Molecular analysis of the interaction between Clostridium perfringens Enterotoxin and Claudins

Doctoral thesis German OPEN
Protze, Jonas (2015)
  • Publisher: Freie Universität Berlin Universitätsbibliothek, Garystr. 39, 14195 Berlin
  • Subject: Protein–protein interactions | Structure–function study | Membrane proteins | Molecular modeling | 572 Biochemistry | 570 Life sciences | Tight junctions | Mutagenesis | 570 Biowissenschaften; Biologie | 572 Biochemie
    • ddc: ddc:570 | ddc:572

Claudins are essential constituents of Tight Junctions (TJs) and responsible for maintenance of these cell-cell contacts. Binding of <i>Clostridium perfringens</i> Enterotoxin’s C-terminal domain (cCPE) to the extracellular loop 2 (EZS2) of claudins, especially Cld3, Cld4 and Cld6-Cld9 causes a reversible opening of TJs. Thus, a structure-function analysis of this system is relevant for biomedical application, since cCPE could be used to enhance paracellular drug uptake. Furthermore cCPE respectively CPE could be used for detection or treatment of Cld3 or Cld4 overexpressing carcinomas. Particularly a Cld5 binding cCPE-variant would be of interest to enhance the permeability of the blood-brain barrier for cytostatics in cases of brain-tumours. However, this needs reprogramming of cCPE’s specificity. Determinants of the cCPE-claudin interaction and cCPE’s specificity for certain Cld-subtyps remained unclear, since no structural information is available for claudins. Therefore, in the course of this work, detailed structural models of cCPE with the EZS2 of Cld3 and Cld4 were created by combining structural bioinformatics and methods of molecular biology in an iterative process. Subsequently, these Models were used to obtain a rational designed cCPE-variant with shifted Cld-subytyp specificity towards Cld5. Furthermore, first steps towards a usage of cCPE as a diagnostic probe to detect overexpression of certain Cld-subtypes were taken by chemical modification of cCPE with fluorophores as well as a cage compound for Xenon. In detail, utilizing cCPE’s crystal structure and the results of cellular binding studies a full mapping and characterisation of cCPE’s Cld-binding pocket was carried out and homology models of the EZS2 of Cld3 and Cld4, as well as cCPE-Cld interaction models, were created. Moreover, by directed substitution of residues in cCPE and claudins, one of the key interactions of cCPE-Cld binding was identified, and the orientation of the EZS2 in cCPE’s Cld-binding pocket was determined. L<SUB>150/151</SUB> (Cld3/4) interacts with a deep and strongly hydrophobic pit within the Cld-binding pocket, which is formed by Y<SUB>306</SUB>, Y<SUB>310</SUB> and Y<SUB>312</SUB> (triple-Tyr pit). Transfer of the knowledge gained by studying cCPE’s binding to Cld3 and Cld4 to Cld1 and Cld5, sequence comparison of the EZS2 of classic claudins, as well as following binding studies, resulted in the identification of two residues each within the EZS2 of Cld1 and Cld5, which weaken (Cld1) or block (Cld5) the interaction with cCPE. Consequently, the interaction models were utilized to design cCPE-variants with a changed Cld-binding pocket, allowing binding to Cld5. By downsizing the triple-Tyr pit of cCPE with a substitution of tyrosine (Y<SUB>306</SUB>) to tryptophan, in combination with a substitution of serine (S<SUB>313</SUB>) to histidine, which narrows the Cld-binding pocket, a cCPE-variant with nanomolar affinity for Cld5 was created. This new cCPE-variant (cCPE<SUB>Y306W/S313H</SUB>) also shows decreased binding to classic CPE-receptor claudins (Cld3, Cld4, Cld6-Cld9). Apart from characterizing the cCPE-Cld interaction, a full model of Cld5 was established. This model is based on experimental data, determined by Jan Rossa as well as evolutionary contacts. It provides first insights into the arrangement of the 4 transmembrane helices of claudins and is also useful to unravel the complex <i>cis</i>- and <i>trans</i>-interactions of claudins, which have a crucial role in organisation and function of tight junctions. The main result of this study is that two rational chosen substitutions (Y<SUB>306</SUB>W/S<SUB>313</SUB>H) change cCPE’s binding behaviour to bind Cld5 with a nanomolar affinity. This new cCPE-Variant (cCPE<SUB>Y306W/S313H</SUB>) provides a good basis for development of a Cld5-specific binder and is a proof of principle that cCPE can be used to design variants targeting non-CPE-re¬cep¬tor claudins. Additionally, two cCPE-variants with improved specificity for Cld3 (cCPE<SUB>L223A/D225A/R227A</SUB>) and Cld4 (cCPE<SUB>L254A/S256A/I258A/D284A</SUB>) were generated in the course of this work. Thus, it is now for the first time possible to selectively target distinct Cld-subtypes with cCPE-variants, an important step for the development of cCPE as TJ-modulator, drug delivery system or tool for detection and treatment of claudin overexpressing tumours. In sum, the results provide detailed insights into the molecular structure and function of claudins and the cCPE-claudin interaction, which on the one hand contribute to new possibilities for the pharmacological development of subtype specific Cld-binders based on CPE and cCPE. On the other hand, the experimentally verified full model of Cld5 now allows to study oligomerisation and organisation of claudins in tight junctions by targeted substitutions.
Share - Bookmark