Long-Term Fetal Microchimerism in Peripheral Blood Mononuclear Cell Subsets in Healthy Women and Women With Scleroderma

Clinical specimens were collected from 48 healthy women with sons including 13 sisters of women with SSc. The mean age was 43.1 years (34 to 71). The mean number of children was 2.2 (1 to 7) with mean number of sons 1.7 (1 to 6). The mean age at birth of the first child was 27.5 years (15 to 39), mean age at birth of first son was 28.1 years (15 to 39), mean age at birth of last child was 30.3 years (15 to 39), and mean age at birth of last son was 29.5 (15 to 39). The mean age of the last child to be born was 12.8 years (0 to 45) and mean age of the last son to be born was 15.1 (0 to 45).

Twenty women with SSc with sons were studied with the mean age of 48.8 years (34 to 71). The mean number of children was 3.8 (1 to 7) with mean number of sons 1.8 (1 to 4). The mean age at birth of the first child was 24.1 years (17 to 36), mean age at birth of first son was 24.8 years (18 to 36), mean age at birth of last child was 28 years (23 to 36), and mean age at birth of last son was 27.3 years (20 to 36). The mean age of the last child to be born was 19.6 years (1 to 40) and the mean age of the last son to be born was 21.1 years (1 to 42). The mean age at diagnosis of SSc was 44.1 years (29 to 62) and the mean time between birth of first child and diagnosis of SSc was 17.3 years (2 to 42). Specimens were also collected from 2 nulligravid women and 22 parous women who had not given birth to a son.

A total of 30 mL acid-citrate-dextrose (ACD)–preserved peripheral blood was purified by ficoll hypaque density centrifugation. DNA was extracted from PBMC using the Isoquick Nucleic Acid Extraction Kit (Orca Research Inc, Bothell, WA) in accordance with the manufacturer’s instructions.

PBMC were purified from ACD–preserved blood as described above and resuspended in phosphate-buffered saline (PBS)/1% fetal calf serum (FCS). Ten to 20 × 10⁶ cells were aliquotted into three tubes and stained with anti-CD3 fluorescein isothiocyanate (FITC) and anti-CD56/16 PE (tube 1), anti-CD14 FITC and anti-CD19 PE (tube 2), and anti-CD34 PE and anti-CD14 FITC (tube 3). Four μL of each fluorescently conjugated antibody (Becton Dickinson, Mountain View, CA) was used for staining a total volume of 500 μL. Tubes were incubated on ice in the dark for 30 minutes then washed twice with 5 mL PBS/1% FCS before FACS sorting. Cells were sorted into CD3⁺, CD56/16⁺, CD14⁺, and CD19⁺ populations. A proportion of sorted cells was then examined by FACS to check purity, which was always 95% to 99%. Cells were then collected by centrifugation and stored above liquid nitrogen. DNA was extracted using the Isoquick Nucleic Acid Extraction Kit (Orca Research Inc) and Y-chromosome–specific PCR was performed. DNA was routinely extracted from 0.5 × 10⁵ cells although occasionally smaller or greater numbers were studied.

Nested PCR for a single-copy Y-chromosome–specific sequence was modified from Lo et al5; reagents were supplied by Perkin Elmer unless otherwise stated. Measures were taken to prevent contamination including dedicated rooms, equipment, and reagents for PCR reaction mix and product analysis and use of three laminar flow hoods for DNA extraction, reaction mix preparation, and transfer of primary products after step 1. For step 1 each reaction tube contained 1 μg template DNA, 50 pmol sense primer (5-CTAGACCGCAGAGGCGCCAT-3; Oligos Etc) and 50 pmol antisense primer (5-TAGTACCCACGCCTGCTCCGG-3; Oligos Etc), 200 μM dNTPs, 1.5mmol/L MgCl₂, 1 × Taq polymerase buffer, 1 μL Perfect Match Enhancer (Stratagene, La Jolla, CA), and 1μL Amplitaq gold. Forty cycles were performed (94°C for 1 minute, 67°C for 1 minute, 72°C for 2 minutes). For step 2 each reaction tube contained 2 μL reaction product from step 1 DNA, 50 pmol sense primer (5-CATCCAGAGCGTCCCTGGCTT-3; Oligos Etc) and 50 pmol antisense primer (5-CTTTCCACAGCCACATTTGTC-3; Oligos Etc), 200μmol/L dNTPs, 1.5 mmol/L MgCl₂, 1 × Taq polymerase buffer, 1 μL Perfect Match Enhancer (Stratagene) and 1μL Amplitaq gold. Twenty five cycles were performed (94°C for 1 minute, 55°C for 1 minute, 72°C for 2 minutes). Positive (male DNA) and negative (water) controls were included in each run. A total of 5 μL PCR product was electrophoresed in a 2% agarose gel (Sigma Chemical Co, St Louis, MO) using TAE buffer (0.04 mol/L Tris-acetate, 0.001 mol/L EDTA). Gels were photographed over ultraviolet light after staining with ethidium bromide. Serial dilutions showed that the DNA equivalent of one male cell could be detected in a background of 4 × 10⁵ female cells. No amplification was observed from negative water controls or from DNA extracted from nulligravid women. Southern blotting was performed onto nylon membrane (Boehringer Mannheim, Mannheim, Germany) following the manufacturer’s instructions to confirm specificity. A PCR product-specific oligonucleotide (5′-CAGCTCGGCTTCGATGTGACTCTT-3′) was end labelled with γ-ATP (Amersham, Arlington Heights, IL) and used to hybridize against blotted PCR product to confirm specificity.

