Final Report Summary - DENAMIC (Developmental neurotoxicity assessment of mixtures in children)
Executive Summary:
Worldwide, serious concern has arisen about the increased incidence of learning and developmental disorders in children and the potential role of exposure to neurotoxic chemicals during early brain development. DENAMIC "Developmental Neurotoxicity Assessment of Mixtures in Children" investigated developmentally neurotoxic effects of low-concentration mixtures of biocides and a number of common environmental pollutants in children. Research in DENAMIC focused on hazard characterization and epidemiology.
The hazard characterisation studied the effects of neurotoxic chemicals and mixtures thereof using novel tools, testing methods and procedures for screening (mixtures of) chemicals for (developmental)neurotoxicity. Rodent studies focused on the effects of low-level exposure to biocides and mixtures on neurobehavior, cognitive and motor function, including evaluations of underlying mechanisms of observed effects with consideration of exposure timing, critical windows during neuronal development and consequences on susceptibility. Early-life exposure during neurodevelopment resulted in persistent effects on behaviour, cognition, and motor activity later in life. Molecular pathway analysis elucidated mechanisms for some of the observed behavior and cognitive effects. Developmental neurotoxicity was also studied in early life-stage zebrafish (larvae) and cellular screening methods for (developmental) neurotoxicity which included assays for neurophysiology and neurodevelopment. DENAMIC contributed to the knowledge needed for optimal comparison and evaluation of the predictive power of in vitro systems. Several of the applied in vitro systems predict neurodevelopmental effects in vivo. Experimental studies with mixtures showed that low-level exposure to chemicals that are not toxic on an individual basis can together lead to a biological or toxic effect if present in a mixture. Also, non-additive interactions occur with combinations of biocides with a different mechanism of action. Another important aspect of the project was the research on potential biomarkers for (developmental) neurotoxicity in animal models using (epi-)genomics, proteomics and metabolomics. Feasibility of biomarkers was analysed by target analysis of specific biomarkers in human tissues.
In the epidemiology part, prenatal and early-childhood exposure was studied in maternal urine, breast milk, cord blood and urine of children in European cohorts. Associations with learning and developmental disorders, including ADHD, ASD and anxiety were explored. Prenatal and neonatal exposure profiles in cohorts (Norway, Netherlands, Slovakia and Spain) showed that the European population is exposed to low concentrations of neurotoxic chemicals (e.g. PCBs, organophosphates, carbamates, pyrethroids, PFAS, methylmercury). A number of associations were found between the exposure of neurotoxic chemical (e.g. DDT, PCBs, PFOA) and neuropsychological and behavioural development in children. Factors affecting the relative differences in socio-demographic and economic impact for (developmental) neurotoxicity resulting from exposure to environmental neurotoxicants were also studied, including sex and gender differences (as observed in cohorts and animal studies) and differences in exposure levels in different EU regions.
New data collected in this project in experimental studies, human exposure analyses and epidemiology was integrated to evaluate current risk for developmental neurotoxicity in the EU population. The experimental data on mixtures concluded that additivity cannot be assumed as a default approach for risk assessment of mixtures, and a uncertainty factor for increased effects in case of mixture exposure has been proposed. The risk assessment showed that the margins of safety (MOS) in the EU are sufficient for neurotoxic and neurobehavioral endpoints for carbaryl and cypermethrin, while for chlorpyrifos, endosulfan and methylmercury such MOS may not be adequate in specific populations. It is therefore recommended to investigate whether further exposure reduction measures are needed. Moreover, current risk assessment assumes that no effect levels are safe based on individual chemical toxicity studies. Some experiments done in DENAMIC indicate that combined exposure to biocides or other contaminants, that individually not resulting in any effect, may in fact cause an effect when present in mixtures. This observation brings doubt in the present approach for risk assessment, which is based on exposure to individual chemicals and may in fact result in an underestimation of the (human) risk.
Project Context and Objectives:
Worldwide, serious concern has arisen about the increased incidence of learning and developmental disorders in children and the potential role of exposure to neurotoxic chemicals during early brain development. The European commission-funded project DENAMIC "Developmental Neurotoxicity Assessment of Mixtures in Children" (www.denamic-project.eu) has investigated neurotoxic effects of low-concentration mixtures of biocides and a number of common environmental pollutants in children. We focused on (subclinical) effects on learning (cognitive skills) and developmental disorders in children (e.g. ADHD, autism spectrum disorders and anxiety disorders). DENAMIC has developed tools and methods for screening of neurotoxic effects. The project duration was from 1st January 2012 to 31st December 2015. The developed methods and obtained results have been used for recommendations for risk management to the EU and WHO and will largely support the EU chemicals legislation for identifying (potential) developmental neurotoxicants.
DENAMIC approach. The research in DENAMIC consisted of two distinct pillars, 1) hazard characterization, and 2) exposure/epidemiology. The first pillar involved hazard characterisation for neurotoxic chemicals and mixtures thereof. Novel tools, testing methods and procedures for screening (mixtures of) chemicals for (developmental) neurotoxicity were developed. In the second pillar (exposure/epidemiology), prenatal and early-childhood exposure was studied in maternal urine, breast milk, cord blood and urine of the child. In the epidemiological part, existing cohorts and a new tailor made cohort of mother/child pairs focused on learning and developmental disorders, including ADHD onset. In both the experimental and epidemiological part of DENAMIC, research was done on potential biomarkers for developmental neurotoxicity by using innovative biotechnology tools. The DENAMIC approach is based on a combination of in vitro, in vivo studies and human exposure assessment to further unravel the pathways and mechanism involved in chemical exposure and neurodevelopmental disorder development.
DENAMIC chemicals-of-interest. Because of recent concerns of cognitive and neurobehavioural effects related to biocide exposure and if proven necessary the possibility to protect future generations by regulatory measures, DENAMIC has primarily focused on possible neurotoxic effects of biocides. From each biocide group of organophosphates, carbamates, pyrethroids and organochlorines one chemical was selected as reference compound; chlorpyrifos, carbaryl, cypermethrin, endosulfan respectively. In most studies, also developmental neurotoxic reference chemical MeHg was included. The lists also included metabolites, as for some biocides these can be more potent than their parent compounds. Nonetheless, several other common environmental pollutants (e.g. flame retardants, plasticizers, perfluorinated compounds, methylmercury) are known or suspected neurotoxicants and were therefore also included. Selection of chemicals was based on representatives of different categories of chemicals that are known or suspected to induce neurotoxic effects in in vivo experimental studies and/or have been reported of relevance in human epidemiological studies, or the occurrence of biocides and metabolites in food. An overview of the chemicals tested in DENAMIC is provided in Table 1. Effects of individual chemicals as well as chemical mixtures were investigated.
DENAMIC objectives. This project focused on (subclinical) effects on learning (cognitive skills) and developmental disorders in children (e.g. ADHD, autism spectrum disorders and anxiety disorders) by exposure to chemicals. One of the aims was to develop better and more sophisticated tools, procedures and testing methods to screen chemicals for (developmental) neurotoxicity (DNT), besides improving the assessment of exposures (environmental contaminants) and identification of biomarker effects (including effects on proteins important for learning and developmental function). Different exposure scenarios were studied, with focus on timing of (early developmental) exposure, bioaccumulation and interactions of complex (low-dose) mixtures. Regarding mixtures, our principal goal was to evaluate the use of dose-additivity for human risk assessment purposes. The effects of the selected chemicals were studied at different biological levels, thereby improving the interpretation and extrapolation of in vitro neurotoxicity data for human risk assessment. In view of possible effects, mechanisms of disease development and individual susceptibility were also included. The validity and usability of these integrated tools were tested in epidemiological studies, focusing especially on learning and developmental disorders in young children. Another objective was to study the differences in prenatal and neonatal exposure of the selected neurotoxic chemicals in four different European regions.
These objectives are the backbone of the DENAMIC project which was organised in 7 work packages (Fig. 1). The scientific work packages focused on development of screening tools for (developmental) neurotoxicity (WP2), timing of exposure in behaviour, cognitive and motor function assessment (WP3), exposure assessment (WP4), epidemiology (WP5), and finally the risk assessment (WP6). WP 1 was dedicated to management and WP7 to dissemination.
The hazard characterisation included in vitro, ex vivo and in vivo studies using innovative models spanning molecular, cellular and multicellular (functional) processes, brain slices, brain and behaviour with a focus on neuronal development. In vitro screening methods for neurotoxicity were developed and applied to study effects of individual model chemicals and mixtures thereof. These methods included assays for neurophysiology and neurodevelopment.
To further investigate the mechanism(s) of action of neurotoxic chemicals, ex vivo, in vivo (behavioural) neurotoxicity, and cognitive and motor function models were developed. In addition, an early life-stage zebrafish model to study neurotoxicity and behaviour development was included. The underlying mechanisms of observed effects on behaviour and cognitive function were studied as well, with consideration of exposure timing, critical windows during neuronal development and consequences on susceptibility. An important aspect of the project was the research on potential biomarkers for (developmental) neurotoxicity in animal models using (epi-)genomics, proteomics and metabolomics. The feasibility of these biomarkers was analysed by a target analysis of specific biomarkers in cohorts. For the study of mixtures, a targeted approach was used, in which data on single chemicals determined the composition of investigated mixtures. Effects of mixtures were also studied in experimental systems with different levels of biological complexity in WP2 (screening tools for (developmental) neurotoxicity) and WP3 (timing of exposure in behaviour, cognitive and motor function assessment). Possible (developmental) neurotoxic potential of complex mixtures extracted from human milk were investigated as well.
Differences in prenatal and neonatal exposure to a selected number of suspected neurotoxic environmental pollutants were studied in four different European regions (WP4). Different exposure profiles were studied in European cohorts in Norway, Netherlands, Slovakia and Spain. In addition, a new cohort of mother-child pairs was set up (WP5). In these cohorts, also possible associations between the selected environmental pollutants with neurodevelopmental outcomes were investigated, in particular with learning (cognitive) and neurobehavioural (ADHD, ASD, anxiety) development or disorders. The relevance of selected biomarkers was also explored in cohort studies.
Associations between neurodevelopmental disorders and chemical exposure observed in the cohorts were also compared with the collected experimental data to explore biological plausibility for the observed associations. Another effort included the analysis of the validity and usability of the integrated screening tools and methods for neurotoxicity developed in the hazard characterization. A specific integration of the different levels (cohorts and experimental models) was the study of effects of exposure to extracts of human milk as a real-life exposure situation in in vitro and in vivo models.
The integration of results from experimental studies, human exposure analyses and epidemiology was used to improve human risk assessment. Based on the experimental data it was concluded that the additivity model cannot be assumed as a default approach for risk assessment of mixtures, and a uncertainty factor for increased effects in case of mixture exposure has been proposed. Factors affecting the relative differences in socio-demographic and economic impact for (developmental) neurotoxicity resulting from exposure to environmental neurotoxicants were also studied. These factors include sex and gender differences and differences in exposure levels in different EU regions.
DENAMIC had the following objectives:
1. Develop tools, testing methods and procedures to screen chemicals for (developmental) neurotoxicity.
2. Improve the predictive power of developmental neurotoxicity testing by developing protein markers, micro-arrays and early screening assays.
3. Determine the significance of differences in exposure timing and critical windows during neuronal development for the induction of neurotoxicological and neurobehavioural effects during early development.
4. Determine the effects and mechanisms on the interaction between low doses of (mixtures of) biocides and a number of common environmental pollutants, together with timing of exposure, on the induction of neurotoxicological and neurobehavioural effects during early development.
5. Determine the consequences of increased susceptibility to neurotoxicants in adults after exposure to (mixtures of) environmental toxicants in early life with respect to the acceleration of neurodegenerative processes.
6. Set-up a tailor-made cohort of mother/child pairs to study learning and developmental disorders.
7. Collect information on the exposure of mothers and children to neurotoxic contaminants, with the focus on but not limited to biocides, exposure across Europe (Norway, Netherlands, Slovakia, and Spain).
8. Improve human risk assessment for developmental neurotoxicity based on human hazard information from toxicological testing and exposure data of neurotoxic chemicals in mothers and children.
9. Carry out a health impact assessment for children based on the relationship between cognitive properties from the tailor-made cohort study and exposure to environmental agents.
10. Predict the cognitive effects for different European regions, US and global situation from the data of the tailor-made cohort and the exposure data of the other cohorts across Europe (Norway, Netherlands, Slovakia, Spain).
11. Environmental justice: integrate the potential impacts possible of (developmental) neurotoxicity from the exposure to biocides and environmental pollutants on the development of environmentally safe livelihoods.
12. Provide a series of recommendations on the use of biocides and other neurotoxic chemicals for agriculture, EU chemical legislation, and sustainable development, to protect the health of future generations (Sustainable Development Strategy).
13. Disseminate the knowledge to stakeholders (producers, users), environmental organisations, policy representatives, EU and national authorities.
Project Results:
1. Screening tools for (developmental) neurotoxicity
In view of ethical considerations and to expand testing capacity, rapid in vitro test systems are urgently needed to (partly) replace animal tests for (developmental) neurotoxicity, e.g. to prioritize for testing using the most appropriate in vivo tests to study (developmental) effects on behaviour and cognitive or motor functions. DENAMIC developed new tools to perform in vitro (developmental) neurotoxicity hazard characterisation of the selected biocides, environmental pollutants and mixtures thereof on a molecular and cellular level. In vitro assays were used to study effects on neurophysiology and neurodevelopment at non-cytotoxic exposure concentrations. Zebrafish larvae were included as an intact organism screening model that is not considered animal research according to the EU Directive 20/63 (2010) on the protection of animals used for scientific purposes, and a behavioural test method was developed. Finally, in the in vivo studies with rodents, also a number of new approaches (refinement in vivo) were included to measure effects and exposure concentrations that are also described below.
During brain development cellular processes are strictly coordinated in time and place in the developing brain. These processes include cell proliferation, migration of neural cells to the appropriate brain region, differentiation to different neuronal subtypes and the formation of neuronal networks. All these processes can be affected by exposure to chemicals and this can be studied using staining of specific immunocytochemical markers. Generating full concentration-response relationships of different chemicals would be hardly feasible using traditional imaging techniques. In DENAMIC, this was however achieved by using a high content imaging system to study the effects of in vitro exposure of cortical neurons to chemicals on the number of neurons, neurites and synapses.
As the main function of the nervous system is to collect, combine and transmit information using chemical messengers (neurotransmitters) and electrical signals (action potentials), a recent effort of the developmental neurotoxicity (DNT) community comprises the inclusion of neurophysiological endpoints in DNT. Therefore, the neurotoxicity studies in DENAMIC included the study of effects of chemicals on several processes underlying neuronal function (neurophysiology). A developmental component was included mainly by expanding the exposure duration in vitro. Effects of acute and (sub) chronic exposure were investigated on calcium homeostasis/signalling in PC12 cells and electrical activity of primary cortical cultures.
In vitro studies on the effects of chemicals on calcium homeostasis commonly investigate the effects of acute exposure. In DENAMIC, investigations of the effects of acute exposure to insecticides on calcium homeostasis in PC12 cells have revealed that several insecticides potently inhibit voltage-gated calcium channels (VGCC). However, subchronic and repeated exposures were also studied as these are more realistic exposure scenarios since exposure to insecticides usually occurs chronically and repeatedly via occupational (e.g. agriculture) settings or consumption of food. It was demonstrated that the potency of most insecticides for inhibition of VGCCs is comparable for acute and repeated exposure, but also that the potency of insecticides for inhibition of VGCCs is lower in subchronic exposure conditions. From these studies it is concluded that investigation of effects of acute exposure to chemicals is generally sufficient for rapid detection of insecticide-induced inhibition of VGCCs. However, some differences in potency of insecticides for inhibition of VGCCs were observed between the exposure scenarios, which argue for the inclusion of repeated exposure conditions for specific studies.
Rat primary cortical cultures grown on multi-well multi-electrode arrays (mwMEAs) were used to investigate possible effects of acute and chronic exposure to chemicals on neuronal activity. Most studies reported in literature have tested only for effects on electrical activity resulting from acute exposure of robustly active cultures. Such an approach was also used in DENAMIC, demonstrating concentration-dependent inhibition or enhancement of electrical activity in the cultures induced by exposure to various insecticides. In addition, effects of chronic exposure on the development of electrical activity were tested, demonstrating lower effective concentrations in some cases. Nevertheless, the acute concentration-response studies reveal in all cases the potential for effects on neurophysiology resulting from chemical exposure. Generating full concentration-response curves with adequate numbers of replicates with regard to effects of chemicals on electrical activity in primary cortical cultures grown on MEAs has the potential to be used in integrated testing strategies. Due to its heterogeneous nature with regard to the neural cell types present (a range of neuronal and glial subtypes), numerous potential molecular targets are present in this system with which chemicals may interact. Interactions of chemicals with key structures involved in the release or receipt of inter- and intercellular signals within a neuronal network will result in changes in the electrical activity. As such, MEA systems provide an integrated measure to screen rapidly and reproducibly for effects on neurotransmission that can be applied for rapid and reproducible screening.
There were also a number of additional technical developments. Firstly, the in vitro neurophysiological development of mouse neural progenitor cells was characterized. Mouse neural progenitor cells differentiate into nervous system–specific cell types and has already proven valuable to detect DNT using biochemical and morphological techniques. A number of functional neuronal parameters were investigated to explore their applicability for neurophysiological in vitro DNT testing. Secondly, experimental approaches for medium- to high-throughput assessment of effects of chemicals on TTR binding and ryanodine receptor mediated calcium homeostasis/signalling with the use of brain microsomes have been developed. Moreover, effect directed analysis was used to investigate the effects of complex mixtures extracted from human milk inducing effects on AChE activity.
