Emergence of a novel mobile colistin resistance gene, mcr-8, in NDM-producing Klebsiella pneumoniae

Abstract

The rapid increase in carbapenem resistance among gram-negative bacteria has renewed focus on the importance of polymyxin antibiotics (colistin or polymyxin E). However, the recent emergence of plasmid-mediated colistin resistance determinants (mcr-1, -2, -3, -4, -5, -6, and -7), especially mcr-1, in carbapenem-resistant Enterobacteriaceae is a serious threat to global health. Here, we characterized a novel mobile colistin resistance gene, mcr-8, located on a transferrable 95,983-bp IncFII-type plasmid in Klebsiella pneumoniae. The deduced amino-acid sequence of MCR-8 showed 31.08%, 30.26%, 39.96%, 37.85%, 33.51%, 30.43%, and 37.46% identity to MCR-1, MCR-2, MCR-3, MCR-4, MCR-5, MCR-6, and MCR-7, respectively. Functional cloning indicated that the acquisition of the single mcr-8 gene significantly increased resistance to colistin in both Escherichia coli and K. pneumoniae. Notably, the coexistence of mcr-8 and the carbapenemase-encoding gene bla_NDM was confirmed in K. pneumoniae isolates of livestock origin. Moreover, BLASTn analysis of mcr-8 revealed that this gene was present in a colistin- and carbapenem-resistant K. pneumoniae strain isolated from the sputum of a patient with pneumonia syndrome in the respiratory intensive care unit of a Chinese hospital in 2016. These findings indicated that mcr-8 has existed for some time and has disseminated among K. pneumoniae of both animal and human origin, further increasing the public health burden of antimicrobial resistance.

Fig. 1. Location of mcr-8 in Klebsiella pneumoniae KP91 and its transconjugants.

a XbaI-digested PFGE of K. pneumoniae KP91, transconjugants, and the recipient E. coli strain J53. b S1-PFGE and c the corresponding Southern hybridization using the mcr-8-specific probe. Lane M, marker H9812; lane 1, K. pneumoniae KP91; lane 2, transconjugant E. coli J53-pKP91; lane 3, recipient E. coli J53

Fig. 2. The genetic contents of plasmid pKP91.

Circles display (outside to inside) (i) size in bp and (ii) the positions of predicted coding sequences transcribed in a clockwise orientation

Fig. 3. WebLogo of amino-acid sequences of MCR-1–8.

A logo represents each column of the alignment in a stack of letters, with the height of each letter proportional to the observed frequency of the corresponding amino acid, and the overall height of each stack proportional to the sequence conservation, measured in bits, at that position. Red asterisks indicate the six conserved cysteine residues that form three disulfide bonds. Black asterisks indicate the conserved active sites among MCR-1–8. The logo was built using WebLogo software (http://weblogo.berkeley.edu/logo.cgi)

Fig. 4. Predicted crystal structure and transmembrane domain of MCR-8.

a Structure prediction for MCR-8 and reference protein MCR-3. Domain 1 was predicted to be a transmembrane domain, while domain 2 was predicted to be a phosphoethanolamine transferase. b The five transmembrane α-helices predicted by the Philius transmembrane prediction server (type confidence, 0.99; topology confidence, 0.88)

Fig. 5. Comparison of the genetic environments of mcr-8.

K. pneumoniae WCHKP1845 was isolated from the sputum of patients in China (accession no. MPOD01000000). The positions and orientations of the genes are indicated by arrows, with the direction of transcription shown by the arrowhead. Gray shading indicates > 90% nucleotide sequence identity

Fig. 6. PFGE analysis of mcr-1- and mcr-8-positive Klebsiella pneumoniae isolates as well as other K. pneumoniae isolates.

XbaI was used for digestion of the genomic DNA. All K. pneumoniae isolates were collected from pig fecal matter and chicken cloacae samples from Shandong, China

Fig. 7

Phylogenetic tree of the deduced amino-acid sequences of 20 putative phosphoethanolamine transferases from different bacterial species and MCR-1–7 with MCR-8 using CLC Genomics Workbench 9 (CLC Bio-Qiagen, Aarhus, Denmark)