T helper cell immunity in pregnancy and influence on autoimmune disease progression.

The immunology of pregnancy underlies the intricate cellular network that is Th cell immunity, and the mechanisms that mediate this coordinated balance of immune responses, from hormone oscillations to paternal alloantigen expression on trophoblasts. While recent investigations have revealed novel pathways that contribute to shaping gestational immunity such as immune checkpoint receptors and modification in epigenetic signatures, further understanding of the mechanisms associated with dysregulation in these networks are necessary for the development of therapeutic interventions for obstetrical complications. Moreover, further investigations into the phenomenon that is pregnancy-induced remission of multiple autoimmune disorders could pave the way for novel therapeutic targets in the treatment of autoimmune disease and post-partal complications.

A successful pregnancy is dependent on a dynamic and responsive immune system that must be able to respond to invading pathogens while simultaneously maintaining tolerance towards the semi-allogeneic fetus. Key to this success is the coordinated shift in balance of T helper (Th) cell immune responses, throughout each stage of gestation. It was in the early 90’s when Wegmann et al. first hypothesized that domination of Th2 immunity during pregnancy superseded maternal Th1 responses, thus ensuring protection and development of the fetoplacental unit [ 1 ]. Sex hormones, in particular androgens and estrogens are critical in shaping Th cell immune responses during pregnancy. The necessity for adequate hormonal signalling and subsequent immunity derives from several clinical and experimental studies describing how abnormal hormonal and Th cell immune responses are associated with multiple obstetrical complications. Aberrant Th cell immunity is commonly associated with the development of autoimmune disorders and frequently presents in women of child-bearing age. Pregnancy and autoimmunity share a bidirectional relationship in that while the presence of an autoimmune disorder may increase the risk of miscarriage, the systemic remodelling of the maternal immunity system during pregnancy has been shown to influence the progression of multiple prototypical autoimmune disorders. This review will focus on the function of Th cell subsets during the stages of gestation and also discuss the evidence of how dysregulation can lead to adverse pregnancy outcomes. More recently, the dichotomy of Th cell immunity has expanded incorporating both Th17 cells and regulatory T cells (Tregs) adding further to our understanding of fetal tolerance resulting in the Th1/Th2/Th17 and Treg paradigm. We will discuss how fluctuations in sex hormones shape this paradigm and how these shifts in gestational cytokine profiles can alter the progression of various autoimmune diseases. Induction and differentiation of Th1 cells is mediated by interleukin (IL) 12, IL-18 and type I interferons (IFN) following antigen presentation ( Fig. 1 ) [ 2 ]. Th1 cells are characterized by their production of pro-inflammatory cytokines IFN-γ, tumor necrosis factor alpha (TNF-α) and IL-2 [ 3 ]. They are the main effectors of phagocyte-mediated host defence protecting against infection sustained by intracellular pathogens [ 4 ]. It had long been suggested that Th1 immunity was associated with negative pregnancy outcomes. However, recent studies refute this concept and have shown that a Th1 pro-inflammatory tissue environment is a pre-requisite for successful human pregnancy, particularly during the peri-implantation period and early placentation ( Fig. 2 ). Following implantation, trophoblasts penetrate the endometrium and differentiate in syncytiotrophoblasts where they modulate the recruitment of leukocytes including T cells to the decidua through secretion of various chemoattractant factors. During this early stage of pregnancy, Th1 cells comprise up to 40% of the decidual CD4 + cell pool [ 5 ]. Natural killer (NK) cells are effector cells activated by Th1 immune responses. During pregnancy, uterine NK (uNK) cells are the most abundant leukocyte population and play vital role in vascular remodelling mediated through the secretion of IFN-γ. uNK cells represent a unique phenotype of NK cell functionally distinct from conventional NK cells and are characterized by high expression of CD56, but expression of CD16 (Fc receptor), which is expressed on most peripheral NK cells is absent, thereby limiting antibody dependent cellular cytotoxicity [ 6 ]. Furthermore, NK cells are traditionally highly cytotoxic in nature, CD16 − uNK cells, display minimal cytotoxicity preventing killing of trophoblasts [ 7 ]. uNK cells mediate blastocyst implantation through production of vascular endothelial growth factor-C, angiopoietin 2, placental growth factor, matrix metalloproteinases (MMP) as well as Th1 associated cytokines and thus are critical in the early stages of pregnancy [ 8 , 9 ]. This early predominant pro-inflammatory Th1 environment is necessary in order to modulate trophoblast invasion, stimulate the adequate repair of the uterine epithelium and removal of cellular debris following blastocyst implantation into the uterus. TNF-α plays a critical a regulatory role in trophoblast invasion by altering trophoblast cell adhesion to laminin, inhibiting trophoblast cell mobility and regulating expression of MMPs [ 10 , 11 ]. IFN-γ signalling also, represents a key pathway in inducing arterial modification during pregnancy with expression being documented at implantation during early pregnancy in both humans and mice [ 12 – 16 ]. Moreover, numbers of circulating and uterine IFN-γ producing cells are elevated during each trimester of pregnancy compared