Publikationsdatenbank

The Impact of Glyphosate on Escherichia coli and Bacterial Communities in vitro and in vivo (2021)

Art

Hochschulschrift

Autor

Bote, Katrin (WE 10)

Quelle

Berlin, 2021 — iv, 83 Seiten

Sprache

Englisch

Verweise

URL (Volltext): https://refubium.fu-berlin.de/handle/fub188/29424

Kontakt

Institut für Tier- und Umwelthygiene

Robert-von-Ostertag-Str. 7-13
14169 Berlin
+49 30 838 51845
tierhygiene@vetmed.fu-berlin.de

Abstract / Zusammenfassung

Glyphosate (N-(phosphonomethyl)glycine) is the most-used herbicide worldwide. Many studies have found residues in feed and food. Naturally, concerns about its safety and side effects on other organisms have been raised. With only insufficient and contradictory data about the susceptibility for the widely used herbicide glyphosate, our study is the first to systematically analyse a large amount of E. coli from livestock isolated at different points in time. According to standards for antimicrobial susceptibility testing, we determined minimum inhibitory concentrations (MICs) by means of broth microdilution for the active ingredient (AI) isopropylamine salt (IPA) and the glyphosate-containing formulation Roundup® LB Plus (RU), commonly used in Germany. In total, 238 E. coli isolates, mainly isolated from poultry, pigs and cattle were investigated. Samples isolated between 2014 and 2015 were compared to historic samples of the standard E. coli collection of reference (ECOR) from 1984. For further statistical analysis, samples were divided into extended-spectrum beta-lactamase (ESBL) and non-ESBL producing E. coli as well as into commensal and pathogenic isolates. Mean and mode for all isolates showed a higher level of tolerance for RU (40 mg/ml IPA) compared to GLY (10 mg/ml IPA). In general, the distribution was narrow, and a clearly resistant subpopulation was lacking. To identify less susceptible isolates, a 95% cut-off was calculated (20 mg/ml for GLY and 40 mg/ml for RU). Isolates above the cut-off were sequenced and their aroA gene, coding for the glyphosate target enzyme, compared. No previously known resistance mechanisms were found, however, most differences occurred close to positions described in the literature. Isolates from poultry showed significantly higher MICs in RU and GLY, both in nonparametric Mann-Whitney U tests and statistical models (multivariable variance analysis for GLY and multivariable proportional-odds regression model for RU). In addition, both pathogenic and non-ESBL isolates showed significantly higher MICs in the GLY group, verified by both statistical methods. Solely in the nonparametric test for GLY, historic isolates were less tolerant than recently sampled isolates. However, with only very few isolates from livestock preserved in the ECOR collection, the sample size is a limiting factor of this model. Hence, future studies should include more equivalent historic isolates. To determine whether the growth and survival of a pathogenic E. coli and a S. Typhimurium isolate in an in vitro ruminal experiment is influenced by 10 mg/l RU as a worst-case concentration, a ’Rumen Simulation System’ (RUSITEC) was used. Fermenters were inoculated with 109 colony forming units (CFU) of each strain, leading to a starting concentration of 106 CFU/ml. Initially, the number of CFU of Salmonella doubled after 2 to 4h. Apart from this brief increase, the number of bacteria continuously declined in all fermenters, without being influenced by the RU application. E. coli was no longer detectable in quantitative tests from day 4 and in qualitative tests from day 5 onwards. S. Typhimurium remained detectable until the end of the experiment on day 7, although only a few CFU survived. MICs for RU did not change after the exposure, while antibiotic susceptibility did not vary significantly. In conclusion, the exposure to RU neither increased the abundance, nor promoted resistance. Considering the fermentation experiment focused only on two Enterobacteriaceae in an artificial environment, a more extensive in vivo experiment with pigs was conducted. Weaned piglets (naturally colonized with ESBL E. coli) were infected with 108 CFU of the same S. Typhimurium DT104 strain used in the RUSITEC experiment, at the age of five weeks. One week later, half of the animals per group (n=14/2) were sacrificed as an internal control. The other half was further exposed to nothing (CTRL), GLY or RU, in worst-case concentrations of 2.85 mg/kg bw/d, based on residue levels described in pig feed. The feeding experiment lasted for two weeks, during which faecal samples were checked twice weekly for Salmonella and weekly for ESBL E. coli. Finally, different organs were investigated and faeces and caecum contents were frozen and sent for 16S rRNA analysis via Illumina MiSeq. Neither the exposure to GLY nor to RU increased the shedding or accumulation in organs. As in the fermentation experiment, MICs of the isolates for RU or GLY did not change. The 16S rRNA analysis revealed some differences between the microbial compositions in the different study groups. In general, the RU group showed greater diversity in both faecal and caecum samples. For the GLY group, a tendency was only observed in faeces. Overall, more differences between the CTRL and the exposed groups were found in caecum samples than in faeces. In both caecum and faeces samples, Lactobacillaceae (genus Lactobacillus) increased in pigs from the RU and Enterobacteriaceae (genus Escherichia) increased in pigs from the GLY group. In contrast to previous reports, the number of Clostridia did not increase, but rather decreased in some samples. Future studies should focus on identifying reasons for inter- and intra-species susceptibility variation by taking a closer look at resistance mechanisms and target structures. Moreover, a possible link between antibiotic and glyphosate tolerance should be investigated. Effects of glyphosate on more vulnerable microbiota that are more sensitive for lack of aromatic amino acids (i.e. after birth or after weaning, infection, antibiotic treatment or immunosuppression) should be investigated.