Impaired Receptor Editing in the Primary B Cell Repertoire of BASH-Deficient Mice

Experimental mice

BASH-deficient mice with a C57BL/6 background were described previously (42). IgH^3H9/+ and Igκ^Vκ4/+ mice were reported previously (13, 15). These mice were crossed to obtain the mice used in this study. Mice were analyzed at 6–12 wk of age.

Flow cytometry

Cells were stained on ice with the following Abs: PE-conjugated anti-mouse B220/CD45R (RA3-6B2; BD Pharmingen, San Diego, CA); FITC-goat anti-mouse κ, FITC- or PE-goat anti-mouse μH, and PE-goat anti-mouse λ (Southern Biotechnology Associates, Birmingham, AL); biotin-goat anti-mouse μH (Cappel, Durham, NC), followed with streptavidin-CyChrome (BD Pharmingen). The cells within the lymphocyte gate defined by light scatters were analyzed through FACSort with CellQuest software (BD Biosciences, Mountain View, CA). Cell sorting was performed by FACSVantage (BD Biosciences) or MACS (Miltenyi Biotec, Auburn, CA) according to the standard procedures.

PCR analysis of recombination sequence (RS) and Ig rearrangements

Genomic DNAs from sorted cells were used as a template for PCR. Equality of the input genomic DNA amount was roughly evaluated by gel electrophoresis in every assay. RS recombination was detected by amplification of a ∼1500-bp product with primers Jκ intron (a in Fig. 1⇓A, 5′-CTGACT GCAGGTAGCGTGGTCTTCTAG-3′) and RS-637 (c in Fig. 1⇓A, 5′-TTGACTGTTTGCTACTTCAGCTCACTG-3′). The 3′ region of RSS (294 bp) was amplified with primers RS-341 (b in Fig. 1⇓A, 5′-CAGTTGAGCTCAGATTTGAGCCCTAATG-3′) and the RS-637. Both amplifications were performed as follows: 94°C for 1 min, 58°C for 1 min, and 72°C for 1 min for 25 cycles. Products were separated by electrophoresis in 1% agarose gel, transferred to nylon membrane, and hybridized with ³²P-labeled probes at 42°C. A DNA fragment amplified by PCR using the primers RS-341 and RS-637 was used as a probe. After washing the membrane, the signals were visualized by exposure to radiographic film (BioMax; Kodak, Rochester, NY). Igκ gene rearrangements were detected as described elsewhere (44), with a modified amplification protocol (95°C for 30 s, 60°C for 1 min, and 72°C for 1 min for 25 cycles). In Fig. 1⇓C (middle), a subset of Vκ genes, not including the Vκ4 gene, was amplified using the primer Vκ1 (d in Fig. 1⇓A) (16). The products were hybridized with the Jκ1-5 probe generated by PCR with tail DNA as a template and the following primers: Jκ1-5′ (5′-TGGACGTTCGGTGGAGGCAC-3′) and Jκ5-as (5′-CTAACATGAAAACCTGTGTCTTACACA-3′). In Fig. 1⇓C (top), the targeted Vκ locus was amplified with the primers Vκ4 (e in Fig. 1⇓A, 5′-CCCATTCACGTTCGGCTCGGGG-3′) and 3′ IRS (g in Fig. 1⇓A, 5′-ACACTGGATAAAGCAGTTTATGCCC-3′). PCR data shown here are the representatives of at least two assays using more than five individuals of each genotype in total, which showed essentially the same results.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Requirement of BASH for receptor editing in anti-DNA Ig knock-in mice. A, Schematic representation of RS recombination and Vκ replacement at Igκ locus in the normal and Vκ4 knock-in alleles. Primers (a–g) for PCR are indicated by arrowheads. iE, intron enhancer; 3′E, 3′ enhancer. B, PCR analysis of RS and VκJκ recombination. Genomic DNA from equal number of sorted immature B cells (IgM^highIgD^low, purity was >90%) from spleens of the indicated mice were subjected to the PCR with primers a and c in Fig. 1A, followed by the hybridization with the probe shown in Fig. 1A for the RS recombination (top) or to the PCR analysis for VκJκ recombination (middle) performed as described previously (44 ). 3′ region of the RSS was amplified using primers b and c (Fig. 1A) to control genomic DNA input (bottom). C, PCR detection of an unedited Vκ4 knock-in locus using the primers e and g (Fig. 1A) in κ⁺ splenic B cells of the indicated mice (top). Recombination of a subset of Vκ genes, not including the knock-in Vκ4 gene, and Jκs was also amplified using primers d (Vκ1) and f shown in Fig. 1A (middle). Amplification of Vκ-Jκ1/Jκ2 recombination was inefficient probably due to the long distance to be amplified. 3′ region of the RSS was amplified as in B to control genomic DNA input (bottom). D, Bone marrow cells from the indicated mice were stained with anti-κ, anti-λ, and anti-μH Abs and analyzed by flow cytometry. To exclude pre-BCR⁺ pre-B cells accumulating in the bone marrow of BASH^−/− mice (37 ,42 ), only μH^high B cells were electrically gated and analyzed for κ and λ expression. Numbers represent percentages of the μH^high B cells.

