In response to encounter with self-Ag, autoreactive B cells may undergo secondary L chain gene rearrangement (receptor editing) and change the specificity of their Ag receptor. Knowing at what differentiative stage(s) developing B cells undergo receptor editing is important for understanding how self-reactive B cells are regulated. In this study, in mice with Ig transgenes coding for anti-self (DNA) Ab, we report dsDNA breaks indicative of ongoing secondary L chain rearrangement not only in bone marrow cells with a pre-B/B cell phenotype but also in immature/transitional splenic B cells with little or no surface IgM (sIgM−/low). L chain-edited transgenic B cells were detectable in spleen but not bone marrow and were still found to produce Ab specific for DNA (and apoptotic cells), albeit with lower affinity for DNA than the unedited transgenic Ab. We conclude that L chain editing in anti-DNA-transgenic B cells is not only ongoing in bone marrow but also in spleen. Indeed, transfer of sIgM−/low anti-DNA splenic B cells into SCID mice resulted in the appearance of a L chain editor (Vλx) in the serum of engrafted recipients. Finally, we also report evidence for ongoing L chain editing in sIgMlow transitional splenic B cells of wild-type mice.

Developing bone marrow (BM)4 B cells that are strongly autoreactive may avoid elimination by changing the specificity of their Ag receptor (1, 2, 3). This process is referred to as receptor editing and usually involves secondary L chain gene rearrangement and expression of a new or additional L chain (4, 5). L chain editing was first demonstrated in mice containing Ig transgenes (tgs) that code for Abs with anti-self specificity (1, 2, 3). Subsequent studies have indicated that L chain editing may occur frequently and help shape the Ab repertoire (6, 7).

L chain receptor editing appears to be initiated as soon as immature B cells encounter self-Ag. For example, in the anti-MHC class I-transgenic mouse model (8), developing B cells that encounter self-Ag show reduced expression of surface IgM (sIgM), elevated RAG expression, and secondary L chain gene rearrangement (1, 9, 10, 11). In the anti-hen egg lysozyme (HEL)/soluble HEL (sHEL)-transgenic mouse model (12), the level of IgM surface expression was shown to correlate inversely with the strength of signaling through the BCR, i.e., the stronger the self-Ag (sHEL) signal, the lower the abundance of sIgM (13). Moreover, cells with the lowest sIgM showed the highest level of secondary L chain rearrangement (13, 14). Even low surface expression of a non-self-reactive BCR has been reported to result in secondary L chain rearrangement (15). In a more recent study, Tze et al. (16) reported that interruption of basal signaling through the BCR results in apparent reversion of affected B cells to an earlier developmental stage and secondary L chain rearrangement. Reversion of immature B cells to an earlier stage during the course of normal B cell differentiation was actually suggested earlier by Mehr et al. (17) based on mathematical modeling of the kinetics of developing B cell subsets in BM.

The above findings have led to the suggestion that down-regulation of sIgM on immature B cells may directly trigger secondary L chain rearrangement (15, 16). Accordingly, one would predict that dsDNA breaks indicative of ongoing secondary L chain gene rearrangement would be present in sIgM−/low autoreactive B cells. Indeed, using a transgenic mouse model (56RVκ8 mice) with tgs coding for anti-DNA Ab (4, 18), we demonstrate in this report that sIgM−/low anti-DNA B cells in BM and spleen (SPL) of 56RVκ8 mice contain dsDNA breaks at their wild-type (wt) κ and λ L chain loci. Splenic sIgM−/low anti-DNA B cells with dsDNA breaks included those with an immature/transitional T3 cell surface phenotype (B220+CD23highCD93+sIgM−/lowsIgD+) (19) and a T3-like (T3′) phenotype (B220+CD23−/lowCD93+sIgM−/lowsIgD+). T3/T3′ splenic B cells were greatly overrepresented in 56RVκ8 mice as these mice contained 3- to 5-fold more such cells than nontransgenic control mice. The observed dsDNA breaks in anti-DNA splenic B cells with a T3/T3′ phenotype suggested that L chain editing is still ongoing outside of the BM relatively late in B cell differentiation. Consistent with such editing, we found that engraftment of SCID mice with T3/T3′ splenic B cells from 56RVκ8 mice along with T cells from mice lacking B cells (JH−/− mice) resulted in the appearance of a L chain editor (Vλx) in the serum of engrafted recipients. Importantly, T3/T3′ splenic B cells from nontransgenic wt mice were also found to contain dsDNA breaks indicative of ongoing L chain rearrangement. This finding is of particular interest because T3 splenic B cells from wt mice have been reported to represent anergic self-reactive B cells unable to give rise to mature B cells (20, 21).

C.B-17 scid/+ mice hemizygous for the site-directed transgenes, 3H9 or 3H9(56R) and Vκ8 (4, 22, 23), were produced by crossing C.B-17 scid/scid mice doubly homozygous for these tgs (e.g., 3H9/3H9, Vκ8/Vκ8, scid/scid mice) with C.B-17 wt mice. The resulting C.B-17 scid/+ mice, hemizygous for 3H9 or 3H9(56R) and Vκ8, are simply designated as 3H9Vκ8 and 56RVκ8 mice, respectively. To produce mice homozygous for 56R and hemizygous for Vκ8 (56R/56R, Vκ8/+ mice), we crossed 56R/56R, Vκ8/Vκ8 wt mice with 56R/56R, scid/scid mice. Similarly, to produce mice hemizygous for 56R and homozygous for Vκ8 (56R/+, Vκ8/Vκ8 mice), we crossed 56R/56R, Vκ8/Vκ8, scid/scid mice with Vκ8/Vκ8 wt mice. C.B-17 scid/+ mice served as nontransgenic controls. Genotyping of transgenic mice was done by PCR as described previously (22, 23, 24). Investigators interested in obtaining mice homozygous for the tgs reported here should contact the Mutant Mouse Regional Resource Center (www.mmrrc.org). All of the mice used in this study, including BALB/c scid/scid (SCID mice), C57BL/6, (C57BL/6 × BALB/c)F1 mice, and C.B-17 mice with deleted JH loci (JH−/− mice) (25), were raised and maintained behind a barrier as specific pathogen-free mice in the Laboratory Animal Facility of the Fox Chase Cancer Center. Mice were used between 8 and 12 wk of age (unless otherwise stated) according to the protocols approved by the Animal Care and Use Committee of this institution.

SPL cells from a 5-mo-old 56RVκ8 mouse were stimulated with LPS (50 μg/ml) for 2 days and then fused with Sp2/0 cells (26) as described previously (27). Screening of culture supernatants by ELISA showed all hybridomas to produce IgM. The IgM protein in hybridoma supernatants was purified by precipitation in 60% saturated (NH4)2SO4 followed by size fractionation of the precipitated IgM using Amicon Ultra-15 Centrifugal Filter Units (Millipore). Purified hybridoma proteins were used for IgM allotyping, anti-dsDNA-binding assays and mass spectrometry. The anti-Vλx-producing hybridoma, 10C5 (28) was generously provided by P.-A. Cazenave (Institute Pasteur, Paris, France). It was grown in a CELLine Bioreactor System (Integra Biosciences) and the Ab was purified using a Melon Gel IgG Purification Kit (Pierce) as directed by the manufacturer.

Cell suspensions of BM and SPL were prepared and stained in the manner previously described (29, 30). Cells were stained for CD43, CD45 (B220), IgM, IgMa, IgMb, IgDa, CD23, CD93, and Vλx using combinations of the following reagents: fluorescein (FL)-anti-CD43 (S7), allophycocyanin-anti-B220 (RA3–6B2), FL-anti-IgDa (AMS9.1), FL-anti-IgM (331.12), biotin-anti-IgM (331.12), FL-anti-IgMa (RS3.1), biotin-anti-IgMb (AF6-78), biotin-anti-Vλx (10C5), Cy7PE-anti-CD23 (B3B4), and PE-anti-CD93 (AA4.1). Biotin-conjugated reagents were visualized by a second-step incubation with QDot605-streptavidin (Invitrogen). Biotin-anti-IgMb (AF6-78), Cy7PE-anti-CD23, PE-anti-CD93, and FL-anti-IgDa were obtained from BD Pharmingen. All other Ab reagents were prepared in this laboratory. Analyses were performed with FACSVantageSE/Diva and LSRII flow cytometers (BD Biosciences) using FlowJo software (Tree Star). All cytometric dot plots are based on the analysis of 105 cells for the indicated gates. Forward and side scatter were set to exclude nonlymphoid cells. Propidium iodide staining was used to exclude dead cells.

Apoptosis of Jurkat cells was induced for 4 h by the addition of 2.0 μM camptothecin (Sigma-Aldrich). At the end of the incubation period, 5 × 105 cells were collected and used in each binding reaction. Cells were washed in HBSS (Mediatech) with 3 mM CaCl2 and fixed in freshly prepared, ice-cold 6% paraformaldehyde (Electron Microscopy Sciences) in the same buffer for 15 min. Fixed cells were washed in HBSS and 3 mM CaCl2, pelleted at 1600 rpm for 5 min. and blocked by resuspension in wash buffer (HBSS containing 3 mM CaCl2, 3% FBS, and 0.02% azide) for 5 min. Cells were pelleted again and resuspended in 20 μg/ml of the primary Ab in wash buffer. Following a 30-min incubation with the primary Ab, cells were washed in wash buffer, pelleted as above, and incubated in a mixture of Alexa Fluor 647 goat anti-mouse IgM (μ-chain- specific) antisera (1/100 dilution), SYTOX Orange DNA stain (1/10,000 dilution), and Alexa Fluor 488/annexin V (1/70 dilution). All secondary reagents and stains were obtained from Invitrogen. Following incubation on ice for 20 min., cells were washed and resuspended in wash buffer containing 50% glycerol before mounting on 24-well, Teflon-printed microscope slides (Electron Microscopy Sciences).

Samples were viewed on a Zeiss LSM 510 laser scanning microscope by using a ×40 Plan-Apochromat oil-immersion lens and excitation at 488, 543, and 633 nm. Detection channels recorded fluorescence emission above 650 nm for Alexa Fluor 647 (displayed as red for consistency with our previous experiments), between 560 and 615 nm for SYTOX Orange (displayed as blue), and between 505 and 530 nm for Alexa Fluor 488 (displayed as green). Stacks of images were collected using between 20 and 30 optical sections taken at intervals of between 0.4 and 0.8 μm.

Hybridoma supernatants were screened by ELISA for the expression of IgM, IgMa, IgMb, Igκ, Igλ1 (from BD Pharmingen), and Vλx using purified mAbs. Assays were performed as described earlier (24) using solutions, buffers, and standards prescribed by BD Pharmingen. Hybridomas secreting IgM molecules that reacted with both anti-IgMa (RS 3.1) and IgMb (AF6-78) were detected using a sandwich ELISA protocol. The protocol involved coating plates with anti-IgMa to allow the binding of IgMa; this was followed by the addition of biotinylated anti-IgMb. To detect IgM anti-dsDNA Ab, we used the ELISA protocol previously described (24). To assay for serum IgVλx, plates were coated with purified anti-Vλx (10C5) followed by the addition of 5-fold diluted sera and biotinylated anti-Vλx. Purified IgM from the hybridoma C6 (see Fig. 9 A) was used as the standard for quantitation of serum Vλx.

FIGURE 9.

Purified IgMa from 56RVκ8 hybridomas binds to dsDNA and apoptotic cells. A, ELISA results for relative binding of IgM as a function of the concentration of IgM added to wells coated with avidin/biotinylated dsDNA. Top panel shows binding curves for unedited 56RVκ8 Ab of four hybridomas (A1–4). Unfilled circle and square symbols close to the abscissa correspond to purified IgM preparations from two IgMb-expressing hybridomas (from group D in Table II). Center panel shows binding curves for 56RVκ21D Ab of six hybridomas (B1–2 and B4–7). Bottom panel shows binding curves for 56RVλx Ab of six hybridomas (C1–4 and C6–7). B, Jurkat cells treated (apoptotic) or untreated (live) with camptothecin were fixed and incubated with nonspecific IgMλ (11E10) or IgM proteins consisting of 56RVκ8 (A4, top row), 56RVκ21D (B1 and B7, middle rows), or 56RVλx (C2 and C3, bottom rows). Bound Ab was detected with anti-IgM secondary reagent (displayed in red). DNA was visualized by binding of Sytox Orange (displayed in blue), whereas annexin V marked the presence of phosphatidylserine (green). These images are individual cross-sections taken from complete three-dimensional serial sections of each apoptotic cell.

