The Role of NK Cells in the Autoimmune Thyroid Disease-associated Pregnancy Loss
- First Online:
- Cite this article as:
- Konova, E. Clinic Rev Allerg Immunol (2010) 39: 176. doi:10.1007/s12016-010-8201-7
- 212 Views
Pregnancy loss is a frequent event. Autoimmune thyroid disorders and altered natural killer (NK) cell functions are two distinct risk factors, which independently could induce adverse pregnancy outcome. Thyroid autoimmunity has been an object of increased attention by investigators in the context of pregnancy loss. Peripheral NK cells and uNK cells comprise distinct cell populations in terms of phenotype and function but they play an important role in the course of a normal human pregnancy via several potential functions. In autoimmune thyroid diseases, several abnormalities of killer cell activity have been described. The functional defects involving NK maturation and/or functional activation observed in Graves’ disease patients could independently influence the reproductive outcome. This suggestion needs extensive investigation and could be important for the therapeutical approach in preventing pregnancy loss in patients with thyroid autoimmunity.
KeywordsPregnancy lossThyroid autoimmunityNatural killer cells
Pregnancy loss is a frequent event. According to the estimation of Edmonds et al., 1982 , fetal viability is only achieved in 30% of all human conceptions, 50% of which are lost prior to the first missed menses. About 25% of implanted embryos are resorbed within 7-14 days after attachment to the uterine endometrium . The loss of clinically recognized pregnancies prior to the 20th week of gestation occurs at a frequency of 15% . Two to five percent of women have recurrent pregnancy loss (RPL), defined as three or more consecutive losses of intrauterine pregnancies before the 28th gestational week [4, 5]. The risk of having another abortion after three consecutive abortions can be as high as 55% . Probably these percentages underestimate the actual frequency of miscarriages.
Even multifactorial, the cause of repeated pregnancy loss can be divided into embryologically driven factors (abnormal embryonic karyotype) and maternally driven factors which affect the endometrium and/or placental development . The identifiable causes of maternal defects include endometrial defects, endocrine disorders, coagulation disorders, and immune factors [7, 8]. Current medical literature suggests that extensive diagnostic screening identifies abnormalities in only 55-65% of patients [9, 10], but it has been postulated that a proportion of identifiable and unidentifiable losses may be due to immune causes.
Thyroid autoimmunity and pregnancy loss
Since 1990, thyroid autoimmunity has been an object of increased attention by investigators in the context of pregnancy loss. It affects 5-10% of the female population of childbearing age and is the most frequent underlying factor leading to, or associated with, thyroid underfunction. With regard to the thyroid, both hypothyroidism and hyperthyroidism have long been associated with increased fetal loss, but such association is usually reversible after appropriate treatment of the underlying thyroid disease. During the first trimester of pregnancy, proper maternal thyroid function is crucial, because the developing fetus is completely dependent on the mother for thyroid hormone. Autoimmune thyroid disease, including Hashimoto thyroiditis, Graves’s disease and primary myxedema, is the most common endocrine disorder encountered during pregnancy. For the first time, Stagnaro-Green and colleagues reported a doubling of the miscarriage rate in unselected euthyroid pregnant women with positive thyroid antibodies . They tested 552 women and found that 20% of them were positive for thyroid antibodies. Of these antibody-positive women, 17% miscarried, compared with only 8.4% of the antibody-negative women. Serum thyroid-stimulating hormone (TSH) levels were abnormal in six out of 17 thyroid antibody-positive women with a miscarriage. This first study on the association between miscarriage and thyroid autoimmunity was rapidly followed and confirmed by other controlled studies [12–28]. In 2004, Prummel and Wiersinga  performed a meta-analysis of 18 controlled studies (eight case-control studies [12–19] and ten prospective [11, 20–28]) performed since 1990 when the association was first described. For all the studies, the authors have calculated the odds ratio (OR) with 95% confidence interval (95% CI) according to Sanderocock . Both types of studies revealed a similar and significant relationship between thyroid autoantibodies and miscarriage: the cumulative OR was 2.73 in the eight case-control studies and 2.30 in the ten prospective studies. Meta-analysis results were further confirmed in 2008 by other controlled study, including 641 patients with RPL . Furthermore, it was established that thyroid autoimmunity was strongly related to unexplained infertility and implantation failure , as well as to significantly lower clinical pregnancies in IVF [28, 33]. Three hypotheses tried to explain the unknown etiology of increased pregnancy loss in women with thyroid autoimmunity [34–36]. The first