A framework for understanding maternal immunity

Introduction:

Maternal immunity from an evolutionary perspective.

Classical immunology suggests that the maternal immune system must be suppressed to support fetal development. Without this modification, there would be no generational passage of genetic material. The counter argument is that the critical shaper of the immune system through evolution is infectious agents¹, not reproduction. Moreover, constraint of the immune system in the face of pathogens is counterproductive from an evolutionary standpoint. Interactions between evolution, reproduction, and environmental stresses, including infection continues to be debated² and the understanding of these may be critical to alleviating disease of pregnancy.

The adaptive immune system arose 450 million years ago with the emergence of jawed fish¹. Expression of molecules recognizing variable protein epitopes and genetic variation in domains of the molecules themselves allowed for increased probing local and internal³ environments¹. This enhanced the response to pathogens⁴, but also raised the problem of differentiation between epitopes indicating a harmful entity from those expressed normally. The genetic locus encoding these proteins likely began as a single gene that was duplicated, deleted, or mutated to give rise to a complex locus⁵.

Over evolutionary time, several mechanisms have evolved to support developing offspring^6,7. Differing complexity and nutritional capacity in egg contents in oviparous species⁶ and different mechanisms supportive of developing offspring in viviparous species⁶ have evolved. This comprises a continuum which has undergone evolutionary revision over time. An example is the placenta. This structure, intended to provide nutritional and hormonal support in addition to oxygenation and waste removal, has appeared and disappeared repeatedly in evolutionary history. Mammalian species exist along a continuum of oviparity, viviparity, and placentation⁸.

Is there a link between evolution of the immune system and that of the modes of support of the developing offspring? Placentation has developed in species without a formal adaptive immune system⁷, and the complexity (e.g., number of genes) within the locus encoding genes for the T cell or B cell receptor do not appear to correlate with the mode of placentation within mammalian species. One interesting locus, however, is that encoding the receptor expressed by gamma-delta T cells⁵. This locus underwent significant changes between the late stages of avian species evolution and the early development of mammals. At some point the gamma delta locus was significantly altered^5,9. Some mammals¹⁰ have comparatively greater numbers and complexity of these cells, especially during pregnancy, than do species with a different placentation^11,12. This allows speculation that innate immunity is more evolutionarily linked to mode of reproduction than is adaptive immunity.

In mice¹³, humans¹⁴, and other species¹⁵ gestational innate immunity is regulated, particularly within the peri-implantation uterus¹⁵. After implantation systemic innate immunity is heightened¹⁴, while local innate immunity is modulated¹⁶, depending on the health of the fetal-placental unit^17,18. Here again is a controversy. One interpretation of existing data is that modification of innate immunity is critical to the health of the developing fetal-placental unit ¹⁹. However, another interpretation is that it is the health of the fetal-placental unit which results in productive collaboration with the innate immune system to generate supportive mechanisms, such angiogenesis^17,20 or expression of needed metabolites or hormones²¹. This may be how reproductive evolution links to immune evolution: via innate host defense and protective mechanisms.

Background

The anatomy of the maternal fetal interface.

There are numerous interactions between mother and fetus (Fig. 1). Mice and humans provide examples. Fetal cells²², cellular fragments (e.g., extracellular vesicles²³) or cell free RNA²⁴ or DNA²⁵ circulate in the maternal blood²², and can implant in lung²⁶, spleen²², heart ²⁷, liver²⁸ or other tissues. Fetal elements may protect or help regenerate ²⁹maternal organs facing stress or infection³⁰. Cargo carried by extracellular vesicles may modify biological processes, like immunity³¹. Fetal proteins may be acquired by maternal antigen presenting cells in uterine lymph nodes or spleen and presented to maternal T cells³².

Figure 1.