FIGURE 9.

Purified IgMa from 56RVκ8 hybridomas binds to dsDNA and apoptotic cells. A, ELISA results for relative binding of IgM as a function of the concentration of IgM added to wells coated with avidin/biotinylated dsDNA. Top panel shows binding curves for unedited 56RVκ8 Ab of four hybridomas (A1–4). Unfilled circle and square symbols close to the abscissa correspond to purified IgM preparations from two IgMb-expressing hybridomas (from group D in Table II). Center panel shows binding curves for 56RVκ21D Ab of six hybridomas (B1–2 and B4–7). Bottom panel shows binding curves for 56RVλx Ab of six hybridomas (C1–4 and C6–7). B, Jurkat cells treated (apoptotic) or untreated (live) with camptothecin were fixed and incubated with nonspecific IgMλ (11E10) or IgM proteins consisting of 56RVκ8 (A4, top row), 56RVκ21D (B1 and B7, middle rows), or 56RVλx (C2 and C3, bottom rows). Bound Ab was detected with anti-IgM secondary reagent (displayed in red). DNA was visualized by binding of Sytox Orange (displayed in blue), whereas annexin V marked the presence of phosphatidylserine (green). These images are individual cross-sections taken from complete three-dimensional serial sections of each apoptotic cell.

Close modal

DNA was prepared from B cell hybridomas: Briefly, cells grown in 24-well plates were harvested by spinning the plates at 1200 rpm for 3 min and washing the cells in 500 μl of PBS (pH 7.4). Washed cells were resuspended in 300 μl of lysis buffer (31) and incubated overnight at 56°C. The next day, lysed cell samples were boiled for 15 min and stored at 4°C. PCR assays were done in a 50-μl volume containing 2 μl of template DNA and 1 U of Platinum TaqDNA polymerase (Invitrogen). For PCR amplification of the 56R tgs, we used the protocol described by Erikson et al. (32) and primers specific for the 56R leader (5′-CTGTCAGGAACTGCAGGTAAGG) and 56R CDR3 region (5′-CATAACATAGGAATATTTACTCCTCGC). To amplify recombining sequence (RS)-mediated rearrangements at the κ locus. a combination of primers was used as described by Retter and Nemazee (6) including primers specific for regions 3′ of the RS element (MB 619; 5′-ACATGGAAGTTTTCCCGGGAGAATATG) and 5′ of the IRS1 element (5′-CAACCTCTTCTTTACAACTGGGTGACC) and primers specific for the Vκ framework region 3 (5′-GGCTGCAGSTTCAGTTGGCAGTGGRTCWGGRAC) (31) and the Vκ8 leader (GGTACCTGTGGGGACATTGTG) (32). To score for DH to JH rearrangement at the wt H chain allele, we used a degenerate primer (5′-GGAATTCGMTTTTTGTSAAGGGATCTACTACTGTG) for DH (33) and primers 3′ of JH2 (5′-GGCTCCCAATGACCCTTTCTGA) or JH4 (5′-CTGTCCTAAAGGCTCTGAGATCC). Unrearranged wt H chain alleles were scored by retention of germline sequence upstream of JH1 using the forward (5′ JH1) and reverse (3′ JH1) primers, 5′-GCCAAGGACTTACCAAGAGG and 5′-GATGCAGGACTCACCTGACC (22), respectively.

Genomic DNA samples for LM-PCR were prepared from cells embedded in agarose as described previously (34). Linker ligation was performed with one-quarter of an agarose block (35). Broken molecules with recombination signal ends were amplified from linker-ligated DNA using one-tenth of the linker ligation reaction and the linker primer (5′-GCTATGTACTACCCGGGAATTCGTG) and a locus-specific primer. Jκ signal ends were amplified with the 5′ Jκ-specific primer (5′-AGTGCCACTAACTGCTGAGCCACCT) and Vλx signal ends were amplified with a 3′ Vλx-specific primer (MB719; 5′-AACATTGTGGCTGTCTCAGTGGCTCA). Jκ coding ends were amplified with a 3′ Jκ1-specific primer (5′-TCTCCAGAGAACATGTCTAGCC) using DNA pretreated with T4 DNA polymerase. RSκ-associated breaks were amplified with a 5′ RSκ-specific primer (MB615; 5′-CAGAAATGAAGGCAGACTCTCTCTAAC). The PCR conditions were 35 of 20 s at 95°C and 30 s at 63°C for DNA breaks at Jκ or 65°C for DNA breaks at Vλx and 30 s at 72°C, with a final 5-min extension at 72°C (see Figs. 3,B and 4 B). To control for the amount of DNA in each linker-ligated sample, the β2-microglobulin (β2m) gene was amplified for 25 cycles using a 5′ (5′-GAATGGGAAGCCGAACATACTGAACTG) and 3′ (5′-TGCTGATCACATGTCTCGATCC) primers with a cycle profile of 30 s at 95°C, 30 s at 61°C, and 45 s at 72°C.

FIGURE 3.

Analysis of 56RVκ8 B cell subsets for RAG expression and dsDNA breaks at Jκ, the RS element downstream of Cκ and Vλx. A, Bar graphs indicate the level of RAG1 expression (as determined by real-time quantitative RT-PCR) in sorted B cell subsets of 56RVκ8 BM and SPL relative to that in pro-B cell fraction of wt BM, which was assigned a value of 1.0. Error bars indicate the SEM for four independent experiments. B, Genomic DNA from ∼7 × 104 sorted B cells was analyzed by LM-PCR for Jκ1, Jκ2, Jκ3, RS, and Vλx recombination intermediates. Sorted cells included those from BM and SPL of 56RVκ8 mice and from control BM of 56R/+, Vκ8/Vκ8 mice (BM Vκ8/Vκ8). Recombination intermediates with Jκ signal, Jκ coding, RS, and Vλx signal ends are indicated in brackets I-IV, respectively. DNA from liver of RAG1−/− mice served as a negative control and β2m as a control for DNA loading. imm. Immature; mat, mature.

FIGURE 3.

Analysis of 56RVκ8 B cell subsets for RAG expression and dsDNA breaks at Jκ, the RS element downstream of Cκ and Vλx. A, Bar graphs indicate the level of RAG1 expression (as determined by real-time quantitative RT-PCR) in sorted B cell subsets of 56RVκ8 BM and SPL relative to that in pro-B cell fraction of wt BM, which was assigned a value of 1.0. Error bars indicate the SEM for four independent experiments. B, Genomic DNA from ∼7 × 104 sorted B cells was analyzed by LM-PCR for Jκ1, Jκ2, Jκ3, RS, and Vλx recombination intermediates. Sorted cells included those from BM and SPL of 56RVκ8 mice and from control BM of 56R/+, Vκ8/Vκ8 mice (BM Vκ8/Vκ8). Recombination intermediates with Jκ signal, Jκ coding, RS, and Vλx signal ends are indicated in brackets I-IV, respectively. DNA from liver of RAG1−/− mice served as a negative control and β2m as a control for DNA loading. imm. Immature; mat, mature.

Close modal
FIGURE 4.

Comparative analysis of T3 and T3′ splenic B cells in 56RVκ8 and C57BL/6 mice. A, Flow cytometric profiles delineating transitional splenic B cell populations in 56RVκ8 and C57BL/6 mice. Gates were set as in Fig. 2. Numbers outside and within the boxed areas and above the horizontal bars correspond to the percentage of cells with the indicated phenotype. B, Genomic DNA from ∼7 × 104 sorted T3 and T3′ (and T1 and T2) splenic B cells was analyzed by LM-PCR for dsDNA breaks at Jκ elements. DNA from BM of 56RVκ8 RAG1−/− mice and from sorted pre-B cells of 56RVκ8 mice served as a negative and positive control, respectively. β2m served as a control for DNA loading.

FIGURE 4.

Comparative analysis of T3 and T3′ splenic B cells in 56RVκ8 and C57BL/6 mice. A, Flow cytometric profiles delineating transitional splenic B cell populations in 56RVκ8 and C57BL/6 mice. Gates were set as in Fig. 2. Numbers outside and within the boxed areas and above the horizontal bars correspond to the percentage of cells with the indicated phenotype. B, Genomic DNA from ∼7 × 104 sorted T3 and T3′ (and T1 and T2) splenic B cells was analyzed by LM-PCR for dsDNA breaks at Jκ elements. DNA from BM of 56RVκ8 RAG1−/− mice and from sorted pre-B cells of 56RVκ8 mice served as a negative and positive control, respectively. β2m served as a control for DNA loading.

Close modal

LM-PCR products were subjected to agarose gel electrophoresis, transferred to nylon membranes, and hybridized to random-primed 32P-labeled probes generated from gel-purified DNA fragments in the manner described earlier (35). The Jκ probe was a 1.7-kb fragment from pJκ and spans the Jκ region (36). The Vλx probe, containing ∼500 bp of 3′ flanking sequence and confirmed by DNA sequence analysis, was a 745-bp fragment amplified from genomic liver DNA with oligonucleotides 5′-ACCTTGAGTAGTCAGCACAG (specific for the Vλx exon) and MB719. The RSκ probe was a 398-bp fragment amplified from genomic liver DNA with oligonucleotides MB615 and MB619. The β2m probe was made using the β2m- specific primers indicated above and gel purified. Clones for nucleotide sequence analysis were produced using the TOPO TA cloning kit (Invitrogen) and underwent plasmid recovery using a Perfectprep plasmid mini kit (Brinkman Instruments). Plasmids were submitted for cycle sequencing using the Applied Biosystems Prism dye terminator reaction kit and a model 3100 genetic analyzer (Applied Biosystems).

Total RNA was prepared using a Micro RNAeasy kit according to the manufacturer’s protocol (Qiagen). cDNA was synthesized by adding 2 μl of oligo(dT)18 primer (50 μM; Ambion) to 10 μl of total RNA, heating at 70°C for 3 min, cooling on ice, adding 2 μl of reverse transcriptase (RT) buffer (Ambion), 1 μl of dNTPs (each dNTP at 10 mM; Invitrogen), 0.6 μl of Superase-in (20 U/μl; Ambion), and 1 μl of ArrayScript RT (200 U/μl; Ambion), and then incubating at 42°C for 1 h. Gene expression was quantitated by real-time PCR. Analyses were performed in triplicate in 25-μl volumes using an Applied Biosystems I7500 thermal cycler. The cDNA was typically diluted 1/3. All probes were purchased from Applied Biosystems. Applied Biosystems software was used to quantify/calculate cycle threshold values and to determine relative gene expression levels using β-actin values for standardization.

Total RNA from 1 × 106 hybridoma cells was obtained using RNAeasy (Qiagen) and mRNA was isolated using Oligotex (Qiagen) according to the manufacturer’s instructions. Full-length cDNA was synthesized using a modified version of the SMART (switching mechanism at 5′ end of RNA transcript) technique (37) with one-tenth of the mRNA, PowerScript RT (BD Biosciences), 1.2 mM SMART oligonucleotide (5′-d(AAGCAGTGGTAACAACGCAGAGTAdCGC)ggg; Biosearch Technologies), 1 mM dNTP mixture (Invitrogen), and 5 μM oligo(dT) primer (Ambion) in a 10-μl reaction volume. The Ig cDNA was amplified via 5′ RACE-anchored RT-PCR using an Advantage 2 PCR kit (BD Clontech), 5′ Universal primer mix (40 nM 5′-CTAATACGACTCACTATAGGGCAAGCAGTGGTAAACAACGCAGAGT and 200 nM 5′-CTAATACGACTCACTATAGGGC) along with a reverse primer complementary to the constant region of the μ (5′-ATGCTCTTGGGAGACAGCAAGACCTGCG)- or κ (5′-CTCGTCCTTGGTCAACGTGAGGGTGCTG)- chain. The 5′ RACE RT-PCR conditions were: 95°C, 30 s, 5 cycles of 95°C for 5 s and 72°C for 1 min, then 5 cycles of 95°C for 5 s, 70°C for 10 s, 72 °C for 1 min, and finally 25 cycles of 95°C for 5 s, 68°C for 10 s, and 72°C for 1 min with a final extension time of 5 min at 72°C.

