The Immunology of Preeclampsia

Introduction

Our understanding of preeclampsia has substantially improved since the original monograph of this series was published. We now know that preeclampsia is caused by syncytiotrophoblast stress, usually secondary to uteroplacental malperfusion. This leads to a maternal vascular inflammatory response that then generates the variable signs and symptoms of preeclampsia. This pathogenic concept has been encapsulated in the two-stage model of preeclampsia. Lack of understanding of the fundamental causes, is the reason why preeclampsia is still defined in terms of a syndrome, comprising maternal and fetal features of the phenotype.

Subtypes of the preeclampsia syndrome including eclampsia, HELLP, and postpartum disease, have long been recognized. Important recent additions include early and late onset preeclampsia, which are distinctive enough to be classified as different subtypes. Early onset is the less common condition, occurring in around 20% of cases, and late onset comprises about 80% of preeclampsia. Compared to early onset preeclampsia, there is a relative lack of placental pathology in late onset disease. This has led to the view that late onset preeclampsia is essentially a maternal problem expressed in the context of a normal placenta, that typically becomes evident at or around term.

An alternative and more recent proposal, which retains the central importance of placental pathology in both early and late onset forms, is that of syncytiotrophoblast stress. It is not easily detected by classical clinical histopathology but appears to be an increasing feature of all pregnancies as the placenta reaches the end of its lifetime at term and postterm (the “twilight” placenta). As do all stressed cells, syncytiotrophoblast releases many stress response factors. From the placenta, these directly enter the maternal circulation and are clearly implicated in the classical vascular inflammatory response of the maternal disease.

For many years, it has been known that restricted remodeling of the spiral arteries, associated with poor placentation, underlies early onset preeclampsia. It is an early cause of uteroplacental malperfusion, placental oxidative stress, and not unexpectedly syncytiotrophoblast stress in the first half of pregnancy.

The evidence has been presented elsewhere that there is also uteroplacental malperfusion in late pregnancy. This develops not because of dysfunctional supply, but because of expansion of the growing placenta within the confined space of the uterus. Sooner or later, it inevitably reaches its limits, causing increasing placental compression, reduction in the intervillous space and declining placental oxygenation.

In this chapter, we present a synthesis of what is known about mechanisms to bring these two apparently different sequences together as a single process sharing some common immune dysfunctions.

The Scope of this Chapter

Preeclampsia is a complex disease. Such disorders are characterized by nonlinear interactions between multiple factors including an individual's genome. However, preeclampsia is more complex because it involves two genomes (mother and fetus), three epigenomes (maternal, paternal, and placental), and their three different phenotypes. The overt disorder is detected by its maternal (and to a lesser extent fetal) features, but the placenta is the cause. Interactions between mother and placental tissue are key to successful or disordered pregnancies. This is important in determining the role of immunological mechanisms in the pathogenesis of preeclampsia, which epitomizes that complexity.

That pregnancy poses special immune challenges in the interactions between the genetically disparate fetus and mother has been recognized for nearly 100 years. The paradox was explicitly outlined by Medawar (1954) in relation to his pioneering studies of transplantation biology (reviewed in Refs. , ). It has long been speculated that preeclampsia could arise from disruption of maternal-fetal immune interactions.

The question of immune dysfunction arose from the consistent evidence that preeclampsia is mainly a disorder of first pregnancies. It was postulated that maternal immune accommodation, or “tolerance,” to fetal antigens exposed to the maternal immune system depends on mechanisms that are learned by immunoregulation after antigenic stimulation. In preeclampsia such adaptation of the immune response may be relatively defective in a first pregnancy but boosted in subsequent pregnancies, because of immunological “memory.” Immunological memory to fetal-placental antigens as a consequence of a first pregnancy seems to afford benefit for the next pregnancy ( Table 7.1 ).

Primiparity as a risk factor has now been superseded by the more specific attribute of primipaternity, after studies suggesting that a change of partner abrogated the protection gained from a previous pregnancy. ^, The evidence was largely anecdotal or uncontrolled but suggested that the issue was relevant and worthy of further investigation. An early systematic review concluded that there were still substantial uncertainties. Later evidence was derived from the use of donor oocytes in assisted reproduction, where there is no prior contact with the donor's alloantigens. This leads to a striking 4.3-fold increase in preeclampsia compared to natural conception. In pregnancies conceived with donor sperm, the risk is also increased, but remarkably this is mitigated by multiple exposures to the same donor's semen.

Such studies provoked two related explorations—of a long interpregnancy interval in parous women with previous preeclampsia, or a short interval between first coitus and conception in women of any parity. A long interpregnancy interval is associated with both a change of partner and an increased risk of recurrent preeclampsia. After the latter association was corrected, the former association disappeared. However a short interval between first coitus and conception in women of any parity has proved to be more interesting and has been confirmed in a prospective study. The concept is that the pregnancy succeeds because of progressive maternal priming and generation of “immunoregulation” toward fetal (paternal) gene proteins. This takes time to generate, and facilitates successful placentation, thereby reducing the risk of preeclampsia. It also implies that preconceptional exposure to the conceiving partner's semen or sperm contributes to building the relevant immunoregulation. This has led to the concept of “prepregnancy immune priming” ( Table 7.1 ).

