SEARCH

SEARCH BY CITATION

Keywords:

  • Decidual NK cell;
  • Granulysin;
  • regulatory T cell;
  • miscarriage;
  • Th17 cell

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Immunopathology
  5. Current clinical approach
  6. Potential clinical approach
  7. References

Immunological dysfunction has been proposed to explain the etiology of recurrent pregnancy loss (RPL). The immunological environment differs between the decidua basalis and decidua parietalis, and also between RPL cases with normal fetal chromosomes and those with abnormal fetal chromosomes. The problem with analyzing decidual tissues from spontaneous abortions is that cause versus effect phenomena are difficult to distinguish. Recent data show that the immune system in a late-stage miscarriage is completely different from that in an early-stage miscarriage. If immunocompetent cells can cause RPL, the immunological environment may be a causative factor, especially in an early-stage miscarriage, at the decidua basalis, and/or in cases of RPL with a normal embryo. Careful examination of the immune system at the decidua basalis in an early-stage miscarriage in RPL cases with normal fetal chromosomes may reveal useful information. This paper aimed at finding a cause of RPL by analyzing the balance of the immune system between T cells and NK cells in an early-stage miscarriage.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Immunopathology
  5. Current clinical approach
  6. Potential clinical approach
  7. References

Recurrent pregnancy loss (RPL) has many known etiologies such as parental cytogenetic anomalies, uterine malformation, endocrine dysfunction, infections, antiphospholipid syndrome, and inherited thrombophilia, yet these identified causes account for only 20–50% of cases.[1-3] A significant proportion of RPL cases are associated with immune etiologies; however, 50% of unexplained RPL is attributed to chromosome abnormalities in the embryo. If immune dysfunction causes RPL, the immunological environment may be important, especially in cases with a normal embryo. The fetus is a semi-allograft for the maternal host, and the maternal immune system is activated by fetal antigens at the feto-maternal interface. In the early stage of pregnancy, fetal tolerance by maternal immune cells is vital to the maintenance of a pregnancy. Maternal tolerance toward fetal alloantigens has been explained by a predominantly Th2-type immunity during pregnancy.[4] Other reports support the idea that Th2-dominant immunity protects the fetus from maternal Th1 cell attack.[5, 6] We previously reported decreases in Th2 and Tc2 cells at the decidua basalis but not decidua parietalis in unexplained RPL with normal chromosomal content.[7] Our data also suggested a shift in the Th1/Th2 balance toward a Th1-predominant stage at the implantation site in unexplained RPL with normal chromosomal content. However, this simplistic characterization did not fully explain the immunopathogenesis of abortion/miscarriage.[8, 9] Indeed, Th2-dominant immunity was reported in recurrent miscarriage cases, and even in the combined absence of IL-5, IL-9, and IL-13, mice showed a normal pregnancy, even in cases of allogeneic pregnancy.[10] However, this work did not include IL-10 – a Th2-type cytokine that has been reported to be of importance in pregnancy.[11-14]

Recently, the classic Th1/Th2 paradigm of effector T cell differentiation has been expanded by the discovery of another lineage of T helper cells that includes Th17 and Th9 cells,[15] and immunoregulatory T cells, termed Treg cells. IL-17 has an important role in induction of the protective immune response against extracellular bacteria or fungal pathogens.[16, 17] Conversely, regulatory T (Treg) cells are potent suppressors of inflammatory immune responses and play pivotal roles in the maintenance of allografted pregnancies in mice and humans.[18-20] Recent findings show the reciprocal development of pathways between the Th17/Treg subsets,[21-23] and an imbalance of Th17/Treg cell development has been reported in recurrent pregnancy loss (RPL).[19, 24, 25] One should therefore discuss again reproductive events from the view point of the new Th1/Th2/Th17 and Treg paradigms. Additionally, we have noticed that immunological changes at the implantation site (decidua basalis) are usually more emphasized compared with those at the decidua parietalis. The distribution of immune cells in the decidua could therefore provide a clue for verifying the immunological dysfunction in cases of miscarriage. Multiple factors are known to correlate with RPL, and it is quite difficult to unravel them. However, NK cells and T cells are clearly at least a part of the cause of abortion/miscarriage in mice and humans. This paper aimed at finding a cause of RPL by analyzing the balance of the immune system between T cells and NK cells in miscarriage.

