Differentiation of ICOS+ and ICOS− recent thymic emigrant regulatory T cells (RTE Tregs) during normal pregnancy, pre‐eclampsia and HELLP syndrome

M I Wagner; M Jöst; J Spratte; M Schaier; K Mahnke; S Meuer; M Zeier; A Steinborn

doi:10.1111/cei.12693

. 2015 Oct 12;183(1):129–142. doi: 10.1111/cei.12693

Differentiation of ICOS⁺ and ICOS⁻ recent thymic emigrant regulatory T cells (RTE T_regs) during normal pregnancy, pre‐eclampsia and HELLP syndrome

M I Wagner ¹, M Jöst ¹, J Spratte ¹, M Schaier ², K Mahnke ³, S Meuer ⁴, M Zeier ², A Steinborn ^1,^✉

PMCID: PMC4687511 PMID: 26285098

Summary

Two different subsets of naturally occurring regulatory T cells (nT_regs), defined by their expression of the inducible co‐stimulatory (ICOS) molecule, are produced by the human thymus. To examine the differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ recent thymic emigrant (RTE) T_regs during normal pregnancy and in the presence of pre‐eclampsia or haemolysis elevated liver enzymes low platelet (HELLP)‐syndrome, we used six‐colour flow cytometric analysis to determine the changes in the composition of the ICOS⁺ and ICOS⁻ T_reg pools with CD45RA⁺CD31⁺ RTE T_regs, CD45RA⁺CD31⁻ mature naive (MN) T_regs, CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs. With the beginning of pregnancy until term, we observed a strong differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE, but not CD45RA⁺CD31⁻ MN T_regs, into CD45RA⁻CD31⁻ memory T_regs. At the end of pregnancy, the onset of spontaneous term labour was associated with a significant breakdown of ICOS⁺CD45RA⁻CD31⁻ memory T_regs. However, in the presence of pre‐eclampsia, there was a significantly increased differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁺ memory T_regs, wherein the lacking differentiation into CD45RA⁻CD31⁻ memory T_regs was partially replaced by the increased differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs into CD45RA⁻CD31⁻ memory T_regs. In patients with HELLP syndrome, this alternatively increased differentiation of CD45RA⁻CD31⁻ MN T_regs seemed to be exaggerated, and presumably restored the suppressive activity of magnetically isolated ICOS⁺ and ICOS⁻ T_regs, which were shown to be significantly less suppressive in pre‐eclampsia patients, but not in HELLP syndrome patients. Hence, our findings propose that the regular differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs ensures a healthy pregnancy course, while their disturbed differentiation is associated with the occurrence of pre‐eclampsia and HELLP syndrome.

Keywords: HELLP syndrome, ICOS⁺ and ICOS⁻ T_regs, immune suppression, pre‐eclampsia, pregnancy

Introduction

In mammalian organisms, the maternal immune system must tolerate the semi‐allogeneic fetus during pregnancy. Numerous murine and human studies have shown that immunosuppressive regulatory T cells (T_regs) have an important role in protecting the fetus from immune‐mediated rejection 1, 2, 3. Interference with respect to their number, differentiation and suppressive function was shown to be responsible for certain pathological conditions, such as infertility 4, 5 and the occurrence of recurrent spontaneous abortions 6, 7, 8, 9. Meanwhile, aberrant T_reg cell homeostasis was also linked to the pathophysiology of most human gestation‐associated diseases that occur in the third trimester and manifest as preterm labour 10, 11, 12, pre‐eclampsia 13, 14, 15, 16 and gestational diabetes 17.

Currently, two different T_reg populations are described in the literature. The naturally occurring T_regs (nT_regs) are induced in the thymus and were shown to suppress immune responses to self‐ and alloantigens 18. The peripherally induced T_regs (iT_regs) are generated by conversion of conventional naive T cells under certain tolerogenic conditions and have the capacity to suppress immune responses towards foreign antigens 19, 20. As it is currently not possible to distinguish between nT_regs and iT_regs phenotypically, it is extremely difficult to recognize the contribution of each T_reg population to tolerance induction during pregnancy and to understand their potential roles for the pathogenesis of the different gestation‐associated diseases. However, functional analysis of different T_reg subsets, which differs with respect to their degree of differentiation, shows that naive CD45RA⁺ T_regs have lower suppressive activity than non‐activated human leucocyte antigen D‐related (HLA‐DR)⁻CD45RA⁻ memory T_regs or activated HLA‐DR⁺CD45RA⁻ memory T_regs 21. Astonishingly, this relation was found to be reversed in healthy pregnant women, whose naive T_reg subset had very high suppressive activity, while the HLA‐DR⁺ memory T_regs were less suppressive 5. These findings suggest that the increase of the suppressive activity of the naive CD45RA⁺ T_reg population may represent a key event in tolerance induction during pregnancy. This assumption is strengthened further by the fact that naive CD45RA⁺ T_regs were shown to be expanded in the periphery during the normal pregnancy course, but were found to be reduced significantly in the periphery of pregnant women suffering from most common pregnancy complications (preterm labour, pre‐eclampsia, gestational diabetes) 11, 12, 17. As naive CD45RA⁺ T_regs always represent nT_regs, it seems that nT_regs are crucial for a successful pregnancy course.

Recently, two different nT_reg populations were detected in the human thymus, which differ concerning their expression of the inducible co‐stimulatory (ICOS) molecule. Their induction by dendritic cells, proliferation and survival were shown to be regulated differentially. It was confirmed that the murine as well as the human T_reg pool consists of a predominant ICOS⁻ T_reg subset, which is much more sensitive to apoptosis, and a minor highly proliferative ICOS⁺ T_reg subset which is resistant to activation‐induced cell death 22, 23.

Like conventional naive CD45RA⁺ T cells, both ICOS⁺ and ICOS⁻ T_regs are released from the thymus as CD31⁺ recent thymic emigrants (RTE T_regs) and develop in a way comparable to conventional RTE T cells. It is known that RTE T cells undergo peripheral post‐thymic proliferation to form a long‐living CD31⁻ mature naive (MN) T cell population, which has the capacity to maintain the naive T cell pool in elderly people. Thereby, both RTE and MN T cells can differentiate into memory T cells after stimulation with appropriate antigens 24. Recently we showed that the normal differentiation of RTE T_regs is changed with the onset of pregnancy. We found that RTE, but not MN T_regs, differentiated increasingly into memory T_regs and that these emerging memory T_regs were maintained during the whole pregnancy course, but disappeared with the onset of spontaneous term labour 25. In the presence of pre‐eclampsia, this differentiation seemed to be impaired, but replaced partly by the increased differentiation of MN T_regs into memory T_regs. Obviously, this phenomenon leads to diminished suppressive activity of the total T_reg pool, which is reduced significantly in pre‐eclampsia patients 11. Currently, it is neither known whether the differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs is regulated differentially during pregnancy, nor whether aberrant differentiation of ICOS⁺ or ICOS⁻CD45RA⁺CD31⁺ RTE T_regs is involved in the pathogenesis of pre‐eclampsia. Therefore, we examined the thymic output and the differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs during the normal pregnancy course and in the presence of gestation‐associated diseases such as pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome.

In this study, we show that normal pregnancy is characterized by the increased differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE but not CD45RA⁺CD31⁻ MN T_regs into CD45RA⁻CD31⁻ memory T_regs. Spontaneous term labour was associated mainly with a sharp decline of ICOS⁺CD45RA⁻CD31⁻ memory T_regs within the total CD4⁺CD127^low+/ ⁻ forkhead box protein 3 (FoxP3)⁺ T_reg pool. Compared to normal pregnancy, the differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁻ memory T_regs was impaired strongly in pre‐eclampsia patients. In contrast, the differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs was relatively normal in patients with HELLP syndrome. However, there was a significantly increased generation of CD45RA⁻CD31⁻ memory cells within the CD4⁺ICOS⁺CD127⁺FoxP3⁻ responder T cell (T_resp) pool.

Materials and methods

Patient collectives and healthy volunteers

Peripheral blood samples were collected from 31 non‐pregnant fertile female volunteers (group 1), 128 pregnant women during the normal pregnancy course (groups 2–5), 41 women with spontaneous term labour (group 6), 45 women 1 day postpartum (group 7), 42 pregnant women affected by pre‐eclampsia (group 8) and 18 women affected with HELLP syndrome (group 9). The diagnosis of pre‐eclampsia was made in the case of blood pressure of more than 140/90 mm Hg occurring on two separate occasions, 6 h apart, along with significant proteinuria (>300 mg/l in a 24‐h collection or a dipstick reading of >1 on a voided random urine sample in the absence of urinary tract infection) in previously normotensive women. The diagnosis of HELLP syndrome was made on the basis of haemolysis, elevated liver enzyme levels (aspartate and alanine aminotransferase >30 U/l) and thrombocytopenia thrombocyte count <150 000/µl). The blood samples from healthy pregnancies (groups 2–5) were collected from women who had routine ultrasonography to exclude fetal malformations (groups 2–4), from women delivering by term elective caesarean section in the absence of labour (group 5), from women in the presence of spontaneous term labour before delivery (group 6) and from women who had delivered spontaneously 1 day before (group 7). The blood samples from women affected by pre‐eclampsia or HELLP syndrome (groups 8 and 9) were taken during their hospitalization. The clinical characteristics of all these women (groups 1–9) are summarized in Table 1. The study was approved by the Regional Ethics Committee. All women were fully informed of the aim of the study and informed consent was obtained from all participants.