Nucleotide sequence analysis of VκJκ joins

To amplify the VκJκ rearrangement upstream of RS recombination, a degenerated primer for Vκ framework region 3 (h in Fig. 2⇓E, 5′-GGCTGC AGSTTCAGTGGCAGTGGRTCWGGRAC-3′) (45) and a primer just 3′ of the RSS (i in Fig. 2⇓E: 5′-ACATGGAAGTTTTCCCGGGAGAATATG-3′) were used in the following amplification protocol: 94°C for 1 min, 58°C for 1 min, and 72°C for 1 min for 35 cycles. The major products (∼1450 bp) corresponding to the alleles that underwent Vκ–Jκ5 and RS recombinations at the same time were gel isolated, cloned into the pGEM-T easy vector (Promega, Madison, WI), and bacterial colonies were screened by hybridization using the IVS probe (45). Thus, selected clones were sequenced with the dideoxy termination method (Beckman Coulter, Fullerton, CA) using the Jκ5-as primer (see above).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Receptor editing in the normal B cell repertoire in wild-type and BASH-deficient mice. A–C, Igλ expression (each right panel) in bone marrow (A) and spleen (B) immature B cell population (μH^highB220^middle and μH^highδH^low, respectively, as indicated by the window in each left panel). Numbers indicate Igλ⁺ cell percentages of the immature B cells, which are summarized in C (n = 6 for each genotype). D, PCR analysis for RS recombination frequency. Genomic DNAs (1 ng/sample) from splenic immature B cells (IgM^high IgD^low) of the indicated genotypes and normal thymocytes (T cell) were used for the PCR with the primers a and c shown in Fig. 1⇑A for the RS recombination (top) or with the primers b and c (Fig. 1⇑A) to control genomic DNA input (bottom). E, Frequency of in-frame VκJκ joins upstream of RS recombination. Genomic DNAs from Igλ⁺ B cells sorted from spleens by MACS were subjected to PCR with a universal Vκ primer and a primer 3′ of RSS (h and i in the left scheme, respectively). VκJκ joins of the products were clonally sequenced. Each bar represents the percentage of the number (n) of clones sequenced. Statistics was determined using the χ² test.