Chemical-induced effects on behaviour of zebrafish larvae were tested at concentrations that not induce embryo toxicity (scored morphologically). Normal swimming behaviour and the response to light-dark transitions were monitored and analysed automatically in a high throughput set-up. Many chemicals were tested and some were shown to have effects at low exposure concentrations. Underlying mechanisms were studied using proteomics and neurotransmitter analyses (using LC-MS and microsensors, see below). A selection of 21 genes was used to explore their combined use (the zebrafish neurotox array) as biomarkers for (developmental) neurotoxicity. Although detailed analysis of predictivity is pending, zebrafish larvae have shown to be sensitive for neurodevelopmental effects and may be used in an alternative strategy to prioritize compounds for further in vivo testing.
Effects of exposure to chemicals during neurodevelopment were studied in rodent models, either in mice exposed neonatally or in rats exposed during gestation and lactation. In these studies, rat dams were exposed to chemicals during gestation and lactation. The offspring was submitted to an extensive set of tests for (effects on) behaviour and cognitive and motor functions (each animal was submitted to all tests). Besides this extensive testing paradigm a number of innovative methods were explored. These include the use of EEG as a non-invasive method to detect neurotoxicity was explored, as well as the use of stochastic microsensors to detect neurotransmitters and cytokines in brain dialysate from freely moving rats.
The electroencephalogram (EEG) is an electrophysiological measure of brain neuronal activity and could therefore be used to detect effects of chemicals on neuronal activity associated with neurological alterations. Possible alterations in EEG after developmental exposure to biocides have not been investigated previously. EEG analysis has the potential to reveal alterations in neuronal activity or effects on the sleep-wake cycle (which is a feature of developmental disorders such as ADHD or ADS).
In vivo brain microdialysis was used to obtain extracellular fluid from different cerebral regions, in which levels of different neurotransmitters and metabolites can be measured. Different neurotransmitters were measured by HPLC although the number of analyses is limited by the sample volume. This limitation has been overcome by the use of stochastic microsensors that have been developed in DENAMIC. These sensors allow the simultaneous measurements of several neurotransmitters and cytokines in a very small amount of sample. In DENAMIC, different types of microsensors were developed and evaluated showing the best performance by stochastic microsensors. These sensors were also used to determine neurotransmitters in zebrafish larvae, different biological fluids from rats, and cerebrospinal fluid and cord blood from children. Very low levels of neurotransmitters can be determined in small sample volumes of different types of matrices for which sample processing is not needed.
A number of other innovative approaches were also used to explore potential biomarkers for developmental neurotoxicity. Effects were investigated on neurotransmitter profiles (LC-MS), and the expression of genes (genomics), proteins (proteomics), metabolites (metabolomics) and glycans in different experimental models. A number of selected biomarkers were evaluated for their use in humans, although their predictive value remains to be determined.
2. Neurodevelopmental outcomes in children after early life exposure and biological plausibility from experimental studies
2.1 Cohort studies
Various recent epidemiological studies have indicated that exposure to low doses of environmental biologically active contaminants during human development can have unwanted effects on cognitive development in childhood. Foetuses and children are especially vulnerable to the exposure to environmental pollutants in comparison to adults, since their organs and systems are still developing and their detoxification mechanisms are not yet fully mature. The exposure to organophosphate biocides (OP) during early development has been associated with impaired child neuropsychological development in previous studies. Because of recent concerns of cognitive and neurobehavioural effects related to biocide exposure and if proven necessary the possibility to protect future generations by regulatory measures, DENAMIC primarily focussed on possible neurotoxic effects of biocides, but also of various other potential neurotoxic compounds such as perfluorinated chemicals (PFOS and PFOA) and methyl mercury (MeHg). Exposure to these different chemicals was determined in children and mothers in different regions of Europe. Prenatal and postnatal exposure to different neurotoxic chemicals was assessed in mother-child cohorts from Norway (HUMIS), Netherlands (LINC), Slovakia (PRENATAL) and Spain (INMA, Valencia and Sabadell sub-cohorts). The associations between the pre- and postnatal exposure to neurotoxic chemicals and child neuropsychological, cognitive and behaviour development in these cohorts were explored.
Exposure to neurotoxic chemicals
Several factors with an influence on exposure levels were observed. The cohorts studied were from different countries with different food and life style habits, which may influence risks of abnormal neurodevelopment. For example, fish consumption in Norway and meat consumption in Slovakia is higher than in the other studied countries. Another example is the distance of people from farm fields.
Urine from mothers and children was investigated on the occurrence of OP metabolites, and carbamates and metabolites. A large number of breast milk samples were analysed for chlorpyrifos, carbamates and metabolites, pyrethroids, organochlorine biocides, PCBs, PBDEs, perfluoroalkyl substances (PFASs) and MeHg. Additionally, in cord blood PFASs, PCBs, organochlorines and mercury were determined.
A limited number of OP and carbamates were detected in urine, and the detection frequencies in urine were low with only three OP metabolites (DMTP, DAP, TCPY) above 30%, and only two carbamates (aldicarb sulfone and 3-hydroxy carbofuran) above 20%. The urine biomonitoring studies showed that difference were observed in the levels and patterns between Spain and Slovakia. The sum of the concentration of OP metabolites in Valencia (25 µg/g creatinine) was higher than in Sabadell and Prenatal (Slovakia; 16 and 13 µg/g creatinine, respectively). The levels of OPs were lower than found in previously reported birth cohorts. An association between OP levels and season of sampling was found (higher levels in summer or fall), and in one cohort the residential proximity to fields was associated with higher OP levels. An association between maternal fruit and vegetable intake was found in the Spanish cohort.
In breast milk carbamates were for the first time ever reported, but the frequency of detection was low (<16%), and the levels were lower than chlorpyrifos. Chlorpyrifos and pyrethroids were for the first time reported in European breast milk; globally only three studies are available. Interestingly, chlorpyrifos was found in most of the breast milk samples from The Netherlands and Slovakia, but was almost absent in Norway. The highest levels were found in Slovakia (median 44 ng/kg ww, maximum 452 ng/kg ww), and lower levels in The Netherlands (median 18 ng/kg ww, maximum 78 ng/kg ww). We found evidence that these non-persistent chemicals such as chlorpyrifos, cypermethrin and permethrin can be found in European breast milk. This indicates that people are exposed either incidentally/episodically, or continuously, perhaps due to indoor use of biocides.
The mean level of methylmercury in breast milk from Norway is 39 ng/kg milk and twice as high as found in the Netherlands and Slovakia. This higher exposure is probably due to the higher consumption of fish in Norway as fish is an important source of methylmercury. Higher consumption of fish in Norway is probably also linked to the higher mean PCB levels in the breast milk from Norway (60 ng/g lw) compared to the Netherlands (39 ng/g lw). In Slovakia the PCB levels were even higher which is probably related to the production of PCB formulations in the former Czechoslovakia. Fortunately levels have been decreasing in the last decades.
The levels of organochlorine biocides, especially DDTs, are much higher in Slovakia than in Norway and The Netherlands. As the ratio of 4,4’-DDE and 4,4’-DDT is lower in Slovakia (25) than in the other countries (35 and 29, respectively) this indicates that DDT exposure was reduced the earliest in the Netherlands, followed by Norway and finally Slovakia.
PFOS and PFOA were detected at high frequency and at relative high levels in cord blood (µg/l range). Lower levels of PFOS and PFOA were found in breast milk (ng/l range); nevertheless PFOA and PFOS were present in all breast milk samples. Median levels of PFOS in breast milk from Norway were 2-5 times higher than in The Netherlands and Slovakia, but levels of PFOA were similar between the countries. The presence of elevated levels of PFOA in some human milk samples from Norway have been suggested to be related to the use of ski wax materials that can contain PFOA.
In general, we can conclude that contaminant patterns and levels are different between Norway, Netherlands and Slovakia. The general public in these countries are therefore exposed to a different mixture of contaminants with a different risk for adverse effects on brain development.
Associations between exposure and neurodevelopmental outcomes
Combining information on measured exposures and neurodevelopmental outcomes in the four cohorts showed that for some contaminants associations were found, for some not, and for some results were less consistent. Each cohort collected a wide range of information on subjects at different time points, using various questionnaires – general, environmental, and food frequency questionnaires. General questionnaires were focused mainly on socio-demographic data, smoking and alcohol exposure, maternal and family history, children’s health and breastfeeding information. Environmental questionnaires were aimed at the household characteristics and environmental exposures to chemicals. Also characteristics of birth (delivery) were reported as well as basic anthropometric measurements in all cohorts.
A meta-analysis of seven European cohorts was conducted to study the effects of persistent organic pollutants on several areas of neuropsychological development (cognitive delay, ADHD, and ASD). Association between exposure levels of organochlorine and bromine compounds (PCB-153, HCB, p,p'-DDE, PFOS, PFOA, PBDE47, PBDE153) and neuropsychological outcomes were investigated. This study showed that no associations were found with any of these compounds. Exploratory studies in the LINC cohort, using stratified data for gender, found positive associations between exposure to PCBs and some organochlorine biocides and ADHD symptoms, externalizing behaviour and aggressive behaviour in girls.
Using novel statistical methods for the analysis of multiple persistent organic pollutants (PCBs, DDT, DDE, HCB, HCH, oxychlordan, PBDEs) in breast milk of the HUMIS cohort, it was demonstrated that p,p’-DDT showed a positive association with increased behavioural problems in children at the age of 12 months. These associations were no longer found at 24 months of age. Exploratory studies at the LINC cohort showed a weak association between prenatal exposure to PCBs and behavioural problems and ADHD/ASD symptoms. Postnatal exposure (via breast milk) showed positive associations between PCBs and ADHD symptoms, externalizing behaviour and aggressive behaviour. Associations were sex-specific.
The cohorts observed no increased risk of abnormal neuropsychological and behavioural development in association with perinatal exposure to PFOS, PFOA. However, exploratory results of the LINC cohort showed an adverse association between PFOA and externalizing behaviour in boys. Results of PFAS associations are less consistent, as of the eight published studies on the association between neuropsychological development and PFOS or PFOA, only two found an increased risk.
Various health effects have been reported associated with pre- and postnatal exposures to low concentrations of organophosphates (OP) in urban and agricultural communities, in which an increased risk of nervous system disorders in children and effects in the neurobehavioural development have been shown. The DENAMIC cohorts did not find associations between OP levels and neuropsychological outcomes (ADHD, ASD, cognitive delay). The absence of the association can be due to the lower levels of OPs found in our study compared to North American studies were in general higher levels of OPs are found.
2.2. Effects observed in animal models
Many different biocides and other environmental pollutants were included in the experimental studies in DENAMIC. After the first phase of the project, a number of reference chemicals were selected to be included in all studies to allow evaluation of the effects in different animal models and cellular systems: chlorpyrifos(-oxon), carbaryl, α-cypermethrin, endosulfan and methylmercury. Rodent models were used to study effects of exposure to these chemicals during critical window of neurodevelopment, mice that were exposed neonatally and rats exposed during gestation and lactation. In these studies effects of developmental exposure were studied on different parameters of spontaneous behaviour (mice and rats) as an apical test of neuronal function, habituation to a novel home environment (mice) which is the integrations of sensory input into motor output, circadian rhythm (rats), anxiety (rats), spatial learning (mice and rats), active avoidance (rats) and motor coordination (rats). In particular spontaneous activity and tests on cognitive functions are of interest as hyperactivity and impaired working memory have been described as symptoms in ADHD patients.
Studies in the neonatal mouse model revealed that exposure to MeHg, endosulfan, cypermethrin, chlorpyrifos or carbaryl can induce developmental neurotoxic effects, when administered as a single oral dose during a defined critical period of brain development. The animals showed distinct neurotoxic effects manifested as persistent aberrant spontaneous behaviour (hyperactivity) in a novel home environment and reduced ability to habituate, which also indicates impaired cognitive function as habituation is the integration of sensory information into motoric output. Exposure to chemicals during a critical period of neonatal brain development can thus be involved in induction of neurofunctional disorders related to ADHD (attention deficit hyperactive disorder) or reduced cognitive capacity. Possible effects on cholinergic susceptibility and neuroprotein levels were investigated as underlying mechanisms for the observed neurobehavioural effects in (young) adults. The studied proteins are involved in essential processes of the developing brain, such as dendritic arborisation, axonal outgrowth, neuronal connections and synaptogenesis. Disturbances of these processes can affect the development of specific neurotransmitter systems as well as the development of the cytoarchitecture of the brain, possibly leading to functional defects later in life and may thus have contributed to the observed developmental neurobehavioural aberrations. Altered levels of neuroproteins may serve as potential markers for developmental neurotoxicity although due to the dynamic and compensatory mechanisms of the developing brain, it is unlikely that a single differentially expressed protein can be used to predict functional neurotoxic effects.
Effects of exposure during gestation and lactation to MeHg, endosulfan, cypermethrin, chlorpyrifos or carbaryl were also studied in rats using a large amount of behavioural tests described above. Different effects of exposure to the reference chemicals during gestation and lactation were observed in the large array of neurobehavioural tests for behaviour, cognition and motor function. Spatial learning was impaired by all biocides, except carbaryl, in male offspring, but not in female offspring. In detail, spatial learning assessed in the Morris water maze is impaired after exposure to endosulfan and chlorpyrifos in male rats, while α-cypermethrin improves spatial learning. In the radial maze spatial learning is impaired after exposure to endosulfan, chlorpyrifos and cypermethrin in male rats. Spatial working memory was impaired in male rats exposed to chlorpyrifos, while spatial learning and cognitive functions appeared to be improved in female rats exposed to chlorpyrifos and α-cypermethrin. Carbaryl does not affect spatial learning in rats. Underlying mechanisms were studied using the measurement of neurotransmitters in brain dialysate and cytokines in brain tissues as a measure for neuroinflammation as this is associated with impaired spatial learning. All biocides induced neuroinflammation, increasing pro-inflammatory IL-1 β content in hippocampus correlated to spatial learning deficits with increased sensitivity of females for this mechanism. Impaired working memory has also been described to occur in ADHD patients. Motor coordination was impaired after exposure to endosulfan, cypermethrin or chlorpyrifos in female rats. There were also effects on motor function of male offspring that were exposed to chlorpyrifos or carbaryl. The impaired motor function of female offspring exposed during gestation and lactation to endosulfan or α-cypermethrin and of male offspring exposed to chlorpyrifos is associated with increased extracellular levels of the inhibitory neurotransmitter GABA in the cerebellum. Motor activity was increased after exposure to chlorpyrifos which is associated to an increased content of NMDA receptor subunits in the hippocampus. Alterations in NMDA receptor function have also been observed in ADHD animal models. Endosulfan is the only biocide inducing anxiety in males. Exposure to endosulfan, α-cypermethrin or chlorpyrifos also altered circadian rhythm of motor activity (hyperactivity) during the sleeping period in males.
In these studies with rats clear sex-differentiated effects resulting from exposure to the reference biocides were observed. This is important with regard to the study of (adverse) health outcome with sex differentiated prevalence such as neurodevelopmental disorders. Such sex specificity has been observed both in cohort studies as in animal models.
In the last phase of the project, a specific study on the chlorpyrifos was performed in a collaboration of different laboratories working in DENAMIC. In this study a neurobehavioural test with mice was included which measures attentional processing. With this test it is possible to study deficits in attention and impulsivity, which are behavioural traits that are analogous to symptoms in ADHD. It was shown in this study that neonatal exposure to chlorpyrifos impairs attention. A number of additional endpoints, including hippocampal synaptic plasticity, AChE activity in plasma and metabolomics were also included to explore underlying mechanisms of the observed neurobehavioural effect.
Experimental studies in DENAMIC and literature demonstrating neurodevelopmental effects by exposure to chlorpyrifos, PCBs, PFAS provide evidence for the biological plausibility of the causal relationship in the observed associations with exposure to these chemicals in the cohort studies, while there is less experimental evidence for developmental neurotoxicity for organochlorines.
2.3. Effects observed in cellular models and zebrafish larvae
Effects of selected biocides and environmental chemicals on morphological and neurophysiological development at non-cytotoxic exposure concentrations were investigated in cellular models. Effects of exposure on the number of neurons, neurites and synapses in mouse cortical cultures using high-content imaging, on calcium homeostasis and signalling in PC12 cells and brain microsomes using fluorescent calcium imaging and on electrical activity (development) of rat cortical cultures were investigated. In addition, behavioural toxicity in zebrafish larvae was used as an alternative in vitro testing system with intact animals. Chemical-induced effects on behaviour of zebrafish larvae were tested at concentrations that not induce embryo toxicity (scored morphologically).
Effects on number of neurons, neurites and synapses in mouse cortical cultures as measures for network development were studied in mouse primary cortical cultures after 14 days of in vitro culture that were treated at the 2nd day in vitro. Over 30 chemicals were tested in this system. No effects or only at the high (10 µM) exposure concentrations were observed for many of the chemicals including chlorpyrifos, chlorpyrifos-oxon, carbaryl and α-cypermethrin. For endosulfan and methylmercury, effects were observed at lower exposure concentrations, as well as for paraoxon-ethyl, paraquat, and 6-OH-BDE-47. Evaluation of the predictive value for developmental neurotoxicity with this system by comparing the identified developmental neurotoxicants in this system with a list of known developmental neurotoxicants demonstrated that approximately half of the known developmental neurotoxicants were predicted with this system.
Effects of acute exposure to insecticides on basal and depolarization-evoked intracellular calcium levels were studied in vitro in rat pheochromacytoma PC12 cells. The data demonstrate that α-cypermethrin, endosulfan, chlorpyrifos and methylmercury concentration-dependently decreased the depolarization-evoked intracellular calcium levels. A particularly interesting observation was that chlorpyrifos-oxon was much less potent compared to parent compound chlorpyrifos. Modest increases in basal intracellular calcium levels were observed for chlorpyrifos, chlorpyrifos-oxon or endosulfan at relatively high concentrations (10 µM) and at lower concentrations for methylmercury. No effects were observed for carbaryl even at a high concentration. Also the effects of subchronic and repeated exposure to the reference biocides on calcium homeostasis were investigated. Except for carbaryl, all selected insecticides inhibited depolarization-evoked intracellular calcium levels after subchronic exposure in PC12 cells while repeated exposure augmented the inhibition of the depolarization-evoked intracellular calcium levels induced by subchronic exposure to cypermethrin, chlorpyrifos, chlorpyrifos-oxon and endosulfan. These data show that calcium homeostasis in PC12 cells can be used as a sensitive readout for insecticide-induced neurotoxicity. From the studies using different exposure paradigms it is concluded that investigation of effects of acute exposure to chemicals is generally sufficient for rapid detection of insecticide (or chemical)-induced inhibition of VGCCs. However, some differences in potency of insecticides for inhibition of VGCCs were observed between the exposure scenarios, which argue for the inclusion of repeated exposure conditions for specific studies. Effects of chemicals on specific mechanisms calcium homeostasis were also investigated using brain microsomes. These studies showed that endosulfan and chlorpyrifos induce ryanodine receptor mediated calcium release from rat brain microsomes to a similar extent as positive control PCB-95. VGCC inhibition in PC12 cells reflected the developmental neurotoxicity of the reference chemicals in DENAMIC although further evaluation of predictivity of this model is pending.
Effects of acute and chronic exposure to insecticides and methylmercury on electrical activity were studied in primary cortical cultures grown on multi-well multi-electrode arrays (MEAs). Concentration-dependent inhibition of electrical activity (spike rate) was observed for acute exposure to methylmercury, chlorpyrifos and its oxon metabolite chlorpyrifos-oxon, α-cypermethrin and carbaryl. Acute exposure to endosulfan resulted in a biphasic response, showing a concentration-dependent increase followed by complete inhibition at the highest exposure concentrations. Lowest observed effect concentrations were in the (low) micromolar range. Chronic exposure to methylmercury, chlorpyrifos and a-cypermethrin inhibited the development of electrical activity, while exposure to endosulfan increased the activity. No significant effects were observed by chronic exposure to chlorpyrifos-oxon and carbaryl. The inhibition of spike activity in primary cortical cultures on MEAs by carbaryl, chlorpyrifos-oxon and methylmercury was observed at lower lowest observed effect concentrations in comparison with earlier studies. These differences may be related to interlaboratory differences in model characteristics, age of the cultures used for testing, (low) numbers of replicates, and differences in exposure paradigms (naïve culture for each replicate or cumulative dosing). The observed modulation of electrical activity by biocides and methylmercury confirms the known potential for developmental neurotoxicity of these chemicals and illustrates the applicability of heterogeneous cortical cultures grown on MEAs for screening purposes.
Also a number of tests were performed to evaluate the potential of chemicals to act via specific mechanisms that may underlie developmental neurotoxicity in vivo. The effects of over 30 chemicals were tested with regard to binding to thyroid hormone binding protein transthyretin (TTR). Thyroid hormones are critical to brain development and disruption has been demonstrated to result in impaired cognitive function. Competitive binding to TTR was observed for only a few of the tested chemicals, showing effects of chlorpyrifos-oxon, 6-OH-BDE47 and PFAS. Another well-known mechanism in particular for organophosphates and carbamates is the inhibition of AChE activity which may result in overactivation of the nervous system. Also in this system over 30 chemicals were tested confirming that in particular oxon metabolites of organophosphates and carbamates inhibit AChE.
Rapid screening using zebrafish larvae was used to determine effects of exposure on spontaneous behaviour and heart beat at exposure concentration ranges lower than those resulting in embryo toxicity. Over 30 chemicals were tested in this system. For several chemicals, including chlorpyrifos-oxon, α-cypermethrin, endosulfan and methylmercury, effects on spontaneous behaviour in zebrafish larvae was observed at low concentrations (<1 µM). Evaluation of the predictive value developmental neurotoxicity with this system by comparing the chemicals inducing developmental neurotoxicity in zebrafish larvae with a list of known mammalian developmental neurotoxicants demonstrated that the majority of known developmental neurotoxicants were predicted with this system.
Several of the applied in vitro systems in DENAMIC correctly predict neurodevelopmental effects. However, additional data is required to firmly establish their predictive power. The DENAMIC project has contributed to the body of knowledge needed for optimal comparison and evaluation of the predictive power of in vitro systems by 1) collecting data on the effects of chemicals and 2) characterization of the physiological relevance of in vitro systems and the measured parameters. As soon as the predictive power of an in vitro system is established it can be used to build an Adverse Outcome Pathway (AOP), in which Molecular Initiating Events and/or Key Events investigated in in vitro systems can be used to predict an ultimate Adverse Outcome in individuals and/or populations. An example of a Molecular Initiating Event could be inhibitory interactions of chemicals with voltage-gated calcium channels. Nevertheless it is currently still unclear which specific organ, organism and population responses an inhibition of VGCCs by chemicals may cause. Primary cortical cultures grown on MEAs can be regarded an integrated system in which effects may be caused by interactions of chemicals with one or more of the many molecular targets present in this system. Electrical (neuronal) activity can thus be regarded as a key event, as this is the inherent function of neurons in networks. In contrast, the neonatal animal model as such does not fit into a distinct AOP for DNT as similar effects are observed for different types of chemicals. Nevertheless, this model has proven to detect DNT effects from widely different compounds with different acute modes of action with delayed observable functional effects. The assays (and models) that are proposed to be used in experiments based on an AOP should be well-established and reliably (reproducibly) give mechanistic information on the effects of chemicals on key events. When generating experimental data to support or build such an AOP, exposure scenarios should also reflect what is relevant for the human exposure situation with regard to exposure levels, exposure timing and exposure duration. To establish AOPs for adverse cognitive outcomes associated with neurotoxicity the pathophysiology of (subclinical) neurodevelopmental disorders and the role of exposure to chemicals should also be further clarified.
2.4. Mechanistic pathways and biomarkers using molecular (omics) studies
The aim of the Omics (genomics, proteomics, glycan analysis and metabolomics) studies was to 1) investigate underlying molecular mechanisms of observed effects on behaviour and cognitive function resulting from exposure to biocides in animal models, and 2) develop possible biomarkers for (developmental) neurotoxicity and cognitive disorders in humans.
In developmental exposure studies, altered regulation of proteins and metabolites by exposure was studied in different parts of rat brain (hippocampus, cortex, cerebellum, and striatum), mouse brain (hippocampus, cortex) and whole brains of zebrafish larvae. More than 100 regulated proteins were observed in rats exposed to biocides, and proteins were regulated differently in the four brain areas. Examples of regulated protein pathways in rats exposed to biocides were glutamate processing, DNA repair, tricarboxylic acid cycle, mitochondrial transport, and sodium/potassium-transporting. In rats exposed to α-cypermethrin, the main effects were seen in striatum, and less in hippocampus, cortex, or cerebellum. Like the effects on neurobehaviour and cognition, also these effects on protein regulation were sex-specific and showed in some cases opposite regulation between sexes. A smaller effect on protein expression in striatum was found in rats exposed to the mixture of α-cypermethrin and endosulfan.
Also in the neonatal mouse model, in which mice were exposed to only a single dose at postnatal day 10, proteins were observed to be regulated, although to a limited extent (low fold-changes), after exposure to biocides, mainly in hippocampus. In mice exposed to chlorpyrifos, carbaryl, or endosulfan, a total of 689 brain proteins were identified of which 67 were significantly regulated. For chlorpyrifos the highest number of regulated proteins was identified in cortex. Pathway analysis identified that chlorpyrifos-induced changes were observed in the cluster of mitochondrial/organelle proteins and neuronal diseases like Alzheimer’s, Parkinson’s and Huntington’s disease. For PFHxS and endosulfan also the strongest regulation was observed in hippocampus and the regulated proteins were related to pathways of neurofilament and intermediate filament cytoskeleton.
Metabolomics studies in the brains of rodents exposed during neurodevelopment focused on four major neurotransmitter pathways, but also untargeted metabolomics studies were carried out. Untargeted metabolomics revealed 400-600 metabolites in the rat brain tissues of which more than 50 were significantly regulated in rats exposed to biocides. In the mouse studies only a limited number of metabolites were regulated. Pathway mapping showed that various metabolomic pathways were affected in rats including those related to serotoninergic, dopaminergic, cholinergic, GABAergic neurotransmitters, sodium/chloride dependent neurotransmitter transporters, transport of nucleosides, and purine metabolism. Also these effects of biocide exposure on metabolite levels of were brain tissue and sex-specific. An example of such a rat brain region specific effect is the upregulation of choline and acetylcholine in striatum only to endosulfan, indicating that there are synapse-related effects. Metabolites were very differently regulated between the compounds studied. As an example α-cypermethrin exposure showed mainly up-regulated metabolites while exposure to endosulfan or chlorpyrifos showed both up- and down-regulated metabolites in brain tissues.
The cognitive and behaviour tests with rats showed that developmental exposure to biocides had effects on anxiety, spatial learning, motor coordination, motor activity, active avoidance, and working memory. These effects were biocide specific and some effects were sex-specific. Metabolomics studies (down and upregulated metabolites) give mechanistic information that may be related to the observed neurobehavioural and cognitive effects. For example, developmental exposure to endosulfan caused a decrease in the neurotransmitter GABA in males (hippocampus), but not in females. This down regulation of GABA in males correlated well to the impairment of spatial learning that was observed in male but not in female rats. Rats exposed to biocides also showed increased levels of cytokines (interleukin) in the hippocampus of males only, suggesting a role of biocides neuroinflammation. The increase of cytokines may very well cause spatial learning and memory impairments. The metabolite succinate can be used as a marker of inflammation, as it is one of the metabolites of the tricarboxylic acid (TCA) cycle. It also plays a role in the GABA shunt pathway. The level of succinate in hippocampus was indeed increased in male offspring after biocide exposure, and levels of GABA and glutamate were increased as well. The increase of succinate may be due to the increased levels of GABA and glutamate available for the GABA shunt pathway. Succinate can stimulate dendritic cells but has also been associated with inflammation and is a metabolite related to innate immune signalling. Cypermethrin exposed rats showed effects on the dopamine pathway, showing effects on the levels of dopamine precursors and metabolites. Dopamine is involved in motor control which was shown to be impaired in female offspring after developmental exposure to α-cypermethrin.
In conclusion, proteomics and metabolomics studies elucidated some of the mechanisms responsible for the observed behaviour and cognitive effects due to chemical exposure. These studies also showed that chemical exposure can results in brain tissue and sex-specific metabolic pathway changes.
Before protein, peptide, glycan and metabolite biomarkers as found in the cell and animal models can be used non-invasively in humans, it needs to be determined whether these can be determined in human tissues. With regard to effects on the developing brain, human cord blood was chosen as the most appropriate non-invasive exposure matrix. Proteomic studies selected a number of candidate biomarkers of chemical exposure related to (developmental) neurotoxicity from the cell and animal models (rat and mice). These candidate biomarkers were verified in human cord blood from three DENAMIC cohorts (LINC, PRENATAL and HUMIS). For this study a novel approach based on a combination of human PC12 or SH-SY5Y cells and a protein labelling method was developed.
DENAMIC identified neuroproteins as potential markers for developmental neurotoxicity. Neuroproteins are involved in essential processes of the developing brain, such as dendritic arborisation, axonal outgrowth, neuronal connections and synaptogenesis. A potential biomarker for neuroinflammation was the level of IL-1β in hippocampus that was related to biocide exposure. Besides these effects on neuroprotein expression, the DENAMIC animal and cell based studies also showed that various neurotransmitter pathways were affected by chemical exposure, and some neurotransmitters are therefore potential markers of exposure to developmental neurotoxicants. An imbalance in the serotonergic and dopaminergic neurotransmitter systems has been suggested to be responsible for hyperactive behaviour in humans. Cord blood from the Dutch mother-child cohort LINC was therefore used to investigate the relationship between neurotransmitters and chemical exposure. This study showed that PFOA and DEHP metabolites were associated with changes in the levels of GABA, acetylcholine, and epinephrine. Whether these changes in neurotransmitter levels are related to abnormal behavioural development, needs to be further explored.
3. Mixture toxicity
Humans are exposed to mixtures of chemicals, either simultaneously or in succession, chronically or repetitively. In the DENAMIC project, therefore also studies on the effects of combined exposure to two or more chemicals were included. Exposure to (mixtures of) chemicals is lifestyle and region dependent and consequently the combined exposure situation can vary strongly. A number of different (types) of mixtures were decided on to streamline the combined exposures tested in the DENAMIC experimental studies on developmental neurotoxicity. Binary mixtures were composed based on 1) occurrence and persistence in humans and 2) differences or similarities in the mechanism of action of the individual constituents. Two types of mechanisms with regard to effect resulting from combined exposure were focused on in the experimental studies with animal and cellular models. Firstly, it was evaluated whether additivity occurs in combined exposure situations. Secondly, it was investigated whether combined low-level exposure to chemical that do not show toxicity on an individual basis results in effects.
Three different binary mixtures were tested in the neonatal mouse model: chlorpyrifos in combination with carbaryl, chlorpyrifos in combination with methylmercury, and PFAS perfluorohexane sulfonate in combination with endosulfan. Neonatal mice were exposed to a single oral dose of the binary mixture and the individual constituents on postnatal day 10. Effects on spontaneous behaviour in a novel home environment and neuroprotein expression were tested at the ages of 2 and 4 months. It was demonstrated that in case of combined exposure interaction of the chemicals resulted in exacerbation of developmental neurotoxic effects in a dose-responsive manner. Observed effects included hyperactivity and impaired habituation as a measure of cognitive function, and these effects were also observed in case of combined exposure at dose were the individual constituents did not cause any effect. Co-exposure to chlorpyrifos and carbaryl did also result in an increased susceptibility of the cholinergic system in adults, as well as in changed neuroprotein levels. Persistent developmental effects in the neonatal mouse model can thus be induced by various persistent and non-persistent chemicals, either individually or in low-level mixtures, by exposure during postnatal day 10 which is a defined critical period in brain development. The behavioural effects observed after exposure to chlorpyrifos in combination with methylmercury, and to PFAS perfluorohexane sulfonate in combination with endosulfan appeared close to additive, despite different mechanisms of action of the individual constituents. In case of exposure to chlorpyrifos in combination with carbaryl, the neurobehaviour effects were exacerbated in a synergistic manner, with chlorpyrifos exposure as the main determinant.
Rats were exposed to a mixture of endosulfan and α-cypermethrin during gestation and lactation. Male offspring demonstrated hyperactivity (associated with decreased NMDA expression in the hippocampus) and their spatial and working memory was impaired by potentially synergistic interactions between the individual constituents, as both of these did not have this effect. This was however not observed for every neurobehavioural parameter. For example, both endosulfan and α-cypermethrin impaired motor coordination in female offspring individually, while this was not observed in animals exposed to the mixture. Impairment of reference memory in male offspring was induced by both of these chemicals, as well as by the mixture.
Several binary (and one tertiary) mixtures of insecticides were investigated for their acute effects on inhibition of VGCCs. The effects of the mixtures were compared with the effects of the individual compounds on inhibition of VGCCs to evaluate potential additivity. In the first phase of the project, LOECs were selected from concentration-response curves of the different insecticides and combined in binary mixtures. It was demonstrated that exposure to chlorpyrifos in combination with either its oxon-analogue, methylmercury or with parathion-ethyl did not increase the degree of inhibition in comparison with the inhibition caused by the individual compounds. In the next studies, exposure levels resulting in 20% inhibition were interpolated from concentration-response curves and used to calculate an expected additivity effect range. For every exposure to a binary mixture (chlorpyrifos in combination with endosulfan, chlorpyrifos in combination with cypermethrin, cypermethrin in combination with endosulfan), the effects of the mixture were within the expected additivity effect range. The observed additivity was however less equivocal for the tertiary mixture of chlorpyrifos, cypermethrin and endosulfan.
Effects of acute exposure to binary mixtures were also tested on electrical activity of primary cortical cultures grown on MEAs. It was concluded that the degree of variation in this type of data precludes the use of an expected additivity effect range to exclude additivity. Therefore, the effect of a mixture was compared with the effects of its individual constituents. It was demonstrated that exposure to chlorpyrifos in combination with methylmercury concentration-dependently inhibited electrical activity at concentrations that did not have significant effects of the individual constituents. Endosulfan and a-cypermethrin have opposite effects at low exposure concentrations on the electrical activity in primary cortical cultures grown on MEAs. Co-exposures to endosulfan (inducing concentration-dependent increases in electrical activity) and cypermethrin (inducing concentration-dependent decreases in electrical activity) demonstrated that the inhibition by cypermethrin was dominating the effect of the binary mixture.
A detailed analysis was obtained for the interaction of chlorpyrifos-oxon and carbaryl with regard to AChE inhibition, a neurotoxic mechanism in particular specific for biocides. When testing binary mixtures at different potency ratios of the individual compounds, effective concentrations resulting in 50% inhibition were somewhat lower than expected in case of additivity. A similar analysis was performed for the interaction of PFAS perfluorooctanesulfonic acid and 6-OH-BDE47 with regard to TTR binding. Dose-response curves of these three mixtures indicate that the combined TTR-binding potency of the two test compounds is well described by concentration addition.
In vitro and in vivo studies performed in the DENAMIC project thus clearly indicate that exposure to combinations of chemicals may result in non-additive effects depending on levels and compounds. Effects of mixtures have been shown to be larger or smaller compared to the expected combined effect of the individual constituents based on additivity. In a number of experiments it has even been shown that combined exposure to chemicals caused developmental neurotoxicity while the exposure levels of the individual chemicals did not result in any effects.
Finally, extracts were prepared of breast milk from various DENAMIC cohorts (LINC, HUMIS, PRENATAL). These extracts contained a complex mixture of real-life chemicals and were tested in three different bioassays, i.e. the acetylcholinesterase (AChE) inhibition assay, the transthyretin (TTR) binding assay, and the endoplasmatic reticulum (ER) calcium-release assay. AChE inhibiting potency was observed in several breast milk extracts. Out of 17 individual breast milk samples from the HUMIS and LINC cohort, highest AChE inhibiting potency was observed for a LINC sample corresponding to 4.4 ng dichlorvos-equivalent per g milk. So far, the observed inhibition could not be explained by target analyzed AChE inhibiting insecticides: all carbamate concentrations were
4. Risk assessment and recommendations on the use of neurotoxic chemicals
The assessment for risk of developmental neurotoxicity in humans was based on the uptake of contaminants from breast milk (postnatal exposure) or (cord) blood levels (prenatal exposure) placed into the context of effective concentrations from the in vivo and in vitro DENAMIC experimental data.
For the in vivo studies the no observed effect levels (NOELs) of the DENAMIC reference chemicals in the neurobehavioural mouse studies were used and compared to the uptake of contaminants by breast milk. These dose – response relationships are relevant for human risk assessment of (developmental) neurotoxicological and neurobehavioural effects. The no-effect doses are used as points-of-departure for the human risk assessment for (developmental) neurotoxicity.
As the mouse studies are single dose experiments, albeit at a postnatal period in which brain development in the mouse is at a critical phase, a number of uncertainty factors (UFs) were applied. Besides the default factors of 10 for interspecies and interhuman variability, an additional safety factor 3 was applied for extrapolating a single dose regime to a chronic exposure. As a result a total composite UF of 10*10*3=300 was applied for the risk assessment of the compounds tested in DENAMIC with these in vivo experiments. In combination with the NOELs from these neurobehavioural mouse studies an interim reference dose (i-RfD) was calculated for the DENAMIC reference compounds. It should be noted that if this i-RfD is used for risk assessment no further safety factors are needed when comparing this with uptake via human milk, because the i-RfD already includes uncertainty factors. Furthermore, for calculations of exposures of the breast fed infant it was assumed that milk consumption is 125 g (ml)/kg bw/day.
For the risk assessment based on the in vitro studies the medium concentrations used in the in vitro experiments resulting in effects on various neurotoxicity endpoints were compared with concentrations in human tissues, in particular human cord blood. For this approach it was necessary to differentiate in uncertainty factors for toxicokinetics and toxicodynamics and the differentiation thereof in interspecies and interindividual uncertainty factors. For risk assessment done based on the results from in vitro experiments and internal exposure levels in plasma the interspecies variability in toxicokinetics was assumed 1. As a result a composite uncertainty factor of 2.5*10=25 was used under these conditions, which was used to calculate an interim reference concentration (i-RfC) for neurotoxicity. It should be noted that if this i-RfC is used for risk assessment no further safety factors are needed when comparing this with internal levels e.g. blood, because this i-RfC already includes uncertainty factors.
The risk assessment of chlorpyrifos indicated that based on in vivo, in vitro and human exposure data from DENAMIC in combination with those from literature on average the margin of safety is likely adequate with a possible exception for high exposure groups. Based on the experimental results from the DENAMIC project no further reductions in human exposure to chlorpyrifos are needed with possible exception for high exposure groups. For carbaryl the in vivo, in vitro and human exposure data from DENAMIC indicated that the margin of safety is more than sufficient and no effects are expected with regard to neurodevelopmental toxicity. No further reductions in human exposure to carbaryl are needed in the EU. Similarly, cypermethrin in vivo and human exposure data from DENAMIC showed that the margin of safety is more than sufficient and no effects are expected. Also for this compound no further reductions in human exposure to cypermethrin are needed in the EU. For endosulfan the in vitro studies and human exposure data from DENAMIC showed that the margin of safety may not be adequate and that therefore neurodevelopmental toxicity cannot be excluded in the EU population. Based on these experimental results from the DENAMIC project further reductions in human exposure to endosulfan are needed in the EU, but as this biocide is already banned in the EU, practical reduction may be difficult to achieve. Also for methylmercury the in vitro studies and human exposure data from DENAMIC concluded that exposure of methylmercury is several orders of magnitude above a calculated i-RfC. Under the assumption that the in vitro medium concentrations of methylmercury are a good reflection of the internal blood levels the i-RfC is several orders of magnitude lower than the internal exposure levels in blood, which indicates significant health risks for neurotoxic effects at present concentrations in some EU countries. As a result, neurotoxic effects based on internal blood levels cannot be excluded in the EU. Based on these experimental results from the DENAMIC project further reductions in human exposure to methylmercury are needed in the EU. A reduction may be difficult as this will need a change in food habits as a large part of the uptake of methyl mercury is by seafood. There are also other food items that can be a source of methylmercury, for example rice.
In vitro and in vivo studies performed in the DENAMIC project clearly indicate that combinations of biocides may give non-additive effects depending on the dose levels and compounds showing that effects of mixtures have been shown to be larger or smaller compared to the expected combined effect of the individual constituents based on additivity. In a number of experiments it has even been shown that combined exposure to chemicals caused developmental neurotoxicity while the exposure levels of the individual chemicals did not result in any effects. This observation in particular brings doubt in the present approach for risk assessment, which is based on exposure to individual biocides and may in fact result in an underestimation of the (human) risk.
It is thus clear that additivity as a default approach in the human risk assessment of biocides cannot a priori be assumed and an additional uncertainty factor for potential mixture effects is proposed. This uncertainty is founded in particular in observations of non-additive mixture effects and sex-specific toxicity. Based on the in vitro mixture neurotoxicity experiments it was concluded that a deviation of a factor 2 from additivity may in fact occur. If the results from the in vivo mouse and rat mixture studies are combined, the uncertainty of non-additive effects appears to be in range between 2 and 5. The application of such an additional safety factor did not change the main conclusions in the risk assessment for the contaminants mentioned above.
5. General conclusions
DENAMIC developed tools and methods to detect chemicals’ effects on the developing nervous system. More than 30 chemicals were tested individually and some in mixtures. Exposure concentrations in the in vivo models (rat, mice, zebrafish) resulting in effects were orders of magnitude lower than tested before. There are critical windows of exposure during neurodevelopment (e.g. during the brain growth spurt). The animal models showed that early-life exposure during neurodevelopment resulted in persistent effects on behaviour, cognition, and motor activity later in life.
The project has contributed to the body of knowledge needed for optimal comparison and evaluation of the predictive power of in vitro systems by 1) collecting data on the effects of chemicals and 2) characterization of the physiological relevance of in vitro systems and the measured parameters. Several of the applied in vitro systems predict neurodevelopmental effects in vivo.
In European cohorts (subclinical) effects on learning (cognitive skills) and developmental disorders in children (e.g. ADHD, autism spectrum disorders and anxiety disorders) were studied. The European population is exposed to low concentrations of neurotoxic chemicals (e.g. PCBs, organophosphate carbamates and pyrethroids biocides, PFAS, methylmercury). The exposure occurs both prenatally (via cord blood) and/or postnatally (via human milk). A number of associations were found between the exposure of these neurotoxic chemical (e.g. DDT, PCBs, PFOA) and neuropsychological and behavioural development in children. The mechanism based cell and animal studies contributed to the understanding of developmental neurotoxicity and low-dose chemical mixture exposures in children. Both the cohort studies as the animal studies showed that effects can be sex-specific. Molecular pathway analysis elucidated mechanisms for some of the observed behaviour and cognitive effects.
Neurotoxic effects of mixtures of chemicals at low concentrations were investigated in the experimental studies. DENAMIC showed that low-level exposure to chemicals that are not toxic on an individual basis can together lead to a biological or toxic effect if present in a mixture. Mixture studies in DENAMIC have indicated that non-additive interactions occur with combinations of biocides with a different mechanism of action resulting in additional uncertainty factors. Deviations of a factor 2-5 from additivity may occur. Co-exposure to a mixture could also cause an increased susceptibility in the adult animal and exacerbate developmental toxic effects that might have consequences for the induction of neurological/neurodegenerative disorders/diseases.
The risk assessment concluded that based on the experimental studies done within the DENAMIC project the margins of safety in the EU are sufficient for neurotoxic and neurobehavioural endpoints for carbaryl and cypermethrin. For chlorpyrifos, endosulfan and methylmercury such margins of safety may not be adequate in specific populations and based on either in vivo and/or in vitro experiments in DENAMIC adverse health effects cannot be excluded. It is therefore recommended to investigate whether further exposure reduction measures are needed.
Current risk assessment assumes that no effect levels are safe based on individual chemical toxicity studies. Some experiments done in the DENAMIC project indicate that combined exposure to biocides or other contaminants, that are individually present at levels not resulting in any effect, may in fact cause an effect when present in mixtures. This observation brings doubt in the present approach for risk assessment, which is based on exposure to individual chemicals and may in fact result in an underestimation of the (human) risk.
Potential Impact:
Socio-demographic and economic impact
Factors affecting the relative differences in socio-demographic and economic impact for (developmental) neurotoxicity resulting from exposure to environmental neurotoxicants were studied in DENAMIC. These factors include sex and gender differences and differences in exposure levels in different EU regions. A number of studies have been conducted on children and pregnant women evaluating the association between organophosphate biocide (OP) exposure and the socio-demographic, environmental and dietary characteristics. Socio-demographic characteristics showed that significant differences can occur between cases (ADHD, cognitive, ASD symptoms) and control groups according to the maternal education, smoking habit during pregnancy, age, and body mass index (BMI) before pregnancy. DENAMIC found that OP biocide levels in urine were associated with season of sampling, being higher in samples taken during summer and fall. Mothers who finished university studies had higher OP levels. Mother who smoked during pregnancy had lower OP levels. The residential proximity to fields was associated in some cohorts with higher OP levels. In the Spanish cohorts an inverse association between OP levels and the maternal BMI before pregnancy and a direct association with maternal intake of fruit and vegetables during pregnancy was found.
In the DENAMIC project a number of epidemiological associations between exposure to chemicals and neurodevelopmental outcomes were observed, and information on causal relationships between exposure to chemicals and (developmental) neurotoxicity was collected in experimental studies. These outcomes can be used in future or more specific calculations for the costs associated with the exposure to chemicals and associated adverse health effects. By studying social and economic aspects of families and the possible relation with developmental effects in children due to exposure to contaminants, information in DENAMIC has provided an informed basis for poverty alleviation efforts and associated health and economic benefits. In that respect the effects of this project could be stronger in Eastern-Europe as exposures may be higher and economical positions are disadvantaged compared to Western-Europe. The DENAMIC inventory of exposure of children to chemicals, can be used to equalizing current inequalities in children’s health across Europe related to the adverse health effects on cognition and neurodevelopmental disorders. These data have significant value for the EU, and for national authorities and other stakeholders involved in health sciences, environmental monitoring, and risk assessment.
There is increasing awareness that the sex of humans, animal models and cellular models matters in biomedical science. Several sex-specific associations between developmental exposure to chemicals and neurodevelopmental outcomes have been observed in DENAMIC. The sex difference in prevalence of neurodevelopmental disorders urges the inclusion of both sexes in experimental studies on the impact of exposure to chemicals on the pathophysiology of such disorders. Several sex-specific effects on neurodevelopmental outcomes have also been observed in developmentally exposed rats in DENAMIC. When sex or gender is not taken into account in experimental or epidemiological studies, important effects may be missed. Taking sex and gender into account will improve reproducibility and translational value of toxicological and epidemiological data.
Societal awareness and risk assessment
There were several activities focused on rising public awareness on exposure of children to neurotoxic compounds in the environment in DENAMIC. Press contacts have actively been sought, with press releases in English, Dutch and Norwegian, in order to explain the benefits of the project for the European citizen. National policy makers were enabled to take advantage of the DENAMIC results through direct approach by newsletters, meetings etc. organized within the dissemination aspect of this project. Stakeholders (health centres, industry, academia) were invited and joined the two DENAMIC workshops.
Another important active dissemination activity for the general public was the recruitment process of finding respondents for the cohort studies. We have discussed the topics of DENAMIC with thousands of people during the recruitment period. Also DENAMIC was discussed with the people at the midwifery clinics, hospitals and health care centers. During the recruitment process a large number of people has been reached for finding parents that wanted to contribute to the cohort studies. During the recruitment phase, pregnant women were approached at the beginning of pregnancy at gynaecological clinics, midwifery clinics, hospitals and health care centers. DENAMIC researchers and experienced and trained health care workers explained to women the main topics of the project and the importance of the prenatal and early life quality for future health of their children. Women received an Informed Consent form, with further information on the project aims and details of follow up, in order to give them the opportunity to discuss the project participation with their families. This way, not only pregnant women, but their partners and relatives as well, were informed on the project topics and reasons for follow-up. Furthermore, women obtained a list of contact persons in case they wanted to ask questions on the project later on. During follow up meetings of children, parents received feedback from the research team on the project results. An Information Letter was sent to parents or presented during a meeting, including general results from the project (description of exposure to selected environmental pollutants in the cohort). This approach helped parents to become more educated about environmental risk factors for neurotoxic developmental effects in children.
DENAMIC generated a large amount of biomonitoring data for biocides, flame retardants, PFASs, methyl mercury and other compounds. This data will be made available for the databases and ECHA, and other European projects.
DENAMIC showed that well developed animal models is an important bridge between in vitro studies and epidemiological studies. From a prospective point of view, the constant development and marketing of new chemicals will constantly alter the human exposure situation via various exposure routes, such as diet, house dust or air, and also in different combinations of chemicals. Therefore, the DENAMIC results, from the neonatal animal model, shows that risk assessment based on single doses of individual chemicals might underestimate the effects of a chemical as agents can interact to exacerbate developmental toxic effects. This implicates that risk assessment needs to be continuously updated in order to meet the present environmental situation. The risk assessment knowledge will be further disseminated to the WHO and EFSA. Also the risk reduction of certain contaminants (methyl mercury and chlorpyrifos) by intake of food will be further discussed with EFSA WHO, and OECD.
Neurotoxicity data (including new tools and testing methods for individual chemicals and mixtures) were discussed with the OECD at the final DENAMIC workshop and can be used for adverse outcome pathways (AOPs) that are related to (developmental) neurotoxicity. Some of the methodologies may also be applied in future studies in other fields of research (e.g. toxicity or health effects in other target organs).
Policy makers were involved in the DENAMIC workshops. The DENAMIC results will be communicated with EFSA, WHO, European Environment Agency, and OECD. Especially the mixture toxicity results, risk assessment, and risk reduction actions for certain chemicals will be discussed with EFSA, and WHO. In addition, the approaches followed by DENAMIC can be applied in a more generic way, such as in comparable studies on health effects of chemicals and studies on other diseases. Additionally there are several collaborations of the consortium with other universities and regulatory agencies that can use the collected information.
Dissemination activities
Dissemination activities undertaken in DENAMIC was by publication of data in peer-reviewed scientific journals, website (www.denamic-project.eu) project logo, summarizing flyers at the start and end of the DENAMIC project, project newsletters, and a midterm and final workshop. In addition, a documentary film was produced, distributed and made available on YouTube to highlight the results of the project. Also two articles with an educational character on topics related to the DENAMIC project were submitted to popular journals. Currently, a total number of 42 peer reviewed articles were published, 7 are under review and more papers are in preparation.
The flyers contained information on the project to create awareness among all researchers, consumer organisations, health care institutes, industries and authorities active in the field of neurotoxicity, developmental disorders, and neurotoxicity testing. A summary of applied technologies and general outcomes of the DENAMIC project is provided in a second flyer. Items in the newsletters were on the introduction of the DENAMIC project, short reports on project meetings, update on work and announcement of midterm scientific workshop, introduction of young scientists in DENAMIC, and reports on the workshops.
Both the midterm and final workshop were two-day scientific workshops. The first workshop focused on novel tools and methods to screen for (developmental) neurotoxicity; in vitro assays, reduction and refinement of animal studies; epidemiological evidence for effects of environmental exposure on child neuropsychological development and the use of biomarkers in developmental neurotoxicity. The second workshop focussed on DENAMIC findings with regard to screening and tool development, mixture toxicity, exposure assessment, neurodevelopmental epidemiology and risk assessment (including socio-economics). Various stakeholders from the EU and North-American industry, academia, non-governmental organizations and organizations for public health and environment participated in the workshops.
The DENAMIC film “Young Minds at Risk?” was made and sets the stage in which neurotoxic effects of low-concentration mixtures of chemicals in children were investigated. The film explains what the benefits are of chemicals, and explores how some of these chemicals can be harmful especially for the developing brain. How DENAMIC approached this problem and some key results are presented. The film was made with the aim to put the project into context for the viewer, explaining why the research work was needed, what problems it was trying to solve, how the work was carried out, who was involved, the collaborative nature of the process, summarise the results and explain how this knowledge can be applied. The film was launched at the final DENAMIC workshop and will be made available for special screenings at conferences and other events. The film is posted on the project website and made available to partner’s websites, other scientific & documentary websites and broadcasters by YouTube: https://youtu.be/drcOBcA2mD0. A film poster an electronic press kit (EPK) are available by the project website and a copy of the DVD can be requested by the coordinator of DENAMIC.
DENAMIC developed also new technologies. A microsensor for direct measurements of neurotransmitters and cytokines in small amounts of sample (e.g. urine, blood) was developed. The microsensor application and commercialization possibilities are studied. An apparatus to study the on-line behaviour of zebrafish was further developed, tested validated, and is commercially available. Both technologies have wider application areas such as the health and medical fields.
List of Websites:
www.denamic-project.eu
pim.leonards@vu.nl
Worldwide, serious concern has arisen about the increased incidence of learning and developmental disorders in children and the potential role of exposure to neurotoxic chemicals during early brain development. DENAMIC "Developmental Neurotoxicity Assessment of Mixtures in Children" investigated developmentally neurotoxic effects of low-concentration mixtures of biocides and a number of common environmental pollutants in children. Research in DENAMIC focused on hazard characterization and epidemiology.
The hazard characterisation studied the effects of neurotoxic chemicals and mixtures thereof using novel tools, testing methods and procedures for screening (mixtures of) chemicals for (developmental)neurotoxicity. Rodent studies focused on the effects of low-level exposure to biocides and mixtures on neurobehavior, cognitive and motor function, including evaluations of underlying mechanisms of observed effects with consideration of exposure timing, critical windows during neuronal development and consequences on susceptibility. Early-life exposure during neurodevelopment resulted in persistent effects on behaviour, cognition, and motor activity later in life. Molecular pathway analysis elucidated mechanisms for some of the observed behavior and cognitive effects. Developmental neurotoxicity was also studied in early life-stage zebrafish (larvae) and cellular screening methods for (developmental) neurotoxicity which included assays for neurophysiology and neurodevelopment. DENAMIC contributed to the knowledge needed for optimal comparison and evaluation of the predictive power of in vitro systems. Several of the applied in vitro systems predict neurodevelopmental effects in vivo. Experimental studies with mixtures showed that low-level exposure to chemicals that are not toxic on an individual basis can together lead to a biological or toxic effect if present in a mixture. Also, non-additive interactions occur with combinations of biocides with a different mechanism of action. Another important aspect of the project was the research on potential biomarkers for (developmental) neurotoxicity in animal models using (epi-)genomics, proteomics and metabolomics. Feasibility of biomarkers was analysed by target analysis of specific biomarkers in human tissues.
In the epidemiology part, prenatal and early-childhood exposure was studied in maternal urine, breast milk, cord blood and urine of children in European cohorts. Associations with learning and developmental disorders, including ADHD, ASD and anxiety were explored. Prenatal and neonatal exposure profiles in cohorts (Norway, Netherlands, Slovakia and Spain) showed that the European population is exposed to low concentrations of neurotoxic chemicals (e.g. PCBs, organophosphates, carbamates, pyrethroids, PFAS, methylmercury). A number of associations were found between the exposure of neurotoxic chemical (e.g. DDT, PCBs, PFOA) and neuropsychological and behavioural development in children. Factors affecting the relative differences in socio-demographic and economic impact for (developmental) neurotoxicity resulting from exposure to environmental neurotoxicants were also studied, including sex and gender differences (as observed in cohorts and animal studies) and differences in exposure levels in different EU regions.
New data collected in this project in experimental studies, human exposure analyses and epidemiology was integrated to evaluate current risk for developmental neurotoxicity in the EU population. The experimental data on mixtures concluded that additivity cannot be assumed as a default approach for risk assessment of mixtures, and a uncertainty factor for increased effects in case of mixture exposure has been proposed. The risk assessment showed that the margins of safety (MOS) in the EU are sufficient for neurotoxic and neurobehavioral endpoints for carbaryl and cypermethrin, while for chlorpyrifos, endosulfan and methylmercury such MOS may not be adequate in specific populations. It is therefore recommended to investigate whether further exposure reduction measures are needed. Moreover, current risk assessment assumes that no effect levels are safe based on individual chemical toxicity studies. Some experiments done in DENAMIC indicate that combined exposure to biocides or other contaminants, that individually not resulting in any effect, may in fact cause an effect when present in mixtures. This observation brings doubt in the present approach for risk assessment, which is based on exposure to individual chemicals and may in fact result in an underestimation of the (human) risk.
Project Context and Objectives:
Worldwide, serious concern has arisen about the increased incidence of learning and developmental disorders in children and the potential role of exposure to neurotoxic chemicals during early brain development. The European commission-funded project DENAMIC "Developmental Neurotoxicity Assessment of Mixtures in Children" (www.denamic-project.eu) has investigated neurotoxic effects of low-concentration mixtures of biocides and a number of common environmental pollutants in children. We focused on (subclinical) effects on learning (cognitive skills) and developmental disorders in children (e.g. ADHD, autism spectrum disorders and anxiety disorders). DENAMIC has developed tools and methods for screening of neurotoxic effects. The project duration was from 1st January 2012 to 31st December 2015. The developed methods and obtained results have been used for recommendations for risk management to the EU and WHO and will largely support the EU chemicals legislation for identifying (potential) developmental neurotoxicants.
DENAMIC approach. The research in DENAMIC consisted of two distinct pillars, 1) hazard characterization, and 2) exposure/epidemiology. The first pillar involved hazard characterisation for neurotoxic chemicals and mixtures thereof. Novel tools, testing methods and procedures for screening (mixtures of) chemicals for (developmental) neurotoxicity were developed. In the second pillar (exposure/epidemiology), prenatal and early-childhood exposure was studied in maternal urine, breast milk, cord blood and urine of the child. In the epidemiological part, existing cohorts and a new tailor made cohort of mother/child pairs focused on learning and developmental disorders, including ADHD onset. In both the experimental and epidemiological part of DENAMIC, research was done on potential biomarkers for developmental neurotoxicity by using innovative biotechnology tools. The DENAMIC approach is based on a combination of in vitro, in vivo studies and human exposure assessment to further unravel the pathways and mechanism involved in chemical exposure and neurodevelopmental disorder development.
DENAMIC chemicals-of-interest. Because of recent concerns of cognitive and neurobehavioural effects related to biocide exposure and if proven necessary the possibility to protect future generations by regulatory measures, DENAMIC has primarily focused on possible neurotoxic effects of biocides. From each biocide group of organophosphates, carbamates, pyrethroids and organochlorines one chemical was selected as reference compound; chlorpyrifos, carbaryl, cypermethrin, endosulfan respectively. In most studies, also developmental neurotoxic reference chemical MeHg was included. The lists also included metabolites, as for some biocides these can be more potent than their parent compounds. Nonetheless, several other common environmental pollutants (e.g. flame retardants, plasticizers, perfluorinated compounds, methylmercury) are known or suspected neurotoxicants and were therefore also included. Selection of chemicals was based on representatives of different categories of chemicals that are known or suspected to induce neurotoxic effects in in vivo experimental studies and/or have been reported of relevance in human epidemiological studies, or the occurrence of biocides and metabolites in food. An overview of the chemicals tested in DENAMIC is provided in Table 1. Effects of individual chemicals as well as chemical mixtures were investigated.
DENAMIC objectives. This project focused on (subclinical) effects on learning (cognitive skills) and developmental disorders in children (e.g. ADHD, autism spectrum disorders and anxiety disorders) by exposure to chemicals. One of the aims was to develop better and more sophisticated tools, procedures and testing methods to screen chemicals for (developmental) neurotoxicity (DNT), besides improving the assessment of exposures (environmental contaminants) and identification of biomarker effects (including effects on proteins important for learning and developmental function). Different exposure scenarios were studied, with focus on timing of (early developmental) exposure, bioaccumulation and interactions of complex (low-dose) mixtures. Regarding mixtures, our principal goal was to evaluate the use of dose-additivity for human risk assessment purposes. The effects of the selected chemicals were studied at different biological levels, thereby improving the interpretation and extrapolation of in vitro neurotoxicity data for human risk assessment. In view of possible effects, mechanisms of disease development and individual susceptibility were also included. The validity and usability of these integrated tools were tested in epidemiological studies, focusing especially on learning and developmental disorders in young children. Another objective was to study the differences in prenatal and neonatal exposure of the selected neurotoxic chemicals in four different European regions.
These objectives are the backbone of the DENAMIC project which was organised in 7 work packages (Fig. 1). The scientific work packages focused on development of screening tools for (developmental) neurotoxicity (WP2), timing of exposure in behaviour, cognitive and motor function assessment (WP3), exposure assessment (WP4), epidemiology (WP5), and finally the risk assessment (WP6). WP 1 was dedicated to management and WP7 to dissemination.
The hazard characterisation included in vitro, ex vivo and in vivo studies using innovative models spanning molecular, cellular and multicellular (functional) processes, brain slices, brain and behaviour with a focus on neuronal development. In vitro screening methods for neurotoxicity were developed and applied to study effects of individual model chemicals and mixtures thereof. These methods included assays for neurophysiology and neurodevelopment.
To further investigate the mechanism(s) of action of neurotoxic chemicals, ex vivo, in vivo (behavioural) neurotoxicity, and cognitive and motor function models were developed. In addition, an early life-stage zebrafish model to study neurotoxicity and behaviour development was included. The underlying mechanisms of observed effects on behaviour and cognitive function were studied as well, with consideration of exposure timing, critical windows during neuronal development and consequences on susceptibility. An important aspect of the project was the research on potential biomarkers for (developmental) neurotoxicity in animal models using (epi-)genomics, proteomics and metabolomics. The feasibility of these biomarkers was analysed by a target analysis of specific biomarkers in cohorts. For the study of mixtures, a targeted approach was used, in which data on single chemicals determined the composition of investigated mixtures. Effects of mixtures were also studied in experimental systems with different levels of biological complexity in WP2 (screening tools for (developmental) neurotoxicity) and WP3 (timing of exposure in behaviour, cognitive and motor function assessment). Possible (developmental) neurotoxic potential of complex mixtures extracted from human milk were investigated as well.
Differences in prenatal and neonatal exposure to a selected number of suspected neurotoxic environmental pollutants were studied in four different European regions (WP4). Different exposure profiles were studied in European cohorts in Norway, Netherlands, Slovakia and Spain. In addition, a new cohort of mother-child pairs was set up (WP5). In these cohorts, also possible associations between the selected environmental pollutants with neurodevelopmental outcomes were investigated, in particular with learning (cognitive) and neurobehavioural (ADHD, ASD, anxiety) development or disorders. The relevance of selected biomarkers was also explored in cohort studies.
Associations between neurodevelopmental disorders and chemical exposure observed in the cohorts were also compared with the collected experimental data to explore biological plausibility for the observed associations. Another effort included the analysis of the validity and usability of the integrated screening tools and methods for neurotoxicity developed in the hazard characterization. A specific integration of the different levels (cohorts and experimental models) was the study of effects of exposure to extracts of human milk as a real-life exposure situation in in vitro and in vivo models.
The integration of results from experimental studies, human exposure analyses and epidemiology was used to improve human risk assessment. Based on the experimental data it was concluded that the additivity model cannot be assumed as a default approach for risk assessment of mixtures, and a uncertainty factor for increased effects in case of mixture exposure has been proposed. Factors affecting the relative differences in socio-demographic and economic impact for (developmental) neurotoxicity resulting from exposure to environmental neurotoxicants were also studied. These factors include sex and gender differences and differences in exposure levels in different EU regions.
DENAMIC had the following objectives:
1. Develop tools, testing methods and procedures to screen chemicals for (developmental) neurotoxicity.
2. Improve the predictive power of developmental neurotoxicity testing by developing protein markers, micro-arrays and early screening assays.
3. Determine the significance of differences in exposure timing and critical windows during neuronal development for the induction of neurotoxicological and neurobehavioural effects during early development.
4. Determine the effects and mechanisms on the interaction between low doses of (mixtures of) biocides and a number of common environmental pollutants, together with timing of exposure, on the induction of neurotoxicological and neurobehavioural effects during early development.
5. Determine the consequences of increased susceptibility to neurotoxicants in adults after exposure to (mixtures of) environmental toxicants in early life with respect to the acceleration of neurodegenerative processes.
6. Set-up a tailor-made cohort of mother/child pairs to study learning and developmental disorders.
7. Collect information on the exposure of mothers and children to neurotoxic contaminants, with the focus on but not limited to biocides, exposure across Europe (Norway, Netherlands, Slovakia, and Spain).
8. Improve human risk assessment for developmental neurotoxicity based on human hazard information from toxicological testing and exposure data of neurotoxic chemicals in mothers and children.
9. Carry out a health impact assessment for children based on the relationship between cognitive properties from the tailor-made cohort study and exposure to environmental agents.
10. Predict the cognitive effects for different European regions, US and global situation from the data of the tailor-made cohort and the exposure data of the other cohorts across Europe (Norway, Netherlands, Slovakia, Spain).
11. Environmental justice: integrate the potential impacts possible of (developmental) neurotoxicity from the exposure to biocides and environmental pollutants on the development of environmentally safe livelihoods.
12. Provide a series of recommendations on the use of biocides and other neurotoxic chemicals for agriculture, EU chemical legislation, and sustainable development, to protect the health of future generations (Sustainable Development Strategy).
13. Disseminate the knowledge to stakeholders (producers, users), environmental organisations, policy representatives, EU and national authorities.
Project Results:
1. Screening tools for (developmental) neurotoxicity
In view of ethical considerations and to expand testing capacity, rapid in vitro test systems are urgently needed to (partly) replace animal tests for (developmental) neurotoxicity, e.g. to prioritize for testing using the most appropriate in vivo tests to study (developmental) effects on behaviour and cognitive or motor functions. DENAMIC developed new tools to perform in vitro (developmental) neurotoxicity hazard characterisation of the selected biocides, environmental pollutants and mixtures thereof on a molecular and cellular level. In vitro assays were used to study effects on neurophysiology and neurodevelopment at non-cytotoxic exposure concentrations. Zebrafish larvae were included as an intact organism screening model that is not considered animal research according to the EU Directive 20/63 (2010) on the protection of animals used for scientific purposes, and a behavioural test method was developed. Finally, in the in vivo studies with rodents, also a number of new approaches (refinement in vivo) were included to measure effects and exposure concentrations that are also described below.
During brain development cellular processes are strictly coordinated in time and place in the developing brain. These processes include cell proliferation, migration of neural cells to the appropriate brain region, differentiation to different neuronal subtypes and the formation of neuronal networks. All these processes can be affected by exposure to chemicals and this can be studied using staining of specific immunocytochemical markers. Generating full concentration-response relationships of different chemicals would be hardly feasible using traditional imaging techniques. In DENAMIC, this was however achieved by using a high content imaging system to study the effects of in vitro exposure of cortical neurons to chemicals on the number of neurons, neurites and synapses.
As the main function of the nervous system is to collect, combine and transmit information using chemical messengers (neurotransmitters) and electrical signals (action potentials), a recent effort of the developmental neurotoxicity (DNT) community comprises the inclusion of neurophysiological endpoints in DNT. Therefore, the neurotoxicity studies in DENAMIC included the study of effects of chemicals on several processes underlying neuronal function (neurophysiology). A developmental component was included mainly by expanding the exposure duration in vitro. Effects of acute and (sub) chronic exposure were investigated on calcium homeostasis/signalling in PC12 cells and electrical activity of primary cortical cultures.
In vitro studies on the effects of chemicals on calcium homeostasis commonly investigate the effects of acute exposure. In DENAMIC, investigations of the effects of acute exposure to insecticides on calcium homeostasis in PC12 cells have revealed that several insecticides potently inhibit voltage-gated calcium channels (VGCC). However, subchronic and repeated exposures were also studied as these are more realistic exposure scenarios since exposure to insecticides usually occurs chronically and repeatedly via occupational (e.g. agriculture) settings or consumption of food. It was demonstrated that the potency of most insecticides for inhibition of VGCCs is comparable for acute and repeated exposure, but also that the potency of insecticides for inhibition of VGCCs is lower in subchronic exposure conditions. From these studies it is concluded that investigation of effects of acute exposure to chemicals is generally sufficient for rapid detection of insecticide-induced inhibition of VGCCs. However, some differences in potency of insecticides for inhibition of VGCCs were observed between the exposure scenarios, which argue for the inclusion of repeated exposure conditions for specific studies.
Rat primary cortical cultures grown on multi-well multi-electrode arrays (mwMEAs) were used to investigate possible effects of acute and chronic exposure to chemicals on neuronal activity. Most studies reported in literature have tested only for effects on electrical activity resulting from acute exposure of robustly active cultures. Such an approach was also used in DENAMIC, demonstrating concentration-dependent inhibition or enhancement of electrical activity in the cultures induced by exposure to various insecticides. In addition, effects of chronic exposure on the development of electrical activity were tested, demonstrating lower effective concentrations in some cases. Nevertheless, the acute concentration-response studies reveal in all cases the potential for effects on neurophysiology resulting from chemical exposure. Generating full concentration-response curves with adequate numbers of replicates with regard to effects of chemicals on electrical activity in primary cortical cultures grown on MEAs has the potential to be used in integrated testing strategies. Due to its heterogeneous nature with regard to the neural cell types present (a range of neuronal and glial subtypes), numerous potential molecular targets are present in this system with which chemicals may interact. Interactions of chemicals with key structures involved in the release or receipt of inter- and intercellular signals within a neuronal network will result in changes in the electrical activity. As such, MEA systems provide an integrated measure to screen rapidly and reproducibly for effects on neurotransmission that can be applied for rapid and reproducible screening.
There were also a number of additional technical developments. Firstly, the in vitro neurophysiological development of mouse neural progenitor cells was characterized. Mouse neural progenitor cells differentiate into nervous system–specific cell types and has already proven valuable to detect DNT using biochemical and morphological techniques. A number of functional neuronal parameters were investigated to explore their applicability for neurophysiological in vitro DNT testing. Secondly, experimental approaches for medium- to high-throughput assessment of effects of chemicals on TTR binding and ryanodine receptor mediated calcium homeostasis/signalling with the use of brain microsomes have been developed. Moreover, effect directed analysis was used to investigate the effects of complex mixtures extracted from human milk inducing effects on AChE activity.
