Plasmid-mediated colistin resistance among human clinical Enterobacterales isolates: national surveillance in the Czech Republic

Abstract

The occurrence of colistin resistance has increased rapidly among Enterobacterales around the world. We performed a national survey of plasmid-mediated colistin resistance in human clinical isolates through a retrospective analysis of samples from 2009 to 2017 and a prospective sampling in 2018-2020. The aim of this study was to identify and characterize isolates with mcr genes from various regions of the Czech Republic using whole genome sequencing (WGS). Of all 1932 colistin-resistant isolates analyzed, 73 (3.8%) were positive for mcr genes. Most isolates carried mcr-1 (48/73) and were identified as Escherichia coli (n = 44) and Klebsiella pneumoniae (n = 4) of various sequence types (ST). Twenty-five isolates, including Enterobacter spp. (n = 24) and Citrobacter freundii (n = 1) carrying the mcr-9 gene were detected; three of them (Enterobacter kobei ST54) co-harbored the mcr-4 and mcr-9 genes. Multi-drug resistance phenotype was a common feature of mcr isolates and 14% (10/73) isolates also co-harbored clinically important beta-lactamases, including two isolates with carbapenemases KPC-2 and OXA-48. Phylogenetic analysis of E. coli ST744, the dominant genotype in this study, with the global collection showed Czech isolates belonged to two major clades, one containing isolates from Europe, while the second composed of isolates from diverse geographical areas. The mcr-1 gene was carried by IncX4 (34/73, 47%), IncHI2/ST4 (6/73, 8%) and IncI2 (8/73, 11%) plasmid groups. Small plasmids belonging to the ColE10 group were associated with mcr-4 in three isolates, while mcr-9 was carried by IncHI2/ST1 plasmids (4/73, 5%) or the chromosome (18/73, 25%). We showed an overall low level of occurrence of mcr genes in colistin-resistant bacteria from human clinical samples in the Czech Republic.

Keywords: Enterobacterales; antibiotic resistance; human; mcr; plasmids.

Figure 1

Phylogenetic tree of E. coli isolates with mcr-1 of Czech clinical origin. The metadata in columns represents phylogenetic group (PG); sequence type (ST); type of sample (Sample type): urine (UR), blood (BL), rectal swab (RS), tonsil swab (TS), wound swab (WS), decubitus swab (DS), urethra swab (US); year of isolation (Year) and city where is the hospital related to the isolate recovery (Hospital): Novy Jicin (NJ), Prague (PR), Ostrava (OS), Karvina (KA), Ostrava-Poruba (OR), Opava (OP), Pribram (PI), Plzen (PL), Usti nad Labem (UL), Brno (BR). The turquoise squares represent presence (full square) or absence (empty square) or respective beta-lactamase encoding genes divided as AmpC (bright turquoise), ESBL (medium) and narrow-spectrum beta-lactamases (dark). The next section (mcr-1 on) reveals which plasmid carried mcr-1 gene; the size of the plasmid (Plasmid size) in bp while approx. Sizes (~) are estimated based by S1-PFGE, while the more precise values come from plasmid sequencing; the success of conjugative transfer is indicated (Transfer of mcr-1). The heat map in the last section indicated the amount of antibiotic resistance genes carried by the respective isolate in specified category of antibiotics from zero (white) to maximum of six (dark purple).

Figure 2

Phylogenetic tree of Enterobacter spp. isolates of Czech clinical origin. The metadata specify the species (Species); sequence type (ST), type of sample (Sample type): urine (UR), throat swab (TSH), sputum (SP), blood (BL), tonque swab (TOS), skin swab (SS), wound swab (WS), rectal swab (RS); year of isolation (Year) and city where is the hospital related to the isolate recovery (Hospital): Ostrava (OS), Pribram (PI), Brno (BR), Karlovy Vary (KV), Prague (PR), Brno-Bohunice (BB), Novy Jicin (NJ), Strakonice (ST). The color squares represent presence (full square) or absence (empty square) or respective antibiotic resistance genes divided to genes encoding resistance to colistin (pink), carbapenemases (blue) and other beta-lactamases (turquoise, see legend Figure 1). The heat map in the last section indicated the amount of antibiotic resistance genes carried by the respective isolate in specified category of antibiotics from zero (white) to maximum of five (dark purple).

Figure 3

Phylogenetic tree of Czech clinical E. coli ST744 isolates with selected sequences from global collection. The red labels indicate isolates coming from this study. The metadata specifies country of origin (Country): China (CH), The United States (US), Poland (PO), Czech Republic (CZ), Germany (GE), Norway (NO), Nigeria (NI), Romania (RO), United Kingdom (UK), Thailand (TH), Ireland (IR), France (FR), Ukraine (UK), Australia (AU), Russia (RU), Switzerland (SW), not available (NA), the source of origin (Source) and the year of isolation (Year). The colour squares represent presence (full square) or absence (empty square) or respective antibiotic resistance genes divided to genes encoding resistance to colistin (pink), carbapenemases (blue) and other beta-lactamases (turquoise, see legend Figure 1).

Figure 4

BRIG comparison of mcr-1.1-encoding IncHI2/ST4 plasmids. The plasmid pMCR1-59496 from K. pneumoniae ST726 identified in our study from urine sample (OP950836) was used as a reference. Two other plasmids originated from our collection, including pMCR1-43934 from E. coli ST8186 from tonsil swab sample and pMCR1-51133 from E. coli ST117 from a urine sample. The sequence alignment includes pMCR_915_C1 (MT929284.1) and pMCR_1085_C1 (MT929286.1) from E. coli recovered from raw turkey meat imported to the Czech Republic from Poland and one of plasmid pKP121-1-mcr (CP031850.1) from K. pneumoniae ST2570 from human blood in China.

Figure 5

Genetic surroundings of mcr-9.1 gene located on chromosomes in Enterobacter. The linearized coding sequences of MCR-9.1 region of four isolates were compared using clinker with identity threshold 90%. The MCR-9.1 (pink) surrounded by mobile genetic elements (turquoise) and other coding sequences (gray) in isolates 50607 and 59720 formed a region corresponding to the mcr-9 region observed in IncHI2/ST1 plasmids. On the other hand, 57166 and 56674 isolates contained downstream sequence similar to the previously described mcr-9 environment in E. asburiae (AP022628). White numbers in links correspond to the similarity of coding sequences.