Unraveling molecular mechanisms involved in the action of Cytotoxic Necrotizing Factors (CNFs)

Cytotoxic necrotizing factors (CNFs) are protein toxins that are produced by a variety of pathogenic bacteria. As AB-type toxins, CNFs consist of a translocation subunit and a catalytic domain that deaminates Rho GTPases and induces toxicity resulting in the formation of stress fibers and multi-nuclei of host cells. Over 40 years ago, the first CNF member, CNF1, was discovered in E. coli isolated from enteritis-infected children. Research into the action mechanism of CNF has been commenced since then, and steps involved in the action process have been roughly studied, but there are still key details that need to be revealed.

CNFs share high sequence similarity, while they exhibit different properties in certain aspects. Therefore, it is interesting to study the mechanism by comparing different members and investigating the reasons for the dissimilarity. In this thesis, the crystal structure of CNF1 has been determined and compared to the crystal structure of the CNF from Yersinia pseudotuberculosis (CNFY), which was published several years ago. Like CNFY, CNF1 can be divided into five domains, although they arrange in a different way resulting in a less compact particle.

The endocytosis of CNFs is a procedure that depends on a receptor and on low pH. The initial step is binding the receptor(s). It is well-known that CNFY does not bind to the receptors that have been identified as the receptors of CNF1, but the reasons have never been clarified before. Hence, the complex of CNF1 and its known receptor Lu/BCAM has been prepared and analyzed by means of X-ray diffraction. From the structure of the complex, three regions located in two loops and one β-sheet, respectively, have been determined as interaction sites between CNF1 and Lu/BCAM. Swapping these three regions in CNF1 with the corresponding sequences of CNFY abolished the affinity of CNF1 for Lu/BCAM.

The catalytic domains of CNFs function as deaminases, specifically removing the amine group of certain glutamines in small GTPases, leading to their constitutively active forms. Although the structures of CNF1 and CNFY catalytic domains have been elucidated, no crystal structures of CNFs in complex with their substrates have been obtained, leaving the question of what determines their distinct substrate specificities open. To shed light on the underlying mechanism of CNFs' distinct substrate specificities, a complex of the CNF1 catalytic domain and its substrate, RhoA, with a linker connecting the two proteins was prepared. After crystallizing the complex, an analysis of the resulting crystal structure suggested that two loops in the catalytic domain are crucial in determining the substrate specificities of CNFs.

An unknown domain that enters the cytosol along with the catalytic domain was discovered within the structures of CNFY and CNF1. Efforts have been made in this thesis to elucidate the function of this domain, but no clear role could be assigned to this domain. However, the conducted experiments exclude some hypotheses and probably could provide some useful information for future work.