Fig 2: Human chorio-decidual interface-on-chip (CDi-on-chip).A Conceptual framework of how the chorio-decidual interface responds to maternal insults. We hypothesize that the decidua and resident innate immune cells counter maternal insults by promoting inflammation, while the chorion produces immunoregulatory molecules such as HLA-G and PGRMC2. With intact barrier functions, the insult should resolve without causing fetal inflammatory response. When HLA-G or PGRMC2 is compromised, this should result in excessive inflammatory response that may lead to premature labor activation. B Schematic diagram of the chorio-decidual interface-on-chip and treatments. Both (A) and (B) are created with BioRender.com. C Immune cell subpopulations in the decidua parietalis at term prior to labor (n = 3) based on Mosebarger et al. 36. D Phase-contrast images of the CDi-on-chip. Yellow arrowheads point to cell line-derived immune cells. E Immunostaining of CD45 (red) and vimentin (green) expression of cells in the CDi-on-chip. DAPI (blue) was used for nuclear staining. Separation between two chambers is highlighted in yellow dashed circle. Inset images show migrating and migrated cell line-derived immune cells (yellow arrowheads). F Timelapse stills at a device microchannel with a cell line-derived immune cell (yellow arrowhead) migrating towards CTC. Scale bar for (D–F), 100 µm. ROS: reactive oxygen species, NK: natural killer, HLA-G: human leukocyte antigen G, PGRMC2: progesterone receptor membrane component 2, CTC: chorion trophoblast cell, DEC: decidual cell, CD45: cluster of differentiation 45, LPS: lipopolysaccharide, pIC: poly(I:C), mOS: maternal oxidative stress.

Fig 3: Transcriptomic and cellular responses of the CDi-on-chip to insults.A Principal component analysis (PCA) plots of cells based on transcriptomic profile. The first two principal components explained 61.46% of the variance. Colored symbols correspond to CTCs and DECs in both chambers of the CDi-on-chip. B Heatmap and hierarchical clustering of top 50 inflammation and immunity gene expression data for CTCs and DECs in CDi-on-chip. Upregulated genes are represented in red, and downregulated genes are represented in blue. Volcano plots of genes involved in inflammation and immunity in DEC and CTC in response to (C) LPS, (D) pIC, or (E) mOS. Significantly downregulated genes (–log10 adj. p-value > 1, log2 fold change < –1) are shown in blue dots, while significantly upregulated genes (–log10 adj. p value > 1, log2 fold change >1) are shown in red dots. Top differentially expressed genes are labeled. Measurement of (F) IL-6, (G) IL-8, (H) IL-10, (I) TNF-ɑ, (J) GM-CSF, (K) IL-1β, (L) sHLA-G and (M) progesterone production in CDi-on-chip. Multiple student’s (at least n = 4 devices from one placenta per treatment group). Data are presented as mean ± standard error of the mean (SEM). Statistics: Multiple student’s t tests (two-tailed). N Quantification of CD45+ cell migration from DEC into CTC (n = 5 devices from one placenta per treatment group). Statistics: Mann Whitney U test. p values are indicated. IL interleukin; TNF tumor necrosis factor; GM-CSF granulocyte-monocyte colony stimulating factor; sHLA-G soluble HLA-G; P4 progesterone.

Fig 4: Effect of PGRMC2 on transcriptomic and cellular responses of CDi-on-chip to various stimulants.A Immunohistochemistry of human fetal membranes, with PGRMC2 staining brown. B Phase-contrast images of WT and PGRMC2 KO CTC. Scale bar, 100 µm. C Western blot analysis of PGRMC2 expression in WT and PGRMC2 KO CTCs. Beta-actin serves as control. D Volcano plots of genes involved in inflammation and immunity in PGRMC2 KO CTC and DEC compared to WT in control conditions. Significantly downregulated genes (–log10 adj. p value > 1, log2 fold change < –1) are shown in blue dots, while significantly upregulated genes (–log10 adj. p value > 1, log2 fold change > 1) are shown in red dots. GSEA dot plots of top significantly-enriched KEGG pathways in (E) PGRMC2 KO CTC and (F) corresponding DEC. Circle size denotes number of enriched genes in a pathway. Gene ratio denotes the quotient between number of enriched genes and total number of genes in the said pathway. Color denotes p value. G Venn diagram of common differentially expressed genes in PGRMC2 KO CTC compared to wild-type CTC and their corresponding DEC in various conditions. GSEA ridge plots of top significantly-enriched KEGG pathways in PGRMC2 KO CTC and their enrichment scores (x-axis) in response to (H) LPS, (I) pIC and (J) mOS. Peak corresponds to enrichment score; enrichment score below 0 denotes suppressed pathways, while enrichment score above zero denotes activated pathways. Color denotes p value. K Immunostaining of and vimentin (green) expression of WT and PGRMC2 KO CTCs in CDi-on-chip. DAPI (blue) was used for nuclear staining. Scale bar, 100 µm. Measurement of (L) chorionic IL-6 and TNF-ɑ, (M) decidual IL-6 and GM-CSF, (N) sHLA-G and (P) progesterone production in CDi-on-chip (at least n = 4 devices from one placenta per treatment group). Data are presented as mean ± SEM. Statistics: Multiple student’s t-tests (two-tailed). O Western blot analysis of HLA-G expression in wild-type and PGRMC1 or PGRMC2 KO CTCs. Beta-actin serves as control. Q Quantification of CD45+ cell migration from DEC into CTC (n = 5 devices from one placenta per treatment group). Data are presented as median (2nd quartile; solid line) and 1st and 3rd quartiles (dashed lines). Statistics: Multiple Mann Whitney U tests (two-tailed). AEC amnion epithelial cell, AMC amnion mesenchymal cell, CMC chorion mesenchymal cell, WT wild-type, KO knockout, GSEA Gene set enrichment analysis.

Fig 5: LD patterns between eight pairs polymorphic sites at the 3’UTR region of HLA-G gene. The image was generated in the Haploview 4.2. D’ and r2, pairwise correlation between polymorphic sites. Areas in red indicated strong LD (r2>0.8), shades of pink indicated moderate LD, and white indicates low LD. The r2 values (×100) for the marker pairs are listed in the corresponding boxes.