Antinuclear Antigen B Cells That Down-Regulate Surface B Cell Receptor during Development to Mature, Follicular Phenotype Do Not Display Fea
http://www.100md.com
免疫学杂志 2005年第8期
Abstract
We previously demonstrated that B cells expressing a transgenic BCR with "dual reactivity" for the hapten arsonate and nuclear autoantigens efficiently complete development to follicular phenotype and stably reside in follicles in vivo. These B cells express very low levels of surface IgM and IgD, suggesting that they avoid central deletion and peripheral anergy by reducing their avidity for autoantigen via surface BCR (sBCR) down-regulation. Since a variety of states of B cell anergy have been previously described, a thorough examination of the functional capabilities of these B cells was required to test this hypothesis. In this study, we show that surface Ig cross-linking induces amounts of proximal BCR signaling in these B cells commensurate with their reduced sBCR levels. Functionally, however, they are comparable to nonautoreactive B cells in cell cycle progression, up-regulation of activation and costimulatory molecules, and Ab-forming cell differentiation when treated with a variety of stimuli in vitro. In addition, these B cells can efficiently process and present Ag and are capable of undergoing cognate interaction with naive TCR-transgenic T cells, resulting in robust IL-2 production. Together, these data reveal a lack of intrinsic anergy involving any known mechanism, supporting the idea that this type of antinuclear Ag B cell becomes indifferent to cognate autoantigen by down-regulating sBCR.
Introduction
Insight into B cell tolerance pathways were first provided by in vitro studies showing that engagement of the BCR on immature B cells or their transformed counterparts results in apoptotic death or unresponsiveness, rather than activation (1, 2, 3). The analysis of Ig-transgenic mice expressing autoreactive BCRs have confirmed and extended these in vitro studies (4). It is now clear that certain types of transgenic autoreactive BCRs are removed from the developing repertoire via clonal deletion and receptor editing, mainly in the bone marrow (4, 5, 6, 7). Nonetheless, deletion or editing mechanisms do not result in a complete "cleansing" of the B cell compartment of autoreactivity. Numerous investigations have shown that Ig-transgenic B cells with particular types of autospecificities, in some cases associated with autoimmune disease, can exit the bone marrow and populate the periphery (8, 9, 10, 11, 12, 13). However, these B cells often display distinct phenotypes and reduced responsiveness (4, 8, 10, 14), a condition commonly referred to as "anergy" (2, 3, 4, 15).
Despite extensive studies over the last decade on the physiological state of peripheral autoreactive B cells, a standard definition of anergy has not been established. Anergic B cells described in different experimental systems often respond to in vitro and in vivo stimuli in distinct ways (8, 14, 16, 17, 18, 19). This argues against a "black and white" definition of anergy, and suggests that autoantigen-induced changes in B cell behavior are graded, depending on factors such as autoantigen structure and concentration, the avidity of the BCR-autoantigen interaction, and the time during B cell development that autoantigen is first encountered. However, essentially all definitions of anergy include hyporesponsiveness to BCR-mediated stimulation. In addition, anergy is usually associated with reduced levels of surface (s) 3 IgM and in some cases sIgD (4). Many anergic B cells have reduced life spans and may be excluded from follicles, perhaps due to reduced ability to "homeostatically compete" with nonanergic B cells for peripheral trophic factors such as BAFF (BLys) (20, 21, 22). Moreover, some investigators have found that while certain peripheral autoreactive B cells can mount proliferative responses to a variety of in vitro stimuli, they inefficiently differentiate to Ab secretory cells (18, 23, 24), suggesting that this step in development may be a target of a tolerance checkpoint.
Despite their reduced responsiveness, many anergic B cells do undergo proliferation and differentiation in vitro and in vivo upon receipt of surrogate or genuine T cell help (25, 26, 27, 28, 29). Activation of innate immune system accessory cells or receptors expressed by B cells themselves may also contribute to "reversal" of the anergic state (30, 31, 32). Also, several recent studies have indicated that some autoreactive B cells complete peripheral development harboring no obvious proliferative or differentiative defects, suggesting that they do not functionally interact with their cognate autoantigen(s). Such B cells are often referred to as being in an "ignorant" or "indifferent" state (12, 33). These observations suggest that the presence of autoreactive B cells in the periphery poses a significant risk factor for the development of autoimmune disease. For this reason, how and to what extent B cells categorized as anergic, ignorant, or indifferent by current standards can respond to BCR, mitogenic, and costimulatory stimuli is worthy of detailed investigation.
We have previously described two lines of VH"knockin" mice that differ only in the presence or absence of a single mutation to arginine (R) at position 55 in the CDR2 subregion of the VH gene used to replace the endogenous JH locus (34, 35). We term these lines of mice HKIR and HKI65, respectively. Both versions of this VH gene, in combination with a single, endogenous L chain gene, encode Abs that we term "canonical." Canonical Abs bind the hapten arsonate and can be specifically detected using the monoclonal anti-clonotypic Ab E4 (36, 37). Abs with the R55 form of the V domain also display reactivity for nuclear autoantigens such as chromatin and can cause kidney dysfunction via glomerular deposition in vivo. B cells expressing BCRs containing both types of V domains develop to mature follicular phenotype, reside in follicles, and are not short-lived (34). However, peripheral B cells expressing the R55 form are characterized by substantially reduced levels of sIgM and sIgD as compared with B cells expressing the form lacking this mutation (34, 35). Along with the results of previous studies discussed above indicating a correlation between sBCR down-regulation and reduced responsiveness of autoreactive B cells, these latter data suggested that the follicular B cells in HKIR mice expressing the canonical R55 BCR might be anergic.
Materials and Methods
Mice
HKIR knockin mice, expressing the canonical R55 VH gene, and HKI65 mice, expressing the canonical VH lacking the R55 mutation, were created in an identical fashion by replacing the entire JH locus with a targeting vector differing only in the sequence of VH codon 55 (34, 35). For many experiments, these lines were crossed to lines in which the entire JH locus had been deleted (JHD mice) (38) (a gift from Dr. R. Hardy, Fox Chase Cancer Center, Philadelphia, PA, USA) or the IgM transmembrane exon had been disrupted (μMT mice) (39) to preclude the development of B cells expressing the endogenous IgH locus. OVA-specific class II-restricted TCR-transgenic mice (OT-II mice) on the C57BL/6 (B6) background were kindly provided by Dr. H. Shen (Department of Microbiology, University of Pennsylvania, Philadelphia, PA (with permission from Dr. W. Heath, Walter and Eliza Hall Institute, Parkville, Victoria, Australia)) (40). All mice were used between 8 and 12 wk of age. Mice were housed under specific pathogen-free conditions and given autoclaved food and water. The use of mice in these studies was conducted in compliance with institute guidelines, and all protocols using animals were approved by the institutional animal care and use committee.
Cells and cell lines
Splenic B cells were isolated as previously described (34). Briefly, single-cell suspensions were prepared from spleens of naive transgenic and age-matched transgene-negative littermates. After removal of dead cells by centrifugation through Lympholyte-M (Cedarlane Laboratories), cells were stained with either biotin-E4 or biotin-anti-IgMb Ab, followed by incubation with paramagnetic streptavidin beads (Miltenyi Biotec). Following the manufacturer’s recommendations, labeled cells were passed over MiniMACS columns (Miltenyi Biotec) and eluted. Two column passes per sample were performed. The purity of the E4+ B cells as assessed by flow cytometry was always >85%. The purity of IgMb+ B cells was always >95%. MHC class II (MHC-II) I-Ab haplotype-restricted T cell hybridoma cell lines BO17.10 (specific for epitope OVA323–339) and BO4 (specific for epitope HEL74–78)) (kindly provided by Dr. G. J. Hammerling, German Cancer Center, DKFZ, Heidelberg, Germany) were maintained as previously described (41).
Flow cytometry and FACS
Cells (106/sample) were stained according to a previously described protocol (34). The following Abs were used: anti-CD45R (RA3-6B2; eBioscience), anti-IgMb (AF6-78; Southern Biotechnology), anti-CD69 (H1.2F3; BD Pharmingen), anti-CD80 (16-10A1; BD Pharmingen), anti-CD86 (GL1; BD Pharmingen), anti-I-A/I-E (2G9; BD Pharmingen), anti- (187.1; Southern Biotechnology Associates), anti-C1qRp (AA4.1; eBioscience), anti-CD21/CD35 (7G6; BD Pharmingen), anti-CD23 (B3B4; BD Pharmingen), anti-CD24 (HSA,M1/69; BD Pharmingen), CD1d (Cd1.1,1B1; BD Pharmingen), anti-CD62L (Mel-14; BD Pharmingen), and anti-idiotypic mAb E4 (prepared in-house). Streptavidin-R670 (BD Pharmingen) was used to detect biotinylated Abs. For evaluating levels of sIgM and total sBCR and Ca2+ flux responses in HKIR E4+, HKI65 E4+IgDhigh, and HKI65 E4+IgDlow populations, splenocytes from HKIR/JHD and HKI65/JHD mice were stained with E4 and anti-IgD and these populations were purified using a MoFlo high-performance cell sorter (DakoCytomation). The sorted cells were then restained with anti-IgM or anti- or stimulated with anti-IgM and subjected to the Ca2+ flux assay (see below). All flow cytometric analyses were performed on an EPICS Elite (Coulter) flow cytometer, and data were analyzed using the FlowJo software (Tree Star).
