Maternal microchimerism in infancy

Despite the dogma of the placental barrier, during pregnancy, maternal cells pass into the fetal circulation, where they can engraft in large numbers in infants with severe combined immunodeficiency (SCID), unperturbed by a defensive host immune system . Maternal microchimerism (MMc) was detected in 40% of infants with SCID and graft-versus-host disease (GVHD) developed in the majority of children , correlating with maternal T and B lymphocytes . Unlike stem cell transplantation, when donor cells functionally replace the host immune system, maternal cells do not function in an immunodeficient child. The functional capacity of chimeric maternal cells is not known. In vitro, chimeric maternal cells respond poorly to alloantigen . One reason for the limited ability of maternal T lymphocytes to respond to specific antigens may be the limited T cell receptor repertoire in engrafted maternal cells . Chimeric maternal T cells do respond to nonspecific mitogens, such as interleukin (IL)-2, anti-CD3 antibodies, or phytohemagglutinin (PHA), and maternal cell lines have been grown from the blood of immunodeficient patients, showing proliferative potential. These maternal cell lines were able to proliferate at normal rates and express maternal human leukocyte antigen (HLA) molecules . Although T lymphocyte activation markers have been identified on chimeric maternal lymphocytes, and MMc has been associated with GVHD, 40% of patients with detectable MMc never developed GVHD, suggesting that regulatory mechanisms can control engrafted maternal lymphocytes. Thus, a limited number of maternal T lymphocyte clones may be transported into the fetus and expand in response to nonspecific stimuli. However, a random selection of maternal cells may travel to the fetus, and specific clones expand by an antigenic stimulus yet to be identified, leading to clonal overrepresentation in the total population. The rate of persistent MMc in patients with SCID is not known, because of the high mortality, and most patients now receive stem cell transplants.

Whereas maternal cells make up the majority of lymphocytes in SCID, MMc is found at lower levels in immunocompetent infants. Maternal cells have been estimated at a rate of 0.02–5% in cord blood . Early studies found MMc in cord blood by labeling maternal blood cells with a fluorescent dye or chromium-51 and injecting them back into the mother hours before delivery In noninvasive studies using fluorescence in situ hybridization (FISH) with probes to the X- and Y-chromosomes, maternal cells with two X-chromosomes were found in 20% of cord blood samples from male infants and in most tissues . MMc was found in both the CD8 ⁺ and CD34 ⁺ leukocyte subsets, as well as hepatocytes, renal tubular cells, and cardiac myocytes, suggesting the transfer of maternal stem cells. In studies, using more sensitive polymerase chain reaction (PCR)-based assays, maternal DNA was found in 24–100% of cord blood samples . Maternal DNA has been reported in the fetal circulation as early as 13 weeks of gestation in blood samples taken prior to elective terminations . By 20–33 weeks of gestation, 53% of fetal blood samples harbored MMc . Lo et al. developed quantitative techniques to study bidirectional traffic, and they found that maternal-to-fetal cell transfer was more common, although less frequent, than from fetus to mother, and at lower levels . MMc has been identified in fetal lymph nodes, where it may play a role in inducing early tolerance to maternal antigens .

MMc in older individuals

Maternal cells derived during gestation are not necessarily eliminated by the child’s immune system early in life, but they can persist into adult life. The original study suggesting that MMc can persist long after birth used PCR for non-inherited non-shared maternal HLA alleles . In men, FISH for X- and Y-chromosomes was also used to identify female cells (presumed to be maternal). These two methods detected MMc in 55% of subjects, as young as 9 years and as old as 49 years. Since then, others have confirmed that a low level of maternal cells can create a state of MMc in the child that persists for decades . The levels of MMc in these studies were only roughly estimated until Lo and Lambert et al. developed a panel of real-time quantitative PCR (Q-PCR) assays specific for highly polymorphic HLA alleles that could be used to accurately measure a low level of maternal DNA by targeting maternal HLA alleles not shared by the child . By assaying genomic DNA isolated from peripheral blood mononuclear cells, evidence for MMc was found in 22% of healthy females aged 13–62 years. The levels of MMc ranged from 0 to 55 genome equivalent (gEq) per million host gEq. Thus, MMc appears to be common in healthy children and adults. The levels of MMc in specific leukocyte subsets were higher than the levels in total peripheral blood cells, rising to as high as 360 per million in T lymphocytes, B lymphocytes, monocytes, and natural killer (NK) cells . Most subjects with MMc detected in at least one cell subset had no MMc in unfractionated peripheral blood mononuclear cells. The presence of MMc in every hematopoietic cell subset suggests that a maternal stem cell may engraft into the fetus and renew multiple cell lineages throughout the child’s life.