hypothesis holds that pregnancy loss is not directly related to the presence of circulating thyroid antibodies. Antibodies only constitute a marker of an underlying more generalized autoimmune imbalance (antiphospholipid syndrome, systemic lupus erythematosus, etc.) that, in turn, causes pregnancy loss. However, in some controlled studies women with other than thyroid antibodies were excluded or were analyzed separately [11–13, 18, 22, 24, 26, 27]. The second hypothesis holds that despite apparent euthyroidism, the presence of thyroid autoimmunity could be associated with a subtle deficiency in thyroid hormone concentrations (euthyroid women with anti-thyroid peroxidase (TPO) antibodies have slightly higher TSH values than those without antibodies); this may indicate less thyroidal reserve in times of greater demand for thyroid hormones, such as in pregnancy . The third hypothesis holds that pregnancy is delayed in such women (thyroid autoimmunity is a risk factor for infertility) and the older age by itself, even in healthy women could constitute an additional risk factor for increased pregnancy loss [29, 38]. Recently, investigators try to evaluate the effect of thyroid antibody (anti-TPO and anti-thyroglobulin) on pregnancy in experimental autoimmune thyroiditis mouse models [39, 40]. Lee Y. and colleagues studied the effects of TPO antibody in female C57bl/6 mice immunized with recombinant mouse TPO on preimplantation embryo development and implantation rate . They established higher serum TSH levels and increased incidence of resorbed fetus in the immunized anti-TPO positive mice compared with the controls. Anti-TPO antibody bound to preimplantation embryos and marginally decreased the formation of 3/4 cell embryos but had no effect on the subsequent development and implantation. Authors conclude that autoimmune thyroiditis is associated with reduced fertility and higher incidence of fetal loss, caused by anti-TPO antibody affection of post-implantation embryo development. In humans, significantly higher anti-TPO antibody avidity was established in pregnant women in their first trimester who had a history of recurrent miscarriage and whose pregnancies failed again later in the first trimester compared to those whose pregnancy continued .
Natural killer cells in pregnancy and pregnancy loss
Cellular constituents underlying immune-mediated losses are another object of increased attention by investigators in the context of pregnancy loss. Оne of the most studied and controversial areas in reproductive medicine is the relationship between natural killer (NK) cells and reproductive failure. NK cells are innate immune effectors, able to exert a prompt cytolitic activity against infected and tumor cells without prior sensitization and without restriction by HLA antigens. NK cells are found in both peripheral blood (peripheral NK cells) and the uterine mucosa (uterine NK cells). Both peripheral NK cells and uNK have been the focus of investigation of many studies trying to understand the physiology of normal pregnancy and pathophysiology of unexplained RPL.
In peripheral blood, NK cells comprise about 10–15% of blood lymphocytes  and can be divided in two populations, based on the intensity of CD56 expression. About 90% of peripheral blood NK cells are CD56dim and express high levels of CD16; the other 10% of peripheral blood NK cells are CD56bright and express low levels or no CD16 . CD56dim cells are more cytotoxic, whereas the CD56bright subset is the main source of NK cell-derived immunoregulatory cytokines . NK cell function is mainly regulated by interleukin (IL)-2 and interferon-gamma (IFN-γ). IL-2 causes both NK cell proliferation and enhanced cytotoxicity whereas IFN-γ augments NK cytolytic activity, but does not cause NK proliferation . Under resting conditions, peripheral NK cells express few cytokines; however, they can be induced to express granulocyte macrophage colony-stimulating factor (GM-CSF), macrophage CSF (M-CSF), IL-3, IFN-γ, TNF-α, and TNF-ß with various stimuli [44, 45]. TNF-α and IFN-γ, secreted by activated NK cells, play important roles in induction of apoptosis and regulation of the immune response. NK cells express a variety of inhibitory receptors that interact with class I molecules and deliver inhibitory signals to NK cells upon recognition of major histocompatibility complex (MHC)-I-expressing cells. These receptors consist of: (1) the killer inhibitory receptors (KIRs), which recognize mainly the classical MHC class I antigens—HLA-A, HLA-B, and HLA-C; and (2) the CD94/NKG2 receptor, which recognizes the nonclassical MHC molecule, HLA-E . Activating NK cell receptors include: CD16, NKp46 and KIR 2DS . Hormonal regulation of peripheral NK cells could be suggested by their changes during normal pregnancy: (1) they decrease in number, mainly as a result of a decrease in the CD16+ subset ; (2) exhibit decreased lytic activity compared with NK cells from controls ; (3) the expression of inhibitory receptors is increased among peripheral NK cells in the first weeks of pregnancy, reaching a maximum within the third month of gestation, with a subsequent decline to basal levels by the end of pregnancy .