Real-time RT-PCR
Ig RNA levels in B cells were measured using an ABI Prism 7000 sequence detection system (PerkinElmer). Purified B cells were activated with LPS for 24 or 72 h, cells were harvested, and RNA was extracted using the RNeasy Mini kit (Qiagen). All RNA templates were subjected to DNase digestion to exclude genomic DNA contamination following the protocol in the kit. The cDNA was subjected to PCR using primers and TaqMan probes (Custom Oligonucleotide Factory) designed to span introns to minimize amplification of genomic DNA and to result in amplicons of <150 bp. Probes were labeled at the 5' end with FAM and at the 3' end with 6-carboxytetramethyl-rhodamine. Relative RNA concentrations were determined by the number of cycles necessary to reach a certain amount of fluorescence in a test sample compared with a designated calibration sample (RNA from littermate IgMb+ B cells) after normalizing to the number of cycles required to amplify a control GAPDH target. The data were analyzed with the ABI Prism 7000 SDS software (Applied Biosystems).
ELISA
Total Ig or E4 Ab was measured in culture supernatants as previously described (42). ELISA using 96-well plates (Immulon-4; Dynatech Laboratories) coated with either goat anti-mouse IgG + IgM (H + L) or E4 Abs were performed.
Ca2+ influx measurements
Fluo-3 AM loading of purified B cells and surface immunostaining of unfractionated splenocytes was performed using the method described before, with minor modifications (43). Cells were resuspended at 2 x 106/ml in HBSS with 1 μM fluo-3-acetoxymethyl ester (fluo-3 AM) and 0.5 μg/ml pluronic F-127 (Molecular Probes) and incubated for 30 min at 37°C. For surface immunostaining of fluo-3-AM-loaded splenocytes, cells were washed three times in HBSS containing 1% FCS and first incubated with biotinylated mAb E4 or anti-IgMb in FACS buffer (PBS/3% BSA) for 20 min at 4°C. Then Streptavidin-R670 was added and the cells were incubated as described above. The cells were then washed twice and diluted to 106 cells/ml with HBSS/1% FCS. The cell suspensions were placed in a water bath at 37°C for 15 min before flow cytometric analysis. Calcium mobilization was induced by adding various concentrations of goat anti-mouse IgM (Fab')2 (Pierce) and monitored by FACS either on total purified B cells or by gating on the E4+ or IgMb+ cells.
Immunoblotting
Levels of protein tyrosine phosphorylation in MACS-purified B cells were detected following stimulation with 10 μg/ml goat anti-mouse IgM (Fab')2 at 37°C for 3 min. Cells were then chilled on ice immediately and pelleted by centrifugation. For Western blotting, cell pellets were lysed in ice-cold lysis buffer (20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM Na2 EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM -glycerophosphate, 1 mM Na3VO4, and 1 μg/ml leupeptin). The buffer was also supplemented with 1 mM PMSF and a complete protease inhibitor tablet (Roche). The whole cell lysates were subjected to 10% SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Millipore) using a Mini Trans-Blot electrophoretic transfer apparatus (Bio-Rad). The blots were blocked with TBS containing 5% milk and incubated with anti-phosphotyrosine-HRP mAb (PY-99; Santa Cruz Biotechnology). Ab binding was detected using the ECL (Pierce) substrate. As a loading control, membranes were stripped and reprobed with an anti -actin Ab (Novus Biological) and developed as described above.
Cell cycle analysis
Splenocyte suspensions were prepared as described above. Cells were washed once with cold PBS with 0.1% sodium azide and labeled with FITC-conjugated B220 and biotinylated E4 or anti-IgMb in cold staining buffer, and biotinylated E4 was revealed by streptavidin-PE. Cells were then washed with cold PBS containing 0.1% sodium azide and resuspended in 0.3 ml of 50% FBS in PBS. Labeled cells were fixed by incubating with 0.9 ml of cold 70% EtOH overnight at 4°C, and cells were then washed twice with cold PBS/azide to remove EtOH and precipitated protein. 7-Aminoactinomycin D (20 μg/ml in PBS) was then added, and E4+ B cells or IgMb+ cells were analyzed using flow cytometry by gating on B220+E4+ and B220+IgMb+ populations. For cell cycle analyses of activated B cells, E4+ or IgMb+ B cells were purified as described above and stimulated with LPS (2.5 μg/ml), anti-IgM (Fab')2 (25 μg/ml), or anti-CD40 (15 μg/ml) plus IL-4 (50 ng/ml) in culture at 37°C for 48 h. Cells were then harvested and stained with 50 μg/ml propidium iodide in PBS. DNA content per cell was then analyzed using flow cytometry and the FlowJo software (Tree Star).
ELISPOT assays
MACS-purified B cells were cultured in 24-well plates at 106 cells/ml for 3 days under the following conditions: 1) medium alone, 2) LPS (2.5 μg/ml), or 3) IL-4 (50 ng/ml) plus anti-CD40 (15 μg/ml). Cells were harvested and plated at 5 x 103 cells/well and diluted serially 1/2 in multiscreen 96-well filtration plates (Millipore) coated with goat anti-mouse IgM (μ-chain specific; Jackson ImmunoResearch Laboratories) or goat anti-mouse IgG (-chain specific; CALTAG Laboratories) for 6 h at 37°C. Ig secreted by the plated cells was detected using biotinylated rabbit anti-mouse mAb (prepared in-house) and streptavidin-alkaline phosphatase (Southern Biotechnology Associates). The spots were visualized using NBT-5-bromo-4-chloro-3-indolyl phosphate substrate (Vector Laboratories).
Ag presentation assays
MACS-purified B cells (105 cells/well) were cocultured with 5 x 104 T cell hybridoma cells and various concentrations of Ag (hen egg lysosome (HEL) or OVA; purchased from Sigma-Aldrich) in RPMI 1640 supplemented with 10% FCS and with or without LPS (2.5 μg/ml). After 20–24 h at 37°C, levels of IL-2 were determined in culture supernatants using a sandwich ELISA with anti-IL-2 capture (JES6-1A12; eBioscience) and biotinylated anti-IL-2 detection (JES6-5H4; eBioscience) Abs. Bound detection Ab was elaborated with streptavidin-HRP and Supersignal Elisa Femto maximum sensitivity substrate (Pierce); luminescence was measured by using the Vector2 1420 multilabel counter (Wallac).
For Ag presentation to primary CD4 T cells, splenic CD4 T cells from OT-II mice were purified using anti-CD4 MACS. The Ag presentation assay to OT-II CD4 T cells was similar to that used for T cell hybridomas described above. In some experiments, polymixin B, which binds to and inactivates the lipid A component of bacterial endotoxin, was added to cultures at 10 μg/ml. At the end of culture, cells were harvested and stained with anti-B220 and anti-CD80 (B7.1), anti-CD86 (B7.2), or anti-MHC-II mAbs, and the expression profile of these molecules was evaluated by flow cytometry after gating on B220+ cells.
Results
Splenic canonical B cells are present in similar numbers in HKIR and HKI65 mice but express very different levels of sBCR
Splenic B cells purified by E4-MACS from HKIR and HKI65 mice and their derivatives obtained from crosses to lines with inactivated endogenous Igh loci (JHD and μMT) were used as sources of follicular B cells expressing canonical BCRs for the in vitro studies described here. Fig. 1A shows that the percentage of splenic B cells that are stained with the E4 anti-clonotypic mAb is similar in the two types of mice. As previously reported (34, 35), such B cells in HKIR mice express substantially reduced levels of sIgM and sIgD as compared with E4+ B cells in HKI65 mice and the bulk B cell compartment in littermate control mice (Fig. 1A, data from HKI/JHD lines are shown). Splenic E4+ B cells in HKIR mice display uniformly low levels of both isoforms of the BCR. However, sIgD levels on splenic HKI65 E4+ B cells are bimodal, and we previously speculated that this might be due to the presence of a subpopulation of slightly immature canonical B cells in these mice (34, 35). Analogous results were obtained from HKIR and HKI65 transgene hemizygous or HKIR/μMT and HKI65/μMT mice (data not shown).
FIGURE 1. Splenic canonical B cells are present in similar numbers in HKIR and HKI65 mice, but express different levels of sBCR. A, Splenocytes from the indicated mice were stained with E4, anti-IgM, and anti-IgD, and cells from nontransgenic littermates were stained with anti-B220, anti-IgM, and anti-IgD and analyzed by flow cytometry. The percentages of E4+ cells in the lymphocyte gate are shown (upper panels). The expression levels of sIgM and sIgD were analyzed after gating on HKIR E4+, HKI65 E4+, or B220+ cells from nontransgenic littermates (lower panels). These data are representative of at least three independent experiments of pooled samples of three mice of each genotype. B, Splenocytes from the indicated mice were stained with E4 and anti-IgD, and cells in the indicated gates were purified by FACS as described in Materials and Methods. The MFI values for sIgD staining of each population were then determined by reanalysis of the populations (right panels), and the cross-hairs in each panel indicate the center of the staining distribution, allowing the relative expression levels of E4 and sIg to be more easily compared. C, The FACS-purified B cells described in B were restained with anti-IgM and anti-, and levels of expression were evaluated by flow cytometry.