MMc in older individuals

Maternal cells derived during gestation are not necessarily eliminated by the child’s immune system early in life, but they can persist into adult life. The original study suggesting that MMc can persist long after birth used PCR for non-inherited non-shared maternal HLA alleles . In men, FISH for X- and Y-chromosomes was also used to identify female cells (presumed to be maternal). These two methods detected MMc in 55% of subjects, as young as 9 years and as old as 49 years. Since then, others have confirmed that a low level of maternal cells can create a state of MMc in the child that persists for decades . The levels of MMc in these studies were only roughly estimated until Lo and Lambert et al. developed a panel of real-time quantitative PCR (Q-PCR) assays specific for highly polymorphic HLA alleles that could be used to accurately measure a low level of maternal DNA by targeting maternal HLA alleles not shared by the child . By assaying genomic DNA isolated from peripheral blood mononuclear cells, evidence for MMc was found in 22% of healthy females aged 13–62 years. The levels of MMc ranged from 0 to 55 genome equivalent (gEq) per million host gEq. Thus, MMc appears to be common in healthy children and adults. The levels of MMc in specific leukocyte subsets were higher than the levels in total peripheral blood cells, rising to as high as 360 per million in T lymphocytes, B lymphocytes, monocytes, and natural killer (NK) cells . Most subjects with MMc detected in at least one cell subset had no MMc in unfractionated peripheral blood mononuclear cells. The presence of MMc in every hematopoietic cell subset suggests that a maternal stem cell may engraft into the fetus and renew multiple cell lineages throughout the child’s life.

Functional MMc

The biological purpose of chimeric maternal cells is not known, but some clues can be derived from characterizing the phenotypes within tissues. Maternal cells can engraft into a child’s tissues . We detected maternal cells in the thymus, heart, liver, kidney, lung, and pancreas using the FISH assay for X- and Y-chromosomes to identify female cells in tissues from males with inflammatory and noninflammatory diseases . MMc constituted 0.1–0.9% of parenchymal cells. A technique was developed by which multiple phenotypic markers could be identified concurrently with FISH in the same cells of a tissue section to simultaneously identify and characterize the maternal cells. As circulating stem cells could have multilineage plasticity , we asked whether cells in the mother could differentiate into tissue-specific phenotypes in her progeny. Female (maternal) cells within male tissues were characterized by simultaneous immunohistochemistry and FISH for X- and Y-chromosomes. Maternal cells expressed sarcomeric α-actin in the hearts of infants with neonatal lupus syndrome, indicating they had differentiated into cardiac myocytes or possibly fused with host cells . Srivatsa et al. identified female cells (presumed maternal) in the tissues of male newborns with congenital anomalies, in the liver, thymus, thyroid, and skin, but not in the spleen . It is not known whether maternal cells function alongside host cells or when allogeneic antigens on maternal cells may be recognized and attacked.

Immune regulation of MMc in a child

How are levels of MMc regulated? The field of transplantation has shown that foreign cells are filtered into the lungs, and then allogeneic cells are eliminated via NK cells and cytolytic T cells. The balance of pro-inflammatory cells to anti-inflammatory regulatory T cells expressing transforming growth factor beta (TGF-β), IL-10, and programmed death 1 (PD1) determine the outcome (survival) of the chimeric cells .

Pregnancy may affect the level of MMc, and it may partially explain the reduced levels of MMc in the younger women studied, who may not have been pregnant before. In a small subset of patients, MMc was found in 45% of parous women in comparison to 22% of nulliparous women . Thus, the same immunoregulatory mechanisms during pregnancy that allow increases in fetal microchimerism may also allow the expansion of MMc.

Tolerance to non-inherited maternal antigens

Why the host immune system does not normally eliminate allogeneic maternal cells is not known. The persistence of maternal cells in a child implies tolerance to maternal antigens, but studies thus far have demonstrated both tolerance and immunity. Tolerance to maternal antigens has been reported in models of heart and skin allografts in the mouse. Maternal T lymphocytes in the lymph nodes, transferred either in utero or through nursing, have been correlated with maternal skin graft survival . An independent study reported a 40–90% reduction in the splenocyte production of IL-2, IL-5, and interferon-γ in response to antigen-presenting cells expressing maternal major histocompatibility complex (MHC) antigens in vitro . More recently, Kinder et al. showed that chimeric maternal cells can drive the postnatal persistence of non-inherited maternal antigen (NIMA)-specific tolerance via antigen-specific regulatory T cells. Overlap between NIMA and fetal antigen during pregnancy highlighted fetal tolerance . In the human system, MMc in the lymph nodes appears to be controlled by regulatory T cells expressing TGF-β .