to non-pregnant controls [ 15 , 17 ]. The necessity of IFN-γ in successful conception has been demonstrated using transgenic murine models, in which absence of IFN-γ or its receptor, resulted in increased fetal resorption rates, decidual lesions and necrosis [ 15 ]. It is clear a non-detrimental role for Th1 immunity exists during early pregnancy however, evidence has implicated aberrant Th1 immunity with the immunopathology of various obstetrical complications ( Fig. 3 ). In a BALB/c mouse model of pre-eclampsia, adoptive transfer of activated Th1-like splenocytes into allogeneically pregnant female mice resulted in pre-eclampsia-like symptoms, however, these symptoms were absent in non-pregnant mice [ 18 ]. Supporting this pathogenic role for Th1 immunity, administration of toll-like receptor (TLR) agonists in mice promoted increased expression of IFN-γ and TNF-α which was associated with increased embryo resorption rates in mice [ 19 ]. In humans, elevations in peripheral ratios of IFN-γ/IL-4, TNF-α/IL-4 and TNF-α/IL-10 in CD3+/CD8− Th cells were also demonstrated in women suffering recurrent spontaneous abortions (RSA) and with repeated in vitro fertilisation (IVF) failures compared with non-pregnant healthy women [ 20 ]. Further investigations revealed evidence of dysregulations in programmed cell death protein (PD-1) expression on Th1 (CD4+/IFN-γ+/CD279+ and CD4+/TNF-α+/CD279+) cells [ 20 ] which could explain the elevation in peripheral inflammatory cells in women who suffer RSA. Elevations in these Th1 associated cytokines such as TNF-α have been shown to contribute to gestational failure through induction of apoptosis in human trophoblast cells [ 21 ]. Moreover, using JEG-3 cells (trophoblast-derived cell line), it was shown TNF-α can inhibit trophoblast integration into maternal endothelial cell complex by blocking MMP-2 expression and stimulating an integrin switch from α6β4 to α1β1 [ 22 ]. More recently, in pre-eclamptic women, expression of a dendritic cell (DC) specific long non-coding (lnc) RNA and signal transducer and activator of transcription 3 (STAT3) were found to be elevated in the decidua, which the authors hypothesized induced the over-maturation of decidual DCs resulting in Th1 cell induction [ 23 ]. Following implantation and placentation, there is a stage of rapid fetal growth and development. At this stage, the mother, placenta and fetus are symbiotic, and the predominant cytokine milieu is quickly shifted from that of a pro-inflammatory Th1 to Th2 type anti-inflammatory environment ( Fig. 2 ). Th2 cells are characterized by production of IL-4, IL-5, and IL-13 and expression of the transcription factor GATA-3 ( Fig. 1 ). Th2 associated cytokines are detectable in all three trimesters of gestation, however, peak levels are found throughout the second trimester where a Th2 environment dominates [ 24 ]. This is the longest phase of pregnancy and inflammatory episodes during this stage are associated with negative pregnancy outcomes and has therefore led to the axiom that pregnancy is as a Th2 dominated biological process [ 1 , 25 ]. Accumulation of Th2 cells to the feto-maternal interface is mediated through several homing and induction mechanisms. Decidual stromal cells secrete chemokine ligand 2 (CCL2) while trophoblasts secrete thymus and activation-regulated chemokine (TARC) also known as CCL17, promoting the accumulation of CCR2 and CCR4 expressing T cells to the decidua [ 26 , 27 ]. Additionally, thymic stromal lymphopoietins (TSLPs) play critical roles in dendritic cell-mediated immune responses particularly in orchestrating differentiation of Th2 cells in asthma and dermatitis [ 28 – 30 ]. Human decidual CD1c(+) DCs (dDCs) highly express the functional TSLP receptor complex and recombinant human TSLP or supernatants from human trophoblasts specifically stimulate dDCs to highly produce IL-10 and CCL17 further promoting Th2 cell differentiation and homing thus, further enhancing a decidual Th2 responses [ 29 , 31 ]. Interestingly, it was found that the protein expression of TSLP in normal pregnancy with strong Th2 bias, was significantly higher than that of miscarriage, displaying a Th1 bias at the maternal-fetal interface [ 29 ]. Furthermore, dDC mediated induction of naive T cells is abolished in the presence of IL-12, which could explain why dysregulation in cytokine networks during the early stages of pregnancy lead to repeated implantation failures and spontaneous abortion [ 32 – 34 ]. More recently, it has been demonstrated that receptor activator for nuclear factor-κ B ligand (RANKL) secreted by human embryonic trophoblasts and maternal decidual stromal cells (DSCs), also plays a critical role in inducing maternal–fetal tolerance through activation of decidual macrophages (Mφ) promoting polarization towards an IL-10 producing M2 phenotype. It was further shown that co-culture of dMϕ pre-cultured with DSCs and trophoblasts and then subsequently naive T cells promoted expression of GATA-3 and IL-10 in T cells [ 35 ]. Together these data highlight the complexity and multifaceted mechanisms utilized by the gestational tissues in order to promote an anti-inflammatory Th2 bias during the fetal growth stage of pregnancy. Despite the strong evidence indicating the necessity of Th2 driven immunity in successful pregnancy, functional redundancy is highly prevalent among the Th2 interleukins and there is evidence suggesting that predominant Th2 type immunity might not be essential for successful pregnancy. Fallon et al., showed that in MHC-mis-matched animals the absence of IL-4, IL-5, and IL-13 in the fetus and IL-4, IL-5, IL-9, and IL-13 in the mother did not impair litter sizes and pregnancy success in mice compared with wild-type. Suggesting that the expression of Th2 cytokines