It is known that paternal genes contribute to the risk of preeclampsia ; and described in Chapter 3 ). This has prompted evaluation of the protective effect of a previous abortion (conceived by the same partner, or acquired preconceptionally by exposure to paternal semen or seminal plasma. ^, Considerable scientific evidence supports a mechanism of prepregnancy priming toward male alloantigens delivered in seminal fluid, and a link between seminal fluid contact and pregnancy outcome (reviewed by Robertson et al. , Saito et al ^, ). The degree of prepregnancy priming therefore appears to be affected by several factors—including duration of exposure to the partner's seminal fluid, previous conceptions and their outcome, and the mode of conception including use of assisted reproductive technologies and donor gametes ( Table 7.1 ). This raises the prospect that the elevated risk of preeclampsia associated with primipaternity is explained by insufficient prepregnancy priming. ^,

The immune system comprises innate and adaptive responses. Innate immunity is more primitive, and acts more rapidly to give immediate immune protection. It protects organisms from “dangers” or stressors that disturb tissue homeostasis ( Table 7.2 ). Danger can take many forms, but placental oxidative stress is a relevant proinflammatory trigger in both normal pregnancy and preeclampsia. An innate immune response propagates through a network of “danger” receptors, called pattern recognition receptors, which are germ line–encoded and recognize many danger signals. The same receptor frequently can recognize both microbial and endogenous danger signals.

Generation of memory to particular paternal genes invokes the adaptive immune response. To activate an adaptive immune response, activated innate immune cells present antigens and co-stimulatory signals, and release cytokines and chemokines, to instruct adaptive immune cells (T or B lymphocytes) to generate either antigen-specific immunity or tolerance. Adaptive immunity delivers slow but precise antigen-specific responses to result in acquired immunological memory. The innate and adaptive systems are asymmetrically interdependent. The innate system does not need the adaptive system to function (although it can be modulated by adaptive immune cells), whereas adaptive immunity has an absolute requirement for signals from the innate system. Innate immunity itself stimulates generic reactions without individual specificity, and so has no capacity for antigen-specific priming or memory.

Most research of the immunology of preeclampsia is focused on the vascular inflammation that causes the maternal syndrome, which comprises features that are secondary to innate immunity and dysfunctional placental signaling. Their complexity is high, but their analysis cannot reveal an immune cause that is individual-specific for the mother or fetus or both. It extends what we know of how the mother's immune system responds to the inflammatory stimuli, but does not explain the mechanisms that underlie primipaternity. Here we focus exclusively on the latter issue—the upstream causes of placental dysfunction that arise in response to dysregulated adaptive immunity. We make no attempt to review the vast literature on the maternal vascular inflammation that is a secondary feature of preeclampsia, but the reader can find these topics discussed in Chapter 10, Chapter 12, Chapter 8, Chapter 9 .

Placentation, Trophoblast, and Human Leukocyte Antigen

Placentation

Placentation is a central focus of this chapter because inadequate placental development at the beginning of the second trimester, associated with poor remodeling of the uteroplacental spiral arteries, has long been considered the origin of maternal-placental malperfusion which leads to syncytiotrophoblast stress and preeclampsia later in pregnancy. Placentation is not confined to this particular stage of pregnancy (weeks 8–16, Fig. 7.1 ). It is an extended process which starts at or even before implantation ; and described in Chapter 4 and continues during the months when the placenta grows, differentiates into its different structural components, and establishes its two circulations—fetoplacental and uteroplacental.

As detailed in Chapter 5 , the main interface between the fetus and mother is the placenta and the principal placental-specific cell is trophoblast. To understand how immunoregulation and prepregnancy priming might operate to protect against preeclampsia, it is important to consider how maternal immune cells could affect placentation. Central questions are where and when during pregnancy are maternal immune cells exposed to trophoblast or trophoblast antigens, how do they respond, do they generate responses specific to paternally inherited fetal antigens, and what are the consequences? A subsidiary question is to what extent does maternal immune exposure to nontrophoblast paternally inherited fetal antigen occurs and what are the consequences?

Before implantation, the primitive trophoblast, trophectoderm, is the first embryonic cell lineage to differentiate. ^, After fertilization, the embryo is temporarily protected by its zona pellucida. When it is shed about five days later, the trophectoderm is the first fetal cell to be exposed to maternal immune cells in the uterine cavity and/or decidua after implantation.