T cell reactivity to maternal antigens in humans has been reported to be decreased in vivo, allowing increased engraftment of maternal tissues in comparison to a semiallogeneic family or unrelated donors in some, but not all, studies . In vitro studies have shown that although peripheral T lymphocyte reactivity to maternal antigens can be detected, it is reduced in some circumstances in comparison to reactivity to unrelated antigens but not others . In addition, the subset of cells responding to maternal antigens has been shown to be different from the cells responding to paternal antigens . While cells responding to paternal stimulator cells were enriched for CD3 ⁺ /CD8 ^high cells, typical of allogeneic cytolytic T lymphocytes (CTL), responders to maternal stimulators were enriched in CD3 ⁻ /CD8 ^dim cells, a phenotype typical of NK cells. Thus, there is evidence for CD8 ⁺ lymphocyte tolerization to maternal cells, but the mechanisms involved are not known. B cell tolerance has been found in patients after multiple blood transfusions, but it is not known whether the B cells are directly tolerized by maternal antigens, or whether they lack T cell help from tolerized T lymphocytes .

Alloreactive CTL and NK cells are crucial for the elimination of foreign cells after solid organ or stem cell transplantation . NK cells, abundant in fetal blood, would be inhibited by HLA class I molecules on maternal cells that are shared by the child, preventing elimination of MMc . T lymphocytes, however, would be expected to react to maternal HLA molecules not inherited or shared by the child. The fetal immune system has been assumed to be too immature to reject maternal cells , but recent studies suggest that fetal CD8 ⁺ T lymphocytes can develop specificity in utero . Anti-maternal CTLs would therefore be expected to eliminate maternal cells.

Thus, mechanisms for developing tolerance to maternal antigens are not known, but thymic selection may be involved . Donor dendritic cells engrafted into the thymus of the recipient can mediate renal allograft tolerance through clonal deletion of alloreactive thymocytes . In addition, intrathymic renal cells have delayed murine lupus nephritis . Although the peripheral versus central mechanisms of tolerance remain to be explored, preliminary evidence suggests that maternal cells in the thymus may play a role in establishing central tolerance to maternal antigens .

Chronic inflammation in autoimmune disease versus transplantation

Chimerism, the state of cells from two genetically distinct individuals living within one body, can occur through multiple mechanisms. Stem cell transplantation, whether from bone marrow or peripheral stem cells, can lead to a spectrum of chronic inflammatory diseases, namely chronic GVHD . Chronic GVHD has clinical similarities with some autoimmune diseases, including systemic sclerosis (SSc), primary biliary cirrhosis (PBC), Sjögren’s syndrome, and some features of systemic lupus erythematosus (SLE) and myositis, although there are also pathological differences . The chances that a patient will develop chronic GVHD are highly dependent upon the HLA genes of the donor and host. Thus, insights from transplantation chimerism contributed to the hypothesis that microchimerism and HLA relationships of host and nonhost cells are involved in spontaneously occurring autoimmune diseases. Clinical similarities of chronic GVHD and autoimmune disease are now considered in the context of cell transfer between fetus and mother during pregnancy. Comparison of fetal/maternal chimerism to transplantation chimerism must, however, consider the significant differences in cell populations. In both hematopoietic and organ-specific lineages, in the case of fetal and maternal microchimerism, foreign cells are present at a frequency of >1%. By contrast, after stem cell transplantation, donor cells completely replace the hematopoietic system and may also constitute a small fraction of organ-specific cells.

MMc in autoimmune diseases

MMc in the newborn is likely benign or may be beneficial, but it may also transmit malignancy or cause GVHD . Just as stem cell transplantation can lead to loss of tolerance to self-antigens , natural transfer of maternal cells may lead to a child’s loss of self-tolerance. MMc has been found to increase in association with some autoimmune diseases. Long-term MMc was first revealed in the peripheral blood of SSc patients and healthy subjects . Since then, it has been identified in patients with additional SSc , as well as in the target organs and blood of those with neonatal lupus syndrome (NLS) and myositis . In an earlier study, maternal DNA was found to be increased in patients with systemic sclerosis. MMc was initially detected in DNA isolated from peripheral blood mononuclear cells from 22% of healthy controls and 72% of women with SSc , ranging from 0 to 68.6 gEq/million. The assay was then used to analyze MMc in organs from a woman who died of systemic sclerosis. In this patient, MMc was found not only in tissues that were targets of disease but also in tissues that were not involved. High levels of MMc were found in the lungs (757 gEq/million), heart (1489 gEq/million), spleen (466 gEq/million), and pancreas (704 gEq/million). Lower levels of MMc were also found in the gut (39 gEq/million) and bone marrow (48 gEq/million). The 10- to 20-fold higher levels of MMc in the tissues than in the blood suggest that future studies into the mechanisms for the role of maternal cells in inflammatory disease may be best directed toward the parenchymal and immunological cells within the target organs.