at the maternal/fetal interface is not obligatory for normal gestation [ 36 ]. However, these data failed to highlight the critical role of IL-10 which has been repeatedly shown to be central for the induction and the maintenance of allograft tolerance [ 37 , 38 ]. Th2 immunity in particular production of IL-4, seemingly enhance IL-10 immune responses [ 39 ] however, many other factors promote production of IL-10. It could therefore be argued that the necessity of Th2 immunity in pregnancy is not absolute when sufficient IL-10 mediated immunity is present. IL-17 producing T cells, termed T helper 17 (Th17) cells, have been characterized as a distinct lineage of CD4 + T helper cells and differentiate from naive T cell precursors by the polarizing cytokines transforming growth factor-β (TGF-β) and IL-6, under the control of the transcription factor retinoic-acid-receptor-related orphan nuclear receptor alpha and gamma (RORα/γ) [ 40 , 41 ]. While considered to be pro-inflammatory in nature, it is thought the role of Th17 cells in pregnancy is to support angiogenesis and protect the materno–fetal interface against extracellular microbes [ 42 ]. Furthermore, reports also indicate Th17 cells serve to promote survival, proliferation and invasion of human trophoblast cells during the first trimester of pregnancy through the secretion of IL-17 [ 43 ]. Th17 are a heterogeneous subset of T cells able to display a significant level of plasticity. Recent data has indicated decidual Th17 cells in normal pregnancy are likely associated with Th2 responses through production of IL-4. A significant population of decidual CD4 + T cells termed Th17/Th2/Th22 cells were shown to produce IL-17, IL-4 and IL-22 [ 44 , 45 ]. These cells are found in the decidua of successful pregnancies whereas IFN-γ producing CD4 + T cells which do not produce IL-4 known as Th17/Th1/Th22+, are prevalent in patients experiencing unexplained recurrent abortion (URA) [ 45 ]. Moreover, these Th17/Th2/IL-22+ and Th17/Th0/IL-22+ cells have been found to reside exclusively at the embryo implantation site where expression levels of IL-4, GATA-3, IL-17A, ROR-C, IL-22, and AHR mRNA are elevated [ 44 , 45 ]. Despite evidence of a protective role during pregnancy, many studies associate Th17 cells with a pathogenic role in allograft rejection implicating a role for aberrant Th17 responses in negative pregnancy outcomes. Indeed several lines of evidence support this assumption with Th17 cells found to play a pathogenic role in RSA and pre-eclampsia [ 46 – 48 ]. Furthermore, in mice, administration of exogenous IL-17 has been shown to increase abortion rates [ 49 ]. Elevations in serum IL-6 and IL-17A and RORγt expressing CD4+IL-17+ cells were also reported in the CBA/J × DBA/2 abortion mouse model compared with healthy pregnancy [ 50 ]. Consistent with these data, in humans, higher proportions of Th17 cells and their associated cytokines, including IL-17 and IL-23, have been described in both peripheral blood and decidual tissue of women suffering RSA [ 51 ]. Finally, in pre-eclampsia patients, multiple studies have reported elevations in Th17 cells and IL-6 skewing the balance of decidual Th17 and T regulatory cells (Tregs) in favor of a pro-inflammatory Th17 predominant milieu [ 46 – 48 ]. The proposed mechanisms perpetuating Th17 responses in pregnancy have pointed towards impaired IL-27 signalling. IL-27 inhibits IL-17 and enhances IL-10 expression in a dose-dependent manner [ 52 ]. Expression of IL-27 is lower in the decidua of patients with unexplained recurrent miscarriage compared with spontaneous miscarriage and normal pregnancy [ 52 ]. Conversely, elevations in IL-27 drive the expansion of IL-10 producing Th17 cells in patients with endometriosis [ 53 ]. More recent data has identified high mobility group protein B1 (HMGB1), a cytokine mediator of inflammation secreted by innate immune cells as a potential indicator of excessive Th17 responses. Activation of TLR4 promotes HMGB1 secretion by macrophages and drives IL-23 and IL-17 expression in T cells [ 54 ] and several studies have described elevations in HMGB1 in women suffering pregnancy morbidities [ 55 – 59 ]. Dysregulation in the expression of the immunoregulatory checkpoint receptor, PD-1 has also been associated with negative pregnancy outcomes; proportions of PD-1+Th17 cells (CD4+/IL17+/CD279+) were significantly lower in women suffering recurrent pregnancy loss (RPL) [ 60 ]. Collectively, these data indicate Th17 immunity plays a critical role in pregnancy. However, aberrant Th17 responses which are the result of a multitude of dysregulated immune pathways can lead to a shift the balance in between Th17/Tregs leading to adverse pregnancy outcomes. Tregs are a subpopulation of T cells that regulate immune homeostasis and maintain immunological self-tolerance by suppressing the activity and proliferation of self-reactive lymphocytes through the production of immunosuppressive cytokines IL-10, TGF-β and IL-35 [ 61 , 62 ]. They are characterized by low/negative expression of CD127 and constitutive expression of IL-2 receptor (CD25) and forkhead box transcription factor (FoxP3), which acts as a major regulator in their development and function [ 63 , 64 ]. Tregs are vital during pregnancy in the maintenance of the maternal immune tolerance against semi-allogeneic fetus [ 65 ]. Upon conception, there is a systemic expansion in Tregs which rapidly localise preferentially to the decidua indicating Tregs provide decidual immunoregulation from the beginning of gestation [ 66 – 68 ]. Functional chemokines and their receptors are widely expressed at the maternal–fetal interface and various chemokine networks mediate this migration of Tregs to the decidua [ 69 , 70 ]. A number of mechanisms mediate local