Fetal maternal interactions via placenta

Extra villous trophoblast (fetal) directly contacts the decidua basalis (maternal) to provide critical structure to the placenta in these species³³. The decidua is generated in the peri-implantation period and contains evolving populations of maternal immune cells. In this tissue, trophoblast infiltrates the endothelium of maternal vessels which feed the developing conceptus and modifies blood flow to the intervillous space. Such flow allows for waste removal and exchange of gas, nutrients, and metabolites to the villous stroma and in turn to the fetal vessels. In the intervillous space, maternal blood percolates through and directly contacts fetal trophoblast (syncytial). In case of cellular damage or specific signals, maternal cells breech the syncytium of the placenta and make their way into fetal stroma and blood vessels. From there, they infiltrate the thymus and other fetal tissues. During development, fetal membrane (chorion) is also in direct contact with maternal decidua (parietalis). Chorion may fuse with this tissue to stabilize the fetal membranes³³. In addition, fetal cells may leave the amnion and chorion and traffic to the maternal decidua^34,35 subsequently arriving in maternal uterine vasculature or lymphatics. Similarly, maternal cells may breech the fetal membranes and, via amnion or amniotic fluid, and reach the fetal circulation. Maternal immune cells or cellular components may then interact with fetal tissue and modify developing immunity³⁶. To an incompletely understood extent, immune modulatory or reactive proteins, including antibodies, cytokines, and related proteins may breech either trophoblast syncytium or fetal membranes or generate signaling effects on extra-villous trophoblast. Thus, far from being an inert, impassable barrier, the maternal-fetal interface is a complex structure that can allow for regulated maternal-fetal trafficking of cells and cellular components. These may have long term effects both on the pregnancy and on the life of either mother or infant³⁷. Thus, a pregnant female is a cellular and molecular chimera of her mother and her offspring. Since cells from semen have been found in the abdominal cavity, partner’s cells, may contribute to maternal chimerism³⁸.

The nature of fetal and placental antigen and maternal autoantigen

The mother is initially exposed to seminal antigen via intercourse. This exposure includes sperm, cellular fragments, and associated immune cells in seminal fluid. Minor antigens in seminal fluid, particularly the male antigen H-Y, are recognized by the female’s immune system and this recognition often leads to the expansion of H-Y specific T cells, including those with a regulatory phenotype³⁹. The reason for this is not completely clear. It may be related to the level of antigen in the vaginal environment⁴⁰, the presence of costimulatory or coinhibitory molecules in vaginal stroma⁴¹ or other factors in subepithelial lymphoid tissue aggregates ⁴² uterus⁴³ uterine draining lymph nodes ³⁴ or peritoneal cavity³⁸.

The fetal-placental unit expresses unique antigens encoded by the paternal genotype. These include MHC molecules inherited only from the father⁴⁴ but also minor antigens (peptides) which are presented by common (fetus and mother) or unrelated MHC. The fetus may also express several proteins which are allelic variants of the protein expressed in the mother. Peptides of these variants can be presented by shared MHC or by unrelated MHC.

Fetus and mother may share significant expression of MHC-plus-peptide elements in their cells. Moreover, antigens uniquely expressed during development and in the placenta may have been “seen” by the mother during her own development and thus would not really be “non-self”, but “former self”⁴⁵. Whether the low frequency exposure to such antigens between the time of the mother’s own development and her index pregnancy makes her recognition of these antigens as “novel” is unclear. Moreover, it is less clear whether “novel” antigens are generated by trophoblast or placental dysregulation as occurs for tumors (for example, ⁴⁶).

It is possible that the high level of shared or even unique paternally encoded antigen might serve as structural “help” for T cell activation against these novel antigens as has been proposed in the context of infection. In contrast, “altered” antigens made by specific modification of local proteins might also serve to downmodulate the outcome of fetal antigen recognition⁴⁷.

Could the prolonged, slowly changing and ultimately high level of local antigen exposure essentially make fetal antigen a “stand-in” for “self” for T cells at the maternal-fetal interface? The implications are interesting. In this case, maternal “tolerance” of the fetus is the same as tolerance for self, potentially through negative feedback (e.g., “tuning” ^48,49) related to signaling through the T cell receptor (for example⁵⁰) leading to a change of state in local T cells.