Detection of Vλ1–Jλ1 excision product

Eight- to 10 wk-old mice were gamma irradiated (5 Gy) and sacrificed after 14 days when reconstitution of the bone marrow by newly generated immature B cells reached maximum (Ref.22 and our unpublished data). The bone marrow cells from three mice of each genotype were combined and first bound on ice with rat anti-mouse Igκ mAb (BioSource International, Camarillo, CA) as a stimulation reagent (as indicated) and then with goat anti-mouse κ polyclonal Ab conjugated with FITC (Southern Biotechnology Associates). Igκ⁺ cells were enriched by MACS using anti-FITC MicroBeads (Miltenyi Biotec) and directly sampled or cultured in RPMI1640 with 10% FCS for 48 h at 1 × 10⁵/100 μl per well along with an equal number of bone marrow cells from RAG2^−/− mice. Just before the culture, the same anti-Igκ mAb (20 μg/ml) was supplemented in the culture (as indicated) to ensure the stimulation. After the culture, cells were lysed in 90 μl of water, incubated at 95°C for 10 min, followed by incubation with 1 mg/ml protease K (Merck, West Point, PA) at 55°C for 1 h and then at 95°C for 10 min. The lysates were used as templates for PCR (94°C for 1 min, 64°C for 1 min, and 72°C for 2 min for 27 cycles) with primers detecting Vλ1–Jλ1 excision product (P46 and P47) (6). Genomic DNA content in the lysates was standardized according to wild-type RAG2 gene dosage estimated by PCR (94°C for 30 s, 55°C for 30 s, and 72°C for 1 min for 27 cycles) with primers RAG2–5′ (5′-TGACATAGCCTGCTTATTGTCTCCT-3′) and RAG2–3′ (5′-CTTGTATAGATTAATAGTGGACCTT-3′). The products were verified by Southern blot hybridization with appropriate probes (6).

Single-cell PCR analysis

Single-cell PCR was performed as previously described (46) with a slight modification. Single cells were picked up by glass capillary under a microscope from a cell suspension of splenic B cells purified by anti-B220 Ab-conjugated magnetic beads for MACS. The single cells were directly transferred to tubes containing 25 μl of 1 mg/ml protease K and incubated for 1 h at 55°C, then for 10 min at 95°C to inactivate the protease K. This lysate was subjected to two rounds of PCR. The first round contained the following six primers: 3H9-5′ (5′-CTGTCAGGAACTGCAGGTAAG G-3′), 3H9-3′ (5′-CATAACATAGGAATATTTACTCCTCGC-3′), Vκ1, Vκ2, Vκ3 (16), and the Jκ5-as (see above). Five-microliter aliquots of the product were subjected to the second PCR to amplify the 3H9 gene and Vκ-Jκ rearrangements separately by the following primers: for 3H9, 3H9–5′ inner (5′-GCATTCAGTAGYTCCTGGATGAAC-3′) and 3H9–3′inner (5′-TCTTGCACAGAAGTAGACCGCAGA-3′); for Vκ-Jκ rearrangements, forward region 3 (see above) and Jκ1-3′-(5′-GCCACAGACATAGACAACGGAAGA-3′), Jκ2-3′ (5′-AATTTGCTAAATCCCTGAAATCTCCT-3′), Jκ4-3′ (5′-CACACAAGTTACCCAAACAGAACC-3′), or Jκ5-3′ (5′-GTGTACTTACGTTTCAGCTCCAGC-3′).

Immunization and anti-DNA Ab titration by ELISA

Mice (8–10 wk old) were immunized by an i.p. injection of 100 μg of an oligonucleotide-avidin complex prepared as follows: biotinylated oligonucleotide (5′-ACAAGCAGGGAGCAGATACTCGAGCGG-3′, a sequence derived from Rous sarcoma virus long terminal repeat) was mixed with avidin (from egg white; Sigma-Aldrich, St. Louis, MO) at a molecular ratio of 4:1 in PBS and either precipitated in alum or emulsified with CFA (H37 Ra; Difco, Detroit, MI). Pre- and postimmunization sera were analyzed for anti-dsDNA or anti-ssDNA IgG titers by ELISA. Plates were precoated for 1 h at room temperature with 0.001% protamine sulfate (Sigma-Aldrich), washed with water, and coated with calf thymus DNA (1 μg/well; Sigma-Aldrich) in 0.15 M NaCl/0.015 M sodium citrate (pH 8.0) by drying at 37°C. The dsDNA was prepared by shearing and the ssDNA by additional boiling. The plates were masked with 3% BSA in PBS and incubated with serially diluted sera. Bound Abs were detected with HRP-conjugated goat anti-mouse IgG (Southern Biotechnology Associates) and a tetramethylbenzidine peroxidase enzyme immunoassay substrate kit (Bio-Rad, Hercules, CA).

