Introduction
Defining microchimerism
The term microchimerism was coined by Liegeois (1) in 1977. The name recalls the Lyon-goat-snake merged mythological figure (Figure 1). Microchimerism is defined by the presence of cells in an individual tissue that derives from another genetically distinct individual. This situation can result from a natural process such as pregnancy or between twins or from the mother (natural microchimerism) or after an artificial intervention such as a tissue transplant or blood transfusion (artificial microchimerism). It has also been suggested that sexual intercourse may involve male cell trafficking and theoretically microchimerism could also arise from an older sibling transferred via the maternal circulation to the foetus of a later pregnancy. Furthermore, long-term persistence of foetal cells in the mother (foetal microchimerism) or maternal cells in her offspring (maternal microchimerism) indicates the prolonged potential for a role in immune modulation and/or disease pathology (Table 1).

Figure 1

Chimera. Etruscan bronze sculpture from Arezzo, Italy, 5th–4th century BC. Archaeological Museum, Florence.

Table 1: Types of microchimerism

The consequences of microchimerism
The role of these alien cells in the health of the recipient is of great interest and has not been fully elucidated. However experimental data support important hypotheses concerning their biological significance. It can be assumed that microchimerism may have adverse, neutral, or beneficial effects, depending on a variety of factors (2):

• The bad microchimerism hypothesis was initially proposed by Nelson (3) who suggested that the presence, in women, of foetal cells following pregnancy leads to a graft-versus-host-like response in parous women. As a consequence the maternal immune response to these ‘foreign’ cells may contribute to autoimmune disease pathogenesis.
• There is potential for a good microchimerism, where persistent foetal cells, instead of inducing an immune reaction, are tolerated and have a positive effect as a resource of progenitor cells that may have the capacity to participate in maternal tissue repair processes.
• In addition, there may be neutral microchimerism, where foetal cells are innocent bystanders. It remains possible that foetal cells act as innocent local observers in a process triggered by other mechanisms and play no role in biology at all.

• It is well known that artificial microchimerism can result in chronic graft-vs-host disease, a disorder not dissimilar to autoimmune disease.
• The contribution of natural microchimerism to the origin or exacerbation of autoimmune diseases has been widely documented (4-9) although not yet universally accepted (10). The mechanism of this association has been based on the degree of HLA disparity between host and foreign cells which may determine the strength of any potential graft v host relationship.
• Controversy exists around the role of microchimeric cells in the pathogenesis of diseases of non-immune origin. Part of this controversy is the consideration that the presence of these foreign cells in tissues may be a consequence rather than the cause of disease. Since there may be a variety of foreign cell types persisting within the host tissue, including microchimeric stem cells, and there is potential for a variety of roles for such cells in disease including host tissue repair (11). Hence, factors that may influence foetal cell involvement include foetal or placental cell activities besides histocompatibility (4).
• The number of foetal cells that are transferred into the maternal circulation is considerably greater than the number of maternal cells transferred into the foetal circulation and cell number may be an important factor in any disease initiation process.

Detecting microchimerism
The commonest approach to detecting foreign cells has been the assay of male-specific gene markers in females such as the SRY gene. This can be studied by Y chromosome-specific Polymerase Cell Reaction (PCR) amplification (12,13) or by Fluorescence In-situ Hybridisation (FISH) with labelling of X and Y-chromosomes (14).

Figure 2

Fluorescence in situ hybridisation to detect the Y chromosome-containing cells in the biopsy specimens of the labial salivary glands. The Y chromosome probe was labelled with fluorescein (green signal), and the X chromosome probe was labelled with rhodamine red (red signal). (A) A female patient without Sjögren's Syndrome (SS). The cell nuclei contain two X chromosomes (red signals). (B) A female patient with SS who tested positive for a Y chromosome-specific sequence by PCR. White arrow - nucleus containing one Y chromosome (green signal); red arrow, nucleus containing two X chromosomes (red signals). From Kuroki et al (55).

Figure 3

Interphase fluorescence in situ hybridization (FISH) of thyroid tissue showing a group of microchimeric cells identified by the presence of X and Y chromosomes (orange and green, respectively). The X or Y chromosome may not be observed in each nucleus, as they may not be in the same plane of focus (magnification X400). The FISH assay employed Cy3-labeled X (orange) and fluorescein isothiocyanate conjugated–labeled Y (green) chromosome probes. Nuclei were counterstained with 4',6-diamidino-2-phenylindole (blue). From Khosrotehrani K et al (40).

