Maternal-Fetal Immune Tolerance, Block by Block

Abstract

How difficult is to go from egg to implanted embryo? The evolution of placentation in eutherian mammals created enormous challenges, in particular to the maternal immune system, as the fetus expresses paternal antigens that are capable of triggering immune rejection. Samstein et al. reveal a role for inducible regulatory T cells in the enforcement of maternal-fetal immune tolerance.

From the perspective of the immune system, pregnancy presents a state akin to organ transplantation, in which fetal antigens are foreign to the mother’s immune system. The fetus expresses combination of antigens from the mother and the father; this combination is referred to, in immunological terms, as semiallogeneic. The immune system of eutherian mammals, therefore, had to evolve mechanisms to prevent immune-mediated rejection of the semiallogeneic fetuses. Regulatory T cells (Treg) are among the most important cells involved in the establishment of immune tolerance against self antigens and antigens encountered in foreign grafts. In this issue, Samstein et al. show that a subset of Treg cells plays an important part in maternal-fetal immune tolerance (Samstein et al., 2012).

Regulatory T cells are characterized by the expression of the transcription factor Foxp3, which is essential for their development and maintenance. Treg cells can follow a developmental path similar to other T cells and emerge from the thymus (referred to as thymic-derived Treg cells [tTreg] or frequently as natural [nTreg]) or differentiate extrathymically from naive CD4⁺ T cells; cells in this latter subset are known as peripherally induced Treg cells (pTreg, or iTreg, for induced). It has been postulated that a division of labor exists between Foxp3⁺ tTreg and pTreg, with tTreg being mainly associated with the maintenance of immune homeostasis and the prevention of autoimmunity; pTreg cells are thought to be most important in establishing tolerance at interfaces such as the mucosal surfaces in the lung and in the gut, where the immune system comes into contact with the ‘‘outside’’ and is chronically exposed to allergens and other foreign antigens. Both tTreg and pTreg subsets are needed to fully protect Foxp3-deficient mice from inflammatory disease (reviewed by Bilate and Lafaille, 2012), and now Samstein et al. make a compelling case for the predominant involvement of pTreg cells in maternal-fetal tolerance.

Three conserved noncoding DNA sequences have been described in the Foxp3 locus, namely CNS1, CNS2, and CNS3, and are shown to regulate and modulate the expression of foxp3 in response to distinct cell-intrinsic and -extrinsic stimuli. Among other sites, CNS1 has binding sites for Smad proteins, which are activated in response to TGFβ, a cytokine that is essential for the differentiation of pTreg cells. CNS1-deficient mice displayed impaired generation of pTreg cells but unaffected tTreg cell development (Zheng et al., 2010). Samstein et al. now report that interstrain breeding of CNS1-deficient mice results in increased resorption of the semiallogeneic fetuses and that this is not seen with syngeneic fetuses, the progeny of intrastrain breedings. This correlated with the impaired generation of paternal antigen-specific pTreg cells at the fetal-maternal interface. Depleting all Treg cells (tTreg and pTreg cells) did not increase resorption rates compared to pTreg deficiency alone, suggesting that tTregs were not involved in maternal-fetal tolerance. Although the experiments did not exclude all involvement of tTreg cells in the acceptance of semiallogeneic fetuses, these results highlighted the importance of pTreg cells that develop in the decidua of pregnant females in the process. These findings thus add the contribution of pTreg to the multiple redundant pathways operating in maternal-fetal tolerance. It should be noted that the absence of all Treg cells does not result in sterility either in semiallogeneic or syngeneic pregnancies.

These findings also bear importance for pregnancy complications such as preeclampsia (Santner-Nanan et al., 2009), in which defective pTreg cell generation may take place. These authors found that women with pre-eclampsia had a lower frequency of Treg cells and a higher frequency of Th17 cells, lymphocytes associated with an inflammatory response, than women undergoing normal pregnancies. Epidemiologically, pre-eclampsia has the hallmarks of an immune response. It occurs mainly in first pregnancies, when the mother has had fewer opportunities to develop tolerance against paternal antigens, and inflammation is a major part of its pathogenesis. Thus, proper development of regulatory T cells is likely to play a role in preventing this condition (Redman and Sargent, 2010).

