Reduced frequency and functional defects of CD4⁺CD25^highCD127^low/− regulatory T cells in patients with unexplained recurrent spontaneous abortion

Recurrent spontaneous abortion (RSA), defined as 2 or more consecutive pregnancy losses, affects approximately 1–2% of prospective mothers [1, 2]. The etiology of RSA is multifactorial; it can be related to fetal chromosomal factors, anatomical abnormalities, and hormonal problems. However, in more than half of RSA cases, the precise cause remains unknown; this phenomenon is termed unexplained RSA (URSA) [3, 4]. Recent evidence suggests that URSA is related to a failure of fetal–maternal immunological tolerance [5].

Successful pregnancy requires various immunological adaptations to facilitate attachment, implantation, placentation, and fetal–maternal tolerance [6]. Disturbances of these adaptations are related to pregnancy complications, such as RSA and preeclampsia [7‐9]. Dynamic immunological changes throughout normal conception include downregulation of T helper (Th) 1 and Th17 responses, general Th2-skewing of T lymphocyte responses, and enrichment of uterine natural killer cells and regulatory T cells (Tregs) in the first and second trimesters of pregnancy [7, 10, 11]. Uterine natural killer cells represent approximately 70% of all uterine lymphocytes; they are abundant in the decidua during the commencement of implantation, and regulate decidual blood vessel remodeling, decidual cell proliferation, and trophoblast invasion [12‐14]. T lymphocyte proportions in the uterus are more variable; they comprise approximately 10 to 20% of decidual leukocytes in the first trimester [15], and their proportion increases over the course of pregnancy [16]. CD25^highFOXP3⁺ Tregs with immunosuppressive properties, which account for 5% of CD4⁺ T cells, are important to sustain a successful pregnancy [17]. Treg expansion can be detected in the follicular phase of the menstrual cycle, and the serum estradiol-driven increase peaks at ovulation [18]. The proportions of both peripheral and decidual Tregs continue to increase in early pregnancy, and reach their peaks at mid-gestation, after which they gradually decline until 6–8 weeks after delivery to basal levels [17, 19, 20]. Tregs exert several biological effects during pregnancy; they induce decidual support for embryo implantation through contact with other immune cells [8], develop the tolerance towards paternal antigens [19, 21], and promote proper maternal vascular remodeling for robust placental development [22]. Although some studies have demonstrated that lower than average proportions and disrupted functions of Tregs are closely related to the occurrence of URSA [23‐26], their findings remain inconclusive due to the limitations of their approaches for identifying Tregs.

Previous studies have typically used the CD4⁺CD25⁺ surface phenotype to investigate the involvement of Tregs in pregnancy loss [27‐30]. However, CD25 is not a unique Treg marker as it is also expressed on activated T cells [31, 32]. Thus, previous findings may have been confounded by the presence of contaminating non-Tregs that express CD25. Only cells that express the highest levels of CD25 (CD25^high) exert a potent immunosuppressive ability [32, 33].

Forkhead box transcription factor P3 (FOXP3)—the main transcriptional regulator for the induction, development, and suppressive ability of Tregs—is unique and specific to Tregs [31, 34‐37]. Dysregulated FOXP3 expression in Tregs is associated with autoimmune diseases, and may induce the failure of immune tolerance in a dose-dependent manner [38]. Upregulation of FOXP3 expression induces the conversion of naïve T cells into Tregs [39], whereas a reduction in FOXP3 expression is related to immunosuppressive Treg dysfunction [40, 41]. Unfortunately, as FOXP3 is an intracellular protein, this marker cannot be used to isolate Tregs and is of limited value in Treg functional studies.

Recent studies have demonstrated that surface expression of the IL-7 receptor (CD127) inversely correlates with FOXP3 expression, such that CD4⁺CD127^low/− cells largely overlap with FOXP3⁺ cells; thus, CD127 can be used as a cell surface marker to distinguish Tregs from activated CD4⁺ T cells [42, 43]. Activated effector T cells that express high levels of CD127 and do not mediate immunosuppressive functions are also present within the CD4⁺CD25^high T cell population [43, 44]. Therefore, instead of the CD4⁺CD25⁺CD127^low/− phenotype for Tregs used in previous studies [25], in this study, we relied on high expression of CD25 and low or no expression of CD127 to strictly distinguish Tregs (CD4⁺CD25^highCD127^low/−) from activated T cells and investigate the number and suppressive capacity of Tregs isolated from the peripheral blood of patients with URSA with the aim of expanding our understanding of their role in URSA.

URSA has been suggested to be largely associated with a failure of fetal–maternal immunological tolerance, in which Tregs play a crucial role [46]. Evidence suggests that URSA is associated with reduced Treg numbers or function. In early studies, the CD4⁺CD25⁺ phenotype was used to identify Tregs, and later studies refined the phenotype to CD25^high or FOXP3⁺ CD4⁺ T cells to more specifically identify Tregs [18, 24]. However, as FOXP3 is a nuclear protein, it cannot be used to isolate Tregs for functional assays. Several studies have shown that FOXP3 expression inversely correlates with CD127 expression, indicating that CD4⁺CD25⁺ T cells with low or no CD127 expression mainly comprise FOXP3-expressing cells [42, 43].