In contrast, since peptide agonists for T cells are more related to “self” ⁵¹ and in the context of previous studies of T cell recognition of peptides, there exists the possibility of structural facilitation by endogenous (“self”) peptide for T cell activation. Thus, maternal autoantigen could drive responsiveness against fetal antigen and fetal antigen could assist in responsiveness to newly expressed maternal autoantigen within local (e.g., uterus) tissues. Thus, both fetal and maternal peptide recognition could drive both anti fetal and anti-maternal “self” reactivity or tolerance, depending on the T cell receptor and the downstream associated signal pathways linked to a specific T cell.

Expression of MHC is also regulated⁵², and this may support early collaboration between cells such as NK or NKT cells and trophoblast and regulate T cell attack of fetal tissue⁵³. Cells within the placenta or in vessels feeding the placenta however can be induced to express polymorphic MHC during infection⁵⁴ or stress⁵⁵ , and this may be critical for immune protection of the whole tissue.

Fetal cells capable of presenting unique fetal antigens to maternal T cells are less abundant in the maternal circulation than are maternal cells capable of processing and presenting fetal antigens to maternal T cells. The relative inefficiency of this “indirect” mode of presentation may protect against harmful anti-fetal responses. Regulated trafficking of uterine maternal dendritic cells to uterine draining lymph nodes⁵⁶ and of maternal T cells into the decidua⁵⁷, a site of direct interaction between maternal cells and trophoblast is also thought to be protective.

The State of The Maternal Immune System

Maternal Immune cell homeostasis

Studies in immune deficient mice and people identified unique phenotypes associated with immune cell homeostasis. Homeostatic proliferation and death (turnover) in the T cell pool is driven by the presence of “space”⁵⁸ (an increase in the functional niche in which immune cells exist⁵⁹), certain cytokines such as IL-7 ⁶⁰and IL-15⁶¹, and “low-level” signaling by molecules such as MHC complexed with “self-antigen” ⁶². Regulators of homeostatic turnover include the expression of molecules such as FASL or FAS⁶³. Homeostatic proliferation supports expansion of T cells that have been positively selected in the thymus⁶³, maintenance of the naïve T cell pool before activation in the periphery, and also maintenance of the memory T cell pool, particularly after viral infection⁶⁴. Disruption of this process is postulated to underlie autoimmune disease, since the driver of homeostatic proliferation is thought to be “self” antigen. The phenotype of these cells has been characterized and includes elevated expression of molecules such as the IL-7 or IL-15 receptors, high expression of the activation markers CD44 and PD1 and increased expression of functional molecules such as granzyme B⁶⁵. Downregulation of CD8 has also been observed—thus leading to high numbers of double (e.g., CD4/CD8) negative T cells.

The maternal immune system undergoes significant homeostatic change ^66–68. Pregnancy hormone-driven changes in blood volume, cardiac output, and decreased systemic vascular resistance⁶⁹ effectively increases the “space” available to the immune system and is reflected in increased size of lymphoid organs such as the spleen and lymph nodes⁷⁰. In contrast, the thymus undergoes a decrease in size⁷¹. This may be due to decreased developing thymocytes or to increased thymic output of fully developed but naïve cells ⁷². Significant changes in the thymus during pregnancy allow speculation that antigen expressed by thymic epithelium also changes and thus may alter the threshold for selection. If this changes the repertoire of naive maternal T cells, this could theoretically alter the capacity for maternal anti-fetal immunity.

Homeostatic turnover in maternal immune cells, not just T cells, has been observed during mouse pregnancy^67,68, with increased proliferation ⁶⁸ and also increased expression of both pro and anti-apoptotic molecules in immune cells⁶⁷. In addition, expression of the receptors for IL-7 and IL-15 are regulated⁶⁷, and increased expression of activation markers such as PD1 and CD44 have been observed⁷³. Furthermore, tissues such as the placenta show regulated expression of IL-7 and IL-15. The antigenic driver of homeostatic proliferation during pregnancy likely not only include “self-antigen” but also fetal antigen. In the decidua, the relatively high level of fetal antigen may further drive this process.