More recently, to identify the foetal cell type, immunological isolation, such as cell sorting with antibody-coated magnetic beads and flow cytometry, followed by detection of male DNA has also been used (6). Single cell RT-PCR after laser captures dissection is also a potential identifier. However, the most straightforward approach is to use fluorescent labelled male mice such that at least half the foetuses will be labelled and their cells easily identified histologically (15).

2.0. Pregnancy

Immunity in pregnancy
For a successful pregnancy outcome the maternal immune system must not reject the foetus. Trophoblast cells serve as physical barrier, expressing several immune modulating molecules as well as secreting a variety of cytokines. Thus placental immune suppression, which includes reduced activity of T regulatory cells (16), helps establish foetal microchimerism.

Table 2: Immune changes in pregnancy

In normal pregnancy there are important changes in maternal immune responses. A physiological immuno-suppression occurs to create an immune-privileged state protecting the foetus from rejection. Both arms of the adaptative immune responses (cell-mediated and humoral) are attenuated as is natural immunity. This placental immune suppression is thought to help establish fetal microchimerism to different degrees in different women.

Pregnancy and microchimerism
Pregnancy is the cause of foetal microchimerism.

Figure 5

Cell traffic during pregnancy is bi-directional for maternal, foetal and placental cells. Microchimeric cells enter the circulation and persist in the host tissues and are tolerated for many years.

Microchimeric cell types in pregnancy
The origin of microchimeric cells in pregnant women can be foetal or placental cells. It is assumed that foetal cells that enter the maternal circulation are predominantly of haematopoietic origin, such as nucleated red blood cells, lymphocytes, dendritic cells or haematopoietic stem cells (4,32,34-37). However trophoblasts and mesenchymal stem cells also circulate within maternal blood (38,39). Microchimeric foetal cells have been found in almost all maternal tissues at variable frequencies (40). One report found 14-60% of microchimeric cells expressing cytokeratin in three thyroid tissues which suggested that foetal microchimeric cells may differentiate in maternal host tissue (40). It has also been suggested, albeit not proven, that fetal stem cells may persist postpartum and have multilineage capacity. Therefore it is also plausible that foetal microchimeric cells may differentiate even into cells with endodermal lineage such as thyroid cells. In that sense it has been reported that foetal cells exist with leukocyte, or hepatocyte markers, in a variety of maternal tissue specimens (40). Immune cells are a significant proportion of these microchimeric cells and may arise directly from foetal or placental origin or may have been derived from pluripotent microchimeric stem cells. Differentiation of such foetal stem cells may be in response to pre-existing tissue injury such as in autoimmunity and infection (6). Furthermore, the microchimeric T cells are immunologically competent as shown by proliferation to maternal antigen (41). Therefore it has been hypothesised that such cells may modulate autoimmune diseases in the postpartum by playing a role in antigen presentation and immunoregulation (8,16).

Autoimmune diseases and pregnancy
Most autoimmune diseases, and especially autoimmune thyroid diseases, are much more frequent in the middle-aged female population and are well known to be modulated by pregnancy (6,16,42). Autoimmune diseases also have an especially high prevalence in the postpartum period (43). Whereas pregnancy leads to an amelioration of autoimmunity, the postpartum period is associated with an exacerbation of autoimmune disease. For example, Graves' disease and postpartum thyroiditis can affect 8-10% of women and up to 40% of TPO-Ab positive women in the post partum (16). It has been presumed that the postpartum exacerbation of the autoimmune diseases is more likely due to loss of placental immune suppression rather than maternal exposure to persisting specific allogeneic foetal antigens but there is no evidence to dismiss a microchimeric cause (6). Up to 60% of reproductive Graves' patients reported the development of Graves' disease within one-year of delivery (44).

