MAIT Cells Balance the Requirements for Immune Tolerance and Anti-Microbial Defense During Pregnancy

Abstract

Mucosal-associated invariant T (MAIT) cells are an innate-like T cell subset with proinflammatory and cytotoxic effector functions. During pregnancy, modulation of the maternal immune system, both at the fetal-maternal interface and systemically, is crucial for a successful outcome and manifests through controlled enhancement of innate and dampening of adaptive responses. Still, immune defenses need to efficiently protect both the mother and the fetus from infection. So far, it is unknown whether MAIT cells are subjected to immunomodulation during pregnancy, and characterization of decidual MAIT cells as well as their functional responses during pregnancy are mainly lacking. We here characterized the presence and phenotype of Vα7.2⁺CD161⁺ MAIT cells in blood and decidua (the uterine endometrium during pregnancy) from women pregnant in the 1^st trimester, i.e., the time point when local immune tolerance develops. We also assessed the phenotype and functional responses of MAIT cells in blood of women pregnant in the 3^rd trimester, i.e., when systemic immunomodulation is most pronounced. Multi-color flow cytometry panels included markers for MAIT subsets, and markers of activation (CD69, HLA-DR, Granzyme B) and immunoregulation (PD-1, CTLA-4). MAIT cells were numerically decreased at the fetal-maternal interface and showed, similar to other T cells in the decidua, increased expression of immune checkpoint markers compared with MAIT cells in blood. During the 3^rd trimester, circulating MAIT cells showed a higher expression of CD69 and CD56, and their functional responses to inflammatory (activating anti-CD3/CD28 antibodies, and IL-12 and IL-18) and microbial stimuli (Escherichia coli, group B streptococci and influenza A virus) were generally increased compared with MAIT cells from non-pregnant women, indicating enhanced antimicrobial defenses during pregnancy. Taken together, our findings indicate dual roles for MAIT cells during pregnancy, with an evidently well-adapted ability to balance the requirements of immune tolerance in parallel with maintained antimicrobial defenses. Since MAIT cells are easily activated, they need to be strictly regulated during pregnancy, and failure to do so could contribute to pregnancy complications.

Keywords: T cells; anti-microbial response; decidua; immune tolerance; immunomodulation; innate-like T cells; mucosal-associated invariant T (MAIT) cells; pregnancy.

Figure 1

MAIT cells are present in 1^st trimester decidua, but are relatively excluded from the decidua. (A) Gating strategy for circulating (blood) and decidual T cells and MAIT cells. PBMCs and decidual cells were isolated from 1^st trimester blood and decidual tissue, respectively, from the same donors, and analyzed by flow cytometry. MAIT cells are identified by their expression of CD45, CD3, Vα7.2 and CD161. (B) NK cell, (C) T cell, (D) MAIT cell, and (E) B cell frequency in PBMCs (blood) and in cells isolated from decidua, expressed as percent of CD45⁺ lymphocytes, or – in the case of MAIT cells – expressed as percent of T cells. Gating strategies for NK and B cells are presented in Supplementary Figure S2. Bars depict median and interquartile range (IQR) from 21 donors in (B, E), 24 donors in (C) and 23 donors in (D). Statistical comparisons were performed with Wilcoxon signed-rank test. ** p < 0.01, **** p < 0.0001.

Figure 2

Different composition of MAIT cell subsets in the decidua compared with blood. Proportions of CD8⁺, CD4⁺ and double-negative (DN) subsets of MAIT cells (A) and non-MAIT T cells (B) in blood and decidua (dec). The gating strategy for CD8⁺, CD4⁺ and DN cells is shown in Supplementary Figure S2. MAIT cells were further analyzed for CD56 expression (C). Gating of CD56⁺ MAIT is shown in Supplementary Figure S3. Bars show median and IQR from 20-24 donors. Statistical comparisons were performed with Wilcoxon signed-rank test. * p < 0.05, *** p < 0.001, **** p < 0.0001.

Figure 3

Decidual MAIT cells express activation and immune checkpoint markers and exhibit a fundamentally different phenotype than circulating MAIT cells from the same women. PBMCs and decidual cells were isolated from blood and decidual tissue (dec), respectively, from the same donors, and analyzed by flow cytometry. (A) illustrates the gating strategy employed for the phenotypic markers HLA-DR, granzyme B, PD-1, and CTLA-4 on Vα7.2⁺CD161⁺ MAIT cells for one representative blood and decidua sample. Positive gates for HLA-DR, granzyme B and PD-1 were set based on isotype controls, whereas the gate for CTLA-4 was set visually. Expression of (B) HLA-DR, (C) granzyme B, (D) PD-1 and (E) CTLA-4 by CD4⁺ T cells (CD4+), CD8⁺ T cells (CD8⁺) and MAIT cells (MAIT). Results for HLA-DR and PD-1 were based on cell surface staining, whereas granzyme B and CTLA-4 were stained intracellularly. From the CD4⁺ and CD8⁺ T cells, Vα7.2⁺CD161⁺ cells were excluded, i.e., the CD4⁺ and CD8⁺ T cell populations do not comprise any MAIT cells. Results for NK cells based on the same samples are shown in Supplementary Figure S4. Bars show median and IQR from 20-22 donors. Statistical comparisons were performed with Wilcoxon signed-rank test. *** p < 0.001, **** p < 0.0001.

Figure 4

The frequency and phenotype of circulating MAIT cells are unaltered during the 3^rd trimester of pregnancy, but in pregnant women more MAIT cells express the activation marker CD69. PBMCs from women pregnant in the 3^rd trimester (n=26) and from non-pregnant, age-matched controls (n=26) were analyzed by flow cytometry for MAIT cells frequencies, MAIT cell subsets and the expression of activation markers. (A) MAIT cells numbers were analyzed using a gating strategy similar to the one shown in Figure 1A (live CD3⁺Vα7.2⁺CD161^hi cells). (B) Proportions of CD8⁺, CD4⁺ and double-negative (DN) subsets of MAIT cells. (C) Proportions of the CD56⁺ subset of MAIT cells. Cell surface expression of the activation markers (D) CD69, (E) HLA-DR and (F) PD-1. Bars show median and IQR and statistical comparisons were performed using Mann-Whitney test. ** p < 0.01, *** p < 0.001.

Figure 5

Circulating MAIT cells from 3^rd trimester pregnant women exhibit a stronger functional response upon stimulation with microbial and inflammatory stimuli compared with MAIT cells from non-pregnant women. PBMCs isolated from the blood of 3^rd trimester pregnant women (n=14) and from non-pregnant, age-matched controls (n=14) were stimulated overnight with plate-bound anti-CD3 and anti-CD28 antibodies, fixed E coli, fixed group B streptococci (GBS), a combination of IL-12 and IL-18, with influenza A virus (IAV) or left unstimulated (Unstim). IAV stimulations were carried out in a separate plate, together with a separate unstimulated control. MAIT cells in unstimulated and stimulated samples were analyzed for their expression of (A) IFNγ, (B) granzyme B (GrB), (C) PD-1 and (D) CTLA-4. For PD-1, cell surface expression was analyzed, whereas IFNγ, granzyme B and CTLA-4 were stained intracellularly. Results for CD69 from the same experiments are shown in Supplementary Figure S6B. Bars show median and IQR and statistical comparisons between the non-pregnant and pregnant group were performed using Mann-Whitney test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. n = 13-14 for both groups.