Diversity of gut microflora is required for the generation of B cell with regulatory properties in a skin graft model

Abstract

B cells have been reported to promote graft rejection through alloantibody production. However, there is growing evidence that B cells can contribute to the maintenance of tolerance. Here, we used a mouse model of MHC-class I mismatched skin transplantation to investigate the contribution of B cells to graft survival. We demonstrate that adoptive transfer of B cells prolongs skin graft survival but only when the B cells were isolated from mice housed in low sterility "conventional" (CV) facilities and not from mice housed in pathogen free facilities (SPF). However, prolongation of skin graft survival was lost when B cells were isolated from IL-10 deficient mice housed in CV facilities. The suppressive function of B cells isolated from mice housed in CV facilities correlated with an anti-inflammatory environment and with the presence of a different gut microflora compared to mice maintained in SPF facilities. Treatment of mice in the CV facility with antibiotics abrogated the regulatory capacity of B cells. Finally, we identified transitional B cells isolated from CV facilities as possessing the regulatory function. These findings demonstrate that B cells, and in particular transitional B cells, can promote prolongation of graft survival, a function dependent on licensing by gut microflora.

Figure 1. Adoptive transfer of B cells isolated from mice kept in CV facilities can prolong MHCI mismatched skin graft survival.

CD8 depleted B6 (H2-K^b) mice received dorsal skin grafts from B6-K^d (H2-K^d) mice. One day prior to skin grafts mice received i.v. injections of 1 × 10⁷ B cells that were FACS purified from naïve B6 mice. Controls received PBS. Plots show skin graft survival when graft and cell transfer are performed on mice kept in (a) SPF facilities (n ≥ 9/group) and (b) CV facilities. (n ≥ 11/group). (c) B cells from naïve B6 mice maintained in SPF facilities were treated (LPS treated) or not (Untreated) with 2 μg/mL LPS for 16 hrs. Then, 1 × 10⁷ B cells were adoptively transfer one day prior to skin grafts, controls received PBS. The mice were kept in SPF facilities (n ≥ 5/group). (d,e) Injections of 1 × 10⁷ B cells that were FACS purified from IL-10^−/− or WT naïve B6 mice kept in CV (d) or SPF (e) facilities. Controls received PBS. The adoptive transfer was made one day prior to skin grafts. Plots show skin graft survival when graft and cell transfer are performed on mice kept in CV (d) or SPF (e) facilities (n ≥ 6/group for each group). (f,g) Plots show skin graft survival when cell transfer of 10⁷ B cells from mice treated with anti-IL10R (aIL-10R) antibody or isotype control (ISO) is performed on mice kept in CV (f) or SPF (g) facilities (n ≥ 6/group for each group). Mean survival in days is denoted between parentheses. Statistics were calculated by log-rank (Mantel-Cox) test + Bonferonni correction for multiple comparisons. ^*p < 0.05, ^**p < 0.01.

Figure 2. Immunological differences between mice kept in SPF & CV facilities.

Spleens, lymph nodes & PPs were collected from B6 mice (6 weeks old) maintained in SPF and CV facilities, and phenotyped for B cell and T cell populations by using antibodies against CD4, CD8, CD25, CD19, GL7, CD38, CD95, CD44, & CD62L. (a) and (b) Histograms show mean percentage ± SEM CD4⁺, CD8⁺ and CD4⁺CD25⁺ T cells in total splenocytes or total lymph node cells, naïve (CD64L^hiCD44^-) and memory (CD62L^hiCD44⁺) CD4⁺ T cells as percentages of total CD4⁺ T cells, and naïve and memory CD8⁺ T cells as percentages of total CD8⁺ T cells in (a) Spleens and (b) Lymph nodes. (c) Histograms show the mean ± SEM of the percentages of GC B cells (GL7^high and CD95⁺, left graph) and absolute number (right graph) in the spleens, MLN and PPs of SPF and CV mice. (d) Immunofluorescence microscopy pictures of B220 and IgD stained frozen sections from spleens (top row) and PPs (bottom row) of mice kept in SPF and CV facilities. Pictures show B220 (green) and IgD (Red). Germinal centres are marked by arrows. (e) Histograms show mean ± SEM of IFNγ, TNFα, and IL-10 expression by splenic CD4⁺ T cells and B cells following 5 hrs in vitro stimulation with PMA & ionomycin. n = 3. Statistics were calculated by t test, ^*P < 0.05 & ^**P < 0.005.

