A multi-organ, feto-maternal interface organ-on-chip, models pregnancy pathology and is a useful preclinical extracellular vesicle drug trial platform

Abstract

Pregnant women and their fetuses are often excluded from clinical trials due to missing drug-related pre-clinical trial information at the human feto-maternal interface (FMi). The two interfaces-placenta/decidua and fetal membranes/decidua are gatekeepers of drug transport; however, testing their functions is impractical during pregnancy. Limitations of current in-vivo/in-vitro models have hampered drug development and testing during pregnancy. Hence, major complications like preterm births and maternal and neonatal mortalities remain high. Advancements in organ-on-chip (OOC) platforms to test drug kinetics and efficacy and novel extracellular vesicle-based fetal drug delivery are expected to accelerate preclinical trials related to pregnancy complications. Here we report the development and testing of a humanized multi-organ fetal membrane/placenta (fetal)-decidua (maternal) interface OOC (FMi-PLA-OOC) that contains seven cell types interconnected through microchannels to maintain intercellular interactions as seen in-utero. Cytotoxicity, propagation, mechanism of action, and efficacy of engineered extracellular vesicles containing anti-inflammatory interleukin (IL)-10 (eIL-10) were evaluated to reduce FMi inflammation associated with preterm birth. A healthy and disease model (lipopolysaccharide-infectious inflammation) of the FMi-PLA-OOC was created and co-treated with eIL-10. eIL-10 propagated from the maternal to fetal side within 72-hours, localized in all cell types, showed no cytotoxicity, activated IL-10 signaling pathways, and reduced lipopolysaccharide-induced inflammation (minimized NF-kB activation and proinflammatory cytokine production). These data recapitulated eIL-10s' ability to reduce inflammation and delay infection-associated preterm birth in mouse models, suggesting FMi-PLA-OOC as an alternative approach to using animal models. Additionally, we report the utility of eIL-10 that can traverse through FMis to reduce inflammation-associated pregnancy complications.

Keywords: EVs; Fetal membrane; Interleukin-10; Microphysiological systems; Placenta; preterm birth.

Figure 1.. Fetal membrane (FMi) and placenta (PLA) feto-maternal interface OOC design and cellular-collagen components

A) The FMi-PLA-OOC mimics the fetal membrane and placenta-maternal interfaces (black box) inside one device. The fetal layers of the FMis are connected by the maternal decidua, which contains maternal blood vessels. B) Each cellular and collagen layer is represented as cell culture chambers in the OOC device. C) A dye-loaded PDMS device shows the cell chamber layer and the media reservoir. D) Cross-section view of the width and height of each cell chamber and connecting microchannels. Microchannels between the HUVEC-CTB chamber are filled with Type I collagen, while the CTC-AMC and AMC-AEC microchannels are filled with Type IV collagen. Collagen is visualized as blue with Masson trichrome stain. 100μM scale bar.

Figure 2:. Analysis of cellular components of the FMi-PLA-OOC.

A) Stitched brightfield image of cells loaded into each cell culture chamber. 600μM scale bar. B) Brightfield image of a multi-layered trophoblast barrier between the CTB-STB chambers. 100μM scale bar. C) Brightfield microscopy documented the cell morphology of each cell type after they were cultured within FMi-PLA-OOC cell chambers for 5–6-days. 20μM scale bar. A variety of immunocytochemistry stains were conducted for cell-specific markers in each cell chamber. HUVECs – Muc18 in green; Cytotrophoblasts (CTB) – cytokeratin 7 (CK7) in red; Syncytiotrophoblasts (STB) – CK7 in red; Decidua (DEC) – vimentin (VIM) in green; Chorion trophoblasts (CTC) - histocompatibility antigen, class I, G (HLA-G) in red; Amnion mesenchymal cell (AMC) – VIM/CK18 in green and red; Amnion epithelial cell (AEC) - VIM/CK18 in green and red. Dapi nuclear stain in blue (N=3). 20μM scale bar. D) An illustration of the gradient of media within the reservoirs of each cell culture chamber of the FMi-PLA OOC. The decidua chamber and reservoir were loaded with 3000kd Dextran beads (N=3). E) 3000kd Dextran beads were tracked every 24-hours for 120-hours to monitor their diffusion from the decidua chamber across the chambers on either side of the FMi-PLA OOC. At 120-hours, 3000kd Dextran beads had diffused from the decidua chamber into the HUVEC chamber on the placental side and into the CTC chamber on the fetal side of the device. F) Plate reader analysis of the supernatants from each chamber showed that by 120-hours, 3000kd Dextran beads had propagated from the decidua chamber into all other chambers of the FMi-PLA-OOC. 2000μM scale bar.