Purified IgM from supernatants of B cell hybridomas was reduced and subjected to 10% SDS-PAGE. The gels were stained with colloidal Coomassie blue and the L chains were excised and digested with trypsin after destaining, reduction, and alkylation (38, 39). Recovered peptides were prepared for MALDI-TOF mass spectrometry by mixing 1 μl of the peptide mixture with 2 μl of 30 μg/ml α-cyano-4-hydroxycinnamic acid and 0.4% trifluoroacetic acid in a 2:1 ethanol:acetone mixture and allowing the droplet to dry on the MALDI sample plate (40). In some cases, the α-cyano-4-hydroxycinnamic acid-affinity sample preparation was used to improve sensitivity (41). Peptide mass maps were obtained using a Bruker Daltonics Reflex IV MALDI-TOF mass spectrometer operated in positive ion reflectron mode. Proteins were identified from the peptide mass maps using Mascot (Matrix Science) (42) to search the protein sequence databases. A special protein database comprised of all know mouse L chain V, J, and C regions was constructed using the ImMunoGeneTics (IMGT) database (http://imgt.cines.fr/textes/IMGTrepertoire/Proteins/#1).

In 3H9(56R)/+, Vκ8/+-transgenic mice, the model system used in this study, the J regions of one H and L (κ) chain allele have been replaced with VDJH and VJκ coding sequences of the 3H9(56R) and Vκ8 tgs (4, 23, 43). H chains coded by 3H9(56R) differ from those coded by 3H9 by one amino acid; the 3H9 chain contains aspartate at position 56, whereas 3H9(56R) has arginine at this position (4). Together, 3H9(56R) and Vκ8 code for a relatively high-affinity anti-DNA Ab, whereas 3H9 and Vκ8 code for a low-affinity anti-DNA Ab (18). In this study, we used H/L chain-transgenic C.B-17 scid/+ mice hemizygous for the above tgs; i.e., 3H9(56R)/+, Vκ8/+, scid/+ and 3H9/+, Vκ8/+, scid/+ mice. The mice are simply designated with the prefix 56RVκ8 and 3H9Vκ8, respectively. C.B-17 scid/+ mice were used as the nontransgenic control; the scid mutation is recessive and these mice have a wt phenotype (44, 45). In some experiments, we included H/L chain-transgenic C.B-17 scid/+ mice homozygous for 56R (56R/56R, Vκ8/+, scid/+ mice) or Vκ8 (56R/+, Vκ8/Vκ8, scid/+ mice).

The primary question posed in this study was: when and where do developing anti-DNA B cells in 56RVκ8 mice undergo L chain receptor editing? To address this question, it was of interest to know whether the normal representation of pro-B, pre-B, and immature/transitional B cell subsets in BM and SPL is markedly altered in 56RVκ8 mice. As shown in Figs. 1 and 2, there were marked differences in the representation of these B cell subsets in 56RVκ8 mice vs 3H9Vκ8 and nontransgenic control mice (wt mice). Such differences were most striking in the SPL. For example, in 56RVκ8 mice, ∼10% of the sIgM gated lymphocytes (arrow in Fig. 1) stained positive for B220 (B220+CD43sIgM cells). In contrast, in wt and 3H9Vκ8 mice, the corresponding B220+CD43sIgM cell population represented ≤0.5% of sIgM gated lymphocytes. Using the additional B cell lineage markers, CD23 (46), CD93 (47), and surface IgD (sIgD), we found most B220+sIgM−/low cells in the SPL of 56RVκ8 mice to display a T3 or T3-like immature/transitional phenotype (Fig. 2). Transitional splenic B cells can be defined as T1, T2, or T3 according to their relative surface expression of IgM, IgD, CD21, CD23, CD24, and CD93 (19, 47, 48, 49). We used the markers B220, CD23, CD93, sIgM, and sIgD to delineate T1, T2, and T3 splenic B cells as designated by Allman et al. (19). The gates used to define B220+CD93+ transitional B cells in the SPL of 3H9Vκ8, 56RVκ8, 56RVκ8 RAG1−/− and wt mice are shown in Fig. 2, A and B. The CD93 staining of splenic B220+ cells (horizontal bar in Fig. 2 B) in wt and transgenic mice was less uniform and intense than that of pre-B cells from wt BM (denoted by the dashed histogram).

FIGURE 1.

Flow cytometric profiles illustrating differences in the representation of B cell subsets in BM and SPL of 56RVκ8 compared with 3H9Vκ8 and wt mice. The percentage of IgM gated lymphocytes displaying a pro-B (B220+CD43+IgM) or pre-B (B220+CD43IgM) cell surface phenotype is denoted within the boxed areas of the B220 vs CD43 dot plots. The percentage of lymphocyte gated cells displaying a B220+sIgM−/low or B220+sIgM+ cell surface phenotype in BM and SPL is denoted within the boxed areas of the B220 vs IgM dot plots. The profiles shown here (and in Figs. 2, 4, and 5) are representative of several independent experiments; in every experiment, at least two animals of each genotype were analyzed.

FIGURE 1.

Flow cytometric profiles illustrating differences in the representation of B cell subsets in BM and SPL of 56RVκ8 compared with 3H9Vκ8 and wt mice. The percentage of IgM gated lymphocytes displaying a pro-B (B220+CD43+IgM) or pre-B (B220+CD43IgM) cell surface phenotype is denoted within the boxed areas of the B220 vs CD43 dot plots. The percentage of lymphocyte gated cells displaying a B220+sIgM−/low or B220+sIgM+ cell surface phenotype in BM and SPL is denoted within the boxed areas of the B220 vs IgM dot plots. The profiles shown here (and in Figs. 2, 4, and 5) are representative of several independent experiments; in every experiment, at least two animals of each genotype were analyzed.

Close modal
FIGURE 2.

Flow cytometric profiles delineating transitional splenic B cells in RAG1−/−, wt, 3H9Vκ8, and 56RVκ8 mice. A, CD93 vs B220 staining of lymphocyte gated cells. B, Histograms of CD93 staining for B220+ gated cells in mice with the indicated genotypes; the dashed histograms correspond to CD93 staining of pre-B cells from wt BM. C, Upper panels, sIgM vs CD23 staining for B220+CD93+ gated cells show representation of T1, T2, T3, and T3′ splenic B cells. Lower panels, Histograms of sIgDa staining for T3 and T3′ splenic cells from 56RVκ8 mice are compared with the histogram for T2 splenic cells from 3H9Vκ8 mice. SPL cells from C.B-17 scid/+ mice (IgDb) served as the negative control for IgDa staining (dashed line). No sIgDa was detectable on BM cells with a pre-B or immature B cell phenotype (data not shown). Numbers above the horizontal bars in B and within the boxed areas in A and C correspond to the percentage of cells with the indicated phenotype in individual mice.

FIGURE 2.

Flow cytometric profiles delineating transitional splenic B cells in RAG1−/−, wt, 3H9Vκ8, and 56RVκ8 mice. A, CD93 vs B220 staining of lymphocyte gated cells. B, Histograms of CD93 staining for B220+ gated cells in mice with the indicated genotypes; the dashed histograms correspond to CD93 staining of pre-B cells from wt BM. C, Upper panels, sIgM vs CD23 staining for B220+CD93+ gated cells show representation of T1, T2, T3, and T3′ splenic B cells. Lower panels, Histograms of sIgDa staining for T3 and T3′ splenic cells from 56RVκ8 mice are compared with the histogram for T2 splenic cells from 3H9Vκ8 mice. SPL cells from C.B-17 scid/+ mice (IgDb) served as the negative control for IgDa staining (dashed line). No sIgDa was detectable on BM cells with a pre-B or immature B cell phenotype (data not shown). Numbers above the horizontal bars in B and within the boxed areas in A and C correspond to the percentage of cells with the indicated phenotype in individual mice.

Close modal

The majority of B220+CD93+ splenic B cells in 56RVκ8 mice displayed a T3 (CD23highIgM−/lowsIgD+) or T3-like (CD23−/lowIgM−/lowsIgD+) transitional phenotype (Fig. 2,C). Cells displaying the latter phenotype will be simply referred to as T3′ cells, although it remains to be determined whether this novel cell population corresponds to an additional subset of nondividing transitional cells. Both T3 and T3′ cells were RAG active (Fig. 3,A) and, consistent with the phenotype of T3 cells (19), were also CD21+ and CD24+ (our unpublished results in collaboration with R. Hardy). Note that sIgD, unlike sIgM, showed little down-regulation in T3 and T3′ splenic B cells of 56RVκ8 mice and was displayed at a level only slightly less than that of sIgD on T2 cells of 3H9Vκ8 mice (see histograms in lower row of Fig. 2,C). This finding is similar to that reported earlier by Goodnow et al. (12) in the anti-HEL/sHEL-transgenic mouse model in which anti-HEL splenic B cells showed persisting sIgD despite markedly reduced levels of sIgM. Finally, it should be noted that T1/T2 cells were underrepresented and T3/T3′ cells were overrepresented in 56RVκ8 SPL compared with 3H9Vκ8 and wt controls (Fig. 2,C). The mean percentages (±SEM) of T1, T2, T3, and T3′ splenic B cells for all mice analyzed, including C57BL/6 wt mice, are given in Table I. As discussed later, one possible explanation of these results is that expression of the strongly self-reactive 56RVκ8 receptor may cause most arising T1 and T2 B cells in C.B-17 mice to down-regulate their sIgM and display a T3′ and T3 cell surface phenotype.

Table I.

Representation of transitional splenic B cell subsets in individual 56RVκ8, 3H9Vκ8, C.B-17 scid/+, and C57BL/6 mice

MiceNo.bSplenic B Cell Subsetsa
T1T2T3T3′
56RVκ8 11 3.5 (±3.1) 2.8 (±2.4) 66.9 (±4.8) 26.3 (±5.0) 
3H9Vκ8 46.4 (±14.4) 43.7 (±15.0) 6.2 (±3.9) 3.8 (±3.5) 
C.B-17 scid/+ 31.2 (±9.5) 43.9 (±13.9) 15.3 (±4.7) 9.6 (±6.0) 
C57BL/6 47.6 (±4.2) 28.1 (±2.8) 12.2 (±2.1) 12.1 (±1.1) 
MiceNo.bSplenic B Cell Subsetsa
T1T2T3T3′
56RVκ8 11 3.5 (±3.1) 2.8 (±2.4) 66.9 (±4.8) 26.3 (±5.0) 
3H9Vκ8 46.4 (±14.4) 43.7 (±15.0) 6.2 (±3.9) 3.8 (±3.5) 
C.B-17 scid/+ 31.2 (±9.5) 43.9 (±13.9) 15.3 (±4.7) 9.6 (±6.0) 
C57BL/6 47.6 (±4.2) 28.1 (±2.8) 12.2 (±2.1) 12.1 (±1.1) 
a

Mean percentages (±SEM) of splenic B cells with a T1, T2, and T3 transitional phenotype and a T3-like (T3′) phenotype.

b

Number of mice analyzed.

To test whether L chain receptor editing was ongoing in both BM and SPL of 56RVκ8 mice, we sorted pro-B (B220+CD43+IgM), pre-B (B220+CD43IgM), and B (B220+IgMlow) cells from BM and immature/transitional T3 (B220+CD23highCD93+ IgM−/low), T3′ (B220+CD23−/lowCD93+IgM−/low) and B (B220+IgM+) cells from SPL. These B cell subsets were first compared for RAG expression relative to that in sorted pro-B cells from wt mice using real-time quantitative RT-PCR. Since similar results were obtained for both RAG1 and RAG2 expression, we show the results for RAG1 only in Fig. 3 A. Strikingly, RAG expression in B lineage cells of 56RVκ8 BM was strongly up-regulated in the pre-B but not the pro-B subset. The basis for relatively low RAG expression in 56RVκ8 pro-B cells is not clear. However, because these cells contain a prerearranged H chain gene, there would be no need for strong up-regulated RAG expression at the pro-B cell stage. RAG expression in B cells of 56RVκ8 BM was ∼7% of the level of that in the wt pro-B cells. In the SPL, RAG expression was strongly up-regulated in T3′ B cells and ∼8-fold higher than in T3 B cells. In splenic B cells of 56RVκ8 mice, RAG expression was ≤1% of that in wt BM pro-B cells.

