Information on all of the strains used in this study is listed in Table 1. All strains were cultured in Luria-Bertani (LB) broth (5 g/L yeast extract, 10 g/L sodium chloride, and 10 g/L tryptone) at 37°C in a shaker at 200 rpm. All primers involved in the construction of plasmid, PCR and RAA assay are listed in Table 2.

Six hundred and seventy-two stool samples were collected from inpatients of Capital Institute of Pediatrics, Beijing, China. After dilution, stool samples with different dilution gradients were plated on LB plates with 2 mg/L of colistin sulfate salt and incubated at 35–37°C for 24 h. The screened single colonies were identified using the VITEK^® 2 compact system (bioMérieux, Nürtingen, Germany). Standard PCR and RAA assay were used simultaneously to detect whether the strains harbored the mcr-1 gene. Meanwhile, a sample of approximately 200 mg was subjected to DNA extraction using a kit (Vazyme Biotech Co., Ltd., Nanjing, China) for further use.

The sequence of the mcr-1 gene was downloaded from the National Center for Biotechnology Information (NCBI) GenBank database (NCBI Reference Sequence: NG_050417.1). The primers and probes were manually designed under the principles of RAA primer and probe. Briefly, the primer size was between 30 and 35 bp, the probe size was between 46 and 52 bp and the final product size was between 100 and 200 bp. The specificity of primers and probes was confirmed by NCBI primer-specific BLAST analysis and the hairpins and primer dimers were analyzed by Primer Primier 5. As an internal positive control, the primers and the probe of the 16S rRNA gene were designed in its conserved region. Related primers and probes involved in this study were synthesized by Sangon Biotech (Shanghai, China) and purified by high-performance liquid chromatography.

The full-length mcr-1 gene was amplified by PCR and cloned into vector pUC57 (Tiangen Biotech Co., Ltd., Beijing, China) by TA cloning, and the recombinant plasmid was called pUC57-mcr-1. The analytical sensitivity of the RAA assay was determined using 10-fold serial dilutions of the recombinant plasmid pUC57-mcr-1 ranging from 10⁷ to 10⁰ copies/μL. The analytical specificity of the RAA assay was evaluated by amplifying the mcr-1 and 16S rRNA genes from 14 different strains, respectively E. coli^mcr–1, Klebsiella pneumoniae 2146, E. coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, Shigella sonnei, Salmonella enteritidis, Acinetobacter baumannii, K. oxytoca, Enterobacter aerogenes, Proteus mirabilis, Enterobacter cloacae, Serratia marcescens, Campylobacter jejuni, and Citrobacter freundii. The E. coli^mcr–1 was used as a positive control and sterile water was used as a negative control.

A commercial RAA kit (Jiangsu Qitian Bio-Tech Co., Ltd., China) was used for the RAA assays. The RAA assays were performed as described previously (Feng et al., 2021). Briefly, a 50 μL reaction mixture was prepared first, which was made of reaction buffer (25 μL), DNase-free water (15.7 μL), 10 μM primer F (2.1 μL), 10 μM primer R (2.1 μL), DNA template (2 μL), 280 mM magnesium acetate (2.5 μL), and 10 μM probe (0.6 μL). Then, the reaction mixture was added to a tube with the lyophilized form of RAA enzyme mix and the tube was mixed briefly and incubated for 4 min in a B6100 Oscillation mixer (QT-RAA-B6100; Jiangsu Qitian Bio-Tech Co., Ltd., China). Finally, a fluorescence detector (QT-RAA-1620; Jiangsu Qitian Bio-Tech Co., Ltd.) was used to measure the fluorescence for 20 min at 39°C.

To detect mcr-1 gene, PCR was performed in a 20 μL reaction mix containing the following: 10 μL of PCR Master Mix reagent (Tiangen Biotech Co., Ltd., Beijing, China), 1 μL of 10 μM mcr-1-F primer (5′-CGTTCAGCAGTCATTATGCCAGTTTCTTTCGCGTGC-3′) and mcr-1-R primer (5′-CTTACGCATATCAGGCTTGGT TGCTTGTACCGC-3′), 1 μL of DNA template and 7 μL of double-distilled water. The PCR cycling conditions were 95°C for 3 min, followed by 35 cycles at 95°C for 30 s, 58°C for 30 s, and 72°C for 1 min. The final extension step was 72°C for 15 min. The PCR products were sent to Sangon Biotech for sequencing.

All mcr-1-positive E. coli isolates were identified using the VITEK^® 2 compact system and screened in accordance with the protocols presented on the MLST website.¹ Eight housekeeping genes, namely, dinB (DNA polymerase), icdA (isocitrate dehydrogenase), pabB (p-aminobenzoate synthase), polB (polymerase PolII), putP (proline permease), trpA (tryptophan synthase subunit A), trpB (tryptophan synthase subunit B), and uidA (beta-glucuronidase), were detected.

