Recurrent miscarriages: What is the role of cytokines?

Abstract

Recurrent miscarriage is defined as three or more consecutive miscarriages before 20 weeks of gestation. This condition is a frequent reproductive problem worldwide, affecting up to 1% of couples. Immune effector cell dysfunction has been implicated in the pathogenesis of early pregnancy loss. This dysfunction may involve defects in cytokines, growth factors and immunosuppressive factors at the maternofetal interface. This is a growing research field, especially the role of cytokines in recurrent miscarriage.

Keywords:

• Recurrent miscarriage
• cytokines
• early pregnancy

Introduction

Recurrent miscarriage is defined as three or more consecutive miscarriages before 20 weeks of gestation [1],[2]. This condition is a frequent reproductive problem worldwide, affecting up to 1% of couples [3],[4]. This incidence is greater than one would expect by chance alone. If up to 15% of all clinically recognized pregnancies end in miscarriage, then the theoretical risk of having three consecutive miscarriages is 0.34%[5]. Therefore, there will be a persistent underlying cause for the pregnancy loss in a proportion of women suffering from recurrent miscarriage [6].

The pathophysiology is complex, and involves both maternal and fetal factors, with the possibility of more than one underlying contributory factor [6]. The two main independent risk factors for recurrent miscarriage are advanced maternal age and greater number of previous miscarriages [2]. Maternal underlying mechanisms include uterine pathology, endocrine dysfunction, thrombophilias and antiphospholipid syndrome [7]. Genetic or developmental abnormalities in the fetus can also contribute to recurrent miscarriage [1]. However, up to 50% of cases are idiopathic [1].

Immune effector cell dysfunction has been implicated in the pathogenesis of early pregnancy loss [8]. This dysfunction may involve defects in cytokines, growth factors and immunosuppressive factors at the maternofetal interface [6]. This is a growing research field, especially the role of cytokines in recurrent miscarriage.

What are cytokines?

Cytokines are soluble proteins or glycoproteins that are produced by leukocytes and in many cases also by other cell types. They act as chemical communicators between cells, but are not effector molecules in their own right. A feature that is common among most cytokines is that they are regulators of host defense against pathogens and/or the inflammatory response. Most cytokines are secreted, but some can also be expressed on the cell membrane, while others are held in reservoirs in the extracellular matrix [9].

Cytokines bind to specific receptors on the surface of target cells. These receptors are coupled to intracellular signal transduction and second messenger pathways. Most cytokines are growth and/or differentiation factors. They generally act on cells within the hematopoietic system [9].

There are four key features displayed by most cytokines.

1. Pleiotropy. Most cytokines have more than one action. This is because their receptors are expressed on multiple cell types. Consequently, the signaling pathways activated will increase gene expression specific for that cell type.

2. Redundancy. Most cytokines have biological effects also observed in other cytokines. This is because of similar signals generated by different cytokines due to similarities in the amino acid sequences within cytokine receptors.

3. Potency. Most cytokines act in the nanomolar to femtomolar range. This is because the cytokine receptors have a very high affinity for their ligands.

4. Action as part of a network or cascade. Most cytokines are part of a cascade of cytokines released in succession. They often act in synergy, and, in most cases, are counter-regulated by inhibitory cytokines or soluble receptors.

Cytokines are grouped into families, depending on the structure of their receptors, which have highly conserved structural features. This provides an explanation for the redundancy in biological effects among cytokines.

There are six major families of cytokine receptors (Table I). Each family is defined by similar sequences in their cytosolic domains. Within each family, there are homologies in receptor sequence both intracellularly and extracellularly. However, there are also differences, particularly extracellularly, where additional domains can be found.