We next tested sorted B cell subsets from 56RVκ8 mice for ongoing secondary L chain rearrangement as indicated by the presence of dsDNA breaks at the wt κ allele and the Vλx gene. Vλx is a known editor of 56R anti-dsDNA Ab (18, 50). LM-PCR (51, 52) was used to detect dsDNA breaks at the borders of Jκ and Vλx coding elements and their associated recombination signal sequences. RAG-mediated cleavage of DNA results in two species of broken molecules (recombination intermediates): those with signal ends and those with coding ends (reviewed in Ref. 53). In 56RVκ8 BM, Jκ signal ends appeared most abundant in the pre-B and B cell fractions; the pro-B cell fraction contained relatively few Jκ signal ends (Fig. 3,B, bracket I). Similar results were obtained for Jκ coding ends (Fig. 3 B, bracket II). Note, however, that coding ends were not detectable in the pro-B cell fraction, a finding that may in part reflect the shorter half-life of coding vs signal ends (54). Jκ signal and coding ends were also detectable in T3 and T3′ splenic B cells (and the less stringently sorted cell fraction of splenic B220+sIgM cells). Coding ends are rapidly joined during V(D)J recombination (55, 56). Thus, the presence of coding ends in RAG-active T3/T3′ splenic B cells is taken as evidence that the observed recombination intermediates were directly generated in these cell populations as opposed to being carried over from an earlier stage of differentiating B cells.

As expected, no Jκ signal or coding ends were detected in the pre-B cell subset from 56R/+, Vκ8/Vκ8 mice (see BM Vκ8/Vκ8 lane in Fig. 3,B, brackets I and II). In these mice, the Jκ region of both alleles contains the inserted Vκ8Jκ5 coding segment (23); thus, there is no Jκ substrate available for RAG targeting. Nonetheless, the κ locus in 56R/+, Vκ8/Vκ8 mice was clearly targeted by the recombinase machinery as indicated by dsDNA breaks at the RS downstream of Cκ (57, 58) (Fig. 3,B, bracket III). dsDNA breaks were not limited to the κ locus because we also found dsDNA breaks at the Vλx gene in BM pre-B/B cells and T3/T3′ splenic B cells (Fig. 3,B, bracket IV). Sequencing of Vλx products amplified by LM-PCR confirmed that they corresponded to DNA molecules cleaved at the Vλx coding/signal border (our unpublished data). From the results of Fig. 3 it is clear that, in 56RVκ8 mice, secondary L chain rearrangement is not only ongoing in BM cells with a pre-B/B cell phenotype, but also outside of the BM in splenic B cells with a T3/T3′ phenotype.

To test for possible ongoing L chain rearrangement in T3/T3′ splenic B cells of nontransgenic wt mice, we sorted these cell populations from the SPL of C57BL/6 mice and tested for dsDNA breaks at the signal/coding border of Jκ elements (Fig. 4). We chose C57BL/6 mice because T3 splenic B cells in these mice have been reported to represent mainly anergic self-reactive B cells (20, 21). As indicated in Fig. 4,A, T3/T3′ splenic B cells from C57BL/6 mice were less abundant and showed less down-regulation of sIgM than from 56RVκ8 mice. Following three independent sorts for T3/T3′ splenic B cells from C57BL/6 mice, we obtained sufficient material to assay for Jκ recombination intermediates. Fig. 4,B shows that Jκ signal ends were present in the T3/T3′ splenic B cell populations of C57BL/6 mice at a level comparable to that seen in the 56RVκ8 controls. We interpret these results to indicate that the T3/T3′ splenic B cell populations in wt mice, as in 56RVk8 mice, contain B cells undergoing L chain editing. As discussed later, some of these B cells may succeed in editing their Ag receptor and give rise to mature B cells. Splenic B cells with a T1 and T2 phenotype were also sorted from C57BL/6 mice and found to contain Jκ signal ends (Fig. 4 B), consistent with a recent report of Jκ signal ends in the T1 and T2 splenic B cells of nontransgenic C.B-17 mice (59). Thus, with respect to the presence of Jκ recombination intermediates, the splenic B cell populations T1 and T2 appear similar to T3 and T3′. As functionally immature B cells, members of the T1-T3 subsets also appear similar with respect to their unresponsiveness to BCR stimulation and rapid turnover (19).

Consistent with the detection of dsDNA breaks at the wt κ allele and Vλx gene, two distinct populations of L chain-edited B cells were readily detectable in the SPL but not the BM of 56RVκ8 mice. One population expressed Vλx (see diagonal profile marked with an arrow in Fig. 5 A) and represented ∼6% of the B220+ gated cells and ∼14% of all IgMa-expressing splenic B cells. Note that ∼3% of Vλx-expressing cells displayed little or no sIgMa; these cells were found to be sIgDhigh for the IgDa allotype (our unpublished data). Vλx-expressing B cells were not detectable in 3H9Vκ8 mice or in wt mice. Thus, the development of Vλx-expressing B cells in 56RVκ8 mice is clearly dependent on the 56R tg.

FIGURE 5.

Detection of Vλx-edited B cells and serum Vλx. A, Profiles of IgMa (the 56R tg allotype) vs Vλx for B220+ gated cells. BM and SPL cells from individual mice of each genotype were analyzed for the indicated markers. The arrow points to doubly stained (IgMa+/Vλx+) splenic B cells. The numbers in the upper right of each panel indicate the percentage of cells in each quadrant. The mean percentage (±SEM) of IgMa+/Vλx+ splenic cells in nine 56RVκ8 mice was 6.5 ± 1.7. B, Levels of serum Vλx. The filled diamonds and triangles in the two panels on the left correspond to serum Vλx concentrations in individual 3H9Vκ8 and 56RVκ8 mice, respectively, at 10–12 wk of age. The mean (±SEM) concentration of serum Vλx in 56RVκ8 mice was 0.672 (±0.29) μg/ml. The circles in the two panels on the right correspond to Vλx serum concentrations in BALB/c SCID mice that received an i.v. injection of 106 sorted T3′ cells (○) or 2 × 106 T3 cells (•) admixed with BM (2 × 106 cells) and thymus (3 × 106 cells) from C.B-17 JH−/− mice. Control BALB/c SCID mice (x) received C.B-17 JH−/− donor cells only. The mean concentration (±SEM) of serum Vλx in the seven SCID recipients with detectable Vλx was 1.10 (±1.41) and 2.11 (±3.27) μg/ml at 7 and 9 wk after cell transfer, respectively.

FIGURE 5.

Detection of Vλx-edited B cells and serum Vλx. A, Profiles of IgMa (the 56R tg allotype) vs Vλx for B220+ gated cells. BM and SPL cells from individual mice of each genotype were analyzed for the indicated markers. The arrow points to doubly stained (IgMa+/Vλx+) splenic B cells. The numbers in the upper right of each panel indicate the percentage of cells in each quadrant. The mean percentage (±SEM) of IgMa+/Vλx+ splenic cells in nine 56RVκ8 mice was 6.5 ± 1.7. B, Levels of serum Vλx. The filled diamonds and triangles in the two panels on the left correspond to serum Vλx concentrations in individual 3H9Vκ8 and 56RVκ8 mice, respectively, at 10–12 wk of age. The mean (±SEM) concentration of serum Vλx in 56RVκ8 mice was 0.672 (±0.29) μg/ml. The circles in the two panels on the right correspond to Vλx serum concentrations in BALB/c SCID mice that received an i.v. injection of 106 sorted T3′ cells (○) or 2 × 106 T3 cells (•) admixed with BM (2 × 106 cells) and thymus (3 × 106 cells) from C.B-17 JH−/− mice. Control BALB/c SCID mice (x) received C.B-17 JH−/− donor cells only. The mean concentration (±SEM) of serum Vλx in the seven SCID recipients with detectable Vλx was 1.10 (±1.41) and 2.11 (±3.27) μg/ml at 7 and 9 wk after cell transfer, respectively.

Close modal

Evidence that Vλx-expressing B cells can differentiate into Vλx-producing plasma cells is shown in Fig. 5,B. 56RVκ8 mice contained very low, but detectable levels of serum Vλx (0.42–1.42 μg/ml), whereas most 3H9Vκ8 mice lacked detectable serum Vλx (≤0.02 μg/ml; Fig. 5,B). No serum Vλx was detectable in C.B-17 scid/+ mice, the nontransgenic control (our unpublished results), consistent with previous reports of little or no detectable Vλx protein in normal mice (28, 50). To test whether the T3 and T3′ subsets contained cells able to differentiate into Vλx-producing plasma cells, we transferred sorted T3 and T3′ cells into SCID recipients in the manner previously described (24). Recipients also received a mixture of BM and thymus cells from C.B-17 JH−/− mice to provide a source of T cell help for any successfully Vλx-edited B cells that might arise from the T3 and T3′ cell populations. Serum Vλx was detectable 5 wk after cell transfer and, as shown in Fig. 5 B, all but one recipient was found to contain low levels of serum Vλx (0.15–9.8 μg/ml) at 7 and 9 wk after cell transfer. No serum Vλx was found in the SCID controls engrafted with C.B-17 JH−/− donor cells only. These results indicate that the T3 and T3′ splenic B cell populations contain cells capable of giving rise to edited Vλx-producing B cells.

The SPL of 56RVκ8 mice also contained a Vκ-edited cell population that was evident from an unexpected cross-reaction with the AF6-78 monoclonal anti-IgMb reagent (Fig. 6). In the course of testing for possible H chain editing (i.e., expression of the wt allotype (IgHb) instead of the tg allotype (IgHa)), we found ∼14% of B220+ gated SPL cells in 56RVκ8 mice (designated as 56R/+, Vκ8/+ mice in Fig. 6) to stain positive for both sIgMa and sIgMb; 35–43% of the remaining B220+ gated cells stained for sIgMa only and 8–9% stained for sIgMb only. The doubly stained cells represented 25–30% of cells expressing the sIgMa allotype; B cells staining positive for both sIgMa and sIgMb or for sIgMb alone were not detectable in 56RVκ8 BM or in SPL of 3H9Vκ8 mice. Further analysis revealed the presence of the doubly stained B cell population in mice homozygous for 56R and hemizygous for Vκ8 (56R/56R, Vκ8/+ mice). The genetic makeup of these mice would preclude any ability to express the IgMb allotype. Strikingly, no doubly stained B cells were detected in mice genetically competent to express both IgMa and IgMb allotypes, i.e., in mice hemizygous for 56R and homozygous for Vκ8 (56R/+, Vκ8/Vκ8 mice). From these results, we inferred that B cells doubly stained for sIgMa and sIgMb express an IgMa molecule with a L chain editor coded by the wt κ allele and that expression of this editor makes such molecules cross-reactive with the AF6-78 anti-IgMb reagent. Validation of this inference and identification of the L chain editor as Vκ21D is given in the next section. It is important to note here that Vκ21D appears not to be used in 3H9Vκ8 mice even though it has been shown to edit (veto) the ability of 3H9-coded H chains to bind dsDNA (4). Splenic B cells doubly stained for sIgMa and sIgMb were not detectable in 3H9Vκ8 mice (Fig. 6). We interpret these results and those of Fig. 5 to reflect little or no L chain editing in 3H9Vκ8 mice, consistent with an earlier report by Casellas et al. (7) showing 3H9Vκ8 mice to lack κ- and λ-edited B cells.

FIGURE 6.

Evidence for a specific Vκ-edited B cell population in the SPL of 56RVκ8 mice. Profiles of IgMa vs IgMb for B220+ gated cells of (C57BL/6 × BALB/c)F1; 56R/+, Vκ8/+; 56R/56R, Vκ8/+; 56R/+, Vκ8/Vκ8 and 3H9/+, Vκ8/+ mice. BM and SPL cells from individual mice of each genotype were analyzed for the indicated markers. The arrows point to splenic B cells that were doubly stained for IgMa and IgMb. The numbers in the upper right of each panel indicate the percentage of cells in each quadrant. The mean percentage (±SEM) of doubly stained splenic cells in six 56R/+, Vκ8/+ mice was 12.2 (±2.5).

