Fetal Response to Maternal Exposures to Environmental Chemicals: Utility of a Four-Cell Human Feto-Maternal Interface Organ-on-Chip

Epidemiological studies suggest that maternal exposures to environmental chemicals may be linked to spontaneous preterm birth. Mechanistic studies are needed to provide support to these hypotheses; however, existing in vitro models do not replicate the human feto-maternal barriers beyond placenta. The recently developed four-cell (human immortalized maternal decidua, chorion trophoblast, amnion mesenchymal, and amnion epithelial cells) Feto-Maternal interface Organ-On-Chip (FMi-OOC) enables studies of chemical effects on the decidua, chorion and amnion, maternal and fetal tissues that are critically important for maintaining full-term pregnancy. We tested four environmental compounds that have been associated with preterm birth – dichlorodiphenyltrichloroethane (DDT-o,p’), bisphenol A (BPA), 2,2'4,4'-tetrabromodiphenyl ether (PBDE-47), and perfluorooctanoic acid (PFOA). First, concentration-response effects of these chemicals were tested on maternal decidua cells in 96-well plates. Then, experiments were performed in FMi-OOC containing four fetal-maternal cell types arranged into four concentric chambers, mimicking the in utero cell topology. Compounds were added to the maternal (i.e., decidua) chambers, and chemical propagation, cell viability and proinflammatory cytokines (IL-6, IL-8, GM-CSF, TNF-α) were measured in decidua, chorion trophoblast, amnion mesenchymal, and amnion epithelial cells for up to 72 hrs. Minimal propagation of test compounds to fetal chambers was observed and tested concentrations were non-cytotoxic. Treatment-associated increase in cytokine production was observed for all compounds, with PFOA and BPA showing the strongest effects and amnion epithelial cells being the most responsive. We demonstrate how the multi-cellular FMi-OOC model can be used to study paracrine signaling in complex human tissues such as fetal-maternal interface. In addition, these studies provide mechanistic support for the plausibility of the epidemiological associations between maternal exposures to tested compounds and preterm birth. We show that upon maternal exposure, albeit at concentrations exceeding reported human blood levels by 1 to 2 orders of magnitude, fetal membranes attain a pro-inflammatory state, a potential trigger for preterm birth.