Table I.  The six different families of cytokine receptors (modified from [9])

Family	Key receptor feature	Members include	Shared function	Cytokine structure
Hematopoietin receptors (type I receptors)	cytosolic box 1/2; WSXWS sequence extracellularly	IL-6R, G-CSFR, IL-12R, IL-2Rβ, IL-2Rγ, IL-4R, IL-3Rα, IL-7Rα, IL-9R, GM-CSFR, EpoR, PRLR, GHR, TpoR	T/B-cell activation; hematopoiesis	four α-helical bundles
Interferon receptors (type II receptors)	cytosolic box 1/2; fibronectin domains extracellularly	IFNα/β, IL-10R	antiviral (not IL-10)	four α-helical bundles
TNF receptors	cytosolic death domain; four cysteine-rich regions extracellularly	p55 TNFR, p75 TNFR, LT(TNF)βR, CD40, CD30, CD27, OX40, TRAMP (DR3), TRAILR (DR4), 4-1BB, NGFR	proinflammatory	jelly roll motif
IL-1/Toll-like receptors	cytosolic Toll/IL-1R domain (IL-1R subgroup) or leucine-rich repeats (TLR subgroup) extracellularly	IL-1RI, IL-1RII, IL-18R α/β-chains, TIGGIR-1, SIGIRR TLR1-TLR10	proinflammatory	β-trefoil
Tyrosine kinase receptors	cytosolic tyrosine kinase domain	M-CSFR, EGFR, TGF, IGFs, FGFs	growth factors	β-sheet
Chemokine receptors	seven transmembrane spanning regions	IL-8, MCPs, RANTES, Eotaxin	chemotactic	triple-stranded antiparallel β-sheet in Greek key motif

In general, the functions of cytokines fitting into each family are also conserved [10]. But since cytokines are so pleiotropic, it is difficult to classify them according to function. General principles on the functioning of cytokines within families have emerged. This can provide a framework from which to approach the complexities involved when studying the function of cytokines.

What is the immunology of pregnancy?

Recent advances in reproductive immunology have aimed to elucidate ways in which the fetus evades specific immunological attack. The maternal immune system in pregnancy adapts in such a way that there is a degree of suppression to avoid rejection of fetal (paternal) tissues, while maintaining immune competence to fight infection. There are, however, other potential functions of immune cells and cytokines, which include aiding fetal–maternal communication and maternal recognition of pregnancy, so that appropriate physiological changes can be made, e.g. in the clotting system and vascular endothelium. In fact, some cytokines also act as local growth factors and encourage fetal growth [11].

Medawar had observed that rejection of skin grafts could be prevented by the administration of cortisone and, because he was aware that steroid levels were elevated in pregnancy, he made the ‘fetal allograft’ hypothesis. He postulated that the anatomical separation of the fetus from its mother, the antigenic immaturity of the fetus and the immunological inertness of the mother led to the survival of the fetus [12]. However, an alternative model of pregnancy was also postulated, leading to the Medawar–Shwartzman paradox.

The localized Shwartzman reaction of hemorrhagic necrosis is induced in some animals by a priming injection of endotoxin followed by a sublethal intravenous endotoxin injection 24 h later [13]. Identical pathological lesions are observed in pregnant animals after just a single intravenous injection of endotoxin, without requiring a priming injection [14]. Pregnant women may be similarly prone to exaggerated responses to endotoxin during pregnancy. The ‘priming agent’ may be cortisone itself, through non-specific mechanisms involving monocytes and natural killer (NK) cells [15]. Tumor necrosis factor-α (TNF-α), interferon-γ (IFN-γ) and interleukin-2 (IL-2) have been identified as key cytokines for the generalized Shwartzman reaction [16]. Thus the maternal immune system is suppressed in one model, but primed to respond in the other. The explanation lies in the differential modulation of the two main arms of the immune system: the innate (non-specific) and adaptive (specific), both of which have cellular and humoral components [11].

The innate system is the first line of defense, leading to ‘signal 1’. It also instigates the adaptive system, which is effective at eliminating infection. The antigen is displayed on the cell surface in association with major histocompatibility complex proteins. This triggers the activation of CD4⁺ helper T-cells (Th cells) which produce cytokines, and in turn coordinate the effector response (‘signal 2’) [17]. In ‘signal 2’, the innate system appears to be able to distinguish between dangerous and non-dangerous [18].