Since sIgD levels on HKI65 E4+ B cells are heterogeneous, more detailed flow cytometric studies were performed on this population. Fig. 1B shows that E4+ B cells in HKI65 mice are composed of two subpopulations, one that is E4highsIgDlow and one that is E4lowsIgDhigh. In contrast, whereas E4+ B cells in HKIR mice display some heterogeneity in levels of expression of the E4 idiotope and sIgD, only one major subpopulation of E4+sIgDlow B cells is evident. The mean fluorescence intensity (MFI) of sIgD staining on even the sIgDlow HKI65 subpopulation is higher than on HKIR E4+ cells.
To directly compare levels of sBCR expression on these three subpopulations, each was purified by FACS, restained with anti-IgM or anti- mAbs, and analyzed by flow cytometry. Fig. 1C shows that HKIR E4+ cells express the lowest levels of sIgD, sIgM, and sIg of all of the subpopulations. Despite expressing different amounts of sIgD, the two HKI65 E4+ subpopulations express similar levels of sIgM. In agreement with these results, the levels of sIg on the HKI65 E4high subpopulation were found to be intermediate between those on the HKIR E4+ and HKI65 E4low subpopulations.
Although the molecular basis for differences in levels of sBCR and E4 on the two subpopulations of HKI65 B cells remains to be determined, we currently favor the idea that this is due to the expression of distinct but related L chain V regions. This would predict that the two subpopulations differ in Ag specificity, and taken along with our previously published results (34) suggests that the E4highsIgDlow subpopulation is more autoreactive than the E4lowsIgDhigh subpopulation, a possibility that we are currently investigating.
These issues notwithstanding, the bulk (MACS-purified) HKI65 E4+ subpopulation would still serve as a useful control for in vitro B cell responsivity of HKIR E4+ B cells, if both E4+ subpopulations in HKI65 mice are composed of mature, follicular B cells. Therefore, we analyzed the levels of HSA (CD24), C1qRp (AA4.1), CD23, CD62L (L-selectin), CD21, and CD1d on these two subpopulations by flow cytometry. Both displayed levels of these markers characteristic of mature, follicular B cells; levels that were indistinguishable from one another (data not shown). Therefore, the bulk, MACS-purified HKI65 E4+ population was used as a control for all subsequent experiments performed on HKIR E4+ B cells.
HKIR E4+ B cells appear to be in a resting state in vivo
Upon encountering autoantigen in vivo, autoreactive B cells can be activated, leading to abortive proliferation and apoptosis (44, 45). If this were the case for a major fraction of HKIR E4+ B cells, the interpretation of in vitro activation and differentiation studies using these cells might be confounded. Therefore, to determine whether HKIR E4+ B cells respond to autoantigens in this way in vivo, we measured the cell cycle status of splenic HKIR E4+ B cells in naive mice. Elevation in the percentages of cycling (S-G2-M) or apoptotic cells was not detected as compared with HKI65 E4+ and nontransgenic IgMb+ control B cells (Table I). Most of the cells were in the G0-G1 phase, as expected of conventional follicular B cells. These data are consistent with our previous observation that HKIR E4+ splenic B cells do not incorporate detectable amounts of BrdU during a 1-h pulse in vivo (34).
Table I. Cell cycle analysis of B cells in vivoa
The R55 mutation does not generically alter the expression of knockin locus-derived mRNA or protein
Since HKIR B cells do not express substantially reduced levels of sBCR early in their primary development, we previously suggested that the increased levels of autoreactivity conferred to canonical BCRs by the R55 mutation was responsible for down-regulation of sBCR levels on mature E4+HKIR follicular B cells (34). However, it remained formally possible that this mutation perturbed the expression of the HKIR knockin locus due to direct effects on transcription, mRNA stability, or H chain translation, processing, or stability.
To further address this issue, MACS-purified HKIR E4+, HKI65 E4+, and littermate splenic B cells were stimulated with LPS to induce polyclonal activation. After 3 days, cells were assayed for amounts of Cμ mRNA by real-time RT-PCR. Fig. 2A shows that these levels were not significantly different in the three types of activated B cells. To evaluate H chain protein expression, supernatants from the LPS cultures were assayed for concentrations of total Ig and E4+ Ab by ELISAs. Fig. 2B shows that concentrations of secreted Ig in these cultures were indistinguishable, as were levels of E4+ Abs in the HKIR and HKI65 supernatants. These data strongly suggest that the presence of the R55 mutation does not intrinsically alter expression of the HKIR knockin locus, supporting our previous suggestion that the reduced levels of sBCR on follicular HKIR B cells results from the influence of this mutation on BCR specificity.
FIGURE 2. The R55 mutation does not directly alter expression of the HKIR knockin locus. Splenic E4+ and IgMb+ B cells were purified from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice by MACS as described in Materials and Methods. These B cells were stimulated with 2.5 μg/ml LPS at 37°C for 72 h. Cellular RNA was then extracted and real-time RT-PCR was performed on the RNA preparations as described in Materials and Methods. The results of quantification of Ig μ RNA are shown in A. The levels of secreted total Ig and E4+ Ig in the 72-h cultures were measured by sandwich ELISA (B, left panel) and competition ELISA (B, are right panel), respectively. These data are representative of two independent were experiments of pooled samples from four mice of each genotype.
HKIR E4+B cells are competent in BCR proximal signaling steps following sIgM cross-linking, but at reduced levels
Defects in BCR-mediated signaling have been suggested to be responsible for the suboptimal proliferation and Ab secretion of anergic B cells in response to anti-BCR stimulation in vitro (8, 13, 18). The molecular mechanism underlying this unresponsiveness was suggested to be Ag receptor complex destabilization, preventing Ig- and Syk phosphorylation (46). We compared the capacity of purified splenic HKIR E4+, HKI65 E4+, and nontransgenic littermate IgMb+ B cells to undergo protein tyrosine phosphorylation upon sIgM cross-linking. Syk and Ig- phosphorylation could be induced in HKIR E4+ B cells, although at reduced levels (Fig. 3).
FIGURE 3. Tyrosine phosphorylation of Syk and Ig- following anti-IgM treatment of splenic B cells. Splenic E4+ and IgMb+ B cells were purified from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice as described in Materials and Methods. B cells were stimulated with 10 μg/ml F(ab')2 goat anti-mouse IgM at 37°C for 3 min. Cells were then lysed, and 106 cell equivalents of whole lysate were resolved by SDS-PAGE and immunoblotted with an anti-phosphotyrosine mAb. The lower panels show a longer exposure of the region of the blot containing Ig- and a -actin loading control. The data are representative of three independent experiments.
Rapid elevation in intracellular Ca2+ concentration is one of the earliest markers of B cell activation (47). In many anergic B cells, stimulation with anti-Ig results in only a transient Ca2+ flux (17, 18, 46). Therefore, we next compared the ability of these B cells to mobilize intracellular calcium after stimulation by anti-IgM. HKIR E4+ B cells were able to flux Ca2+ rapidly upon anti-IgM stimulation and sustain this response for the same duration as control B cells (Fig. 4A). However, the amplitude of this response was reduced. Analogous results were obtained from HKIR and HKI65 E4+ B cells purified by E4-MACS (Fig. 4A, left panel) or present in bulk cultures that were stained with E4 at the time of anti-IgM addition (Fig. 4A, right panel). This suggests that the use of E4-MACS to purify these cells does not alter subsequent responses to BCR stimuli.
FIGURE 4. Intracellular calcium flux in HKIR E4+, HKI65 E4+, and littermate IgMb+ B cells in response to anti-IgM. A, Splenic E4+ or IgMb+ B cells isolated by MACS purification from HKIR/JHD, HKI65/JHD, or nontransgenic littermates mice (see Materials and Methods) were loaded with fluo-3-AM. The basal level of intracellular Ca2+ was monitored for the first 30 s before 25 μg/ml anti-IgM (Fab')2 was added (arrow). Alternatively, total splenocytes were used and E4+ or IgMb+ cells were elaborated by staining at 4°C just before the assay, followed by analysis via flow cytometry. Basal Ca2+ levels varied subtly in both cases from experiment to experiment but no reproducible trends were noted. The results are representative of four independent experiments; cells purified from two mice for each sample were used. The data are consistent with results obtained from transgene hemizygous HKIR and HKI65 mice in two other independent experiments. B, Experiments were conducted as described above on flow cytometric gated B cells from HKIR/μMT and HKI65/μMT mice and littermates, except using the indicated doses of anti-IgM (Fab')2. C, Experiments were conducted as described for A and B, but using the indicated B cell subpopulations that had been purified by FACS from HKIR/JHD and HKI65/JHD mice as described in Materials and Methods.