Figure 3:. Establishment of an infectious inflammation disease state within the FMi-PLA-OOC

A) An anatomical illustration and 3D model of the maternal and fetal components of both FMis (black box: placenta-decidua and fetal membrane-decidua with the introduction of inflammation at the level of the decidua). B) Immunocytochemistry showing P-NF-kB expression (green) and localization. LPS significantly induced P-NF-kB expression in DEC after 30 min of treatment. Blue – DAPI. Graph representing (N=3) data as a mean± SEM. 20μM scale bar. C) Immunocytochemistry followed by fluorescent microscopy shows LPS (red dots; white arrows) propagation from the decidua into all other cell culture chambers of the FMi-PLA OOC by 6-days (N=3). Blue – DAPI. 20μM scale bar. D) Cell-specific immunocytochemistry stains were assessed in each cell layer after 6-day LPS treatment. HUVECs – Muc18 in green; CTB – cytokeratin 7 (CK7) in red; STB – CK7 in red; DEC – HLA-DR in green; CTC - histocompatibility antigen, class I, G (HLA-G) in red; AMC – VIM/CK18 in green and red; AEC - VIM/CK18 in green and red. DAPI nuclear stain in blue (N=3). 20μM scale bar. Graph representing N=3 data as a mean± SEM. E) Cytokine multiplex assay showing LPS-induced pro-inflammatory cytokine production, IL-6 or IL-8, in different cell layers. Graphs representing N=3 data as a mean± SEM.

Figure 4:. The effect of eIL-10 at the fetal membrane and placenta feto-maternal interface

A) Schematic showing the different treatment groups used to evaluate eIL-10 toxicity, mechanism of action, propagation, and efficacy. B) LDH cytotoxicity assay showing the percent of dead FMis cells within the FMi-PLA-OOC after 6-day treatments. Data are shown as mean±SEM (N=3). C) Quant Blue IL-10 functional assay documenting eIL-10 induced signaling in DECs after 30 min treatment on days 0 and 3. Data are shown as mean±SEM (N=3). D) Immunocytochemistry showing P-NF-kB expression (green) and localization. LPS significantly induced P-NF-kB expression in DEC after 30 min of treatment, while eIL-10 pre- or post-LPS treatment significantly reduced P-NFkB expression. Blue – DAPI. Graph representing (N=3) data as a mean± SEM. 20μM scale bar. E) Immunocytochemistry followed by fluorescent microscopy shows CFSE labeled EVs (green dots; white arrows) propagation from the decidua into all other cell culture chambers of the FMi-PLA OOC by 3-days (N=3). Blue – DAPI. 10μM scale bar.

Figure 5:. eIL-10 treatment reduces inflammation at the fetal membrane and placenta feto-maternal interfaces

Cytokine multiplex assay showing LPS-induced anti-inflammatory cytokine (IL-10) or pro-inflammatory cytokine (IL-8, TNF-α, IL-6) production or eIL-10 pre- or post-treatment suppressing pro-inflammatory cytokines in different cell layers. Graphs representing N=3 data as a mean± SEM.