T-helper 1 versus T-helper 2 response

Activated Th cells can differentiate into Th1 and Th2. These have different immune responses via different patterns of cytokine production [19],[20]. Studies on murine and human pregnancy show a strong association between maternal Th1-type immunity and pregnancy loss on one hand and, on the other, Th2-type immunity, together with transforming growth factor-β (TGF-β) secreted by Th3 cells, and successful pregnancy.

Th1 cells secrete TNF-α, TNF-β, IFN-γ and IL-2. Low levels of TNF-α are also derived from some Th2 cells and cytotoxic T-cells. However, since much higher levels of TNF-α are produced during Th1 responses and due to its cytolytic effects contributing to cellular immunity, this cytokine can be regarded as exclusively associated with a Th1 or Th1-like response. Th1-associated cytokines activate macrophages and cell-mediated reactions involved in resisting infections due to intracellular pathogens and in cytotoxic and delayed-type hypersensitivity reactions. This type of reaction is also known as a type 1 or cell-mediated immune response [20],[21].

The Th2 cells stimulate B-cells and eosinophils. The Th2-type cytokines are the following interleukins: IL-4, IL-5, IL-6, IL-10 and IL-13. These are associated with strong antibody responses to infections with extracellular organisms. Such immune responses are also known as type 2 or humoral reactions [20],[21].

Th3 cells are a unique subset of CD4⁺ helper T-cells usually found in mucosa and are involved in the downregulation of Th1 and other immune cells. In addition, these two responses are known to be antagonistic to each other. In fact, IL-10 acts on the antigen-presenting cells thus inhibiting development of Th1 cells. Similarly, IFN-γ prevents the activation of Th2 cells.

It is hypothesized that during pregnancy, there is a subtle immunological shift to the Th2-type cytokine responses that would suppress the potential harmful effects of the cell-mediated (Th1-type) immune system [22]. Animal studies have shown that powerful Th2-promoting factors such as progesterone [23] and prostaglandin E₂ [24],[25] are present in high concentrations at the maternal–fetal interface. In humans, there is a wide variety of cytokines (both Th1 and Th2 types) at the maternal–fetal interface [26],[27]. Cell-specific methods have shown preferential production by trophoblast cells of IL-4 (Th2-type cytokine) rather than IFN-γ, IL-2 or TNF-α (Th1-type) [28].

Thus, pregnancy has been labeled a ‘Th2 phenomenon’. This is substantiated by the evidence that pregnancy tends to ameliorate pre-existing autoimmune cell-mediated conditions [22]. In normal pregnant women, there is significantly higher IL-10 and other Th2 cytokine production by mitogen-activated peripheral blood mononuclear cells [29],[30] and reduced IL-2 [27] and IFN-γ mRNA expression [31] and their corresponding CD4⁺ and CD8⁺ T-cells [32].

The Th1 response, especially TNF-α, IFN-γ and IL-2, has been postulated to be detrimental to the survival of the conceptus [33-35]. When administered to pregnant mice, TNF-α, IFN-γ and IL-2 cause miscarriage [36], which is then reversed by anti-TNF-α. In vitro, TNF-α and IL-2 inhibit outgrowth of human trophoblast cells [37] and synergistically stimulate apoptosis of human primary villous trophoblast cells [38]. One hypothesis states that the conceptus protects itself by secreting type 2 cytokines which downregulate harmful type 1 cytokines [22].

What is the role of cytokines in recurrent miscarriage?

