Using 3D gastrointestinal tract in vitro models with microfold cells and mucus secreting ability to assess the hazard of copper oxide nanomaterials

BACKGROUND: Copper oxide nanomaterials (CuO NMs) are exploited in many products including inks, cosmetics, textiles, wood preservatives and food contact materials. Their incorporation into these products may enhance oral exposure in consumer, environmental and occupational settings. Undifferentiated and differentiated monocultures of Caco-2 cells are commonly used to assess NM toxicity to the intestine in vitro. However, the integration of other cell types into Caco-2 in vitro models increases their physiological relevance. Therefore, the aim of this study is to evaluate the toxicity of CuO NMs and copper sulphate (CuSO(4)) to intestinal microfold (M) cell (Caco-2/Raji B) and mucus secreting (Caco-2/HT29-MTX) co-culture in vitro models via assessment of their impact on barrier integrity, viability and interleukin (IL)-8 secretion. The translocation of CuO NMs and CuSO(4) across the intestinal barrier was also investigated in vitro. RESULTS: CuO NMs and CuSO(4) impaired the function of the intestinal barrier in the co-culture models [as indicated by a reduction in transepithelial electrical resistance (TEER) and Zonular occludens (ZO-1) staining intensity]. Cu translocation was observed in both models but was greatest in the Caco-2/Raji B co-culture. CuO NMs and CuSO(4) stimulated an increase in IL-8 secretion, which was greatest in the Caco-2/HT29-MTX co-culture model. CuO NMs and CuSO(4) did not stimulate a loss of cell viability, when assessed using light microscopy, nuclei counts and scanning electron microscopy. CuO NMs demonstrated a relatively similar level of toxicity to CuO(4) in both Caco-2/Raji B and Caco-2/HT29-MTX co- culture models. CONCLUSIONS: The Caco-2/Raji B co-culture model was more sensitive to CuO NM and CuSO(4) toxicity than the Caco-2/HT29-MTX co-culture model. However, both co-culture models were less sensitive to CuO NM and CuSO(4) toxicity than simple monocultures of undifferentiated and differentiated Caco-2 cells, which are more routinely used to investigate NM toxicity to the intestine. Obtained data can therefore feed into the design of future studies which assess the toxicity of substances (e.g. NMs) and pathogens to the intestine (e.g. by informing model and endpoint selection). However, more testing with a wider panel of NMs would be beneficial in order to help select which in vitro models and endpoints to prioritise when screening the safety of ingested NMs. Comparisons with in vivo findings will also be essential to identify the most suitable in vitro model to screen the safety of ingested NMs. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12951-019-0503-1) contains supplementary material, which is available to authorized users.