To optimize the isolation of cells that actually function as Tregs, we utilized the specific CD4⁺CD25^highCD127^low/− phenotype to isolate Tregs with the aim of expanding our understanding of their role in URSA. We confirmed that CD4⁺CD25^highCD127^low/− T cells are a highly purified population of FOXP3⁺ Tregs, which allowed us to examine the frequency and function of Tregs in the circulation of URSA patients. URSA patients had a lower proportion of CD4⁺CD25^highCD127^low/− Tregs among the CD4⁺ T cells than normal controls Furthermore, we observed lower expression of FOXP3 on a per-Treg basis at the protein and mRNA levels in URSA patients, which is consistent with the lower frequency of circulating Tregs. These findings are in agreement with reports of lower-than-normal levels of Tregs in patients with URSA [18, 23, 24, 47]. Yang et al. [23] reported that the proportions of CD4⁺CD25^bright T cells in both the peripheral blood and decidua of URSA patients were lower than those in normal pregnant control subjects. Wang et al. [47] observed that the frequencies of CD4⁺CD25⁺CD127^low/− T cells in the peripheral blood and decidua of URSA patients were lower than those in normal women in early pregnancy. In addition, Mei et al. [24] found that the percentage of CD4⁺CD25^high Tregs in the peripheral blood of nonpregnant URSA patients was lower than that in URSA patients who had early miscarriages and normal nonpregnant women. In the decidua, they observed a lower frequency of CD4⁺CD25^high Tregs and lower FOXP3 expression in Tregs in URSA patients with early miscarriages than in normal women in early pregnancy.

A few studies have described the peripheral and decidual expansion of CD4⁺CD25⁺FOXP3⁺ Treg populations in women during the first 2 trimesters of pregnancy [48, 49]. Qian et al. [50] reported that the proportion of peripheral CD4⁺CD25⁺CD127^low Tregs was not higher in pregnant women with a history of URSA than in nonpregnant women with URSA, whereas the percentage of Tregs, the expression level of CTLA-4 in CD4⁺CD25⁺CD127^low cells, and the proportion of CD4⁺FOXP3⁺ cells in the decidua were significantly lower in patients with URSA than in normal controls. On the other hand, Quan et al. [51] reported no differences in the percentages of peripheral CD4⁺CD25⁺ cells among total CD4⁺ cells or FOXP3⁺ cells among CD4⁺CD25⁺ cells between nonpregnant URSA patients, early-term URSA patients, normal women who underwent planned abortions, and normal nonpregnant controls. These studies used different antibody panels to identify Tregs at various phases of the menstrual cycle [18]; both these factors are likely to have affected the proportions of Tregs observed.

Our study is the first to report the isolation of CD4⁺CD25^highCD127^low/− Tregs and CD4⁺CD25⁻ Tresps from the peripheral blood of URSA patients to conclusively determine the localization of the T cell-inhibitory defect. We found that Tregs isolated from patients with URSA suppressed autologous Tresps stimulated with anti-CD3/CD28 beads in a concentration-dependent manner, with the greatest suppression occurring in coculture at a 1:1 Treg:Tresp ratio; however, patient-derived Tregs had poorer suppressive capacity than Tregs from normal controls. Anergy, a hallmark of CD4⁺CD25^highCD127^low/− Tregs, was evident in Tregs from both URSA patients and normal fertile women. In addition, Tresps from URSA patients and control subjects had similar proliferation rates. This result corroborated that the defects in the suppressive activity of Tregs from URSA patients were not due to fluctuations in CD4⁺CD25⁻ Tresp proliferation. Other groups have assessed the function of Tregs in the context of URSA [18, 25]. Arruvito et al. [18] found that RSA patients showed similarly lower proportions of CD4⁺CD25⁺, CD4⁺CD25^high, and FOXP3⁺ cells within the CD4⁺ population in both the follicular and luteal phases, comparable to the proportions in postmenopausal women. Unlike fertile controls, RSA patients did not exhibit fluctuations in these cell populations during their menstrual cycles. Moreover, the authors found that CD4⁺CD25⁺ cells—deemed Tregs in this study—from RSA patients had lower suppressive activity than those from fertile controls. Bao et al. [25] reported that CD4⁺CD25⁺CD127^dim/− Tregs isolated from the decidua of URSA patients had lower suppressive potency than those from normal pregnant controls. They also found that only 2 and 1% of Tregs from URSA patients expressed IL-10 and TGF-β, respectively, compared with 9 and 14%, respectively, in Tregs from gestational controls, and the immunosuppressive function of the Tregs could be partially blocked with neutralizing anti-IL-10 and anti-TGF-β antibodies. However, they did not explore whether the functional T cell suppression deficit was specific to Tregs or intrinsic to Tresps.

We analyzed whether the apparent functional impairment of Tresp suppression observed in URSA patients was partly due to resistance of the patients’ Tresps to suppression. We found that Tregs from patients with URSA were less potent in suppressing the proliferation of Tresps from normal fertile women than Tregs from the normal controls. Tregs from URSA patients also showed less inhibitory activity in cocultures with autologous Tresps than Tregs from the normal controls. Taken together, our experiments demonstrated, for the first time, that the failure of Tresp suppression in URSA patients is predominantly due to an intrinsic Treg defect, rather than the resistance of Tresps to suppression by functional CD4⁺CD25^highCD127^low/− Tregs.

From a clinical perspective, our findings have 2 important implications. First, the phenotype CD4⁺CD25^highCD127^low/− can be used for Treg selection and purification, particularly for clinical immunotherapy applications. Second, CD4⁺CD25^highCD127^low/− Tregs may contribute to the immunopathogenesis of URSA and may be an attractive target for the prevention and treatment of URSA. Strategies to boost Treg abundance and suppressive function for achieving successful gestation in patients with URSA should be evaluated.

We demonstrated that the CD4⁺CD25^highCD127^low/− phenotype is highly specific for the isolation and purification of Tregs. In addition, we demonstrated deficiencies in peripheral CD4⁺CD25^highCD127^low/− Treg numbers, suppressive function, and FOXP3 expression in patients with URSA. The impaired immunoregulation in patients with URSA is primarily localized to Tregs, not Tresps. Further studies are needed to identify the specific Treg defects in patients with URSA and the mechanisms underlying Treg dysfunction.