IgD class switching is initiated by microbiota and limited to mucosa-associated lymphoid tissue in mice

Abstract

Class-switch recombination (CSR) alters the Ig isotype to diversify antibody effector functions. IgD CSR is a rare event, and its regulation is poorly understood. We report that deficiency of 53BP1, a DNA damage-response protein, caused age-dependent overproduction of secreted IgD resulting from increased IgD CSR exclusively within B cells of mucosa-associated lymphoid tissues. IgD overproduction was dependent on activation-induced cytidine deaminase, hematopoietic MyD88 expression, and an intact microbiome, against which circulating IgD, but not IgM, was reactive. IgD CSR occurred via both alternative nonhomologous end-joining and homologous recombination pathways. Microbiota-dependent IgD CSR also was detected in nasal-associated lymphoid tissue of WT mice. These results identify a pathway, present in WT mice and hyperactivated in 53BP1-deficient mice, by which microbiota signal via Toll-like receptors to elicit IgD CSR.

Keywords: 53BP1; IgD; Toll-like receptor; class-switch recombination; microbiota.

Fig. 1.

Age-dependent hyper-IgD syndrome caused by an ENU-induced Trp53bp1 mutation in mice. (A) MFI of surface IgD immunostaining (IgD MFI) on peripheral blood B cells from WT C57BL/6J mice (WT), or from three pedigrees of third generation (G3) descendants of a single ENU-mutagenized male mouse, with REF (^+/+), HET (+/lentil), or VAR (lentil/lentil) genotypes for Trp53bp1. Data were normalized to average IgD MFI of age matched C57BL/6 mice at the time of experiment. (B) Manhattan plot. –Log₁₀ P values plotted vs. the chromosomal positions of mutations identified in the three G1 founders of the affected pedigrees. (Insets) Representative flow cytometry analysis of IgD expression by peripheral blood B cells from a homozygous lentil mouse and a WT littermate. (C) IgD MFI on peripheral blood B cells from 12-wk-old Trp53bp1^−/− or Pla2g4b^−/− mice generated by the CRISPR/Cas9 system. (D) IgD MFI on peripheral blood B cells from mice of the indicated ages and genotypes. Cells were stained in the presence of serum. (E) IgD MFI on peripheral blood B cells from 12-wk-old Trp53bp1^−/− and WT littermates; cells were stained in the presence or absence of serum. (F) IgD MFI on WT peripheral blood B cells after immunostaining in the presence of serum from the indicated mouse strains. (Insets) Representative flow cytometry analysis of IgD expression under each serum condition. (G) Serum IgD concentration in mice of the indicated ages and genotypes. In A and C–G, data points represent individual mice. In C–G, P values were determined by Student’s t test, and unless indicated otherwise correspond to differences between the marked group and WT. Results are representative of two to three independent experiments with n = 5–23 mice/genotype or strain; error bars indicate SD.

Fig. S1.

Nature of the lentil mutation. (A) Diagram of the 1,969-aa mouse 53BP1 protein based on Uniprot annotations. The location of the lentil A317S variant is indicated in red. KBD, kinetochore-binding domain; OLIGO, oligomerization domain; GAR, glycine/arginine-rich region; UDR, ubiquitin-dependent recruitment motif; NLS, nuclear localization signal; BRCT, BRCA1 C terminus. The phosphoinositide 3-kinase-like kinase (PIKK) S/TQ phosphorylation sites are labeled (Ser6, Ser25, Ser29, Ser166, Ser176/Ser178, Thr302, Ser452, Ser831, and Ser1219), as is the site of ubiquitination (Lys1523). (B) RT-PCR analysis of Trp53bp1 splicing in C57BL/6J and lentil blood using primers spanning exons 6–9. Sequence analysis of the cDNA amplified from lentil blood demonstrated a 106-bp deletion within the 167-bp exon 8 (ENSMUST00000110648). (C) Diagram of the predicted truncated mutant 53BP1^lentil protein. The 106-bp deletion within exon 8 is predicted to cause a frameshift that creates a premature stop codon in exon 11 (aberrant amino acids after position 285, truncation after position 362). (D) Immunoblot showing 53BP1 expression in B cells isolated from lentil, Trp53bp1^−/−, and C57BL6J mice.