FIGURE 6.

Evidence for a specific Vκ-edited B cell population in the SPL of 56RVκ8 mice. Profiles of IgMa vs IgMb for B220+ gated cells of (C57BL/6 × BALB/c)F1; 56R/+, Vκ8/+; 56R/56R, Vκ8/+; 56R/+, Vκ8/Vκ8 and 3H9/+, Vκ8/+ mice. BM and SPL cells from individual mice of each genotype were analyzed for the indicated markers. The arrows point to splenic B cells that were doubly stained for IgMa and IgMb. The numbers in the upper right of each panel indicate the percentage of cells in each quadrant. The mean percentage (±SEM) of doubly stained splenic cells in six 56R/+, Vκ8/+ mice was 12.2 (±2.5).

Close modal

To test for the representation and identity of L chain editors in individual B cells, we fused SPL cells from a single 56RVκ8 mouse with the SP2 cell line (26). Twenty-six hybridomas were obtained and all secreted IgM (Table II). Hybridoma culture supernatants were screened by ELISA for expression of the H chain allotype of the 56R (IgMa) and wt (IgMb) alleles; 13 typed positive for IgMa (groups A and C) and 6 for IgMb (group D). Seven hybridomas typed positive with both anti-IgMa and anti-IgMb (group B). Purified IgM from culture supernatants was analyzed by ELISA to confirm the cross-reactivity of secreted IgM molecules recognized by both anti-IgMa and anti-IgMb (results shown in Table II). Extensive analysis of the hybridomas in group B showed they all retained an unaltered 56R tg (see B1–7 in Fig. 8,B) and lacked a productive VDJ rearrangement at their wt H chain allele (our unpublished data). Thus, they could not have actually produced IgMb; instead, they appeared to produce IgMa molecules that cross-reacted with the anti-IgMb reagent as was also inferred for splenic B cells doubly stained for IgMa and IgMb in Fig. 6.

Table II.

IgM allotyping of B cell hybridomas from 56RVκ8 mouse; demonstration of IgMa molecules cross-reactive with a monoclonal anti-IgMb (AF6-78)a

HybridomasAbsorbance (mean ± SEM)
GroupNo.Anti-IgM/anti-IgMaAnti-IgM/anti-IgMbAnti-IgMa/anti-IgMb
1.30 ± 0.03 <0.04 <0.05 
1.26 ± 0.04 0.73 ± 0.20 1.11 ± 0.03 
1.41 ± 0.18 <0.05 <0.07 
<0.05 1.28 ± 0.23 <0.07 
HybridomasAbsorbance (mean ± SEM)
GroupNo.Anti-IgM/anti-IgMaAnti-IgM/anti-IgMbAnti-IgMa/anti-IgMb
1.30 ± 0.03 <0.04 <0.05 
1.26 ± 0.04 0.73 ± 0.20 1.11 ± 0.03 
1.41 ± 0.18 <0.05 <0.07 
<0.05 1.28 ± 0.23 <0.07 
a

Hybridoma groups are based on isotype, allotype, and mass spectrometry analyses of the IgM proteins secreted by each hybridoma. ELISA allotyping was initially done using hybridoma culture supernatants and then subsequently confirmed using purified hybridoma proteins. The ELISA results shown were obtained by coating plates with anti-IgM followed by sequential addition of excess hybridoma IgM (≥10 μg/ml) and biotinylated anti-IgMa or anti-IgMb. Detection of hybridoma IgMa cross-reactive with anti-IgMb involved coating of plates with anti-IgMa followed by sequential addition of excess hybridoma IgM and biotinylated anti-IgMb. For details see Materials and Methods.

FIGURE 8.

B cell hybridomas expressing Vκ21D- and Vλx-edited anti-dsDNA Abs represent distinct B cell clones. A, DNA sequences of DJH junctions of rearranged wt H chain alleles in hybridomas B1–7 and C1–7. In two hybridomas, the wt H chain locus was in germline configuration; all other hybridomas showed a rearranged wt allele. The hybridomas, B5 and C4, each contained two clones, designated B5a and B5b and C4a and C4b. B, Amino acid sequence for VH residues 40–80 of the original 56R tg and for the corresponding residues in the 56R tg of hybridomas B1–7 and C1–7. Amino acid sequences were deduced from the nucleotide sequence of 56R cDNA from each hybridoma (see Materials and Methods for details).

FIGURE 8.

B cell hybridomas expressing Vκ21D- and Vλx-edited anti-dsDNA Abs represent distinct B cell clones. A, DNA sequences of DJH junctions of rearranged wt H chain alleles in hybridomas B1–7 and C1–7. In two hybridomas, the wt H chain locus was in germline configuration; all other hybridomas showed a rearranged wt allele. The hybridomas, B5 and C4, each contained two clones, designated B5a and B5b and C4a and C4b. B, Amino acid sequence for VH residues 40–80 of the original 56R tg and for the corresponding residues in the 56R tg of hybridomas B1–7 and C1–7. Amino acid sequences were deduced from the nucleotide sequence of 56R cDNA from each hybridoma (see Materials and Methods for details).

Close modal

All 20 hybridomas expressing IgMa were screened for expression of Vκ8 and the two major L chain editors of anti-dsDNA Ab: Vλx and Vκ21D (4, 18, 50). Screening was done by mass spectrometric analysis of the L chains in purified IgM molecules from each hybridoma. Distinguishing mass spectra “signature” peptides for Vκ8, Vκ21D. and Vλx are shown in Fig. 7,A. As indicated, the IgM secreted by hybridomas A1–4, B1–7, and C1–7 contained L chains with Vκ8, Vκ21D, and Vλx variable regions, respectively. Two hybridomas produced IgM molecules containing two L chains, one with Vκ8 and Vλx and the other with Vλx and Vλ1 (our unpublished results). To confirm our mass spectrometric typing results, we used the SMART 5′ RACE assay to synthesize full-length cDNA from members of groups B and C and sequenced the expressed rearranged κ gene in B3, B4, and B5 and the rearranged Vλx gene in C1–7. Among all tested members of groups B and C, we recovered L chain sequences corresponding to a Vκ21D and Vλx gene, respectively (Fig. 7,B). Thus, all seven hybridomas (B1–7) that secreted IgMa molecules cross-reactive with AF6-78 anti-IgMb (shown in Table II) were found to express Vκ21D instead of Vκ8. Based on the results in Figs. 4–6, we estimate that Vλx- and Vκ21D-edited B cells represent ≥40% of IgMa-expressing splenic B cells in 56RVκ8 mice.

FIGURE 7.

Identification of L chains secreted by B cell hybridomas of 56RVκ8 mice. A, Vκ8, Vκ21D, and Vλx L chains produced by hybridomas of 56RVκ8 mice were identified by mass spectrometry. L chains of reduced hybridoma IgM proteins (illustrated in the gel lane to the right of each spectra) were excised from the gels and processed as described in Materials and Methods. In the mass spectra for Vκ8 and Vκ21D, the common tryptic peptide RADAAPTVSIFPPSSEQLTSGGASVVCFLNNFYPK derives from the κ constant region (residue C at position 27 is carbamidomethyl). The variable region peptides LLIYWASTR and LLIYAASNLESGIPAR are specific for Vκ8 and Vκ21D, respectively. Specific peptides for Vλx, VTVLGQPK, and KDGSHSTGDGIPDR are indicated in the Vλx mass spectra. B, The predicted amino acid sequence of the VJ junctional regions as deduced from the nucleotide sequences of rearranged L chain genes for three members in the Vκ21D group (B3–5) and for the expressed rearranged L chain genes of members in the Vλx group (C1–7). These results confirmed our typing results by mass spectrometry. Note that the CDR3 region (underlined) for the Vλx hybridoma in C2 is two amino acids (YN) longer than that of the other Vλx-producing hybridomas (C1 and C3–7).

FIGURE 7.

Identification of L chains secreted by B cell hybridomas of 56RVκ8 mice. A, Vκ8, Vκ21D, and Vλx L chains produced by hybridomas of 56RVκ8 mice were identified by mass spectrometry. L chains of reduced hybridoma IgM proteins (illustrated in the gel lane to the right of each spectra) were excised from the gels and processed as described in Materials and Methods. In the mass spectra for Vκ8 and Vκ21D, the common tryptic peptide RADAAPTVSIFPPSSEQLTSGGASVVCFLNNFYPK derives from the κ constant region (residue C at position 27 is carbamidomethyl). The variable region peptides LLIYWASTR and LLIYAASNLESGIPAR are specific for Vκ8 and Vκ21D, respectively. Specific peptides for Vλx, VTVLGQPK, and KDGSHSTGDGIPDR are indicated in the Vλx mass spectra. B, The predicted amino acid sequence of the VJ junctional regions as deduced from the nucleotide sequences of rearranged L chain genes for three members in the Vκ21D group (B3–5) and for the expressed rearranged L chain genes of members in the Vλx group (C1–7). These results confirmed our typing results by mass spectrometry. Note that the CDR3 region (underlined) for the Vλx hybridoma in C2 is two amino acids (YN) longer than that of the other Vλx-producing hybridomas (C1 and C3–7).

Close modal

IgMa-expressing hybridomas were also examined for rearrangements involving the RS element downstream of the Cκ constant region gene (57, 58). Using primers specific for the Vκ8 leader, JκCκ intronic recombining sequence (IRS1) and RS (see Materials and Methods), we recovered PCR products corresponding to an IRS1-RS rearrangement at the tg allele in all Vλx-expressing hybridomas of group C (Table III). We also recovered Vκ-RS rearrangements at the wt κ allele in four members of group C using primers specific for RS and the Vκ framework 3 region. Thus, expression of Vλx in the 56RVκ8 B cell hybridomas was found to correlate with a RS-mediated deletion of Cκ at the Vκ8 allele, consistent with earlier findings that λ-expressing B cells frequently show RS-mediated deletions of Cκ (57, 58). Rearrangements involving RS were also observed in four of seven Vκ21D-expressing hybridomas (Table III).

Table III.

L chain production and RSκ rearrangements in 56RVκ8 hybridomasa

Hybridoma GroupIgL Chain IsotypesRSκ Rearrangements
κλ1λxVκ8-IRS1-RSVκ-RS
B1 +(Vκ21D) − − − − 
B2 +(Vκ21D) − − − 
B3 +(Vκ21D) − − − − 
B4 +(Vκ21D) − − − 
B5 +(Vκ21D) − − − − 
B6 +(Vκ21D) − − − 
B7 +(Vκ21D) − − − 
C1 − − − 
C2 − − − 
C3 − − − 
C4 − − 
C5 − − 
C6 − − 
C7 − − 
Hybridoma GroupIgL Chain IsotypesRSκ Rearrangements
κλ1λxVκ8-IRS1-RSVκ-RS
B1 +(Vκ21D) − − − − 
B2 +(Vκ21D) − − − 
B3 +(Vκ21D) − − − − 
B4 +(Vκ21D) − − − 
B5 +(Vκ21D) − − − − 
B6 +(Vκ21D) − − − 
B7 +(Vκ21D) − − − 
C1 − − − 
C2 − − − 
C3 − − − 
C4 − − 
C5 − − 
C6 − − 
C7 − − 
a

L chain isotypes in hybridoma supernatants as assayed by ELISA and mass spectrometry (see Fig. 7). Usage of Vκ21D was also identified by SMART cDNA cloning and sequencing. Rearrangement of IRS1 on the Vκ8 tg allele and Vκ on the wt allele to the RSS downstream of Cκ (RS) were scored by PCR as described in Materials and Methods.

Because most of the hybridomas in groups B and C showed indistinguishable Vκ21D-Jκ and Vλx-Jλ rearrangements, respectively (Fig. 7,B), and all retained an unaltered 56R tg (Fig. 8,B), we considered that they may have represented only a few clones of expanded B cells. However, when we examined the status of the wt H chain allele in these hybridomas, we found clear evidence of multiclonality (Fig. 8 A). Five distinct B cell clones were represented among the seven hybridomas in the Vκ21D series; i.e., three showed different DJ rearrangements or junctions (B1, B4, and B5b), four were clonally related and contained the same DJ rearrangement (B2, B3, B5a, and B6), and one retained a germline H chain allele (B7). Seven distinct cell clones were represented among the hybridomas in the Vλx series; six contained different DJ rearrangements or junctions (C1, C2, C4a, C4b, C5, and C7) and one retained a germline H chain allele (C6). Therefore, most Vκ21D- and Vλx-expressing B cell hybridomas were clonally distinct and represented independently edited B cells.