Impaired editing of anti-DNA Ag receptors in Ig knock-in mice lacking BASH

To examine whether BASH is necessary for BCR to signal receptor editing, we analyzed the BASH-deficient mice carrying a targeted IgH locus in which D_Q52 and all the J_H gene segments had been replaced by a rearranged V_H3H9 gene encoding an H chain of anti-dsDNA Ab (IgH^3H9/+) (13). In the IgH^3H9/+ mice, the 3H9 H chain forms dsDNA-binding receptors in combination with ∼60% of endogenous mouse κ-chains (47). Such receptors are edited mostly through repeated L chain gene rearrangements and partly through recombination between the V_H3H9 gene and the upstream endogenous V_H or D_H genes as previously shown by analyses of splenic B cell hybridomas (13, 48). The editing at the L chain gene loci often results in a deletion of the responsible κ gene through RS recombination, a recombination between an intron-recombining sequence (IRS) located in the Jκ-Cκ intron and a recombination signal sequence (RSS) located ∼25 kb downstream of the Cκ exon (Fig. 1⇑A) (16, 49). Therefore, the extent of the RS recombination was first assessed by PCR with genomic DNA from splenic immature B cells as a template. Robust amplification of RS recombination at κ loci was observed in splenic immature B cells from BASH^+/+IgH^3H9/+ mice, whereas essentially no RS recombination was detected in the same subset of cells from BASH^−/−IgH^3H9/+ mice (Fig. 1⇑B, top), as compared with the similar level of overall Vκ-Jκ recombination between the two genotypes (Fig. 1⇑B, middle). This indicates that the RS recombination due to the editing of DNA-binding BCR is strongly dependent on BASH.

To determine the extent of the V_H gene replacement, the integrity of the knocked-in V_H3H9 locus in single splenic B cells was analyzed by PCR. Vκ-Jκ-rearranged alleles were simultaneously amplified to identify the cells as B cells. Among the cells having undergone Vκ-Jκ rearrangement, the V_H3H9 gene was undetectable in 38% of the BASH^+/− B cells, whereas it was undetectable in only 13% of the BASH^−/− B cells (Table I⇓), suggesting that receptor editing through V_H gene replacement is also defective in BASH^−/− mice. These results combine to suggest that BASH is required for receptor editing caused by self-Ag recognition.

To confirm the latter conclusion, we further introduced a κ allele carrying targeted integration of a rearranged Vκ4Jκ4 gene (Vκ4) (15) into mice carrying V_H3H9 and BASH⁻ alleles. The Vκ4-encoded κ-chain forms a specific receptor for dsDNA in combination with the 3H9 H chain and is edited in essentially all B cells in mice (12, 14). In contrast to the innocuous V_H/Vκ gene knock-in mice deficient for BASH, in which the number of immature B cells in the bone marrow was recovered to the level of wild-type mice (43), BASH-deficient V_H3H9/Vκ4 knock-in mice failed to recover the B cell number (Table II⇓). This suggests that the majority of the dsDNA-binding B cells in the BASH-deficient knock-in mice were deleted in the bone marrow, possibly because of inefficient receptor editing. By contrast, in BASH-sufficient V_H3H9/Vκ4 knock-in mice, B cells were not rigorously deleted, but presumably edited as efficiently as previously reported (12, 13, 14, 15, 16, 17, 18). These assumptions were tested as follows.