To determine whether the quantitatively reduced Ca2+ flux responses displayed by HKIR E4+ B cells might be a direct effect of their low levels of sIgM, flow cytometry gated HKIR E4+, HKI65 E4+, and littermate splenic B cells were stimulated with different concentrations of anti-IgM and mobilization of intracellular Ca2+ was monitored. Fig. 4B shows that as the dose of anti-IgM was increased, the differences in the magnitude of Ca2+ flux in the three types of B cells decreased. This idea was further tested by comparing the Ca2+ responses of FACS-purified HKIR E4+, HKI65 E4highsIgDlow, and HKI65 E4lowsIgDhigh splenic B cells, since the latter two subpopulations express similar amounts of sIgM (Fig. 1C). Fig. 4C shows that at a low dose of anti-IgM both HKI65 subpopulations fluxed Ca2+ to similar extents and in amounts higher than observed in HKIR E4+ B cells. In contrast, raising the dose of anti-IgM 3-fold resulted in indistinguishable Ca2+ flux responses among the three types of B cells.
HKIR E4+ B cells display somewhat reduced cell cycle progression in response to anti-IgM but not other mitogenic stimuli
We previously found that HKIR E4+ B cells proliferate suboptimally upon anti-IgM stimulation, as indicated by a [3H]TdR incorporation assay. We suggested that this was attributable to their low levels of sBCR, since these B cells responded to non-BCR stimuli in this assay as well as HKI65 E4+ and nontransgenic IgMb+ B cells. To evaluate the possibility that the reduced proliferation of these B cells was due to activation-induced cell death or proliferation of only a subpopulation of B cells in the cultures, we evaluated the cell cycle status of HKIR E4+ B cells via flow cytometry after stimulation with anti-IgM for 48 h. The percentage of apoptotic and necrotic cells in HKIR E4+ cultures was found to be comparable to that of control cultures, but the percentage of cycling HKIR E4+ B cells was somewhat lower than that in the control cultures (Table II). This suggests that HKIR E4+ B cells are indeed stimulated to undergo cell cycle progression to a slightly lesser extent than control B cells, but this is not due to increased levels of activation-induced cell death. As expected from our previous results, when HKIR E4+ B cells were stimulated with LPS or anti-CD40 plus IL-4, these B cells displayed levels of cell cycle progression comparable to those of control B cells.
Table II. Cell cycle analysis of B cells after in vitro stimulationa
HKIR E4+ B cells differentiate efficiently to IgM- and IgG-producing Ab-forming cells (AFCs)
Previous studies have demonstrated that most anergic B cells inefficiently differentiate to secretory phenotype following in vitro stimulation (48, 49, 50, 51, 52, 53). However, the data presented in Fig. 2 suggested this might not be the case for HKIR E4+ B cells, and therefore to further investigate this question we stimulated purified HKIR E4+, HKI65 E4+, and nontransgenic IgMb+ B cells with LPS, anti-IgM, and anti-IgM plus anti-CD40 and IL-4. IgM- or IgG-secreting cells in the cultures were enumerated 72 h later. As shown in Fig. 5, HKIR E4+ B cells did not differ from control B cells in their ability to produce IgM or IgG AFCs in response to LPS. All three types of B cells failed to give rise to many AFCs in response to anti-IgM alone (data not shown). However, all cells activated with anti-IgM in the presence of mitogenic anti-CD40 and IL-4 produced a robust AFC response, and HKIR E4+ B cells produced IgM- and IgG-secreting AFCs in numbers comparable to HKI65 E4+ and nontransgenic IgMb+ B cells.
FIGURE 5. In vitro differentiation of B cells to AFCs. AFC formation by splenic HKIR E4+ (), HKI65 E4+(), and nontransgenic littermate B cells () purified from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice were evaluated by ELISPOT assay. B cells were incubated in vitro in medium alone or with either LPS (2.5 μg/ml) or 25 μg/ml F(ab')2 anti-IgM plus 15 μg/ml anti-CD40 and 50 ng/ml IL-4 for 72 h at 37°C. Total IgM (A)- and IgG (B)-secreting B cells were detected by ELISPOT. Triplicate samples from each B cell population were analyzed. The data obtained are representative of three separate experiments, and pools of cells purified from three mice of each genotype for each experiment were used. These data are also consistent with results obtained from transgene hemizygous HKIR and HKI65 in three other independent experiments.
Up-regulation of costimulatory molecules on HKIR E4+, HKI65 E4+, and nontransgenic littermate IgMb+ B cells
B cells can act as potent APCs, in part, because of the ability of the BCR to mediate rapid and specific Ag uptake (54, 55, 56). However, productive T cell activation requires signals delivered by costimulatory molecules, which naive B cells appear to lack (57, 58, 59). Previous studies have shown that anergic B cells are deficient in their ability to up-regulate costimulatory molecules after anti-BCR stimulation (18, 60, 61, 62, 63). To examine the effect of BCR engagement on the expression of the CD28 counterreceptors CD80 (B7.1) and CD86 (B7.2) and other activation molecules, purified HKIR E4+, HKI65 E4+, and nontransgenic IgMb+ B cells were stimulated with anti-IgM for 16 h and surface levels of expression of these molecules were assayed by flow cytometry. Parallel cultures were also stimulated with LPS or anti-CD40 plus IL-4. All of these stimuli resulted in increased surface expression of both CD80 and CD86 as well as the induction of MHC-II and CD69 on HKIR E4+ B cells. Furthermore, the levels of surface expression of these molecules were comparable to nonautoreactive B cells both before and after stimulation (Fig. 6).
FIGURE 6. Up-regulation of costimulatory and activation markers on HKIR E4+ B cells. MACS-purified B cells from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice were activated in vitro for 16 h using the indicated stimuli and assayed for surface levels of the indicated activation or costimulatory molecules by flow cytometry. The data obtained from experiments in which polymixin B was not added to the medium are representative of those obtained in three independent experiments, and pools of cells purified from three mice per sample per experiment were used. These data are also consistent with results obtained from transgene hemizygous HKIR and HKI65 mice in one other independent experiment.
Since HKIR E4+ B cells respond to LPS in a manner indistinguishable from HKI65 E4+ and littermate B cells, contamination of the in vitro cultures by endotoxin might mask potential qualitative differences in responsiveness of HKIR E4+ and the control B cell populations to stimulation via the BCR. To address this possibility, the anti-IgM stimulations were repeated in the presence of polymixin B, a compound that blocks the effects of the lipid A component of LPS on B cells (64). The lower panels in Fig. 6 illustrate that the relative levels of induction of costimulatory molecules on HKIR E4+ B cells were unaltered by the presence of this agent.
HKIR E4+ B cells efficiently present Ag to a T cell hybridoma taken up nonspecifically, but not via the BCR
To further evaluate whether the reduced levels of sBCR characteristic of HKIR E4+ B cells might alter their ability to undergo cognate interaction with T cells, we measured their efficiency of presentation of OVA to the T cell hybridoma B17.10 (specific for OVA323–329 in the context of I-Ab). Purified HKIR E4+, HKI65 E4+, and littermate IgMb+ splenic B cells were incubated with B17.10 cells and varying concentrations of OVA for 20–24 h, and levels of IL-2 in culture supernatants were quantified via ELISA. HKIR E4+ B cells showed an Ag dose-dependent ability to induce IL-2 production from the T cell hybridoma comparable to that of control B cells. This was true either without (Fig. 7A) or with (Fig. 7B) addition of LPS to the cultures. Similar data were obtained using another Ag (HEL) and T cell hybridoma (BO4, specific for the epitope HEL74–78 in the context of I-Ab) (data not shown). This suggests that HKIR E4+ B cells have no intrinsic defect in their ability to process and present protein Ags.
FIGURE 7. B cell presentation of Ag to a T cell hybridoma. HKIR E4+, HKI65 E4+, and littermate IgMb+ splenic B cells were purified from HKIR/μMT, HKI65/μMT mice or nontransgenic littermates and incubated with varying concentrations of the indicated Ag and 5 x 104 BO17.10 T cell hybridoma cells for 20–24 h before quantifying the levels of IL-2 in culture supernatants (see Materials and Methods). Background levels of IL-2 secretion were evaluated by incubation of nontransgenic IgMb+ B cells with BO17.10 cells and various concentrations of BSA. The assay was performed in the absence (A and C) and presence (B and D) of 2.5 μg/ml LPS. The data are representative of three independent experiments. Pooled B cells purified from three mice of each genotype were used for each experiment.
To assess the potential of HKIR E4+ B cells to process and present Ag taken up via the BCR, we made the use of the fact that canonical BCRs containing and lacking the R55 mutation are "dual reactive" for both the hapten p-azophenylarsonate (Ars) and nuclear autoantigens. At low doses of an Ars-OVA conjugate, HKIR E4+ B cells gave rise to reduced levels of activation of the B17.10 hybridoma as compared with HKIR65 E4+ B cells (Fig. 6C). As expected, IgMb+ B cells were also inefficient in presenting this Ag at low concentration. When B cells were activated with LPS, the deficiency of HKIR E4+ and IgMb+ B cells in presentation was somewhat diminished (Fig. 6D).
