Regulation of AID expression in the immune response

Abstract

The B cell-specific enzyme activation-induced cytidine deaminase (AID) has been shown to be essential for isotype switching and affinity maturation of antibody genes during the immune response. Conversely, AID activity has also been linked to autoimmunity and tumorigenesis. Determining how AID expression is regulated in vivo is therefore central to understanding its role in health and disease. Here we use phylogenetic footprinting and high-resolution histone acetylation mapping to accurately demarcate AID gene regulatory boundaries. Based on this strategy, we identify a novel, positive regulatory element required for AID transcription. Furthermore, we generate two AID indicator mouse strains using bacterial artificial chromosomes that faithfully recapitulate endogenous AID expression. The first strain uses a green fluorescent protein reporter to identify B cells that actively express AID during the immune response. In the second strain, AID transcription affects the permanent expression of a yellow fluorescent protein reporter in post-germinal center and terminally differentiated lymphocytes. We demonstrate the usefulness of these novel strains by resolving recent contradictory observations on AID expression during B cell ontogeny.

Figure 1.

Phylogenetic analysis of Aicda. AID locus in human chromosome 12. Homology “peaks and valleys” graph showing percent conservation between human, dog, and mouse AID genomic loci. The nucleotide sequence is plotted on the horizontal axis. The vertical axis depicts the percent identity between the three species. Only conserved regions containing at least one peak with percent identity >75% within a shifting window of 100 bp in all three species are represented. Red peaks correspond to gene exons, whereas black peaks denote CNSs (I–XI). Genes and their orientation are represented by their RefSeq name and by arrows, respectively. As in the human, the mouse and dog Mfap5 and Kiaa1238 orthologs are located 5′ of Aicda and form a synteny group within the 1 MB interval analyzed. Conversely, the mouse and dog Clecsf9 and Clecsf8 genes are located beyond the 1 MB sequence analyzed here, and no homology is therefore assigned to these genes. Alignment was obtained using Family Relations (http://family.caltech.edu/) and VISTA software (http://www-gsd.lbl.gov/vista/index.shtml).

Figure 2.

H3 acetylation status at the AID locus. (A) Percent homology identity between the dog, human, and mouse genomes as a function of the mouse AID locus in chromosome 6. Homology peaks and genes are depicted as in Fig. 1. Black dots represent the position of quantitative PCR primers used to assess H3 acetylation by CHIP. AID exons are marked with yellow boxes. E, EcoRV; S, SalI; X, XbaI. (B) Absolute percentage of histone H3 acetylation from resting (blue circles) or LPS-activated (red squares) B cells. (C) DNaseI hypersensitivity at CNS VII and X. Nuclei prepared from LPS-activated mouse B cells were untreated (lane 1) or treated with increased concentrations of DNaseI (lanes 2–6). DNA was extracted, digested, and analyzed by Southern blot hybridization using VII- or X-specific probes (PVII and PX, respectively). Horizontal arrows indicate DNaseI hypersensitive sites that correlate with CNS-X. (D) Detailed map of mouse CNS X. Elements A and B (red boxes) depict two highly conserved sequences between mouse and humans that correlate with the Southern blot bands obtained by DNaseI hypersensitivity. Putative transcription factor binding sites in elements A and B, predicted using VISTA software, are shown.

Figure 3.

AID expression in AID-GFP and AID-GFP-ΔX transgenic mice. (A) Molecular schemes of AID-GFP and AID-GFP-ΔX transgenes. The position of CNS X is indicated. (B) FACS analysis of AID expression in wild-type (left) and transgenic mice (middle and right) immunized with NP-CGG and CFA. Cells were stained with CD95-PE and gated on B220 (PercP-Cy5.5). Apoptotic cells were excluded by To-Pro-3 or DAPI staining. The GC nature of GFP⁺ cells was confirmed by GL7-APC staining (not depicted). (C) Endogenous AID (gray bars), AID-GFP (green bars), and AID-GFP-ΔX (red bars) transcripts were measured by quantitative PCR from resting, LPS, or LPS plus IL-4 stimulated B lymphocytes. The diagram on the left depicts primer positions.

Figure 4.