uNK cells resemble the dominant CD56 population of peripheral blood NK cells in some phenotypic characteristics and probably derive from a small peripheral blood CD56bright population; however, they have some phenotypic and functional differences. uNK express CD56 as well as the killer activatory and inhibitory receptors, but lack expression of other NK markers (CD16 or CD57)  and in contrast to the majority of peripheral NK cells, express CD69, an early activation marker [43, 44]. Additionally, uNK cells express a different cytokine profile, compared with resting peripheral NK cells. mRNAs for TNF-α, IFN-γ, TGF-ß, G-CSF, M-CSF, GM-CSF, and leukemia inhibitory factor have been found in uNK cells [45, 46]. In contrast, in the resting state of peripheral blood NK cells mRNAs for only TGF-ß1 and TNF-α are detected . In the endometrium, NK cells constitute the predominant leukocyte population present at the time of implantation and in early pregnancy and comprise over 70% of endometrial leukocytes in first trimester decidua . Not only the number but also uNK cell phenotype changes during the normal menstrual cycle and early pregnancy . For example, expression of the activation antigens CD69 and HLA-DR is highest in the proliferative phase and decreases gradually during the menstrual cycle .
Peripheral NK cells and uNK cells comprise distinct cell populations, in terms of phenotype and function but they play an important role in the course of a normal human pregnancy via several potential functions: peripheral NK cells down-regulate the activity of uNK cells in a normal pregnancy and uNK cells provide appropriate cytokine support and local immunomodulation leading to: initiation of the process of decidualization , regulation of placental and trophoblast growth by cytokines [52, 53]; local immunomodulation by glycodelin and galectin-1  and control of trophoblast invasion .
There is considerable evidence of hormonal regulation of peripheral and uNK cells. In peripheral NK cells, hormonal regulation may occur through direct actions of estrogen through estrogen receptors , through direct actions of progesterone via an as yet unidentified receptor or receptor-independent mechanism , or through indirect pathways, involving progesterone action on peripheral T cells which could act through cytokines on peripheral NK cells . In vitro studies provide supporting evidence for progesterone regulation of proliferation and differentiation of uNK cells  as well as of expression and production by them of immunomodulatory substances (glycodelin, galectin-1) [54, 60]. This regulation may occur through direct actions on uNK—via regulation of gene expression of immunomodulatory proteins and through indirect pathways—via hormonal effects on intermediary cells (T lymphocytes and endometrial stromal cells ) and chemokines (macrophage inflammatory protein-1ß; MIP-1ß) and vascular endothelial growth factor (VEGF) . During the early stages of normal pregnancy, progesterone enhances migration of peripheral CD56bright NK cells to the uterus both by directly inducing expression of molecules facilitating such homing and by up-regulation of MIP-1ß and VEGF production by endometrial stromal cells.
A lot of observations suggest that reproductive outcome may be influenced by changes in peripheral and uterine NK cell numbers, phenotype, or function. Several studies have attempted to associate peripheral NK cell activity or elevated NK cell numbers with reproductive failure. Ntrivalas et al.  established significantly higher expression of CD69 on all studied subsets of NK cells (CD56dim, CD56bright, and CD16neg) from women with RPL compared with controls. The expression of the CD94/NKG2 inhibitory receptor was significantly decreased compared with controls. An increased preconceptional NK cell activity (measured by a chromium-51 release cytotoxicity assay) was reported by Aoki et al.  in 68 women with unexplained RPL compared with 47 controls. Among women with a history of RPL, enhanced NK cell activity attributed a relative risk of 3.5 for miscarriage in the next pregnancy, compared with women with normal NK cell activity . In another study, the authors noted an increase in NK cell cytotoxicity in early pregnancy, peaking at 8 weeks, as well as an increase in the percentage of CD56+CD16+ cells in 43 women with RPL . Probably, the creation of a Th1 cytokine environment in the periphery in RPL may lead to NK cell activation and proliferation (via the actions of Th1 cytokines Il-2 and IFN-γ.), which could result in migration of cytotoxic NK cells into the uterus and cause miscarriage. Additionally, an imbalance of CD69 and CD94 expression on peripheral blood NK cells in women with RPL may account for their pathology .
