Molecular detection of colistin resistance genes (mcr-1 to mcr-5) in human vaginal swabs

Objective Colistin resistance has emerged worldwide and has been threatening the efficacy of one of the last-resort antimicrobials used for treatment of multidrug resistant Gram-negative bacteria. While five colistin resistance genes (mcr-1, mcr-2, mcr-3, mcr-4 and mcr-5) have been described, few data are available on the prevalence of mcr-genes other than mcr-1 in human samples. Results In this study, the presence of five currently described colistin resistance genes (mcr 1–5) in vaginal swabs of women undergoing infertility evaluation was reported. Most samples were found to be positive for the mcr-4 (12.7%), followed by two for the mcr-2 (1.5%), two for the mcr-3 (1.5%), one for the mcr-1 (0.7%), and one for the mcr-5 (0.7%). Phylogenetic comparison demonstrated identical (mcr-1, mcr-2, mcr-3, mcr-5) or similar (mcr-4) nucleotide sequences of human samples and those of animal origins from the same city, suggesting the potential transmission of mcr genes from animals to humans. This is the first detection of mcr-2, mcr-4 and mcr-5 genes in human samples, and warrants further research to determine the spread of the mcr genes and elucidate the full epidemiology of colistin resistance genes in humans.


Introduction
Antimicrobial resistance is one of the most serious global threats to human health, especially the multiple drug resistant-pathogens of ESKAPE group (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa and Enterobacter spp.). The reintroduction of the older and less user-friendly antibiotics such as colistin is an option for treatment of the infections with bacteria of ESKAPE group. However, the efficiency of colistin treatment is compromised by the presence of an increasing number of mobile colistin resistance (mcr) genes. Recent findings indicate a low prevalence of mcr-1 in Enterobacteriaceae from inpatients and healthy volunteers (≤ 1%) [1]. Up to the preparation of this manuscript, there are five colistin resistance genes described (mcr-1, mcr-2, mcr-3, mcr-4 and mcr-5) [2][3][4][5][6]. However, few data are available on the prevalence of mcr-genes other than mcr-1 in human samples [1,2,[7][8][9]. In this study, PCRs were used to determine the presence of five mcr genes in human vaginal swabs, and phylogenetic comparison was performed the nucleotide similarity of the mcr genes from human and animals in the same city.

Materials and methods
This study received permission from the patients and was approved by the Institutional Review Board of Subei People's Hospital. In 2016, vaginal swabs were collected from 134 women attending Subei People's Hospital of Yangzhou in China for first or second infertility evaluation. Previous study demonstrated that none of vaginal swabs were positive for Neisseria gonorrhoeae and Treponema pallidum, but 18.8% of these swabs were positive for Chlamydia trachomatis and 17.3% of the swabs were positive for Mycoplasma species by PCRs [10]. All swabs

BMC Research Notes
*Correspondence: wangche@auburn.edu 6 College of Veterinary Medicine, Auburn University, Auburn, AL, USA Full list of author information is available at the end of the article were positive for tetracycline resistance gene tet(M) which is the most effective antibiotic for bacterial sexually transmitted infections [10].
Collection and DNA extraction of the human vaginal swabs were performed as described before [10]. Nucleotides of mcr-1, mcr-2 and mcr-3 genes in the samples were amplified with primers described before [11]. Meanwhile, using the Clustal Multiple Alignment Algorithm, we developed and validated a mcr-4-PCR (forward primer: 5′-AAT TGT CGT GGG AAA AGC CGC-3′; reverse primer: 5′-CTG CTG ACT GGG CTA TTA CCG TCA T-3′; amplicon size 1062 bp), and a mcr-5-PCR (forward primer: 5′-GTG AAA CAG GTG ATC GTG ACT TAC CG-3′; reverse primer: 5′-CGT GCT TTA CAC CGA TCA TGT GCT -3′; amplicon size 271 bp) in this study. The specificity of the primers for mcr-4 and mcr-5 PCRs were verified by BLASTN and DNA sequencing of the obtained PCR products. The sensitivity of the mcr-4-PCR and mcr-5-PCR was determined by amplifying dilutions of synthesized plasmids containing portions of the target mcr-4 and mcr-5 which were linearized with Sac I (Takara Biothechnology, Dalian, China). The quantitative standards were quantified using the PicoGreen DNA fluorescence assay (Molecular Probes, Eugene, OR, USA) for preparation of standards (10 4 , 10 3 , 10 2 , 10 1 , and 10 0 copies/reaction). The PCR products were confirmed by gel electrophoresis, and DNA sequencing with both PCR primers (BGI, Beijing, China) after purification with QIAquick Gel Extraction Kit (Qiagen, Valencia, CA, USA).
In this study, every 24th samples tested consisted of diethylpyrocarbonate-treated ddH 2 O, serving as a negative extraction control to confirm the absence of contamination between samples during DNA extraction and carry-over contamination. Additionally, control swabs from pipettes, experimental benches and centrifuges were frequently processed for five mcr-PCRs to verify that no false amplification occurred during this study resulting from carry-over contamination.
All reported human mcr sequences, representative mcr sequences from animals, and the mcr sequences from animals at Yangzhou were aligned with the obtained mcr sequences in this study. Based on these alignments, phylogenetic trees were constructed by the neighbor-joining method using the Kimura 2-parameter model with MEGA 6.0. The Bootstrap values were calculated using 500 replicates. The BLASTN was performed to determine new mcr variants by comparing the mcr sequences from this study and those available in GenBank.

