Circulating Immune Cells from Early- and Late-onset Pre-eclampsia Displays Distinct Profiles with Differential Impact on Endothelial Activation

Abstract

Pre-eclampsia (PE) affects 5-8% of pregnancies and has detrimental effects on maternal-fetal health. PE is characterized by de novo hypertension after 20 wk of gestation and end-organ damage. Systemic inflammatory imbalance has been associated with PE, but its contribution to the pathology is poorly understood. Our objective was to investigate maternal systemic immune changes in early-onset PE (EOPE) and late-onset PE (LOPE) versus uncomplicated pregnancies (control [CTRL]), and their contribution to endothelial activation, hallmark of hypertension. Blood samples were analyzed by flow cytometry, multiplex assay, intracellular cytokine staining, and single-cell RNA sequencing. We performed cocultures between circulating immune cells and HUVECs to assess endothelial activation. We found that EOPE had decreased regulatory T cells (4.64±0.33, p < 0.05) and monocytes (33.92±3.08, p < 0.01), whereas LOPE had decreased regulatory T cells (4.60±0.30, p < 0.05) and Th2 cells (7.50±0.62, p < 0.01) versus CTRL. Compared to CTRL, elevated cytokines/chemokines, and growth factors were observed in LOPE, whereas EOPE primarily showed decreased levels. Using intracellular cytokine staining, we observed more monocytes producing IL-12, TNF-α, and IL-1β (all p < 0.05) in LOPE versus CTRL. At the transcriptomic level, we found differentially expressed genes between EOPE and CTRL, predominantly related to upregulation of immune activation pathways. Lastly, EOPE PBMCs induced heightened endothelial activation in vitro observed by increased ICAM-1 and ET-1 (p < 0.05), whereas LOPE PBMCs required LPS stimulation. Although significant proteomic changes are observed in the LOPE group, the EOPE displayed changes mostly at the transcriptomic levels and could induce endothelial activation in vitro.

Graphical abstract

FIGURE 1.

Blood pressure measurement and maternal end-organ damages. (A–D) Higher SBP and DBP measurements were taken at specific timepoints during pregnancy in EOPE (A and C) and LOPE (B and D) versus uncomplicated pregnancies (CTRL). (E) A higher proportion of individuals with severe hypertension in the EOPE group, most of them had at least two organs affected, predominantly including liver dysfunction. The data are presented as the means ± SEM with a line plot or number (%). Dark blue indicates CTRL, white indicates EOPE, and light blue indicates LOPE. **p < 0.01, ***p < 0.001 versus CTRL by Kruskal–Wallis test with Dunn’s multiple comparisons test for each timepoint. Timepoint 1 was the first prenatal consultation; timepoint 2 was the second prenatal consultation; timepoint 3 was the last prenatal consultation or EOPE diagnosis; timepoint 4 was the CTRL delivery or LOPE diagnosis; and timepoint 5 was discharge from the hospital. #p < 0.05, ##p < 0.01 versus LOPE by Fisher’s exact test.

FIGURE 2.

Immune cell proportions and levels of inflammatory mediators in the maternal circulation. (A and B) Decreased proportion of Treg cells (CD3⁺CD4⁺CD25⁺) in EOPE and LOPE versus CTRL (A), leading to the decrease of the Treg/Th17 ratio (B). (C and D) LOPE individuals have a decreased proportion of Th2 cells (CD3⁺CD4⁺CD194⁺CD196⁻) (C), whereas EOPE has decreased proportion of monocytes (CD3⁻CD14⁺) (D). (E–H) EOPE individuals had decreased levels of proinflammatory cytokine IL-5 (E), chemokines CCL22 (F), anti-inflammatory cytokine IL-4 and IL-1Ra (G), as well as growth factor Flt3L (H). As for LOPE, they had decreased levels of IL12p70 (E) and increased levels of IL-6 (E), CXCL9 (F), CXCL10 (F), and IL-10 (G). The data are presented as histograms and means ± SEM with scatter plot with bar, or as the mean of normalized value with heat map or mean ± SEM with scatter dot plot. Dark blue indicates CTRL, white indicates EOPE, and light blue indicates LOPE. *p < 0.05, **p < 0.01, ***p < 0.001 versus CTRL by one-way ANOVA with Dunnett’s multiple comparisons test or Kruskal–Wallis test with Dunn’s multiple comparisons test as appropriate.

FIGURE 3.