induction of Tregs at the maternal-fetal interface. These mechanisms include expression of indoleamine 2,3-dioxygenase (IDO) which is central regulator in the induction of maternal tolerance. IDO is an enzyme expressed by decidual trophoblasts and macrophages, which catabolises the degradation of tryptophan (Trp) to kynurenines (Kyn). Kyn exerts immunomodulatory effects via activation of the transcription factor aryl hydrocarbon receptor (AhR) expressed by T cells, which upon ligation promotes the differentiation of Tregs [ 71 – 73 ]. Moreover, interaction between decidual NK and myelomonocytic CD14 + (dCD14+)+ cells results in the production of IFN-γ, a potent inducer of IDO expression in dCD14+ cells thereby further enhancing IDO mediated induction of Treg within decidual tissues [ 74 ]. Other mechanisms include expansion and recruitment of granulocytic myeloid-derived suppressor cells (G-MDSC) to the fetal-maternal interface. G-MDSCs are enriched in the decidua and possess a variety of immunomodulatory properties [ 75 , 76 ] and at the feto-maternal interface have been shown to induce Treg expansion through induction of Foxp3 in CD4 + CD25 − T cells [ 77 ]. Murine experiments have demonstrated evidence of alloantigen independent Treg expansion with increased levels of Tregs being found in both syngeneic and allogeneic matings. However, contrary to this Zhao et al., found a significantly higher number of Tregs in pregnant mice from allogeneic versus syngeneic matings, suggesting a significant involvement of paternal antigens in Treg expansion. Observations in mice exposed to seminal fluid in the absence of conceptus have demonstrated the ability of paternal antigens to drive antigen specific-Treg activation and expansion [ 78 , 79 ]. Furthermore, following parturition these paternal antigen-specific Tregs are able to persist at elevated levels and maintain immunosuppressive capacity. Upon subsequent pregnancies with the same paternal background, these cells rapidly re-accumulate and expansion correlates with decreased fetal resorption [ 80 ]. The importance of Tregs for successful pregnancy was first described by Aluvihare et al., who provided the first compelling evidence indicating that Tregs are essential for embryo implantation and progression of pregnancy. Adoptive transfer of lymphocytes depleted of CD25 + cells into T cell deficient BALB/c mice resulted in gestation failure [ 68 ]. Moreover, anti-CD25 mediated depletion of Tregs in the implantation phase and early pregnancy increases the frequency of activated CD8 + and CD4 + T cells in para-aortic lymph nodes resulting in an increase in resorption rates in vivo [ 68 , 81 ]. Although, depletion in later stages of pregnancy does not induce any abnormal pregnancy outcomes unless mice receive a second-hit inflammatory challenge, suggests that Treg cells may not be essential for immune tolerance in the later phases of pregnancy [ 82 ]. Consistent with the findings of Aluvihare et al., stimulation of Tregs, either directly by administration of low dose of IL-2 or indirectly by Fms-related tyrosine kinase 3 ligand resulted in normal pregnancy rates in CBA × DBA/2J abortion-prone mice [ 83 ]. Similarly, observations in humans have demonstrated the critical role for Tregs in pregnancy with maldistribution and functional impairment of Tregs being associated in implantation failure, miscarriage, and preeclampsia in humans [ 67 , 84 , 85 ]. For example, reduced FoxP3 mRNA expression was found in the uterine endometrium in primary unexplained infertility patients [ 86 ]. Moreover, the frequency of Tregs is decreased in peripheral and decidual lymphocytes in human miscarriage, with peripheral Treg frequencies shown to be a predictor of miscarriage risk in newly pregnant women with a history of conception failure [ 84 , 85 ]. The findings of these experiments and clinical observations implicate that in allogenic matings, Tregs are a necessity for the prevention of adverse maternal immune responses against the fetus. The oscillations in systemic immunity between inflammation and tolerance during pregnancy are accompanied by many hormonal fluctuations ( Fig. 5 ). Sex hormones have profound effects on the immune system and play a critical role in shaping Th cell immunity throughout stages of pregnancy. Androgens are considered to promote anti-inflammatory responses whereas estrogens can exhibit both pro- and anti-inflammatory roles depending the relative expression of estrogen receptor (ER) isoforms. However, significantly high dosages of estrogens such as those observed in pregnancy typically suppress immune responses. Three predominant naturally occurring estrogens have been described in women, namely estrone (E1), estradiol (E2), and estriol (E3). Estrogens function primarily through binding to two intracellular receptors, estrogen receptor α (ERα) and ERβ, members of the nuclear receptor family and are ubiquitously expressed in a variety of cell types and tissues including CD4 + and CD8 + T cells. E2 is the predominant estrogen produced during the reproductive years and during pregnancy, E2 levels increase continuously throughout gestation. Estrogens have profound immunomodulatory properties and it is thought that elevated levels of E2 during pregnancy is one of the critical mechanisms in the induction of maternal tolerance towards the fetus. Indeed, pregnant and E2 treated mice show enhanced expression of FoxP3 in CD4 + CD25 + T cells and increased suppressive Treg capacity [ 87 , 88 ]. Pregnancy levels of E2 also influence CD4 + T cell polarization through enhanced expression of Th2 associated (GATA3, IL-4) and Treg associated genes (Foxp3, PD-1, IL-10, and TGF-β) while suppressing the expression of Th1 associated (T-bet, IL-2, TNF-α, IFN-γ) and Th17 associated