Results and discussion
The mcr PCRs described before (mcr-1, mcr-2 and mcr-3) and established in this study (mcr-4 and mcr-5) were confirmed to specifically amplify the intended mcr gene, but not other mcr genes. The detection limit for mcr-4 PCR and mcr-5 PCR were determined to be 10 copies per 20 µL reaction.
Of 134 human swab samples, 22 of the vaginal swab (16.4%) were positive for at least one mcr gene. Most samples were found to be positive for the mcr-4 gene (12.7%), followed by two for the mcr-2 gene (1.5%), two for the mcr-3 gene (1.5%), one for the mcr-1 gene (0.7%), and one for the mcr-5 gene (0.7%) ( Table 1). The single swab sample positive for mcr-5 was also positive for mcr-4. This is a much higher prevalence of mcr genes than in every other human sample taken to date, but a similar prevalence for the mcr-1 gene previously reported in bacterial isolates from humans in China (Table 1) [1,2,12].
Furthermore, the mcr-4 gene was also reported in Salmonella strains isolated from human of Italy [13]. While direct PCR testing is ideal for the rapid estimation of risk and risk analysis, it does not readily enable investigations of movement of resistance between bacteria of the same or different species. As we did not base the study on the isolation of the bacteria, we are not sure which bacteria are carrying these genes, or on what mobile genetic elements these genes are carried. It is striking that the mcr-4 gene was so prevalent and this urges for more studies into the importance of this gene in colistin resistance.
In this study, we identified two new mcr-2 variants (MG520400, MG520401) and two new mcr-4 variants (MG520403, MG520404), while nucleotide sequences of mcr-1 (MG520399), mcr-3 (MG520402) and mcr-5 (MG520405) were identical to the original description of those genes (mcr-1: NG_054417; mcr-3: NG_055505; mcr-5: KY807920) (Fig. 1) [2,4,6]. Phylogenetic analysis demonstrates the identical (mcr-1, mcr-2, mcr -3, mcr-5) or similar (mcr-4) nucleotide sequences of human samples from Yangzhou in this study and those of animal origins in the same city (Fig. 1). This suggests the potential transmission of mcr-positive bacteria and/or mcr genes from animals to human beings. Therefore, it is more important for the government and companies to keep the food-productive animals from contaminating with mcr genes to ensure the food safety for human beings. In conclusion, we found all the current described colistin resistance genes in vaginal swabs with a surprisingly high prevalence of mcr-4. This is the first detection of mcr-2, mcr-4 and mcr-5 genes in human samples. Further studies on other samples and including cultivation of the mcr-carrying bacteria should be performed to determine the exact spread of these genes in bacteria from humans and elucidate the full epidemiology of colistin resistance genes in humans.

Limitations
The main limitation of this study is that the detection of mcr genes in vaginal swabs was solely based on quantitative PCRs, DNA sequencing and phylogenetic comparison. The background information such as prior treatment history of the patients was not available to this investigation. In future investigations, it would be useful to isolate mcr positive bacteria, and determine the species of the mcr-carrying bacteria, and whether the mcr genes are carried by plasmids or on the chromosome.

Authors' contributions
WC and ZJ participated in the design of the study, supervision of data collection, data analysis, and draft of the manuscript. CL, JW and BP contributed in the study operation and data analysis. JW, HK, QH, ZX and GW contributed in sample collection. All authors read and approved the final manuscript.  Fig. 1 Phylogenetic tree of sequences in five mcr genes. The nucleotide sequences of colistin resistance genes (mcr-1, 342 bp; mcr-2, 282 bp; mcr-3, 267 bp; mcr-4, 1062 bp; mcr-5, 197 bp) in human beings obtained this study (in bold font) were compared with those of representative mcr variants obtained from NCBI. The sequences from animals in Yangzhou, the same city for human samples, are shown with red font. The evolutionary history was inferred using the Neighbor-Joining method. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree