Impacts of chronic exposure to inorganic food additives from in utero life to adulthood on the development of neurodevelopmental and metabolic disorders: role of the microbiota – DevADDIRisk

The DevADDiRisk project used laboratory mice fed diets enriched or not with the food additive of interest (E171, E551, E554, alone or in combination) at human exposure doses based on exposure data published by the European Food Safety Authority (EFSA). The mothers were exposed from 40 days before conception, then during gestation and lactation. The offspring were then exposed to the same diet from weaning onwards. The exposure protocols (7 months in total) are based on the OECD guidelines for chronic studies (Extended One Generation Repro-Toxicity Study, EOGRTS). They also follow EFSA recommendations on nanotoxicology requirements applied to food to be useful in assessing health risks to humans. The fate of food additives was studied by ICP-MS measurements of elements of interest (titanium, silica, aluminium) in foetuses and offspring (liver, brain), supplemented by electron microscopy coupled with X-ray spectrometry (EDX) to measure the size of particles (TiO2, SiO2, Al) accumulated in tissues. The impacts along the microbiota-immune system axis and on the integrity of the intestinal barrier, brain functions and metabolism in young and adult offspring are studied using a combination of targeted or high-throughput molecular approaches (metagenomics by 16S sequencing of bacteria, transcriptomics for gene expression, metabolomics by NMR on faeces to assess the production of bacterial metabolites by the microbiota), inflammatory status (cytokines profiles), flow cytometry (immune cell frequency), histology (intestine, liver, brain), behavioural studies (cognitive functions) and metabolism (glucose tolerance, fasting insulin levels) in both sexes. Finally, approaches involving the transfer of gut microbiota from exposed mice to recipient mice (axenic) that had not been treated with additives made it possible to determine the role of gut microbiota in the most significant deleterious effects previously identified in the project.

In both sexes, TiO2 NPs accumulate in the foetuses of mothers exposed to E171, as well as in the livers of adult offspring who have consumed the same diet as their mothers since weaning. No similar accumulation was observed with the other additives (foetus, liver, brain), while only E554 caused a decrease in foetal and placental weight. These results confirmed the previously suggested maternal-foetal transfer of E171 NPs in humans and validate the relevance of the mouse model of chronic exposure from the in utero stage to assess the risk in the general population.

At the immune level, a breakdown in oral tolerance to food antigens appears in male offspring exposed to E171 or E551. For E171 (TiO2), this effect results in increased susceptibility of juveniles to cow's milk protein allergy, accompanied by increased intestinal permeability and dysbiosis. In adulthood, males exposed to E551 (SiO2) develop low-grade colic inflammation and systemic immunosuppression, verified after septic challenge with LPS. This vulnerability correlates with a decrease in dendritic cells and lymphocytes (Th1/Th17/Tregs), responsible for a decrease in cytokine production after in vitro stimulation. In contrast, E554 (Al) is associated with systemic pro-inflammatory effects, which are more pronounced in females and dominate when combined with E171 and E551.

In vitro, E171 induces dysbiosis of the human microbiota, affecting taxa linked to immune signalling. In mice, E171 and E551 induce dysbiosis in males alone (currently being analysed for E554), accompanied by alterations in glucose metabolism, exacerbated by a high-fat diet. The transfer of flora from E171 males to axenic mice reproduces glucose intolerance, demonstrating the causal role of the altered microbiota alone. For E551, supplementation with A. muciniphila and B. pseudolongum, whose abundance was reduced after treatment, protects males from metabolic disorders.

In terms of neurobehaviour, adult mice exposed to E171 and E551, alone or in combination, show spatial memory deficits, which vary according to sex and mode of exposure. Social behaviour deficits are observed in both sexes exposed to the mixture and in females exposed to E171 alone. Memory disorders appear as soon as weaning; in males exposed to E551, they correlate with inflammation in the hippocampus, while in females exposed to E171, a decrease in BDNF expression is observed, suggesting early neurobiological alterations. Advanced physical development is observed in both sexes exposed to the mixture of additives, while sensorimotor development is impaired in males exposed to E171 alone, but improved in females under all conditions, all of which indicates marked sexual dimorphism in behavioural consequences.

The DévADDiRisk project opens up scientific and regulatory perspectives, starting with demonstrating the causal link between TiO2-induced dysbiosis and metabolic disorders following chronic exposure to the colouring agent E171. It encourages the consideration of the microbiome in the EFSA's risk assessment, which may lead to new recommendations in the definition of toxic reference values for additives with biocidal properties. For E171, our data support the decision to ban it in the EU since 2022, and are a strong argument in favour of its re-evaluation in countries where it remains authorised, including in the pharmaceutical and cosmetics sectors. Regardless of the additive (E171/E551/E554), alone or in combination, the data opens up avenues for research into their immunotoxicity, in particular the need for dedicated protocols (in vivo/ex vivo challenges) to reveal breaks in oral tolerance to antigens that promote a ‘risk’ of food hypersensitivity from an early age, which is not detectable under baseline conditions. Taken together, these data encourage epidemiological studies to explore these links in humans.