Figure 3. Mice kept in CV facility have different composition of microbiota than mice housed in SPF facility.

Faecal samples were collected from B6 mice kept in SPF & CV facilities, or kept in CV facility and treated with antibiotics for 2 weeks. Different bacterial groups (Bac = Bacteroides and Prevotella, Erec = Lachnospiraceae, Bif = Bifidobacterium, Lab = Lactobacillus & Streptococcus) were identified by FISH-Flow. Histograms show mean ± SEM for (a) proportions of the dominant fecal microbiota in B6 mice maintained in SPF and CV facilities, and (b) ratio of total microbiota in antibiotic treated (ATB) and non-treated (CV) B6 mice maintained in CV facilities (the mean total bacteria in untreated mice is defined as 100%). (c) Pictures of caeca recovered from antibiotics treated and non-treated B6 mice maintained in CV facilities. (d) CD8 depleted B6 (H2-K^b) mice received dorsal skin grafts from B6-K^d (H2-K^d) mice. One day prior to skin grafts mice received i.v. Injections of 1 × 10⁷ B cells that were FACS purified from antibiotics (ATB) or control (CV) naïve B6 mice kept in CV facilities. Plots show skin graft survival when graft and cell transfer are performed on mice kept in SPF facilities. n ≥ 6/group for each group. Mean survival in days is denoted between parentheses. (e) Histograms show mean ± SEM of IFNγ, TNFα, and IL-10 expression by splenic CD4⁺ T cells and B cells from mice treated with antibiotics (ATB) or non-treated (CV) naïve B6 mice kept in CV facilities, 2 independent experiments with 10 mice per group. (f) B cells were isolated from spleens of B6 mice treated with antibiotics or not and maintained in CV facilities. B cells were co-cultured with CD4⁺ T cells and irradiated allo-DCs (25 CD4 T cells:25 B cells:1 allo-DC) for 48 hrs. Summary data showing percentage of TNFα supression by B cell from spleens of B6 mice antibiotics treated or non-treated and housed in CV facilities. Histograms display mean ± SEM, 2 independent experiments with 10 mice per group. Statistics were calculated by log-rank (Mantel-Cox) test + Bonferonni correction for multiple comparisons and t test, ^*P < 0.05, ^**P < 0.005 & ^***P < 0.001.

Figure 4. Adoptive transfer of T2 B cells isolated from mice kept in CV facilities can prolong MHC I mismatched skin graft survival.

CD8 depleted B6 (H2-K^b) mice received dorsal skin grafts from B6-K^d (H2-K^d) mice. One day prior to skin grafts mice received i.v. Injections of 1 × 10⁶ T2, T1, FO, or MZ B cells that were FACS purified from naïve B6 mice. Controls received PBS. Plots show skin graft survival when graft and cell transfer are performed on mice kept in (a) SPF facilities (n ≥ 6/group for each experiment) and (b) CV facilities. (n ≥ 8/group for each experiment). Mean survival in days is denoted between parentheses. Statistics were calculated by log-rank (Mantel-Cox) test + Bonferonni correction for multiple comparisons. (C–G) B cells were isolated from spleens of B6 mice maintained in SPF (c,d) or CV (E-G) facilities by magnetic sorting. B cell subsets were purified by FACS and co-cultured with negatively isolated CD4⁺ T cells and irradiated allo-DCs (25 CD4 T cells:25 B cells:1 allo-DC) for 48 hrs. PMA, Ionomycin and brefeldin A were added for the last 4 hours of culture. (c and e) Representative FACS plots of CD4⁺ T cell TNF-α expression, (d and f) summary data showing CD4⁺ T cells TNF-α expression, (g) summary data showing IL-10 expression by B cell subsets isolated from spleens of B6 mice housed in CV facility. Histograms display mean ± SEM, (n = 3). (h) Summary data showing IL-10 secretion in cultures of T cell alone and with each subset of B cells from mice treated with antibiotics (ATB) or saline solution (CV) and kept in CV facilities. Statistics were calculated by one-way ANOVA and Bonferroni post-test, ns- not significant, ^*P < 0.05, ^**P < 0.005, ^***P = 0.0005.