Microchimerism and pregnancy related abnormalities
Some pregnancy abnormalities such as pre-term labour, pre-eclampsia and aneuploidy have been related to increased numbers of circulating foetal cells in women (9). A significant association between foetal loss and microchimerism has also been observed (45). Polymorphic eruptions of pregnancy are pruritic, non-follicular erytematous papules in the skin where male foetal cells have been detected (31). In each of these situations there is a reported association but no mechanistic relationship has been described.

3.0. Autoimmune Disease and Microchimerism

Autoimmune disease and maternal microchimerism
Many scientists had linked cells transferred from foetus to mothers to the development of some autoimmune diseases, but the evidence linking microchimeric maternal cells to the development of these diseases is less clear. However new discoveries have shed light on the physiological activity of maternal microchimeric cells in offspring. Nelson et al have recently shown that cells which passed from mother to child during pregnancy can differentiate into pancreatic islet beta cells in type 1 diabetes patients (46). The authors speculated that these cells may play a role in aiding the production of insulin. They found that 51% of the blood samples from 94 patients with autoimmune type 1 diabetes had maternal microchimeric cells, as did 33% of 54 unaffected siblings and 17% of 24 healthy controls. The differences between groups were significant. The researchers also found maternal cells in pancreatic islet tissue of four male cadavers. They concluded that maternal cells help to regenerate damaged tissue. In this paper, in contrast with their previous opinion (25), Nelson et al. hypothesised that maternal microchimeric cells do not appear to trigger the autoimmune response that contributes to the development of type 1 diabetes. However this interesting finding, while confirming previous results, does not rule out other possible capabilities for maternal microchimeric cells. It seems, therefore, premature to conclude that microchimeric maternal cells were not causing an autoimmune response as well as facilitating repair of the pancreas tissue. Other authors have found maternal cells in inflammatory lesions of scleroderma patients which express similar antigens to those of the autologous cells in the same lesion, supporting the contribution of maternal microchimeric cells in the origin of autoimmune diseases in their children (9).

Autoimmune disease and foetal microchimerism
The presence of microchimeric male cells in blood samples of women with previous male pregnancies ranges from 25 to 31% (8). This frequency increases in those women who harbour any autoimmune disease, ranging from 45 to 60%. The existence of foetal microchimerism in relation to human autoimmune disease was first confirmed in systemic sclerosis patients (47). Studies demonstrated the presence of male DNA in skin lesions of women with systemic sclerosis (48). However this association it is not as clear in other autoimmune diseases with female preponderance such as primary biliary cirrhosis, Sjögren's syndrome or erythematosus systemic lupus (49). In addition some reports have shown that foetal cell microchimerism is a relatively common occurrence in women with both autoimmune and non-autoimmune diseases (2). Therefore, such data are contradictory in relation to the health consequences of persistent foetal cells in maternal tissues.

Autoimmune thyroid disease and microchimerism
The thyroid is the most common target for autoimmunity in humans. The high female preponderance and the high prevalence in women after childbearing suggest that pregnancy related factors have a strong influence on thyroid autoimmune disorders such as Graves' disease and Hashimoto's thyroidtis. Autoimmune thyroid disease may initiate or exacerbate in the postpartum period, whereas during pregnancy the activity of these disorders is reduced in relation to placental immune suppression.
The influence of pregnancy in autoimmune thyroid disease was investigated in our laboratory in a murine model of experimental autoimmune thyroiditis, showing that thyroglobulin immunisation leads to foetal loss in specific allogenic pregnancies (42). In later studies we focused on the investigation of the relationship between foetal microchimerism and autoimmune thyroiditis. The results demonstrated a significantly higher accumulation of foetal cells in the maternal thyroids from thyroglobulin immunised animals (46%) when compared with control non-immunised pregnant mice (20%) (15). The microchimeric cells included cells of foetal T cell and dendritic cell lineage. These results suggested that the inflamed thyroid gland was capable of accumulating foetal immune cells that may have a regulatory role on maternal autoimmune thyroiditis. We also found that the presence of foetal cells in the maternal thyroids decline rapidly after delivery, indicating the recovery of the maternal immune system in the postpartum period. However, in allogeneic pregnancies this was followed by an increase in the intrathyroidal infiltrate, as a consequence of the conclusion of the pregnancy related immune-privileged status (15). In contrast to the mouse model, many women show persistence of fetal cells for many years and this is reflected in the thyroid gland of those women with autoimmune thyroid disease. The reasons for the persistence of apparently foreign cells in the human thyroid are not completely understood. Currently the only known factor that determines the persistence of microchimeric cells in the host is the Human Leukocyte Antigen compatibility between mother and foetus (50). In addition the immunogenetic susceptibility markers, HLA DQA1*0501-DQB1*0201 and DQB1*0301, that are more frequent in patients with thyroid autoimmunity are also more common in patients of mother-child pairs with microchimerism (51).
In theory, the presence of immune foetal cells within the maternal thyroid gland may elicit a response once the immune-privileged pregnancy status is finished in the postpartum period (8). Therefore, it has been hypothesised that immune foetal cells interact with maternal immune cells during the postpartum to initiate or exaggerate autoimmune postpartum thyroid disease in a host versus graft reaction.