genes (ROR-γt, IL-6, IL-17, IL-23) [ 87 – 90 ]. The similarities observed in expansion of suppressive T cell populations and inhibition of effector T cells in pregnant and E2 treated mice suggest that estrogens are likely a major contributing factor in shifting the balance between Th1/Th2 and Th17/Treg cells observed in pregnancy. Progesterone (P4) is a female sex hormone commonly referred to as ‘the pregnancy hormone’ critical for the establishment and maintenance of successful pregnancy. P4 is a nuclear hormone receptor ligand with pleiotropic functions including strong immunomodulatory properties [ 91 ]. T cells express the progesterone receptors (PR-A, PR-B, and PR-C) and also membrane progestin receptors (mPRa, mPRb, and mPRg) [ 92 ]. During pregnancy P4 induces anti-inflammatory responses and promotes tolerance through the induction of Th2 and Tregs subsets [ 93 ]. In mice, administration of P4 not only increased the proportion of CD4 + CD25 + T cells, but also enhanced suppressive function through increases in IL-10 expression. At physiological doses equivalent to those observed in mid-term pregnancy, P4 stimulated conversion of CD4 + CD25 − T cells into CD4 + CD25 + T cells [ 94 ]. Evidence indicates that one mechanism in which P4 induces Tregs is through suppression of the mammalian target of rapamycin (mTOR) pathway which has a key role on T cell differentiation. Lee et al., demonstrated that P4 decreases the phosphorylation of S6 ribosomal protein, a major substrate of p70 S6 kinase that is a downstream target of the mTORC1 kinase [ 95 ]. During pregnancy, the placenta becomes a major organ for P4 production but is also produced by cord blood (CB) erythroblasts and is therefore present at high levels within CB [ 96 ]. Interestingly, within the CB, P4 has been shown to further promote immunomodulatory responses through the selectively inducing the differentiation of naive CB T cells into Tregs, while suppressing their differentiation into inflammatory Th17 cells, potentially through suppression of the IL-6 receptor expression [ 97 ]. Finally, the immunoregulatory properties of P4 have been utilized in the treatment of adverse pregnancy outcomes. P4 or progestogens (PR agonists) have been administered therapeutically at all stages of pregnancy including luteal-phase support prior to pregnancy, threatened miscarriage, unexplained recurrent miscarriage, and to prevent preterm labor [ 98 – 102 ]. Prolactin is a 23-kD peptide hormone secreted in the pituitary gland. During pregnancy, levels increase and reach peak values, remaining elevated throughout breastfeeding [ 103 ]. Prolactin is involved in multiple physiological functions including immune modulation. It has been shown to favor the survival and differentiation of T-cell progenitors [ 104 ]. Moreover, at high levels such as those observed in pregnancy, prolactin is capable of inducing the interferon regulatory factor (IRF-1) expression, a key transcription factor driving the Th1 phenotype in T cells [ 105 ]. Consistent with these data, coculture of T cells with prolactin promotes expression of a Th1 cytokine profile with increased production of TNF-α, IFN-γ and IL-2 in CD4 + and CD8 + T effector cells [ 106 – 108 ]. However, whereas other sex hormones appear to promote immunosuppression, prolactin appears to have a negative effect on the suppressive function of Tregs [ 108 ]. It is possible, that the spike in prolactin levels observed in the third trimester of pregnancy aid the switch to a Th1 pro-inflammatory environment that has been shown to be indispensable for initiating parturition [ 109 , 110 ]. hCG is a placental glycoprotein hormone required to establish pregnancy. Active secretion of hCG is initially produced by the blastocyst 6–8 days following fertilization [ 111 ] then later by the trophoblasts [ 112 , 113 ]. hCG levels peak during the first trimester and decrease to approximately 10% of the peak value during the second and third trimesters [ 114 ]. The immunoregulatory properties of hCG have been well described in multiple leukocyte populations highlighting a critical role in programming maternal tolerance towards the embryo especially during the early stages of pregnancy. Th cell immunity is strongly affected by hCG as it has been proposed to be the primary driver in facilitating the switch from a Th1 inflammatory to a Th2 ant-inflammatory environment [ 115 ]. Treatment of activated dendritic cells with hCG results in an up-regulation of MHC class II, IL-10 and IDO expression, reducing the ability to stimulate T cell proliferation and promoting expansion of Tregs [ 116 ]. Exposure to hCG also enhances the suppressive capacity of Tregs through increased secretion of IL-10 and TGF-β [ 117 – 120 ]. In mice, hCG can stimulate expansion of decidual and peripheral CD4+FoxP3+ T cells [ 120 ]. Similarly, coculture of human CD4+FoxP3-T Cells with hCG producing JEG-3 trophoblast cells promoted conversion into suppressive CD4+FoxP3+ T Cells [ 121 ]. Additionally, hCG has been demonstrated to possess chemoattractant properties [ 118 , 122 ] and is able to attract Treg cells to the fetal–maternal interface, with recent data indicating that hCG mediated induction of CCL2 in human decidual stromal cells, triggers recruitment of CCR2+ Tregs to the endometrium [ 122 ]. The necessity of hCG for manipulating Th cell immunity during the early stages of pregnancy has become ever more apparent due to emerging data demonstrating it to be an effective therapeutic intervention in women who have suffered previous adverse pregnancy outcomes. Sha et al. reported that in women who had suffered multiple unexplained RSA before week 20 of gestation, following hCG therapy, all pregnancies were ongoing at 28 weeks of gestation and were associated with a reduction