It is clear from the literature that cytokines do play an important role in recurrent miscarriage. In women with a history of recurrent miscarriage, embryotoxic activity was detected in supernatants from trophoblast-activated peripheral blood mononuclear cells, which were positive for TNF-α, TNF-β and IFN-γ. In contrast, normal women and men neither were embryotoxic nor contained Th1-type cytokines, but most contained the Th2-type cytokines IL-10 and IL-4 [39]. Thus type-1 immunity to trophoblast antigens is associated with recurrent miscarriage. Mitogen-activated peripheral blood mononuclear cells derived from women with a history of recurrent miscarriage had higher levels of IL-2, IFN-γ and TNF-α, and lower levels of IL-4, IL-5, IL-6 and IL-10 than in normal pregnant women, indicating a greater type 1 bias in recurrent miscarriage and a greater type 2 bias in normal pregnancy [40],[41].

It has been shown that there is a pro type-2 shift in ratios of cytokine expression at the maternal–fetal interface in women with normal pregnancy as compared with women who suffer recurrent miscarriage [27],[42]. Th2-type cytokines are released spontaneously from placental tissue in culture [43]. This has been shown in animal studies, together with the observation that pregnancy loss can be caused by Th1 cytokines [36] and prevented by Th2 cytokines [44].

An important factor in driving naive CD4⁺ T-cells toward Th1- or Th2-dominated populations is the predominance of a given cytokine in the microenvironment at the time of antigen presentation. Th1 differentiation is promoted by IFN-γ and IL-12, while IL-4 determines Th2 polarization [20],[21]. IL-12 is produced mainly by macrophages in response to certain microbial infections; therefore even subclinical infections can lead to increased IL-12 production resulting in an overall shift towards type 1 bias. Th1 development is also enhanced by IFN-γ. IL-12 and viral infections both lead to increased IFN-γ synthesis by NK cells, in turn enhancing Th1 response.

The conceptus may be considered an ‘innocent bystander’ while the mother fights infection with a vigorous Th1-type proinflammatory response [34]. Women with recurrent miscarriage have significantly high levels of IL-12 compared with normal pregnant women [45]. Hormones like relaxin (produced by the corpus luteum) may also contribute to differentiation of Th cells or in favoring the shift of already differentiated Th cells to a type 1 profile [46]. On the other hand, progesterone has been shown to promote the development of a Th2 response, via an immunomodulatory protein known as progesterone-induced blocking factor, which inhibits several Th1-type responses in vitro [47],[48]. If there is a lack of suppression at the maternofetal interface, due to reduced levels of TGF-β2, a Th1 shift may be induced [49].

It has been proposed that Th1 immune-mediated damage could be due to NK cells, similar to activated Th1 cells, releasing cytokines harmful to the trophoblast [50]. Decidual NK cells produce IFN-γ, activating decidual macrophages to produce toxic levels of nitric oxide and TNF-α[51]. Th1 cytokines may convert the NK cells to lymphokine-activated killer cells that have been shown to lyse trophoblasts [19].

TNF-α and IFN-γ can lead to apoptosis of trophoblast cells [38], inhibition of secretion of granulocyte macrophage colony-stimulating factor from uterine epithelium [22] and upregulation of the prothrombinase fg12 (procoagulant). The latter may be the most important effect as anti-fg12 antibodies significantly reduce the effect of TNF-α and IFN-γ and prevent miscarriage. It is proposed that miscarriage of normal karyotype human embryos could be due to a process of ‘cytokine-triggered vascular auto-amputation’[52]. This involves activation of coagulation mechanisms leading to vasculitis, affecting the maternal blood supply to the implanted embryo and thrombus formation [53].

Evidence shows that TNF-α, IFN-γ and NK cells cannot induce miscarriage separately, but a Th1–NK cell–macrophage triad is known to bring about miscarriage, which can in turn be suppressed by a Th2 cytokine response [22],[54]. One should note that, depending on their time of expression, stage of gestation, relative concentrations and receptor occupancy threshold, some type-1 cytokines may actually be beneficial rather than harmful to the pregnancy [42].

Conclusion

The role of cytokines in recurrent miscarriage has become clearer over recent years. However, more research still needs to be done before one can devise treatment options for patients with recurrent miscarriage, based on the information available so far.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.