One of the outcomes of a high level of homeostatic proliferation is a state that shares elements of what is called “exhaustion”⁶⁵. Also attributed to chronic antigen activation, this state is described as having a mixed signature that includes inflammatory proteins and decreased ability to generate certain responses⁶⁵. If exhausted cells are not removed by interactions such as FAS/FAS-ligand, they will persist until an alteration in their tissue milieu which could then cause them to regain their functional capacity⁷⁴. At the maternal fetal interface, the presence of cells with an exhausted phenotype have been observed^67,75,76. Other mechanisms to relieve this state are currently under investigation in several disease states⁷⁷. One manipulation relevant to this state is the removal of the cytokine IL-10⁷⁸. Moreover, as trophoblast expresses IL-10, this could enhance the exhausted state in local cells, particularly those specific for fetal antigen⁷⁹. Viral infection at the maternal-fetal interface modifies IL-10 expression^78,80, and this may be relevant to the T cell immunity at that site.

The unique phenotype of maternal immune cells.

Pregnancy is associated with functioning innate and adaptive immunity.

Maternal B cells make antibody to proteins to which the mother is exposed, including to fetal proteins such as MHC⁸¹ and blood group antigens⁸² This suggests that, at least in lymphoid tissue, B-cells can respond to T dependent and T independent antigen, and that CD4 T cells and possibly CD8 T cells can assist in B cell responses⁸³. Pregnant mice and humans can generate cytotoxic T cells against fetal, major, and minor antigens such as H-Y^22,84. Further, the generation of effective memory to infectious agents is similar to that of nonpregnant animals⁸⁵. How does one explain data suggesting a class shift in the systemic presence of cytokines (e.g., Th2) over Th1? The relative increase in regulatory T cells over Th17 cells? The answers may not be that pregnancy is an inherent state of T cell deviation (e.g., ⁸⁶). The rules for development of T cell responses, individual setpoints and the specific the role of so-called self-peptide may be the answer. For example, initial antigen load can lead to differences in final responsiveness⁸⁷. The lowest antigen load leads to a delayed hypersensitivity (DTH) response (or a cytotoxic T cell response) while a higher load leads to a limited DTH and a prolonged antibody response. At still higher loads, the predominant response is an antibody response. Some theories in this realm rest on the Th1 versus Th2 dichotomy, whereby low dose leads to Th1 response and high dose leads to Th2 responses^88,89. The setpoint for the switch between one and the other is hypothesized to be related to genetics. This makes the tendency for one type versus the other individually or genetically regulated⁹⁰. The role of pregnancy then would be in timing and the antigen load³². Remembering that the type of antigen “seen’ by the immune system is multifaceted (“self”, ‘former- self’ ‘shared-self’ and unique”) the relative presence of these antigens might lead to differences in the observed state of the immune system.

In addition, presentation of “endogenous” or “self” or “high-frequency” antigen may either assist in getting an immune response out of a low frequency antigen or a T cell whose receptor generates a low strength signal. Regulatory T cells are driven by moderate strength interactions and recognition of self-peptide. In this regard, it is interesting that regulatory T cells expand, even in syngeneic pregnancies⁹¹, and can be driven by auto antigen both in the generation of “natural T regs in the thymus⁴⁵, and in induced T regs in the periphery of pregnant animals⁹². Thus “fetal antigen” can act as self, and self-antigen can drive both Th2 and regulatory T cell responses⁹³. The difference is the T cell repertoire, specificity and pool size going into pregnancy. This may explain cross-sectional studies that have observed a shift during pregnancy.

Another facilitator of this shift could be the volume expansion-driven homeostatic proliferation experienced by immune cells during pregnancy. It would be difficult to suggest however that these changes are part of an inherent drive to immune cell deviation. They could however fit in the context of a fallout of the normal vascular biology of pregnancy⁶⁹. In this case the inherent cause of abnormalities in the systemic T cells pool is not “loss of tolerance” or loss of inherent deviation, but inherent deficiency in the mechanisms which should lead to normal hormone-mediated endothelial cell function, decreased vascular resistance and increased volume expansion.