uNK cells assayed from women with miscarriages may be uninformative, since uNK cells may undergo apoptosis when progesterone levels decrease during reproductive failure. The changes observed in miscarriage could be due to the necrosis and inflammation following the death of the embryo, thus they might be the result, rather than the cause of miscarriage. Flow cytometric and immunohistochemical analyses of uNK cells in RPL have conflicting results and must be interpreted with caution. Flow cytometry analysis of midluteal endometrial samples from 20 women with a history of two or more unexplained miscarriages showed no difference in the percentages of uNK cells, CD56+CD16−, or CD56+CD16+ uNK cell subsets compared with samples from control subjects, while observing a significant decrease in the percentage of T lymphocytes in RPL patients . Lachapelle et al.  analyzed by flow cytometry the phenotype of endometrial T, B, and NK cells from secretory phase endometrial specimens from 20 women with RPL compared with 15 fertile controls. In contrast, women with RPL had a greater percentage of CD16+CD56dim NK cells and a smaller percentage of CD16– CD56bright cells, compared with control patients . Immunohistochemical analyses of uNK cells in RPL are also controversial. When frozen sections of luteal-phase endometrial biopsy specimens of women with unexplained RPL were evaluated, an increase in the numbers of CD56+ cells compared with controls was observed in two studies [68, 69], and one of these also reported an increase in CD16+ cells . Another immunohistochemistry study that used formalin-fixed tissue embedded in paraffin showed no difference in the numbers of the endometrial CD56+ and CD16+ cell populations .
It is possible, that the uNK cells that normally reside in the endometrium and support the development of a healthy pregnancy are not involved in the pathogenesis of RPL. Upon activation and proliferation, peripheral NK cells that are normally not present in substantial numbers in the uterus could infiltrate it and result in increased numbers of CD56dimCD16+ cells in the endometrium. Any dysregulation of immune response, including hormonal, at any level, could be responsible for these effects in RPL.
NK cells in autoimmune thyroid disorders
Autoimmune thyroid disorders and altered NK cell functions are two distinct risk factors, which independently could induce adverse pregnancy outcome. However, if the pathogenesis of thyroid autoimmune disorders is associated with alterations in NK cell function, or if altered by other factors NK cells could affect thyroid function and autoimmunity, it is important to know if these two factors could influence each other during pregnancy and increase the risk of pregnancy loss.
NK cell cytotoxicity (NKCC) mainly results in cytolysis of tumor, virus-infected and microbial cells, without a prior sensitization with target cell antigen. NK cells could function also as killer cells that mediate antibody-dependent cellular cytotoxicity, in particular during cytokine modulation . Howеver, it has been suggested that NK cells play an immunoregulatory role in the prevention of some autoimmune diseases . The mechanism by which NK cells could influence autoimmunity is still controversial . In multiple sclerosis  and Crohn’s disease , a role for viral antigen was suspected to cause disturbances in NK cell activity. Decreased NK activity, due to antilymphocyte antibodies (ALA) in patients’ sera, has been found in rheumatoid arthritis , lupus erythematosus , and Sjogren’s syndrome . It is possible that autoimmune diseases could be dependent on chronic viral infections due to the decreased NKCC against virus-infected cells or to NK modulation of autoimmune responses by the regulation of B and T cells survival and expansion. NK cells produce some Th2 cytokines, such as IL-5 and IL-13, that may enhance B cell activity and indirectly suppress Th1 autoimmune cell-mediated responses .
