Toxicity Ranking and Toxic Mode of Action Evaluation of Commonly Used Agricultural Adjuvants on the Basis of Bacterial Gene Expression Profiles

Toxicity and toxic mode of action of adjuvants

Adjuvants comprise of three major groups: surfactants, solvents and synergists and are often referred to as “inert ingredients”. A consumer survey performed by US EPA learned that many consumers are mislead by the term “inert ingredient”, believing it to mean harmless [17]. This certainly is not the case and in fact they can be toxic to humans, may have biological activity of its own [18], [19]. Nevertheless, up till recently adjuvants were not taken into account for the risk evaluation of pesticides. The upcoming new EU regulation concerning the placing of plant protection products on the market (EC1107/2009) includes for the first time the demand for information on the possible negative effects of not only the active ingredients but also the used adjuvants. This new regulation requires basic toxicological information that is used to decide on the use, ban or preferential use of available adjuvants [8].

This study provides information on the toxicity and toxic mode of action of the selected compounds. The ranking of the adjuvants based on their toxicity (growth inhibition) showed that the surfactants are far more toxic than the selected solvents in the assay. Ethoxylated tallow alkyl amine is the most toxic compound tested. High toxicity after exposure to ethoxylated tallow alkyl amine was already reported for several species e.g. tadpoles and green algae [20]–[23]. Within the group of surfactants toxicity varies by three orders of a magnitude, with ethoxylated fatty acid (isotridecanol) and trisiloxaan ethoxylate tenside as the least toxic compounds. The toxicity results illustrate the importance of reporting toxicity in different ways (here IC50 and NOEC-LOEC) to characterise the toxicity of a compound. If only IC50 values are determined EO NP would be regarded as a non-toxic compound while growth inhibition already occurs at low concentrations. For several compounds IC50 and LOEC values could not be calculated due to limited water solubility. We preferred not to use other solvents than water since in realistic conditions (sprays and tank-mixes) water is used as a diluent or solvent.

Organosilicone surfactants, a fairly new class of non-ionic wetting agents, do not act like classical surfactants through the membranes but they provide a faster penetration of the pesticide in the plant through a specific mode of action i.e. by facilitating stomatal infiltration of solutions [24]. They are considered as promising compounds since improved spreading of the active ingredient can lead to a reduction of the latter in formulations. Two organosilicone surfactants were tested in this study, i.e. trisiloxane ethoxylate tenside (tri EO) and trisiloxane ethoxy-propoxylate tenside (tri EO-PO). Both compounds increase the uptake and efficacy of pesticides in a similar way [25], though this study demonstrates that they differ by one order of magnitude at the toxicity level. Stark and Walthall (2003) investigated the acute toxicity of several agricultural adjuvants, including organosilicone surfactants, with Daphnia pulex. They found different LC50 values for different organosilicone surfactants: Silwet L-77® 3 mg/L and Kinetic® 111 mg/L. The results from our study at the gene expression level confirm that the main mode of action of the tested organosilicone surfactants is not through membrane damage (MicF, OsmY and ClpB) since these genes are not significantly induced. The toxic mode of action of organosilicone surfactants is mainly oxidative damage through part of the SoxRS pathway (Zwf and Soi28). Both compounds are not grouped together with paraquat, the model compound for SoxRS mediated oxidative damage, in the PCA analysis the reason for this is that not the whole SoxRS pathway is induced as can be observed from the FI(S) dataset.

In a mode of action and QSAR (quantitative structure activity relationships) context, non-ionic surfactants are described as compounds that provoke toxicity through non-specific mechanisms, the toxic potency of these compounds correlates well with their hydrophobicity. Such a mode of action is defined as narcosis, one of the four mode of action categories (narcotics, non-polar narcotics, reactive chemicals and specifically acting reactive chemicals) in the Verhaar classification scheme (Verhaar et al., 2000). Exposure to narcotics typically results in disruption of the biological membrane integrity [26], [27]. Several of the non-ionic surfactants included in this study induced membrane damage (MicF and OsmY) and general cell lesions (UspA and ClpB). Narcosis was already described for several of the adjuvants tested, dichlormethane, ethoxylated nonylphenol, ethoxylated alcohol [20], [28]. The results in our study confirmed these results for EA and EO NP and also revealed membrane damage after exposure to POEA, EO PE and gamma-butyrolactone. In our study, no membrane damage is found after exposure to dichloromethane, but the test concentrations were low due to the limited solubility.

Membrane damage and general cell lesions were not the only pathways affected after exposure to these compounds, DNA damage and oxidative stress are induced as well. The induced DNA damage markers are part of the SOS response, a well described repair mechanism in bacteria [29]. Valuable markers for the SOS response are RecA, UmuDC and SfiA, they can be considered as indicators for potential genotoxic compounds like the model compound methylmethane sulphonate (MMS) [12]–[14]. SfiA, is also part of the validated SOS chromotest [14]. The observed DNA damage (both FIS and NOEC) demonstrated that in the reporter assay the SOS response pathway is induced as described in literature, mild SOS response only RecA induction and severe SOS response both RecA and UmuDC inductions [29]. Previous studies already pointed out that the induction of SfiA could be related to oxidative DNA damage [12]. This is also the case in our study since together with the high induction of the oxidative damage markers the induction of SfiA was observed.

Several of the surfactants (POEA, AE, tri-EO, EO FA and EO NP) and one solvent (gamma-butyrolactone) that were tested showed significant inductions for the SOS response pathway. The FIS showed for several compounds inductions of up to 50% of the MMS signal for RecA, indicating that POEA, AE, EO FA and tri EO are half as potent as MMS to induce RecA. These results were observed at mg/L range for POEA, EA and tri EO and in g/L range for EO FA. Environmental concentrations of the selected compounds are not routinely monitored so little data are available, Belanger and colleagues found concentrations of AE (sum of all) in European effluents of 6.8 µg/L, far below the LOEC at the gene expression level, nevertheless further research on the potential genotoxic effects of these compounds is needed as there are no threshold levels for genotoxic compounds [30].

The most recent US-EPA classification of adjuvants lists gamma butyrolactone as harmless and the usage in pesticides is unlimited, though the report lists genotoxic effects at high concentrations [31]. The concentrations that were tested in this study are very high and unlikely to occur in the environment or food chain. Information on possible genotoxic potential of the other tested compounds is not found in literature.

Ethoxylated fatty alcohol is considered as an alternative for the endocrine disruptor ethoxylated nonylphenol which is banned in Europe. Nevertheless, toxicity results in this study show that NOEC-LOEC values are comparable [20]. At the gene expression level, both compounds induce several stress genes, the LOEC at the gene expression level is even lower for ethoxylated fatty alcohol than for ethoxylated nonylphenol. Based on the results from this study other surfactants seem more appropriate to replace ethoxylated nonylphenol i.e. organosilicone surfactants or ethoxylated tristyryl phenol. The results show a first ranking based on toxicity and toxic mode of action of adjuvants, but additional information concerning other relevant endpoints like endocrine disruption potential is needed.