Environmental Risk Assessment of Antibiotics: Investigations into Cyanobacteria Interspecies Sensitivities and Establishing Appropriate Protection Limits

Abstract

Antibiotics have been described as a ‘wonder drug’ that have transformed medicine since their discovery at the beginning of the 20th century and are used globally in safeguarding human and animal health. Environmental risk assessment (ERA) aims to ensure their environmental safety by setting protection limits that seek to prevent adverse effects upon populations and ecosystem function. In the case of antibiotics however, there is concern that ERA may not be fully protective of bacterial populations. This thesis examines the ERA of antibiotics and highlights that protection limits may in some cases be under-protective or over-protective for bacteria populations (including cyanobacteria), depending on the antibiotic mode of action and the species on which the protection limit is based. The first section of the thesis contains a systematic review including a meta-analysis of all publicly available aquatic ecotoxicity data. The results illustrate that generally bacteria are the most sensitive taxa to antibiotics compared with eukaryotes but that interspecies variability in sensitivity among bacteria can range by up to five orders of magnitude. This far exceeds the assessment factor of 10 used to account for such uncertainty in protection limits. It also shows that the costly testing on fish may not be required and in accordance with the principle of the 3Rs could be excluded from the ERA of antibiotics, as they are not likely to drive the protection limit. Further, it demonstrates that protection limits established for antimicrobial resistance (AMR) may not always be protective of environmental health and that both protection limits should be determined in ERA. Next, the thesis reports on the development and validation of a microplate assay for the rapid screening of chemical effects (here antibiotics) on cyanobacteria. The microplate assay is optimised to allow for the direct comparison of species sensitivity, ensuring consistent test conditions and thus limiting differences in antibiotic behaviour between assays. Reference toxicity testing with potassium dichromate demonstrates reproducibility over time and comparability with the standard shake flask test used in ERA. The microplate assay is then used to experimentally confirm the findings of the meta-analysis. In this work, interspecies sensitivity across eight cyanobacteria species was demonstrated to vary by up to 70 fold following exposure to β-lactam antibiotics but only by an order of magnitude for macrolides. Cyanobacteria were not sensitive to sulfonamides and thus are not likely to be suitable for the setting of protection limits for this antibiotic class. Finally in this thesis, species sensitivity distributions were created to examine how effective the protection limit currently derived in ERA is for antibiotics. For cephalosporins, there was a higher probability of under-protection whilst the protection limits were over-protective for classes of antibiotics with less interspecies differences in sensitivity, such as macrolides. Further, a probabilistic ecological risk assessment suggested that 60 to 100% of cyanobacteria species might be adversely affected at the higher measured environmental concentrations in the literature, while no significant risk was found at average concentrations. The findings from this thesis illustrate that protection limits, as currently determined in ERA for antibiotics, may not be suitable for the adequate protection of cyanobacteria populations and most likely other bacterial taxa. The thesis proposes several approaches for improving ERA, including incorporating greater bacterial diversity in ecotoxicity testing, the inclusion of functional and/or community testing and the use of probabilistic methods to derive protection limits.