UC Berkeley

Drug resistant infections with Gram negative bacteria have become increasingly common. Some of these Gram negative bacteria are pan resistant. To better understand the spread of drug resistant infections we used E. coli as a model system. By using this system we investigated both the spread of drug resistance genes and bacteria that are drug resistant.

Drug resistance genes and their mobile genetic elements are frequently identified from environmental saprophytic organisms. It is widely accepted that the use of antibiotics in agriculture selects for drug resistant microorganisms, which are then spread from the farm environment to humans through the consumption of contaminated food products. We wished to identify novel drug resistance genes from microbial communities on retail food products. To do this, we created metagenomic plasmid libraries from microbiota isolated from retail spinach samples. From these libraries we identified five unique plasmids that increased resistance to antimicrobial drugs. These plasmids were identified in E. coli that grew on plates that contained ampicillin (pAMP), aztreonam (pAZT), ciprofloxacin (pCIP), trimethoprim (pTRM), and trimethoprim-sulfamethoxazole (pSXT). We identified open reading frames with similarity to known classes of drug resistance genes in the DNA inserts of all five plasmids. These drug resistance genes conferred resistance to fluoroquinolones, cephalosporins, and trimethoprim, which are classes of antimicrobial drugs frequently used to treat human Gram negative bacterial infections. These results show that novel drug resistance genes are found in microbiota on retail produce items. Food saprophytes may serve as an important reservoir for new drug-resistance determinants in human pathogens.

The clinical management of infections caused by E. coli, including meningitis, is greatly complicated when the organism becomes resistant to broad-spectrum antibiotics. We sought to characterize the antimicrobial susceptibility, multilocus sequence type (MLST), and presence of known drug resistance genes of E. coli that caused meningitis between 1996 and 2011 in Salvador, Brazil. We then compared these findings to E. coli isolates from community acquired urinary tract infections (UTI) that occurred during the same time period and in the same city. We found 19% of E. coli from cases of meningitis and less than 1% of isolates from UTI to be resistant to third-generation cephalosporins. The sequence types of E. coli from cases of meningitis included ST 131, ST 69, ST 405, and ST 62, which were also found among isolates from UTI. These sequence types of E. coli have previously been isolated from produce items and food animals. Additionally, among the E. coli isolates that were resistant to third-generation cephalosporins, we found genes that encode the extended-spectrum beta-lactamases CTX-M-2, CTX-M-14, and CTX-M-15. These observations demonstrate that, compared to E. coli isolated from cases of community acquired UTI, those isolated from cases of meningitis are more resistant to third-generation cephalosporins, even though the same sequence types are shared between the two forms of extraintestinal infections.

The results of our investigation of retail food products indicate that drug resistant genes are frequently found in these products and that these food items may be a reservoir for drug resistance genes found in human pathogens. We also found that many of the sequence types of E. coli that cause both UTI and meningitis have previously been isolated from retail food products. These findings suggest that bacteria that are considered to be normal flora of food products may play an important role of the spread of drug resistant genes and drug resistant bacteria.