Immunopathology

  1. Top of page
  2. Abstract
  3. Introduction
  4. Immunopathology
  5. Current clinical approach
  6. Potential clinical approach
  7. References

Decidual Natural Killer Cells are Regulated by Hormones

Decidualization of the human endometrium following embryo implantation is normally associated with the massive recruitment of CD16CD56bright natural killer (NK) cells. In the early pregnant decidua, CD16CD56bright NK cells constitute the major immune cell population accounting for more than 70% of lymphocytes, whereas CD4+ T cells and CD8+ T cells are a minor population (<5%).[26, 27] NK cells interact with target cells via a series of inhibitory and activating NK cell receptors constitutively expressed by decidual NK cells. Furthermore, CD16CD56bright NK cells recognize the major histocompatibility complex class I molecules HLA-C, HLA-E, and HLA-G via these inhibitory and activating receptors.[28] To clarify the characteristics of decidual NK cells necessary for maintaining pregnancy, we compared the cytokine expression patterns in decidual CD16-CD56bright NK cells between normal pregnancy and miscarriage. Prior to this analysis, decidual NK cells were divided into four groups: NK1: Th1 cytokine-producing cells, NK2: Th2 cytokine-producing cells, NK3: TGF-β-producing cells, and NKr1: IL-10-producing cells. Normal pregnant women had a significantly increased number of NK3 cells among decidual lymphocytes, compared with women who miscarried.[29] Additionally, a significant increase in NKr1 cells was also observed among peripheral blood lymphocytes in normal pregnant women.[29] There were no significant differences in the proportion of NK1 or NK2 cells between the normal pregnant groups and miscarriage groups. NK3 cells were induced to form from peripheral blood mononuclear cells by prolactin and soluble HLA-G1 in vitro (data not shown). Conversely, NKr1 cells were generated from peripheral blood mononuclear cells by human chorionic gonadotropin and progesterone in vitro. Progesterone is known to play a crucial role in peripheral blood NK cell accumulation to the uterus.[30] Thus, progesterone, prolactin, human chorionic gonadotropin, and soluble HLA-G1 could contribute to fetal tolerance by inducing the production of immunosuppressive NK subsets (Fig. 1).

image

Figure 1. Decidual CD56bright NK cell subsets.

Download figure to PowerPoint

NK Cell-Derived Granulysin Induces Apoptosis of Extravillous Trophoblasts

As mentioned before, the proportions of immunosuppressive NK subsets, NK3 and NKr1, were suppressed in miscarriage cases, compared with those of normal pregnancy. In other words, once the immune balance of NK cells is lost, excessive activation of these cells may induce miscarriage.[31, 32] Immune changes are well known to occur in RPL, but it was unclear whether maternal NK cells directly attack fetus-derived trophoblasts in humans. To clarify the immunological causes of miscarriage, we examined the relationship between cytotoxic granule proteins, such as granulysin, granzyme B, and perforin, in decidual lymphocytes and the induction of apoptosis in extravillous trophoblasts (EVTs). In the decidua basalis, the number of granulysin-positive CD56bright NK cells significantly increased in miscarriage compared with normal pregnancy; however, granzyme B- and perforin-positive cell numbers did not change.[33] Additionally, granulysin was mainly detected in CD56bright NK cells, but not CD3+ T cells in decidual lymphocytes, and the proportion of granulysin in CD56bright NK cells was significantly higher in cases of miscarriage than normal pregnancy. In vitro study showed that IL-2-stimulated CD56bright NK cells induced apoptosis in EVT cells (HTR-8/SV40neo) dependent on cell to cell contact and the expression of perforin in decidual NK cells. Furthermore, transfected cells expressing a GFP-granulysin fusion protein induced apoptosis in HTR-8/SV40neo cells independently of caspases. Our results suggest that granulysin-positive decidual NK cells attack EVTs, causing the death of EVTs because of apoptosis (Fig. 2). The data support a new apoptotic pathway for trophoblasts via decidual NK cell-derived granulysin, suggesting that granulysin is involved in miscarriage. These NK cell-related studies clearly showed the bilateral character for inducing abortion. Conversely, the results suggest the suppression of NK cell activity to be a potential therapeutic target for RPL.

