Better understanding of the exposure of the general population to mixtures of pesticides at low concentrations and the potentially associated risk, is at the heart of TOXALIM's concerns
To assess more precisely the exposure of the general population to pesticides in mixtures and better characterize the hazards that could lead from this exposure, are at the center of concerns several research teams at TOXALIM, especially teams E2-MeX, E3-PEP, E6-TMR and E10-TCMX and the AXIOM-MetaToul technological platform.
With several multidisciplinary projects supported by the Midi-Pyrénées Regional Council, by the ANR (National Research Agency) or the ANSES (French Food Safety Authority), a dense collaborative network, several PhD grants addressing this issue and the development of effective and innovative tools, significant advances have been obtained during the past two years.
1. To define qualitatively and quantitatively the mixtures of substances ("cocktails") to which we are exposed is a research topic in its own right. One goal of the project PERICLES (Exposure to mixtures of active substances and possible combined effects on human cells) supported by the ANR and coordinated by the ANSES, in which two teams of TOXALIM contributed (E2 and E10) was to identify pesticide cocktails to which consumers are more likely to be exposed. This assessment is usually based on the intersection of food consumption data with those for food contamination.
- The originality of the Pericles program was to address the issue of cocktails starting from the actual exposure of the French population to understand the nature of cocktails on human cells in vitro. A statistical classification method was developed and applied to these exposures to determine the main cocktails and the main food vectors. Seven cocktails containing from 2 to 6 pesticides were identified and a battery of in vitro toxicology tests, assessing cell viability, genotoxicity, activation of xeno-hormone receptors and other biomarkers were performed on cells representative of different human organs (liver, kidney, intestine, brain, colon). The first results were used to focus the investigations on three cocktails with the most significant effects. The toxicological effects of different mixtures were compared with those of "dose-additivity" and "response-additivity" models to determine whether it is possible to predict the effect of a cocktail based on the knowledge of the toxicological effects of pesticides only (Crépet et al., 2013). The Effects of 7 selected cocktails in the Pericles study highlighted more or less important cytotoxic effects depending on the cocktails and the cell lines . A cocktail was genotoxic in a liver cell line (HepG2), with a greater effect than that resulting from the sum of the effects of each of the mixture components (Graillot et al., 2012a) effect. Other modes of action, such as transactivation of PXR receptor in liver cells showed "infra-additive" effects of the mixtures tested (Rouimi et al., 2012). The results show a wide variation in the sensitivity of cell models and tests. Thus, these cocktails produce additive effects but also supra or infra-additive effects, which makes the prediction of the response of cocktails from the results of the responses and doses of pesticides alone difficult.
2. To measure the concentration of contaminants in samples of blood or urine collected from the population is another way to estimate the exposure of the latter to chemicals. Although this method has limitations (apart from the case of persistent and bioaccumulative pollutants, substances detected are only representative of a recent exhibition, also residues in blood or urine are more often metabolites that are not always known), it has the merit of being conducted at the level of the individual.
- Following the establishment of a collaboration with INSERM, TOXALIM began to characterize pesticide exposure of individuals recruited through the cohort PELAGIE (Endocrine Disruptors Longitudinal Study on Abnormal Pregnancy, the infertility and Childhood). The constitution of this cohort of nearly 3,500 pregnant women of Brittany (France) was to assess the consequences of exposure to various contaminants, especially pesticides, in the course of pregnancy, the outcome and psychomotor development the child. From a limited number of urine from these pregnant women it was possible to determine the spectrum of contamination of these individuals samples. Initially a list of 47 pesticides was developed from data on agricultural practices related to residential places of the women at the time of urine collection, on the one hand, and recommendations for use of pesticides by Chambers of Agriculture and plant index published by ACTA, on the other hand. In a second step, the metabolites corresponding to the list of selected pesticides were found in the data generated from the analysis of urine samples (378 metabolites). The approach relies on the identification of specific signals to a pesticide or its transformation products (metabolites) based on its exact mass and confirmation of chemical structures by comparison with standards, where they exist, or biotransformation products isolated from separate animal experiments. The signal of a substance detected by mass spectrometry being proportional to its concentration, the overall method is termed semi-quantitative. It discriminates exposure depending on the area devoted to cereal crops near the residence of individuals groups. The data obtained indicate that the metabolites are mainly of two pesticides, azoxystrobin and fenpropimorph, which allow to discriminate these exposure groups. This work led by the AXIOM-MetaToul platform of TOXALIM on urine samples from pregnant women has needed an important methodological development to determine, from the analysis of urinary metabolites, the specter of exposure of these women to pesticides. It required the implementation of last generation separative and spectrometric methods (ultra high performance liquid chromatography coupled to high resolution mass spectrometry), but also a deep understanding of the metabolism of pesticides (Jamin et al., 2013).