The presence of fetal microchimerism was substantiated further in some patients using HLA-specific PCR. PCR reactions were performed in a volume of 50 μL containing 1.5 μg genomic DNA, 10 mmol/L Tris-HCl (pH 8.3), 50 mmol/L KCl, 1.5 mmol/L MgCl₂, 0.001% wt/vol gelatin, 260 μmol/L of each deoxynucleotide, 0.5 U Perfect Match Enhancer (Stratagene), 2.5 U Amplitaq Gold DNA Polymerase (Perkin Elmer-Cetus, Norwalk, CT) and 20 pmol of each of the HLA-specific primers. Amplification consisted of 5 minutes at 96°C followed by 35 cycles at 95°C for 35 seconds, 55 to 65°C for 35 seconds and 72°C for 1 minute with a final extension step at 72°C for 10 minutes using a Gene Amp System 9600 (Perkin Elmer-Cetus). Optimum amplification sensitivity and specificity was achieved for each primer set by titrating MgCl₂ and primer concentrations and optimizing annealing temperature and number of thermocycles. To control for nonspecific amplification of background HLA alleles, DNA was extracted from control PBMC or from Epstein-Barr virus (EBV)–transformed human B-lymphoblastoid cell lines expressing HLA alleles of the mother (and not the child). Negative controls comprising all PCR reagents without DNA were also included. DNA from the subject’s child served as a positive control for HLA-specific PCR.

Calculations of statistical significance were done using the Mantel-Haenszel test of the null hypothesis that the odds ratio is equal to one (StatXact program).

Thirty-one percent (11 of 35) of normal healthy women had Y chromosome DNA in PBMC. Patients with SSc were more frequently positive for Y chromosome DNA in PBMC than healthy women; 60% of SSc patients were positive, P = .042 (Table 1). The frequency of Y chromosome PCR-positive PBMC was not significantly different amongst sisters of women with SSc and healthy women. Patients with SSc were more frequently positive for Y chromosome DNA in PBMC than the combined control population of healthy women and sisters of women with SSc; 60% of SSc patients were positive compared with 33% controls, P = .046.

In control experiments we found that two nulligravid women were consistently negative for Y-chromosome–specific DNA. However, 10 of 22 parous women who had never given birth to a male were positive (5 patients and 5 controls). Eight of the 10 had prior pregnancy loss and/or transfusions that represent a potential source of Y chromosome DNA in these subjects.

The presence of DNA from a son as detected by Y-chromosome–specific DNA PCR was confirmed in some families by use of HLA-specific PCR. In these families HLA differences of the son were exploited to study DNA extracted from PBMC of the patient. In one family DRB1*01-specific primers were used, in two families DRB5-specific primers, and in another family B44 (HLA class I)-specific primers were used.

PCR for Y chromosome DNA was performed on FACS sorted subsets for 20 women. Eleven controls comprised 9 normal healthy women and 2 healthy sisters of women with SSc. All subjects were positive for Y chromosome DNA in unsorted PBMC. Table 2 shows that Y-chromosome–positive cells were frequently detected in PBMC subsets for both controls and patients. Three of 10 (30%) controls and 3 of 9 (33%) scleroderma patients were positive in CD3⁺ cells. In CD56/CD16⁺ cells 4 of 9 (44%) and 5 of 8 (63%) were positive. In CD14⁺ cells 4 of 11 (36%) and 2 of 9 (22%) were positive. In CD19⁺ cells 5 of 11 (45%) and 2 of 8 (25%) were positive respectively. Two patients had not given birth to a male child, however, patient SSc4 had two prior pregnancy losses and received transfusions at the time of childbirth and patient SSc33 had prior pregnancy loss.

Family HLA studies were completed for 17 women (sons of 1 woman were not available and 2 did not have a son as noted above). Sons were HLA class-II incompatible with their mothers for DRB1, DQA1, and DQB1 in the majority of families (53%). Nevertheless, persistent microchimerism was found in PBMC subsets in all of these women (Table3). Persistent microchimerism was detected in each of the subsets, CD3, CD56/16, CD14, and CD19 in some women. No significant difference was apparent for 8 women with a son who was compatible for 1 or more of the class-II loci, DRB1, DQA1, and/or DQB1. Again, persistent microchimerism was detected in some women in each of the subsets, CD3, CD56/16, CD14, and CD19.

