Evert den Drijver

138 Chapter 7 present study. All encountered variants seem to result in a sequence that resembles the E. coli consensus σ70 promoter more than the wild-type sequence they are derived from (Tracz et al. 2007). Next to the −42C>T promoter mutation, we detected 23 other positions in our analysis that are associated with CTX resistance and have extremely high levels of homoplasy. Most of these are synonymous mutations, with only two missense mutations (thiM and gspL) found. It is remarkable that one missense mutation (p.Ser330Thr) is located in gspL that encodes a protein of the type II secretion system. The type II secretion system is used by many Gram-negative bacteria to translocate folded proteins from the periplasm, through the outer membrane, into the extracellular milieu (Korotkov, Sandkvist, and Hol 2012). The system is composed of 12–15 different general secretory pathway (Gsp) proteins and is related to virulence of various pathogenic E. coli , e.g. EHEC (enterohaemorrhagic E. coli ) and UPEC (uropathogenic E. coli ) (Ho et al. 2008; Baldi et al. 2012; Kulkarni et al. 2009). It could be that in our selection of mainly clinical samples a certain predilection has occurred towards isolates with particular virulence traits and not based on mechanistic benefits. The gspL gene has been described as being part of the accessory genome of E. coli (Dunne et al. 2017). Our study supports this finding as some strains did not harbour this gene. Additionally, we found evidence that recombination events in the type II secretion system could be the underlying cause of the extreme homoplasy levels. Still, it is remarkable that missense mutation p.Ser330Thr in the gspL gene correlates with the CTX-resistance trait even though it is most likely caused by a recombination event. To the best of our knowledge, no relationship between the type II secretion system and CTX resistance has been observed before. One could hypothesize that the mutation is a secondary adaptation needed to cope with the elevated AmpC production, as the peptidoglycan (PG) layer is affected by AmpC hyperproduction and the type II secretion system contains proteins that are partly localized in the periplasm (Vanderlinde et al. 2017; Juan et al. 2018). The use of genomic data to detect homoplasy events is an accepted scientific technique (Read and Massey 2014; Chen and Shapiro 2015; B. J. Shapiro et al. 2009). In Mycobacterium tuberculosis , it is a well-known method for identifying advantageous mutations, as they are likely to be associated with phenotypes such as drug resistance, heightened transmissibility or host adaptation (Farhat et al. 2013; Mortimer, Weber, and Pepperell 2018; Ruesen et al. 2018; Miotto et al. 2014). In other species, e.g. Staphylococcus aureus and Burkholderia pseudomallei , the method has been effectively applied to identify mutations associated with antibiotic resistance or virulence-associated genes (Alam et al. 2014; Sahl et al. 2015). Homoplasy-based

RkJQdWJsaXNoZXIy MTk4NDMw