CD4+ T cell heterogeneity in gestational age and preeclampsia using single-cell RNA sequencing

Abstract

A balance between pro-inflammatory decidual CD4⁺ T cells and FOXP3⁺ regulatory T cells (FOXP3⁺ Tregs) is important for maintaining fetomaternal tolerance. Using single-cell RNA-sequencing and T cell receptor repertoire analysis, we determined that diversity and clonality of decidual CD4⁺ T cell subsets depend on gestational age. Th1/Th2 intermediate and Th1 subsets of CD4⁺ T cells were clonally expanded in both early and late gestation, whereas FOXP3⁺ Tregs were clonally expanded in late gestation. Th1/Th2 intermediate and FOXP3⁺ Treg subsets showed altered gene expression in preeclampsia (PE) compared to healthy late gestation. The Th1/Th2 intermediate subset exhibited elevated levels of cytotoxicity-related gene expression in PE. Moreover, increased Treg exhaustion was observed in the PE group, and FOXP3⁺ Treg subcluster analysis revealed that the effector Treg like subset drove the Treg exhaustion signatures in PE. The Th1/Th2 intermediate and effector Treg like subsets are possible inflammation-driving subsets in PE.

Keywords: CD4 + T cell; FOXP3; T cell receptor; immune tolerance; preeclampsia; pregnancy; regulatory T cell; single-cell RNA-sequencing.

Figure 1

Landscape of decidual CD4⁺ T cells. (A) UMAP plot of 24,002 cells from 11 samples showing CD4⁺ T cell subsets. Thirteen distinct clusters were annotated using gene expression markers. (B) Violin plots represent the normalized expression levels (y-axis) of key cluster-defining genes in CD4⁺ T cells. (C) UMAP plots of CD4⁺ T cell subsets based on sample origin: healthy early gestation decidua (n = 4, 8472 cells) (upper left), healthy late gestation decidua (n = 3, 6572 cells) (upper right), healthy late gestation PBMC (n = 1, 2628 cells) (lower left), and PE decidua delivered at term (n = 3, 6330 cells) (lower right). (D) Stacked bar graph comparing the distribution of CD4⁺ T-cell clusters based on sample origin. (E) Floating box plots showing the abundance of cells in 0-Tn, 3-Tfh like, 9-PRF⁺ Tct, 11-Th17, and 12-FOXP3⁺ Treg clusters. (F) Heatmap showing percentages of expanded clones (size = 2 or more cells) in each cluster. The columns indicate CD4⁺ T cell clusters. Rows indicate sample origins.

Figure 2

Distributions of expanded clonotypes in each CD4+ T cell cluster. (A) Pie charts show the ratio of expanded clones (n ≥ 2 clones) in each cluster. Gray parts represent unique clonotypes (n = 1 clones). Colored parts represent expanded clonotypes. The same color does not mean the same clonotype. (B–E) UpSetR graphs showing inter-cluster overlap of clonotypes that have clone sizes of 2 or more for each subject. Histogram bar (top) shows number of clonotypes shared between individual clusters (bottom left). Each dot (bottom) represents a cluster in which clonotypes exist. Connected lines represent clusters that share clonotypes.

Figure 3

Different transcriptomic signatures between healthy early vs. healthy late and healthy late vs. PE decidua. (A) Number of DEGs that are upregulated or downregulated in healthy late gestation compared to healthy early gestation decidua. DEG: adjusted p-value < 0.05, |log2 fold change| > 0.5. (B) GOs of downregulated genes in 5-Tm, 7-Th1/Th2 int, and 12-FOXP3⁺ Treg clusters identified by Metascape in healthy late decidua compared to healthy early decidua. Shared GOs between clusters and similar GOs are marked with blue and sky blue, respectively. (C) Number of DEGs that are upregulated or downregulated in PE compared to healthy late gestation decidua. DEG: adjusted p-value < 0.05, |log2 fold change| > 0.5. (D) GOs of upregulated genes in 5-Tm, 7-Th1/Th2 int, and 12-FOXP3⁺ Treg clusters identified by Metascape in PE compared to healthy late decidua. Shared GOs between clusters, similar GOs, and unique key GOs are marked with red, pink, and yellow, respectively. (E) Volcano plot representing DEGs in the 12-FOXP3⁺ Treg cluster shared between healthy late gestation and PE. (F) Violon plot showing the module score of Treg exhaustion in the 12-FOXP3⁺ Treg cluster from term and PE (Mann–Whitney U test).

Figure 4

FOXP3⁺ Treg sub-cluster analysis and alternative Treg functional signature in PE. (A) UMAP plot of 1354 cells from 11 samples showing Treg subsets. Five distinct sub-clusters were annotated based on unique gene-expression markers. (B) Violin plots representing the normalized expression level (y-axis) of key cluster-defining genes in the FOXP3⁺ Treg sub-clusters. (C) Heatmap showing percentages of expanded clones (size = 2 or more cells) in each cluster. Columns indicate FOXP3⁺ Treg sub-clusters. Rows indicate sample origins. (D) Heatmap showing the number of DEGs that are upregulated or downregulated in PE compared to healthy late gestation decidua in each Treg sub-cluster. DEG: adjusted p-value < 0.05, |log2 fold change| > 0.2. (E) Violin plots showing selected genes in Treg4 effector like cluster compared with healthy late and PE. * represents differentially expressed genes (adjusted p-value < 0.05, |log2 fold change| > 0.5) in the whole 12-FOXP3⁺ Treg cluster or Treg4 effector like cluster subset.