We investigated parenchymal MMc in the context of an autoimmune disease that develops in utero, NLS . Infants born to mothers with anti-SSA antibodies are at a risk of developing NLS, with the life-threatening complication of inflammation of the atrioventricular node leading to congenital heart block . Maternal (female) cells were detected and quantified in NLS and control male heart tissues by FISH for X- and Y-chromosome-specific sequences. In blinded studies, maternal cells were found in 15 of 15 sections of heart tissue examined from four NLS patients, ranging from 0.025% to 2.2% of host myocardial cells. Maternal cells were also found in two of eight control sections at lower levels (0.05–0.1%). Because recent studies in transplantation indicate that donor cells can differentiate into somatic tissue cells, we asked whether maternal cells transferred in utero have cellular plasticity. A small minority of maternal cells expressed the hematopoietic cell marker CD45. Eighty-six percent of maternal cells expressed sarcomeric α-actin, a specific marker for cardiac myocytes. These results suggest that differentiated tissue-specific MMc can occur in the neonate. Thus, semi-allogeneic maternal cells could be targets of an immune response. Alternatively, maternal cells could contribute to a secondary process of tissue repair.

In older children, two groups have reported the presence of female cells (presumed maternal) in muscle biopsies from male patients with idiopathic myositis . Age-matched controls who underwent biopsies for other muscle disorders carried fewer female cells. In addition, MMc in the blood was also increased in patients with myositis, as detected by nested PCR assay for maternal HLA alleles not shared with the patient . MMc has also been reported in biliary atresia in the form of hepatocytes and hematopoietic cells , and pityriasis lichenoides, where female cells in the form of keratinocytes were found in skin biopsies from males aged 2–13 years . No female hematopoietic or Langerhans cells were identified. The MMc in the skin was found in 11 of 12 patients at an average level of 99 per million host cells. Maternal cells were also found in some controls, but the levels were much lower (five per million).

Two additional case reports have suggested that MMc may be common in chronic idiopathic inflammatory disease. In the first report, maternal cells were found in an 11-year-old boy with dermatitis, sclerodactyly, myositis, and hepatitis with features of SLE and dermatomyositis . By FISH for X- and Y-chromosomes, maternal nuclei were found in blood and muscle. A second case report described a 21-year-old man exposed to volatile chemicals in a tire factory who developed lymphocytic infiltrates and fibrosis in skin, lungs, and intestinal mucosa pathologically resembling chronic GVHD . HLA typing in this patient showed not only an SSc-associated gene DRB1*1104 but also a bidirectional compatibility at all HLA class II loci with his mother. Maternal cells were found in his blood at a rate of 0.0017%. The patient also showed active T lymphocyte activity against his mother. In vitro mixed lymphocyte cultures demonstrated a two- to threefold increase in CD4 ⁺ and CD8 ⁺ T lymphocyte activation to maternal rather than unrelated donor antigen-presenting cells, as demonstrated by increased HLA DR and CD25 expression. An environmental trigger has also been implicated in mice, where fetal microchimerism was found to be expanded and lead to sclerotic disease in response to polyvinyl chloride administration . Thus, naturally acquired maternal cells, normally present at a low level tolerogenic to the immune system, may be activated by an environmental stimulus to proliferate and expand in blood and affected tissues. At expanded numbers, the maternal cells would provide antigen levels adequate to overcome the threshold for activation of the host immune system. An alternative hypothesis suggests that maternal T lymphocytes may be reactive to the child’s antigens.

T lymphocytes have been found not only in the affected skin of localized scleroderma patients but also in antigen-presenting cells and B lymphocytes . Thus, the range of allogeneic cell roles that maternal cells may play in the blood or tissues of children with autoimmune diseases is not clear. MMc is not found in every suspected disease. Infantile hemangioma, hypothesized to be placental-derived maternal endothelial cells, was investigated for MMc. By FISH for X- and Y-chromosomes, no female cells were identified in hemangiomas from eight patients, although the amount of tissue assayed was not clear, and it may have been low enough that rare cells present were not detected .

Only one other functional study of MMc has been reported . Chimeric maternal T lymphocytes were isolated from patients with myositis and shown to react to the child’s cells in vitro by producing interferon-γ. Maternal cells isolated from siblings did not react to the sibling’s antigen-presenting cells. Thus, although cells may change through culture conditions in vitro, maternal T lymphocytes may recognize the child’s cells expressing non-shared MHC class I or class II molecules in vivo.