It has been known for decades that the neuroendocrine system can both directly and indirectly influence the functional activity of the immune system. This concerns particularly the immune–endocrine interactions of the hypothalamus-pituitary-thyroid axis, which primary function is to regulate thyroid hormone synthesis and production. Interactions between pituitary–thyroid hormones and the immune system are mainly based on the existence of receptors for thyreotropic and thyroid hormones on lymphocytes. In contrast, far less is known about the mechanisms by which the immune system collaborates in the regulation of thyroid hormone activity in physiological conditions as well as during immunological stress. Evidence for the production of TSH by cells of the immune system was first demonstrated by Smith E. et al.  and Kruger T. et al. . Initial studies used human leukocytes, which produced TSH following stimulation with staphylococcus enterotoxin A and thyrotropin-releasing hormone [78–80]. TSH can be produced by many types of extra-pituitary cells, including bone marrow hematopoietic cells, splenic dendritic cells, T cells, B cells, and intestinal epithelial cells [81–86].
The data for a hematopoietic source of TSH raises the question as to how immune system TSH takes part in physiological and pathological immune process. It is reasonable to assume that if TSH can be produced by leukocytes, it may act as a cytokine-like regulatory molecule within the immune system. Studies of Coutelier J et al.  and Bagriacik E. et al.  gave evidence for this demonstrating the expression of TSHR on lymphoid and myeloid cells, which could mediate the ability of TSH to influence lymphocyte functional behavior [89, 90].
Provinciali M et al.  established that TSH has a co-stimulatory activity for NK cells in combination with Il-2. They investigated the effect of TSH on both the proliferative capacity and the NK cell activity of murine spleen lymphocytes and found that TSH at various concentrations significantly increased the proliferative response of mouse lymphocytes to both concanavalin A and phytohemagglutinin. The administration to cell cultures of TSH alone did not induce a significant stimulation of proliferative capacity. The authors studied the mechanism by which TSH improved the mitogen-induced lymphocyte proliferation by testing the effect of TSH on lymphocytes directly stimulated with recombinant Il-2 and established that there was a great increase in Il-2-induced lymphocyte proliferation by TSH. The studies carried out on the cytotoxic activity of NK cells showed that TSH was able to significantly increase the Il-2-induced NK cell activity without modifying the basal levels of cytotoxicity.
It is well established that the proinflammatory Th1-type cytokine Il-2, together with TNF-α and IFN-γ, play pivotal roles in the pathogenesis of organ-specific autoimmune diseases, including Graves’ disease and Hashimoto's thyroiditis. Migita K. et al.  investigated the sensitivity of thyroid epithelial cells (thyrocytes) to IL-2 activated NK cells. The thyrocytes were lysed by autologous and allogenic IL-2-activated killer cells and there were no differences in sensitivity to the NK cells between normal thyrocytes and thyrocytes from patients with Graves’ diseases. Thyrocyte lysis caused by various humoral and cellular factors is a characteristic feature of autoimmune thyroiditis. Findings of Migita K. et al. suggest that the cellular interaction between thyrocytes and IL-2-activated NK cells may play a role in the modulation of thyrocyte destruction in Hashimoto thyroiditis. This is supported by the observation that treatment with IL-2 and lymphokine-activated killer cells caused hypothyroidisms in patients with advanced neolasmas . It could be suggested based on the studies of Provinciali  and Migita  that increased serum TSH levels could be additional factor enhancing the IL-2-mediated NK cell activation.
Benhadi N et al.  examined the relationship between maternal TSH concentrations in early pregnancy and the risk of miscarriage, fetal, or neonatal death in a cohort of 2,497 pregnant women without overt thyroid dysfunction. The incidence of child loss increased by 60% for every doubling in TSH concentration. This association remained after adjustment for smoking, age, parity, diabetes mellitus, hypertension, previous preterm deliveries, and previous preterm stillbirth/miscarriage. The authors concluded that the risk of child loss increased with higher levels of maternal TSH. Maternal FT4 concentrations and child loss were not associated.
