Th17 and Regulatory T cells in Women with Recurrent Pregnancy Loss
Sung Ki Lee, Department of Obstetrics and Gynecology, College of Medicine, Konyang University, 685 Gasoowon-dong, Seo-gu, Daejeon 302-718, South Korea.
The immune system of pregnant women is tightly controlled to defend against microbial infections and at the same time, to accept an embryo or the fetus, which are expressing semi-allogenic paternal antigens. Furthermore, inflammation-like processes are crucial for tissue growth, remodeling, and differentiation of the decidua during pregnancy. Dysregulation of elaborate immune control may lead reproductive failure, such as implantation failure, recurrent pregnancy loss (RPL), preterm birth, intrauterine fetal growth restriction, and preeclampsia. Until recent years, a balance between Th1 and Th2 cells was believed to be the key immune regulatory mechanism of T-cell immunology especially during pregnancy. Since the identification of regulatory T cells was made, the mechanism of immune regulation has become a major issue in immunologic research. Also, the recent identification of Th17 cells has drawn our attention to a new immune effector. The balance between Th17 and regulatory T cells may explain more about the pathophysiology of reproductive failure. This review will discuss relevant human literature on regulatory T and Th17 cells in normal reproductive physiology and in women with RPL and infertility.
During pregnancy, the immune system of the mother is tightly controlled to defend against microbial infections and to accept an embryo and a fetus, which are expressing semi-allogenic paternal antigens. Furthermore, immune-mediated processes such as tissue growth, remodeling, and differentiation are crucial to maintain pregnancy. Dysregulation of this elaborate immune control may lead reproductive failure, such as implantation failure, pregnancy loss, preterm birth, intrauterine fetal growth restriction, and preeclampsia.
Recurrent pregnancy loss (RPL), commonly defined as three or more spontaneous pregnancy losses before 20 weeks of gestation, is as frequent as in 1–2% of reproductive couples. The expected prevalence of pregnancy loss following three or more episodes is one in 300 pregnancies, 0.3%. The etiology of RPL is multifactorial, and sometimes women with RPL showed multiple causative factors following thorough evaluation. In general, more than half of women with RPL have autoimmune or alloimmune abnormalities. Antiphospholipid syndrome is a well-known autoimmune factor, which causes thrombosis in the uterine vessels and decrease in blood supply to the fetomaternal interface. Alloimmune abnormalities seem to significantly contribute to the pathogenesis of RPL, even though the exact extent of these abnormalities remains to be defined. Natural killer cells have been extensively studied in RPL. High proportion and high cytotoxicity of NK cells have been reported as poor prognostic factors.[5-7] In addition, an increased population of CD4+ Th1 cells is also thought to be harmful in early pregnancy.[8-10]
Recent advances in immunologic studies have widened our knowledge of how the immune response is regulated. Regulatory T cells are considered the most important immune regulator, especially in the peripheral immune system.[11, 12] Recently, a new T-cell subset was introduced as another key effector T cell. These Th17 cells, which secrete IL-17, are thought to play a role in chronic inflammation and protection from fungal infection.[13, 14] There is growing evidence that regulatory T and Th17 cells are involved in establishment and maintenance of pregnancy as regulator and effector cells, respectively. Many researchers suggest that an immune imbalance between effectors and regulatory cells may lead implantation failure and many other pregnancy disorders. This review will discuss recent and review recent studies concerning regulatory T and Th17 cells in RPL and infertility.
Foxp3+ regulatory T cells as a key immune regulator
For immune homeostasis, the balance between effector cells and regulator cells is necessary. Some conditions such as microbial infection trigger immune activation to defend against microorganisms or repair tissue damage. However, this activated immune response should be downregulated and return to the same normal state as prior to activation.
The idea of immune regulation by thymic lymphocytes was introduced by Gershon and Kondo in 1970, and T lymphocytes that were capable of suppressing an immune response were named as suppressor T cells. Even though there were many efforts to identify these cells, the search for the elusive suppressor T cells was not successful for a few decades. In 1995, CD4+ CD25+ T cells were reported as a particular T-cell subset with regulatory function in mice. However, CD25 expression did not differentiate regulatory T cells from activated CD4+ T cells in humans. In 2003, a transcription factor, forkhead boxP3 (Foxp3), was reported as a master control gene for development and function of regulatory T cells in mice and humans.[17, 18] Foxp3+ regulatory T cells originate from the thymus (naturally occurring regulatory T cells) and in the periphery by activation of naïve CD4+ T cells following antigen stimulation under the influence of TGF-β (inducible regulatory T cells). Both regulatory T cells have demonstrated suppressive function against immune effectors including CD4+, CD8+, B, NK, NKT, and dendritic cells (DCs). Even though several cell surface markers have been proposed as specific markers for regulatory T cells such as CD25high, CD127low, and CD62L, Foxp3 is still the most reliable marker for regulatory T cells so far.