Dynamics of AID expression in T cell–dependent and –independent immune responses. (A) AID-GFP mice were immunized with NP-CGG, and CD4⁻CD8⁻IgD⁻PI (or DAPI)⁻ lymphocytes recognizing NP were subdivided by B220 expression (left) and assessed for CD138 and AID-GFP expression 7–10 d after immunization (middle and right). Antibodies were IgD-biotin/Streptavidin-PE-Cy5, CD4-PE-Cy5, CD8-PE-Cy5, NP-APC, CD95-PE-Cy7, B220-APC-Cy7, and CD138-PE. (B) Left: NP staining of blood lymphocytes (IgM⁻IgD⁻Gr-1⁻F4/80⁻) recovered from unimmunized (day 0) and NP-CGG–immunized (day 7) wild-type mice. Staining was To-Pro-3, F4/80-Alexa 647, GR1-Alexa 647, IgM-APC, IgD-Alexa 647 (Dump channel) and NP-PE, IgG1-Biotin-Streptavidin-PercPCy5.5, and B220-FITC. Right: Analysis of AID transcription by SC-RT-PCR. NP⁺ cells were recovered by cell sorting from blood (72 cells analyzed) or lymph node GCs (72 cells) isolated from wild-type or AID^−/− mice. (C) Left: AID-GFP animals carrying the pre-recombined B1-8^hi heavy chain gene were immunized with NP-haptenated Ficoll and splenic B cells analyzed 5 d after immunization. AID-GFP expression in resting (B220⁺CD138⁻), T cell–independent GC (B220⁺CD138^low), plasmablasts (B220⁺CD138^high), and plasma cells (B220⁻CD138^high) is indicated with histograms. Bottom histograms show CD95 expression in AID-GFP⁺ cells from population 3 or population 4. Right: Analysis of AID transcription by SC-RT-PCR in the indicated populations isolated from B1-8^hi mice.

Figure 5.

Permanent labeling of B cells expressing AID. (A) Upper schematics: molecular maps of the AID-Cre (top) and Rosa-Neo-EYFP (bottom) transgenes (S, SalI; P, PacI; E, EcoRI). Lower Southern blot: three independent AID-Cre transgenic lines were generated, with copy number ranging between 1 to 5 copies. (B) FACS pseudocolor plots showing CD95-PE and YFP expression in B cells (gated on B220 (PercP-Cy5.5)) from AID-Cre-YFP compound mice. Left panel: axillary lymph node cells from nonimmunized mice; right panel: PPs. (C) Immunofluorescence analysis of gut lamina propria CD138⁺ plasma cells from AID-Cre-YFP mice. Scale bar = 20μm. (D) Mutation analysis of the J_H4 intron from CD95^lowYFP⁺, CD95^highGFP⁺, CD95^highGL7⁺ lymph node cell sorted from immunized AID-Cre-YFP, AID-GFP⁺, and AID^−/− mice, respectively. Segments in pie charts are sized in proportion to the number of sequences carrying mutations as indicated. The number of sequences analyzed is given in the center of the pie charts. Mutation frequency and P values are indicated below. Statistical significance was assessed by a two-tailed t test assuming unequal variance and comparing values to mutation background determined in AID^−/− sequences.

Figure 6.

AID expression in the GALT. AID expression in IgM (APC) or IgA (biotin/Streptavidin PercP-Cy5.5) B lymphocytes isolated from PPs (A) and lamina propria (LP; B) from AID-GFP mice (B220 was APC-Cy7 conjugated). Histograms show CD95 (PE) expression in the indicated populations. (C) Left micrographs: PNA staining of mesenteric lymph nodes (MLNs) isolated from OcaB^+/+ and OcaB^−/− mice (bar, 150 μm). Right density plot shows AID-GFP expression of IgA⁺ cells isolated from the lamina propria of AID-GFP-OcaB^−/− mice. Bottom gel shows analysis of AID expression by SC-RT-PCR in the indicated B cell populations recovered from OcaB^−/− mice as indicated.

Figure 7.

AID expression in immature, bone marrow B lymphocytes. (A) AID expression determined by flow cytometry from bone marrow B220^lowIgM⁻ pro–/pre–B (population 1), transitional B220^lowIgM⁺ (population 2), and recirculating B220^highIgM⁺ (population 3) B lymphocytes obtained from AID-GFP (left histograms) or AID-Cre-YFP (right histograms) mice. All immature B cells were also CD93⁺ (not depicted). Antibodies were IgM-APC, B220-PercP-Cy5.5, CD93-biotin/Streptavidin-PE, and To-pro-3. (B) Real-time PCR analysis of AID transcripts in pro–, pre–, transitional, mature, and PP B lymphocytes cell sorted from wild-type mice. For comparative purposes, cDNA synthesized from GL7⁺CD95^high GC B cells was serially diluted, as indicated below each column. cDNA was amplified using PCR primers annealing to AID exons 2 and 3. All B cell populations were sorted to >95% purity before RNA extraction and cDNA synthesis. Vertical bars represent the standard deviation of three independent experiments. (C) Bone marrow cells from AID-Cre-YFP mice were infected with Ab-MLV in vitro. The appearance of YFP⁺ cells was monitored in the infected cultures as a function of time (panels 1–4; only days 1 and 7 after infection are shown). The images are representative of three independent experiments. Panel 5 represents bone marrow B220⁺ cells isolated from AID-Cre-YFP mice and cultured for 7 d with IL-7 and in the presence of S17 cells. The following antibodies were used: CD43-PE, IgM-APC, B220-PercP-Cy5.5, and DAPI. Bar graph depicts the relative AID expression assessed by real-time PCR in Ab-MLV cultures.