The findings by Samstein et al. also contain important evolutionary implications. The authors’ hypothesis of a role for pTregs in maternal-fetal tolerance came from phylogenetic analysis of CNS1. By aligning the CNS1 sequences of a number of vertebrate species, including noneutherian mammals such as wallaby and opposum, they found that the presence of CNS1 correlates with the evolution of placentation; CNS1 emerges abruptly in evolution and is present only in eutherian mammals. Interestingly, the CNS1 element is contained within a retrotransposon. The findings by Samstein et al. indicate the possibility that the acquisition of the TGFβ-responsive CNS1 element and the increased development of pTreg cells enhanced maternal-fetal immune tolerance (Figure 1). Important implications of these findings are that acquisition of CNS1 via a transposable element eventually provided eutherian mammals with the advantage of peripheral Treg differentiation and that the dominant selective pressure for the existence of pTreg cells was to improve acceptance by the maternal immune system of fetuses presenting paternal antigens foreign to the mother. Other functions of pTreg cells, such as their role in maintaining immune tolerance at mucosal interfaces such as the lung and the gut, may have come as a later adaptation. One of the many interesting future directions will be to investigate the differences in pTreg generation and function between eutherian and noneutherian vertebrates, which have the Foxp3 gene but lack a recognizable CNS1 element.

Figure 1. CNS1 Is Present in Placental Mammals and Confers Foxp3+ Peripheral Treg Differentiation Capacity to T Cells.

(A) The thymus of mammals exports naive T cells (Foxp3⁻) and thymus-generated Treg cells (tTreg).

(B) The CNS1 element in the Foxp3 gene is contained within a retrotransposon that is present in all eutherian mammals, but not in other vertebrates. CNS1 contains, among other binding sites, a Smad-binding site, allowing pTreg Foxp3⁺ differentiation from peripheral Foxp3⁻ T cells in response to TGFβ. Due to its highly conserved sequence and role in maternal-fetal tolerance, CNS1 acquisition is likely to have played a role during the evolution of placental mammals.

(C). Noneutherian mammals (and other vertebrates) may have less prominent peripheral Treg cells (pTreg) and, consequently, a higher proportion of thymic-derived tTreg cells.

Evolution by mobile genetic elements allows sudden changes in gene expression or changes in the expression of blocks of genes; both the quickness and the block changes are suitable for major adaptation. From egg to implanted embryo, eutherian mammals accumulated many changes in their physiology to allow the development of a fetus in the uterus. For example, the body of the mother had to become aware of pregnancy and, accordingly, secrete multiple hormones that control the large transformation of the uterus architecture necessary for a successful pregnancy (Wildman, 2011). How did these changes occur during evolution? There is evidence that many of these changes came in blocks. For example, analysis of gene expression in the uterine endometrium showed that more than 1,500 genes were recruited into endometrial expression and that about 13% of them were regulated by a single eutherian-specific transposable element, Mer20 (Lynch et al., 2011). Other genes important to placentation, such as the Syncytins, are also encoded by endogenous retroviral sequences (Nahmias et al., 2011). The findings by Samstein et al. now add a new layer to the importance of transposable elements in the evolution of eutherian mammals, involving in this case the adaptation of the maternal immune system to pregnancy.

The finding that the acquisition of CNS1 was likely via a transposon insertion into the Foxp3 locus adds to other transposable events that have influenced the immune system. An ancestral transposon insertion might have brought recombination activating genes (RAG) 1 and 2 into the genomes of jawed vertebrates; this insertion was advantageous once RAG1 and RAG2 became an integral part of the diversification of antigen receptors (Hirano et al., 2011). The generation of highly diverse antigen receptors required the coevolution of mechanisms to control immune responses targeting the individual’s own tissues; in this light, it is interesting to consider that the gene encoding Foxp3 appeared early in vertebrate evolution and is present in all jawed vertebrates, with the possible exception of birds. Interestingly, the coding sequence of Foxp3 became highly conserved in eutherian mammals (Andersen et al., 2012). Thus, several transposable elements acted as building blocks in the evolution of eutherian mammals, Mer20, RAG1/2, and CNS1, each one contributing to the evolutionary fitness of placental animals.

The findings by Samstein et al. provide unique insight into the evolution of mammals. The study raises questions about the TGFβ dependence of pTreg in noneutherian mammals and other Foxp3-expressing vertebrates. Finally, it raises the possibility that defects in the generation of pTreg may be at the root of conditions such as pre-eclampsia.