Genomic epidemiology of global VIM-producing Enterobacteriaceae

Abstract

Background: International data on the molecular epidemiology of Enterobacteriaceae with VIM carbapenemases are limited.

Methods: We performed short read (Illumina) WGS on a global collection of 89 VIM-producing clinical Enterobacteriaceae (2008-14).

Results: VIM-producing (11 varieties within 21 different integrons) isolates were mostly obtained from Europe. Certain integrons with bla VIM were specific to a country in different species and clonal complexes (CCs) (In 87 , In 624 , In 916 and In 1323 ), while others had spread globally among various Enterobacteriaceae species (In 110 and In 1209 ). Klebsiella pneumoniae was the most common species ( n = 45); CC147 from Greece was the most prevalent clone and contained In 590 -like integrons with four different bla VIM s. Enterobacter cloacae complex was the second most common species and mainly consisted of Enterobacter hormaechei ( Enterobacter xiangfangensis , subsp. steigerwaltii and Hoffmann cluster III). CC200 (from Croatia and Turkey), CC114 (Croatia, Greece, Italy and the USA) and CC78 (from Greece, Italy and Spain) containing bla VIM-1 were the most common clones among the E. cloacae complex.

Conclusions: This study highlights the importance of surveillance programmes using the latest molecular techniques in providing insight into the characteristics and global distribution of Enterobacteriaceae with bla VIM s.

Figure 1

Global distribution of VIM-producing Enterobacteriaceae isolates in this study.

Figure 2

Phylogenetic tree of VIM-producing Klebsiella spp. This maximum-likelihood phylogram is based on a 3737 806 bp core genome and a total of 369 829 SNPs. Core genome was identified using K. pneumoniae subsp. pneumoniae ATCC BAA-2146 as a reference genome. Tree includes 47 study isolates and five reference strains (marked with asterisks). Tree is rooted by using the outgroup of K. quasipneumoniae ATCC 700603 and asterisks indicate bootstrap support >90% from 100 replicates. In the ‘Subspecies’ column, K. variicola and K. quasipneumoniae (marked with asterisks) are not subspecies of K. pneumoniae, but distinct species. STs 2287–2292 were novel types found in this study. A CC marked with an asterisk was distributed internationally. Integron numbers with asterisks were partially characterized (Dataset S3). ‘OmpK35’ and ‘OmpK36’ columns indicate predicted mutation of porins: W, WT; D, deficient (due to premature stop codon); V, variant associated with increased MIC of carbapenems; U, variant with unknown significance. Virulence genes of clbA-R (colibactin), iroBCDN (salmochelin) and rmpA were sought, but not found. UD, undetermined.

Figure 3

Phylogenetic tree of VIM-producing Enterobacter spp. This maximum-likelihood phylogram is based on a 1738 728 bp core genome and a total of 511 679 SNPs. Core genome was identified using E. cloacae subsp. cloacae ATCC 13047 as a reference genome. Tree includes 33 study isolates and 19 reference strains (marked with asterisks). Tree is rooted by using the outgroup of E. aerogenes KCTC 2190 and asterisks indicate bootstrap support >90% from 100 replicates. ‘Group’ column indicates E. cloacae complex groups defined by Chavda et al. ST512, ST514 and ST520 were novel types found in this study. CC marked with an asterisk was distributed internationally. Integron numbers with asterisks were partially characterized (Dataset S3). UD, undetermined.

Figure 4

Phylogenetic tree of VIM-producing Citrobacter spp. This maximum-likelihood phylogram is based on a 2 406 029 bp core genome and a total of 594 405 SNPs. Core genome was identified using C. freundii CAV1321 as a reference genome. Tree includes six study isolates and eight reference strains (marked with asterisks). Tree is rooted by using the outgroup of Citrobacter koseri ATCC BAA-895 and asterisks indicate bootstrap support >90% from 100 replicates. An ST marked with an asterisk was distributed internationally. STs 95, 96, 98 and 101 were novel types found in this study. UD, undetermined.