Fig. S2.

Trp53bp1 deficiency has no effect on surface IgD and IgM expression in B-cell subsets other than peripheral blood B cells. IgD MFI (A) and IgM MFI (B) on B-cell subsets in BM, spleen, and peritoneal cavity (PC) isolated from 12-wk-old Trp53bp1^+/− mice, Trp53bp1^−/− mice, and WT littermates. The following surface markers were used to identify B cells: pre-B (B220⁺IgD⁻IgM⁻), immature B (B220⁺IgD⁻IgM⁺), pro-B (B220⁺IgD^medIgM⁺), mature recirculating B (B220⁺IgD⁺IgM⁺), T1 (CD19⁺B220⁺CD93⁺IgM⁺CD23⁻), T2 (CD19⁺B220⁺CD93⁺IgM⁺CD23⁺), T3 (CD19⁺B220⁺CD93⁺IgM^lowCD23⁺), marginal zone B (CD19⁺CD21^highCD23^low), follicular B (CD19⁺CD21^lowCD23^high), B1 (CD19⁺B220^low), and B2 (CD19⁺B220^high) cells. (C) IgM MFI on peripheral blood B cells of mice of the indicated ages and genotypes. No significant difference was found among Trp53bp1^+/−, Trp53bp1^−/−, and WT littermates (A–C). (D) Serum concentrations of IgM, IgG1, IgG2b, IgA, and IgE in 12-wk-old Trp53bp1^−/− mice and WT littermates. Data points represent individual mice. P values were determined by Student’s t test. Results are representative of more than three independent experiments with at least n = 3 mice/genotype. Error bars indicate SD.

Fig. S3.

Recapitulation of hyper-IgD syndrome in BM chimeras. (A) IgD MFI on peripheral blood B cells of BM chimeras rescued from lethal irradiation (15 Gy for WT recipients and 9 Gy for Trp53bp1^−/− recipients). Cells were stained in the presence of serum. (B) IgD MFI on peripheral blood B cells from mixed BM chimeras measured at 14 wk after BM transplantation. Black dots represent IgD MFI on CD45.1⁺ WT cells, and white dots represent IgD MFI on CD45.2⁺ Trp53bp1^−/− cells. Lower-case labels (a–d) correspond to FACS plots with the same labels in C. (C) Representative flow cytometry analysis of surface IgD expression on B cells of WT (CD45.1⁺) or Trp53bp1^−/− (CD45.2⁺) origin in mixed BM chimeric mice. All mice used in the experiments demonstrated at least 95% hematopoietic engraftment. (D) Serum IgD concentration in the chimeras at 14 wk after BM transplantation. Data points represent individual mice. P values were determined by Student’s t test. In A, P values correspond to differences between the marked group and WT recipients of WT BM. Results are representative of two independent experiments with n = 3–5 mice/group. Error bars indicate SD.

Fig. 2.

Trp53bp1-deficient B cells of MALT undergo IgD CSR. (A, C, and F) IgD MFI on peripheral blood B cells from mice carrying homozygous ENU-induced mutations in Tnfsf13b, Cd40, or Dock8, or mice with null mutations of Aicda, Atm, Trp53bp1, Trp53bp1;Aicda, or Trp53bp1;Atm. REF, homozygous for the WT allele; VAR, homozygous for the mutant allele. Cells were stained in the presence of serum. (B, D, and G) Serum IgD concentration in 14-wk-old mice of the indicated strains or genotypes. Results are representative of two independent experiments, n = 2–13 mice/genotype or strain. (E) Southern blot analysis of S_µ-σ_δ junctions in B cells from BM, spleen, MLNs, bronchoalveolar lavage fluid (BAL), NALT (a subtype of MALT), and SMLNs of 14-wk-old Trp53bp1^−/− (KO), Trp53bp1^−/−;Aicda^−/− (DKO), and WT littermates. Results are representative of three independent experiments using pooled samples from mice of the indicated genotypes.