HKIR E4+ B cells are efficient in presenting Ag to and activating primary CD4 T cells, whether or not the Ag is taken up via the BCR
The activation of T cell hybridomas by APCs does not require nor is enhanced by costimulation (60). Therefore, we next tested the ability of HKIR E4+ B cells to present Ags to primary CD4 T cells ex vivo. For this purpose, we used splenic CD4 T cells isolated from OVA-specific, class II-restricted TCR-transgenic mice (OT-II mice). Purified HKIR E4+ B cells presenting both OVA and Ars-OVA activated OT-II T cells to levels comparable to control B cells (Fig. 8A). As expected, control IgMb+ B cells presented OVA with comparable efficiency, but were far less capable in presentation of Ars-OVA. We next investigated the levels of surface expression of CD80 (B7.1), CD86 (B7.2), and MHC-II molecules on HKIR E4+ and control B cells after their incubation with OT-II T cells and Ag. Induction of all of these molecules on the surface of HKIR E4+ B cells reached levels similar to those displayed by control B cells (Table III and data not shown). Thus, in a situation mimicking cognate T cell-B cell interaction in vivo, HKIR E4+ B cells act as APCs as efficiently as nonautoreactive B cells.
FIGURE 8. B cell Ag presentation to primary T cells. A, The indicated types of MACS-purified splenic B cells from HKIR/μMT, HKI65/μMT, and littermate mice were incubated with OVA or Ars-OVA and purified splenic T cells from OT-II TCR-transgenic mice, and IL-2 secretion was assayed as described in Materials and Methods. The data are representative of three independent experiments. B, The same assay was performed as illustrated in A, with the exception that polymixin B (10 μg/ml) was used to inhibit endotoxin activity. Pooled CD4 T cells from two OT-II mice and pooled B cells purified from three mice of each genotype were used for each experiment.
Table III. B cell surface expression of CD80 (B7.1), CD86 (B7.2), and MHC-II molecules after incubation with Ag and Ag-specific primary T cellsa
As discussed above, contamination of cultures with endotoxin might mask differences in the ability of HKIR E4+ B cells to be functionally activated via the BCR as compared with control B cells. Since sBCR levels on HKIR E4+ and HKI65 E4+ follicular B cells (Fig. 1) correlate with degree of sBCR-mediated Ag processing and presentation to T cell hybridomas (Fig. 7), but not primary T cells (Fig. 8A), we addressed the issue of whether endotoxin contamination might account for this difference using polymixin B. Fig. 8B illustrates that although the addition of this compound appeared to have a generic inhibitory effect on IL-2 production in the cultures, the levels of IL-2 induced by HKIR E4+ and HKI65 E4+ B cells presenting either OVA or Ars-OVA to OT-II T cells were indistinguishable.
Discussion
Induction of B cell tolerance is a multifaceted process, involving a variety of "checkpoint" mechanisms operative at numerous stages of primary and Ag-driven B cell development (4, 6, 8, 10, 14). Among these, receptor editing and anergy have been documented to be associated with down-modulation of sBCR, particularly sIgM, levels (9, 50, 65, 66, 67). Unfortunately, past studies have not revealed whether there are functional consequences of this down-regulation or whether it simply represents an epiphenomenon of autoantigen engagement. Nonetheless, a variety of observations have suggested that the fate of developing B cells is critically dependent on both the amount and specificity of BCR that is expressed (4, 68). Indeed, peripheral survival requires the expression of a signaling-competent BCR complex (69), immature B cells expressing suboptimal levels of sBCR fail to progress in development (70, 71), and positive selection based on BCR specificity appears to play an important role in shaping the composition of the peripheral clonotype repertoire (72, 73, 74, 75, 76). Moreover, a body of anecdotal evidence has indicated that developing B cells are capable of altering levels of sBCR expression in response to changes in the signaling capacity of the BCR complex (68, 77, 78, 79, 80, 81, 82).
Our previous data on the fate of developing canonical R55 B cells in HKIR mice provided no evidence for their clonal deletion or developmental arrest, and these cells were observed to acquire a mature, follicular phenotype (34). Taken together with the results discussed above, this led us to propose that active down-regulation of sIgM on these cells upon autoantigen engagement in the bone marrow resulted in an adjustment of sBCR expression to levels resulting in signaling sufficient to promote developmental progression and survival, but inadequate to induce deletion, development arrest, or receptor editing. Subsequently, we imagined that levels of sIgD were also regulated in this fashion to ensure peripheral stability under conditions of chronic autoantigen engagement (34). However, these past studies did not carefully examine the question of whether follicular canonical R55 B cells harbor deficiencies in responsiveness to BCR or other stimuli, a distinct possibility given the previous results of others on the fate of peripheral, autoreactive clonotypes (8, 9, 10, 19, 24).
The data presented herein strongly suggest that the only limitation on the responsiveness of canonical R55 B cells in vitro results from their low levels of sBCR. Stimulation of HKIR E4+ splenic B cells with anti-IgM resulted in reduced levels of Syk and Ig- phosphorylation and Ca2+ responses that were blunted in magnitude as compared with HKI65 E4+ and nontransgenic control B cells. As expected from these reductions in proximal BCR signaling, levels of ensuing proliferation were somewhat reduced. Nonetheless, such stimulation did induce an intracellular Ca2+ response that was sustained and resulted in induction of costimulatory and activation markers in amounts comparable to those of controls. Moreover, increasing the dose of anti-IgM resulted in minimization of the amplitude differences in Ca2+ flux observed in canonical HKIR and control B cell populations.
Whether the quantitative differences in BCR responsiveness noted for HKIR E4+ B cells in vitro translates into alterations in in vivo responsiveness during an Ag-driven immune response remain to be thoroughly investigated. Treatment with anti-CD40 and IL-4 resulted in functional outcomes indistinguishable from control B cells. Analogous results were obtained with LPS stimulation. It might be argued that such conditions are not an accurate representation of the in vivo situation, as B cells with reduced sBCR levels might fail to compete effectively for the limiting amounts of T or other accessory cell help most probably available during the initial stages of the immune response to a TD Ag (83, 84, 85, 86). Indeed, we observed that HKIR E4+ B cells were less efficient than HKI65 E4+ B cells at presenting an Ars-protein conjugate to T cell hybridomas. However, this difference was not detected in Ag presentation assays where CD4 T cells from TCR-transgenic mice were used, suggesting that under conditions of cognate T-B interaction in which Ag-specific T cells are not limiting, the impact of this functional deficiency is mollified. Nonetheless, it is unlikely that the in vitro conditions we used for our studies truly mimic the events that result in recruitment of B cells into a T cell-dependent immune response in vivo.
Another important issue for future studies is whether the low levels of sBCR characteristic of canonical R55 B cells is reflective of a state of reduced "competitiveness" with other B cells for peripheral trophic factors required for normal behavior during immune responses in vivo (21, 22, 87). Our previous studies on HKIR mice containing a functional, endogenous Igh locus showed that 50% of peripheral B cells exclusively expressed this locus, apparently as a consequence of inactivation of the knockin locus due to Rag-mediated DNA rearrangements (34). Even when the development of such B cells is precluded by crosses to mice with inactivated Igh loci, the resulting B cell compartment is "semidiverse," due to expression of endogenous L chain V genes. Despite the presence of large "competitor" populations in both cases, canonical R55 B cells stably inhabit follicles in substantial numbers (35). Moreover, most of the experiments reported here were performed with canonical B cells obtained from both transgene hemizygous mice and mice in which the transgene was present in the context of an inactivated endogenous Igh locus. In all such experiments, the B cells from the different sources yielded analogous results (see figure legends). This suggests that potential alterations in trophic factor-driven processes in vivo do not overtly influence the ability of canonical R55 B cells to mount vigorous proliferative and differentiative responses in vitro.
Given the many previously characterized mechanisms contributing to B cell tolerance, what might be the utility of an additional pathway allowing the normal follicular development of certain autoreactive B cell clonotypes via sBCR down-regulation? Given the arguments above, we suggest that follicular B cells expressing very low levels of sBCR will not be efficiently recruited into the early AFC response because of reduced sBCR cross-linking and their inability to garner large amounts of T cell help as a consequence of low levels of Ag capture, processing, and presentation (88). Nonetheless, the data we present here suggest that such B cells are fully competent to undergo productive cognate interaction with CD4 T cells, a process that could result in their recruitment into the germinal center (GC) response. Once in the GC microenvironment, V gene hypermutation, combined with stringent clonal selection processes, might result in the emergence of mutant derivatives of the founding, autoreactive clones that had lost affinity for autoantigen, resulting in expression of high levels of sBCR, proficient Ag capture, processing, and presentation to GC T cells, and induction of memory cells and secondary AFCs. Such a process of "specificity maturation" (89, 90) might also facilitate the nucleation of the GC response by ensuring that not all B cells are driven to terminal differentiation during the initial stages of a TD immune response. We are presently testing these ideas by determining whether canonical R55 B cells exclusively enter the GC response after Ars immunization in vivo, and whether hypermutated progeny of this clonotype can contribute to the anamnestic AFC response and give rise to memory B cells expressing high levels of sBCR.
