Structural basis of bacterial toxin activation by prodrug-activating peptidases

Abstract

Bacteria produce specialized metabolites to compete with other microbes for limited resources. These metabolites give the producing organism an advantage by either allowing it to scavenge essential nutrients as in the case of siderophores, or by directly killing other organisms as in the case of toxins. Many of the molecular targets of toxins are conserved molecular machines involved in vital processes such as transcription of DNA or translation, so bacteria must protect their own integrity while producing these toxins. In one resistance mechanism, bacteria produce non-toxic precursors in the cytoplasm and control activation of these precursors so that it happens after they are exported from the cell. In this thesis I focus on one such mechanism, in which cleavage of N-acyl-D-asparagine prodrug motifs from the precursors by membrane-bound peptidases (prodrug-activating peptidases) either exposes chemical groups necessary for toxicity or enables chemical reactions that generate the toxic species. According to their architecture, prodrug-activating peptidases can be classified as either type I, which have an S12 hydrolase domain followed by a transmembrane domain of three helices, or type II, which contain an additional half ABC transporter. In this thesis I present our work to understand the structure, biological assembly, and function of both types of prodrug-activating peptidases. In Chapter 2, I describe the structure and the basis of substrate specificity of ClbP, the type I peptidase that activates colibactin. I discovered that the transmembrane domain of ClbP adopts a novel fold and plays an essential role by binding to the acyl tail in the prodrug motif of its substrate. I also discovered that ClbP forms a homodimer, which may be important for processing pseudodimeric precolibactin. In Chapter 3, I describe the structure and biological assembly of the type II peptidase that activates zwittermicin ZmaM. We discovered that the two peptidase modules in the ZmaM dimer are oriented like in the ClbP dimer, and that the fused ABC transporter is encased by the cavity formed by this dimer. The interactions between the ClbP-like module and the ABC transporter in ZmaM suggest that there may be allosteric communication and coordination of function between the two modules. In Chapter 4, I discuss the implications of our work on ClbP and ZmaM for understanding bacterial toxin secretion and activation systems, including in a potential association between type I peptidases and ABC transporter partners in the cell