Study of the first trimester thyroid function values and association with TPO antibody in a cohort of 668 pregnant women without known thyroid disease established that serum TSH levels were higher (p < 0.001) and serum T4 levels and free T4 index values were marginally lower in TPO antibody positive women (p = 0.03 and p = 0.06, respectively) compared with TPO antibody negative women . The study supports preliminary findings [11, 37] that TPO antibody levels are associated with higher TSH values. TPO antibody is a risk factor for postpartum thyroiditis, miscarriage, and premature birth [29, 96, 97]. Most women with detectable TPO antibody do not have clinical hypothyroidism, although, they do tend to have higher TSH and lower free T4 index values in the first trimester than women without thyroid antibodies test values [95, 98, 99]. Several studies confirmed that not only is overt hypothyroidism associated with maternal and fetal adverse consequences, but also subclinical hypothyroidism or euthyroidism in patients affected by thyroid autoimmunity may adversely affect the mother or fetus [95, 98–101].
In autoimmune thyroid diseases several additional abnormalities of killer cell activity have been described [102, 103]. The studies of NK activity in PBL from Graves’ disease (GD) patients by phenotypic analysis or cytolytic assay have produced widely controversial results, with reports of the activity being decreased [104–106], enhanced , or normal [108, 109]. ALA, which constitute anti-asialo ganglioside membrane 1 antibodies, a marker for NK cells, have been detected in sera of patients with GD and Hashimoto's thyroiditis (HT) [103, 110]. The reduced effector activity in PBL from hyperthyroid patients would seem to be due to a functional defect rather than to a decreased NK cell count, and the incubation of PBL of GD patients with recombinant human interleukin-2 promptly reverses the NK cell defect . Increased activity was found in hyperthyroid Graves’ and HT patients , whereas NK cell activity was found to be reduced in GD and HT patients . All these observations suggest that in GD patients, there is a functional defect involving NK maturation and/or functional activation.
Solerte S. et al.  evaluated the functional alterations of spontaneous and IL-2-/IFN-ß-mediated NKCC and of TNF-α release from circulating NK cells in subjects with GD and HT. They established that in subjects with GD and HT the percentage of CD16 +/CD56+ cells was similar to that found in healthy subjects, but NKCC was depressed and the secretion by NK of the inflammatory cytokine TNF-α was reduced under stimulation with LPS and IL-2. The depression of NK cells was, therefore, related to all of the functional aspects linked to their immune activity (i.e., cytolytic and secretory functions). As suggested in another study , the NK defect would seem to be due to a functional alteration rather than to a decreased number of NK cells. Authors suggest, in agreement with this and prevoius studies [111, 115], that in subjects with clinical thyroid autoimmunity there is a functional defect involving a subpopulation of mature cytotoxic NK lymphocytes either in the stage of basal pre-activated function (spontaneous NKCC) or during the specific dose-dependent activation with cytokines and LPS. Since the study demonstrated the absence of correlations between the serum levels of thyroid hormones and the immunological parameters, NK alterations would seem to be directly associated with the autoimmune condition linked to GD and HT pathogenesis. This evidence was also supported by the demonstration of the persistance of the NK defect even during the normalization of thyroid metabolic patterns with methimazole in GD subjects and with L-thyroxine in HT patients. Other studies also found no effect on NK cell function during in vitro pharmacologic exposure with relevant concentrations of methimazole . Authors conclude that the depression of NK activity could imply the potential expansion of T/B cell functions with a consequent up-regulation of auto reactive T lymphocytes, the production of thyroid-specific auto antibodies and lymphocytic migration and infiltration into the thyroid gland. Therefore, the complexity of NK functional depression could potentially be related to the pathophysiology of thyroid autoimmunity, also suggesting NK dysregulation as a trigger factor for GD and HT immunopathogenesis. These functional disorders are present before treatment and persisted during the normalization of thyroid function by methimazole and replacement therapy with L-thyroxine. NK immune cells are altered either during spontaneous conditions and IL-2/IFN-ß modulation or during the intracellular pathway leading to synthesis and release of the inflammatory cytokine TNF-α .
In conclusion, the alterations in NK cell functions present in autoimmune thyroid disorders such as Graves’ and Hashimoto’s diseases could be another suggestion trying to explain the unknown etiology of increased pregnancy loss in women with thyroid autoimmunity. Peripheral NK cells play an important role in the course of a normal human pregnancy via down-regulation of the activity of uNK cells and providing of appropriate cytokine support and immunomodulation. The functional defects involving NK maturation and/or functional activation observed in GD patients could independently influence the reproductive outcome. This suggestion needs extensive investigation and could be important for the therapeutical approach in preventing pregnancy loss in patients with thyroid autoimmunity.