Both innate and adaptive immune cells in the decidua have a unique or “mixed” phenotype. For example, there are CD8 T cells with high expression of activation and memory markers, the capacity-in terms of protein expression-to potentially kill trophoblast, yet decreased expression of critical cytokine markers of cytotoxic T cells⁹⁴. They are seen in both normal and abnormal pregnancy^95,96 and have the potential to kill virally infected cells⁹⁴. For these cells fetal or maternal antigen specificity may be important. For example, it has been observed that high specificity for fetal antigen leads to an “exhausted state”, yet, as in other states of “exhaustion” these cells can be modified and become quite effective antiviral killers⁷⁸. For another, this site is where one finds a mixed population of regulatory T cells- some which are apparently developmentally determined in this site ⁹⁷ while others are apparently induced at the maternal-fetal interface. Interestingly, such regulatory T cells are quite reversible in their phenotype, as infection with agents such as listeria cause them to “switch” to TH-17 cells and this may be why infections may lead to decreased Treg and fetal loss⁹⁸.

Innate cells, including macrophages⁹⁹, NK, and other innate lymphocytes (ILC) cells at this site have a unique phenotype. NK cells here comprise both tissue resident and trafficking (from the maternal blood) conventional cells^100,101. These NK cells have a mixed phenotype as well. Rather than kill trophoblast, which might be expected due to the expression of paternal MHC or the relative absence of self MHC, they collaborate with trophoblast and other cells to promote the development of the blood supply to the placenta¹⁰².

Gamma- delta cells are expanded during pregnancy¹¹ and found in both placental villi ¹² and in decidua. Interestingly, their strong Th-17 phenotype is not thought to be harmful. Instead, this developmentally regulated population¹⁰³ may be critical for supporting angiogenesis and further development of placental blood supply.

In mouse⁶⁸ and in human pregnancy¹⁰⁴, B cells undergo homeostatic turnover likely driven by similar mechanisms at T cells, and the mother mounts antibody responses to fetal (e.g., MHC) and environmental antigens (allergens, vaccine). The idea that the generation of antibodies, rather than cytotoxic T cells, as the “default” state during pregnancy may have limited the search for unique B cell phenotypes. However, potentially harmful autoantibodies can be generated and for example and can lead to anticardiolipin mediated placental damage and trophoblast loss. General examination of B cell types has revealed the existence of long-lived, self-replenishing B-cells, termed B1 cells. They produce repertoire-conserved, “natural” IgM antibodies, arise during early development and typically reside in the peritoneal and pleural cavities¹⁰⁵, but can traffic to perivascular tissue¹⁰⁶ the spleen¹⁰⁵ and other tissues. Cytokines, such as IL-9¹⁰⁷ IL-10¹⁰⁸ IL-5^109,110 contribute to B1 cell development and/or function. They may play a role in recognition and removal of apoptotic¹¹¹ or senescent cells, cell debris, and other “self-antigens” generated by cellular homeostasis. With this capacity, and their production of IL-10¹¹², they can help the immune system delineate non dangerous signals and processes and prevent autoimmunity¹¹³. Pregnancy is a time of remodeling, cell turnover, and changes in metabolism in both the fetal-placental unit and in the mother, and fetal cellular components in maternal blood^22,114,24, ²³ may drive B1 development. B1 with regulatory function expand in human pregnancy¹¹⁵. However, B1 cells are increased in recurrent pregnancy loss¹¹⁶. This may represent a compensatory response to cellular dysfunction of the fetal-placental unit. Other B cells with regulatory potential, called B-reg or B10 cells have a different surface marker phenotype^117,118, inclusive of molecules such as TIM-1, and can moderate transplant rejection¹¹⁸. B10 are genetically and anatomically heterogeneous. They express IL-10, TGF-β and IL-35. In mice, B10 expand in normal but not abnormal pregnancy¹¹⁹, and B10 manipulation modifies pregnancy outcome¹²⁰. In humans B10 are not modified by labor¹²¹, but are increased in the postpartum¹²². Deficiency in IL-10- expressing B cells is associated with increased susceptibility to LPS-induced preterm birth in mice¹²³. Moreover, recent data suggests that local B cells may be regulated via expression of molecules such as trophoblast glycan ligands for B cell CD22-LYN ¹²⁴. However, non-modified circulating antigens, such as may be generated by placental damage, may bind to the B cell receptor and not engage this regulatory signal construct⁴⁷.

