Microchimerism is strongly correlated with tolerance to noninherited maternal antigens in mice

Abstract

In mice and humans, the immunologic effects of developmental exposure to noninherited maternal antigens (NIMAs) are quite variable. This heterogeneity likely reflects differences in the relative levels of NIMA-specific T regulatory (T(R)) versus T effector (T(E)) cells. We hypothesized that maintenance of NIMA-specific T(R) cells in the adult requires continuous exposure to maternal cells and antigens (eg, maternal microchimerism [MMc]). To test this idea, we used 2 sensitive quantitative polymerase chain reaction (qPCR) tests to detect MMc in different organs of NIMA(d)-exposed H2(b) mice. MMc was detected in 100% of neonates and a majority (61%) of adults; nursing by a NIMA+ mother was essential for preserving MMc into adulthood. MMc was most prevalent in heart, lungs, liver, and blood, but was rarely detected in unfractionated lymphoid tissues. However, MMc was detectable in isolated CD4+, CD11b+, and CD11c+ cell subsets of spleen, and in lineage-positive cells in heart. Suppression of delayed type hypersensitivity (DTH) and in vivo lymphoproliferation correlated with MMc levels, suggesting a link between T(R) and maternal cell engraftment. In the absence of neonatal exposure to NIMA via breastfeeding, MMc was lost, which was accompanied by sensitization to NIMA in some offspring, indicating a role of oral exposure in maintaining a favorable T(R) > T(E) balance.

Figure 1

Microchimerism in F1 backcross offspring. (A) The left panel shows a representative PCR curve showing standard and negative controls. The right panel shows H2D^d standard curve. Each data point is shown as mean ± range of values. qPCR was performed with BDF₁ DNA diluted into B6 DNA in different concentrations. B6 DNA and blank (DNAase-free water) were used as negative controls. Primers and probe specific for maternal H2D^d were used to detect maternal DNA. A standard curve was made from 15 independent experiments. (B) Histogram showing variability in the levels of Mc in NIMA^d-exposed versus control F1 backcross mice (standard breeding), plotted by percentage of offspring with indicated number of organs positive by H2D^d qPCR. (C) Top panel shows tissue distribution of Mc in different organs of NIMA^d-exposed (n = 41 [*n = 12 for blood only]) and NIPA^d (n = 16) control mice expressed as H2D^d gene equivalents (GEq) per 10⁵ cells. The y-axis shows estimated GEq per 10⁵ cells. Bottom panel shows GFP Mc in different organs of NIMA^d NIMA^GFP-exposed (n = 11) and NIPA^d NIPA^GFP control (n = 6) mice. (D) Summary of tissue distribution data in panel C expressed as a percentage of NIMA-exposed mice with detectable levels of Mc in each organ.

Figure 2

MMc in hematopoetic versus parenchymal cell lineages. (A) Cells from spleen and bone marrow of 6 NIMA^d-exposed mice (standard breeding) were separated by expression of CD4, CD8, CD11b, and CD11c using MACS beads, and DNA was extracted from separated populations and tested for H2D^d MMc. The figure shows levels of MMc (GEq/10⁵ sorted cells) in the sorted cell populations. Individual mice are represented by symbols as shown. (B) Cells of hematopoietic lineages (lineage⁺) and nonhematopoietic lineages (lineage⁻) were sorted from single-cell suspensions of heart tissue pooled from 8 NIMA^d-exposed mice. Lineage⁺ cells were sorted with a lineage cell depletion kit containing a cocktail of monoclonal antibodies against CD5, CD45R (B220), CD11b, Gr-1 (Ly-6G/C), 7-4, and Ter-119. Maternal cells were detected using H2D^d qPCR from the sorted fractions. The dots represent replicate values of analysis of the pooled lineage⁺ or lineage⁻ cell DNA sample.

Figure 3

Detection of NIMA^d-specific T_R cells using a DTH bystander suppression assay. Splenocytes were collected from NIMA^d-exposed and NIPA^d control offspring immunized with tetanus toxoid and diphtheria (TT/DT) vaccine 2 weeks previously. A total of 10 million splenocytes were injected into footpads of naive B6 recipients with coinjection inoculums: PBS, BDF₁, or B6C3F₁ Ag, or TT/DT. To measure bystander suppression, splenocytes were coinjected with BDF₁Ag and TT/DT. Changes in footpad thickness were measured after 24 hours of injection to measure DTH reaction. (A) Examples of DTH bystander suppression assays. Spleen cells from 2 NIMA^d-exposed mice (nos. 7 and 8) and 1 NIPA^d control mouse (no. 1) that had been immunized with TT were coinjected into footpads with antigens shown on x-axis. Splenocytes from NIMA^d no. 7 had a strong DTH response to TT but not to BDF₁ antigen; the TT response was suppressed by 50% in the presence of BDF₁ (maternal) Ag, but not in presence of third-party (B6C3F₁) Ag. NIMA^d no. 8 and the NIPA^d control (no. 1) splenocytes did not suppress their strong anti-TT DTH responses in presence of the BDF₁ Ag. (B) Summary of bystander suppression values in n = 54 mice. The horizontal lines represent mean values.