Chemical-induced effects on behaviour of zebrafish larvae were tested at concentrations that not induce embryo toxicity (scored morphologically). Normal swimming behaviour and the response to light-dark transitions were monitored and analysed automatically in a high throughput set-up. Many chemicals were tested and some were shown to have effects at low exposure concentrations. Underlying mechanisms were studied using proteomics and neurotransmitter analyses (using LC-MS and microsensors, see below). A selection of 21 genes was used to explore their combined use (the zebrafish neurotox array) as biomarkers for (developmental) neurotoxicity. Although detailed analysis of predictivity is pending, zebrafish larvae have shown to be sensitive for neurodevelopmental effects and may be used in an alternative strategy to prioritize compounds for further in vivo testing.
Effects of exposure to chemicals during neurodevelopment were studied in rodent models, either in mice exposed neonatally or in rats exposed during gestation and lactation. In these studies, rat dams were exposed to chemicals during gestation and lactation. The offspring was submitted to an extensive set of tests for (effects on) behaviour and cognitive and motor functions (each animal was submitted to all tests). Besides this extensive testing paradigm a number of innovative methods were explored. These include the use of EEG as a non-invasive method to detect neurotoxicity was explored, as well as the use of stochastic microsensors to detect neurotransmitters and cytokines in brain dialysate from freely moving rats.
The electroencephalogram (EEG) is an electrophysiological measure of brain neuronal activity and could therefore be used to detect effects of chemicals on neuronal activity associated with neurological alterations. Possible alterations in EEG after developmental exposure to biocides have not been investigated previously. EEG analysis has the potential to reveal alterations in neuronal activity or effects on the sleep-wake cycle (which is a feature of developmental disorders such as ADHD or ADS).
In vivo brain microdialysis was used to obtain extracellular fluid from different cerebral regions, in which levels of different neurotransmitters and metabolites can be measured. Different neurotransmitters were measured by HPLC although the number of analyses is limited by the sample volume. This limitation has been overcome by the use of stochastic microsensors that have been developed in DENAMIC. These sensors allow the simultaneous measurements of several neurotransmitters and cytokines in a very small amount of sample. In DENAMIC, different types of microsensors were developed and evaluated showing the best performance by stochastic microsensors. These sensors were also used to determine neurotransmitters in zebrafish larvae, different biological fluids from rats, and cerebrospinal fluid and cord blood from children. Very low levels of neurotransmitters can be determined in small sample volumes of different types of matrices for which sample processing is not needed.
A number of other innovative approaches were also used to explore potential biomarkers for developmental neurotoxicity. Effects were investigated on neurotransmitter profiles (LC-MS), and the expression of genes (genomics), proteins (proteomics), metabolites (metabolomics) and glycans in different experimental models. A number of selected biomarkers were evaluated for their use in humans, although their predictive value remains to be determined.
2. Neurodevelopmental outcomes in children after early life exposure and biological plausibility from experimental studies
2.1 Cohort studies
Various recent epidemiological studies have indicated that exposure to low doses of environmental biologically active contaminants during human development can have unwanted effects on cognitive development in childhood. Foetuses and children are especially vulnerable to the exposure to environmental pollutants in comparison to adults, since their organs and systems are still developing and their detoxification mechanisms are not yet fully mature. The exposure to organophosphate biocides (OP) during early development has been associated with impaired child neuropsychological development in previous studies. Because of recent concerns of cognitive and neurobehavioural effects related to biocide exposure and if proven necessary the possibility to protect future generations by regulatory measures, DENAMIC primarily focussed on possible neurotoxic effects of biocides, but also of various other potential neurotoxic compounds such as perfluorinated chemicals (PFOS and PFOA) and methyl mercury (MeHg). Exposure to these different chemicals was determined in children and mothers in different regions of Europe. Prenatal and postnatal exposure to different neurotoxic chemicals was assessed in mother-child cohorts from Norway (HUMIS), Netherlands (LINC), Slovakia (PRENATAL) and Spain (INMA, Valencia and Sabadell sub-cohorts). The associations between the pre- and postnatal exposure to neurotoxic chemicals and child neuropsychological, cognitive and behaviour development in these cohorts were explored.
Exposure to neurotoxic chemicals
Several factors with an influence on exposure levels were observed. The cohorts studied were from different countries with different food and life style habits, which may influence risks of abnormal neurodevelopment. For example, fish consumption in Norway and meat consumption in Slovakia is higher than in the other studied countries. Another example is the distance of people from farm fields.
Urine from mothers and children was investigated on the occurrence of OP metabolites, and carbamates and metabolites. A large number of breast milk samples were analysed for chlorpyrifos, carbamates and metabolites, pyrethroids, organochlorine biocides, PCBs, PBDEs, perfluoroalkyl substances (PFASs) and MeHg. Additionally, in cord blood PFASs, PCBs, organochlorines and mercury were determined.
A limited number of OP and carbamates were detected in urine, and the detection frequencies in urine were low with only three OP metabolites (DMTP, DAP, TCPY) above 30%, and only two carbamates (aldicarb sulfone and 3-hydroxy carbofuran) above 20%. The urine biomonitoring studies showed that difference were observed in the levels and patterns between Spain and Slovakia. The sum of the concentration of OP metabolites in Valencia (25 µg/g creatinine) was higher than in Sabadell and Prenatal (Slovakia; 16 and 13 µg/g creatinine, respectively). The levels of OPs were lower than found in previously reported birth cohorts. An association between OP levels and season of sampling was found (higher levels in summer or fall), and in one cohort the residential proximity to fields was associated with higher OP levels. An association between maternal fruit and vegetable intake was found in the Spanish cohort.
In breast milk carbamates were for the first time ever reported, but the frequency of detection was low (<16%), and the levels were lower than chlorpyrifos. Chlorpyrifos and pyrethroids were for the first time reported in European breast milk; globally only three studies are available. Interestingly, chlorpyrifos was found in most of the breast milk samples from The Netherlands and Slovakia, but was almost absent in Norway. The highest levels were found in Slovakia (median 44 ng/kg ww, maximum 452 ng/kg ww), and lower levels in The Netherlands (median 18 ng/kg ww, maximum 78 ng/kg ww). We found evidence that these non-persistent chemicals such as chlorpyrifos, cypermethrin and permethrin can be found in European breast milk. This indicates that people are exposed either incidentally/episodically, or continuously, perhaps due to indoor use of biocides.
The mean level of methylmercury in breast milk from Norway is 39 ng/kg milk and twice as high as found in the Netherlands and Slovakia. This higher exposure is probably due to the higher consumption of fish in Norway as fish is an important source of methylmercury. Higher consumption of fish in Norway is probably also linked to the higher mean PCB levels in the breast milk from Norway (60 ng/g lw) compared to the Netherlands (39 ng/g lw). In Slovakia the PCB levels were even higher which is probably related to the production of PCB formulations in the former Czechoslovakia. Fortunately levels have been decreasing in the last decades.
The levels of organochlorine biocides, especially DDTs, are much higher in Slovakia than in Norway and The Netherlands. As the ratio of 4,4’-DDE and 4,4’-DDT is lower in Slovakia (25) than in the other countries (35 and 29, respectively) this indicates that DDT exposure was reduced the earliest in the Netherlands, followed by Norway and finally Slovakia.
PFOS and PFOA were detected at high frequency and at relative high levels in cord blood (µg/l range). Lower levels of PFOS and PFOA were found in breast milk (ng/l range); nevertheless PFOA and PFOS were present in all breast milk samples. Median levels of PFOS in breast milk from Norway were 2-5 times higher than in The Netherlands and Slovakia, but levels of PFOA were similar between the countries. The presence of elevated levels of PFOA in some human milk samples from Norway have been suggested to be related to the use of ski wax materials that can contain PFOA.
In general, we can conclude that contaminant patterns and levels are different between Norway, Netherlands and Slovakia. The general public in these countries are therefore exposed to a different mixture of contaminants with a different risk for adverse effects on brain development.
Associations between exposure and neurodevelopmental outcomes
Combining information on measured exposures and neurodevelopmental outcomes in the four cohorts showed that for some contaminants associations were found, for some not, and for some results were less consistent. Each cohort collected a wide range of information on subjects at different time points, using various questionnaires – general, environmental, and food frequency questionnaires. General questionnaires were focused mainly on socio-demographic data, smoking and alcohol exposure, maternal and family history, children’s health and breastfeeding information. Environmental questionnaires were aimed at the household characteristics and environmental exposures to chemicals. Also characteristics of birth (delivery) were reported as well as basic anthropometric measurements in all cohorts.
A meta-analysis of seven European cohorts was conducted to study the effects of persistent organic pollutants on several areas of neuropsychological development (cognitive delay, ADHD, and ASD). Association between exposure levels of organochlorine and bromine compounds (PCB-153, HCB, p,p'-DDE, PFOS, PFOA, PBDE47, PBDE153) and neuropsychological outcomes were investigated. This study showed that no associations were found with any of these compounds. Exploratory studies in the LINC cohort, using stratified data for gender, found positive associations between exposure to PCBs and some organochlorine biocides and ADHD symptoms, externalizing behaviour and aggressive behaviour in girls.
Using novel statistical methods for the analysis of multiple persistent organic pollutants (PCBs, DDT, DDE, HCB, HCH, oxychlordan, PBDEs) in breast milk of the HUMIS cohort, it was demonstrated that p,p’-DDT showed a positive association with increased behavioural problems in children at the age of 12 months. These associations were no longer found at 24 months of age. Exploratory studies at the LINC cohort showed a weak association between prenatal exposure to PCBs and behavioural problems and ADHD/ASD symptoms. Postnatal exposure (via breast milk) showed positive associations between PCBs and ADHD symptoms, externalizing behaviour and aggressive behaviour. Associations were sex-specific.
The cohorts observed no increased risk of abnormal neuropsychological and behavioural development in association with perinatal exposure to PFOS, PFOA. However, exploratory results of the LINC cohort showed an adverse association between PFOA and externalizing behaviour in boys. Results of PFAS associations are less consistent, as of the eight published studies on the association between neuropsychological development and PFOS or PFOA, only two found an increased risk.
Various health effects have been reported associated with pre- and postnatal exposures to low concentrations of organophosphates (OP) in urban and agricultural communities, in which an increased risk of nervous system disorders in children and effects in the neurobehavioural development have been shown. The DENAMIC cohorts did not find associations between OP levels and neuropsychological outcomes (ADHD, ASD, cognitive delay). The absence of the association can be due to the lower levels of OPs found in our study compared to North American studies were in general higher levels of OPs are found.
2.2. Effects observed in animal models
Many different biocides and other environmental pollutants were included in the experimental studies in DENAMIC. After the first phase of the project, a number of reference chemicals were selected to be included in all studies to allow evaluation of the effects in different animal models and cellular systems: chlorpyrifos(-oxon), carbaryl, α-cypermethrin, endosulfan and methylmercury. Rodent models were used to study effects of exposure to these chemicals during critical window of neurodevelopment, mice that were exposed neonatally and rats exposed during gestation and lactation. In these studies effects of developmental exposure were studied on different parameters of spontaneous behaviour (mice and rats) as an apical test of neuronal function, habituation to a novel home environment (mice) which is the integrations of sensory input into motor output, circadian rhythm (rats), anxiety (rats), spatial learning (mice and rats), active avoidance (rats) and motor coordination (rats). In particular spontaneous activity and tests on cognitive functions are of interest as hyperactivity and impaired working memory have been described as symptoms in ADHD patients.
Studies in the neonatal mouse model revealed that exposure to MeHg, endosulfan, cypermethrin, chlorpyrifos or carbaryl can induce developmental neurotoxic effects, when administered as a single oral dose during a defined critical period of brain development. The animals showed distinct neurotoxic effects manifested as persistent aberrant spontaneous behaviour (hyperactivity) in a novel home environment and reduced ability to habituate, which also indicates impaired cognitive function as habituation is the integration of sensory information into motoric output. Exposure to chemicals during a critical period of neonatal brain development can thus be involved in induction of neurofunctional disorders related to ADHD (attention deficit hyperactive disorder) or reduced cognitive capacity. Possible effects on cholinergic susceptibility and neuroprotein levels were investigated as underlying mechanisms for the observed neurobehavioural effects in (young) adults. The studied proteins are involved in essential processes of the developing brain, such as dendritic arborisation, axonal outgrowth, neuronal connections and synaptogenesis. Disturbances of these processes can affect the development of specific neurotransmitter systems as well as the development of the cytoarchitecture of the brain, possibly leading to functional defects later in life and may thus have contributed to the observed developmental neurobehavioural aberrations. Altered levels of neuroproteins may serve as potential markers for developmental neurotoxicity although due to the dynamic and compensatory mechanisms of the developing brain, it is unlikely that a single differentially expressed protein can be used to predict functional neurotoxic effects.
Effects of exposure during gestation and lactation to MeHg, endosulfan, cypermethrin, chlorpyrifos or carbaryl were also studied in rats using a large amount of behavioural tests described above. Different effects of exposure to the reference chemicals during gestation and lactation were observed in the large array of neurobehavioural tests for behaviour, cognition and motor function. Spatial learning was impaired by all biocides, except carbaryl, in male offspring, but not in female offspring. In detail, spatial learning assessed in the Morris water maze is impaired after exposure to endosulfan and chlorpyrifos in male rats, while α-cypermethrin improves spatial learning. In the radial maze spatial learning is impaired after exposure to endosulfan, chlorpyrifos and cypermethrin in male rats. Spatial working memory was impaired in male rats exposed to chlorpyrifos, while spatial learning and cognitive functions appeared to be improved in female rats exposed to chlorpyrifos and α-cypermethrin. Carbaryl does not affect spatial learning in rats. Underlying mechanisms were studied using the measurement of neurotransmitters in brain dialysate and cytokines in brain tissues as a measure for neuroinflammation as this is associated with impaired spatial learning. All biocides induced neuroinflammation, increasing pro-inflammatory IL-1 β content in hippocampus correlated to spatial learning deficits with increased sensitivity of females for this mechanism. Impaired working memory has also been described to occur in ADHD patients. Motor coordination was impaired after exposure to endosulfan, cypermethrin or chlorpyrifos in female rats. There were also effects on motor function of male offspring that were exposed to chlorpyrifos or carbaryl. The impaired motor function of female offspring exposed during gestation and lactation to endosulfan or α-cypermethrin and of male offspring exposed to chlorpyrifos is associated with increased extracellular levels of the inhibitory neurotransmitter GABA in the cerebellum. Motor activity was increased after exposure to chlorpyrifos which is associated to an increased content of NMDA receptor subunits in the hippocampus. Alterations in NMDA receptor function have also been observed in ADHD animal models. Endosulfan is the only biocide inducing anxiety in males. Exposure to endosulfan, α-cypermethrin or chlorpyrifos also altered circadian rhythm of motor activity (hyperactivity) during the sleeping period in males.
In these studies with rats clear sex-differentiated effects resulting from exposure to the reference biocides were observed. This is important with regard to the study of (adverse) health outcome with sex differentiated prevalence such as neurodevelopmental disorders. Such sex specificity has been observed both in cohort studies as in animal models.
In the last phase of the project, a specific study on the chlorpyrifos was performed in a collaboration of different laboratories working in DENAMIC. In this study a neurobehavioural test with mice was included which measures attentional processing. With this test it is possible to study deficits in attention and impulsivity, which are behavioural traits that are analogous to symptoms in ADHD. It was shown in this study that neonatal exposure to chlorpyrifos impairs attention. A number of additional endpoints, including hippocampal synaptic plasticity, AChE activity in plasma and metabolomics were also included to explore underlying mechanisms of the observed neurobehavioural effect.
Experimental studies in DENAMIC and literature demonstrating neurodevelopmental effects by exposure to chlorpyrifos, PCBs, PFAS provide evidence for the biological plausibility of the causal relationship in the observed associations with exposure to these chemicals in the cohort studies, while there is less experimental evidence for developmental neurotoxicity for organochlorines.
2.3. Effects observed in cellular models and zebrafish larvae
Effects of selected biocides and environmental chemicals on morphological and neurophysiological development at non-cytotoxic exposure concentrations were investigated in cellular models. Effects of exposure on the number of neurons, neurites and synapses in mouse cortical cultures using high-content imaging, on calcium homeostasis and signalling in PC12 cells and brain microsomes using fluorescent calcium imaging and on electrical activity (development) of rat cortical cultures were investigated. In addition, behavioural toxicity in zebrafish larvae was used as an alternative in vitro testing system with intact animals. Chemical-induced effects on behaviour of zebrafish larvae were tested at concentrations that not induce embryo toxicity (scored morphologically).
Effects on number of neurons, neurites and synapses in mouse cortical cultures as measures for network development were studied in mouse primary cortical cultures after 14 days of in vitro culture that were treated at the 2nd day in vitro. Over 30 chemicals were tested in this system. No effects or only at the high (10 µM) exposure concentrations were observed for many of the chemicals including chlorpyrifos, chlorpyrifos-oxon, carbaryl and α-cypermethrin. For endosulfan and methylmercury, effects were observed at lower exposure concentrations, as well as for paraoxon-ethyl, paraquat, and 6-OH-BDE-47. Evaluation of the predictive value for developmental neurotoxicity with this system by comparing the identified developmental neurotoxicants in this system with a list of known developmental neurotoxicants demonstrated that approximately half of the known developmental neurotoxicants were predicted with this system.