3. In addition to studying the presence of pesticide residues in human biological fluids, one of the possible approaches to take into account these multiple exposures is to study the biological changes in the exposed body to better understand the mechanisms involved in the exposure to pollutants.
- These changes can be identified through a comprehensive analysis of endogenous metabolites in biological fluids (metabolic footprint). These metabolites are small molecules such as eg amino acids, (hippuric acid, citric) small organic acids, or sugars, or simple fatty acids. On urine samples from the same cohort, it was therefore performed a metabolomic analysis to understand the impact of exposure to a complex mixture of pesticides on the metabolome (all metabolites present in the sample). NMR analyzes were performed and data were processed by partial least squares regression (PLS-DA). Exposure groups previously formed from cereal crops were discriminated by this approach. This discrimination is based primarily on metabolites such as glycine, threonine, glycerophosphocholine, or lactic and citric acids. These results suggest that exposure to complex mixtures of pesticides at low doses is likely to have an impact on the metabolic pathways of the body by affecting energy metabolism and oxidative stress (Bonvallot et al., 2013). This metabolomics study is the first of its kind to examine the impact of exposure to pesticides and to have been conducted in humans.
4. Metabolomic approaches have also been developed from experimental studies in the context of the programs EXPOMATPEST (Impact of maternal exposure to a mixture of pesticides on immunity, hematopoiesis and the central nervous system of the offspring) and SPICY (Exposure to individual pesticides or in combination: evaluation of biomarkers), supported respectively by the ANR and the ANSES and coordinated by L. Payrastre (Team E6-TMR).
- In this case, three pesticides (chlorpyrifos, atrazine and endosulfan) alone or in combination were incorporated at low-dose (corresponding to the TDI) into the food distributed to mice from mating until weaning of pups. These were then fed with the food received by their mother for an additional 11 weeks. A comparison of the effects of the mixture with those observed with each of the individual substances was carried out not only on the metabolome, but also on hematopoiesis, immunity, liver function, and central nervous system. The metabolomic analysis of plasma of adults showed differences between males and females. Regardless of gender, the metabolic fingerprint obtained following dietary exposure to a mixture of pesticides is different from that observed in animals consuming food contaminated with the pesticides, and they are themselves different from those not exposed to pesticides. Metabolic disturbances between different groups of animals were visible from weaning (Demur et al., 2013). Metabolomics is not only suitable for the study of biological fluids such as urine or plasma, but can also be used in tissue extracts. Liver extracts were analyzed by NMR in mice exposed in utero to low doses of endosulfan, confirming that metabolic disturbances identified from plasma could be associated with oxidant stress and impairment of glucose metabolism observed in the liver (Canlet et al. 2013). Furthermore this work, carried out in collaboration with a team from INSERM (CHU Purpan, Toulouse, France), show that at doses which individually are supposed not to engage in harmful health effects, of atrazine, chlorpyrifos, endosulfan, are capable of disrupting hematopoiesis, immunity and the expression of certain liver genes related to stress and cellular toxicity. These effects were observed after pre-and postnatal exposure but some effects were observed in mice after weaning. At hematopoietic level the mixture does not exert a higher effect than that of the individual pesticides. In contrast, on synaptic plasticity, it has a higher effect than pesticides taken individually in some animals at 14 weeks of age. In the liver, the mixture generally reproduced the modulation induced by indivudual pesticides in terms of expression of stress and toxicity genes. In general, the impact of mixitures can hardly be predicted from the impact of individual pesticides according to the parameter studied since the mixture may or may not produce an effect greater than or equal to that of individual pesticides (Demur et al. 2013). The in vitro tests with the same mixture on murine or human hepatocytes in primary culture indicate that the effects observed with the mixture are mainly due to the effects of the most active of the pesticide compound (endosulfan and chlorpyrifos depending on the test) (Rouimi et al., 2012).