Table 1.

Clinical characteristics

Group	Number	Age, median (range)	Weeks’ gestation, median (range)	Diagnosis
1 (non‐pregnant)	31	25 (18–38)	–	Healthy
2 (1st trimester)	31	32 (32–39)	13 (13–14)	Healthy
3 (2nd trimester)	31	31 (21–41)	22 (20–23)	Healthy
4 (3rd trimester)	33	31 (19–42)	31 (27–35)	Healthy
5 (Term, caes. section)	33	32 (20–39)	38 (37–39)	Healthy
6 (Term, spont. delivery)	41	32 (19–38)	40 (38–42)	Healthy
7 (1 day postpartum)	45	30 (19–41)	–	Healthy
8 (Pre‐eclampsia)	42	33 (21–45)	35 (23–41)	Pre‐eclampsia
9 (HELLP syndrome)	18	32 (31–43)	36 (30–40)	HELLP syndrome

Open in a new tab

caes. = caesarean; spont. = spontaneous; HELLP = haemolysis elevated liver enzymes low platelet.

Changes in the composition of the total CD4⁺CD127^low+/ ⁻FoxP3⁺ T_reg pool/CD4⁺CD127⁺FoxP3⁻ T_resp pool with ICOS⁺ and ICOS⁻ T cells were examined during the normal pregnancy course, in the presence of spontaneous term labour and in the presence of pre‐eclampsia or HELLP syndrome. To examine whether the differentiation of ICOS⁺ and ICOS⁻ T_regs/T_resps was regulated differentially during normal pregnancy, or regulated aberrantly in the presence of pre‐eclampsia and HELLP syndrome, we calculated the percentages of CD45RA⁺CD31⁺ RTE cells, CD45RA⁺CD31⁻ MN cells, CD45RA⁻CD31⁺ memory cells and CD45RA⁻CD31⁻ memory cells, both within the total ICOS⁺ and ICOS⁻ T_reg/T_resp pools and within the total T_reg pool. These measurements were performed by six‐colour flow cytometric analysis for all patient groups. Moreover, comparative analyses concerning the suppressive activity of separated ICOS⁺ and ICOS⁻ T_regs were performed for non‐pregnant women (group 1), healthy pregnant women (groups 4 and 5) pre‐eclampsia patients (group 8) and women affected with HELLP syndrome (group 9).

Fluorescence activated cell sorter (FACS) staining

Venous blood samples (10 ml) were collected from all participants into ethylenediamine tetraacetic acid (EDTA)‐containing tubes. Whole peripheral blood mononuclear cells (PBMCs) were isolated by Lymphodex (Inno‐Train Diagnostik GMBH, Kronberg, Germany) gradient centrifugation and analysed by six‐colour flow cytometric analysis. Briefly, the PBMCs (8 × 10⁶ cells) were surface‐stained with 10 µl peridinin chlorophyll (PerCP)‐conjugated anti‐CD4 (BD Bioscience, Heidelberg, Germany), 20 µl phycoerythrin (PE)‐conjugated anti‐ICOS (BD Bioscience), 5 µl PE cyanin (Cy)7‐conjugated anti‐CD127 (eBioscience, Frankfurt, Germany), 5 µl Alexa Fluor 647‐conjugated anti‐CD31 (BD Bioscience) and 5 µl allophycocyanin (APC)‐conjugated anti‐CD45RA (BD Bioscience) mouse monoclonal antibodies. Subsequently, intracellular staining was performed for the detection of FoxP3 using a fluorescein isothiocyanate (FITC)‐labelled anti‐human FoxP3 staining set (clone PCH101; eBioscience), according to the manufacturer's instructions. Negative control samples were incubated with isotype‐matched antibodies. Dead cells were excluded by forward‐ and side‐scatter characteristics. Cells were analysed by a FACS Canto cytometer (BD Bioscience). Statistical analysis was based on at least 100 000 gated CD4⁺ T cells.

Positive selection of CD4⁺CD127^low+/−CD25⁺ T_reg cells

Whole peripheral blood mononuclear cells (PBMCs) were isolated from peripheral blood (50 ml) drawn in EDTA tubes by Lymphodex (Inno‐Train Diagnostik GMBH) gradient centrifugation. CD4⁺CD127^low+/ ⁻CD25⁺ T_reg cells were purified using the Regulatory T Cell Isolation Kit II (Miltenyi Biotec, Bergisch Gladbach, Germany), according to the manufacturer's instructions. First, CD4⁺CD127^low+/ ⁻ T cells were isolated by magnetic depletion of non‐CD4⁺ and CD127^high+ cells. In the second step, the CD4⁺CD127^low+/ ⁻CD25⁺ T_regs were isolated by positive selection over two consecutive columns. The CD4⁺CD127^low+/ ⁻CD25⁻ T cells were obtained in the flow‐through fraction and used as T_resps. The CD4⁺CD127^low+/ ⁻CD25⁺ T_regs were subsequently retrieved from the columns.

Sorting and functional testing of T_reg subsets

For the sorting of the isolated CD4⁺CD127^low+/ ⁻CD25⁺ T_regs into ICOS⁺ and ICOS⁻ T_regs, cells were stained with 20 µl PE‐conjugated anti‐ICOS (BD Bioscience) and 20 µl FITC‐conjugated anti‐CD4 (BD Bioscience) mouse monoclonal antibodies. Total CD4⁺CD127^low+/ ⁻CD25⁺ T_regs were obtained from 11 non‐pregnant fertile women, 10 healthy pregnant women (groups 4 and 5), 11 pregnant women affected by pre‐eclampsia (group 8) and six pregnant women affected by HELLP syndrome (group 9). In all experiments, dead cells were excluded, while the remaining CD4⁺CD127^low+/ ⁻CD25⁺ T_regs were sorted using a FACS Vantage SE Sorter (BD Bioscience).

To analyse the suppressive activity of each isolated T_reg population, 2 × 10⁴ T_resps were co‐cultured with the purified T_reg subsets at ratios of 1 : 2 to 1 : 256 in 96‐well U‐bottomed plates. Suppression assays were performed in a final volume of 100 μl/well of X‐VIVO15 medium (Lonza, Verviers, Belgium). For T cell stimulation, the medium was supplemented with 1 μg/ml anti‐CD3 and 2 μg/ml anti‐CD28 antibodies (eBioscience). As controls, CD4⁺CD127^low+/ ⁻CD25⁺ T_regs and T_resps alone were cultured both with and without stimulus. Cells were incubated at 37 °C in 5% CO₂. After 4 days, 1 μCi [³H]‐thymidine (Hartmann Analytic, Braunschweig, Germany) was added to the cultures and cells were incubated for a further 16 h. Then, the cells were harvested and [³H] incorporation was measured by scintillation counting. According to the cell numbers achieved after cell sorting, the assays were performed as single or multiple determinations. In order to compare the suppressive activity of the different T_reg subsets in non‐pregnant women, healthy pregnant women and pre‐eclampsia patients, the maximum suppressive activity (ratio of T_regs to T_resps 1/2) and the minimum ratio of T_regs to T_resps at which a suppression of at least 15% could be achieved were calculated 11.

Statistical analysis

Statistical comparison of the percentages of ICOS⁺ and ICOS⁻ T_regs/T_resps within the total T_reg/T_resp pool between the different patient groups was performed using the non‐parametric Wilcoxon–Mann–Whitney U‐test. This test was also used for statistical comparison of the percentages of CD45RA⁺CD31⁺ RTE cells, CD45RA⁺CD31⁻ MN cells, CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory cells within the ICOS⁺ and ICOS⁻ T_reg/T_resp pool. Comparison of the suppressive activity of the different T_reg subsets between non‐pregnant women, healthy pregnant women and pre‐eclampsia or HELLP syndrome patients was also performed using the non‐parametric Wilcoxon–Mann–Whitney U‐test. P < 0·05 was considered significant. For all tests, the software package BiAS for Windows (version 10·06) was used.

Results

ICOS⁺ and ICOS⁻ T_regs show similar differentiation during the normal pregnancy course

In this study we examined whether there were differences in the differentiation of ICOS⁺ and ICOS⁻ T_regs/T_resps during the normal pregnancy course and whether the presence of pre‐eclampsia or HELLP syndrome was associated with deficient differentiation of either ICOS⁺ or ICOS⁻ T_regs/T_resps compared to healthy pregnancies. Therefore, we first divided the total CD4⁺CD127^low+/ ⁻FoxP3⁺ T_reg pool/CD4⁺CD127⁺FoxP3⁻ T_resp pool into ICOS⁺ and ICOS⁻ T_regs/T_resps and then determined the percentages of CD45RA⁺CD31⁺ RTE and CD45RA⁺CD31⁻ MN T_regs/T_resps, as well as CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs/T_resps within both the total ICOS⁺ and ICOS⁻ T_reg/T_resp pools. Figure 1a–h shows the gating strategy that was used in all experiments. These measurements were performed with venous blood samples of non‐pregnant fertile women (group 1), healthy non‐labouring pregnant women during the normal pregnancy course (groups 2–5), spontaneously term labouring women (group 6) and women having delivered spontaneously 1 day before (group 7). The data obtained from healthy non‐labouring third‐trimester women (groups 4 and 5) were compared with third‐trimester women affected by pre‐eclampsia (group 8) or HELLP syndrome (group 9). Table 1 summarizes the clinical characteristics of all participants in this study.

