Neonicotinoids are the most commonly used insecticides globally of the past three decades (Borsuah et al., 2020), but are harmful to non-target organisms like pollinators (Goulson, 2013; Pisa et al., 2014) and soil invertebrates (de Lima e Silva et al., 2017, 2020, 2021). As a consequence, ecosystem services crucial for sustainable agriculture, such as nutrient cycling, pest control and pollination, are under threat by the use of neonicotinoid insecticides (EASAC, 2015; FAO, 2020; Gunstone et al., 2021).
Current environmental risk assessment (ERA) and policy regarding pesticides is based on phenotypic toxicity tests that measure effects on the survival and reproduction of model organisms after exposure to individual pesticides. Extrapolation of these findings to ecotoxicological effects in the field is difficult as most agricultural soils are polluted by pesticide mixtures (Pelosi et al., 2021; Silva et al., 2019), and the synergistic interactions between pesticides within mixtures is a major knowledge gap (Gunstone et al., 2021). Furthermore, the predicted effect concentrations derived from these phenotypic tests can only be used in ERA after measuring the exposure concentration of the pollutants in soil, a laborious and costly procedure. In contrast, gene expression responses can be used to determine the type of pollution even under varying mixture composition (Fontanetti, Carmem et al., 2011; Shi et al., 2017). Determining the effects of the near infinite number of possible soil pollution mixtures on the gene expression of model organisms is unfeasible. Therefore, reliable genetic responses, i.e. biomarkers, have to be identified that remain indicative for a group of soil pollutants even under synergistic interaction with other pollutants. Gene expression biomarkers, in turn, can be used in biomonitoring; a cost-effective tool to screen for samples that, in case of detecting a potential risk, may be subjected to subsequent chemical analysis to identify the chemical(s) of concern. In this way, gene-expression assays may provide ERA with more accurate metrics of adverse effects by pesticides than traditional toxicity tests.
The selection of candidate gene expression patterns requires an understanding of the molecular mediators behind pesticide toxicity in a relevant non-target model organism. Most studies on the molecular mechanisms that mediate neonicotinoid toxicity in invertebrates have been carried out in honey bees. However, the honey bee is not an ideal representative for non-target soil invertebrates because it does not live in the soil, its genome is limited in its detoxification capacity (Claudianos et al., 2006), and it has an unusual life history due to its social lifestyle (Gradish et al., 2019). Folsomia candida is a more suitable representative for non-target soil invertebrates because (1) it belongs to the springtails (Collembola), which is one of the most prevalent non-target invertebrate groups (Rusek, 1998), and a key component of the soil food web by promoting nutrient cycling (FAO, ITPS, GSBI, 2020); (2) F. candida is well established as a soil ecotoxicological model species since the 1960s (van Gestel, 2012); (3) its genome has been sequenced and annotated facilitating the development of molecular tools for studying its genomic responses to pollution (Faddeeva-Vakhrusheva et al., 2017), and (4) F. candida is representative for the sensitivity to neonicotinoids of other springtail species (de Lima e Silva et al., 2021). Together, these aspects make F. candida an ideal candidate for the development of biomarker assays for the monitoring of pesticide exposure in soil.
For the successful applications of neonicotinoid biomonitoring, gene-expression patterns have to be identified that are indicative for the exposure to a variety of neonicotinoids and remain to do so even under synergistic interaction with other pollutants. Neonicotinoids are commonly subdivided in two groups, depending on the inclusion of either nitro- or cyano-moieties into their chemical structure (Buszewski et al., 2019). Although both groups share the same mode-of-action, the nitro-substituted neonicotinoids are more toxic than the cyano-substituted ones to the fecundity and survival of various springtail species (de Lima e Silva et al., 2017, 2020; 2021). In the honey bee, the differential toxicity of the two groups of neonicotinoids has been attributed to an increased detoxification rate of the cyano-substituted ones by CYP enzymes (Iwasa et al., 2004; Manjon et al., 2018). Moreover, CYP inhibition has also been proposed to trigger synergistic interactions between neonicotinoids and other pesticides such as triazole fungicides (Feyereisen, 2018; Glavan & Bozic, 2013; Raimets et al., 2017; Sgolastra et al., 2017). Finally, various studies on the genomic response of F. candida to various pollutants have identified CYP genes as biomarkers for a variety of chemicals (Chen et al., 2014; de Boer et al., 2009; Nota et al., 2009; Qiao et al., 2015; Roelofs et al., 2012). Based on these findings, CYPs have emerged as promising biomarkers for the toxicity of neonicotinoid exposure. Yet, it remains to be confirmed if gene expression patterns of CYP genes provide a reliable indication for the toxicity of both cyano- and nitro-substituted neonicotinoids, as well as for synergistic interaction with other pesticides. This also needs to be confirmed still for other biomarkers identified for neonicotinoid exposure in the honey bee (Christen et al., 2016; Fent et al., 2020; Manjon et al., 2018). Given the central role of CYPs in mediating differential effects of the two major classes of the neonicotinoid family and its role in mediating synergy, we propose inhibition of CYPs could serve as “stress-test” to assess biomarker robustness. For this we applied piperonyl butoxide (PBO), which is a CYP inhibitor that forms a metabolite-inhibitory complex with CYPs and thereby prevents the binding of other substrates (Hodgson & Levi, 1999). By choosing PBO over toxicants, we can ensure that observed effects on biomarker gene-expression is the result CYP inhibition, rather than, other synergistic interactions.
The range of effects soil pollution has on organisms is diverse and, hence, the integration of multiple biomarkers into a panel for biomonitoring and ERA is highly recommended (Lionetto et al., 2019). The aim of this study was to assess the suitability of candidate genes to construct a panel of biomarkers for the assessment of soil polluted with neonicotinoids. For this we considered three criteria: (1) the panel should indicate exposure of both nitro- and cyano-substituted neonicotinoids, (2) the response of the panel should relate in a concentration-dependent manner with the adverse fitness effect of neonicotinoid exposure on F. candida, and (3) the expression patterns of biomarkers in the panel should be reliable under synergistic interaction caused by CYP inhibition by PBO. To represent the two major classes of neonicotinoids we selected imidacloprid and thiacloprid, as representatives of nitro- and cyano-substituted neonicotinoids, respectively. First, we determined the effect of PBO on the fecundity of springtails and its potency-enhancing effects when combined with thiacloprid and imidacloprid. Then, we screened the expression of eight candidate biomarker genes at various PBO and neonicotinoid concentrations using RT-qPCR. These were derived from previous studies on the genomic response of F. candida to various pollution types which have identified gene expression patterns that may have potential to be applied as biomarkers (de Boer et al., 2009; Nota et al., 2009; Qiao et al., 2015; Roelofs et al., 2012).