There is a theme here. The uniqueness of the phenotype of these immune cells at the maternal-fetal interface is not tied to tolerance of fetal antigens. It is tied to development and functioning of the tissue. These unique cells may play a role in the postpartum return of maternal physiology to its pre-pregnancy state ^125,126.

Metabolism and the maternal immune system

Energy balance and metabolism may influence contextual elements of the immune response in these tissues and may support fetal tolerance. For example, if there is adequate energy and support for metabolism of the fetus, there are not the signals, including increased expression of co-stimulation, decreased co-inhibition, or altered expression of cytokines which would increase effector function or T cell survival that would support local anti-fetal immunity. Global maternal energy balance may also dictate the capacity of regulatory T cell responses. This might regulate maternal tolerance, as, for example, lack of the presence of such cells is associated with poor pregnancy outcome¹²⁷.

Recent focus on the metabolism of immune cells includes examination of the links between such elements as glucose and fatty acid metabolism, mitochondrial function and immune activation, regulation, and homeostasis. For example T cells activation generates an upregulation of the glycolytic machinery¹²⁸, glutaminolysis¹²⁹ uptake of large neutral amino acids, such as tryptophan and arginine¹³⁰. Th1 or TH17 predominance involves increase in the activities of the TCA cycle, mitochondrial fatty acid oxidation, and oxidative phosphorylation¹³¹. Regulatory T cells have increased uptake in exogenous fatty acids and downregulated glycolytic machinery¹²⁸. Memory CD8 cells engage in a cycle of fatty acid synthesis, lipolysis, and oxidation to drive oxidative phosphorylation¹³². Other immune cells show different linkage between metabolism, activation, and homeostasis. For example, dendritic cells ¹³³ and decidual NK cells ¹³⁴are particularly regulated by glucose metabolism.

Early data suggested the potential role played by the metabolism in tumors or tissues of the maternal-fetal interface context in the regulation of immunity. Disruption of tryptophan uptake from the tissue environment results in activation of immune cells and decreased pregnancy success¹³⁵. This was considered to reveal an inherent mechanism of maternal tolerance that extended to other tissues¹³⁶. However, the specificity of the critical immune response is still in question, since metabolic disruption has a broad effect on tissue integrity and this in turn may drive immune activation¹³⁷. Other studies of tissues from normal and abnormal pregnancies have observed altered metabolism in diseases¹³⁸ where presumed loss of immune tolerance has also occurred ¹³⁹. The critical remaining question is the timing of altered tissue metabolism with respect to immune activation¹⁴⁰. The links between immunometabolism pregnancy and maternal tolerance are just beginning¹⁴¹. Studies of administration of N acetyl cysteine ^142,143, an inhibitor of the reactive oxygen species generated in cytotoxic T cells, and of dietary short chain fatty acids to enhance regulatory T cell function¹⁴⁴ are also under investigation in studies of reproductive outcome¹⁴⁵. Metformin, used to treat diabetes, has multiple activities, including decreasing inflammation¹⁴⁶. It is studied for the treatment of other pregnancy related than diabetes¹⁴⁷.

Discussion

Theories of maternal tolerance

The definition of immune tolerance is a complex topic. It can be the absence of an expected response, e.g., cytotoxicity, generation of specific antibody, or a functional outcome, e.g., graft non-rejection. It can be the absence of any response, or it can be the alteration or downmodulation of a response. It can be an expected positive response that is decreased in magnitude or lifespan. The list of what could comprise tolerance is long. The list of underlying potential mechanisms is even longer. For T cells this includes (discussed elsewhere) everything from “holes” in the T cell repertoire or negative selection in the thymus. It includes lack of contact with its cognate antigen, lack of responsiveness to its cognate antigen, and responsiveness that is truncated, altered, or hyper-stimulatory. Tolerance in T cells may lead to a state that is paralyzed, exhausted⁷⁷, inactive, or dead.

