High Prevalence and Diversity Characteristics of blaNDM, mcr, and blaESBLs Harboring Multidrug-Resistant Escherichia coli From Chicken, Pig, and Cattle in China

The objective of this study was to understand the diversity characteristics of ESBL-producing Escherichia coli (ESBL-EC) in chicken, pig, and cattle. A high prevalence of ESBL-EC (260/344) was observed in all food animals with prevalence rates of 78.6% (110/140) for chicken, 70.7% (58/82) for cattle, and 75.4% (92/122) for swine. However, the resistance rates presented significant differences in different animal origin ESBL-EC, where resistance to CTX, GEN, IMP, NEO, and OFL was the highest in chicken ESBL-EC, then in cattle, and the lowest in swine. Seriously, most ESBL-EC harbor multidrug resistance to antibiotics (MDR, ≥3 antibiotic categories), and the MDR rates of ESBL-EC were the highest in chicken (98.18%), followed by swine (93.48%), and the lowest in cow (58.62%), while the same trend also was observed in MDR of ≥5 antibiotic categories. This high prevalence and resistance can be partly interpreted by the high carriage rates of the β-lactamases CTX-M (n = 89), OXA (n = 59), SHV (n = 7), and TEM (n = 259). A significant difference of β-lactamase genes also presented in different animal species isolates, where the chicken origin ESBL-EC possessed higher carriage rates of almost all genes tested than cattle and swine. Notably, eight chicken origin ESBL-EC carried transferable plasmid-mediated blaNDM-1 or blaNDM-5, especially, of which four ESBL-EC also contained the colistin resistance gene mcr-1, as confirmed by genomic analysis. More interestingly, two deletion events with a 500-bp deletion in ΔISAba125 and a 180-bp deletion in dsbC were observed in three blaNDM-5 IncX3 plasmids, which, as far as we know, is the first discovery. This showed the instability and horizontal transfer of blaNDM genetic context, suggesting that blaNDM is evolving to “pack light” to facilitate rapid and stable horizontal transfer. Sequence types (STs) and PFGE showed diversity patterns. The most prevalent STs were ST48 (n = 5), ST189 (n = 5), ST206 (n = 4), ST6396 (n = 3), ST10 (n = 3), and ST155 (n = 3), where ST48 ESBL-EC originated from three food animal species. The STs of all blaNDM-positive ESBL-EC were attributed to three STs, namely, ST6396 (n = 2), ST206 (n = 2), and ST189 (n = 4), where ST189 was also the unique type for four mcr-1-carrying ESBL-EC. In conclusion, we suggest that the three animal species ESBL-EC show similar high prevalence, diversity in isolate lineages, and significant discrepancies in antibiotic resistance and resistance genes. This suggests that monitoring and anti-infection of different food animal origin ESBL-EC need different designs, which deserves more attention and further surveillance.