Gating strategy for six‐colour flow cytometric detection of CD45RA⁺CD31⁺ recent thymic emigrant (RTE), CD45RA⁺CD31⁻mature naive (MN) T_regs, CD45RA⁻CD31⁺ memory T cells and CD45RA⁻CD31⁻ memory T cells within the inducible co‐stimulatory (ICOS)⁺ and ICOS⁻ regulatory T cell (T_reg) pool, as well as within the ICOS⁺ and ICOS⁻ responder T cell (T_resp) pool. At first, CD4⁺ T cells (P1) were gated by fluorescence intensity of CD4 *versus* side‐scatter (SSC) (a). The CD4⁺CD127^low+/− forkhead box protein 3 (FoxP3)⁺ T_regs (P2) and the CD4⁺CD127⁺FoxP3⁻ T_resps (P13) were gated by fluorescence intensity of FoxP3 *versus* CD127 (b). The ICOS⁺ (P3) and ICOS⁻ T_regs (P4) were gated by fluorescence intensity of CD45RA *versus* ICOS (c). The percentages of CD45RA⁺CD31⁺ RTE T_regs (P5, P9), CD45RA⁺CD31⁻ MN T_regs (P6, P10), CD45RA⁻ CD31⁺ memory T_regs (P7, P11) and CD45RA⁻ CD31⁻ memory T_regs (P8, P12) were estimated by analysing the total ICOS⁺ (P3) and ICOS⁻ T_reg pool (P4) for its expression of CD45RA and CD31 (d,e). Moreover, the ICOS⁺ (P14) and ICOS⁻ T_resps (P15) were gated by fluorescence intensity of CD45RA *versus* ICOS (f). The percentages of CD45RA⁺CD31⁺ RTE T_resps (P16, P20), CD45RA⁺CD31⁻ MN T_resps (P17, P21), CD45RA⁻ CD31⁺ memory T_resps (P18, P22) and CD45RA⁻ CD31⁻ memory T_resps (P19, P23) were estimated by analysing the total ICOS⁺ (P14) and ICOS⁻ T_resp pool (P15) for its expression of CD45RA and CD31 (g,h).

We found that the onset of normal pregnancy was associated with a significant decrease of CD45RA⁺CD31⁺ RTE (Fig. 2a,e), as well as CD45RA⁻CD31⁺ memory T_regs (Fig. 2c,g) within both the ICOS⁺ and the ICOS⁻ T_reg pools. The percentages of CD45RA⁺CD31⁻ MN T_regs did not change, neither within the ICOS⁺ nor within the ICOS⁻ T_reg pool (Fig. 2b,f). However, there was a strong increase in the percentage of CD45RA⁻CD31⁻ memory T_regs within both T_reg pools (Fig. 2d,h). During the further pregnancy course (groups 2–5) this situation was almost maintained until term, when with the onset of normal spontaneous term labour (group 6) the original distribution of CD45RA⁺CD31⁺ RTE T_regs, CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs was restored according with that of non‐pregnant women (group 1). Surprisingly, after spontaneous term delivery (group 7), there was a new, significant, decrease of CD45RA⁺CD31⁺ RTE and CD45RA⁻CD31⁺ memory T_regs (Fig. 2a,c) and a complementary increase of CD45RA⁻CD31⁻ memory T_regs within the ICOS⁺ T_reg pool (Fig. 2d). A similar decrease of CD45RA⁺CD31⁺ RTE and CD45RA⁻CD31⁺ memory T_regs was also observed within the ICOS⁻ T_reg pool (Fig. 2e,g), but instead of the CD45RA⁻CD31⁻ memory T_regs, the CD45RA⁺CD31⁻ MN T_regs increased complementary postpartum (Fig. 2f). Supporting information, Fig. S1 also shows the confidence intervals for the median for these results.

Changes in the differentiation of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻ regulatory T cells (T_regs) during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentages of CD45RA⁺CD31⁺ recent thymic emigrant (RTE) T_regs (a,e), CD45RA⁺CD31⁻mature naive (MN) T_regs (b,f), CD45RA⁻CD31⁺ memory T_regs (c,g) and CD45RA⁻CD31⁻ memory T_regs (d,h) within the ICOS⁺ and ICOS⁻ T_reg pool were estimated in healthy non‐pregnant fertile women (group 1, ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). Significant differences concerning the percentages of the different T_reg subsets were detected between women during the normal pregnancy course (groups 2–5) and non‐pregnant women (group 1) or women with spontaneous term labour (group 6). Significant differences were also observed between spontaneously term labouring women (group 6) and women 1 day postpartum (group 7). In addition, significant differences were detected between healthy third‐trimester women (groups 4 and 5) and women with pre‐eclampsia (group 8) or HELLP syndrome (group 9).

Inline graphic — Changes in the differentiation of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻ regulatory T cells (T_regs) during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentages of CD45RA⁺CD31⁺ recent thymic emigrant (RTE) T_regs (a,e), CD45RA⁺CD31⁻mature naive (MN) T_regs (b,f), CD45RA⁻CD31⁺ memory T_regs (c,g) and CD45RA⁻CD31⁻ memory T_regs (d,h) within the ICOS⁺ and ICOS⁻ T_reg pool were estimated in healthy non‐pregnant fertile women (group 1, ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). Significant differences concerning the percentages of the different T_reg subsets were detected between women during the normal pregnancy course (groups 2–5) and non‐pregnant women (group 1) or women with spontaneous term labour (group 6). Significant differences were also observed between spontaneously term labouring women (group 6) and women 1 day postpartum (group 7). In addition, significant differences were detected between healthy third‐trimester women (groups 4 and 5) and women with pre‐eclampsia (group 8) or HELLP syndrome (group 9).

In summary, these data show that the onset of pregnancy is associated with a decreased thymic distribution of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs and that the already distributed ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs differentiate increasingly into CD45RA⁻CD31⁻ memory T_regs. This situation seems to be maintained during the entire healthy pregnancy course. With the onset of spontaneous term labour, the percentage of CD45RA⁻CD31⁻ memory T_regs declined, while the percentage of CD45RA⁻CD31⁺ memory T_regs increased strongly. Furthermore, there was a significant redistribution of CD45RA⁺CD31⁺ RTE T_regs, both within the ICOS⁺ and ICOS⁻ T_reg pools.

The decrease of ICOS⁺CD45RA⁻CD31⁻ but not ICOS⁻CD45RA⁻CD31⁻ memory T_regs is crucial for the onset of spontaneous term labour

To examine whether the onset of normal spontaneous term labour was associated primarily with a loss of either ICOS⁺ or ICOS⁻ T_regs, we estimated their percentages within the total CD4⁺CD127^low+/ ⁻FoxP3⁺ T_reg pool during the whole pregnancy course. In addition, we examined whether there were considerable changes in the composition of the total CD4⁺CD127^low+/ ⁻FoxP3⁺ T_reg pool with CD45RA⁺CD31⁺ RTE and CD45RA⁺CD31⁻ MN T_regs, as well as CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs, each of ICOS⁺ and ICOS⁻ T_regs (Fig. 3). We found that total T_regs decreased continuously (Fig. 3a) and that the percentages of ICOS⁺ and ICOS⁻ T_regs did not change markedly during the normal pregnancy course (Fig. 3b,c). However, the presence of spontaneous term labour was associated with a significant decrease of ICOS⁺ T_regs and a complementary increase of ICOS⁻ T_regs (Fig. 3b,c). Thereby, the percentage of ICOS⁺CD45RA⁻CD31⁻ memory T_regs decreased strongly (Fig. 3g), while the percentage of ICOS⁻CD45RA⁻CD31⁻ memory T_regs did not change perceptibly (Fig. 3k). Instead, there was a significant increase in the percentages of ICOS⁻CD45RA⁺CD31⁺ RTE and ICOS⁻CD45RA⁻CD31⁺ memory T_regs (Fig. 3h,j). Supporting information, Fig. S2 also shows the confidence intervals for the median for these results. Therefore, it seems that the occurrence of normal spontaneous term labour is characterized mainly by a significant decrease of ICOS⁺CD45RA⁻CD31⁻ memory T_regs and a significant redistribution of ICOS⁻CD45RA⁺CD31⁺ RTE T_regs within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool.