In 1982, a case report about an X-linked autoimmune disease that killed 17 male infants in a family before their first birthday was published. Affected male infants showed many symptoms involving multiple organs, such as diarrhea, DM, hemolytic anemia, thyroiditis, etc. This family disease named IPEX syndrome (immunodysregulation, polyendocrinopathy, enteropathy, and X-linked syndrome) is known to be caused by mutations of Foxp3 gene.
In humans, Foxp3 is mainly expressed in CD4+ CD25+ T cells, but other T-cell subsets such as CD4+ CD25− and CD8+ T cells also express Foxp3. In general, Foxp3-expressing T cells are considered as possessing suppressive function. Recently, Lee et al. have reported that CD4+ Foxp3high and CD4+ Foxp3low T cells correlated with different lymphocyte subsets. CD4+ Foxp3+ cells negatively correlated with CD3− CD56+ NK cells. On the other hand, CD4+ Foxp3high regulatory T cells positively correlated with CD3+ CD4+ TNF-α+ cells and the ratio of type 1/2 cytokine-producing CD3+ CD8+ cells, but negatively correlated with CD3+ CD8+ IL-10+ T cells. These findings indicate that each Foxp3+ regulatory T-cell subpopulation may have unique immune interaction, which controls particular subsets of lymphocytes.
The precise molecular mechanisms of Foxp3+ regulatory T cell-mediated suppression are not elucidated yet. Many putative mechanisms to control effector cells have been demonstrated. One of them is mediated by direct cell-to-cell contact between regulatory T and target cells. Several molecules have been identified, for example, FAS-FASL and granzyme A in human. Cytokine-mediated mechanisms such as IL-10 and interaction with DCs have also been suggested. Indoleamine dioxygenase 2,3-dioxygenase (IDO), a potent immune regulator, is secreted following interaction of regulatory T cells with DCs and induces pro-apoptotic molecules from the catabolism of tryptophan, resulting in suppression of effector T cells. In mice, many other molecular mechanisms are likely involved in regulatory T cell-mediated suppression: cytotoxic T lymphocyte activation 4 (CTLA4), CD73-CD39-mediated suppression, lymphocyte activation gene 3 (LAG3), granzyme B, TGF-β, latency-associated peptide (LAP), galectin 1, CD25, and IL-35 etc.
The mechanism of regulation of Foxp3+ regulatory T cells remains elusive. The thymus supplies naturally occurring regulatory T cells constantly but the proportion of naturally occurring regulatory T cells seems to decrease by aging. Contrarily, in the periphery, inducible regulatory T cells increase in proportion in elderly people. This finding suggests that inducible regulatory T cells may replenish depletion of the naturally occurring regulatory T cells or are induced as a response to chronic inflammation as a person ages.
Regulatory T cells expand during pregnancy in mice and humans and play a key role in protection when maternal immune cells first contact fetal antigens associated with invading trophoblasts. Draining lymph nodes from the uterus have been implicated as the predominant site of regulatory T-cell expansion and fetal alloantigen. Neither estrogen nor progesterone was suggested as responsible for regulatory T-cell expansion in mice. Mice seminal fluid was reported to contribute to the accumulation of Foxp3+ regulatory T cells in the preimplantation uterus, and insufficient expansion of regulatory T cells against paternal antigens may trigger spontaneous abortion in mice. Additionally, the induction of paternal specific regulatory T cells has been demonstrated in pregnant women at 24–28 weeks of gestation. Furthermore, fetus-specific CD4+ CD25bright regulatory T cells are selectively recruited from the peripheral blood into the deciduas in human pregnancy. These findings suggest that paternal antigens in the seminal fluid and fetal allogeneic antigens may induce the expansion of maternal regulatory T cells in the periphery, which are preferentially recruited toward the fetomaternal interface so as to control the maternal immune response to fetal antigens and lead to favorable pregnancy outcome.
Recently, several reports have indicated that transient or low expression of Foxp3 did not confer regulatory function to the cells[31, 32] and those cells are converted into effector T cells producing pro-inflammatory cytokines such as IL-2, interferon-γ (IFN-γ) and IL-17.[31, 33-35] Intriguingly, IDO may play a role in T-cell differentiation. The presence of IDO is known to develop inducible regulatory T cells, but its absence reprograms regulatory T cells into the effectors such as Th17 cells.