5. Along with the metabolomics studies, several recent studies at TOXALIM were aimed at deepening the mechanism of the toxic action of pesticides at low doses.
- Among these, a series of studies concerned Fipronil, active substance used as an insecticide and as an external parasite treatment for dogs and cats. These studies in rats have shown that the transformation of fipronil to fipronil sulfone, under the action of the liver P450 cytochrome, could play a major role in the ability of this insecticide to disrupt thyroid function (Roques et al. , 2012). Two nuclear receptors PXR and CAR have been identified as targets of fipronil in the liver. They act as key regulators of the molecular pathways involved on the one hand in the metabolism of fipronil to fipronil sulfone, and secondly in the metabolism and transport of thyroid hormones (Roques et al., 2013). The study of the tissue distribution of radiolabeled Fipronil in rats indicates that it is the sulfone form (active) that is mainly found in tissues such as adipose tissue, adrenal, liver, kidney, testes or the brain 72 hours after oral administration of fipronil, while the latter has disappeared (Cravedi et al., in press). The team E10-TCMX, led by Roger Rahmani, has obtained several important results on organochlorine pesticides from work in human hepatocytes in primary culture or cell lines derived from human hepatocellular carcinoma (HepG2 line). It has highlighted the profound changes in the phenotype of hepatocytes when they were treated with pesticides such as DDT, heptachlor and endosulfan, causing cell dedifferentiation and induction of epithelial-mesenchymal transition phenomena known to be related to tumor development (Zucchini-Pascal et al., 2012). It also showed that endosulfan could disrupt cell adhesion and migration mechanisms (Peyre et al., 2012).
6. Having developed a high throughput in vitro genotoxicity assay based on the phosphorylation of the histone H2AX reflecting double-strand DNA breaks, the team E2-MeX led by Daniel Zalko, applied it to the pesticide mixtures identified in the Pericles project.
- Several cell lines were used for this screening, but only HepG2 gave positive results for the 7 tested mixtures consisting of procymidone, iprodione, cyprodinil, fludioxonil and-cyhalothrin (see above Graillot et al. , 2012a). This test was also used to highlight the genotoxic effect of a family of pesticides: methyl pyrrazoles (Graillot et al, 2012b).
In conclusion:
The richness of this scientific production makes TOXALIM a leading player among European laboratories working on the risk associated with pesticide residues.
- Two researchers of Toxalim (L. Payrastre and R. Rahmani) took part in the collective INSERM expertise "Pesticides: health effects".
- Using unconventional approaches (outside the guidelines of regulatory toxicology) teams of this unit could dissect some mechanisms of action at the cellular and molecular levels and thereby contribute to the development of the body of knowledge necessary to a better assessment of the risk associated with pesticide exposure at low doses.
- Several studies aim to provide answers to the difficult problem of mixtures which requires the implementation of new approaches tailored to high throughput.
- The overall metabolomics results show that global untargeted approaches can help not only to characterize exposure to complex mixtures of pesticides at low doses, but also to discriminate against people based on the level of exposure. The subtle changes in metabolism, although they can not be considered a priori as adverse effects, however, indicate that metabolic pathways can be modified by exposure to pesticides at concentrations found in the environment and / or food . The interpretation of these data is a field of investigation still barely covered by toxicology.