Disclosures
The authors have no financial conflict of interest.
Acknowledgments
We thank Scot Fenn for technical assistance, Bice Perussia for polymixin B, Kishore Kunapuli, Dr. G. J. Hammerling for the T cell hybridomas, Connie Krawczyk Jones, Dr. H. Shen for the OT-II mice, and all members of the Manser laboratory for their indirect contributions.
Footnotes
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 study was supported by National Institutes of Health Grant AI 38965 (to T.M.).
2 Address correspondence and reprint requests to Dr. Tim Manser, Department of Microbiology and Immunology and Kimmel Cancer Center, Jefferson Medical College, Bluemle Life Sciences Building, 708, 233 South 10th Street, Philadelphia, PA 19017-5541. E-mail address: manser{at}mail.jci.tju.edu
3 Abbreviations used in this paper: s, surface; AFC: Ab-forming cell; BAFF (BLys): B Cell-activating factor belonging to the TNF family (B lymphocyte stimulator); GC, germinal center; HEL, hen egg lysozyme; Ars, p-azophenylarsonate; fluo-3-AM, fluo-3-acetoxymethyl ester; MFI, mean fluorescence intensity; MHC-II, MHC class II; WT, wild type.
Received for publication July 26, 2004. Accepted for publication January 31, 2005.
References
Monroe, J. G.. 1996. Tolerance sensitivity of immature-stage B cells: can developmentally regulated B cell antigen receptor (BCR) signal transduction play a role?. J. Immunol. 156:2657
Nossal, G. J.. 1992. A lifetime‘s flirtation with repertoire purging. Res. Immunol. 143:268.
Klinman, N. R.. 1996. The "clonal selection hypothesis" and current concepts of B cell tolerance. Immunity 5:189.
Goodnow, C. C., J. G. Cyster, S. B. Hartley, S. E. Bell, M. P. Cooke, J. I. Healy, S. Akkaraju, J. C. Rathmell, S. L. Pogue, K. P. Shokat. 1995. Self-tolerance checkpoints in B lymphocyte development. Adv. Immunol. 59:279.
Tiegs, S. L., D. M. Russell, D. Nemazee. 1993. Receptor editing in self-reactive bone marrow B cells. J. Exp. Med. 177:1009.
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.
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.
Santulli-Marotto, S., M. W. Retter, R. Gee, M. J. Mamula, S. H. Clarke. 1998. Autoreactive B cell regulation: peripheral induction of developmental arrest by lupus-associated autoantigens. Immunity 8:209.
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.
Mandik-Nayak, L., A. Bui, H. Noorchashm, A. Eaton, J. Erikson. 1997. Regulation of anti-double-stranded DNA B cells in nonautoimmune mice: localization to the T-B interface of the splenic follicle. J. Exp. Med. 186:1257.
Lang, J., D. Nemazee. 2000. B cell clonal elimination induced by membrane-bound self-antigen may require repeated antigen encounter or cell competition. Eur. J. Immunol. 30:689.
Aplin, B. D., C. L. Keech, A. L. de Kauwe, T. P. Gordon, D. Cavill, J. McCluskey. 2003. Tolerance through indifference: autoreactive B cells to the nuclear antigen La show no evidence of tolerance in a transgenic model. J. Immunol. 171:5890.
Jenkins, M. K., R. H. Schwartz. 1987. Antigen presentation by chemically modified splenocytes induces antigen-specific T cell unresponsiveness in vitro and in vivo. J. Exp. Med. 165:302.
Li, Y., H. Li, M. Weigert. 2002. Autoreactive B cells in the marginal zone that express dual receptors. J. Exp. Med. 195:181.
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.
Cooke, M. P., A. W. Heath, K. M. Shokat, Y. Zeng, F. D. Finkelman, P. S. Linsley, M. Howard, C. C. Goodnow. 1994. Immunoglobulin signal transduction guides the specificity of B cell-T cell interactions and is blocked in tolerant self-reactive B cells. J. Exp. Med. 179:425.
Adams, E., A. Basten, C. C. Goodnow. 1990. Intrinsic B-cell hyporesponsiveness accounts for self-tolerance in lysozyme/anti-lysozyme double-transgenic mice. Proc. Natl. Acad. Sci. USA 87:5687
Thien, M., T. G. Phan, S. Gardam, M. Amesbury, A. Basten, F. Mackay, R. Brink. 2004. Excess BAFF rescues self-reactive B cells from peripheral deletion and allows them to enter forbidden follicular and marginal zone niches. Immunity 20:785.
Engel, P., L. J. Zhou, D. C. Ord, S. Sato, B. Koller, T. F. Tedder. 1995. Abnormal B lymphocyte development, activation, and differentiation in mice that lack or overexpress the CD19 signal transduction molecule. Immunity 3:39.
Kraus, M., L. I. Pao, A. Reichlin, Y. Hu, B. Canono, J. C. Cambier, M. C. Nussenzweig, K. Rajewsky. 2001. Interference with immunoglobulin (Ig) immunoreceptor tyrosine-based activation motif (ITAM) phosphorylation modulates or blocks B cell development, depending on the availability of an Ig cytoplasmic tail. J. Exp. Med. 194:455.
Otipoby, K. L., K. B. Andersson, K. E. Draves, S. J. Klaus, A. G. Farr, J. D. Kerner, R. M. Perlmutter, C. L. Law, E. A. Clark. 1996. CD22 regulates thymus-independent responses and the lifespan of B cells. Nature 384:634.
Wang, L. D., J. Lopes, A. B. Cooper, M. Dang-Lawson, L. Matsuuchi, M. R. Clark. 2004. Selection of B lymphocytes in the periphery is determined by the functional capacity of the B cell antigen receptor. Proc. Natl. Acad. Sci. USA 101:1027.
Manser, T., K. M. Tumas-Brundage, L. P. Casson, A. M. Giusti, S. Hande, E. Notidis, K. A. Vora. 1998. The roles of antibody variable region hypermutation and selection in the development of the memory B-cell compartment. Immunol. Rev. 162:183.
Hande, S., E. Notidis, T. Manser. 1998. Bcl-2 obstructs negative selection of autoreactive, hypermutated antibody V regions during memory B cell development. Immunity 8:189.(Xiaohe Liu and Tim Manser)
We previously demonstrated that B cells expressing a transgenic BCR with "dual reactivity" for the hapten arsonate and nuclear autoantigens efficiently complete development to follicular phenotype and stably reside in follicles in vivo. These B cells express very low levels of surface IgM and IgD, suggesting that they avoid central deletion and peripheral anergy by reducing their avidity for autoantigen via surface BCR (sBCR) down-regulation. Since a variety of states of B cell anergy have been previously described, a thorough examination of the functional capabilities of these B cells was required to test this hypothesis. In this study, we show that surface Ig cross-linking induces amounts of proximal BCR signaling in these B cells commensurate with their reduced sBCR levels. Functionally, however, they are comparable to nonautoreactive B cells in cell cycle progression, up-regulation of activation and costimulatory molecules, and Ab-forming cell differentiation when treated with a variety of stimuli in vitro. In addition, these B cells can efficiently process and present Ag and are capable of undergoing cognate interaction with naive TCR-transgenic T cells, resulting in robust IL-2 production. Together, these data reveal a lack of intrinsic anergy involving any known mechanism, supporting the idea that this type of antinuclear Ag B cell becomes indifferent to cognate autoantigen by down-regulating sBCR.
Introduction
Insight into B cell tolerance pathways were first provided by in vitro studies showing that engagement of the BCR on immature B cells or their transformed counterparts results in apoptotic death or unresponsiveness, rather than activation (1, 2, 3). The analysis of Ig-transgenic mice expressing autoreactive BCRs have confirmed and extended these in vitro studies (4). It is now clear that certain types of transgenic autoreactive BCRs are removed from the developing repertoire via clonal deletion and receptor editing, mainly in the bone marrow (4, 5, 6, 7). Nonetheless, deletion or editing mechanisms do not result in a complete "cleansing" of the B cell compartment of autoreactivity. Numerous investigations have shown that Ig-transgenic B cells with particular types of autospecificities, in some cases associated with autoimmune disease, can exit the bone marrow and populate the periphery (8, 9, 10, 11, 12, 13). However, these B cells often display distinct phenotypes and reduced responsiveness (4, 8, 10, 14), a condition commonly referred to as "anergy" (2, 3, 4, 15).
Despite extensive studies over the last decade on the physiological state of peripheral autoreactive B cells, a standard definition of anergy has not been established. Anergic B cells described in different experimental systems often respond to in vitro and in vivo stimuli in distinct ways (8, 14, 16, 17, 18, 19). This argues against a "black and white" definition of anergy, and suggests that autoantigen-induced changes in B cell behavior are graded, depending on factors such as autoantigen structure and concentration, the avidity of the BCR-autoantigen interaction, and the time during B cell development that autoantigen is first encountered. However, essentially all definitions of anergy include hyporesponsiveness to BCR-mediated stimulation. In addition, anergy is usually associated with reduced levels of surface (s) 3 IgM and in some cases sIgD (4). Many anergic B cells have reduced life spans and may be excluded from follicles, perhaps due to reduced ability to "homeostatically compete" with nonanergic B cells for peripheral trophic factors such as BAFF (BLys) (20, 21, 22). Moreover, some investigators have found that while certain peripheral autoreactive B cells can mount proliferative responses to a variety of in vitro stimuli, they inefficiently differentiate to Ab secretory cells (18, 23, 24), suggesting that this step in development may be a target of a tolerance checkpoint.