To test the effect of Vκ21D and Vλx on the ability of the 56R H chain to bind dsDNA, we compared the relative anti-dsDNA binding of IgM from hybridomas in groups A, B, and C. This comparison was done with the same purified IgM preparations used for ELISA and mass spectrometric analysis of L chains in Table I and Fig. 7. The relative binding of IgM to dsDNA was plotted as a function of the concentration of IgM added to wells containing biotinylated dsDNA attached to avidin D. As shown in Fig. 9,A, all tested IgM preparations from the Vκ21D and Vλx series retained binding specificity for dsDNA but showed less affinity for dsDNA than IgM preparations from members of the Vκ8 group (A1–4). The binding curves for members of the Vκ21D group (B1–7) were displaced ∼5- to 8-fold to the right of those for the Vκ8 group. IgM from B2 and B6 gave similar binding curves consistent with our evidence that these hybridomas represented one and the same clone (Fig. 8). With one exception, the binding curves for members of the Vλx group (C1–7) were displaced ∼50- to 100-fold to the right of those for members of the Vκ8 group. The one exceptional member (C2) showed much stronger binding to dsDNA than other members in this group. Interestingly, the Vλx-Jλ2 junction of C2 contained two additional amino acids (tyrosine and asparagine) not present in other members of the group (Fig. 7,B). Thus, the lengthening of the Vλx-Jλ2 junction by these two amino acids correlated with an increased strength of anti-dsDNA binding. IgM from all six of the 56RVκ8 hybridomas that produced IgMb failed to bind dsDNA; this is illustrated for two such hybridomas in the top panel of Fig. 9 A. Our binding results are consistent with those of Doyle et al. (60) who also found 56RVλx-expressing B cell hybridomas to produce Ab with specificity for dsDNA, but they differ from those of earlier reports showing the Vκ21D and Vλx editors to inhibit completely the binding of 56R to dsDNA (4, 18, 50). The basis for the difference is unclear.

Because 56RVκ21D- and 56RVλx-expressing hybridomas secreted IgM Abs that bind dsDNA, albeit with low affinity, it was of interest to know whether these Abs would also bind to apoptotic Jurkat cells since anti-dsDNA Abs often bind to apoptotic cells (61). Specific targets of anti-dsDNA Abs include blebs formed by the protrusion of nuclear fragments from the cell surface. Anti-dsDNA Abs may also cross-react with phospholipid structures associated with the inner plasma membrane that become externalized during apoptosis (61, 62, 63, 64, 65). As shown in Fig. 9,B, IgM Abs from 56RVκ8-, 56RVκ21D-, and 56RVλx-expressing hybridomas showed preferential binding to apoptotic cells over live cells, although the pattern of Ab binding to apoptotic cells was distinct for each group of hybridomas. No appreciable IgM binding to apoptotic (or live) cells was evident with the nonspecific IgMλ control (11E10). 56RVκ8 Abs seemed to bind focal sites on apoptotic plasma membranes, possibly reflecting the recognition of phospholipid domains whose distribution in the membrane was altered by apoptosis (see A4, Fig. 9,B). In contrast, 56RVκ21D Abs displayed a preference for individual apoptotic blebs surrounding nuclear fragments or associated with the plasma membrane (B1 and B7, Fig. 9,B). 56RVλx Abs showed relatively sparse binding to the plasma membrane, yet bound dispersed sites in the cytoplasm (C2 and C3, Fig. 9 B). Unlike authentic anti-DNA autoantibodies (61), none of the above Abs bound to the surface of nuclear fragments, implying that these Abs do not recognize DNA-histone, subnucleosomal complexes, or intact nucleosomes.

Using sorted B cell subsets from 56RVκ8 mice and scoring for dsDNA breaks at the wt κ allele and Vλx gene, we obtained direct evidence for ongoing secondary L chain rearrangement in BM and splenic anti-DNA B cells. In the BM, dsDNA breaks were prominent in the B cell fraction (B220+sIgMlow) and in cells displaying a pre-B phenotype (B220+CD43IgM). With little or no cell surface Ag receptor, it is not obvious how sIgM−lo cells in 56RVκ8 BM could have been directly induced by self-Ag to undergo L chain rearrangement. Thus, we suggest these cells represent immature anti-DNA B cells that down-regulated their expression of sIgM due to encounter with self-Ag and then greatly elevated RAG expression. L chain editing may be initiated at the onset of Ag encounter and then continue in cells with down-regulated sIgM, consistent with our finding of abundant dsDNA breaks in both B220+sIgMlow and B220+sIgM BM B lineage cells (Fig. 3 B). Cells that fail to edit their Ag receptor successfully are presumably deleted, consistent with earlier evidence for deletion of anti-DNA B cells at the pre-B/B transitional stage (66).

In the SPL, dsDNA breaks were detectable in sIgM−low anti-DNA B cells with a T3 and T3′ phenotype. The immediate source of T3/T3′ splenic B cells in 56RVκ8 mice is not clear and could be BM and/or SPL. These cells might, for example, represent immature/transitional anti-DNA B cells that encountered self-Ag in the BM, down-regulated their sIgM, and then emigrated to SPL where they continued to undergo secondary L chain rearrangement as T3/T3′ cells. Alternatively, these cells could derive from immature/transitional anti-DNA B cells that encountered self-Ag in the SPL and then rapidly down-regulated expression of their sIgM (e.g., from that of a T1 cell to that of a T3′ cell). Although T3/T3′ anti-DNA B cells were deficient for sIgM, they nonetheless expressed intermediate levels of sIgD (Fig. 2 C), consistent with the previously described phenotype for T3 cells (19). Since this IgD receptor (sIgDa) would have the same anti-self specificity as sIgMa, it remains an open question as to whether it plays a role in signaling L chain editing in transitional anti-DNA splenic B cells.

RAG expression and L chain editing in the SPL have been reported earlier in immunized mice (67, 68, 69, 70, 71) and in LPS/IL-4-stimulated splenic B cells (72). However, such RAG expression and L chain editing may not occur, as initially thought, in mature splenic (germinal center) B cells. Using transgenic mice with a RAG2-GFP fusion gene inserted into the RAG2 locus (RAG2-GFP mice), Gartner et al. (73) found that most RAG2+ cells appearing in the SPL after immunization corresponded to migrant pre-B/immature B cells (B220lowCD43lowsIgMlow) from the BM. No GFP expression was detectable in sIgMhigh transitional splenic B cells or mature B cells (73, 74). In transgenic mice with GFP under the control of RAG2 regulatory genes, Yu et al. (75) reported that GFP expression does not occur in mature B cells during an immune response although it does occur in immature/transitional splenic B cells (B220low CD24+CD93+). The highest GFP expression was seen in sIgMlow transitional splenic B cells; sIgMhigh transitional splenic B cells showed lower GFP expression (75). Low RAG expression in sIgMhigh transitional splenic B cells has been reported in Ig-transgenic mice (6-1/Vκ1A mice) (59). Anti-Igβ stimulation of sIgMhigh transitional splenic B cells from 6-1/Vκ1A mice was found to result in slight to moderate up-regulation of RAG expression but not L chain rearrangement as scored by dsDNA breaks at Jκ coding elements (59). However, when BCR-induced apoptosis of cultured sIgMhigh transitional splenic B cells is inhibited by coculture with Thy1dullDX5+ cells from normal BM, up-regulated expression of both RAG and dsDNA breaks at Jκ coding elements is observed (76, 77). These results suggest that BCR-stimulated sIgMhigh transitional splenic B cells may undergo apoptosis or receptor editing, depending on the microenvironment in which these cells encounter Ag (76, 77). Our findings with 56RVκ8 and nontransgenic C57BL/6 mice extend the studies cited above and show that sIgMlow T3 transitional SPL B cells contain dsDNA breaks indicative of ongoing L chain rearrangement. Similar results were obtained with T3′ splenic B cells. Some T3/T3′ cells may successfully edit their receptor based on our detection of a L chain editor (Vλx) in the serum of SCID mice engrafted with sorted T3/T3′ splenic B cells from 56RVκ8 mice.

The presence of Vκ21D- and Vλx-edited B cells in the SPL of 56RVκ8 but not 3H9Vκ8 mice is attributable to the single amino acid substitution in the H chains encoded by 56R and 3H9 (arginine for aspartate at position 56 in the VH region) and correlates with the known difference in affinity of these H chains for DNA. The 56RVκ8 Ab has high affinity for DNA (binds both ssDNA and dsDNA), whereas the 3H9Vκ8 Ab has low affinity for DNA (binds ssDNA but not dsDNA) (18). Using a human κ polymorphism to detect L chain editing at the wt κ allele in 3H9Vκ8 mice, Casellas et al. (7) also found that κ (as well as λ)-edited B cells were not detectable in these mice. In contrast, κ- and λ-edited B cells were readily detectable in 3H9Vκ4 mice (7), which express an anti-dsDNA Ab with comparable affinity for dsDNA as that of the 56RVκ8 Ab (18). Yachimovich et al. (78) reported a lack of L chain editing in anti-DNA double-transgenic mice bearing the D42H (79) and Vκ8 tgs, but not in mice bearing the tgs, D42H and Vκ4 or Vκ1. L chain editing at the active tg allele in the D42H model correlated with the number of available Jκ gene segments on the tg alleles (three in Vκ1Jκ1, one in Vκ4Jκ4, and none in Vκ8Jκ5). All of these mice expressed anti-DNA Abs with moderate affinity for dsDNA (78). Although the Vκ8 tg can be inactivated by secondary RS-mediated rearrangements (Table II), it cannot express an edited L chain because there are no available Jκ coding segments (23). However, when the Vκ8 L chain is paired with the strongly self-reactive 56R H chain, as in 56RVκ8 mice, L chain editing readily occurs at the wt κ allele and the λ locus.

No edited B cells expressing Vκ21D or Vλx L chains were detectable in 56RVκ8 BM despite clear evidence that L chain editing is ongoing in this tissue. This finding is not really surprising. There are 96 known functional VL genes in the mouse genome (80) and, given that the RAG proteins do not preferentially target Vκ21D and Vλx genes, one would not expect a high frequency of 56RVκ21D- and 56RVλx-expressing B cells to arise in the BM. Moreover, given that edited B cells rapidly exit the BM, there would not be sufficient time for them to accumulate to detectable numbers. Vλx- and Vκ21D-edited B cells were, however, detectable in the SPL and represented ∼40% of all IgMa-expressing splenic B cells in 56RVκ8 mice. A higher percentage of edited B cells expressing Vκ21D or Vλx (∼70%) were recovered in our relatively small sample of IgMa-expressing splenic hybridomas. Strikingly, most of the hybridomas were clonally unrelated as evidenced by distinct DJ rearrangements at their wt H chain allele (Fig. 8). Thus, we infer that most Vκ21D- and Vλx-expressing cells in the SPL represent independently edited B cells.

IgM proteins from 56RVκ21D- and 56RVλx-expressing B cell hybridomas showed binding specificity for apoptotic cells (Fig. 9,B). The same proteins were also able to bind to dsDNA, albeit with much less affinity than unedited 56RVκ8 Ab as the binding curves of IgM to dsDNA for 56RVκ21D- and 56RVλx-expressing B cell hybridomas were displaced ∼5- to 8- and 50- to 100-fold, respectively, to the right of those for 56RVκ8-expressing B cell hybridomas (Fig. 9 A). Whether the edited B cells represented by these hybridomas retained sufficient autoreactivity to be positively selected by self-Ag is open to question. Low self-reactivity could be essential for the survival and maintenance of 56RVκ21D- and 56RVλx-expressing B cells. Others have previously shown that functional autoreactive cells can be generated in the presence of self-Ag (81, 82) and positively selected (82). Moreover, there is evidence most peripheral B cells may be positively selected by internal and external ligands (83) and that signaling through the BCR is essential for maintenance of mature B cells (84). With the previous findings of others serving as precedent, we suggest that edited anti-DNA B cells expressing Vκ21D or Vλx may retain sufficient self-reactivity to allow for their positive selection and maintenance in the presence of self-Ag.