Figure 6

Hypothetical mechanism by which foetal cells may modulate autoimmune thyroid disease in the postpartum period. Foetal microchimerism is established during pregnancy by placental immune suppression. The mechanism that attracts foetal immune cells (or other cells as foetal stem cells) to migrate into the maternal thyroid gland has not been fully elucidated but is exaggerated by inflammation. Cytokines, chemokines, and adherent factors may be involved. Due to placental immune suppression, immunological interaction between maternal and foetal immune cells is minimal and/or neglectful during pregnancy. After delivery, partially sustained immunosuppressive effects facilitate the survival of foetal cells, but eventual loss of placental immune suppression results in activation intrathyroidal foetal immune cells. Activated foetal immune cells initiate a graft vs. host reaction against maternal antigens by secreting immunomodulatory cytokines and/or expressing immunomodulatory molecules, which activates intrathyroidal maternal autoreactive T cells and eventually initiates and/or exacerbates postpartum autoimmune thyroid disease. Modified from Ando and Davies (8).

Table 3: Summary of studies of foetal microchimeric male cells in the thyroid gland

This table summarises the population of females with different thyroid diseases in whom the presence of microchimeric male cells has been studied. The information was collected and combined from five very different studies: Ando et al (12), Klintschar et al (13) Srivasta et al (14), Renne et al (52) and Klintschar et al (53). For the correct interpretation of this information it is important to considerer that the design and techniques of these studies were diverse. Ando et al and Klintschar et al used a PCR approach while Srivasta et al and Renne et al used FISH technique. It, therefore, provides only a rough guide to the situation.

FISH: Fluorescence In-situ Hybridisation. PCR: Polymerase Cell Reaction.

Hashimoto's thyroiditis
Four studies have found male cells in samples of Hashimoto's thyroiditis, assuming they were microchimeric cells related to a previous pregnancy (13,14,52,53). The presence of these cells in the studied females ranged between 38 to 83%. This discrepancy probably reflects the low number of samples examined, the differences in the study designs and the variation in the evaluation techniques (FISH or PCR). The studies also showed a wide discrepancy in the quantification of the presence of male cells per subject. The two laboratories which utilised FISH showed results that range from 0 to 35 male cells per slide. Analysis by quantitative real time PCR showed that thyroid of Hashimoto's thyroiditis patients contained male cells that ranged from 0 to 4,900 per 100,000 female cells. The median of male cells detected in those women with positive results was 97 male cells/100,000 female cells. The differences between the Hashimoto's thyroiditis patients and control groups in individual studies are constant. In the first study published by Klintschar et al (13), sequences were derived from the Y-chromosome in 8 out of 17 thyroids of women diagnosed with Hashimoto thyroiditis. In contrast these cells were found only in 1 of 25 nodular goitres that were used as a control group. This highly significant difference between both groups, despite similar characteristics in age and number and gender of children, was strong evidence of an etiological role of microchimerism in the pathogenesis of Hashimoto's disease (13). When compared both autoimmune entities (Graves' disease and Hashimoto's thyroiditis), the frequency of microchimerism was highest in patients with Hashimoto's thyroiditis (52). Similarly, when compared, microchimeric cells were found in more subjects with Hashimoto's thyroiditis (83%) than in any other thyroid diseases (14).

