An Immunologic Mode of Multigenerational Transmission Governs a Gut Treg Setpoint

Abstract

At the species level, immunity depends on the selection and transmission of protective components of the immune system. A microbe-induced population of RORγ-expressing regulatory T cells (Tregs) is essential in controlling gut inflammation. We uncovered a non-genetic, non-epigenetic, non-microbial mode of transmission of their homeostatic setpoint. RORγ⁺ Treg proportions varied between inbred mouse strains, a trait transmitted by the mother during a tight age window after birth but stable for life, resistant to many microbial or cellular perturbations, then further transferred by females for multiple generations. RORγ⁺ Treg proportions negatively correlated with IgA production and coating of gut commensals, traits also subject to maternal transmission, in an immunoglobulin- and RORγ⁺ Treg-dependent manner. We propose a model based on a double-negative feedback loop, vertically transmitted via the entero-mammary axis. This immunologic mode of multi-generational transmission may provide adaptability and modulate the genetic tuning of gut immune responses and inflammatory disease susceptibility.

Keywords: IgA; Rorγ+; Tregs; colonic Tregs; entero-mammary axis; maternal transmission.

Figure 1:. Proportions of colonic RORγ+ Tregs are different in inbred mouse strains and these proportions are determined maternally.

A. Proportion of colon RORγ+ Tregs in different inbred mice, gated as shown in left. B. Proportions of colon RORγ+ Tregs in GF B6 or Balb/c mice mono-colonized with indicated microbes. (throughout t.test p.value, **p<0.01, *** p<0.001, **** p<10⁻⁴). C. Representative pedigree chart showing colon RORγ+ Treg proportions (color-coded as indicated; sex denoted by shape) in F1 offspring of B6 and Balb/c mothers (for space, only 3 representative F1 are shown for each of 14 breeding pairs, all values summarized). D. Proportions of RORγ+ Tregs in (B6 x Balb/c) F1 offspring at different ages. E. Pedigree chart of colon RORγ+ Tregs (color-code and sex as in C) in F2 mice resulting from crossing (B6 × Balb/c) F1 females to (B6 × Balb/c) F1 males as shown. F. Representative pedigree chart of multiple generation backcross against B6 (left) or Balb/c (right) males, females for breeding picked randomly. Representative colonic RORγ+ Tregs at 6 weeks of age in non-bred littermates are shown (unknown in breeder females). Data representative of >3 independent experiments, bars in plots indicate mean

Figure 2:. RORγ+ Treg proportions are maternally imparted during an early neonatal window, and the learned phenotype is transferable.

A. Proportions of RORγ+ Tregs at 6 weeks of age in B6 or Balb/c mice fostered at birth by Balb/c or B6 mothers, compiled from 6 different litters ****t.test p<10⁻⁴. Shading indicates the usual range of RORγ+ Tregs in B6 and Balb/c mice. B. Top: experimental schematic. B6 or Balb/c females were fostered at birth by a B6 or Balb/c mother, and later crossed to a syngeneic male. Colon RORγ+ Tregs were quantitated in their offspring at 6 weeks of age. Data compiled from 3 litters, as indicated. C. B6 mice were fostered by Balb/c mothers (left) or the reverse (right) starting at birth or at indicated ages, and their RORγ+ Tregs quantitated at 6 weeks of age. Shading as above. Data representative of >3 independent experiments, bars in plots indicate mean

Figure 3:. Impact and stability of the maternally-transmitted RORγ+ Treg setpoint

A–D. B6 mice fostered at birth by B6 or Balb/c mothers were infected with 10⁹ cfu of C. rodentium at 6 weeks of age. A. Intestine length after 2 weeks (t test ***p < 0.001); B. C. rodentium counts in stool over the course of infection (***t-test p<0.001, mean±SD) and, C. in spleen on day 15 (Mann-Whitney ** p<0.01); D. Flow cytometry plots of IFNγ− and IL17-producing colonic CD4⁺ T cells, quantitated at right (t.test p.values). E. B6 Foxp3^DTR mice were fostered by B6 or Balb/c mothers at birth; colon RORγ+ Tregs were quantitated at 6 weeks of age (Pre), or after two doses of DT, or after 4 weeks recovery (t test *p < 0.05, **p < 0.01). F. B6 and Balb/c mice were fostered by B6 or Balb/c mothers at birth; colon RORγ+ Tregs were quantitated at 6 weeks of age, or after antibiotic treatment (VMNA) for 3 weeks, or after 4 weeks’ recovery. G. B6 mice fostered by B6 or Balb/c mothers at birth were treated with 2.5% DSS in drinking water for 6 days at 7 weeks, and colon RORγ+ Tregs analyzed 10 days post treatment. H. F1 females born to B6 or Balb/c mothers were treated with 4% DSS in drinking water for 6 days at 7 weeks and RORγ+ Tregs analyzed 10 days post treatment. I. Some of their similarly treated littermates were then bred to naive F1 males, and RORγ+ Tregs analyzed in their 6-week-old progeny (t test ****p < 10⁻⁴). Data representative of >3 independent experiments, bars in plots indicate mean.

Figure 4:. Maternal transfer of RORγ+ Treg proportions is not dependent on transmission of specific microbial taxa.