Detection of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻CD45RA⁺CD31⁺ recent thymic emigrant (RTE) regulatory T cells (T_regs), ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ mature naive (MN) T_regs, ICOS⁺ and ICOS⁻CD45RA⁻CD31⁺ memory T_regs and ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs within the total CD4⁺CD127^low+/−forkhead box protein 3 (FoxP3)⁺ T_reg pool during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentage of total CD4⁺CD127^low+/−FoxP3⁺ T_regs within CD4⁺ T cells (a), the percentages of ICOS⁺ (b) and ICOS⁻ T_regs (c), ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs (d,h), ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs (e,i), ICOS⁺ and ICOS⁻CD45RA⁻CD31⁺ memory T_regs (f,j) and ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs (g,k) within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool were determined in healthy non‐pregnant fertile women (group 1, ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). Significant differences concerning the percentages of the different T_reg subsets were detected between women during the normal pregnancy course (groups 2–5) and non‐pregnant women (group 1) or women with spontaneous term labour (group 6). In addition, significant differences were detected between healthy third‐trimester women (groups 4 and 5) and women with pre‐eclampsia (group 8) or HELLP syndrome (group 9).

Pre‐eclampsia patients show a significantly increased differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺RTE T_regs into CD45RA⁻CD31⁺ instead of CD45RA⁻CD31⁻ memory T_regs

To examine whether ICOS⁺ or ICOS⁻ T_regs show deficient differentiation in the presence of pre‐eclampsia or HELLP syndrome, we determined the percentages of CD45RA⁺CD31⁺ RTE and CD45RA⁺CD31⁻ MN T_regs, as well as CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs, within the ICOS⁺ and the ICOS⁻ T_reg pool of both healthy third‐trimester women (groups 4 and 5) and patients affected with pre‐eclampsia (group 8) or HELLP syndrome (group 9). We found that the differentiation of ICOS⁺ and ICOS⁻ T_regs was disturbed similarly in pre‐eclampsia patients (Fig. 2a–h). Compared to healthy pregnancies, the percentages of CD45RA⁺CD31⁺ RTE and CD45RA⁺CD31⁻ MN T_regs (Fig. 2a,b,e,f) were reduced significantly, while the percentages of CD45RA⁻CD31⁺ memory T_regs (Fig. 2c,g) were increased significantly in both T_reg pools. The only difference was found in the ICOS⁺ T_reg pool, as the percentages of CD45RA⁻CD31⁻ memory T_regs were decreased significantly in pre‐eclampsia patients (Fig. 2d), but unchanged in the ICOS⁻ T_reg pool (Fig. 2h). Surprisingly, compared to healthy pregnancies, HELLP syndrome patients showed only significantly decreased percentages of CD45RA⁺CD31⁻ MN T_regs within the ICOS⁺ T_reg pool as well as within the ICOS⁻ T_reg pool (Fig. 2b,f). The percentages of CD45RA⁺CD31⁺ RTE T_regs, CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs again reached values in the range of healthy pregnancies in both T_reg pools.

The total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool was not decreased in the presence of pre‐eclampsia and HELLP syndrome (Fig. 3a). When detecting total ICOS⁺ and ICOS⁻ T_regs within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool, we found that the relationship between these two T_reg subsets was not changed (Fig. 3b,c). Rather, we could see that the deviations in the differentiation of the ICOS⁺ and ICOS⁻ T_reg pool in pre‐eclampsia and HELLP syndrome patients were also observable when detecting ICOS⁺ and ICOS⁻ T_regs as CD45RA⁺CD31⁺ RTE, CD45RA⁺CD31⁻ MN, CD45RA⁻CD31⁺ memory and CD45RA⁻CD31⁻ memory T_regs within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool (Fig. 3d–k). The percentages of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs (Fig. 3d,h), as well as ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs (Fig. 3e,i), were reduced significantly, while the percentages of ICOS⁺ and ICOS⁻CD45RA⁻CD31⁺ memory T_regs (Fig. 3f,j) were increased significantly in the presence of pre‐eclampsia. However, the percentages of ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs remained unchanged within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool. In the presence of HELLP syndrome, we only detected significantly reduced percentages of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool (Fig. 3e,i). All other T_reg subsets again reached values in the range of healthy pregnancies. Supporting information, Figs S1 and S2 show the confidential intervals for these results.

These data suggest that the occurrence of pre‐eclampsia is associated with an increased differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁺ memory T_regs which presumably do not proliferate enough to increase CD45RA⁻CD31⁻ memory T_regs. As the CD45RA⁻CD31⁻ memory T_regs are not reduced in these patients, it seems that there is an alternatively increased differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs into CD45RA⁻CD31⁻ memory T_regsin the presence of pre‐eclampsia. With the occurrence of HELLP syndrome, this increased differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs into CD45RA⁻CD31⁻ memory T_regs seems to be exaggerated, and presumably induces an increasingly enhanced differentiation of previously non‐activated ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁺ memory T_regs with stronger proliferative capacity. Thus, the percentages of CD45RA⁺CD31⁺ RTE T_regs, CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T_regs seem to adjust again to normal values detected in healthy pregnancies.

ICOS⁺ and ICOS⁻ T_regs have similar suppressive activity that is reduced strongly in patients affected with pre‐eclampsia, but not HELLP syndrome

To examine whether the suppressive activity of ICOS⁺ and ICOS⁻ T_regs decreases in the presence of pre‐eclampsia or HELLP syndrome, the total CD4⁺CD127^low+/−CD25⁺ T_reg pool obtained from 11 non‐pregnant fertile women (group 1), 10 healthy pregnant women (groups 4 and 5), 11 pregnant women affected by pre‐eclampsia (group 8) and six women affected by HELLP syndrome (group 9) was isolated by magnetic activated cell sorting (MACS) and sorted into ICOS⁺ and ICOS⁻ T_regs (Fig. 4a). Subsequently, both ICOS⁺ (Fig. 4a, P1) and ICOS⁻ T_regs (Fig. 4a, P2) obtained from each women were analysed separately for their suppressive capacity. Figure 4b–e shows the results of the maximum suppressive activity (ratio T_regs/T_resps = 1/2) and the ratio of the T_regs/T_resps up to which the T_regs could be diluted to achieve a minimum suppressive activity of at least 15%. Depending on the number of the separated ICOS⁺ and ICOS⁻ T_reg cells the suppression assays were performed as single or multiple determinations. Both parameters did not differ between ICOS⁺ and ICOS⁻ T_regs, either in non‐pregnant or healthy pregnant women. There were also no differences concerning these parameters between non‐pregnant women and healthy pregnant women. However, ICOS⁺ and ICOS⁻ T_regs showed a significantly decreased suppressive activity regarding both parameters in pre‐eclampsia patients compared to non‐pregnant and pregnant women. In contrast, both ICOS⁺ and ICOS⁻ T_regs obtained from patients with HELLP syndrome showed a similar high suppressive activity as detected in healthy pregnant and non‐pregnant women (Fig. 4b–e).

Cell sorting and suppressive activity of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻ regulatory T cells (T_regs). CD4⁺CD127^low+/−CD25⁺ T_regs were isolated by magnetic activated cell sorting (MACS), stained with anti‐CD4 and anti‐ICOS‐specific monoclonal antibodies and sorted into ICOS⁺ (P1) and ICOS⁻ T_regs (P2) (a). The suppressive activity of the different ICOS⁺ and ICOS⁻ T_reg subsets was estimated by suppression assays. Blood samples were obtained from non‐pregnant women (), healthy third‐trimester women (groups 4 and 5) (◆), pre‐eclampsia patients () and women with haemolysis elevated liver enzymes low platelet (HELLP) syndrome (). The figure shows the individual and median values of the maximum suppressive activity (T_reg/responder T cell (T_resp) = 1/2) (b,c) and of the ratio of T_reg/T_resp up to which the purified T_reg cells could be diluted to achieve a minimum suppressive activity of at least 15% (d,e). Compared to non‐pregnant and healthy third‐trimester women, the suppressive activity of both ICOS⁺ and ICOS⁻ T_regs was decreased significantly in pre‐eclampsia patients, but not in HELLP syndrome patients.

These findings suggest that the increased differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁺ instead of CD45RA⁻CD31⁻ memory T_regs has a negative influence on the suppressive activity of both the ICOS⁺ and ICOS⁻ T_reg pools. As the percentages of all T_reg subsets, with the exception of the CD45RA⁺CD31⁻ MN T_regs, were found to be in the normal range in patients with HELLP syndrome, it seems that the reinforced differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into sufficiently proliferating CD45RA⁻CD31⁺ memory T_regs causes the restoration of the suppressive activity of both T_reg pools.