image

Figure 2. The apoptotic pathways in EVT

Download figure to PowerPoint

IL-17 in the Pathogenesis of Miscarriage

The identification of a novel helper T (Th)-cell subset, the IL-17-producing Th (Th17) cells, has provided new insight into our understanding of the molecular mechanisms of reproduction. Most of the IL-17-producing cells are CD4+ T cells in the peripheral blood and deciduas.[34-36] Our results showed that Th17 levels in peripheral blood lymphocytes did not change during normal pregnancy,[36] and the proportion of IL-17+ decidual lymphocytes was significantly higher than that of peripheral blood lymphocytes early in normal pregnancy.[36] But Santner-Nanan et al.[37] reported that the population of Th17 cells in the late stages of pregnancy was significantly smaller than that in non-pregnant subjects. These findings suggest IL-17 to play some role in the establishment of pregnancy. We therefore explored the correlation between IL-17 and miscarriage. IL-17 expression was observed in the decidua parietalis as well as decidua basalis in miscarriage cases.[35] IL-17-stained cells were restricted to lymphocytes, and staining for IL-17A was not observed in villous trophoblasts, EVTs, endometrial gland cells, or endometrial stromal cells.[34] Immunohistochemistry showed that the number of IL-17+ cells in inevitable abortion cases, which were accompanied with genital bleeding or severe abdominal pain, was significantly higher than that in normal pregnancy cases. IL-17+ cells were distributed over the entire region of the decidua, the cell column in the decidua basalis, as well as the decidua parietalis.[35] Importantly, there was no significant difference in the number of IL-17+ cells between missed abortion cases, which were accompanied with no symptoms, and normal pregnancy cases. Additionally, a significant positive correlation was observed between the number of IL-17+ cells and the number of neutrophils in the abortion cases. These findings suggest IL-17 to be involved in the appearance of symptoms in cases of miscarriage and that IL-17 expression might not be the cause, but rather an effect, of miscarriage.

The Role of Treg Cells in Pregnant Mice

We established pregnant mice with lower numbers of regulatory T cells by an administration of anti-CD25 monoclonal antibodies at days 4.5 and 7.5 of gestation. The ratio of regulatory T cells in CD4+ T cells was decreased proportionally to the dosage of anti-CD25 monoclonal antibody. The miscarriage ratio was significantly increased in proportion to the decrease in regulatory T cells in allogeneic pregnancies, but not in syngeneic pregnancies. Thus, regulatory T cells prevent fetal rejection at the early phase in an allogeneic pregnancy. Our preliminary data showed that paternal antigen-reactive regulatory T (PA-Treg) cells increased in regional lymph nodes of the uterus at day 3.5 and increased in the uterus from day 5.5 in an allogeneic pregnancy. Implantation occurs at day 4.5 in mice, and the increase in PA-Treg cells occurred before implantation in regional lymph nodes. This result suggests that implantation might enhance the migration of PA-Treg cells to the uterus for maintaining pregnancy. Additionally, seminal fluid seems to contribute to the priming by the paternal antigen of maternal immune cells.[38] Taken together, maternal immune cells recognize the paternal antigen in seminal fluid. Subsequently, PA-Treg cells increase in regional lymph nodes before implantation and help to maintain the pregnancy after implantation. Adversely, the decrease in Treg cells solely induces miscarriage in an allogeneic pregnancy. In other words, a decrease in maternal regulatory T cells could cause the fetus to be rejected, suggesting a cause of miscarriage in humans.

Current clinical approach

  1. Top of page
  2. Abstract
  3. Introduction
  4. Immunopathology
  5. Current clinical approach
  6. Potential clinical approach
  7. References

The Regulation of Peripheral NK Cell Activity Helps to Prevent RPL

We performed a clinical retrospective study to analyze the relationship between peripheral NK cell activity and RPL. Outpatients with unexplained RPL, which all had high NK cell activity in peripheral blood (>40%, n = 32), were treated with prednisolone p.o. These patients were then divided into two groups after peripheral NK cell activity was re-estimated: a normalized group (<40%, n = 21) and a high cell activity group (>40%, n = 11). The number of recurrent pregnancy losses was significantly decreased in the normalized group, compared with the high activity group (P = 0.001). All abortive fetuses were analyzed as to karyotype, and miscarriage cases with an abnormal chromosomal content were excluded. These results suggest peripheral NK cell activity to be one of the markers for disruption of fetal tolerance. However, a randomized control study is needed to estimate the relationship.