Subsequent milestones include differentiation into villous and extravillous trophoblast, formation of the cytotrophoblastic shell, and growth of the chorionic villous from which trophoblast cell columns give rise to invasive extravillous cytotrophoblast. The transition from histiotrophic to hematotrophic nutrition, when the intervillous circulation opens after unplugging of the spiral arteries, causes a sudden change in oxygen tensions that brings the risk of oxidative stress and concurrently unmasks the syncytial epithelium of the chorionic villi to maternal systemic immune surveillance (summarized in Ref. ). Impaired placentation may in theory develop at different stages of the first four months of pregnancy. This may cause a spectrum of adverse outcomes ranging from infertility, through early pregnancy loss, to late pregnancy syndromes that include fetal growth restriction and early onset preeclampsia, usually in combination. ^, Placentation is the time most vulnerable to development of dysfunctional maternal immune responses to trophoblast, which is consistent with the known importance of poor placentation as a cause of preeclampsia ( Fig. 7.1 ).

The Importance of Human Leukocyte Antigens

The human leukocyte antigen (HLA) family comprises extremely polymorphic proteins that allow individual immune systems to distinguish between self, nonself, altered self, or infected self. As such HLA are major drivers of adaptive immunity and likely to be crucial players in the success of human pregnancy and the phenomenon of primipaternity. The patterns of expression of different HLA components on trophoblasts are described below and summarized in Fig. 7.2 .

Class I and II HLA differ structurally and functionally, and in their cellular expression patterns. HLA-A, HLA-B, and HLA-C are highly polymorphic members of the class Ia subfamily expressed by most nucleated cells, which interact with T cell receptors (TCRs) to trigger cytotoxic T cells. HLA-E, HLA-F, and HLA-G are weakly polymorphic members of the nonclassical class Ib family, with more limited cellular expression. Like class Ia, class II HLA (HLA-D) is extremely polymorphic, but differs in its receptor interactions. TCRs only recognize non-self antigens when they are processed into short peptides and complexed with HLA-D on the surface of a limited range of cell lineages. This is called antigen presentation. The ability to present antigen is normally restricted to dendritic cells, macrophages, B cells, and some endothelium, but can be extended to other cells by so-called cross-presentation. The lymphocyte responds to both the antigenic peptide and the HLA to which it is bound. This is called HLA restriction.

As far as it is known, no type of trophoblast expresses the strongly polymorphic HLA-A, HLA-B (HLA class I), or HLA-D (HLA class II) antigens, the principal stimulators of T cell–dependent graft-rejection. New data are now available that suggest that this might not be entirely true in the rare inflammatory condition of chronic histiocytic intervillositis (see below). In all cases, extravillous cytotrophoblast does express HLA-C.

Where and when is Paternally Inherited Fetal Human Leukocyte Antigen Exposed to the Maternal Immune System?

The requirement for antigen exposure in order to generate immunoregulation is different for uNK cells and T cells. uNK cells do not require foreign antigen to be activated, while T cells do. Antigens are commonly peptides that interact with TCRs (on the surface of T cells), after presentation by antigen-presenting cells (in the context of HLA class II/HLA-D) or target cells (in the context of HLA class I/HLA-A, HLA-B, or HLA-C). The antigens are derived from the intracellular breakdown of proteins that can arise externally (in the context of pregnancy, this means “nonself” antigens encoded by genes of the fetus and placenta, or the father) or may be encoded internally (e.g., so-called “self” antigens encoded by the maternal host genome). TCR binding of antigen is necessary to activate naïve T cells, and promote their proliferation and ability to exert effector functions.

Immune recognition of fetal-paternal polymorphic HLA is essential to explain primipaternity, which must include immune “memory” to change subsequent responses in pregnancies by the same father. T cells generate memory that is specific for the eliciting antigen. But NK cells also have their own form of memory—sometimes called NK cell education. These two cell types are the main players to explain primipaternity, but there is as yet no conclusive evidence that they cause or protect from preeclampsia. Furthermore if priming of maternal T cells to paternally inherited alloantigens can influence placental development, an unanswered question is how can prior memory of antigen be protective? Stimulation of effector (or “cytotoxic”) T cells would be expected to have an adverse effect—not a benefit—for placentation. However, the fact that antigen priming is required to generate not just effector T cells, but also Treg cells required for immune tolerance, solves this conundrum. Maternal systemic and uterine Treg cell function are both diminished in preeclampsia as discussed in Section Effects of T cells on Placental and Fetal Tissues below.

There are two critical interfaces where maternal immune cells have the potential to directly interact with placental trophoblasts—in the decidua, where extravillous cytotrophoblasts interact with decidual immune cells, and at the surface of the placental villi, where multinuclear syncytiotrophoblast interacts with immune cells circulating in maternal blood ( Fig. 7.3 ). However, a key issue is that maternal exposure to paternally inherited fetal HLA is not restricted to trophoblast, but can also be stimulated by male partner seminal fluid, and fetal microchimerism. These potential routes of exposure to paternally inherited fetal alloantigens are detailed below.