F TAGINI¹, A KOUTSOKERA², T PILLONEL¹, A LOVIS², G GREUB<sup>1

1</sup>Institute of Microbiology, Department of Laboratory, University of Lausanne and University Hospital Center, Lausanne, Switzerland, ²Service of Pneumology, Department of Medicine, University Hospital Center Lausanne, Switzerland

Corynebacterium diphtheriae is the causal agent of diphtheria. Classical diphtheria is due to the production of a prophage encoded toxin (tox), against which potent vaccines have shown efficacy since 1923. However, there is a recent increase of non-toxigenic C. diphtheriae infections. We performed here a genomic analysis of a strain isolated in a bronchoalveolar lavage taken from a twenty-two year-old immunocompetent male with tracheo-bronchitis. Genome assembly was done using SPAdes Genome Assembler 3.6.2. 174 contigs of more than 1000 base pairs (bp) were obtained for a total length of 2.89 megabases (Mb) and a GC content of 53.76%. In comparison, genomes sizes of the other 14 complete genomes available on NCBI range from 2.395 Mb to 2.535 Mb. Moreover, the Average Nucleotide Identity (ANI) between our strain and the other complete genomes range from 95.15% to 95.36% whereas the ANI of the reference genome (NCTC13129) compared to the others range from 98.24% to 98.56%. Ten prophages sequences were predicted using PHAST and could partially explain the higher amount of DNA found in the genome. Interestingly, an incomplete prophage, which does not bear the toxin gene, is present at the same locus where the usual tox+ corynephage inserts. We found only 1 putative Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) region with 3 direct repeats and 2 spacers. C. diphtheriae is usually associated with 2 CRISPR regions of a larger size, which could be a reason of the elevated number of predicted prophages. After BLASTing the Virulence Factor Database on our assembly, we found 39 hits with > 75.0% identity and an e-value < 10^-5. Thirty-one of them are involved in the iron-uptake or siderophore system and we noted the presence of a conserved diphtheria toxin regulator (dtxR) gene. Two interesting genomic islands were found with IslandViewer 3 and one of them bears 8 genes coding for the siderophore system. In conclusion, thanks to sequencing, we were able (i) to determine the absence of the toxin quickly, which had a major impact on patient care and (ii) to characterize this very divergent pathogenic strain holding an unusual high number of prophages.