beta]-helical fold. PerA plays an important role in envelope homeostasis. Additionally, I identified two envelope-remodeling genes whose upregulation confers resistance to Cdz-mediated killing. Taken together, my work has expanded the repertoire of bacteriocins, demonstrating that these antimicrobials can be contact-dependent and, consequently, advantageous in a wider range of environments than previously anticipated.