At the heart of immune theories is the very nature of the decision between activation and tolerance. What is the driving force for that decision? One signal or two- the debate continues⁵⁰. The first signal is through the T cell receptor which gives specificity to the response while the second is provided by the environment or context in which the T cell resides. In a one signal model the encounter with cognate antigen sets the stage for all subsequent states of all molecular pathways. In a two-signal model, “context” also provides critical licensing, feedback, regulation, and support⁵⁰. The struggle for the precise delineation of these features continues. Most immunologists assert that a critical piece of the decision between activation and tolerance is the delineation of “self” from “non-self”. The difficulties of this terminology have been discussed previously¹⁴⁸ and the problem with even the concept of “infectious non-self from non-infectious self¹⁴⁹ is highlighted by recent recognition of the importance of the microbiome in both homeostasis and disease^150,151. “Self” even as “expressed during development” is problematic¹⁴⁸ since for examples animals undergo significant change with developmental time (puberty, pregnancy) and neonates can mount good immune responses depending on the formulation of the antigen given^152,153.

A single amino acid substitution in an otherwise antigenic peptide can lead to different outcome in T cell development¹⁵⁴. So, the question remains: what is the molecular basis of detection of self, versus non-self? Is it related to signaling through the T cell receptor? or, is it related to the context in which that signal occurs? If it is related to T cell receptor signaling, is it the strength of signal? the timing of signal? the sum of signal strength over time? the pattern of signaling (e.g., mixture of short versus long interactions)? or the ability for the signal to reach a developmentally determined threshold? If the context is important, then what does it provide that leads to both the initial decision and the subsequent response?

Immune system activation versus tolerance may not critically depend on a “self-non-self” decision. A holistic theory¹⁵⁵ suggests that the decision for activation versus tolerance is a summation of all the signals seen by an individual or population of T cells and the subsequent state of the T cell in response to these signals. Thus, maternal “tolerance” of the fetus is the sum of signals received by all maternal T cells present in the mother during pregnancy and is driven by the presence and strength of signal given by both “self” and fetal antigen, as well as the “context” which is basically the physiological health of mother, fetus, and placenta.

Pregnancy does not fit easily into models of immune activation and tolerance. It is thus often marginalized as a “special” case or ignored altogether. Two questions remain. First, how did “self, versus non-self” become central to thinking in immunology, and how did pregnancy become ignored or marginalized? Both might be rooted in the philosophical and historical nature of science. “Self-versus -other” is also a philosophical question with non-euro-centric philosophies have a differing view¹⁵⁶. Modern immunology’s, focus on self-non-self-determination may be the result of a particular intellectual and philosophical framework that itself should be open to query¹⁵⁷. Similarly, the evolutionary theory maternal-fetal conflict¹⁵⁸ may have unduly driven the intellectual framework around maternal tolerance of the fetus¹⁵⁹. misalignment between evolutionary conflict and self-non-self-theories and the actual observations of pregnancy may drive the marginalization observed.

Summary

Alternative views of the maternal-fetal relationship^160,161 might reframe the immunological “problem” of mother(“self”)/fetus (“non-self”). The experimentally observed immunological relationship between mother and fetus should be used to interrogate several themes relevant to immunity. Those who see things slightly differently may be able to help in this process.

Key points:

• Evolution may contradict the assertion of maternal immune suppression

• The maternal-fetal “interface” comprises complex and multilayered interactions between fetal and maternal cells

• Fetal/placental and maternal antigen interact in the maternal immune response

• The maternal immune system undergoes homeostatic changes that drive unique phenotypes; metabolism may contribute to this

• Pregnancy may not fit into classic immune theory which itself may be rooted in a cultural framework that should be interrogated

Synopsis:

This is an alternative and controversial framing of the data relevant to maternal immunity. It argues for a departure from classical theory to view, interrogate and interpret existing data.

Clinical care points:

• Pregnancy is not a state of global suppression or deviation, and multiple factors drive the immune response of the individual mother. Do not paint every mother with the same brush.

• Give recommended vaccines and encourage participation in clinical trials relevant to the immune system.

• Examine each woman with a pregnancy complication with a broad perspective and be sure to examine her not only during pregnancy, but in the postpartum period as well.