In IgH^3H9/+Igκ^Vκ4/+ mice, it was reported that receptor editing at the Igκ locus includes a deletion of the inserted Vκ4 gene by a recombination between upstream Vκ segments and a downstream Jκ5 segment, the RS recombination on the targeted allele, and/or a recombination on the untargeted allele (14). Thus, we assessed the Vκ4 deletion and the RS recombination on the targeted allele simultaneously by PCR using a primer specific for the Vκ4 gene (e in Fig. 1⇑A) and another for the 3′ region of the IRS (3′ IRS, g in Fig. 1⇑A). Such recombination was greatly reduced; namely, far more intact Vκ4 knock-in locus was retained in BASH^−/−IgH^3H9/+Igκ^Vκ4/+ mice as compared with BASH^+/+IgH^3H9/+Igκ^Vκ4/+ mice (Fig. 1⇑C, top). This was confirmed by PCR using a primer specific for a neomycin-resistant gene located upstream of the Vκ4 gene and the 3′ IRS primer (data not shown). In addition, recombination of the endogenous Vκ and Jκ segments was less frequent in the former mice than in the latter (Fig. 1⇑C, middle).

It has been shown previously that, in anti-self IgH/κ transgenic/knock-in mice, receptor editing results in λ-chain gene rearrangement and appearance of λ⁺ cells, probably due to a failure to produce a functional and innocuous receptor despite repeated rearrangements of κ-chain loci (6, 12, 16, 20). In agreement with this finding, a markedly high proportion (22%) of the bone marrow B cells was λ⁺ in IgH^3H9/+Igκ^Vκ4/+BASH^+/+ mice. However, the appearance of λ⁺ B cells was suppressed to 5% in IgH^3H9/+Igκ^Vκ4/+BASH^−/− mice (Fig. 1⇑D). This indicates that editing of the dsDNA-binding BCR resulting in the replacement of κ-chain with λ was also hampered in BASH^−/− mice.

Taken together, in BASH^−/−IgH^3H9/+Igκ^Vκ4/+ mice, the dsDNA-binding BCR is not efficiently edited and the unedited B cells are considerably deleted but somewhat remain in the periphery, although direct measurement of the frequency of such cells was hampered by unavailability of a clonotypic Ab. Aside from this quantitative uncertainty, these results collectively indicate that the editing of the anti-dsDNA receptor is impaired in BASH-deficient mice and that BASH is crucial for BCR signaling that directs receptor editing.

Impaired receptor editing in the normal B cell repertoire of BASH-deficient mice

Although appearance of Igλ⁺ B cells and RS rearrangements have been considered as hallmarks of receptor editing in the IgH/κ transgenic/knock-in mouse systems, it remains unclear whether the generation of Igλ⁺ B cells and RS rearrangement in normal mice results from the editing of self-reactive BCR. Thus, it has not been proven whether receptor editing indeed operates in the development of normal Ig repertoire. Since receptor editing was impaired in the BASH^−/− mice with the knock-in Ig loci as so far described, we next analyzed BASH^−/− mice with unmanipulated Ig loci to determine whether such BASH-dependent receptor editing operates in mice with normal Ig repertoire. As known for most experimental mice, ∼5–6% of immature B cells in the bone marrow as well as in spleen were λ⁺ in BASH^+/+ and BASH^+/− mice (Fig. 2⇑, A–C). In BASH^−/− mice, the proportion of λ⁺ B cells within the comparable immature B cell population was decreased to 1.5 and 1.3% on average in the bone marrow and spleen, respectively. This suggests that BASH is necessary for the generation of a large part of λ-bearing B cells, possibly through receptor editing.