Effects of acute exposure to insecticides on basal and depolarization-evoked intracellular calcium levels were studied in vitro in rat pheochromacytoma PC12 cells. The data demonstrate that α-cypermethrin, endosulfan, chlorpyrifos and methylmercury concentration-dependently decreased the depolarization-evoked intracellular calcium levels. A particularly interesting observation was that chlorpyrifos-oxon was much less potent compared to parent compound chlorpyrifos. Modest increases in basal intracellular calcium levels were observed for chlorpyrifos, chlorpyrifos-oxon or endosulfan at relatively high concentrations (10 µM) and at lower concentrations for methylmercury. No effects were observed for carbaryl even at a high concentration. Also the effects of subchronic and repeated exposure to the reference biocides on calcium homeostasis were investigated. Except for carbaryl, all selected insecticides inhibited depolarization-evoked intracellular calcium levels after subchronic exposure in PC12 cells while repeated exposure augmented the inhibition of the depolarization-evoked intracellular calcium levels induced by subchronic exposure to cypermethrin, chlorpyrifos, chlorpyrifos-oxon and endosulfan. These data show that calcium homeostasis in PC12 cells can be used as a sensitive readout for insecticide-induced neurotoxicity. From the studies using different exposure paradigms it is concluded that investigation of effects of acute exposure to chemicals is generally sufficient for rapid detection of insecticide (or chemical)-induced inhibition of VGCCs. However, some differences in potency of insecticides for inhibition of VGCCs were observed between the exposure scenarios, which argue for the inclusion of repeated exposure conditions for specific studies. Effects of chemicals on specific mechanisms calcium homeostasis were also investigated using brain microsomes. These studies showed that endosulfan and chlorpyrifos induce ryanodine receptor mediated calcium release from rat brain microsomes to a similar extent as positive control PCB-95. VGCC inhibition in PC12 cells reflected the developmental neurotoxicity of the reference chemicals in DENAMIC although further evaluation of predictivity of this model is pending.
Effects of acute and chronic exposure to insecticides and methylmercury on electrical activity were studied in primary cortical cultures grown on multi-well multi-electrode arrays (MEAs). Concentration-dependent inhibition of electrical activity (spike rate) was observed for acute exposure to methylmercury, chlorpyrifos and its oxon metabolite chlorpyrifos-oxon, α-cypermethrin and carbaryl. Acute exposure to endosulfan resulted in a biphasic response, showing a concentration-dependent increase followed by complete inhibition at the highest exposure concentrations. Lowest observed effect concentrations were in the (low) micromolar range. Chronic exposure to methylmercury, chlorpyrifos and a-cypermethrin inhibited the development of electrical activity, while exposure to endosulfan increased the activity. No significant effects were observed by chronic exposure to chlorpyrifos-oxon and carbaryl. The inhibition of spike activity in primary cortical cultures on MEAs by carbaryl, chlorpyrifos-oxon and methylmercury was observed at lower lowest observed effect concentrations in comparison with earlier studies. These differences may be related to interlaboratory differences in model characteristics, age of the cultures used for testing, (low) numbers of replicates, and differences in exposure paradigms (naïve culture for each replicate or cumulative dosing). The observed modulation of electrical activity by biocides and methylmercury confirms the known potential for developmental neurotoxicity of these chemicals and illustrates the applicability of heterogeneous cortical cultures grown on MEAs for screening purposes.
Also a number of tests were performed to evaluate the potential of chemicals to act via specific mechanisms that may underlie developmental neurotoxicity in vivo. The effects of over 30 chemicals were tested with regard to binding to thyroid hormone binding protein transthyretin (TTR). Thyroid hormones are critical to brain development and disruption has been demonstrated to result in impaired cognitive function. Competitive binding to TTR was observed for only a few of the tested chemicals, showing effects of chlorpyrifos-oxon, 6-OH-BDE47 and PFAS. Another well-known mechanism in particular for organophosphates and carbamates is the inhibition of AChE activity which may result in overactivation of the nervous system. Also in this system over 30 chemicals were tested confirming that in particular oxon metabolites of organophosphates and carbamates inhibit AChE.
Rapid screening using zebrafish larvae was used to determine effects of exposure on spontaneous behaviour and heart beat at exposure concentration ranges lower than those resulting in embryo toxicity. Over 30 chemicals were tested in this system. For several chemicals, including chlorpyrifos-oxon, α-cypermethrin, endosulfan and methylmercury, effects on spontaneous behaviour in zebrafish larvae was observed at low concentrations (<1 µM). Evaluation of the predictive value developmental neurotoxicity with this system by comparing the chemicals inducing developmental neurotoxicity in zebrafish larvae with a list of known mammalian developmental neurotoxicants demonstrated that the majority of known developmental neurotoxicants were predicted with this system.
Several of the applied in vitro systems in DENAMIC correctly predict neurodevelopmental effects. However, additional data is required to firmly establish their predictive power. The DENAMIC project has contributed to the body of knowledge needed for optimal comparison and evaluation of the predictive power of in vitro systems by 1) collecting data on the effects of chemicals and 2) characterization of the physiological relevance of in vitro systems and the measured parameters. As soon as the predictive power of an in vitro system is established it can be used to build an Adverse Outcome Pathway (AOP), in which Molecular Initiating Events and/or Key Events investigated in in vitro systems can be used to predict an ultimate Adverse Outcome in individuals and/or populations. An example of a Molecular Initiating Event could be inhibitory interactions of chemicals with voltage-gated calcium channels. Nevertheless it is currently still unclear which specific organ, organism and population responses an inhibition of VGCCs by chemicals may cause. Primary cortical cultures grown on MEAs can be regarded an integrated system in which effects may be caused by interactions of chemicals with one or more of the many molecular targets present in this system. Electrical (neuronal) activity can thus be regarded as a key event, as this is the inherent function of neurons in networks. In contrast, the neonatal animal model as such does not fit into a distinct AOP for DNT as similar effects are observed for different types of chemicals. Nevertheless, this model has proven to detect DNT effects from widely different compounds with different acute modes of action with delayed observable functional effects. The assays (and models) that are proposed to be used in experiments based on an AOP should be well-established and reliably (reproducibly) give mechanistic information on the effects of chemicals on key events. When generating experimental data to support or build such an AOP, exposure scenarios should also reflect what is relevant for the human exposure situation with regard to exposure levels, exposure timing and exposure duration. To establish AOPs for adverse cognitive outcomes associated with neurotoxicity the pathophysiology of (subclinical) neurodevelopmental disorders and the role of exposure to chemicals should also be further clarified.
2.4. Mechanistic pathways and biomarkers using molecular (omics) studies
The aim of the Omics (genomics, proteomics, glycan analysis and metabolomics) studies was to 1) investigate underlying molecular mechanisms of observed effects on behaviour and cognitive function resulting from exposure to biocides in animal models, and 2) develop possible biomarkers for (developmental) neurotoxicity and cognitive disorders in humans.
In developmental exposure studies, altered regulation of proteins and metabolites by exposure was studied in different parts of rat brain (hippocampus, cortex, cerebellum, and striatum), mouse brain (hippocampus, cortex) and whole brains of zebrafish larvae. More than 100 regulated proteins were observed in rats exposed to biocides, and proteins were regulated differently in the four brain areas. Examples of regulated protein pathways in rats exposed to biocides were glutamate processing, DNA repair, tricarboxylic acid cycle, mitochondrial transport, and sodium/potassium-transporting. In rats exposed to α-cypermethrin, the main effects were seen in striatum, and less in hippocampus, cortex, or cerebellum. Like the effects on neurobehaviour and cognition, also these effects on protein regulation were sex-specific and showed in some cases opposite regulation between sexes. A smaller effect on protein expression in striatum was found in rats exposed to the mixture of α-cypermethrin and endosulfan.
Also in the neonatal mouse model, in which mice were exposed to only a single dose at postnatal day 10, proteins were observed to be regulated, although to a limited extent (low fold-changes), after exposure to biocides, mainly in hippocampus. In mice exposed to chlorpyrifos, carbaryl, or endosulfan, a total of 689 brain proteins were identified of which 67 were significantly regulated. For chlorpyrifos the highest number of regulated proteins was identified in cortex. Pathway analysis identified that chlorpyrifos-induced changes were observed in the cluster of mitochondrial/organelle proteins and neuronal diseases like Alzheimer’s, Parkinson’s and Huntington’s disease. For PFHxS and endosulfan also the strongest regulation was observed in hippocampus and the regulated proteins were related to pathways of neurofilament and intermediate filament cytoskeleton.
Metabolomics studies in the brains of rodents exposed during neurodevelopment focused on four major neurotransmitter pathways, but also untargeted metabolomics studies were carried out. Untargeted metabolomics revealed 400-600 metabolites in the rat brain tissues of which more than 50 were significantly regulated in rats exposed to biocides. In the mouse studies only a limited number of metabolites were regulated. Pathway mapping showed that various metabolomic pathways were affected in rats including those related to serotoninergic, dopaminergic, cholinergic, GABAergic neurotransmitters, sodium/chloride dependent neurotransmitter transporters, transport of nucleosides, and purine metabolism. Also these effects of biocide exposure on metabolite levels of were brain tissue and sex-specific. An example of such a rat brain region specific effect is the upregulation of choline and acetylcholine in striatum only to endosulfan, indicating that there are synapse-related effects. Metabolites were very differently regulated between the compounds studied. As an example α-cypermethrin exposure showed mainly up-regulated metabolites while exposure to endosulfan or chlorpyrifos showed both up- and down-regulated metabolites in brain tissues.
The cognitive and behaviour tests with rats showed that developmental exposure to biocides had effects on anxiety, spatial learning, motor coordination, motor activity, active avoidance, and working memory. These effects were biocide specific and some effects were sex-specific. Metabolomics studies (down and upregulated metabolites) give mechanistic information that may be related to the observed neurobehavioural and cognitive effects. For example, developmental exposure to endosulfan caused a decrease in the neurotransmitter GABA in males (hippocampus), but not in females. This down regulation of GABA in males correlated well to the impairment of spatial learning that was observed in male but not in female rats. Rats exposed to biocides also showed increased levels of cytokines (interleukin) in the hippocampus of males only, suggesting a role of biocides neuroinflammation. The increase of cytokines may very well cause spatial learning and memory impairments. The metabolite succinate can be used as a marker of inflammation, as it is one of the metabolites of the tricarboxylic acid (TCA) cycle. It also plays a role in the GABA shunt pathway. The level of succinate in hippocampus was indeed increased in male offspring after biocide exposure, and levels of GABA and glutamate were increased as well. The increase of succinate may be due to the increased levels of GABA and glutamate available for the GABA shunt pathway. Succinate can stimulate dendritic cells but has also been associated with inflammation and is a metabolite related to innate immune signalling. Cypermethrin exposed rats showed effects on the dopamine pathway, showing effects on the levels of dopamine precursors and metabolites. Dopamine is involved in motor control which was shown to be impaired in female offspring after developmental exposure to α-cypermethrin.
In conclusion, proteomics and metabolomics studies elucidated some of the mechanisms responsible for the observed behaviour and cognitive effects due to chemical exposure. These studies also showed that chemical exposure can results in brain tissue and sex-specific metabolic pathway changes.
Before protein, peptide, glycan and metabolite biomarkers as found in the cell and animal models can be used non-invasively in humans, it needs to be determined whether these can be determined in human tissues. With regard to effects on the developing brain, human cord blood was chosen as the most appropriate non-invasive exposure matrix. Proteomic studies selected a number of candidate biomarkers of chemical exposure related to (developmental) neurotoxicity from the cell and animal models (rat and mice). These candidate biomarkers were verified in human cord blood from three DENAMIC cohorts (LINC, PRENATAL and HUMIS). For this study a novel approach based on a combination of human PC12 or SH-SY5Y cells and a protein labelling method was developed.
DENAMIC identified neuroproteins as potential markers for developmental neurotoxicity. Neuroproteins are involved in essential processes of the developing brain, such as dendritic arborisation, axonal outgrowth, neuronal connections and synaptogenesis. A potential biomarker for neuroinflammation was the level of IL-1β in hippocampus that was related to biocide exposure. Besides these effects on neuroprotein expression, the DENAMIC animal and cell based studies also showed that various neurotransmitter pathways were affected by chemical exposure, and some neurotransmitters are therefore potential markers of exposure to developmental neurotoxicants. An imbalance in the serotonergic and dopaminergic neurotransmitter systems has been suggested to be responsible for hyperactive behaviour in humans. Cord blood from the Dutch mother-child cohort LINC was therefore used to investigate the relationship between neurotransmitters and chemical exposure. This study showed that PFOA and DEHP metabolites were associated with changes in the levels of GABA, acetylcholine, and epinephrine. Whether these changes in neurotransmitter levels are related to abnormal behavioural development, needs to be further explored.
3. Mixture toxicity
Humans are exposed to mixtures of chemicals, either simultaneously or in succession, chronically or repetitively. In the DENAMIC project, therefore also studies on the effects of combined exposure to two or more chemicals were included. Exposure to (mixtures of) chemicals is lifestyle and region dependent and consequently the combined exposure situation can vary strongly. A number of different (types) of mixtures were decided on to streamline the combined exposures tested in the DENAMIC experimental studies on developmental neurotoxicity. Binary mixtures were composed based on 1) occurrence and persistence in humans and 2) differences or similarities in the mechanism of action of the individual constituents. Two types of mechanisms with regard to effect resulting from combined exposure were focused on in the experimental studies with animal and cellular models. Firstly, it was evaluated whether additivity occurs in combined exposure situations. Secondly, it was investigated whether combined low-level exposure to chemical that do not show toxicity on an individual basis results in effects.
Three different binary mixtures were tested in the neonatal mouse model: chlorpyrifos in combination with carbaryl, chlorpyrifos in combination with methylmercury, and PFAS perfluorohexane sulfonate in combination with endosulfan. Neonatal mice were exposed to a single oral dose of the binary mixture and the individual constituents on postnatal day 10. Effects on spontaneous behaviour in a novel home environment and neuroprotein expression were tested at the ages of 2 and 4 months. It was demonstrated that in case of combined exposure interaction of the chemicals resulted in exacerbation of developmental neurotoxic effects in a dose-responsive manner. Observed effects included hyperactivity and impaired habituation as a measure of cognitive function, and these effects were also observed in case of combined exposure at dose were the individual constituents did not cause any effect. Co-exposure to chlorpyrifos and carbaryl did also result in an increased susceptibility of the cholinergic system in adults, as well as in changed neuroprotein levels. Persistent developmental effects in the neonatal mouse model can thus be induced by various persistent and non-persistent chemicals, either individually or in low-level mixtures, by exposure during postnatal day 10 which is a defined critical period in brain development. The behavioural effects observed after exposure to chlorpyrifos in combination with methylmercury, and to PFAS perfluorohexane sulfonate in combination with endosulfan appeared close to additive, despite different mechanisms of action of the individual constituents. In case of exposure to chlorpyrifos in combination with carbaryl, the neurobehaviour effects were exacerbated in a synergistic manner, with chlorpyrifos exposure as the main determinant.
Rats were exposed to a mixture of endosulfan and α-cypermethrin during gestation and lactation. Male offspring demonstrated hyperactivity (associated with decreased NMDA expression in the hippocampus) and their spatial and working memory was impaired by potentially synergistic interactions between the individual constituents, as both of these did not have this effect. This was however not observed for every neurobehavioural parameter. For example, both endosulfan and α-cypermethrin impaired motor coordination in female offspring individually, while this was not observed in animals exposed to the mixture. Impairment of reference memory in male offspring was induced by both of these chemicals, as well as by the mixture.
Several binary (and one tertiary) mixtures of insecticides were investigated for their acute effects on inhibition of VGCCs. The effects of the mixtures were compared with the effects of the individual compounds on inhibition of VGCCs to evaluate potential additivity. In the first phase of the project, LOECs were selected from concentration-response curves of the different insecticides and combined in binary mixtures. It was demonstrated that exposure to chlorpyrifos in combination with either its oxon-analogue, methylmercury or with parathion-ethyl did not increase the degree of inhibition in comparison with the inhibition caused by the individual compounds. In the next studies, exposure levels resulting in 20% inhibition were interpolated from concentration-response curves and used to calculate an expected additivity effect range. For every exposure to a binary mixture (chlorpyrifos in combination with endosulfan, chlorpyrifos in combination with cypermethrin, cypermethrin in combination with endosulfan), the effects of the mixture were within the expected additivity effect range. The observed additivity was however less equivocal for the tertiary mixture of chlorpyrifos, cypermethrin and endosulfan.
Effects of acute exposure to binary mixtures were also tested on electrical activity of primary cortical cultures grown on MEAs. It was concluded that the degree of variation in this type of data precludes the use of an expected additivity effect range to exclude additivity. Therefore, the effect of a mixture was compared with the effects of its individual constituents. It was demonstrated that exposure to chlorpyrifos in combination with methylmercury concentration-dependently inhibited electrical activity at concentrations that did not have significant effects of the individual constituents. Endosulfan and a-cypermethrin have opposite effects at low exposure concentrations on the electrical activity in primary cortical cultures grown on MEAs. Co-exposures to endosulfan (inducing concentration-dependent increases in electrical activity) and cypermethrin (inducing concentration-dependent decreases in electrical activity) demonstrated that the inhibition by cypermethrin was dominating the effect of the binary mixture.
A detailed analysis was obtained for the interaction of chlorpyrifos-oxon and carbaryl with regard to AChE inhibition, a neurotoxic mechanism in particular specific for biocides. When testing binary mixtures at different potency ratios of the individual compounds, effective concentrations resulting in 50% inhibition were somewhat lower than expected in case of additivity. A similar analysis was performed for the interaction of PFAS perfluorooctanesulfonic acid and 6-OH-BDE47 with regard to TTR binding. Dose-response curves of these three mixtures indicate that the combined TTR-binding potency of the two test compounds is well described by concentration addition.
In vitro and in vivo studies performed in the DENAMIC project thus clearly indicate that exposure to combinations of chemicals may result in non-additive effects depending on levels and compounds. Effects of mixtures have been shown to be larger or smaller compared to the expected combined effect of the individual constituents based on additivity. In a number of experiments it has even been shown that combined exposure to chemicals caused developmental neurotoxicity while the exposure levels of the individual chemicals did not result in any effects.
