Evaluation of microbial shifts caused by a silver nanomaterial: comparison of four test systems

Abstract Background Before chemicals, pesticides and biocides are registered and approved, their effects on soil microorganisms must be tested, specifically their impact on nitrogen transformation. Following a request from the European Food Safety Authority (EFSA), the Panel on Plant Protection Products and their Residues provided an opinion document evaluating the science behind the risk assessment of plant protection products in the context of soil-dwelling organisms. The EFSA document concludes that the most relevant community-based microbial test systems should cover the widest possible range of metabolic processes without compromising test sensitivity. The EFSA document refers to the MicroResp test system, stating that although it has not been used to study the effects of pesticides on soil microbial processes, its capability should be investigated in the future. In the scope of harmonization approaches, the recommendations in the EFSA document covering pesticides could also influence the risk assessment and regulation of other kinds of chemicals, including silver nanomaterials. We therefore used the silver nanomaterial NM-300K as a model substance to evaluate the sensitivity of three functional tests covering the activities of different microbial fractions: (1) the potential ammonium oxidation (PAO) test, which considers the first step in nitrification; (2) the MicroResp test, which determines respiratory activity by measuring CO2 evolution; and (3) a colorimetric test system for exoenzyme activity. We also surveyed bacterial 16S rRNA sequence diversity by next-generation sequencing (NGS). Results There was no major difference in the general sensitivity of the tests, each of which revealed significant effects at silver nanomaterial concentrations of at least 1.67 mg/kg. The PAO test was a robust and sensitive indicator of toxicity, and concentration–effect relationships were calculated for every time interval. The effects on respiration and exoenzyme activities were more variable. Among the three functional tests, the selected exoenzyme activities showed the weakest concentration–effect relationships, although silver concentrations were clearly related to two of the four activities we tested (glucosidase and arylsulfatase). We also observed a relationship between silver concentrations and respiration activity on glucose, cellobiose and alanine substrates. The bacterial orders identified by NGS differed in sensitivity to the silver nanomaterial. We found that the adverse impact on nitrifiers matched the inhibition of PAO activity. EC50 values calculated for each functional test did not identify a generally superior method. Conclusion We found that all four test approaches were similar in sensitivity towards the model silver nanomaterial, an ion-releasing substance. We observed advantages and limitations for each test, which must be considered when selecting tests for the registration or approval of substances. It is unclear whether the sensitivity of the tests would be comparable when testing substances that do not release ions. The regulatory assessment of metabolic profiles requires further consideration in terms of substrate selection. Finally, for tests performed in multiwell plates with small quantities of soil, the quality of the concentration–effect relationships must be studied in more detail.