Despite their reduced responsiveness, many anergic B cells do undergo proliferation and differentiation in vitro and in vivo upon receipt of surrogate or genuine T cell help (25, 26, 27, 28, 29). Activation of innate immune system accessory cells or receptors expressed by B cells themselves may also contribute to "reversal" of the anergic state (30, 31, 32). Also, several recent studies have indicated that some autoreactive B cells complete peripheral development harboring no obvious proliferative or differentiative defects, suggesting that they do not functionally interact with their cognate autoantigen(s). Such B cells are often referred to as being in an "ignorant" or "indifferent" state (12, 33). These observations suggest that the presence of autoreactive B cells in the periphery poses a significant risk factor for the development of autoimmune disease. For this reason, how and to what extent B cells categorized as anergic, ignorant, or indifferent by current standards can respond to BCR, mitogenic, and costimulatory stimuli is worthy of detailed investigation.
We have previously described two lines of VH"knockin" mice that differ only in the presence or absence of a single mutation to arginine (R) at position 55 in the CDR2 subregion of the VH gene used to replace the endogenous JH locus (34, 35). We term these lines of mice HKIR and HKI65, respectively. Both versions of this VH gene, in combination with a single, endogenous L chain gene, encode Abs that we term "canonical." Canonical Abs bind the hapten arsonate and can be specifically detected using the monoclonal anti-clonotypic Ab E4 (36, 37). Abs with the R55 form of the V domain also display reactivity for nuclear autoantigens such as chromatin and can cause kidney dysfunction via glomerular deposition in vivo. B cells expressing BCRs containing both types of V domains develop to mature follicular phenotype, reside in follicles, and are not short-lived (34). However, peripheral B cells expressing the R55 form are characterized by substantially reduced levels of sIgM and sIgD as compared with B cells expressing the form lacking this mutation (34, 35). Along with the results of previous studies discussed above indicating a correlation between sBCR down-regulation and reduced responsiveness of autoreactive B cells, these latter data suggested that the follicular B cells in HKIR mice expressing the canonical R55 BCR might be anergic.
Materials and Methods
Mice
HKIR knockin mice, expressing the canonical R55 VH gene, and HKI65 mice, expressing the canonical VH lacking the R55 mutation, were created in an identical fashion by replacing the entire JH locus with a targeting vector differing only in the sequence of VH codon 55 (34, 35). For many experiments, these lines were crossed to lines in which the entire JH locus had been deleted (JHD mice) (38) (a gift from Dr. R. Hardy, Fox Chase Cancer Center, Philadelphia, PA, USA) or the IgM transmembrane exon had been disrupted (μMT mice) (39) to preclude the development of B cells expressing the endogenous IgH locus. OVA-specific class II-restricted TCR-transgenic mice (OT-II mice) on the C57BL/6 (B6) background were kindly provided by Dr. H. Shen (Department of Microbiology, University of Pennsylvania, Philadelphia, PA (with permission from Dr. W. Heath, Walter and Eliza Hall Institute, Parkville, Victoria, Australia)) (40). All mice were used between 8 and 12 wk of age. Mice were housed under specific pathogen-free conditions and given autoclaved food and water. The use of mice in these studies was conducted in compliance with institute guidelines, and all protocols using animals were approved by the institutional animal care and use committee.
Cells and cell lines
Splenic B cells were isolated as previously described (34). Briefly, single-cell suspensions were prepared from spleens of naive transgenic and age-matched transgene-negative littermates. After removal of dead cells by centrifugation through Lympholyte-M (Cedarlane Laboratories), cells were stained with either biotin-E4 or biotin-anti-IgMb Ab, followed by incubation with paramagnetic streptavidin beads (Miltenyi Biotec). Following the manufacturer’s recommendations, labeled cells were passed over MiniMACS columns (Miltenyi Biotec) and eluted. Two column passes per sample were performed. The purity of the E4+ B cells as assessed by flow cytometry was always >85%. The purity of IgMb+ B cells was always >95%. MHC class II (MHC-II) I-Ab haplotype-restricted T cell hybridoma cell lines BO17.10 (specific for epitope OVA323–339) and BO4 (specific for epitope HEL74–78)) (kindly provided by Dr. G. J. Hammerling, German Cancer Center, DKFZ, Heidelberg, Germany) were maintained as previously described (41).
Flow cytometry and FACS
Cells (106/sample) were stained according to a previously described protocol (34). The following Abs were used: anti-CD45R (RA3-6B2; eBioscience), anti-IgMb (AF6-78; Southern Biotechnology), anti-CD69 (H1.2F3; BD Pharmingen), anti-CD80 (16-10A1; BD Pharmingen), anti-CD86 (GL1; BD Pharmingen), anti-I-A/I-E (2G9; BD Pharmingen), anti- (187.1; Southern Biotechnology Associates), anti-C1qRp (AA4.1; eBioscience), anti-CD21/CD35 (7G6; BD Pharmingen), anti-CD23 (B3B4; BD Pharmingen), anti-CD24 (HSA,M1/69; BD Pharmingen), CD1d (Cd1.1,1B1; BD Pharmingen), anti-CD62L (Mel-14; BD Pharmingen), and anti-idiotypic mAb E4 (prepared in-house). Streptavidin-R670 (BD Pharmingen) was used to detect biotinylated Abs. For evaluating levels of sIgM and total sBCR and Ca2+ flux responses in HKIR E4+, HKI65 E4+IgDhigh, and HKI65 E4+IgDlow populations, splenocytes from HKIR/JHD and HKI65/JHD mice were stained with E4 and anti-IgD and these populations were purified using a MoFlo high-performance cell sorter (DakoCytomation). The sorted cells were then restained with anti-IgM or anti- or stimulated with anti-IgM and subjected to the Ca2+ flux assay (see below). All flow cytometric analyses were performed on an EPICS Elite (Coulter) flow cytometer, and data were analyzed using the FlowJo software (Tree Star).
Real-time RT-PCR
Ig RNA levels in B cells were measured using an ABI Prism 7000 sequence detection system (PerkinElmer). Purified B cells were activated with LPS for 24 or 72 h, cells were harvested, and RNA was extracted using the RNeasy Mini kit (Qiagen). All RNA templates were subjected to DNase digestion to exclude genomic DNA contamination following the protocol in the kit. The cDNA was subjected to PCR using primers and TaqMan probes (Custom Oligonucleotide Factory) designed to span introns to minimize amplification of genomic DNA and to result in amplicons of <150 bp. Probes were labeled at the 5' end with FAM and at the 3' end with 6-carboxytetramethyl-rhodamine. Relative RNA concentrations were determined by the number of cycles necessary to reach a certain amount of fluorescence in a test sample compared with a designated calibration sample (RNA from littermate IgMb+ B cells) after normalizing to the number of cycles required to amplify a control GAPDH target. The data were analyzed with the ABI Prism 7000 SDS software (Applied Biosystems).
ELISA
Total Ig or E4 Ab was measured in culture supernatants as previously described (42). ELISA using 96-well plates (Immulon-4; Dynatech Laboratories) coated with either goat anti-mouse IgG + IgM (H + L) or E4 Abs were performed.
Ca2+ influx measurements
Fluo-3 AM loading of purified B cells and surface immunostaining of unfractionated splenocytes was performed using the method described before, with minor modifications (43). Cells were resuspended at 2 x 106/ml in HBSS with 1 μM fluo-3-acetoxymethyl ester (fluo-3 AM) and 0.5 μg/ml pluronic F-127 (Molecular Probes) and incubated for 30 min at 37°C. For surface immunostaining of fluo-3-AM-loaded splenocytes, cells were washed three times in HBSS containing 1% FCS and first incubated with biotinylated mAb E4 or anti-IgMb in FACS buffer (PBS/3% BSA) for 20 min at 4°C. Then Streptavidin-R670 was added and the cells were incubated as described above. The cells were then washed twice and diluted to 106 cells/ml with HBSS/1% FCS. The cell suspensions were placed in a water bath at 37°C for 15 min before flow cytometric analysis. Calcium mobilization was induced by adding various concentrations of goat anti-mouse IgM (Fab')2 (Pierce) and monitored by FACS either on total purified B cells or by gating on the E4+ or IgMb+ cells.