The project also opens up new research into the adverse effects of nanoparticulate additives, particularly along the microbiota-immune system axis. It calls for examination of long-term consequences, particularly cognitive development and gender differences. It shows that accumulation in the brain is not necessary to observe behavioural disorders in offspring.

With regard to mixed effects, no additivity or synergy was reported in mice exposed to levels corresponding to the last percentile of the maximum human exposure scenario. This suggests that biological responses are already greatly amplified, which may mask combined effects. These results argue for a dose-dependent approach that takes into account the variability of exposure levels (low, moderate, high) according to age group and additive, thereby better reflecting actual exposure levels in the population.

On a societal level, the harmful effects on immune and neurobehavioural status have particularly affected young people, highlighting the need for specific nutritional recommendations for children and adolescents, thereby limiting their exposure to additives with immunosuppressive (E551) or inflammatory (E171/E554) potential. For manufacturers, our data encourages the development of safe-by-design alternatives for common applications (colourings, anti-caking agents). Finally, better information for the general public on the potential risks associated with the consumption of ultra-processed foods rich in these additives could promote a shift in eating habits towards greater transparency and health safety.

- An oral communication (E171) selected for the annual congress of the American Society of Toxicology (2022): Evariste L, Gaultier E, Cartier C, Noireaux J, Chassaing B, Houdeau E, Lamas B. Mice exposed to food-grade TiO2 from in utero life to adulthood show sex-specific gut microbiota and metabolic disorders. Society of Toxicology (SOT) 61st Annual meeting, San Diego, USA, March 2022.

- A review article (E171) in the Cahiers de Nutrition et de Diététique (2023): Lamas B, Evariste L, Houdeau E. Interactions of dietary titanium dioxide with the microbiota-immune system axis: a new player in the development of metabolic disorders? 58: 70-81.

Concerns are raising about the potential health consequences for Humans daily exposed to inorganic nanoparticles (NPs) present in common food additives. Recent works showed that NP ingestion alters intestinal flora, brain structures and behaviour. These effects are related to the physico-chemical properties of NPs enable them to cross biological barriers, like the intestine, blood-brain-barrier and placenta. Among the food additives, titanium dioxide (TiO2, E171 in UE), silver (Ag, E174) and silicon dioxide (SiO2, E551) are representative of manufactured nanomaterials exposing daily the general population through the diet, including pregnant women. Ag (E174) is used to colour the surfaces of cakes, ice creams and chocolates, with exposure levels ranging 0.03-2.6µg/kg body weight (bw)/day (d). TiO2 (E171) used in confectionery, white sauces and icing, have exposure levels from 0.3 to 10.4mg/kg bw/d, while SiO2 (E551) is added to powdered food as an anticaking agent (salt, sugar, spices, dried milk), with exposure level range of 0.9-31.2mg/kg bw/d. The French High Council of Public Health highlights in 2018 that the human risk of a transfer of TiO2-NPs to the foetus and health consequences for the new-born are not documented while animal studies showed that TiO2-, Ag- and SiO2-NP models cross the placental barrier and are detected in the newborn brain, with consequences on offspring development and behaviour. SiO2-, TiO2- and Ag-NPs accumulation in rodent newborns’ liver also suggests potential alteration of energy metabolism. However, these studies have been carried out at high doses with nanomodels (i.e., 100% nanosized) administered to dams by inhalation or subcutaneous injections. Within the scope of food additives exhibiting a mix of micro- and nanosized particles in their composition, studies on developmental toxicity should consider perinatal exposure to foodborne NPs (E171, E174 and E551) orally consumed by mothers. Because TiO2-, SiO2- and Ag-NPs showed anti-bacterial activities, a chronic oral exposure to E171, E174 and E551 could alter the gut microbiota leading to dysbiosis, with deleterious effects in the offspring. Indeed, the microbiota play a key role in several physiological functions such as energy metabolism, brain development and behaviour, but is often neglected in food toxicology. The dialogue between the gut microbiota and distant organs are complex and involve bacterial metabolites, such as the short-chain fatty acids (SCFA) and the aryl hydrocarbon receptor (AhR) ligands, which are essential for the maintenance of intestinal, nervous system and liver functions.
Our preliminary data show that circulating Ti(O2) in pregnant women accumulates in the placenta then reach the meconium suggesting foetal exposure, and a transplacental passage of E171 particles occurred ex vivo on human perfused placentae. Interestingly, in adult rats orally exposed to E171, we observed a microglia activation and neurogenesis alteration in the hippocampus essential for memory and learning. Altered microbiota metabolism activity leading to impaired production of SCFA and AhR ligands was observed in rodents orally exposed to E171 for 60 or 100 days. These preliminary data suggest that chronic oral exposure of mothers to inorganic food additives (E171, E174 and E551) could be involved in the occurrence of neurodevelopmental and/or metabolic disorders in the offspring. These effects may be synergistic or even potentiated by intestinal dysbiosis transmitted by the mothers to the offspring and/or directly induced in the offspring by altering the microbiota colonization of the gut. The aim of this project through a longitudinal study from in utero life to adulthood in mice is to explore the perinatal fate of the food additives E171, E174 and E551, and the consequences of long-term exposure on neurological and metabolic functions of the descendants, including the potentially synergistic role of the gut microbiota in these alterations.