The immune landscape of fetal chorionic villous tissue in term placenta

Abstract

Introduction: The immune compartment within fetal chorionic villi is comprised of fetal Hofbauer cells (HBC) and invading placenta-associated maternal monocytes and macrophages (PAMM). Recent studies have characterized the transcriptional profile of the first trimester (T1) placenta; however, the phenotypic and functional diversity of chorionic villous immune cells at term (T3) remain poorly understood.

Methods: To address this knowledge gap, immune cells from human chorionic villous tissues obtained from full-term, uncomplicated pregnancies were deeply phenotyped using a combination of flow cytometry, single-cell RNA sequencing (scRNA-seq, CITE-seq) and chromatin accessibility profiling (snATAC-seq).

Results: Our results indicate that, relative to the first trimester, the frequency of fetal macrophages (HBC, proliferating HBC) is significantly reduced, whereas that of infiltrating maternal monocytes/macrophages (PAMM1b, PAMM1a, PAMM2, MAC_1) increased in T3. PAMM1b and HBCs exhibit the most phagocytic capacity at term highlighting their regulatory role in tissue homeostasis in late pregnancy. The transcriptional profiles of resident villous immune subsets exhibit a heightened activation state relative to the relative to T1, likely to support labor and parturition. Additionally, we provide one of the first insights into the chromatin accessibility profile of villous myeloid cells at term. We next stratified our findings by pre-pregnancy BMI since maternal pregravid obesity is associated with several adverse pregnancy outcomes. Pregravid obesity increased inflammatory gene expression, particularly among HBC and PAMM1a subsets, but dampened the expression of antimicrobial genes, supporting a tolerant-like phenotype of chorionic villous myeloid cells. We report a decline in HBC abundance accompanied by an increase in infiltrating maternal macrophages, which aligns with reports of heightened chorionic villous inflammatory pathologies with pregravid obesity. Finally, given the shared fetal yolk-sac origin of HBCs and microglia, we leveraged an in vitro model of umbilical cord blood-derived microglia to investigate the impact of pregravid obesity on fetal neurodevelopment. Our findings reveal increased expression of activation markers albeit dampened phagocytic capacity in microglia with pregravid obesity.

Discussion: Overall, our study highlights immune adaptations in the fetal chorionic villous with gestational age and pregravid obesity, as well as insight towards microglia dysfunction possibly underlying poor neurodevelopmental outcomes in offspring of women with pregravid obesity.

Keywords: epigenomics; macrophage; microglia; monocyte; obesity; placenta; pregnancy; transcriptomics.

Figure 1

Integration of first trimester and term chorionic villous transcriptional profiles. (A) Overarching experimental design – leukocytes were isolated from chorionic villous tissues from term cesarean deliveries (N=46 total, 24 lean and 22 obese). Chorionic villous leukocytes were FACS sorted and subjected to gene expression profiling using 10x 3’single-cell gene expression protocol (scRNA-seq), transcriptome-based profiling using CITE-seq, or snATAC-seq for chromatin accessibility. Expression of activation markers, phagocytic capacity, and responses to bacterial TLR ligands by total chorionic villous leukocytes were assayed using flow cytometry. Spontaneous cytokine production by chorionic villous leukocytes was measured by Luminex assay. Finally, microglia-like cells were derived from UCBMC (iMGL) and expression of activation markers and phagocytic capacity of iMGLs were assayed using flow cytometry. (B) Uniform Manifold Approximation and Projection (UMAP) representation of integrated data with 36,598 leukocytes with an embedded feature plot of the expression of CD14. (C) Violin plots showing expression levels of markers genes used for cluster identification. (D) Bar graph of gene ontology (GO) terms associated with cluster marker genes. Length of the bar indicates the number of genes associated with the GO term and the color indicates log10(Q-value). (E) Violin plots showing module scores for the indicated pathways for each cell cluster. All comparisons between chorionic villous subsets were statistically significant (p<0.05).

Figure 2

Comparison of T1 and T3 chorionic villous immune landscape. (A) Stacked bar graphs of cluster frequencies for T1 (left) and T3 (right) chorionic villous from scRNA-seq data. (B) Violin plots of module scores for the indicated terms in T1 and T3. (C) Bubble plot of select GO terms for DEGs upregulated in T1 or T3 for each cell cluster. The size of the bubble denotes the number of genes mapping to each gene ontology (GO) term, and the intensity of color denotes the -log10(Q-value).

Figure 3

Functional capacity of chorionic villous immune cell subsets at term with pregravid obesity. (A) Experimental design – total chorionic villous leukocytes were assayed for the expression of activation markers, phagocytic capacity, and response to bacterial TLR ligands using flow cytometry. (B) Gating strategy used for the identification of chorionic villous leukocyte populations by flow cytometry. (C) Bar plots of mean fluorescence intensity (MFI) of CD62L (SELL), CD64, CD11c, CD86, and CD163 expression by resting chorionic villous leukocyte subsets. (D) Bar plot representing the percentage of E. coli pHrodo+ cells in each chorionic villous subset. (E) Bar plot of fetal villous (left) and maternal decidua (right) leukocyte responses to stimulation with bacterial TLR ligand cocktail measured by %TNFα+IL-6+ (N=6 for chorionic villous, N=9-10 for maternal decidual tissue). (****p<0.0001, ***p<0.001, **p<0.01, *p<0.05, #p<0.1).

Figure 4

Impact of obesity on the transcriptional profile of term chorionic villous cellular subsets. (A) Stacked bar graph of chorionic villous cluster frequencies between groups (lean: 10,647 cells and obese: 13,087 cells). (B) Module scores for the terms indicated between lean and obese groups. (C) Bubble plot of select gene ontology (GO) terms for DEGs upregulated with obesity in the MAC_2 and HBC clusters. The size of the bubble denotes the number of genes mapping to each GO term, and the intensity of color denotes the -log10(Q-value). (D) Violin plots of select DEGs from MAC_2 and HBC subsets.

Figure 5

Chorionic villous cytokine and growth factor production with pregravid obesity. (A) Experimental design for chorionic villous tissue homogenate Luminex assay. (B) Scatter plot of the concentration of select cytokines and growth factors detected in the supernatant of chorionic villous tissue homogenate with or without pregravid obesity (N=19/group). (*p<0.05, #p<0.1).

Figure 6

Comparison of the epigenetic regulation of term chorionic villous myeloid subsets by snATACseq. (A) UMAP projection of cell subsets within the term chorionic villous (37,316 cells). (B) Gene scores and (C) Motif Matrix for key transcription factors. (D, E) Bubble plot of select GO terms within each subset for (D) promoter and (E) intergenic regions. The size of the bubble denotes the number of genes mapping to each gene ontology (GO) term, and the intensity of color denotes the -log10(Qvalue). (F, G) Bubble plot of select GO terms within the PAMM2 subset between lean and obese groups for (F) promoter and (G) intergenic regions. The size of the bubble denotes the number of genes mapping to each gene ontology (GO) term, and the intensity of color denotes the -log10(Qvalue).

Figure 7

Impact of pregravid obesity on UCB-derived microglia-like cell phenotype and function. (A) Experimental design for the derivation of microglia-like cells (iMGL) from umbilical cord blood mononuclear cells (UCBMC) of newborns of lean subjects and those with obesity. The expression of activation markers and phagocytic capacity of iMGLs were then assayed using flow cytometry. (B, C) Bar plots of (B) the percent of CD163, CX3CR1, TMEM119, and HLA-DR (C) percentage of E. coli pHrodo+ cells expressed by P2RY12+ iMGL in lean and obese groups (N=5/group). (*p<0.05, #p<0.1).