Profile of cytokines production within the immune cells and endothelial activation. (A) Increased in the proportion of IL-12–producing T CD4⁺ lymphocytes in the LOPE group. (B–D) This group also has a higher proportion of monocytes producing IL-12 (B), TNF-α (C), and IL-1β (D), with the latter being also increased in EOPE (D). (E–G) Moreover, an increase in ICAM-1 (E) and ET-1 (F) production was driven by the PBMCs of EOPE versus CTRL individuals, whereas no change was observed in VCAM-1 production (G). The data are presented as contour plots and as means ± SEM with bar graph or with scatter plot with bar graph. Dark blue indicates CTRL, white indicates EOPE, and light blue indicates LOPE. *p < 0.05 versus CTRL by one-way ANOVA with Dunnett’s multiple comparisons test or Kruskal–Wallis test with Dunn’s multiple comparisons test as appropriate. FSC, forward light scatter.

FIGURE 4.

Single-cell RNA sequencing of circulating immune cells. (A) UMAP plot showing cell types found in the circulation of these individuals. (B) UMAP plots showing the expression of selected genes of cell clusters (enclosed by a dotted line); expression is indicated by a colored bar. (C) Stack bar graphs showing increased percentage of classical monocytes in both PE groups. We also observed an increase of plasma cells, as well as a decrease of nonclassical monocytes and pDC in the EOPE group solely (C). *p < 0.05, ***p < 0.001 versus CTRL by Kruskal–Wallis test with Dunn’s multiple comparisons test.

FIGURE 5.

Transcriptional changes in EOPE T lymphocytes. (A) Naive CD4⁺ T cells in EOPE had 19 DEGs all upregulated represented in the heat map. (B) Naive CD8⁺ T cells in EOPE had 30 upregulated DEGs and 1 downregulated DEG represented in the heat map. (C) Pathway analysis revealed similar enrichment in both naive subtypes using the BioPlanet catalog. (D–F) EOPE CD8⁺ effector T cells had 6 upregulated DEGs displayed on a heat map (D), whereas EOPE CD4⁺ effector T cells cluster 1 and cluster 2 had 4 (E) and 50 (F) upregulated DEGs, respectively. (G) Pathway analysis revealed enrichment only in CD8⁺ T cells and CD4⁺ T cells cluster 2. DEGs meeting the cut-off criteria (logFC ± 1.5 and adjusted p < 0.05).

FIGURE 6.

Transcriptional changes in EOPE monocytes. (A–C) Classical clusters 1 (A) and 2 (B) display similar number of DEGs represented in heat maps, which resulted in mediator-dependent regulation pathways enrichment (C). (D–F) Intermediate monocyte clusters 1 (D) and 2 (E) also had similar numbers of DEGs represented in heat maps, and pathways analysis revealed enrichment of mediator-dependent regulation pathways in both clusters (F). (G) Nonclassical monocytes had 59 upregulated DEGs and 12 downregulated DEGs represented in the heat map. (H) EOPE nonclassical monocytes had enhancement in complement cascade and cytokine interaction pathways. DEGs meeting the cut-off criteria (logFC ± 1.5 and adjusted p < 0.05).

FIGURE 7.

Impact of classical inflammatory stimuli on the cytokines production within the immune cells and their endothelial activation capacity in vitro. (A and B) Stimulation with PMA-IONO leads to a decreased proportion of CD4⁺ T cells producing IL-12 (A) within the LOPE group, whereas their CD8⁺ T cells produce more TNF-α (B). (C and D) The proportion of TNF-α–producing monocytes was increased in the EOPE and CTRL groups following PMA-IONO stimulation (C), whereas LOPE monocytes displayed higher levels of TNF-α production under LPS stimulation (D). (E and F) An increased proportion of IFN-γ–producing monocytes was observed in the LOPE group upon PMA-IONO stimulation (E), whereas IL-10–producing monocytes were decreased in EOPE and CTRL groups after stimulation with LPS (F). (G) Decreased in the production of ICAM-1 in HUVECs cocultured with EOPE PBMCs with PMA-IONO. (G and H) Increased in the production of ICAM-1 in HUVECs cocultured with CTRL (H) or LOPE PBMCs stimulated with LPS (G). The LPS/NS of ICAM-1 production was higher in LOPE versus CTRL. (I) PBMCs of all three groups lead to enhancement of VCAM-1 production when stimulated with LPS or PMA-IONO. The data are presented as means ± SEM with bar graphs with or without scatter plot. Dark blue indicates CTRL (for nonstimulated [NS]); white indicates EOPE (for NS); light blue indicates LOPE (for NS); dark gray indicates NS; and light gray indicates PMA-IONO. #p < 0.05, ##p < 0.01 versus NS or **p < 0.01 versus CTRL by one-way ANOVA with Dunnett’s multiple comparisons test or Kruskal–Wallis test with Dunn’s multiple comparisons test as appropriate.