Tagging of the type three secretion system basal apparatus of enterohaemorrhagic Escherichia coli 0157:H7

• Main Author: Emmerson, James R.
• Published: University of Edinburgh 2008
• Subjects: 571.29
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.650016

The aim of this project was to label inner membrane basal apparatus proteins with fluorescent markers or immunogenic tags in order to investigate their function, regulation and localization. Cloning strategies were designed to insert tags at the 3’ end of eight genes encoding putative inner membrane proteins and to use these fusions to replace the wild type sequences by allelic exchange. From the eight strategies, <i>escR </i>and <i>escU</i> were taken forward to produce five EHEC O157:H7 strains. One strain contained EscR labelled at the C-terminus with an HA (haemagglutinin) epitope tag and the other four strains contained different labelled versions of EscU. Work with these strains demonstrated that the T3SS could not be visualized using fluorescence microscopy. However, Western blot analysis did show that the EscU protein was cleaved into 30kDa and 10kDa peptides, both of which localized to the membrane fraction of the bacterial cell. This cleavage was most likely occurring at the conserved cleavage site NPTH. Interestingly all the mutant strains constructed did not secrete a detectable level of EspD, apart from one fusion that had previously been shown to cleave the tag from EscU. This indicated that, despite the small size of the HA tag, all the tags interfered with the function of EscR or EscU. Fusions to the 10kDa fragment of EscU, along with a deletion of this domain, were used to elucidate its function in the T3SS. All the EscU mutants did not form EspA filaments, secrete EspD at wild type levels or secrete detectable levels of Tir, whilst the expression from the LEE1-5 promoters remained unaffected. These phenotypes could not be restored upon supplying the 10kDa peptide <i>in trans. </i>It is proposed that the uncleaved EscU protein is needed to secrete EscF. After a defined period EscU is cleaved (possibly auto-catalytically) and this allows the secretion of EspA, B and D. The lack of wild type EscU in the mutant strains may not allow EscF to be secreted or assembled correctly and this in turn leads to the inhibition of EspA, B and D translocation. It was evident that even minimal changes were not tolerated and inhibited secretion by the system. Future research will have to proceed with either alternative target proteins or the generation of high affinity antibodies coupled with sensitive imaging technology.