Application of molecular biological techniques to the study of human pregnancy has resulted in the appreciation that there is bidirectional traffic of cells between mother and child.1 Moreover, fetal progenitor cells have been found to persist in the maternal peripheral blood for decades after pregnancy.2 Because progenitor cells can differentiate into other immune-competent cells populations, we asked whether persistent microchimerism occurs in PBMC subpopulations. Y chromosome DNA served as a marker for persistent male fetal cells in women who had previously given birth to a son. Healthy women frequently had male DNA in their PBMC despite a mean age of the last son to be born of 15 years.

Long-term persistence of fetal cells has potential biological significance,1 including the possibility that these cells could be involved in some autoimmune diseases.6 We studied women with SSc because of the hypothesized link between microchimerism and development of this autoimmune disease.6 SSc is a progressive and often fatal connective tissue disease characterized by inflammation, fibrosis, and obliterative vasculopathy of skin, lung, heart, kidney, and gut.7 SSc shares a number of characteristics with cGVHD that may arise after hematopoietic stem-cell transplantation,4 has a higher incidence among women than men, and rises sharply following childbearing years.3 A significantly greater proportion of women with SSc had male DNA in PBMC compared with control women with sons. The mean age since last birth of a son in SSc patients was 21 years. The nested PCR test employed was not quantitative, but the difference between healthy women and women with SSc could reflect a quantitative difference because negative subjects may harbor fetal cells at a level below the sensitivity of the test. This is consistent with our previous report of a smaller series that was limited to testing of whole blood samples in which it was shown that women with SSc compared with controls harbor a greater quantity of DNA of fetal origin in peripheral blood.8 

In PBMC subpopulations we found persistent microchimerism in the majority of healthy women and also in women with scleroderma. Ninety percent of women had microchimerism in some PBMC subpopulation, either in T lymphoctyes, B lymphoctyes, monocytes, and/or in NK cell populations. Only one woman was positive for all subpopulations, and this subject had the most recent delivery of a son. Thus, most women evidenced persistent microchimerism in some, but not all PBMC subpopulations. Microchimerism may be detected more readily after application of FACS because greater concentrations of specific PBMC subsets can be tested than would be present in whole PBMC. Due to the relative purity with which PBMC subpopulations can be attained with FACS it is possible that microchimerism detected could be from a contaminating-cell subset. The nested PCR was sensitive, capable of detecting the DNA equivalent of a single cell so that if, for example, sorting purity was 95% there could be a 1 in 20 chance that microchimerism detected was not from the subset of interest. Whereas it is possible that an individual determination could be in error, a systemic error is unlikely, purity was always greater than 95%, and most subjects were studied on more than one occasion and with multiple aliquots of cells. It is therefore unlikely that this consideration impacts the overall study results.

Long-term microchimerism of fetal cells in the peripheral blood of parous women is a recent concept.2 Persistence of fetal cells after nonterm pregnancies has not been specifically studied, but is the most likely explanation for positive results in some women without a son. Fetal cells have been detected in maternal peripheral blood as early as 5-weeks gestation.9 In control experiments we found Y chromosome positivity in 10 parous women without sons, all but two of whom had history of prior pregnancy losses and/or had received blood transfusions. Another potential source is through transfusion that has, in some cases, led to the development of GVHD.10,11 A potential source of male chimeric cells remains unexplained for two women who may have experienced an unrecognized early pregnancy loss. Although the possibility of contamination cannot be entirely ruled out, stringent measures were employed to minimize this risk and negative controls were consistently negative. A final potential source of microchimerism is engraftment of cells from a twin that could occur early in pregnancy with later unrecognized loss of the twin. Chimeric cells were first described by Owen12 who detected red blood cell antigen sharing between dizygotic twin cattle. Interestingly, one control woman in the present study, with two daughters and one early pregnancy loss also had a twin brother, and was consistently positive for Y chromosome DNA.

The factors facilitating microchimerism are poorly understood. It can be envisaged that patient age and/or time since childbirth could influence fetal-cell microchimerism; however, there was no readily apparent correlation of these variables with results in our study. Total parity also did not correlate with microchimerism.

Persistent fetal microchimerism raises the issue of maternal tolerance to fetal paternally inherited HLA antigens. Maternal T-cell13 and humoral14 awareness of paternally inherited antigens has been shown during and after pregnancy. In a murine model it was found that fetal cells are cleared from the maternal circulation more rapidly after allogeneic matings than syngeneic.15 In this study there was no apparent correlation between HLA compatibility of a son and the long-term persistence of male DNA in PBMC subsets. We found evidence for persistent microchimerism in CD3⁺, CD56/CD16⁺, CD14⁺, and CD19⁺ PBMC subsets in healthy women and in women with SSc for whom all previously born sons were HLA-incompatible.

In summary, our findings show that persistent fetal microchimerism is not uncommon in normal healthy women in T-lymphocyte, B-lymphocyte, NK, and monocyte-cell populations. Significantly, more women with SSc had microchimerism in unsorted PBMC than controls, however, additional studies will be necessary to address the potential role of microchimerism in the pathogenesis of SSc and other autoimmune diseases.