Th17 cells, a new effector T cells
About 25 years ago, the Th1 and Th2 hypothesis was first introduced.[37, 38] In this concept, type 1 CD4+ T helper cells (Th1 cells) that secrete IL-2 and IFN-γ induce cell-mediated immune reaction related to tissue damage, and type 2 CD4+ Th cells (Th2) lead to antibody-mediated immune responses, such as allergy. This theory of Th1/Th2 had been accepted as a solid dichotomy of effector T-cell immunity. However, it was challenged by some questions that arose in studies using animal autoimmune disease models. Experimental autoimmune encephalomyelitis (EAE) had been believed to be a Th1-mediated disease. Unexpectedly, IFN-γ did not worsen the EAE and antibody to IFN-γ could not protect it but made EAE worse. In contrast, IL-17-producing T cells caused EAE in adaptive transfer experiment.
The discovery of IL-17 secreting CD4+ T (Th17) cells was a major step toward resolving a puzzle of EAE. In humans, Th17 cells are known to develop from naïve CD4+ T cells by TGF-β, IL-6, IL-23, and IL-1 and secrete IL-17A, IL-17F, IL-22, and IL-26.[14, 41-43] The transcriptional factor to develop Th17 cells is retinoic acid-related orphan receptor γt (RORγt) in humans and mice. Early Th17 cell studies were focused in autoimmune diseases such as EAE, rheumatoid arthritis, asthma, inflammatory bowel diseases, and lupus.[44, 45] Thereafter, studies of Th17 cells have been expanded to allograft rejection, host defense, metabolic disorders, and tumor immunology.[44, 46, 47]
IL-17 is known to induce inflammation via neutrophil infiltration and stimulation of IL-1, IL-6, IL-8, TNF-α, nitric oxide, matrix metalloproteinase, receptor activator for nuclear factor κB ligand (RANKL) and granulocyte-macrophage colony stimulating factor (GM-CSF) production.[48, 49] Major source of IL-17 production is CD4+ T cells, but other immune cells including CD8+ cells, γδ T cells, CD14+ monocytes, lymphoid tissue inducer (LTi) cells, and NK-like cells also secrete IL-17.[50, 51] These IL-17-producing cells are believed to play a role in defense against viruses, some bacteria, fungi, and chronic inflammation.
Human regulatory T and Th17 cells during a menstrual cycle and pregnancy
There is little information regarding the expression of peripheral blood and uterine regulatory T cells during a menstrual cycle. Arruvito et al. have reported that the proportion of peripheral blood Foxp3+ T cells was significantly increased in the late follicular phase as compared to that in the luteal phase (Table 1). They also presented a positive correlation between the level of regulatory T cells and the serum estradiol concentration. This finding may indicate that estradiol positively affects the expansion of regulatory T cells. However, other studies did not find any significant association between the estradiol level and the percentage of CD4+ CD25high T cells during a menstrual cycle. There is an indirect regulatory T-cell study carried out in the human endometrium. The density of endometrial Foxp3+ regulatory T cells rose gradually throughout the proliferative phase. The authors suggested that the increase in peripheral blood and endometrial Foxp3+ regulatory T cells may play a role in the implantation of an embryo in the mid-secretory phase.
Table 1. Summary of Regulatory T (Treg) Cells and Th17 Cells in Women with Normal Reproductive Physiology and Reproductive Failure
|Menstrual cycle||PB: ↑ in late follicular phase?EM: ↑during proliferative phase||?||?|
|Pregnancy||PB: ↑ during 1st and 2nd trimester, ↓ in 3rd trimester||PB: → or ↓||PB: ↓|
EM: ↓ in the mid-secretory phase
|Unexplained recurrent pregnancy loss|
PB: ↓ in number and function
For regulatory T-cell recruitment into the endometrium and deciduas, some chemokine receptors and their ligands are likely involved. Chemokine receptor 5 (CCR5) expressed in regulatory T cells plays a role in the accumulation of regulatory T cells in the uterus. This CCR5 is related to a highly suppressive phenotype and may be a marker for those cells activated by paternal alloantigens. Chemokine ligand 4 (CCL4), a CCR5 ligand, is intensively expressed in the pregnant uterus and is involved in the further selective accumulation of CCR5+ regulatory T cells during pregnancy. Additionally, human chorionic gonadotropin (hCG) is suggested as a hormone trafficking regulatory T cells in the fetomaternal interface. As regulatory T cells have LH/CG receptors, both hCG-producing JEG3 cells and first trimester trophoblast cells efficiently attracted regulatory T cells. This is another mechanism attracting regulatory T cells in the embryo-implanted deciduas.