Next, we assessed by PCR the extent of RS recombination of Igκ loci in splenic immature B cells from mice of the three genotypes (Fig. 2⇑D). The frequency of RS recombination per genome was markedly reduced in BASH^−/− mice as compared with BASH^+/+ or BASH^+/− mice, suggesting that the RS recombination in the normal repertoire is induced largely through BASH-mediated BCR signaling. RS recombination could also happen to delete a nonfunctional Igκ allele. In such a case, a rearranged VκJκ sequence upstream of the RS recombination would be likely to have an out-of-frame joining, whereas it should be in-frame if RS recombination had inactivated a self-reactive κ-chain gene (45). To distinguish these possibilities for the residual RS recombination in BASH^−/− B cells, the VκJκ joins upstream of the RS recombination were amplified from splenic λ⁺ B cells in which the edited cells should be enriched, and their nucleotide sequences were determined as described previously (45). As shown in Fig. 2⇑E, the frequency of in-frame VκJκ joins was 38% of such VκJκ joins in the B cells from wild-type mice. In those from BASH^−/− mice, it significantly deceased (21%), suggesting RS recombination had inactivated a self-reactive κ-chain gene less frequently than in wild-type mice. The remaining in-frame joins may still include nonfunctional genes such as Vκ pseudogenes or those producing mispairing κ-chains. Taken together, these results indicate that receptor editing, and perhaps the editing of self-reactive BCR, is greatly impaired in a normal B cell repertoire in the absence of BASH. This provides the first genetic evidence, to our knowledge, that BCR-signaled receptor editing indeed operates in the development of the normal B cell repertoire of mice.

To determine a role for BASH in the BCR signaling that directs secondary recombination of L chain loci in vitro, a population enriched for immature κ⁺ B cells was obtained from the bone marrow of irradiated, self-reconstituted BASH^+/+ and BASH^−/− mice, and anti-κ Ab-induced Vλ-Jλ recombination was estimated by detecting excised circle DNA (λ circle) by PCR (6). Lymphocyte-free bone marrow cells from RAG2-deficient mice were added to the culture to prevent BCR-signaled apoptosis of the immature B cells (22). Before culture, the background level of λ circle could be detected in the immature B cells, possibly due to contamination of λ⁺ cells having just finished the λ rearrangement, and such cells appeared to be less in BASH^−/− immature B cells than in BASH^+/+ cells (Fig. 3⇓A, lanes 5 and 6) in line with the result shown in Fig. 2⇑, A and C. Anti-κ stimulation during the culture resulted in the increase of the λ circles in the BASH^+/+ cells as compared with unstimulated cells (Fig. 3⇓A, lanes 1 and 2), but not in the BASH^−/− cells (lanes 3 and 4). In the latter, the λ circles were rather decreased by the anti-κ stimulation, which might be due to a prolonged survival of stimulated κ⁺ B cells in this culture condition. Therefore, the 2-fold increase of λ circle in the BASH^+/+ cells in this data might be underestimated. As previously reported (43), the immature B cells differentiated into the mature stage (B220^high) during the culture without anti-κ Ab, with some retardation in BASH^−/− cells (Fig. 3⇓B, middle). Anti-κ stimulation suppressed the differentiation in BASH^+/+ cells, but not in BASH^−/− cells (Fig. 3⇓B, right). Thus, the BCR-signaled delay of differentiation at the immature B cell stage may correlate to the induction of λ recombination. These results indicate that BASH is necessary for BCR ligation-induced recombination of λ-chain gene in vitro and therefore suggest a pivotal role for BASH in the BCR-signaling pathway directing receptor editing through secondary L gene rearrangement.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Anti-κ Ab-induced λ rearrangement and maturation block in vitro are defective in BASH-deficient mice. A, Self-reconstituted bone marrow κ⁺ immature B cells of the indicated BASH genotypes were sampled before culture (0 h) or cultured with (+) or without (−) anti-mouse κ mAbs for 48 h in the presence of bone marrow cells from RAG2^−/− mice. The cell lysates were subjected to PCR for amplification of Vλ1–Jλ1 excision products (λ circle; top) or wild-type RAG2 gene (bottom). The latter does not amplify the targeted RAG2 gene in the RAG2^−/− mice, therefore it was used to normalize the number of input B cells in the samples. Bone marrow cells from normal C57BL/6 and RAG2^−/− mice were used as positive and negative controls for the PCRs, respectively. Numbers reflect the relative intensity of the hybridization bands representing λ circle DNA, which was normalized according to those representing RAG2 gene. B, After the same culture as in A, beside that five times more cells were used in the culture, the cells were stained with anti-B220 Ab and analyzed by flow cytometry. Numbers indicate percentages of the cells in the lymphocyte gate (middle and right panels). Before the culture, the sorted immature B cells were stained for κ L chain and B220 and analyzed by flow cytometry to show the similar homogeneity of the sorted samples (left panels).