A. Relative frequencies of TCR Vβ usage in splenic CD4+ T cells from B6 or Balb/c mice fostered by Balb/c or B6 mothers at birth. B. Proportions of RORγ+ Tregs in 6 week-old B6 and Balb/c mice which received, as neonates (housed with birth mother), milk or stool from Balb/c and B6 females in the first three days of life. C. RORγ+ Tregs in 6 week-old B6 and Balb/c mice co-housed at 28 days of age for 2 weeks (left), or on the day of birth with adult virgin B6 or Balb/c females (right). D. RORγ+ Tregs in GF mice colonized with stool from adult B6, CBA/J, or Balb/c SPF mice for 2 weeks. E. Neonate microbiota: relative abundance of bacterial species in stool from 3-day old mice (metagenomic analysis). Mice were B6 or Balb/c pups cross-fostered onto B6 or Balb/c mothers, or F1s from B6 or Balb/c mothers. F. Adult microbiota: bacterial population analysis (16S rDNA) on 24+24 6-week old F1 mice from B6 or Balb/c mothers (each from 12 breeding cages). Frequencies for the 545 OTUs with frequencies >10⁻⁵ in at least 5 sample are shown, ordered by the mean differential representation in offspring of B6 or Balb/c mothers. OTUs with KS.test for B6 vs Balb/c origin p<10⁻⁵ marked at top). G. Four representative OTUs from F. Frequency in stool of F1 mice from B6 or Balb/c mothers plotted against RORγ+ Treg proportions, each dot an individual mouse. KS.test for B6 vs Balb/c origin is shown. H. RORγ+ Treg proportions after reciprocal fostering at birth, where pregnant foster mothers were treated with broad-spectrum antibiotics (VMNA, 2–5 days before delivery) ****t-test p<10⁻⁴. Data representative of >3 independent experiments, bars in plots indicate mean

Figure 5:. Immuno-profiling reveals that F1 mice born to B6 and Balb/c mothers display differences in their T cell activation states and IgA levels.

A. Frequency of conventional T cells expressing the activation markers CD69 and Sca-1 (left), and of RORγ+ or Helios+ Treg cells expressing Sca-1 (right) in adult F1 mice born to B6 or Balb/c mothers. B. Representative flow cytometry plots and quantification of colonic IgA+ B220- plasma cells in adult F1 mice (throughout t test, **p < 0.01, ****p < 10⁻⁴). C. Total IgA (ELISA) in serum (top) and stool (bottom) of 6-week old B6 mice, Balb/c mice, and F1 mice born to B6 or Balb/c mothers. D. Representative flow cytometry plots (top) and quantification (bottom left) of IgA-coated bacteria in stool of adult F1 mice born to B6 or Balb/c mothers, and correlation with colonic RORγ+ Tregs (bottom right). E. Correlation plot between colonic RORγ+ Tregs and IgA-coated bacteria in stool of adult F1 mice born to CBA or NOD mothers. F. Proportions of IgA coated bacteria in 3 day old F1 mice born to B6 or Balb/c mothers (left), and in 3-day-old B6 or Balb/c pups fostered by Balb/c or B6 mothers at birth (right). G. Total IgA (ELISA) in milk from B6 or Balb/c mothers at day 3 of lactation. H. Proportion of IgA-coated bacteria in representative mice from the backcross pedigrees of Fig. 1F (color-coded as indicated; sex denoted by shape). Data representative of >3 independent experiments, bars in plots indicate mean

Figure 6:. Expansion and migration of IgA+ plasma cells

A. Number of IgA+ plasma cells in the colon, small intestine and mammary gland of B6 mothers during late gestation (E18.5) or early days of lactation (L1–L5), vs virgin controls (V). B. Experimental design: intestines of Kaede+ mice were photoconverted (green to red) by illumination after laparotomy on day L1, and migration of cells to the mammary gland examined 48h later (L3). C. Representative plots of mammary gland plasma cells at L3 for Kaede green (baseline) vs red (result of illumination of the intestine at L1), in a non-photoconverted vs photoconverted mouse. D. Numbers of migrated Kaede red plasma cells of gut origin, 48 hrs after intestinal illumination, in the mammary gland, spleen or bone marrow of L3 females. E. Proportion of Kaede red plasma cells of gut origin in the mammary gland of L3 F1 females from B6 or Balb/c mothers. F. Total number of IgA+ plasma cells in the mammary glands at L3 of F1 females from B6 or Balb/c mothers. G. Total IgA (ELISA) in milk at L3 of F1 females from B6 or Balb/c mothers. Data representative of >3 independent experiments, bars in plots indicate mean

Figure 7:. Colonic RORγ+ Tregs and secretory IgA regulate each other in a double negative feedback loop.

A. Proposed model: Mothers transfer variable amounts of IgA to their offspring, leading to differences in IgA coating of gut microbes in neonates, which condition RORγ+ Treg proportions in adults, which in turn regulate levels of intestinal IgA+ plasma cells. In female offspring, the resulting IgA differences are reflected in their milk, thus allowing this phenotype to be transferred through multiple generations. B. Correlation between proportions of colon RORγ+ Tregs and IgA-coated bacteria in stool of B6 mice fostered by Balb/c or Balb/c.Jh−/− mothers at birth. C. Correlation between proportions of colon RORγ+ Tregs and IgA-coated bacteria in stool of (B6 x Balb/c)F1 mice born of B6, Balb/c or Balb/c.Jh−/− mothers. D. Proportions of colon RORγ+ Tregs in stool of (B6 x Balb/c)F1 mice born of B6, Balb/c or Balb/c.Iga−/− mothers. E. Proportions of IgA-coated bacteria in stool of Treg-specific conditional knockout mice deficient in Rorc (encodes RORγ) or Ikzf2 (encodes Helios) and their littermate controls (t test, ***p < 0.001). Data representative of >3 independent experiments, bars in plots indicate mean.