T3 (and T3′) splenic B cells were 3- to 10-fold more frequent in 56RVκ8 mice than in 3H9Vκ8 and nontransgenic C.B-17 scid/+ mice. Overrepresentation of T3 splenic B cells has also been observed in transgenic B6.56R mice (85) and in two other Ig-transgenic models with autoreactive B cells (20, 86). In the latter two models, anti-Ars/A1 and anti-HEL/sHEL-transgenic mice, maintenance of the T3 phenotype was found to depend on continuous exposure to self-Ag. Moreover, the T3 cells in anti-HEL/sHEL C57BL/6 mice were shown to be anergic in that they did not mobilize calcium and initiate tyrosine phosphorylation in response to BCR stimulation (20). The T3 cell population in nontransgenic C57BL/6 mice has also been reported to consist mainly of anergic B cells (20, 21).

Does the anergic phenotype of T3 cells necessarily imply that these cells are (all) inactivated and unable to give rise to mature B cells? We think not based on our finding that the splenic B cell population with a T3 (and T3′) phenotype in 56RVκ8 and C57BL/6 normal mice contained cells with dsDNA breaks indicative of ongoing L chain rearrangement. Thus, we infer that these B cell populations contain cells attempting to edit their Ag receptor. Assuming T3 splenic B cells undergoing L chain rearrangement in normal wt mice are self-reactive, only those cells that successfully edit their Ag receptor would be expected to survive and give rise to mature B cells. As many L chain rearrangements presumably fail to result in expression of a suitable L chain editor, most T3 splenic B cells would be deleted, consistent with their rapid turnover and anergic phenotype (19, 20, 21).

We thank Dr. P-A. Cazenave for providing the 10C5 hybridoma, M. Weigert for the original BALB/c mice with the 3H9, 3H9(56R), and Vκ8 sd-tgs; R. Hardy for JH−/− mice; D. Douek for help with the SMART 5′RACE technique; D. Allman, R. Hardy, D. Kappes, and T. Manser for review of this manuscript; and R. Brooks and K. Trush for help in formatting the text and figures. Assistance from personnel in the following CORE facilities of the Fox Chase Cancer Center is gratefully acknowledged: the Flow Cytometry and Cell Sorting Facility, Laboratory Animal Resources, DNA Sequencing Facility, Biochemistry and Biotechnology Facility, and the Hybridoma Facility.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by National Institutes of Health Grants CA06927 and CA04946 and an appropriation from the Commonwealth of Pennsylvania.

4

Abreviations used in this paper: BM, bone marrow; tg, transgene; HEL, hen egg lysosome; sHEL, soluble hen egg lysozyme; sIgM, surface IgM; wt, wild type; FL, fluorescein; LM-PCR, ligation-mediated PCR; RS, recombining sequence; β2m, β2-microglobulin; RT, reverse transcriptase; SPL, spleen; IRS1, intronic recombining sequence 1.