ratio of peripheral Th17/Tregs [ 117 ]. Consistent with these data, recurrent implantation failure (RIF) patients receiving IVF, hCG therapy improves the clinical pregnancy rate, implantation rate and live birth rate by increasing peripheral Tregs [ 123 ]. While hCG therapy has been shown to ameliorate the peripheral Treg imbalances, there is also evidence that hCG augments local Treg imbalances. Women experiencing RIF exhibit lower endometrial Tregs frequencies however, intrauterine infusion of 2000IU hCG significantly increased the number of endometrial Tregs to levels comparable to controls [ 122 ]. By this means hCG likely has pleiotropic functions in facilitating the expansion of systemic and local Treg pools through induction of Tregs, conversion of conventional T cells and stimulating active Treg recruitment mechanisms. The distinction between foreign and self by the mammalian immune system is not absolute, and aberrant responses characterized by self-reactivity towards self-organs and tissues has been implicated in more than 80 inflammatory disorders. Autoimmune diseases are a clinically heterogeneous group of chronic diseases that individually, affects only a small number of individuals, but as a whole, it is estimated that prevalence is between 7.6 and 9.4% [ 124 ]. Gender dimorphism represents one of the most enigmatic observations among the labyrinth of autoimmunity with many autoimmune conditions displaying a strong female bias. The immunoregulatory properties elicited by sex hormones are believed to play a crucial role in this dimorphism. CD4 + Th cells play central roles in maintaining immune homeostasis and host defence. However, propagations and dysregulations in Th cell mediated immunity have been well documented as a major driver of several autoimmune diseases. As described above, pregnancy and the concomitant changes in sex hormones levels, induces profound changes in systemic Th cell immunity suppressing Th1 and Th17 responses and favoring a shift towards a Th2/Treg dominant cytokine environment. For many Th1/Th17 associated autoimmune conditions this results in favorable disease outcomes whereas Th2 associated diseases frequently suffer flares or relapse in disease. Indeed, several animal models and clinical studies have described the influence of pregnancy on disease progression and Th cell immunity in multiple autoimmune disorders ( Table 1 .). Multiple Sclerosis (MS) is a chronic neuroinflammatory disorder of the central nervous system (CNS) affecting up to 2.5 million people worldwide, with a female/male sex ratio of 3:1 [ 125 , 126 ]. Pro-inflammatory Th1 and Th17 cells targeting CNS self-antigens, have been highly implicated in the pathogenesis of MS and are found to be enriched in MS lesions, whereas Tregs are dysfunctional and reduced in number [ 127 – 130 ]. In murine models, it is most frequently represented by experimental autoimmune encephalomyelitis (EAE), where activation of myelin-specific CD4 + Th1 and Th17 cells is sufficient to induce MS-like disease [ 131 – 133 ]. Pregnancy is protective in MS, substantially reducing the relapse rate, particularly during the third trimester, however, the short-term positive effect of pregnancy is then followed by a temporary rebound of disease activity post-partum [ 134 – 137 ]. Expansion of antigen specific Tregs and inhibition of effector T cells driven by pregnancy associated hormone fluctuations have been associated with significant improvements in disease activity in mice ( Fig. 4 ) [ 138 ]. Indeed, several studies support this assumption with evidence indicating estrogen as the primary driver of pregnancy-induced protection in MS. In the EAE model, administration of estrogens has been shown to elicit anti-inflammatory responses promoting expression of immunosuppressive cytokines [ 139 ], upregulation of checkpoint receptors [ 140 ] and expansion of Tregs resulting in inhibition of autoantigen specific Th1 and Th17 cell responses. These data prompted the use of estrogens as a therapeutic approach in an attempt recapitulate the beneficial effects of pregnancy on MS. A pivotal clinical study almost two decades ago showed administration of 8 mg/day for 6 months of E3 improved disease activity in non-pregnant females [ 141 ]. More recently, a longitudinal study by Voskuhl et al., showed that E3 in combination with glatiramer acetate was effective in reducing relapse rates in women over a period of 24 months [ 142 ]. The exact mechanisms of how estrogen exerts immunoregulatory effects on Th cell immunity in MS remain unclear. However, recent data has described how estrogen can alter epigenetic signatures in Th17 and Tregs in humans. Using an integrative data analysis, Iannello et al., identified cell-type regulatory regions (CSR) regulated by Erα. Treatment with 17-estadiol induced active histone marks enrichment at FoxP3-CSR and repressive histone marks enrichment at RORC in Th17 polarised cells [ 143 ]. Within relapse-remitting MS patients, a disease-specific epigenetic profile was identified during pregnancy which lead the authors to suggest a FOXP3 positive regulation and a RORC negative regulation in the third trimester of pregnancy [ 143 ]. A recent study has also implicated the existence of antigen-specific immunomodulatory mechanisms during pregnancy. Ramien et al., characterized immunomodulation at the single-clone level by sequencing the T cell repertoire in both female MS patients and healthy women over the course of pregnancy. Their findings revealed that T cell clonality is significantly reduced from the first to third trimester in terms of both absolute and relative change in MS patients compared with controls. This lead the authors to suggest that the T cell repertoire becomes more evenly distributed and less dominated by