ICOS⁺ and ICOS⁻ T_resp cells differentiate similar to ICOS⁺ and ICOS⁻ T_regs during the normal pregnancy course

Regarding the differentiation of T_resps, we also could not detect any shift in the composition of the total CD4⁺CD127⁺FoxP3⁻ T_resp pool with ICOS⁺ or ICOS⁻ T_resps during the normal pregnancy course (Fig. 5a,b). We could even see a similar differentiation of the ICOS⁺ and ICOS⁻ T_resps, whose CD45RA⁺CD31⁺ RTE and CD45RA⁻CD31⁺ memory T cells also decreased strongly with the beginning of pregnancy (Fig. 5c,g,e,i). Their percentages of CD45RA⁺CD31⁻ MN T cells either did not change (Fig. 5d) or even rose (Fig. 5h), while their CD45RA⁻CD31⁻ memory T cells increased significantly (Fig. 5f,j). Similar to T_reg cell differentiation, this condition was maintained during the entire pregnancy course. However, with the onset of spontaneous term labour, the original distribution of CD45RA⁺CD31⁺ RTE, CD45RA⁻CD31⁺ and CD45RA⁻CD31⁻ memory T cells was restored for ICOS⁻ T_resps (Fig. 5g,i,j), but not for ICOS⁺ T_resps (Fig. 5c,e,f). These findings reveal that beside the increased generation of ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs, there is also an increased generation of ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_resps during the normal pregnancy course. With the beginning of spontaneous term labour, ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs, as well as ICOS⁻CD45RA⁻CD31⁻ memory T_resps, diminish while the ICOS⁺CD45RA⁻CD31⁻ memory T_resps do not decrease. Therefore, it appears that the continued existence of ICOS⁺CD45RA⁻CD31⁻ memory T_resps may account for the induction of normal spontaneous term labour.

Changes in the differentiation of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻responder T cells (T_resps) during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentages of ICOS⁺ (a) and ICOS⁻ T_resps (b) within the total T_resp pool, as well as CD45RA⁺CD31⁺RTE T_resps (c,g), CD45RA⁺CD31⁻MN T_resps (d,h), CD45RA⁻CD31⁺ memory T_resps (e,i) and CD45RA⁻CD31⁻ memory T_resps (f,j) within the ICOS⁺ and ICOS⁻ T_resp pool were estimated in healthy non‐pregnant fertile women (group 1, ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). Significant differences concerning the percentages of the different T_resp subsets were detected between women during the normal pregnancy course (groups 2–5) and non‐pregnant women (group 1) or women with spontaneous term labour (group 6). In addition, significant differences were detected between healthy third‐trimester women (groups 4 and 5) and women with pre‐eclampsia (group 8) or HELLP syndrome (group 9).

Surprisingly, the presence of pre‐eclampsia and HELLP syndrome was associated with a clear shift in favour of ICOS⁻ T_resps within the total CD4⁺CD127⁺FoxP3⁻ T_resp pool (Fig. 5a,b). Thereby, it was striking that the changes concerning the composition of the total ICOS⁺ T_resp pool were the same as those found in the ICOS⁺ T_reg pool. In the presence of pre‐eclampsia, the percentages of CD45RA⁺CD31⁺ RTE and CD45RA⁺CD31⁻ MN T cells decreased, while the percentages of CD45RA⁻CD31⁺ memory T cells increased significantly (Fig. 5c–e). An increased percentage of CD45RA⁻CD31⁺ memory T cells was also found within the ICOS⁻ T_resp pool (Fig. 5i). Therefore, it seems that particularly the ICOS⁺CD45RA⁺CD31⁺ RTE T_resps and presumably also the ICOS⁻CD45RA⁺CD31⁺ RTE T_resps show an increased differentiation into CD45RA⁻CD31⁺ instead of CD45RA⁻CD31⁻ memory T_resps in the presence of pre‐eclampsia. In contrast, there was a significantly increased differentiation of ICOS⁺CD45RA⁺CD31⁺ RTE T_resps into ICOS⁺CD45RA⁻CD31⁻ memory T_resps in the presence of HELLP syndrome (Fig. 5c,d,f). Supporting information, Fig. S3 also shows the confidential intervals for these results.

In summary, these findings suggest that an adequate differentiation of both T_reg and T_resp cells may be necessary for a successful pregnancy course. Obviously, the continued or enhanced generation of especially ICOS⁺CD45RA⁻CD31⁻ memory T_resps plays an important role in spontaneous term delivery or HELLP syndrome.

Discussion

Our previous reports demonstrated that thymus‐derived naturally occurring naive CD45RA⁺ T_regs are of potential importance for successful pregnancy course 5, 10, 11, 12, 17. Due to the fact that the human thymus contains two subsets of nT_regs, defined by their expression of the co‐stimulatory molecule ICOS, we examined whether the thymic output of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs and their subsequent differentiation in the periphery were regulated differentially during pregnancy and whether aberrant differentiation could be observed in the presence of pre‐eclampsia or HELLP syndrome. For that, we compared the changes in the composition of the ICOS⁺ and the ICOS⁻ T_reg pool with CD45RA⁺CD31⁺ RTE, CD45RA⁺CD31⁻ MN, CD45RA⁻CD31⁺ memory and CD45RA⁻CD31⁻ memory T_regs during normal and complicated pregnancy and examined whether the characteristic changes observed within the ICOS⁺ and the ICOS⁻ T_reg pool could be confirmed when detecting the different ICOS⁺ and ICOS⁻ T_reg subsets within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool. In addition, we examined whether there were similar changes in the composition of the total CD4⁺CD127⁺FoxP3⁻ T_resp pool with the corresponding T_resp subsets.

We found that both CD45RA⁺CD31⁺ RTE and CD45RA⁻CD31⁺ memory T_regs decreased strongly with the onset of pregnancy, both within the ICOS⁺ and ICOS⁻ T_reg pools, indicating that the thymic output of both ICOS⁺ and ICOS⁻ RTE T_regswas greatly reduced. It is already known that the thymic activity is restricted severely during puberty in both men and women (thymic involution), due to the increased secretion of gonadal hormones 26. Therefore, it seems likely that the excessive production of pregnancy hormones may cause an additional reduction of the thymic activity, which seems to be limited temporally for the duration of pregnancy. In mice, such transient thymic involution during pregnancy has already been shown and it was demonstrated that progesterone and oestrogens had the capacity to decrease thymocyte proliferation at the DN stage without inducing apoptosis 26, 27.

Our findings show that the decrease of CD45RA⁺CD31⁺ RTE and CD45RA⁻CD31⁺ memory T_regs at the onset of pregnancy was accompanied by a concurrent increase of CD45RA⁻CD31⁻ memory T_regs, while CD45RA⁺CD31⁻ MN T_regs did not change, both within the ICOS⁺ and the ICOS⁻ T_reg pools. Therefore, we propose that both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs, but not CD45RA⁺CD31⁻ MN T_regs, differentiate increasingly into CD45RA⁻CD31⁻ memory T_regs. Presumably, the restricted thymic release of RTE T_regs at the onset of pregnancy may cause their enhanced proliferation and force their direct differentiation into CD45RA⁻CD31⁻ memory T_regs during the entire pregnancy course. At the end of pregnancy, these newly generated populations of both ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs were found to break down with the onset of normal spontaneous term labour. This phenomenon may be explained by the fact that both naive ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs may not have the capacity to proliferate indefinitely. Therefore, both ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs may decline at the end of pregnancy, when the ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs lose the capacity to proliferate any longer.

Although ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs differentiated similarly during the pregnancy course, we found a significant shift in the proportion of ICOS⁺ and ICOS⁻ T_regs within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool in favour of ICOS⁻ T_regs with the onset of normal spontaneous term labour. This was due mainly to a strong decrease of ICOS⁺CD45RA⁻CD31⁻ memory T_regs and a potential increase of ICOS⁻CD45RA⁺CD31⁺ RTE and ICOS⁻CD45RA⁻CD31⁺ memory T_regs. As the ICOS⁺ T_regs were shown to have a greater capacity for proliferation than ICOS⁻ T_regs 22, 23, the loss of ICOS⁺CD45RA⁻CD31⁻ memory T_regs may have a greater impact on the composition of the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool than the loss of ICOS⁻CD45RA⁻CD31⁻ memory T_regs. Therefore, our findings may propose that spontaneous term delivery may be induced primarily by the breakdown of ICOS⁺CD45RA⁻CD31⁻ memory T_regs, but not ICOS⁻CD45RA⁻CD31⁻ memory T_regs. Clearly, there is a significant thymic redistribution of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs with the onset of spontaneous term labour. Thereby, the redistribution of ICOS⁻CD45RA⁺CD31⁺ RTE T_regs may be much more noticeable than that of ICOS⁺CD45RA⁺CD31⁺ RTE T_regs, as these cells account for only a vanishingly small fraction within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool.

Interestingly, after delivery, these newly released CD45RA⁺CD31⁺ RTE T_regs differentiated preferably into CD45RA⁺CD31⁻ MN T_regs in the ICOS⁻ T_reg pool but into CD45RA⁻CD31⁻ memory T_regs in the ICOS⁺ T_reg pool. Such findings may propose that placental hormones such as oestrogens and progesterone may be necessary for the direct differentiation of ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into ICOS⁻CD45RA⁻CD31⁻ memory T_regs and that the postpartum loss of these hormones may, rather, favour the development of ICOS⁻CD45RA⁺CD31⁻ MN T_regs instead of ICOS⁻CD45RA⁻CD31⁻ memory T_regs. In contrast, it seems that there is a hormone‐independent differentiation of ICOS⁺CD45RA⁺CD31⁺ RTE T_regs into ICOS⁺CD45RA⁻CD31⁻ memory T_regs. A hormone‐dependent differentiation was already reported for conventional T cells, as it was shown that there is a profound loss of naive T cells during puberty, accompanied by a strong expansion of T cells with memory phenotype 28, 29, 30, 31. In addition, it was shown in mice that ovarian hormone ablation leads to phenotypical alterations in the major peripheral blood lymphocyte (PBL) and splenic T cell subsets by diminishing the peripubertal changes in the frequency of RTE, MN and memory T cells 32.