To analyze the characteristics of strains harboring mcr-1 genes, MICs were mainly determined using a VITEK^® 2 system (bioMérieux, Nürtingen, Germany). Escherichia coli ATCC25922 was used for quality control. The 2020 Clinical Laboratory Standards Institute’s threshold was used to as reference. In addition, the MICs of colistin and polymyxin B were determined by the twofold serial dilution method, as previously described (Fan et al., 2021). Briefly, the strains harboring mcr-1 genes were grown in LB broth at 37°C until the optical density at 600 nm (OD600) reached 1.0. Next, 100 μL of the bacterial suspension (5 × 10⁵ CFU/mL) and different concentrations of diluted antibiotics were added to each well of a 96-well plate (Corning). The 96-well plate was incubated without agitation at 37°C for 24 h. The minimum antibiotic concentration to visibly inhibit bacterial growth was recorded as the MIC.

To detect the HGT of mcr-1, colistin-resistant isolates served as donor strains, with sodium azide-resistant E. coli J53 as recipient. Briefly, 500 μL of each donor strain and 500 μL of the recipient in LB broth were mixed, centrifuged at 8,000 rpm, and the bacterial precipitates were resuspended on 50 μL of LB broth. The resuspended bacteria were then added to round filter papers pre-placed on nutritional agar plates and cultured overnight at 37°C. Transconjugant bacteria were selected on LB plates containing sodium azide (100 mg/mL) and colistin (4 μg/mL). Meanwhile, the same number of bacteria were plated on LB plates only containing sodium azide (100 mg/mL). The number of bacterial cells was determined by serial dilution and plating. The transfer efficiency was calculated by dividing the number of successfully transformed strains by the total number of receptor strains.

All trials were conducted three times. The p-values and kappa values of the RAA and standard PCR assays were calculated. The statistical analysis was conducted with SPSS 21.0 (IBM, Armonk, NY, United States).

Since the first mcr-1 sequence was released in 2015 (NCBI Reference Sequence: NG_050417.1), 31 variants of it have been identified. The genome sequences of all mcr-1 genes are almost identical. The primers and probes for this study were manually designed on the specific and conserved region (Figure 1 and Table 2).

Fourteen different strains were used as templates to amplify the mcr-1 gene (Table 1). Only from E. coli^mcr–1 did we succeed in detecting amplification signals, while the others were all negative (Figure 2A). As an internal control, the amplification signals of the 16S rRNA were detected from all bacteria (Figure 2B). Therefore, the RAA assay for the detection of mcr-1 was 100% specific.

Furthermore, a 10-fold gradient dilution series of recombinant plasmids pUC57-mcr-1 was used to detect the sensitivity of the RAA assay. As the copies/μL increased from 1 × 10¹ to 1 × 10⁷, the fluorescence signal increased (Figure 3). The detection limit of the RAA assay was 20 copies per reaction.

The RAA assay and the standard PCR assay were simultaneously used to detect mcr-1 in 672 samples. Among these, the presence of the mcr-1 gene was found in five samples. The two methods gave the same experimental results (Figure 4 and Table 3).

By screening for colistin resistance, five colistin-resistant E. coli were isolated from five clinical samples. All of these E. coli had the mcr-1 gene, as determined by PCR and sequencing (Table 3). MLST analysis showed that the five E. coli belonged to different ST types: E. coli/mcr-1 1 (ST21), E. coli/mcr-1 2 (ST740), E. coli/mcr-1 3 (ST48), E. coli/mcr-1 5 (ST809), and E. coli/mcr-1 4 (a new ST type).

Following testing with the VITEK^® 2 compact system, most of the mcr-1-positive E. coli were shown to be resistant to colistin, ciprofloxacin, levofloxacin, doxycycline, and trimethoprim, and sensitive to imipenem, amikacin, and meropenem (Table 4). E. coli/mcr-1 2 was resistant to ticarcillin/clavulanic acid, piperacillin/tazobactam, and cefoperazone, while the other bacteria were sensitive to these antibiotics. The MICs of colistin and polymyxin B were simultaneously determined by the twofold serial dilution method (Table 4).

To investigate the transferability of mcr-1, HGT assays were performed using colistin-resistant isolates as donors and sodium azide-resistant E. coli J53 as recipient. mcr-1 from all five colistin-resistant strains was successfully transferred to E. coli J53 recipient strain with some efficiency (Table 3). In contrast to the parental recipient strain, E. coli J53 transconjugants showed varying degrees of resistance to colistin and polymyxin B (Table 4). It was not difficult to find that other resistance did not transfer with the mcr-1 gene transfer (Table 4). Interestingly, the transfer efficiency of mcr-1 from E. coli/mcr-1 4 and E. coli/mcr-1 5 was much higher than that from other isolated E. coli strains, at up to 57–58%. The mechanisms related to this are under investigation.