Type I interferon exposure of an implantation-on-a-chip device alters invasive extravillous trophoblast function

Abstract

Inappropriate type I interferon (IFN) signaling during embryo implantation and placentation is linked to poor pregnancy outcomes. Here, we evaluate the consequence of elevated type I IFN exposure on implantation using a human implantation in an organ-on-a-chip device. We reveal that type I IFN reduces extravillous trophoblast (EVT) invasion capacity. Analyzing single-cell transcriptomes, we uncover that IFN truncates invasive EVT emergence in the implantation-on-a-chip device by stunting EVT epithelial-to-mesenchymal transition. Disruptions to the epithelial-to-mesenchymal transition are associated with the pathogenesis of preeclampsia, a life-threatening disorder of pregnancy. Strikingly, IFN stimulation induces genes associated with increased preeclampsia risk in EVTs. These dysregulated EVT phenotypes ultimately reduce EVT-mediated endothelial cell vascular remodeling in the implantation-on-a-chip device. Overall, our work implicates unwarranted type I IFN as a maternal disturbance that can result in abnormal EVT function that could trigger preeclampsia.

Keywords: EVT invasion; extravillous trophoblast; implantation; implantation-on-a-chip; preeclampsia; reproductive immunology; type I interferon.

Graphical abstract

Figure 1

Type I interferon exposure abrogates EVT invasion (A) A visual portrayal of human embryo implantation, where cytotrophoblasts, present within the fetal chorionic villi, differentiate into extravillous trophoblasts (EVTs) that invade into the maternal uterus toward maternal vasculature. (B) Top left: schematic of the implantation-on-a-chip (IOC) microfluidics device. The center and two side lanes have dimensions of 0.5 mm (width) × 0.3 mm (height) and 0.25 mm (width) × 0.3 mm (height), respectively. Top right: compartmentalized design of the IOC device allows EVTs to migrate through an extracellular matrix (ECM) toward maternal endothelial cells (ECs). Bottom row: 3D (left) and 2D top-down (right) representative images of EVTs (green: CellTracker Green) migrating across the ECM hydrogel toward maternal ECs (red: CD31). (C) Graphic of experimental timeline shown in days with start of type I interferon (IFN-β) treatment at day 0. (D) Representative image of IOC device at day 3 (top row) and day 5 (bottom row) with EVTs (green: CellTracker Green) migrating toward maternal ECs (red: CD31) in control and after IFN-β (1,000 IU/mL) exposure; scale bars, 200 μm. Representative images are from three independent experiments. (E) Quantification of EVT area invasion at day 1, 3, and 5. (F and G) Quantification of number of invaded EVTs (F) and depth of invasion (G) at day 5 post treatment. Data are presented as mean ± SEM (n = 3 independent devices per group). two-way ANOVA with Tukey’s multiple comparison test (E). One-way ANOVA with Dunnett’s multiple comparison test (F and G). ns, p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001 (n = 3 independent devices per group).

Figure 2

Elevated type I interferon limits invading extravillous trophoblast emergence (A) Uniform manifold approximation and projection (UMAP) plot of scRNA-seq of cells within 36 implantation-on-chip (IOC) devices (n = 38,495 cells) colored by cell type and state. (B) UMAP plot of marker genes characteristic of trophoblasts (left) and endothelial cells (right). (C) Violin plots showing normalized and log-transformed expression of genes characteristic of endovascular EVTs in individual EVT subsets (x axis). (D) Violin plots showing normalized expression of the type I IFN receptor subunits (IFNAR1 and IFNAR2) and representative interferon (IFN)-stimulated genes (x axis) from scRNA-seq data in all cells in the IOC device separated by control (white) and type I IFN-exposed cells (gray). (E) Bar plot representing the proportion of EVT subsets separated by control and IFN-stimulated cells. Data are presented as mean ± SEM. (F) Minimum spanning tree computed by Slingshot, visualized on the UMAP of EVT subsets, separated by control and type I IFN-stimulated cells.

Figure 3

Elevated type I interferon alters epithelial-to-mesenchymal transition and promotes a preeclamptic gene signature in extravillous trophoblasts (A) Schematic describing the dynamic spectrum of epithelial-to-mesenchymal transition (EMT) in cytotrophoblast to extravillous trophoblast (EVTs) transition. Small arrows indicate that first trimester EVTs display more developed mesenchymal features during active endometrial remodeling as compared with third trimester EVTs. (B) Violin plots showing normalized expression of representative canonical epithelial genes (left) and mesenchymal genes (right) in all EVT cells. (C) Dot plots showing normalized, log-transformed, and variance-scaled expression of genes characteristic of mesenchymal genes and regulatory, inducers of EMT (y axis) in individual EVT subsets (x axis). (D) Boxplot showing normalized, classical mesenchymal genes in all EVT cells separated by control (white) and IFN-stimulated cells (gray). (E) Dot plots showing normalized, log-transformed, and variance-scaled expression of preeclampsia risk factors (x axis) in all EVT separated by control and IFN-stimulated cells (y axis). (F) Venn diagram of dysregulated genes in EVTs from patients with PE with early- and late-onset preeclampsia with our list of differentially expressed genes in IFN-exposed EVTs (upregulated and downregulated genes had a false detection rate less 5% and log fold change greater than one). Representative genes that overlap are listed to the right.

Figure 4

EVT-directed vascular remodeling is limited by type I interferon (A) Minimum spanning tree computed by Slingshot, visualized on the UMAP of EC subsets, EC1 (dark blue) and EC2 (light blue). (B) Violin plots showing normalized expression of VE-cadherin (CDH5) in individual EC subsets (x axis). (C) Dot plots showing normalized, log-transformed, and variance-scaled expression of genes characteristic of progression toward an apoptotic state and endothelial function (y axis) in individual EC subsets (x axis). (D) Percent mitochondrial gene expression (x axis) in individual EC subsets, separated by control (bottom) and IFN-stimulated cells (top) (y axis). (E–G) (E) Visualization and (F and G) quantification of caspase-3 (magenta) expression. Scale bars, 50 μm. The representative images are from three independent experiments. Two-sided t test (n = 3 independent devices per group). Data are presented as mean ± SEM. ∗p < 0.05; ∗∗p < 0.01.