In contrast to spontaneous term delivery, the presence of pre‐eclampsia or HELLP syndrome was not associated with a significant shift in the proportion of ICOS⁺ and ICOS⁻ T_regs within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool. However, there was an increased differentiation of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁺ memory T_regs instead of CD45RA⁻CD31⁻ memory T_regs, indicating that both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs may have deficiencies concerning their proliferative capacity. As these patients showed decreased percentages of both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs, the deficient differentiation of CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁻ memory T_regs may have been replaced partially by the increased differentiation of CD45RA⁺CD31⁻ MN T_regs into CD45RA⁻CD31⁻ memory T_regs. Nevertheless, pre‐eclampsia patients showed decreased suppressive activity, both of their ICOS⁺ and ICOS⁻ T_reg pools. In contrast, the suppressive activity of both T_reg pools was in the normal range in patients with HELLP syndrome, a phenomenon which was confirmed when testing the suppressive activity of the total T_reg pool of these patients 11. As these patients showed only decreased percentages of CD45RA⁺CD31⁻ MN T_regs, both within the ICOS⁺ and the ICOS⁻ T_reg pool and within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool, we propose that the alternatively increased differentiation of CD45RA⁺CD31⁻ MN T_regs into CD45RA⁻CD31⁻ memory T_regs in pre‐eclampsia patients may be exaggerated in patients with HELLP syndrome. The strong decrease of CD45RA⁺CD31⁻ MN T_regs could induce the additional differentiation of so far non‐activated CD45RA⁺CD31⁺ RTE T_regs and could cause their further reinforced differentiation into CD45RA⁻CD31⁻ memory T_regs with stronger proliferation capacity. Presumably, this reinforced differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁻ memory T_regs ensures that the suppressive activity of both T_reg pools is restored in patients with HELLP syndrome. Nevertheless, such mechanisms may constitute an emergency response causing a large consumption of functional RTE T_regs.

In addition, our results reveal that the reduction of thymic activity during pregnancy also affects the differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_resp cells. It was striking that similar to T_reg cells, both ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_resps, but not ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_resps, differentiated into ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_resps during the normal pregnancy course. Currently, we are not able to explain how these mechanisms are realized. There are some indications in the literature that the CD31 molecule itself could play a role in the post‐thymic homeostatic proliferation of conventional T_resps, and presumably of T_regs as well 24. Similar to ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs, the population of ICOS⁻CD45RA⁻CD31⁻ memory T_resps, but not that of ICOS⁺CD45RA⁻CD31⁻ memory T_resps, broke down with the onset of normal spontaneous term labour. Such findings suggest that ICOS⁺CD45RA⁺CD31⁺ RTE T_resps may have the ability to proliferate for longer than ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs or ICOS⁻CD45RA⁺CD31⁺ RTE T_resps. Presumably, at the end of pregnancy these conditions are responsible for the development of normal spontaneous term labour.

Regarding the presence of pre‐eclampsia, our results propose that the proliferation capacity of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs as well as ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_resps is impaired, so that both the suppressive effect of T_regs and the repellent effect of T_regs may be disturbed. However, in the presence of HELLP syndrome, there was a significant excessive production of ICOS⁺CD45RA⁻CD31⁺ memory T_resps, whose effector functions could, presumably, have caused the reinforced differentiation of ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs into CD45RA⁻CD31⁻ memory T_regs. Meanwhile, it is known that the ICOS molecule is expressed on follicular helper T cells (Tfh). These cells are necessary for B cell antibody isotype‐switching, germinal centre (GC) formation and high‐affinity antibody production 33. Recently, the frequency of these cells was found to be increased significantly in patients with autoimmune diseases such as systemic lupus erythematosus (SLE) and ankylosing spondylitis 34, 35. In addition, ICOS signalling was shown to be required for the generation of both central and effector CD4⁺ memory T cell populations 36. Currently, it is not known whether ICOS⁺ and ICOS⁻ T_resps also differ concerning their induction, proliferation and survival. Further studies may be necessary to clarify the special role of ICOS⁺CD45RA⁻CD31⁻ memory T_regs/T_resps for tolerance induction during healthy and affected pregnancies.

Author contributions

M. W. and A. S. designed the study, M. W. and M. J. performed the study, J. S., M. S., K. M., S. M. and M. Z. contributed important methods and patients. M. W. and A. S. analysed the data and wrote the manuscript. All authors contributed to the final version of the manuscript and approved it.

Disclosure

None of the authors have any conflicts of interest related to this manuscript.

Supporting information

Additional Supporting information may be found in the online version of this article at the publisher's web site:

Fig. 1. Changes in the differentiation of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻ regulatory T cells (T_regs) during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentages of CD45RA⁺CD31⁺ recent thymic emigrant (RTE) T_regs (a,e), CD45RA⁺CD31⁻ mature naive (MN) T_regs (b,f), CD45RA⁻CD31⁺ memory T_regs (c,g) and CD45RA⁻CD31⁻ memory T_regs (d,h) within the ICOS⁺ and ICOS⁻ T_reg pool were estimated in healthy non‐pregnant fertile women (group 1, Inline graphic ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). The figures show the 95% confidence intervals and the median of all patient groups for each T_reg subset.

Click here for additional data file.^{(372.5KB, ppt)}

Fig. 2. Detection of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻CD45RA⁺CD31⁺recent thymic emigrant (RTE) T_regs, ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ mature naive (MN) T_regs, ICOS⁺ and ICOS⁻CD45RA⁻CD31⁺ memory T_regs and ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ T_regs within the total CD4⁺CD127^low+/−forkhead box protein 3 (FoxP3)⁺ T_reg pool during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentage of total CD4⁺CD127^low+/−FoxP3⁺ T_regs within CD4⁺ T cells (a), the percentages of ICOS⁺ (b) and ICOS⁻ T_regs (c), ICOS⁺ and ICOS⁻CD45RA⁺CD31⁺ RTE T_regs (d,h), ICOS⁺ and ICOS⁻CD45RA⁺CD31⁻ MN T_regs (e,i), ICOS⁺ and ICOS⁻CD45RA⁻CD31⁺ memory T_regs (f,j) and ICOS⁺ and ICOS⁻CD45RA⁻CD31⁻ memory T_regs (g,k) within the total CD4⁺CD127^low+/−FoxP3⁺ T_reg pool were determined in healthy non‐pregnant fertile women (group 1, Inline graphic ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). The figures show the 95% confidence intervals and the median of all patient groups for each T_reg subset.

Click here for additional data file.^{(449.5KB, ppt)}

Fig. 3. Changes in the differentiation of inducible co‐stimulatory (ICOS)⁺ and ICOS⁻ responder T cells (T_resps) during the normal pregnancy course and in the presence of pre‐eclampsia and haemolysis elevated liver enzymes low platelet (HELLP) syndrome. The percentages of ICOS⁺ (a) and ICOS⁻ T_resps (b) within the total T_resp pool, as well as CD45RA⁺CD31⁺ recent thymic emigrant (RTE) T_resps (c,g), CD45RA⁺CD31⁻ mature naive (MN) T_resps (d,h), CD45RA⁻CD31⁺ memory T_resps (e,i) and CD45RA⁻CD31⁻ memory T_resps (f,j) within the ICOS⁺ and ICOS⁻ T_resp pool were estimated in healthy non‐pregnant fertile women (group 1, Inline graphic ), healthy pregnant women during their pregnancy course (groups 2–5, ◆), spontaneously term labouring women (group 6, ), 1 day postpartum (group 7, ◆) and in women with pre‐eclampsia (group 8, ) or HELLP syndrome (group 9, ). The figures show the 95% confidence intervals and the median of all patient groups for each T_reg subset.

Click here for additional data file.^{(346KB, ppt)}

Acknowledgements

This work was supported by the Dietmar Hopp Foundation, Sankt Leon‐Rot, Germany. The authors would like to thank the nursing staff of the Department of Obstetrics and Gynecology for arranging the collection of the blood samples. In addition, they would like to thank Dieter Stefan (Institute of Immunology, University of Heidelberg, Heidelberg, Germany) for professional help with FACS sorting of the T_reg subsets, and Larissa Schoenhoff (Department of Medicine I (Nephrology), University of Heidelberg, Heidelberg, Germany) for excellent technical assistance.