1
Tiegs, S. L., D. M. Russell, D. Nemazee.
1993
. Receptor editing in self-reactive bone marrow B cells.
J. Exp. Med.
177
:
1009
-1020.
2
Radic, M. Z., J. Erickson, S. Litwin, M. Weigert.
1993
. B lymphocytes may escape tolerance by revising their receptors.
J. Exp. Med.
177
:
1165
-1173.
3
Gay, D., T. Saunders, S. Camper, M. Weigert.
1993
. Receptor editing: an approach by autoreactive B cells to escape tolerance.
J. Exp. Med.
177
:
999
-1008.
4
Li, H., Y. Jiang, E. L. Prak, M. Z. Radic, M. Weigert.
2001
. Editors and editing of anti-DNA receptors.
Immunity
15
:
947
-957.
5
Casellas, R., Q. Zhang, N. Y. Zheng, M. D. Mathias, K. Smith, P. C. Wilson.
2007
. Igκ allelic inclusion is a consequence of receptor editing.
J. Exp. Med.
204
:
153
-160.
6
Retter, M. W., D. Nemazee.
1998
. Receptor editing occurs frequently during normal B cell development.
J. Exp. Med.
188
:
1231
-1238.
7
Casellas, R., T. A. Shih, M. Kleinewietfeld, J. Rakonjac, D. Nemazee, K. Rajewsky, M. C. Nussenzweig.
2001
. Contribution of receptor editing to the antibody repertoire.
Science
291
:
1541
-1544.
8
Nemazee, D., K. Burki.
1989
. Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes.
Nature
337
:
562
-566.
9
Melamed, D., D. Nemazee.
1997
. Self-antigen does not accelerate immature B cell apoptosis, but stimulates receptor editing as a consequence of developmental arrest.
Proc. Natl. Acad. Sci. USA
94
:
9267
-9272.
10
Hertz, M., D. Nemazee.
1997
. BCR ligation induces receptor editing in IgM+IgD bone marrow B cells in vitro.
Immunity
6
:
429
-436.
11
Pelanda, R., S. Schwers, E. Sonoda, R. M. Torres, D. Nemazee, K. Rajewsky.
1997
. Receptor editing in a transgenic mouse model: site, efficiency, and role in B cell tolerance and antibody diversification.
Immnunity
7
:
765
-775.
12
Goodnow, C. C., J. Crosbie, S. Adelstein, T. B. Lavoie, S. J. Smith-Gill, R. A. Brink, H. Pritchard-Briscoe, J. S. Wotherspoon, R. H. Loblay, K. Raphael, et al
1988
. Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice.
Nature
334
:
676
-682.
13
Tze, L. E., E. A. Baness, K. L. Hippen, T. W. Behrens.
2000
. Ig light chain receptor editing in anergic B cells.
J. Immunol.
165
:
6796
-6802.
14
Tze, L. E., K. L. Hippen, T. W. Behrens.
2003
. Late immature B cells (IgMhighIgDneg) undergo a light chain receptor editing response to soluble self-antigen.
J. Immunol.
171
:
678
-682.
15
Kouskoff, V., G. Lacaud, K. Pape, M. Retter, D. Nemazee.
2000
. B cell receptor expression level determines the fate of developing B lymphocytes: receptor editing versus selection.
Proc. Natl. Acad. Sci. USA
97
:
7435
-7439.
16
Tze, L. E., B. R. Schram, K. P. Lam, K. A. Hogquist, K. L. Hippen, J. Liu, S. A. Shinton, K. L. Otipoby, P. R. Rodine, A. L. Vegoe, et al
2005
. Basal immunoglobulin signaling actively maintains developmental stage in immature B cells.
PLoS Biol.
3
:
e82
17
Mehr, R., G. Shahaf, A. Sah, M. Cancro.
2003
. Asynchronous differentiation models explain bone marrow labeling kinetics and predict reflux between the pre- and immature B cell pools.
Int. Immunol.
15
:
301
-312.
18
Chen, C., M. Z. Radic, J. Erikson, S. A. Camper, S. Litwin, R. R. Hardy, M. Weigert.
1994
. Deletion and editing of B cells that express antibodies to DNA.
J. Immunol.
152
:
1970
-1982.
19
Allman, D., R. C. Lindsley, W. DeMuth, K. Rudd, S. A. Shinton, R. R. Hardy.
2001
. Resolution of three nonproliferative immature splenic B cell subsets reveals multiple selection points during peripheral B cell maturation.
J. Immunol.
167
:
6834
-6840.
20
Merrell, K. T., R. J. Benschop, S. B. Gauld, K. Aviszus, D. Decote-Ricardo, L. J. Wysocki, J. C. Cambier.
2006
. Identification of anergic B cells within a wild-type repertoire.
Immunity
25
:
953
-962.
21
Teague, B. N., Y. Pan, P. A. Mudd, B. Nakken, Q. Zhang, P. Szodoray, X. Kim-Howard, P. C. Wilson, A. D. Farris.
2007
. Cutting edge: transitional T3 B cells do not give rise to mature B cells, have undergone selection, and are reduced in murine lupus.
J. Immunol.
178
:
7511
-7515.
22
Chen, C., Z. Nagy, E. L. Prak, M. Weigert.
1995
. Immunoglobulin heavy chain gene replacement: a mechanism of receptor editing.
Immunity
3
:
747
-755.
23
Prak, E. L., M. Weigert.
1995
. Light chain replacement: a new model for antibody gene rearrangement.
J. Exp. Med.
182
:
541
-548.
24
Bosma, G. C., J. Oshinsky, K. Kiefer, P. B. Nakajima, D. Charan, C. Congelton, M. Radic, M. J. Bosma.
2006
. Development of functional B cells in a line of SCID mice with transgenes coding for anti-double-stranded DNA antibody.
J. Immunol.
176
:
889
-898.
25
Chen, J., M. Trounstine, F. W. Alt, F. Young, C. Kurahara, J. F. Loring, D. Huszar.
1993
. Immunoglobulin gene rearrangement in B cell deficient mice generated by targeted deletion of the JH locus.
Int. Immunol.
5
:
647
-656.
26
Shulman, M., C. D. Wilde, G. Kohler.
1978
. A better cell line for making hybridomas secreting specific antibodies.
Nature
276
:
269
-270.
27
Kotloff, D. B., M. J. Bosma, N. R. Ruetsch.
1993
. V(D)J recombination in peritoneal B cells of leaky scid mice.
J. Exp. Med.
178
:
1981
-1994.
28
Sanchez, P., B. Nadel, P.-A. Cazenave.
1991
. Vλ-Jλ rearrangements are restricted within a V-J-C recombination unit in the mouse.
Eur. J. Immunol.
21
:
907
-911.
29
Bosma, G. C., J. Kim, T. Urich, D. M. Fath, M. G. Cotticelli, N. R. Ruetsch, M. Z. Radic, M. J. Bosma.
2002
. DNA-dependent protein kinase activity is not required for immunoglobulin class switching.
J. Exp. Med.
196
:
1483
-1495.
30
Chang, Y., M. J. Bosma, G. C. Bosma.
1999
. Extended duration of DH-JH rearrangement in immunoglobulin heavy chain transgenic mice: implications for regulation of allelic exclusion.
J. Exp. Med.
189
:
1295
-1305.
31
Schlissel, M. S., D. Baltimore.
1989
. Activation of immunoglobulin κ gene rearrangement correlates with induction of germline κ gene transcription.
Cell
58
:
1001
-1007.
32
Erikson, J., M. Z. Radic, S. A. Camper, R. R. Hardy, C. Carmack, M. Weigert.
1991
. Expression of anti-DNA immunoglobulin transgenes in non-autoimmune mice.
Nature
349
:
331
-334.
33
Schlissel, M. S., L. M. Corcoran, D. Baltimore.
1991
. Virus-transformed pre-B cells show ordered activation but not inactivation of immunoglobulin gene rearrangement and transcription.
J. Exp. Med.
173
:
711
-720.
34
Nakajima, P. B., J. P. Menetski, D. B. Roth, M. Gellert, M. J. Bosma.
1995
. V-D-J rearrangements at the T cell receptor δ locus in mouse thymocytes of the αβ lineage.
Immunity
3
:
609
-621.
35
Nakajima, P. B., M. J. Bosma.
2002
. Variable diversity joining recombination: nonhairpin coding ends in thymocytes of SCID and wild-type mice.
J. Immunol.
169
:
3094
-3104.
36
Coleclough, C., R. P. Perry, K. Karjalainen, M. Weigert.
1981
. Aberrant rearrangements contribute significantly to the allelic exclusion of immunoglobulin gene expression.
Nature
290
:
372
-378.
37
Chenchik, A., Y. Zhu, L. Diatchenko, R. Li, J. Hill, P. D. Siebert.
1998
. Generation and use of high quality cDNA from small amounts of total RNA by SMART PCR. P. Siebert, and J. larrick, eds.
RT-PCR Methods for Gene Cloning and Analysis
305
-319. BioTechniques Books, Natick, MA.
38
Shevchenko, A., M. Wilm, O. Vorm, M. Mann.
1996
. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels.
Anal. Chem.
68
:
850
-858.
39
Spodik, B., S. H. Seeholzer, T. R. Coleman.
2002
. Using Xenopus egg extracts to modify recombinant proteins. E.A. Coleman, ed.
Protein-Protein Interactions
355
-374. Cold Spring Harbor Lab. Press, Cold Spring Harbor.
40
Schuerenberg, M., C. Luebbert, H. Eickhoff, M. Kalkum, H. Lehrach, E. Nordhoff.
2000
. Prestructured MALDI-MS sample supports.
Anal. Chem.
72
:
3436
-3442.
41
Gobom, J., M. Schuerenberg, M. Mueller, D. Theiss, H. Lehrach, E. Nordhoff.
2001
. α-Cyano-4-hydroxycinnamic acid affinity sample preparation: a protocol for MALDI-MS peptide analysis in proteomics.
Anal. Chem.
73
:
434
-438.
42
Perkins, D. N., D. J. Pappin, D. M. Creasy, J. S. Cottrell.
1999
. Probability-based protein identification by searching sequence databases using mass spectrometry data.
Electrophoresis
20
:
3551
-3567.
43
Carmack, C. E., S. A. Camper, J. J. Mackle, W. U. Gerhard, M. G. Weigert.
1991
. Influence of a Vκ8 L chain transgene on endogenous rearrangements and the immune response to the HA(Sb) determinant on influenza virus.
J. Immunol.
147
:
2024
-2033.
44
Bosma, G. C., R. P. Custer, M. J. Bosma.
1983
. A severe combined immunodeficiency mutation in the mouse.
Nature
301
:
527
-530.
45
Kiefer, K., J. Oshinsky, J. Kim, P. B. Nakajima, G. C. Bosma, M. J. Bosma.
2007
. The catalytic subunit of DNA-protein kinase (DNA-PKCs) is not required for Ig class-switch recombination.
Proc. Natl. Acad. Sci. USA
104
:
2843
-2848.
46
Rao, M., W. T. Lee, D. H. Conrad.
1987
. Characterization of a monoclonal antibody directed against the murine B lymphocyte receptor for IgE.
J. Immunol.
138
:
1845
-1851.
47
Rolink, A. G., J. Andersson, F. Melchers.
1998
. Characterization of immature B cells by a novel monoclonal antibody, by turnover and by mitogen reactivity.
Eur. J. Immunol.
28
:
3738
-3748.
48
Loder, F., B. Mutschler, R. J. Ray, C. J. Paige, P. Sideras, R. Torres, M. C. Lamers, R. Carsetti.
1999
. B cell development in the spleen takes place in discrete steps and is determined by the quality of B cell receptor-derived signals.
J. Exp. Med.
190
:
75
-89.
49
Chung, J. B., R. A. Sater, M. L. Fields, J. Erikson, J. G. Monroe.
2002
. CD23 defines two distinct subsets of immature B cells which differ in their responses to T cell help signals.
Int. Immunol.
14
:
157
-166.
50
Li, Y., Y. Louzoun, M. Weigert.
2004
. Editing anti-DNA B cells by Vlambdax.
J. Exp. Med.
199
:
337
-346.
51
Roth, D. B., C. Zhu, M. Gellert.
1993
. Characterization of broken DNA molecules associated with V(D)J recombination.
Proc. Natl. Acad. Sci. USA
90
:
10788
-10792.
52
Schlissel, M., A. Constantinescu, T. Morrow, M. Baxter, A. Peng.
1993
. Double-strand signal sequence breaks in V(D)J recombination are blunt, 5′-phosphorylated, RAG-dependent, and cell cycle regulated.
Genes Dev.
7
:
2520
-2532.
53
Gellert, M..
2002
. V(D)J recombination: RAG proteins, repair factors, and regulation.
Annu. Rev. Biochem.
71
:
101
-132.
54
Ramsden, D. A., M. Gellert.
1995
. Formation and resolution of double-strand break intermediates in V(D)J rearrangement.
Genes Dev.
9
:
2409
-2420.
55
Roth, D. B., P. B. Nakajima, J. P. Menetski, M. J. Bosma, M. Gellert.
1992
. V(D)J recombination in mouse thymocytes: double-strand breaks near T cell receptor δ rearrangement signals.
Cell
69
:
41
-53.
56
Roth, D. B., J. P. Menetski, P. B. Nakajima, M. J. Bosma, M. Gellert.
1992
. V(D)J recombination: broken DNA molecules with covalently sealed (hairpin) coding ends in scid mouse thymocytes.
Cell
70
:
983
-991.
57
Durdik, J., M. W. Moore, E. Selsing.
1984
. Novel κ light-chain gene rearrangements in mouse λ light chain-producing B lymphocytes.
Nature
307
:
749
-752.
58
Moore, M. W., J. Durdik, D. M. Persiani, E. Selsing.
1985
. Deletions of κ chain constant region genes in mouse λ chain-producing B cells involve intrachromosomal DNA recombinations similar to V-J joining.
Proc. Natl. Acad. Sci. USA
82
:
6211
-6215.
59
Wang, H., J. Feng, C. F. Qi, Z. Li, H. C. Morse, III, S. H. Clarke.
2007
. Transitional B cells lose their ability to receptor edit but retain their potential for positive and negative selection.
J. Immunol.
179
:
7544
-7552.
60
Doyle, C. M., J. Han, M. G. Weigert, E. T. Prak.
2006
. Consequences of receptor editing at the l locus: multireactivity and light chain secretion.
Proc. Natl. Acad. Sci. USA
103
:
11264
-11269.
61
Radic, M., T. Marion, M. Monestier.
2004
. Nucleosomes are exposed at the cell surface in apoptosis.
J. Immunol.
172
:
6692
-6700.
62
Casciola-Rosen, L. A., G. Anhalt, A. Rosen.
1994
. Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes.
J. Exp. Med.
179
:
1317
-1330.
63
Casciola-Rosen, L. A., G. J. Anhalt, A. Rosen.
1995
. DNA-dependent protein kinase is one of a subset of autoantigens specifically cleaved early during apoptosis.
J. Exp. Med.
182
:
1625
-1634.
64
Cocca, B. A., S. N. Seal, P. D’Agnillo, Y. M. Mueller, P. D. Katsikis, J. Rauch, M. Weigert, M. Z. Radic.
2001
. Structural basis for autoantibody recognition of phosphatidylserine-β2 glycoprotein I and apoptotic cells.
Proc. Natl. Acad. Sci. USA
98
:
13826
-13831.
65
Cocca, B. A., A. M. Cline, M. Z. Radic.
2002
. Blebs and apoptotic bodies are B cell autoantigens.
J. Immunol.
169
:
159
-166.
66
Chen, C., Z. Nagy, M. Z. Radic, R. R. Hardy, D. Huszar, S. A. Camper, M. Weigert.
1995
. The site and stage of anti-DNA B-cell deletion.
Nature
373
:
252
-255.
67
Han, S., B. Zheng, D. G. Schatz, E. Spanopoulou, G. Kelsoe.
1996
. Neoteny in lymphocytes: Rag1 and Rag2 expression in germinal center B cells.
Science
274
:
2094
-2097.
68
Han, S., S. R. Dillon, B. Zheng, M. Shimoda, M. S. Schlissel, G. Kelsoe.
1997
. V(D)J recombinase activity in a subset of germinal center B lymphocytes.
Science
278
:
301
-305.
69
Hikida, M., M. Mori, T. Takai, K. Tomochika, K. Hamatani, H. Ohmori.
1996
. Reexpression of RAG-1 and RAG-2 genes in activated mature mouse B cells.
Science
274
:
2092
-2094.
70
Hikida, M., M. Mori, T. Kawabata, T. Takai, H. Ohmori.
1997
. Characterization of B cells expressing recombination activating genes in germinal centers of immunized mouse lymph nodes.
J. Immunol.
158
:
2509
-2512.
71
Hikida, M., H. Ohmori.
1998
. Rearrangement of λ light chain genes in mature B cells in vitro and in vivo: function of reexpressed recombination-activating gene (RAG) products.
J. Exp. Med.
187
:
795
-799.
72
Papavasiliou, F., R. Casellas, H. Suh, X. F. Qin, E. Besmer, R. Pelanda, D. Nemazee, K. Rajewsky, M. C. Nussenzweig.
1997
. V(D)J recombination in mature B cells: a mechanism for altering antibody responses.
Science
278
:
298
-301.
73
Gartner, F., F. W. Alt, R. J. Monroe, K. J. Seidl.
2000
. Antigen-independent appearance of recombination activating gene (RAG)-positive bone marrow B cells in the spleens of immunized mice.
J. Exp. Med.
192
:
1745
-1754.
74
Monroe, R. J., K. J. Seidl, F. Gaertner, S. Han, F. Chen, J. Sekiguchi, J. Wang, R. Ferrini, L. Davidson, G. Kelsoe, F. W. Alt.
1999
. RAG2:GFP knockin mice reveal novel aspects of RAG2 expression in primary and peripheral lymphoid tissues.
Immunity
11
:
201
-212.
75
Yu, W., H. Nagaoka, M. Jankovic, Z. Misulovin, H. Suh, A. Rolink, F. Melchers, E. Meffre, M. C. Nussenzweig.
1999
. Continued RAG expression in late stages of B cell development and no apparent re-induction after immunization.
Nature
400
:
682
-687.
76
Sandel, P. C., J. G. Monroe.
1999
. Negative selection of immature B cells by receptor editing or deletion is determined by site of antigen encounter.
Immunity
10
:
289
-299.
77
Sandel, P. C., M. Gendelman, G. Kelsoe, J. G. Monroe.
2001
. Definition of a novel cellular constituent of the bone marrow that regulates the response of immature B cells to B cell antigen receptor engagement.
J. Immunol.
166
:
5935
-5944.
78
Yachimovich, N., G. Mostoslavsky, Y. Yarkoni, I. Verbovetski, D. Eilat.
2002
. The efficiency of B cell receptor (BCR) editing is dependent on BCR light chain rearrangement status.
Eur. J. Immunol.
32
:
1164
-1174.
79
Pewzner-Jung, Y., D. Friedmann, E. Sonoda, S. Jung, K. Rajewsky, D. Eilat.
1998
. B cell deletion, anergy, and receptor editing in “knock in” mice targeted with a germline-encoded or somatically mutated anti-DNA heavy chain.
J. Immunol.
161
:
4634
-4645.
80
Martinez-Jean, C., G. Folch, M. P. Lefranc.
2001
. Nomenclature and overview of the mouse (Mus musculus and Mus sp.) immunoglobulin κ (IGK) genes.
Exp. Clin. Immunogenet.
18
:
255
-279.
81
Hannum, L. G., D. Ni, A. M. Haberman, M. G. Weigert, M. J. Shlomchik.
1996
. A disease-related rheumatoid factor autoantibody is not tolerized in a normal mouse: implications for the origins of autoantibodies in autoimmune disease.
J. Exp. Med.
184
:
1269
-1278.
82
Hayakawa, K., M. Asano, S. A. Shinton, M. Gui, D. Allman, C. L. Stewart, J. Silver, R. R. Hardy.
1999
. Positive selection of natural autoreactive B cells.
Science
285
:
113
-116.
83
Gu, H., D. Tarlinton, W. Muller, K. Rajewsky, I. Forster.
1991
. Most peripheral B cells in mice are ligand selected.
J. Exp. Med.
173
:
1357
-1371.
84
Lam, K.-P., R. Kuhn, K. Rajewsky.
1997
. In vivo ablation of surface immunoglobulin on mature B cells by inducible gene targeting results in rapid cell death.
Cell
90
:
1073
-1083.
85
Sekiguchi, D. R., L. Yunk, D. Gary, D. Charan, B. Srivastava, D. Allman, M. G. Weigert, E. T. Prak.
2006
. Development and selection of edited B cells in B6.56R mice.
J. Immunol.
176
:
6879
-6887.
86
Benschop, R. J., K. Aviszus, X. Zhang, T. Manser, J. C. Cambier, L. J. Wysocki.
2001
. Activation and anergy in bone marrow B cells of a novel immunoglobulin transgenic mouse that is both hapten specific and autoreactive.
Immunity
14
:
33
-43.