Finally, extracts were prepared of breast milk from various DENAMIC cohorts (LINC, HUMIS, PRENATAL). These extracts contained a complex mixture of real-life chemicals and were tested in three different bioassays, i.e. the acetylcholinesterase (AChE) inhibition assay, the transthyretin (TTR) binding assay, and the endoplasmatic reticulum (ER) calcium-release assay. AChE inhibiting potency was observed in several breast milk extracts. Out of 17 individual breast milk samples from the HUMIS and LINC cohort, highest AChE inhibiting potency was observed for a LINC sample corresponding to 4.4 ng dichlorvos-equivalent per g milk. So far, the observed inhibition could not be explained by target analyzed AChE inhibiting insecticides: all carbamate concentrations were
4. Risk assessment and recommendations on the use of neurotoxic chemicals
The assessment for risk of developmental neurotoxicity in humans was based on the uptake of contaminants from breast milk (postnatal exposure) or (cord) blood levels (prenatal exposure) placed into the context of effective concentrations from the in vivo and in vitro DENAMIC experimental data.
For the in vivo studies the no observed effect levels (NOELs) of the DENAMIC reference chemicals in the neurobehavioural mouse studies were used and compared to the uptake of contaminants by breast milk. These dose – response relationships are relevant for human risk assessment of (developmental) neurotoxicological and neurobehavioural effects. The no-effect doses are used as points-of-departure for the human risk assessment for (developmental) neurotoxicity.
As the mouse studies are single dose experiments, albeit at a postnatal period in which brain development in the mouse is at a critical phase, a number of uncertainty factors (UFs) were applied. Besides the default factors of 10 for interspecies and interhuman variability, an additional safety factor 3 was applied for extrapolating a single dose regime to a chronic exposure. As a result a total composite UF of 10*10*3=300 was applied for the risk assessment of the compounds tested in DENAMIC with these in vivo experiments. In combination with the NOELs from these neurobehavioural mouse studies an interim reference dose (i-RfD) was calculated for the DENAMIC reference compounds. It should be noted that if this i-RfD is used for risk assessment no further safety factors are needed when comparing this with uptake via human milk, because the i-RfD already includes uncertainty factors. Furthermore, for calculations of exposures of the breast fed infant it was assumed that milk consumption is 125 g (ml)/kg bw/day.
For the risk assessment based on the in vitro studies the medium concentrations used in the in vitro experiments resulting in effects on various neurotoxicity endpoints were compared with concentrations in human tissues, in particular human cord blood. For this approach it was necessary to differentiate in uncertainty factors for toxicokinetics and toxicodynamics and the differentiation thereof in interspecies and interindividual uncertainty factors. For risk assessment done based on the results from in vitro experiments and internal exposure levels in plasma the interspecies variability in toxicokinetics was assumed 1. As a result a composite uncertainty factor of 2.5*10=25 was used under these conditions, which was used to calculate an interim reference concentration (i-RfC) for neurotoxicity. It should be noted that if this i-RfC is used for risk assessment no further safety factors are needed when comparing this with internal levels e.g. blood, because this i-RfC already includes uncertainty factors.
The risk assessment of chlorpyrifos indicated that based on in vivo, in vitro and human exposure data from DENAMIC in combination with those from literature on average the margin of safety is likely adequate with a possible exception for high exposure groups. Based on the experimental results from the DENAMIC project no further reductions in human exposure to chlorpyrifos are needed with possible exception for high exposure groups. For carbaryl the in vivo, in vitro and human exposure data from DENAMIC indicated that the margin of safety is more than sufficient and no effects are expected with regard to neurodevelopmental toxicity. No further reductions in human exposure to carbaryl are needed in the EU. Similarly, cypermethrin in vivo and human exposure data from DENAMIC showed that the margin of safety is more than sufficient and no effects are expected. Also for this compound no further reductions in human exposure to cypermethrin are needed in the EU. For endosulfan the in vitro studies and human exposure data from DENAMIC showed that the margin of safety may not be adequate and that therefore neurodevelopmental toxicity cannot be excluded in the EU population. Based on these experimental results from the DENAMIC project further reductions in human exposure to endosulfan are needed in the EU, but as this biocide is already banned in the EU, practical reduction may be difficult to achieve. Also for methylmercury the in vitro studies and human exposure data from DENAMIC concluded that exposure of methylmercury is several orders of magnitude above a calculated i-RfC. Under the assumption that the in vitro medium concentrations of methylmercury are a good reflection of the internal blood levels the i-RfC is several orders of magnitude lower than the internal exposure levels in blood, which indicates significant health risks for neurotoxic effects at present concentrations in some EU countries. As a result, neurotoxic effects based on internal blood levels cannot be excluded in the EU. Based on these experimental results from the DENAMIC project further reductions in human exposure to methylmercury are needed in the EU. A reduction may be difficult as this will need a change in food habits as a large part of the uptake of methyl mercury is by seafood. There are also other food items that can be a source of methylmercury, for example rice.
In vitro and in vivo studies performed in the DENAMIC project clearly indicate that combinations of biocides may give non-additive effects depending on the dose levels and compounds showing that effects of mixtures have been shown to be larger or smaller compared to the expected combined effect of the individual constituents based on additivity. In a number of experiments it has even been shown that combined exposure to chemicals caused developmental neurotoxicity while the exposure levels of the individual chemicals did not result in any effects. This observation in particular brings doubt in the present approach for risk assessment, which is based on exposure to individual biocides and may in fact result in an underestimation of the (human) risk.
It is thus clear that additivity as a default approach in the human risk assessment of biocides cannot a priori be assumed and an additional uncertainty factor for potential mixture effects is proposed. This uncertainty is founded in particular in observations of non-additive mixture effects and sex-specific toxicity. Based on the in vitro mixture neurotoxicity experiments it was concluded that a deviation of a factor 2 from additivity may in fact occur. If the results from the in vivo mouse and rat mixture studies are combined, the uncertainty of non-additive effects appears to be in range between 2 and 5. The application of such an additional safety factor did not change the main conclusions in the risk assessment for the contaminants mentioned above.
5. General conclusions
DENAMIC developed tools and methods to detect chemicals’ effects on the developing nervous system. More than 30 chemicals were tested individually and some in mixtures. Exposure concentrations in the in vivo models (rat, mice, zebrafish) resulting in effects were orders of magnitude lower than tested before. There are critical windows of exposure during neurodevelopment (e.g. during the brain growth spurt). The animal models showed that early-life exposure during neurodevelopment resulted in persistent effects on behaviour, cognition, and motor activity later in life.
The project has contributed to the body of knowledge needed for optimal comparison and evaluation of the predictive power of in vitro systems by 1) collecting data on the effects of chemicals and 2) characterization of the physiological relevance of in vitro systems and the measured parameters. Several of the applied in vitro systems predict neurodevelopmental effects in vivo.
In European cohorts (subclinical) effects on learning (cognitive skills) and developmental disorders in children (e.g. ADHD, autism spectrum disorders and anxiety disorders) were studied. The European population is exposed to low concentrations of neurotoxic chemicals (e.g. PCBs, organophosphate carbamates and pyrethroids biocides, PFAS, methylmercury). The exposure occurs both prenatally (via cord blood) and/or postnatally (via human milk). A number of associations were found between the exposure of these neurotoxic chemical (e.g. DDT, PCBs, PFOA) and neuropsychological and behavioural development in children. The mechanism based cell and animal studies contributed to the understanding of developmental neurotoxicity and low-dose chemical mixture exposures in children. Both the cohort studies as the animal studies showed that effects can be sex-specific. Molecular pathway analysis elucidated mechanisms for some of the observed behaviour and cognitive effects.
Neurotoxic effects of mixtures of chemicals at low concentrations were investigated in the experimental studies. DENAMIC showed that low-level exposure to chemicals that are not toxic on an individual basis can together lead to a biological or toxic effect if present in a mixture. Mixture studies in DENAMIC have indicated that non-additive interactions occur with combinations of biocides with a different mechanism of action resulting in additional uncertainty factors. Deviations of a factor 2-5 from additivity may occur. Co-exposure to a mixture could also cause an increased susceptibility in the adult animal and exacerbate developmental toxic effects that might have consequences for the induction of neurological/neurodegenerative disorders/diseases.
The risk assessment concluded that based on the experimental studies done within the DENAMIC project the margins of safety in the EU are sufficient for neurotoxic and neurobehavioural endpoints for carbaryl and cypermethrin. For chlorpyrifos, endosulfan and methylmercury such margins of safety may not be adequate in specific populations and based on either in vivo and/or in vitro experiments in DENAMIC adverse health effects cannot be excluded. It is therefore recommended to investigate whether further exposure reduction measures are needed.
Current risk assessment assumes that no effect levels are safe based on individual chemical toxicity studies. Some experiments done in the DENAMIC project indicate that combined exposure to biocides or other contaminants, that are individually present at levels not resulting in any effect, may in fact cause an effect when present in mixtures. This observation brings doubt in the present approach for risk assessment, which is based on exposure to individual chemicals and may in fact result in an underestimation of the (human) risk.
Potential Impact:
Socio-demographic and economic impact
Factors affecting the relative differences in socio-demographic and economic impact for (developmental) neurotoxicity resulting from exposure to environmental neurotoxicants were studied in DENAMIC. These factors include sex and gender differences and differences in exposure levels in different EU regions. A number of studies have been conducted on children and pregnant women evaluating the association between organophosphate biocide (OP) exposure and the socio-demographic, environmental and dietary characteristics. Socio-demographic characteristics showed that significant differences can occur between cases (ADHD, cognitive, ASD symptoms) and control groups according to the maternal education, smoking habit during pregnancy, age, and body mass index (BMI) before pregnancy. DENAMIC found that OP biocide levels in urine were associated with season of sampling, being higher in samples taken during summer and fall. Mothers who finished university studies had higher OP levels. Mother who smoked during pregnancy had lower OP levels. The residential proximity to fields was associated in some cohorts with higher OP levels. In the Spanish cohorts an inverse association between OP levels and the maternal BMI before pregnancy and a direct association with maternal intake of fruit and vegetables during pregnancy was found.
In the DENAMIC project a number of epidemiological associations between exposure to chemicals and neurodevelopmental outcomes were observed, and information on causal relationships between exposure to chemicals and (developmental) neurotoxicity was collected in experimental studies. These outcomes can be used in future or more specific calculations for the costs associated with the exposure to chemicals and associated adverse health effects. By studying social and economic aspects of families and the possible relation with developmental effects in children due to exposure to contaminants, information in DENAMIC has provided an informed basis for poverty alleviation efforts and associated health and economic benefits. In that respect the effects of this project could be stronger in Eastern-Europe as exposures may be higher and economical positions are disadvantaged compared to Western-Europe. The DENAMIC inventory of exposure of children to chemicals, can be used to equalizing current inequalities in children’s health across Europe related to the adverse health effects on cognition and neurodevelopmental disorders. These data have significant value for the EU, and for national authorities and other stakeholders involved in health sciences, environmental monitoring, and risk assessment.
There is increasing awareness that the sex of humans, animal models and cellular models matters in biomedical science. Several sex-specific associations between developmental exposure to chemicals and neurodevelopmental outcomes have been observed in DENAMIC. The sex difference in prevalence of neurodevelopmental disorders urges the inclusion of both sexes in experimental studies on the impact of exposure to chemicals on the pathophysiology of such disorders. Several sex-specific effects on neurodevelopmental outcomes have also been observed in developmentally exposed rats in DENAMIC. When sex or gender is not taken into account in experimental or epidemiological studies, important effects may be missed. Taking sex and gender into account will improve reproducibility and translational value of toxicological and epidemiological data.
Societal awareness and risk assessment
There were several activities focused on rising public awareness on exposure of children to neurotoxic compounds in the environment in DENAMIC. Press contacts have actively been sought, with press releases in English, Dutch and Norwegian, in order to explain the benefits of the project for the European citizen. National policy makers were enabled to take advantage of the DENAMIC results through direct approach by newsletters, meetings etc. organized within the dissemination aspect of this project. Stakeholders (health centres, industry, academia) were invited and joined the two DENAMIC workshops.
Another important active dissemination activity for the general public was the recruitment process of finding respondents for the cohort studies. We have discussed the topics of DENAMIC with thousands of people during the recruitment period. Also DENAMIC was discussed with the people at the midwifery clinics, hospitals and health care centers. During the recruitment process a large number of people has been reached for finding parents that wanted to contribute to the cohort studies. During the recruitment phase, pregnant women were approached at the beginning of pregnancy at gynaecological clinics, midwifery clinics, hospitals and health care centers. DENAMIC researchers and experienced and trained health care workers explained to women the main topics of the project and the importance of the prenatal and early life quality for future health of their children. Women received an Informed Consent form, with further information on the project aims and details of follow up, in order to give them the opportunity to discuss the project participation with their families. This way, not only pregnant women, but their partners and relatives as well, were informed on the project topics and reasons for follow-up. Furthermore, women obtained a list of contact persons in case they wanted to ask questions on the project later on. During follow up meetings of children, parents received feedback from the research team on the project results. An Information Letter was sent to parents or presented during a meeting, including general results from the project (description of exposure to selected environmental pollutants in the cohort). This approach helped parents to become more educated about environmental risk factors for neurotoxic developmental effects in children.
DENAMIC generated a large amount of biomonitoring data for biocides, flame retardants, PFASs, methyl mercury and other compounds. This data will be made available for the databases and ECHA, and other European projects.
DENAMIC showed that well developed animal models is an important bridge between in vitro studies and epidemiological studies. From a prospective point of view, the constant development and marketing of new chemicals will constantly alter the human exposure situation via various exposure routes, such as diet, house dust or air, and also in different combinations of chemicals. Therefore, the DENAMIC results, from the neonatal animal model, shows that risk assessment based on single doses of individual chemicals might underestimate the effects of a chemical as agents can interact to exacerbate developmental toxic effects. This implicates that risk assessment needs to be continuously updated in order to meet the present environmental situation. The risk assessment knowledge will be further disseminated to the WHO and EFSA. Also the risk reduction of certain contaminants (methyl mercury and chlorpyrifos) by intake of food will be further discussed with EFSA WHO, and OECD.
Neurotoxicity data (including new tools and testing methods for individual chemicals and mixtures) were discussed with the OECD at the final DENAMIC workshop and can be used for adverse outcome pathways (AOPs) that are related to (developmental) neurotoxicity. Some of the methodologies may also be applied in future studies in other fields of research (e.g. toxicity or health effects in other target organs).
Policy makers were involved in the DENAMIC workshops. The DENAMIC results will be communicated with EFSA, WHO, European Environment Agency, and OECD. Especially the mixture toxicity results, risk assessment, and risk reduction actions for certain chemicals will be discussed with EFSA, and WHO. In addition, the approaches followed by DENAMIC can be applied in a more generic way, such as in comparable studies on health effects of chemicals and studies on other diseases. Additionally there are several collaborations of the consortium with other universities and regulatory agencies that can use the collected information.
Dissemination activities
Dissemination activities undertaken in DENAMIC was by publication of data in peer-reviewed scientific journals, website (www.denamic-project.eu) project logo, summarizing flyers at the start and end of the DENAMIC project, project newsletters, and a midterm and final workshop. In addition, a documentary film was produced, distributed and made available on YouTube to highlight the results of the project. Also two articles with an educational character on topics related to the DENAMIC project were submitted to popular journals. Currently, a total number of 42 peer reviewed articles were published, 7 are under review and more papers are in preparation.
The flyers contained information on the project to create awareness among all researchers, consumer organisations, health care institutes, industries and authorities active in the field of neurotoxicity, developmental disorders, and neurotoxicity testing. A summary of applied technologies and general outcomes of the DENAMIC project is provided in a second flyer. Items in the newsletters were on the introduction of the DENAMIC project, short reports on project meetings, update on work and announcement of midterm scientific workshop, introduction of young scientists in DENAMIC, and reports on the workshops.
Both the midterm and final workshop were two-day scientific workshops. The first workshop focused on novel tools and methods to screen for (developmental) neurotoxicity; in vitro assays, reduction and refinement of animal studies; epidemiological evidence for effects of environmental exposure on child neuropsychological development and the use of biomarkers in developmental neurotoxicity. The second workshop focussed on DENAMIC findings with regard to screening and tool development, mixture toxicity, exposure assessment, neurodevelopmental epidemiology and risk assessment (including socio-economics). Various stakeholders from the EU and North-American industry, academia, non-governmental organizations and organizations for public health and environment participated in the workshops.
The DENAMIC film “Young Minds at Risk?” was made and sets the stage in which neurotoxic effects of low-concentration mixtures of chemicals in children were investigated. The film explains what the benefits are of chemicals, and explores how some of these chemicals can be harmful especially for the developing brain. How DENAMIC approached this problem and some key results are presented. The film was made with the aim to put the project into context for the viewer, explaining why the research work was needed, what problems it was trying to solve, how the work was carried out, who was involved, the collaborative nature of the process, summarise the results and explain how this knowledge can be applied. The film was launched at the final DENAMIC workshop and will be made available for special screenings at conferences and other events. The film is posted on the project website and made available to partner’s websites, other scientific & documentary websites and broadcasters by YouTube: https://youtu.be/drcOBcA2mD0. A film poster an electronic press kit (EPK) are available by the project website and a copy of the DVD can be requested by the coordinator of DENAMIC.
DENAMIC developed also new technologies. A microsensor for direct measurements of neurotransmitters and cytokines in small amounts of sample (e.g. urine, blood) was developed. The microsensor application and commercialization possibilities are studied. An apparatus to study the on-line behaviour of zebrafish was further developed, tested validated, and is commercially available. Both technologies have wider application areas such as the health and medical fields.
List of Websites:
www.denamic-project.eu
pim.leonards@vu.nl