During pregnancy, peripheral blood CD4+ CD25+ and CD4+ CD25+ Foxp3+ regulatory T cells increase gradually during 1st and 2nd trimester and then decrease in the 3rd trimester and postpartum.[58, 59] A recent study has found that suppressive activity of regulatory T cells from normal pregnant women was significantly increased in 1st and 2nd trimester, but significantly weak in 3rd trimester and at term as compared with that of non-pregnant women. Published data comparing endometrial and decidual regulatory T cells between non-pregnant and pregnant women or during pregnancy were not found. In a study in women with spontaneous pregnancy loss, CD4+ CD25high regulatory T cells were preferentially recruited into the deciduas as compared to circulating regulatory T cells. Some ex vivo studies have demonstrated that high estradiol concentration during pregnancy promoted proliferation of human regulatory T cells without altering suppressive phenotypes and pregnancy estradiol level expanded regulatory T cells and increased Foxp3 expression in mice.
It is still unknown whether Th17 cells fluctuate during a menstrual cycle. The findings of Th17 cells during pregnancy are inconsistent. Santner-Nanan et al. have found lower Th17/regulatory T-cell ratio and lower Th17 cell level during pregnancy than those of non-pregnant women. However, several reports have published that circulating Th17 cells were not different between non-pregnant state and each trimester or between non-pregnant period and a certain period of pregnancy.[64, 65] Nakashima et al. showed that the proportion of decidual Th17 cells was significantly higher than that of circulating Th17 cells in the first trimester. Furthermore, the Th17/Foxp3+ regulatory T-cell ratio was decreased in normal 2nd and 3rd trimester pregnant women as compared to that in healthy non-pregnant women. Further studies are warranted regarding normal physiology of Th17 cells in women in reproductive age.
Regulatory T and Th17 cells in women with infertility
Only a few regulatory T-cell studies in women with infertility have been published so far. Peripheral blood mononuclear cells (PBMCs) obtained from women with unexplained infertility undergoing in vitro fertilization (IVF) secreted less IL-10 than those from fertile controls. Foxp3 mRNA expression was significantly decreased, but not mRNAs for T-bet (Th1 cells), GATA3 (Th2 cells), and TGF-β, in the endometrium of mid-secretory phase in women with primary unexplained infertility comparing with that in fertile controls. These findings implicate that decreased immune regulatory function may have negative influence on fertility. Recently, Boomsma et al. have demonstrated that cytokines from aspirated endometrial secretion including type 1 and type 2 cytokines, and IL-17 were not significantly different between women with IVF and controls in terms of pregnancy rate.
Regulatory T and Th17 cells in women with RPL
Several reports have demonstrated that regulatory T cells decreased in the peripheral blood and/or deciduas in women with RPL.[9, 52, 61, 64, 70] In 2004, Sasaki et al. first reported the association between regulatory T cells and spontaneous abortion. CD4+ CD25high regulatory T cells in the peripheral blood and deciduas decreased in spontaneous abortion group as compared to induced abortion group. Furthermore, the percentage of circulating CD4+ CD25+ regulatory T cells significantly increased in early pregnancy comparing to non-pregnant state. However, women with spontaneous abortions did not demonstrate the increase in regulatory T cells during pregnancy. In addition, decidual CD4+ CD25high T cells were significantly lower in women with spontaneous abortion than women undergoing induced abortion. They also observed that decidual and peripheral blood CD4+ CD25+ regulatory T cells were anergic and suppressed the proliferation of CD4+ CD25− T cells via cell contact manner. Arruvito et al. have published a wonderful regulatory T-cell study comparing women with RPL with fertile controls. Opposite to fertile controls, in women with RPL, CD4+ CD25+, CD4+ CD25high, and Foxp3+ regulatory T cells did not show any significant fluctuation during a menstrual cycle. CD4+ CD25high and Foxp3+ T cells regulatory T cells in women with RPL not only significantly decreased as compared to those of controls, but also were as low as those of postmenopausal women. Moreover, regulatory T cells from women with RPL showed suppressive, but significantly lower in function as compared to those of fertile controls. Lymphocyte immunotherapy (LIT) with paternal or third-party lymphocytes has been demonstrated to increase CD4+ CD25bright T cells. The proportion of these CD4+ CD25bright T cells was higher in women with a successful pregnancy than in women with pregnancy loss after LIT. The presence of intravenous immunoglobulin with human CD4+ CD25+ regulatory T cells in culture significantly increased the expression of Foxp3, TGF-β, and IL-10. These findings suggest that deceased number and defective function of regulatory T cells in women with RPL results in reproductive failure, and immunotherapy may reverse the decreased number and function of regulatory T cells.