expanded clones with only a few T cell clones, in particular antigen experienced CD45RO expressing CD4 + and CD8 + T cells being substantially modulated during pregnancy in each patient [ 144 ]. Moreover, relapse-associated T cell clones identified in an individual patient seemingly contract during pregnancy and expand during a postpartum relapse [ 144 ] thereby suggesting that profiling the T cell repertoire during pregnancy could provide a valuable tool to highlight and track “private” T cell clones associated with disease activity in MS and other autoimmune conditions. Systemic lupus erythematosus (SLE) is an autoimmune disorder affecting several organs, the skin and the kidneys being the most distressed. Hormones elicit a strong influence on SLE evident by fluctuations in disease activity during hormonal events such as menses, menopause, and pregnancy. The synergic changes pregnancy and the associated hormonal alterations exert on the balance of Th1 and Th2 cytokines favor disease outcomes in Th1 driven autoimmune conditions. However, SLE is a Th2 associated autoimmune disease [ 145 ] and therefore the strong Th2 bias observed during pregnancy should have negative implication for SLE activity. Indeed, using the lupus-prone mouse model MRL-1 (MLR/lpr), it was described that administration of estrogen or pregnancy were both able to induce exacerbations in kidney disease and subsequent survival was dramatically reduced [ 146 , 147 ]. Consistent with these data, clinical studies on SLE patients show 20–68% of women experience a flare in disease during pregnancy [ 148 – 152 ], however, there are conflicting reports suggesting disease course does not differ in pregnant versus non-pregnant patients. Discrepancies between these data could be attributed to the inconsistency in which flares were defined, making it difficult to make comparisons across studies. Many studies suffer with small sample sizes, which reduced power to determine differences in the rate of flares between pregnant and non-pregnant patients. A recent study by Eudy et al., on 1349 SLE patients (398 pregnancies in 304 patients) showed that the incidence of flare (defined by Physician Global Assessment) was increased during pregnancy and within the 3-months post-partum [ 152 ]. However, this risk was modified by hydroxychloroquine (HCQ) use. HCQ has also been shown to reduce the risk for pre-eclampsia in pregnant SLE patients [ 153 ] thus, further advocating the continued use of HCQ during pregnancy and post-partum. In addition to exacerbations in disease activity during pregnancy, patients suffering from SLE and lupus nephritis are also associated with an increased risk of negative pregnancy outcomes. A systematic review which included 1842 patients and 2751 pregnancies revealed that unsuccessful pregnancy rate was 23.4%, and premature birth rate of 39.4% among pregnanct SLE women [ 149 ]. Furthermore, a more recent study showed that the presence of lupus anticoagulant, antihypertensive use, PGA score greater than 1, and a low platelet count were all risk factors associated with adverse pregnancy outcomes with absence of these baseline predictors conferring more favorable pregnancy outcomes in SLE patients [ 154 ]. Rheumatoid arthritis (RA) is a chronic, progressive autoimmune disease primarily affecting the lining of the synovial joints. The precise aetiology of RA remains unknown, however the relative abundance of T cells in the inflammatory infiltrate of affected joints, together with the strong association of the disease with molecules involved in T cell activation (such as HLA–DR and protein tyrosine phosphatase N22 [PTPN22]), indicate a pathological role for T cells. Indeed, several lines of evidence describe infiltration and activation of Th1 and Th17 immune responses within the inflamed joints of RA patients and in animal models [ 155 – 160 ]. RA predominantly affects post-menopausal women, however, is also found in women of child-bearing age. As first described by Hench in 1938, the majority of RA patients experience spontaneous improvement in disease activity during pregnancy however, this effect is transient, and aggravation of disease symptoms is commonly observed after delivery [ 161 ]. As described in MS, a systemic expansion in Tregs and skewing towards a Th2 dominant cytokine profile has been associated with pregnancy-induced remission in RA ( Fig. 4 ). Indeed, frequencies of CD4 + CD25 + Tregs have been shown to be inversely correlated with disease activity in the third trimester and post-partum in pregnancy RA patients [ 162 ]. The necessity for Tregs in suppressing RA during pregnancy was highlighted by Munoz-Suano et al., who firstly showed that pregnancy protected mice from collagen induced arthritis (CIA), but also remarkably demonstrated that adoptive transfer of CD4 + CD25 + Tregs from these pregnant mice, was sufficient to confer protection to non-pregnant mice [ 163 ]. Sex hormones are thought to be the primary mediators of RA disease activity during pregnancy; estrogens have been repeatedly shown to confer protection through suppression of Th17 responses [ 164 – 166 ], while prolactin and lactation have been associated with post-partum flares, disease onset and relapse in disease [ 167 ]. This has been exemplified in women who breastfed as they display a higher risk of developing de novo RA and around 90% will flare within the first 3 months postpartum [ 168 , 169 ]. Moreover, severity of disease has been shown to correlate with duration of breastfeeding and higher number of breast fed children [ 170 ] implicating prolactin-mediated immune stimulation. While administration of bromocriptine suppresses post-partum flares in CIA mice [ 171 , 172 ], treatment in humans has yielded inconsistent findings [ 173 – 175 ] and