mcr-1-Mediated In Vitro Inhibition of Plasmid Transfer Is Reversed by the Intestinal Environment

Abstract

Colistin is regarded as an antibiotic of last resort against multidrug-resistant Gram-negative bacteria, including Klebsiella pneumoniae and Escherichia coli. Colistin resistance is acquired by microorganisms via chromosome-mediated mutations or plasmid-mediated mobile colistin resistance (mcr) gene, in which the transfer of mcr is the predominant factor underlying the spread of colistin resistance. However, the factors that are responsible for the spread of the mcr gene are still unclear. In this study, we observed that mcr-1 inhibited the transfer of the pHNSHP45 backbone in liquid mating. Similar inhibitory effect of mcr-1.6 and chromosomal mutant ΔmgrB suggested that colistin resistance, acquired from either plasmid or chromosomal mutation, hindered the transfer of colistin resistance-related plasmid in vitro. Dual plasmid system further proved that co-existing plasmid transfer was reduced too. However, this inhibitory effect was reversed in vivo. Some factors in the gut, including bile salt and anaerobic conditions, could increase the transfer frequency of the mcr-1-containing plasmid. Our results demonstrated the potential risk for the spread of colistin resistance in the intestine, provide a scientific basis against the transmission of colistin resistance threat.

Keywords: Escherichia coli; Klebsiella pneumoniae; bile salt; colistin resistance; conjugation; mcr-1.

Figure 1

Different conjugation systems for quantitative measurement of plasmid transfer rate. The pHNSHP45 plasmid was transferred from K. pneumoniae A2312NM to K. pneumoniae D20-2 through a filter or liquid mating. Cultures were incubated till the log phase. Donor and recipient strains were mixed in equal proportions and concentrated 50-fold. The mixture was incubated on a filter or suspended in a liquid medium for 4 h at 37 °C, after which the plasmid transfer frequency was measured. Data are the mean and SD of three independent experiments. Significance was determined using t-test; ** p < 0.01.

Figure 2

mcr-1 effect on conjugal transfer of pHNSHP45 plasmid. (A) Growth curve of K. pneumoniae A2312NM carrying different plasmids. (B) Growth competition assay between K. pneumoniae A2312NM harboring mcr-1 plasmid or Δmcr-1::Km^R plasmid. Samples were transferred to fresh LB medium every day, and CFU was calculated with appropriate antibiotic. Transfer frequency of mcr-1 plasmid between different strains, including K. pneumoniae A2312NM to K. pneumoniae D20-2 (red) or A1502 (black) (C), E. coli MG1655 to E. coli Nissle 1917 (D), and K. pneumoniae A2312NM to E. coli MP13 (E). An equal volume of donor strain harboring the mcr-1 plasmid, Δmcr-1::Km^R plasmid or mcr-1^c plasmid, and recipient strain were mixed and concentrated 50-fold. Conjugation was performed for 4 h at 37 °C.

Figure 3

The impact of plasmid or chromosome mediated colistin resistance on plasmid transfer. (A) Transfer frequency of mcr-1.6 plasmid and Δmcr-1.6::Apra^R plasmid from K. pneumoniae A2312NM to E. coli MP13. (B) mcr-1-related plasmid transfer frequency. Donor strains were A2312NM ΔmgrB containing Δmcr-1::Km^R plasmid and A2312NM containing mcr-1 plasmid or Δmcr-1::Km^R plasmid. They were conjugated individually with the recipient strain K. pneumoniae D20-2. Data are mean and SD of three independent experiments. Significance was determined using t-test; * p < 0.05, ** p < 0.01.

Figure 4

The impact of colistin resistance on the transfer frequency of co-existing plasmid. (A) A dual plasmid system was constructed with pRK2013 combined with an mcr-1 plasmid (orange) or Δmcr-1::Apra^R plasmid (purple) in the donor strain K. pneumoniae A2312NM. (B) Conjugation frequency of mcr-1 or Δmcr-1::Apra^R plasmid (left) and pRK2013 in the dual plasmid system was measured separately. (C) Conjugation frequency of pRK2013 with donor K. pneumoniae A2312NM WT or A2312NM ΔmgrB and recipient K. pneumoniae D20-2. Data are the mean and SD of more than three independent experiments. The significance was determined using t-test; ** p < 0.01.

Figure 5

Influence of oxygen and bile salt on plasmid transfer. (A) Plasmid transfer in vivo. Five-week-old CD-1 mice were pretreated with streptomycin. Approximately 10⁸ cells of different donor strains, namely, K. pneumoniae A2312NM harboring mcr-1 plasmid, A2312NM containing the Δmcr-1::Km^R plasmid, were mixed with recipient strain K. pneumoniae MP13 separately and immediately administered to each mouse intragastrical. Fecal samples were collected after 3 days, and the transfer frequency was calculated. (B) The effect of mucin and bile salt on the transfer of mcr-1 plasmid. Conjugation was performed with or without additional mucin or bile salt. (C) Conjugation was performed with or without bile salt under different oxygen concentrations. Data are the mean and SD of three and more independent experiments. The significance was determined using t-test; ns, no significance; ** p < 0.01, *** p < 0.001.