References

1. Erlebacher A. Mechanisms of T cell tolerance towards the allogeneic fetus. Nat Rev Immunol 2013; 13:23–33. [DOI] [PubMed] [Google Scholar]
2. Williams N. Inducing tolerance to pregnancy. N Engl J Med 2012; 7:1159–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Jiang TT, Chaturvedi V, Ertelt JM et al Regulatory T cells: new keys for further unlocking the enigma of fetal tolerance and pregnancy complications. J Immunol 2014; 192:4949–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Zhou J, Wang Z, Zhao X, Wang J, Sun H, Hu Y. An increase of Treg cells in the peripheral blood is associated with a better in vitro fertilization treatment outcome. Am J Reprod Immunol 2012; 68:100–6. [DOI] [PubMed] [Google Scholar]
5. Schlossberger V, Schober L, Rehnitz J et al The success of assisted reproduction technologies in relation to composition of the total regulatory T cell (Treg) pool and different Treg subsets. Hum Reprod 2013; 28:3062–73. [DOI] [PubMed] [Google Scholar]
6. Yang H, Qiu L, Chen G, Ye Z, Lü C, Lin Q. Proportional change of CD4⁺CD25⁺ regulatory T cells in decidua and peripheral blood in unexplained recurrent spontaneous abortion patients. Fertil Steril 2008; 92:301–5. [DOI] [PubMed] [Google Scholar]
7. Winger EE, Reed JL. Low circulating CD4⁺ CD25⁺ Foxp3⁺ T regulatory cell levels predict miscarriage in newly pregnant women with a history of failure. Am J Reprod Immunol 2011; 66:320–8. [DOI] [PubMed] [Google Scholar]
8. Lee S, Kim JY, Lee M, Gilman‐Sachs A, Kwak‐Kim J. Th17 and regulatory T cells in women with recurrent pregnancy loss. Am J Reprod Immunol 2012; 67:311–8. [DOI] [PubMed] [Google Scholar]
9. Wang WJ, Liu FJ, Xin‐Liu Hao CF, Bao HC, Qu QL Liu XM. Adoptive transfer of pregnancy‐induced CD4⁺CD25⁺‐ regulatory T cells reverses the increase in abortion rate caused by interleukin 17 in the CBA/JxBALB/c mouse model. Hum Reprod 2014; 29:946–52. [DOI] [PubMed] [Google Scholar]
10. Kisielewicz A, Schaier M, Schmitt E et al A distinct subset of HLA‐DR⁺‐regulatory T cells is involved in the induction of preterm labor during pregnancy and in the induction of organ rejection after transplantation. Clin Immunol 2010; 137:209–20. [DOI] [PubMed] [Google Scholar]
11. Steinborn A, Schmitt E, Kisielewicz A et al Pregnancy‐associated diseases are characterized by the composition of the systemic regulatory T cell (Treg) pool with distinct subsets of Tregs. Clin Exp Immunol 2012; 167:84–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Schober L, Radnai D, Schmitt E, Mahnke K, Sohn C, Steinborn A. Term and preterm labor: decreased suppressive activity and changes in composition of the regulatory T‐cell pool. Immunol Cell Biol 2012; 90:935–44. [DOI] [PubMed] [Google Scholar]
13. Santner‐Nanan B, Peek MJ, Khanam R et al Systemic increase in the ratio between Foxp3⁺ and IL‐17‐producing CD4⁺ T cells in healthy pregnancy but not in preeclampsia. J Immunol 2009; 183:7023–30. [DOI] [PubMed] [Google Scholar]
14. Laresgoiti‐Servitje E, Gomez‐Lopez N, Olson DM. An immunological insight into the origins of pre‐eclampsia. Hum Reprod Update 2010; 16:510–24. [DOI] [PubMed] [Google Scholar]
15. Darmochwal‐Kolarz J, Kludka‐Sternik M, Tabarkiewicz J et al The predominance of Th 17 lymphocytes and decreased number and function of Treg cells in preeclampsia. Reprod Immunol 2012; 93:75–81. [DOI] [PubMed] [Google Scholar]
16. Hsu P, Santner‐Nanan B, Dahlstrom JE et al Altered decidual DC‐SIGN⁺ antigen‐presenting cells and impaired regulatory T‐cell induction in preeclampsia. Am J Pathol 2012; 181:2149–60. [DOI] [PubMed] [Google Scholar]
17. Schober L, Radnai D, Spratte J et al The role of regulatory T cell (Treg) subsets in gestational diabetes mellitus. Clin Exp Immunol 2014; 177:76–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Aluvihare VR, Kallikourdis M, Betz AG. Regulatory T cells mediate maternal tolerance to the fetus. Nat Immunol 2004; 5:266–71. [DOI] [PubMed] [Google Scholar]
19. Samstein RM, Josefowicz SZ, Arvey A, Treuting PM, Rudensky AY. Extrathymic generation of regulatory T cells in placental mammals mitigates maternal–fetal conflict. Cell 2012; 150:29–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Rowe JH, Ertelt JM, Xin L, Way SS. Pregnancy imprints regulatory memory that sustains anergy to fetal antigen. Nature 2012; 490:102–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Schaier M, Seissler N, Schmitt E et al DR^high+CD45RA−‐Tregs potentially affect the suppressive activity of the total Treg pool in renal transplant patients. PLOS ONE 2012; 7:e34208. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Ito T, Hanabuchi S, Wang Y et al Two functional subsets of FOXP3⁺ regulatory T cells in human thymus and periphery. Immunity 2008; 28:870–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Chen Y, Shen S, Gorentla BK, Gao J, Zhong XP. Murine regulatory T cells contain hyperproliferative and death‐prone subsets with differential ICOS expression. J Immunol 2012; 188:1698–707. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Kohler S, Thiel A. Life after the thymus: CD31⁺ and CD31 human naive CD4⁺ T‐cell subsets. Blood 2009; 113:769–74. [DOI] [PubMed] [Google Scholar]
25. Wagner MI, Mai C, Schmitt E et al The role of recent thymic emigrant‐regulatory T cell (RTE‐Treg) differentiation during pregnancy. Immunol Cell Biol 2015. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
26. Dooley J, Liston A. Molecular control over thymic involution: from cytokines and microRNA to aging and adipose tissue. Eur J Immunol 2012; 42:1073–79. [DOI] [PubMed] [Google Scholar]
27. Zoller AL, Schnell FJ, Kersh GJ. Murine pregnancy leads to reduced proliferation of maternal thymocytes and decreased thymic emigration. Immunology 2007; 121:207–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Clambey ET, Kappler JW, Marrack P. CD8 T cell clonal expansions and aging: a heterogeneous phenomenon with a common outcome. Exp Gerontol 2007; 42:407–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Czesnikiewicz‐Guzik M, Lee WW, Cui D et al T cell subset‐specific susceptibility to aging. Clin Immunol 2008; 127:107–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Mason D, Powrie F. Memory CD4⁺ T cells in man form two distinct subpopulations, defined by their expression of isoforms of the leucocyte common antigen, CD45. Immunology 1990; 70:427–33. [PMC free article] [PubMed] [Google Scholar]
31. Ericsson PO, Linden O, Dohlsten M, Sjogren HO, Hedlund G. Functions of rat CD4⁺ T cell subsets defined by CD45RB: CD45RB‐ cells have a much stronger response to recall antigens, whereas polyclonally activated cells of both subsets are equally efficient producers of IFN in the presence of exogenous IL‐2. Cell Immunol 1991; 132:391–99. [DOI] [PubMed] [Google Scholar]
32. Perisic M, Stojic‐Vukanic Z, Pilipovic I et al Role of ovarian hormones in T‐cell homeostasis: from the thymus to the periphery. Immunobiology 2013; 218:353–67. [DOI] [PubMed] [Google Scholar]
33. Akiba H, Takeda K, Kojima Y et al The role of ICOS in the CXCR5+ follicular B helper T cell maintenance in vivo . J Immunol 2005; 175:2340–48. [DOI] [PubMed] [Google Scholar]
34. Choi JY, Ho JH, Pasoto SG et al Circulating follicular helper‐like T cells in systemic lupus erythematosus: association with disease activity. Arthritis Rheumatol 2015; 67:988–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Wu S, Yang T, Pan F et al Increased frequency of circulating follicular helper T cells in patients with ankylosing spondylitis. Mod Rheumatol 2015; 25:110–5. [DOI] [PubMed] [Google Scholar]
36. Marriott CL, Carlesso G, Herbst R, Withers DR. ICOS is required for the generation of both central and effector CD4+ memory T‐cell populations following acute bacterial infection. Eur J Immunol 2015; 45:1706–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Additional Supporting information may be found in the online version of this article at the publisher's web site:

Click here for additional data file.^{(372.5KB, ppt)}

Click here for additional data file.^{(449.5KB, ppt)}

Click here for additional data file.^{(346KB, ppt)}

[cei12693-bib-0001] 1. Erlebacher A. Mechanisms of T cell tolerance towards the allogeneic fetus. Nat Rev Immunol 2013; 13:23–33. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0002] 2. Williams N. Inducing tolerance to pregnancy. N Engl J Med 2012; 7:1159–61. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0003] 3. Jiang TT, Chaturvedi V, Ertelt JM et al Regulatory T cells: new keys for further unlocking the enigma of fetal tolerance and pregnancy complications. J Immunol 2014; 192:4949–56. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0004] 4. Zhou J, Wang Z, Zhao X, Wang J, Sun H, Hu Y. An increase of Treg cells in the peripheral blood is associated with a better in vitro fertilization treatment outcome. Am J Reprod Immunol 2012; 68:100–6. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0005] 5. Schlossberger V, Schober L, Rehnitz J et al The success of assisted reproduction technologies in relation to composition of the total regulatory T cell (Treg) pool and different Treg subsets. Hum Reprod 2013; 28:3062–73. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0006] 6. Yang H, Qiu L, Chen G, Ye Z, Lü C, Lin Q. Proportional change of CD4⁺CD25⁺ regulatory T cells in decidua and peripheral blood in unexplained recurrent spontaneous abortion patients. Fertil Steril 2008; 92:301–5. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0007] 7. Winger EE, Reed JL. Low circulating CD4⁺ CD25⁺ Foxp3⁺ T regulatory cell levels predict miscarriage in newly pregnant women with a history of failure. Am J Reprod Immunol 2011; 66:320–8. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0008] 8. Lee S, Kim JY, Lee M, Gilman‐Sachs A, Kwak‐Kim J. Th17 and regulatory T cells in women with recurrent pregnancy loss. Am J Reprod Immunol 2012; 67:311–8. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0009] 9. Wang WJ, Liu FJ, Xin‐Liu Hao CF, Bao HC, Qu QL Liu XM. Adoptive transfer of pregnancy‐induced CD4⁺CD25⁺‐ regulatory T cells reverses the increase in abortion rate caused by interleukin 17 in the CBA/JxBALB/c mouse model. Hum Reprod 2014; 29:946–52. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0010] 10. Kisielewicz A, Schaier M, Schmitt E et al A distinct subset of HLA‐DR⁺‐regulatory T cells is involved in the induction of preterm labor during pregnancy and in the induction of organ rejection after transplantation. Clin Immunol 2010; 137:209–20. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0011] 11. Steinborn A, Schmitt E, Kisielewicz A et al Pregnancy‐associated diseases are characterized by the composition of the systemic regulatory T cell (Treg) pool with distinct subsets of Tregs. Clin Exp Immunol 2012; 167:84–98. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0012] 12. Schober L, Radnai D, Schmitt E, Mahnke K, Sohn C, Steinborn A. Term and preterm labor: decreased suppressive activity and changes in composition of the regulatory T‐cell pool. Immunol Cell Biol 2012; 90:935–44. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0013] 13. Santner‐Nanan B, Peek MJ, Khanam R et al Systemic increase in the ratio between Foxp3⁺ and IL‐17‐producing CD4⁺ T cells in healthy pregnancy but not in preeclampsia. J Immunol 2009; 183:7023–30. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0014] 14. Laresgoiti‐Servitje E, Gomez‐Lopez N, Olson DM. An immunological insight into the origins of pre‐eclampsia. Hum Reprod Update 2010; 16:510–24. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0015] 15. Darmochwal‐Kolarz J, Kludka‐Sternik M, Tabarkiewicz J et al The predominance of Th 17 lymphocytes and decreased number and function of Treg cells in preeclampsia. Reprod Immunol 2012; 93:75–81. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0016] 16. Hsu P, Santner‐Nanan B, Dahlstrom JE et al Altered decidual DC‐SIGN⁺ antigen‐presenting cells and impaired regulatory T‐cell induction in preeclampsia. Am J Pathol 2012; 181:2149–60. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0017] 17. Schober L, Radnai D, Spratte J et al The role of regulatory T cell (Treg) subsets in gestational diabetes mellitus. Clin Exp Immunol 2014; 177:76–85. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0018] 18. Aluvihare VR, Kallikourdis M, Betz AG. Regulatory T cells mediate maternal tolerance to the fetus. Nat Immunol 2004; 5:266–71. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0019] 19. Samstein RM, Josefowicz SZ, Arvey A, Treuting PM, Rudensky AY. Extrathymic generation of regulatory T cells in placental mammals mitigates maternal–fetal conflict. Cell 2012; 150:29–39. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0020] 20. Rowe JH, Ertelt JM, Xin L, Way SS. Pregnancy imprints regulatory memory that sustains anergy to fetal antigen. Nature 2012; 490:102–6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0021] 21. Schaier M, Seissler N, Schmitt E et al DR^high+CD45RA−‐Tregs potentially affect the suppressive activity of the total Treg pool in renal transplant patients. PLOS ONE 2012; 7:e34208. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0022] 22. Ito T, Hanabuchi S, Wang Y et al Two functional subsets of FOXP3⁺ regulatory T cells in human thymus and periphery. Immunity 2008; 28:870–80. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0023] 23. Chen Y, Shen S, Gorentla BK, Gao J, Zhong XP. Murine regulatory T cells contain hyperproliferative and death‐prone subsets with differential ICOS expression. J Immunol 2012; 188:1698–707. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0024] 24. Kohler S, Thiel A. Life after the thymus: CD31⁺ and CD31 human naive CD4⁺ T‐cell subsets. Blood 2009; 113:769–74. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0025] 25. Wagner MI, Mai C, Schmitt E et al The role of recent thymic emigrant‐regulatory T cell (RTE‐Treg) differentiation during pregnancy. Immunol Cell Biol 2015. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0026] 26. Dooley J, Liston A. Molecular control over thymic involution: from cytokines and microRNA to aging and adipose tissue. Eur J Immunol 2012; 42:1073–79. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0027] 27. Zoller AL, Schnell FJ, Kersh GJ. Murine pregnancy leads to reduced proliferation of maternal thymocytes and decreased thymic emigration. Immunology 2007; 121:207–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0028] 28. Clambey ET, Kappler JW, Marrack P. CD8 T cell clonal expansions and aging: a heterogeneous phenomenon with a common outcome. Exp Gerontol 2007; 42:407–11. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0029] 29. Czesnikiewicz‐Guzik M, Lee WW, Cui D et al T cell subset‐specific susceptibility to aging. Clin Immunol 2008; 127:107–18. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0030] 30. Mason D, Powrie F. Memory CD4⁺ T cells in man form two distinct subpopulations, defined by their expression of isoforms of the leucocyte common antigen, CD45. Immunology 1990; 70:427–33. [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0031] 31. Ericsson PO, Linden O, Dohlsten M, Sjogren HO, Hedlund G. Functions of rat CD4⁺ T cell subsets defined by CD45RB: CD45RB‐ cells have a much stronger response to recall antigens, whereas polyclonally activated cells of both subsets are equally efficient producers of IFN in the presence of exogenous IL‐2. Cell Immunol 1991; 132:391–99. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0032] 32. Perisic M, Stojic‐Vukanic Z, Pilipovic I et al Role of ovarian hormones in T‐cell homeostasis: from the thymus to the periphery. Immunobiology 2013; 218:353–67. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0033] 33. Akiba H, Takeda K, Kojima Y et al The role of ICOS in the CXCR5+ follicular B helper T cell maintenance in vivo . J Immunol 2005; 175:2340–48. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0034] 34. Choi JY, Ho JH, Pasoto SG et al Circulating follicular helper‐like T cells in systemic lupus erythematosus: association with disease activity. Arthritis Rheumatol 2015; 67:988–99. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cei12693-bib-0035] 35. Wu S, Yang T, Pan F et al Increased frequency of circulating follicular helper T cells in patients with ankylosing spondylitis. Mod Rheumatol 2015; 25:110–5. [DOI] [PubMed] [Google Scholar]

[cei12693-bib-0036] 36. Marriott CL, Carlesso G, Herbst R, Withers DR. ICOS is required for the generation of both central and effector CD4+ memory T‐cell populations following acute bacterial infection. Eur J Immunol 2015; 45:1706–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Differentiation of ICOS+ and ICOS− recent thymic emigrant regulatory T cells (RTE Tregs) during normal pregnancy, pre‐eclampsia and HELLP syndrome

M I Wagner

M Jöst

J Spratte

M Schaier

K Mahnke

S Meuer

M Zeier

A Steinborn

Summary

Introduction

Materials and methods

Patient collectives and healthy volunteers

Table 1.

Fluorescence activated cell sorter (FACS) staining

Positive selection of CD4+CD127low+/−CD25+ Treg cells

Sorting and functional testing of Treg subsets

Statistical analysis

Results

ICOS+ and ICOS− Tregs show similar differentiation during the normal pregnancy course

Figure 1.

Figure 2.

The decrease of ICOS+CD45RA−CD31− but not ICOS−CD45RA−CD31− memory Tregs is crucial for the onset of spontaneous term labour

Figure 3.

Pre‐eclampsia patients show a significantly increased differentiation of both ICOS+ and ICOS−CD45RA+CD31+RTE Tregs into CD45RA−CD31+ instead of CD45RA−CD31− memory Tregs

ICOS+ and ICOS− Tregs have similar suppressive activity that is reduced strongly in patients affected with pre‐eclampsia, but not HELLP syndrome

Figure 4.

ICOS+ and ICOS− Tresp cells differentiate similar to ICOS+ and ICOS− Tregs during the normal pregnancy course

Figure 5.