The Distribution of Treg Cells in the Decidua of Fetuses with Normal and Abnormal Chromosomal Content

We first reported that CD4+ CD25high regulatory T cells increased in human peripheral blood and further increased in the early pregnant decidua to three times the level found in peripheral blood.[19] Somerset et al.[39] reported that circulating CD4+ CD25+ T cells increase during early pregnancy, peak during the second trimester, and then decline postpartum. Interestingly, these increased peripheral and decidual CD4+ CD25high Treg cell ratios decreased to non-pregnancy levels in miscarriage or RPL cases,[19] suggesting that increased CD4+ CD25high Treg cell numbers are needed to maintain a pregnancy in humans. To further clarify whether immunological dysfunction closely relates to miscarriage in humans, a comparison between fetuses with normal and abnormal chromosomal content would be required. Our preliminary data showed the proportion of Treg cells among decidual lymphocytes to be significantly lower in cases of miscarriage with normal fetal chromosomes than in normal pregnancies. Immunohistochemistry also showed that the number of Treg cells at the decidua basalis was significantly smaller in cases of miscarriage with normal chromosomal content than those with abnormal chromosomal content. These results suggest the decrease in Treg cells at the implantation site to be the cause of miscarriage in humans and are consistent with the data obtained from the pregnant mouse model. However, further studies are needed to test this hypothesis.

Immunological Environment at the Decidua Basalis and Decidua Parietalis

Immune changes are known to occur in RPL, although it is difficult to obtain proof. From our data, numbers of decidual Th17 cells were higher in inevitable abortion cases, but not in missed abortion cases.[35] We presume that missed abortions, which have no symptoms, are an early stage of abortion, whereas inevitable abortions, which do have clinical symptoms such as genital bleeding and severe abdominal pain, are a late stage of abortion. An increase in Th17 cells was observed in inevitable abortion cases but not missed abortion cases and was observed at the decidua parietalis as well as decidua basalis in inevitable abortion cases. Th17 cells therefore might not be the cause, but rather an effect, of miscarriage. Granulysin expression in decidual CD56bright NK cells was significantly higher in miscarriage, compared with that in normal pregnancy. Immunohistochemistry showed that granulysin expression was quite significantly higher in the deicuda basalis than decidua parietalis in missed abortion cases.[33] Our preliminary findings show that there are no significant differences in the number of granulysin-positive cells in the deicuda basalis between miscarriage cases with normal chromosome content and those with abnormal chromosomal content. These results may indicate granulysin to be in a higher cascade than Th17 cells, but this might just be the effect of miscarriage. Conversely, the decrease in Treg cells was observed only in cases of miscarriage with normal chromosomal content.

Potential clinical approach

  1. Top of page
  2. Abstract
  3. Introduction
  4. Immunopathology
  5. Current clinical approach
  6. Potential clinical approach
  7. References

We propose that (i) A decrease of Treg cells disturbs fetal tolerance by activating maternal immune cells, which might be a cause of miscarriage in fetuses with a normal chromosomal content. (ii) Granulysin-positive CD56bright NK cells directly attack fetal-derived EVT cells, resulting in excessive inflammation at the feto-maternal interface. Reducing peripheral NK cell activity could decrease the rate of miscarriage for RPL patients. Further studies are needed to clarify the correlation between Granulysin-positive CD56bright NK cells and higher peripheral NK cell activity in RPL. (iii) Th17 cells directly abrogate the conceptus from the uterus at the final stage of miscarriage (Fig. 3).

image

Figure 3. The cascade of miscarriage with immune etiology.

Download figure to PowerPoint

To clarify the pathophysiology of unexplained RPL associated with immune etiologies, cases of miscarriage should be classified according to chromosomal content. With regard to immune status, a comparison between cases involving fetuses with a normal versus abnormal chromosomal content would provide useful information. To verify the role of immune cells in pregnancy, not only the proportion but also the distribution of cells should be analyzed. Immunohistochemistry would be useful to clarify the distribution of maternal immune cells in between the decidua basalis and decidua parietalis. Cause versus effect phenomena are difficult to distinguish, but the finding that immunological changes occurred only in missed abortion cases (early stage) with a normal chromosomal content is important to establishing a cause of miscarriage. The onset of miscarriage continues to be a problem. More research will explain the pathophysiology of unexplained RPL associated with immune etiology and may elucidate useful treatments.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Immunopathology
  5. Current clinical approach
  6. Potential clinical approach
  7. References