Modular Domain Compatibility Among Cytotoxic Necrotizing Factors – Finding a Universal Platform for BTIDD

Bacterial toxin‐inspired drug delivery (BTIDD) exploits the cellular entry and cytosolic delivery mechanisms of bacterial protein toxins to facilitate intracellular delivery of therapeutic agents. The Gln‐deamidase activity domains of the cytotoxic necrotizing factor (CNF) family are found in a diverse array of protein toxins (6 known AB‐type toxins), effectors of specialized secretion systems (T3SS and T6SS), and a number of putative homologs. This suggests that CNF‐like domains might be universal toxic cargos that can be swapped readily among various delivery systems. The CNF family of highly conserved modular toxins presents a unique opportunity to gain insights on how a delivery platform evolved for efficient delivery of its cognate cargo. We have swapped functional domains among CNF1, CNF2, CNF3 and CNFy at various joining sites to determine the effect on efficiency of cargo delivery into the host cell cytosol, using a cell‐based SRE‐luciferase downstream reporter assay. We first refined the limits of modularity among the CNF domains by testing three amino acid joining sites and found the most effective joining site is not the canonical catalytic domain start site, but rather sites surrounding a C‐terminal secondary binding domain. Under our defined assay conditions, CNFy is less efficient than CNF1 or CNF2. CNFy1 or CNFy2 chimeras with a CNFy delivery domain are less efficient than CNF1 or CNF2, suggesting the delivery domain dictates delivery efficiency. However, the delivery domains of CNF1 and CNF2 were unable to enhance delivery efficiency of the CNFy cargo. Combined, these results show that CNFy is limiting as both the cargo and the delivery system. We hypothesized that rather than one domain being responsible for the efficiency of cargo delivery, that cooperativity between the delivery domain and its cognate cargo is rate limiting, whereby specific amino acid residues within these domains recognize and interact with each other to facilitate an efficient translocation event. To test this, we generated chimeric constructs with CNF3 cargo delivered by CNF1, CNF2, and CNFy. Since CNF3 is most closely related to CNFy, as expected CNFy is most efficient at delivering CNF3 cargo. Although the CNF toxin family has high amino acid identity, the seemingly small variations in each peptide may account for subtle compensatory changes that enhance the delivery of cognate cargo over related domains. Using our cellular assays coupled with examining the alignment of these high identity proteins, we can trace these evolutionary adaptations and gain insight into the protein determinants that define toxin‐based delivery of cargo. This knowledge offers basis for generating tunable BTIDD systems to enhance delivery of specific cargo for clinical and research applications. Support or Funding Information DeBoer Fellowship UIUC Research Assistantship UIUC NIH/NAID AI038396 Chemistry Biology Interface Training Program Fellowship UIUC National Institute Of General Medical Sciences of the National Institutes of Health T32GM070421