Figure 4

Correlation between the level of MMc and percentage of DTH inhibition. (A) Linear correlation between the percentage of DTH suppression and percentage of positive organs (out of a total of 8 or 9 organs tested per mouse) by qPCR. (B) NIMA^d-exposed mice were divided into 2 groups on the basis of percentage of DTH suppression (50% and more and less than 50%) and plotted against the level (GEq/10⁵ cells) of MMc. The y-axis shows average estimated GEq per 10⁵ cells (average of 8 or 9 organs tested per mouse). (C) Percentage of DTH suppression plotted against levels of MMc averaged from 8 or 9 organs tested per mouse. The diagonal lines represent the best fit regression lines. Horizontal lines represent mean values.

Figure 5

“In vivo MLR” analysis of lymphocytes from NIMA^d-exposed versus NIPA^d control offspring. Splenocytes were harvested from NIMA^d-exposed and NIPA^d control mice and labeled with CFSE. CFSE-labeled splenocytes (50 × 10⁶) were injected intravenously into a BDF₁ recipient, which is the maternal type in the F₁ backcross breeding system. After 3 days, splenocytes and ILN cells were harvested from the BDF₁ recipients. (A) Flow cytometric analysis of lymphoproliferation in BDF₁ hosts. H2K^d-negative CFSE-labeled splenic lymphocytes found in BDF₁ ILNs were gated as shown. Total lymphocyte proliferation from MMc⁺ mouse NIMA^d no. 9 was less than MMc^neg NIMA^d no. 10 and NIPA control (no. 2). (B) Inverse correlation between the level of MMc and percentage of proliferated responder lymphocytes recovered from BDF₁ lymph nodes (n = 22). (C) Gating strategy for H2K^d-negative (donor) CD4⁺ T cells. (D) Surface TGF-β and (E) intracellular Foxp3 staining of donor CD4⁺ T cells. NIMA^d no. 9–derived CD4⁺ T cells expressed a higher amount of surface TGF-β staining on nonproliferated CD4⁺ T cells than NIMA^d no. 10 and NIPA control (no. 2). The horizontal line indicates the level of staining with isotype control antibody. (F) Correlation between the level of MMc and percentage of donor CD4⁺ T cells expressing surface TGF-β (n = 15). The diagonal lines represent the best fit regression lines in panels B and F.

Figure 6

Cells of NIMA^d-exposed mice express low levels of maternal MHC class I antigen. Cells from thymus, spleen, pooled LNs, bone marrow, and blood of NIMA^d-exposed and NIPA^d control mice were stained with anti-H2K^d antibody and analyzed by flow cytometry. A minimum of 10⁶ events were collected for each sample. All cells in the live gate (propidium iodide–excluding) were included in the analysis. (A) An example of flow histograms of H2K^d expression on peripheral blood mononuclear cells (PBMCs) from a NIMA^d-exposed, NIPA^d control, B6 negative control, and BDF₁ positive control mouse. (B) Summary data of percentage of H2K^d-dim cells in various lymphoid organs of NIMA^d-exposed mice (▵) and NIPA^d control mice (●). B6 background values have been subtracted. The horizontal line indicates the lower limit of sensitivity by flow cytometry. (C) Example of levels of maternal H2K^d-dim offspring cells in blood at 4 different time points after weaning. (D) Comparison of peak percentages of maternal antigen-dim offspring cells (from 4 different time points tested) among 3 groups of offspring. The error bars represent SEM.

Figure 7

Oral exposure to maternal antigens is important in maintaining tolerance and preventing NIMA-induced sensitization. (A) Left panel shows average levels of MMc found in the heart, liver, and lungs in newborn offspring (less than 1 day old). Levels of MMc in heart, liver, and lungs of the newborn offspring are depicted in the right panel. (B) Newborn NIMA^d-exposed mice were either nursed by the birth mother, or separated from their BDF₁ mothers within 6 hours of birth and foster-nursed (F.N.) by B6 females. A control group of 3 NIMA^d-exposed newborns were separated from their mothers and foster-nursed by another BDF₁ female. Levels of H2D^d Mc (average of the organs tested) of the adult NIMA^d-exposed mice (6-8 weeks old) nursed by their own mothers (NIMA^d), or foster-nursed by B6 (F.N. by B6) or BDF₁ mice (F.N. by BDF₁) are shown. (C) Comparison of DTH responses to BDF₁ antigens between NIMA^d and NIMA^d foster-nursed by B6. For a positive control response to BDF₁ alloantigen, some NIPA^d control mice were injected intravenously with BDF₁ splenocytes 2 weeks before DTH assay and used as a source of sensitized splenocytes. The horizontal lines represent mean values of the observations.