thus far has not been trialled in pregnant patients. Despite these data, avoidance of breastfeeding may be beneficial for at risk RA patients following pregnancy. Inflammatory bowel disease (IBD) is a group of chronic diseases characterized by inflammation of the gastrointestinal (GI) tract and affects men and women most commonly in their reproductive years. The most common of these are Crohn disease (CD) and ulcerative colitis (UC). The pathogenic mechanism of IBD which include genetic predisposition, environmental factors, microbiome alterations and dysregulations in GI immunity [ 176 – 179 ], overlap strongly with those of RA [ 180 ] and therefore, many treatments options are effective across both disease [ 181 ]. Yet, whereas RA has a well-documented clinical course during pregnancy characterized by disease ablation commonly followed by post-partum flares, the data surrounding pregnancy in IBD is far from conclusive. One might postulate as observed in RA, pregnancy would impel positive outcomes on IBD activity however, conflicting reports of the effect of pregnancy in IBD have been documented. In general, the majority of women with IBD (up to 80%), whether CD or UC, conceive during disease remission which could explain why a significant proportion of pregnant IBD patients experience a similar clinical course as non-pregnant women. Yet, while some studies have indicated pregnancy stimulates improvements in clinical activity and post-partum relapse rates [ 182 ] others have refuted these findings reporting no improvements and even worsening in disease course [ 183 , 184 ]. These improvements have been attributed to HLA disparity and cessation of smoking during gestation and may not be associated with pregnancy-induced changes in immune cell dynamics [ 185 , 186 ]. Furthermore, disparity in disease behaviour between CD and UC has been documented with UC patients displaying a higher risk of relapse during pregnancy and post-partum compared to non-pregnant UC patients, whereas CD patients display a similar disease course as non-pregnant controls [ 183 ]. Differences in disease behaviour may potentially be explained by intrinsic differences in the immunopathology associated with each disease [ 187 ]. While involvement of Th17 cells have been reported in both CD and UC [ 188 – 191 ], CD has been shown to be a predominantly Th1 driven disease [ 192 – 194 ] which would favor the pregnancy induced shift to an anti-inflammatory Th2 cytokine profile. However, UC has frequently been described as a Th2 dominant disease [ 195 , 196 ] and thus pregnancy-induced cytokine shift risks aggravating disease activity. Several factors seemingly can influence disease course of IBD during pregnancy and post-partum. While the majority of IBD medications are safe to take during pregnancy, data suggests an association between combination therapy and the risk of preterm births in UC and caesarean section in CD patients compared with monotherapy [ 183 ]. However, it is possible that this may merely reflect a more aggressive, complicated disease course requiring modulation in anti-inflammatory therapies. Therefore, predicting whether patients experience positive or negative effects may depend on individual patient characteristics, however, further exploration into pathology of pregnancy and IBD is required. Autoimmune hepatitis (AIH) is a chronic inflammatory liver disease characterized by immune mediated injury affecting the hepatocytes [ 197 ]. AIH has a strong female predominance and can affect women of child-bearing age [ 197 ]. While there exists substantial data from both clinical and animal studies concerning autoimmune diseases such as RA, MS and SLE, studies investigating the influence of pregnancy on AIH are comparatively sparse, therefore, little is known regarding the evolution of AIH during pregnancy. This could be attributed to the lack of a reliable disease animal model that mirrors the pattern injury observed in humans and also, the rarity of disease in comparison to other autoimmune conditions. Of the limited studies describing pregnancies with AIH, retrospective analyses have documented conflicting reports. Like other prototypical autoimmune disorders, the pathogenesis of AIH implicates aberrant Th1 and Th17 immune responses [ 198 – 200 ], therefore the pregnancy-induced switch from a Th1 dominant to a Th2 dominant cytokine prolife should favor AIH disease progression ( Fig. 4 ). Indeed, high rates of remission have been documented during pregnancy albeit in small case cohorts [ 201 – 205 ] however, Braga et al., only observed clinical improvements in during gestation in 11% of pregnancies [ 206 ]. Additionally, flares in disease during pregnancy can occur in 20–30% of patients [ 203 , 207 – 209 ], particularly in those who were not on therapy or who had a disease flare in the year prior to conception [ 207 ]. Post-partum episodes of disease flares are also frequently described [ 201 , 203 ], likely a consequence of augmentations in Th1 immunity mediated by immune stimulatory prolactin. Post-partal development of AIH has also been described in multiple centres [ 210 – 212 ]. Pre-emptive increase of the immunosuppressive therapy is sometimes advocated after delivery. While some authors have suggested pregnancy induces favorable maternal outcomes for AIH during gestation, the relationship is not reciprocal, with the presence of AIH being associated with high fetal loss rates of 30–40% [ 206 , 207 , 209 , 213 ] and lower live birth rates, particularly in patients with cirrhosis [ 207 ]. Discussions between AIH patients and their obstetrician and hepatologist are therefore advised prior to conception to ensure disease is well controlled in order to limit the both the fetal and maternal risks associated with pregnancy in AIH.