Enhanced anti-DNA Ab production upon DNA immunization in BASH-deficient mice

Assuming that receptor editing, in cooperation with clonal deletion, is indeed responsible for purging self-reactivity from the repertoire of newly generated B cells, then supposedly remaining, unedited B cells in BASH^−/− mice would respond to self-Ags if they are functional and T cell help is provided appropriately. To verify whether such self-reactive B cells are present in the periphery of BASH^−/− mice, we assessed anti-DNA Ab production in the mice upon immunization of DNA. Since BASH^−/− mice were shown to be deficient in a T cell-independent immune response (38, 40), we prepared a novel T cell-dependent DNA Ag, biotinylated oligonucleotides coupled with avidin as a non-self carrier protein. This oligonucleotide itself was not mitogenic for B cells in vitro (data not shown). When this Ag was given in an alum-precipitated form, approximately four times as much anti-dsDNA IgG Ab was produced in BASH^−/− mice as in BASH^+/+ or BASH^+/− mice 3 wk after immunization, and this Ab remained at a higher level at a later time course in the former mice than in the latter (Fig. 4⇓A). The results for anti-ssDNA IgG Ab production were essentially the same (data not shown). When the Ag was given in CFA, a rapid and strong anti-dsDNA IgG response was observed only in BASH^−/− mice at 1 wk after immunization; however, after 2 wk, anti-dsDNA IgG was significantly produced also in wild-type mice, perhaps by low-affinity poly-reactive B cells (Fig. 4⇓B). In both experiments, anti-avidin IgG responses were equivalent among the mouse genotypes (Fig. 4⇓A and data not shown). These results are consistent with the proposition that DNA-binding BCR are not efficiently edited and functional B cells carrying such BCR remain, if mostly deleted, in the periphery of BASH^−/− mice. Interestingly, an anti-dsDNA IgG response did not recur in either of the genotypes upon secondary immunization at 7 wk in the former experiment, or 13 wk in the latter (data not shown), although a memory response to T-dependent Ag is not impaired in BASH^−/− mice (50). This suggests that strong tolerance to dsDNA was induced after the primary immune response by some unknown mechanism.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Excessive production of anti-dsDNA IgG Ab upon DNA immunization in BASH-deficient mice. A, Serum anti-dsDNA IgG (top) and anti-avidin IgG (bottom) titers 1 wk before (preimmune) or the indicated weeks after immunization with oligonucleotide-avidin complex precipitated in alum. One sample from a wild-type mouse showed high titer for both DNA and avidin for unknown reasons. B, Serum anti-dsDNA IgG titers 1 wk before (preimmune) or the indicated weeks after the immunization with oligonucleotide-avidin complex emulsified with CFA. A and B, Shown are the relative serum IgG concentrations of the individual mice of the indicated genotypes. They are expressed as relative to a value of a standard (set as 1000), namely, an autoimmune serum from a mice suffered from experimental graft-vs-host disease. The titer of a serum from a typical autoimmune NZB/W F₁ mouse was 1771 U when assayed in B.