Th17 cells have been studied in women with reproductive failure during the past few years. Nakashima et al. showed the accumulation of IL-17+ T cells in the deciduas in women with inevitable abortion. Decidual IL-17+ T cells were mostly CD4+ T cells and a few CD8+ cells also expressed IL-17 in this study. In addition, the number of decidual IL-17+ cells was positively correlated with the number of decidual neutrophils. However, they could not find any difference in the number of decidual IL-17+ T cells between women with missed abortion and normal pregnancy. From these results, the authors concluded that decidual IL-17+ cells might be involved in the inflammation of the late stage of abortive process, not the causative factor of abortion. Because their data of IL-17+ cells were limited to inevitable abortion, not to RPL, it may be difficult to generalize the results as the immunologic mechanism of RPL.
A series of studies concerning Th17 cells have been reported regarding RPL in the past 2 years. Wang et al. found an increase in Th17 cells in the peripheral blood and decidua of women with unexplained RPL as compared to normal pregnant women. Serum IL-17 and IL-23 levels were significantly higher in women with RPL. Furthermore, Th17-related molecules such as IL-17, IL-23, and retinoid orphan receptor C (RORC) were significantly expressed in the deciduas of women with RPL. The number of Th17 cells inversely correlated with that of regulatory T cells in the peripheral blood and deciduas. The same group has reported another Th17 cell study in women with RPL. They found that the proportions of peripheral blood CCR6+ CD4+ T and CCR6+ IL17+ T cells were significantly elevated in women with RPL as compared to healthy pregnant women undergoing elective abortion. In ex vivo culture study, IL-17 production from CD4+ T cells was significantly higher in women with RPL and regulatory T cells from women with RPL were less suppressive to the expression of IL-17 as compared to control women. Similarly, a decrease in CD4+ CD25bright Foxp3+ regulatory T cells and increase in Th17 cells have been reported in the peripheral blood of women with RPL in comparison with normal healthy pregnant women. The ratio of Th17/regulatory T cells was significantly increased in women with RPL as compared to normal pregnant and non-pregnant women. The proportion of regulatory T cells negatively correlated with the proportion of Th17 cells (Table 1). Serum IL-17 levels correlated positively with Th17 cells and the ratio of Th17/regulatory T cells. These results suggest that regulatory T cells inhibit IL-17 expression and suppressive function of regulatory T cells on Th17 cells may decrease in women with RPL.
Our group recently published a study that investigated pro-inflammatory cytokines (TNF-α, IFN-γ, and IL-17), anti-inflammatory cytokine IL-10, and Foxp3 in the PBMCs of idiopathic women with RPL. IL-17+ T cells including Th17 and Tc17 cells significantly increased in non-pregnant women with RPL as compared to fertile women. Among Foxp3+ regulatory T-cell subpopulations, Foxp3+, Foxp3low, and CD4+ Foxp3+ T cells were significantly decreased in women with RPL, but Foxp3high and CD4−Foxp3+ T cells were not different. However, each ratio of IL-17+ cells/Foxp3+ T-cell subsets was significantly elevated in women with RPL as compared to fertile women. Interestingly, the level of IL-17+ T cells was positively correlated with CD3+ CD4+ TNF-α+ T cells and the ratios of Th1/Th2 CD3+ CD4+ TNF-α+cells/CD3+ CD4+ IL-10+ cells and CD3+ CD4+ IFN-γ+ cells/CD3+ CD4+ IL-10+ cells. These results suggest that women with RPL have propensity of pro-inflammation via Th1 and Th17 immunity and decreased immune regulatory function by Foxp3+ regulatory T cells.
To achieve successful pregnancy, both immune tolerance and an effective immune defense are required. A new immune effector, Th17 cells, may be the missing component in the Th1/Th2 paradigm and be responsible for the inflammatory processes that cannot be explained by Th1 or Th2 immunity. Regulatory T cells play a role as a key regulator to counteract the effector cells such as Th17 cells. An elaborate immune balance between immune effectors and immune regulators is crucial to achieve implantation and maintain pregnancy until term. In addition to Th1 and Th2 immunity, it becomes evident that Th17 immunity and regulatory T cell-mediated immune regulation are deeply involved in pathogenesis of RPL. Further studies are needed to elucidate the immune mechanism operating during implantation and pregnancy.