Fig. S4.

Normal I_μ-C_δ and membrane C_μ transcript levels in B cells of Trp53bp1^−/− mice and schematic overview of detection of S_μ-σ_δ junctions by Southern blot analysis. (A and B) Relative abundance of I_μ-C_δ (secretory form) transcripts (A) and I_μ-C_μ (membrane C_μ) transcripts (B) measured by quantitative RT-PCR in B cells isolated from BM, spleen, and peripheral blood of Trp53bp1^−/− and WT littermates. Red lines represent mean ± SEM levels of I_μ-C_δ and I_μ-C_μ mRNA (normalized to Pax5), expressed as arbitrary units. Data points represent individual mice. P values were determined using Student’s t test. Results are representative of three independent experiments with n = 6–9 mice/genotype. (C) Partial diagram of the Ighm exon structure before and after S_μ-σ_δ CSR. AID-initiated S_μ-σ_δ CSR events were detected using a nested touchdown PCR protocol as described previously (1, 7). Genomic DNA was extracted from B cells of various tissues, PCR-amplified, and hybridized with a 5′ C_δ probe. S_μ-σ_δ junctions of variable size (0.5–2 kb) were expected in the current experimental protocol. Arrows indicate primer binding sites for PCR reactions. The probe binding site is highlighted in red.

Fig. 3.

IgD produced by Trp53bp1-deficient B cells recognizes intestinal bacteria. (A) Binding of serum IgD, IgM, and IgA to intestinal bacteria. Serum from 12- to 14-wk-old Trp53bp1^−/−, Ighm^−/−, Trp53bp1^−/−;Ighm^−/−, and WT littermates was incubated in vitro with bacteria isolated from the gut of Ighm^−/− mice. Unbound antibodies were removed by extensive washing, and IgD-bound bacteria were stained and measured by flow cytometry. (B and C) IgD-coated (B) and IgA-coated (C) intestinal bacteria from 12- to 14-wk-old indicated mouse strains were analyzed by flow cytometry. Note that Trp53bp1^−/− mice produce serum IgA at concentrations lower than those in WT mice, but higher than those in Ighm^−/− mice. Data points represent individual mice. P values were determined by Student’s t test. Results are representative of two independent experiments with n = 4–11 mice/genotype or strain. Error bars indicate SD.

Fig. 4.

Hyper-IgD production in Trp53bp1^−/− mice requires microbiota. (A) Serum NP-specific IgG, IgM, and IgD titers measured at 14 d after immunization with NP-LPS. Igh^B1-8+ transgenic mice express a recombined variable region derived from an NP-binding antibody in place of the endogenous 3′ Igh-D element (DQ52) and the Igh-J elements (15). (B and E) Number of bacteria in stool samples freshly isolated from conventionally reared (CNV) mice, broad-spectrum antibiotic-treated (Abx) mice, and GF-Rag1^−/− recipients. Samples were diluted and plated on BHI blood agar plates. (C) Serum IgD concentrations in 10-wk-old CNV and Abx-treated mice of the indicated genotypes. (D) Serum IgD concentrations in CNV and GF-Rag1^−/− recipients before and after microbiota colonization. GF-Rag1^−/− recipients were colonized with normal intestinal bacteria at 8 wk after BM engraftment. (F and G) Fecal samples collected from the small intestine of Trp53bp1^+/− mice, Trp53bp1^−/− mice, and WT littermates from three independent litters were subjected to 16S rRNA metagenomic sequencing. (F) Heat map of fold differences in relative abundance of commensal bacteria. The top 20 commensal genera that differed among the indicated genotypes are presented. (G) Principal coordinates analysis plot of fecal microbiota. In A–E, P values were determined by Student’s t test. Results are representative of two independent experiments with n = 3–11 mice/genotype or strain. Error bars indicate SD.