Graves' disease
The main study of the association of Graves' disease with microchimerism was performed in a sample of 27 glands of affected subjects by ELISA-PCR technique (12). Twenty thyroids of Graves' disease patients had been preserved in paraffin block while seven had been frozen. In the first group only 4 (20%) samples were positive for SRY gene, while 6 (86%) of the frozen samples showed the Y chromosome gene. In addition intrathyroidal microchimerism was more common and profound in samples from patients with autoimmune Graves' disease than in benign adenoma (12). Surprisingly the study also showed that many women with male cells in their thyroids declared no past history of male pregnancies. However this did not necessarily exclude the possibility of undetected first trimester pregnancies, because it has been demonstrated that foetal microchimerism can be established in the first month of pregnancy (12,27). Some authors have argued that as they only detected male cells in mothers with at least one son, pregnancy appeared to be (in absence of transplants or transfusions) a condition sine qua non for microchimerism (53). However we believe that the sources of natural microchimerism are much more complex and involve other factors such as cell trafficking from maternal origin or intercourse. At the same time, the techniques for detecting microchimeric cells are limited, and the majority of the studies have looked only for male cells. This procedure is unable to detect microchimeric cells from daughters, which presumably are the source of similar amounts of microchimeric cells as sons.

Microchimerism and non-autoimmune diseases
It is also conceivable that microchimerism contributes to non autoimmune thyroid disease (52). The presence of microchimeric cells in nodular disease might result from anatomical changes or associated with local expression of growth factors re-lated to goitrous growth and repair processes (14). The relationship between foetal cell microchimerism and non-autoimmune diseases, including infectious disease and cancer, has led to speculation that foetal cells may provide a rejuvenating source of progenitor cells. These cells may potentiality participate in maternal tissue repair processes. It has been hypothesised that microchimeric cells have the ability to migrate from the circulation, home to diverse tissues, and differentiate or fuse with host cells (2). In fact male cells of presumably foetal origin with follicular morphology have been observed in a thyroid of a woman affected with a multinodular goitre (14). Foetal cells have also been identified in thyroid adenomas and adenomatous goitres although a local inflammatory infiltrate may have been responsible. The presence of microchimeric cells ranges from 45 to 57% of cases of nodular thyroid disease although it is unclear if this was related to an associated perinodular thyroiditis (13,14,52,53). The highest presence was detected in a FISH study where up to 156 male cells were found in a sample from a multinodular goitre. This was fourfold higher than that found in a Hashimoto's sample. Klintschar et al (13) found that only one of the 25 (4%) thyroids of women with multinodular goitre was positive for male gene PCR. Ando et al (12) did not show SRY gene in the adenomas which were preserved in paraffin but their SRY PCR came out with positive results in 25% of the frozen adenoma samples. Renne et al (52) showed similar results (22%) for microchimeric cells in nodular thyroid disease. The role of these cells in adenoma formation is unknown.

Points of controversy
Although some data conflict with the hypothesis of a direct relationship between microchimerism and autoimmune disease it remains a consistent observation with many examples. The absence of foetal cell microchimerism in some female preponderant autoimmune disorders such as primary biliary cirrhosis or Sjögren's syndrome may simply imply a different immunopathologic mechanism at work in these disorders. Furthermore, the presence of microchimerism in non-autoimmune diseases including infectious hepatitis and even thyroid adenomas may simply be a reflection of an inflammatory response in susceptible individuals attracting foetal cell accumulation (2,14). An Australian study of more than one thousand women of varying parity concluded that pregnancy was not a risk factor for autoimmune disease (54). But that does not exclude the fact that that pregnancy modulates autoimmune disease when it exists. Nevertheless, it remains to be shown that the presence of foetal cells directly influences the natural history of autoimmune disease.

4.0. Conclusions
There is a large body of evidence that indicates the thyroid gland as an important organ where foetal microchimeric cells settle and persist and this is more common in autoimmune than in non-autoimmune thyroid diseases. The biological consequences of foetal immune cells within the maternal thyroid gland, which may become activated in the postpartum period, as maternal immune suppression is lost, remains an attractive explanation for the postpartum exacerbation of the autoimmune thyroid diseases. Understanding the types of foetal cells transferred to mothers, the location of these cells within diverse maternal organs and their function may provide incite into their role in immunopathology and tissue repair and allow us to take advantage of new therapeutic procedures.