Fig. S5.

Composition of microbiota in the colon of Trp53bp1^−/− mice. Fecal samples isolated from the colon of Trp53bp1^+/−, Trp53bp1^−/−, and WT littermates from three independent litters were subjected to 16S rRNA metagenomic sequencing. (A) Heat map of fold differences in relative abundance of commensal bacteria. The top 18 commensal genera that differed among the indicated genotypes are presented. (B) Principal coordinates analysis plot of fecal microbiota.

Fig. 5.

Hematopoietic-intrinsic MyD88 expression is required for IgD production by Trp53bp1^−/− B cells. (A) IgD MFI on peripheral blood B cells in BM chimeras rescued from lethal irradiation. WT or Trp53bp1^−/− BM was used to reconstitute WT, Tlr4^lps3/lps3, and Ticam1^lps2/lps2;Irak4^{otiose/otiose} mutant recipients. Chimerism was assessed using congenic CD45 markers. All mice used in the experiments demonstrated at least 95% hematopoietic engraftment. Cells were stained in the presence of serum. (B) Serum IgD concentration in the indicated chimeric mice at 14 wk after BM transplantation. (C) IgD MFI on peripheral blood B cells from mice of the indicated ages and genotypes. Cells were stained in the presence of serum. (D) Serum IgD concentration in 10-wk-old mice of the indicated genotypes. P values were determined by Student’s t test and correspond to differences between the marked group and WT recipients for a given time point. In A, asterisks for P values apply similarly to green, yellow, and red data points. In C, black asterisks for P values apply similarly to white and orange data points. Data points represent individual mice. Results are representative of two independent experiments with n = 3–10 mice/genotype or strain. Error bars indicate SD.

Fig. 6.

Requirement for microbiota in IgD-secreting B-cell development. (A) Representative flow cytometry analysis of IgD⁺IgM⁻ cells in NALT of CNV and GF C57BL/6J mice. Cells were gated first on IgM⁻ cells and then on IgD⁺ cells. The percentage of cells in each gate is indicated. (B) Percentage of IgD⁺IgM⁻ B cells isolated from NALT. (C) Serum IgD concentration in 12-wk-old mice. (D and E) Flow cytometry analysis of B220, CD19, CD23, CD138, BAFF-R, and TACI expression on IgD⁺IgM⁻ B cells from NALT of CNV and GF WT mice. Representative FACS plots are shown in (D). (F) GF-Rag1^−/− mice were engrafted with Trp53bp1^−/− or WT BM and 8 wk later colonized with normal intestinal bacteria; they were then conventionally housed for 5 mo (ex-GF-Rag1^−/− plus indicated BM). Southern blot analysis of S_µ-σ_δ junctions in B cells from MLNs, NALT, and SMLNs of the indicated mice. Data points represent individual mice (B, C, E). P values were determined by Student’s t test. Results are representative of three independent experiments with n = 3–14 mice/genotype or strain. Error bars indicate SEM.

Fig. S6.

Serum IgE concentration in GF-WT mice and chimeric GF-Rag1^−/− recipients before and after conventionalization. (A) Serum IgE concentration in CNV and GF WT mice. (B) Serum IgE concentration in BM chimeras (CNV, GF, or GF mice after conventionalization). GF-Rag1^−/− mice were engrafted with Trp53bp1^−/− or WT BM and 8 wk later colonized with normal intestinal bacteria. They were then conventionally housed for 5 mo (ex-GF-Rag1^−/− plus indicated BM). P values were determined by Student’s t test. n = 3–14 mice/genotype or strain. Error bars indicate SEM.