Immunoblotting
Levels of protein tyrosine phosphorylation in MACS-purified B cells were detected following stimulation with 10 μg/ml goat anti-mouse IgM (Fab')2 at 37°C for 3 min. Cells were then chilled on ice immediately and pelleted by centrifugation. For Western blotting, cell pellets were lysed in ice-cold lysis buffer (20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM Na2 EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM -glycerophosphate, 1 mM Na3VO4, and 1 μg/ml leupeptin). The buffer was also supplemented with 1 mM PMSF and a complete protease inhibitor tablet (Roche). The whole cell lysates were subjected to 10% SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Millipore) using a Mini Trans-Blot electrophoretic transfer apparatus (Bio-Rad). The blots were blocked with TBS containing 5% milk and incubated with anti-phosphotyrosine-HRP mAb (PY-99; Santa Cruz Biotechnology). Ab binding was detected using the ECL (Pierce) substrate. As a loading control, membranes were stripped and reprobed with an anti -actin Ab (Novus Biological) and developed as described above.
Cell cycle analysis
Splenocyte suspensions were prepared as described above. Cells were washed once with cold PBS with 0.1% sodium azide and labeled with FITC-conjugated B220 and biotinylated E4 or anti-IgMb in cold staining buffer, and biotinylated E4 was revealed by streptavidin-PE. Cells were then washed with cold PBS containing 0.1% sodium azide and resuspended in 0.3 ml of 50% FBS in PBS. Labeled cells were fixed by incubating with 0.9 ml of cold 70% EtOH overnight at 4°C, and cells were then washed twice with cold PBS/azide to remove EtOH and precipitated protein. 7-Aminoactinomycin D (20 μg/ml in PBS) was then added, and E4+ B cells or IgMb+ cells were analyzed using flow cytometry by gating on B220+E4+ and B220+IgMb+ populations. For cell cycle analyses of activated B cells, E4+ or IgMb+ B cells were purified as described above and stimulated with LPS (2.5 μg/ml), anti-IgM (Fab')2 (25 μg/ml), or anti-CD40 (15 μg/ml) plus IL-4 (50 ng/ml) in culture at 37°C for 48 h. Cells were then harvested and stained with 50 μg/ml propidium iodide in PBS. DNA content per cell was then analyzed using flow cytometry and the FlowJo software (Tree Star).
ELISPOT assays
MACS-purified B cells were cultured in 24-well plates at 106 cells/ml for 3 days under the following conditions: 1) medium alone, 2) LPS (2.5 μg/ml), or 3) IL-4 (50 ng/ml) plus anti-CD40 (15 μg/ml). Cells were harvested and plated at 5 x 103 cells/well and diluted serially 1/2 in multiscreen 96-well filtration plates (Millipore) coated with goat anti-mouse IgM (μ-chain specific; Jackson ImmunoResearch Laboratories) or goat anti-mouse IgG (-chain specific; CALTAG Laboratories) for 6 h at 37°C. Ig secreted by the plated cells was detected using biotinylated rabbit anti-mouse mAb (prepared in-house) and streptavidin-alkaline phosphatase (Southern Biotechnology Associates). The spots were visualized using NBT-5-bromo-4-chloro-3-indolyl phosphate substrate (Vector Laboratories).
Ag presentation assays
MACS-purified B cells (105 cells/well) were cocultured with 5 x 104 T cell hybridoma cells and various concentrations of Ag (hen egg lysosome (HEL) or OVA; purchased from Sigma-Aldrich) in RPMI 1640 supplemented with 10% FCS and with or without LPS (2.5 μg/ml). After 20–24 h at 37°C, levels of IL-2 were determined in culture supernatants using a sandwich ELISA with anti-IL-2 capture (JES6-1A12; eBioscience) and biotinylated anti-IL-2 detection (JES6-5H4; eBioscience) Abs. Bound detection Ab was elaborated with streptavidin-HRP and Supersignal Elisa Femto maximum sensitivity substrate (Pierce); luminescence was measured by using the Vector2 1420 multilabel counter (Wallac).
For Ag presentation to primary CD4 T cells, splenic CD4 T cells from OT-II mice were purified using anti-CD4 MACS. The Ag presentation assay to OT-II CD4 T cells was similar to that used for T cell hybridomas described above. In some experiments, polymixin B, which binds to and inactivates the lipid A component of bacterial endotoxin, was added to cultures at 10 μg/ml. At the end of culture, cells were harvested and stained with anti-B220 and anti-CD80 (B7.1), anti-CD86 (B7.2), or anti-MHC-II mAbs, and the expression profile of these molecules was evaluated by flow cytometry after gating on B220+ cells.
Results
Splenic canonical B cells are present in similar numbers in HKIR and HKI65 mice but express very different levels of sBCR
Splenic B cells purified by E4-MACS from HKIR and HKI65 mice and their derivatives obtained from crosses to lines with inactivated endogenous Igh loci (JHD and μMT) were used as sources of follicular B cells expressing canonical BCRs for the in vitro studies described here. Fig. 1A shows that the percentage of splenic B cells that are stained with the E4 anti-clonotypic mAb is similar in the two types of mice. As previously reported (34, 35), such B cells in HKIR mice express substantially reduced levels of sIgM and sIgD as compared with E4+ B cells in HKI65 mice and the bulk B cell compartment in littermate control mice (Fig. 1A, data from HKI/JHD lines are shown). Splenic E4+ B cells in HKIR mice display uniformly low levels of both isoforms of the BCR. However, sIgD levels on splenic HKI65 E4+ B cells are bimodal, and we previously speculated that this might be due to the presence of a subpopulation of slightly immature canonical B cells in these mice (34, 35). Analogous results were obtained from HKIR and HKI65 transgene hemizygous or HKIR/μMT and HKI65/μMT mice (data not shown).
FIGURE 1. Splenic canonical B cells are present in similar numbers in HKIR and HKI65 mice, but express different levels of sBCR. A, Splenocytes from the indicated mice were stained with E4, anti-IgM, and anti-IgD, and cells from nontransgenic littermates were stained with anti-B220, anti-IgM, and anti-IgD and analyzed by flow cytometry. The percentages of E4+ cells in the lymphocyte gate are shown (upper panels). The expression levels of sIgM and sIgD were analyzed after gating on HKIR E4+, HKI65 E4+, or B220+ cells from nontransgenic littermates (lower panels). These data are representative of at least three independent experiments of pooled samples of three mice of each genotype. B, Splenocytes from the indicated mice were stained with E4 and anti-IgD, and cells in the indicated gates were purified by FACS as described in Materials and Methods. The MFI values for sIgD staining of each population were then determined by reanalysis of the populations (right panels), and the cross-hairs in each panel indicate the center of the staining distribution, allowing the relative expression levels of E4 and sIg to be more easily compared. C, The FACS-purified B cells described in B were restained with anti-IgM and anti-, and levels of expression were evaluated by flow cytometry.
Since sIgD levels on HKI65 E4+ B cells are heterogeneous, more detailed flow cytometric studies were performed on this population. Fig. 1B shows that E4+ B cells in HKI65 mice are composed of two subpopulations, one that is E4highsIgDlow and one that is E4lowsIgDhigh. In contrast, whereas E4+ B cells in HKIR mice display some heterogeneity in levels of expression of the E4 idiotope and sIgD, only one major subpopulation of E4+sIgDlow B cells is evident. The mean fluorescence intensity (MFI) of sIgD staining on even the sIgDlow HKI65 subpopulation is higher than on HKIR E4+ cells.
To directly compare levels of sBCR expression on these three subpopulations, each was purified by FACS, restained with anti-IgM or anti- mAbs, and analyzed by flow cytometry. Fig. 1C shows that HKIR E4+ cells express the lowest levels of sIgD, sIgM, and sIg of all of the subpopulations. Despite expressing different amounts of sIgD, the two HKI65 E4+ subpopulations express similar levels of sIgM. In agreement with these results, the levels of sIg on the HKI65 E4high subpopulation were found to be intermediate between those on the HKIR E4+ and HKI65 E4low subpopulations.
Although the molecular basis for differences in levels of sBCR and E4 on the two subpopulations of HKI65 B cells remains to be determined, we currently favor the idea that this is due to the expression of distinct but related L chain V regions. This would predict that the two subpopulations differ in Ag specificity, and taken along with our previously published results (34) suggests that the E4highsIgDlow subpopulation is more autoreactive than the E4lowsIgDhigh subpopulation, a possibility that we are currently investigating.
These issues notwithstanding, the bulk (MACS-purified) HKI65 E4+ subpopulation would still serve as a useful control for in vitro B cell responsivity of HKIR E4+ B cells, if both E4+ subpopulations in HKI65 mice are composed of mature, follicular B cells. Therefore, we analyzed the levels of HSA (CD24), C1qRp (AA4.1), CD23, CD62L (L-selectin), CD21, and CD1d on these two subpopulations by flow cytometry. Both displayed levels of these markers characteristic of mature, follicular B cells; levels that were indistinguishable from one another (data not shown). Therefore, the bulk, MACS-purified HKI65 E4+ population was used as a control for all subsequent experiments performed on HKIR E4+ B cells.
HKIR E4+ B cells appear to be in a resting state in vivo
Upon encountering autoantigen in vivo, autoreactive B cells can be activated, leading to abortive proliferation and apoptosis (44, 45). If this were the case for a major fraction of HKIR E4+ B cells, the interpretation of in vitro activation and differentiation studies using these cells might be confounded. Therefore, to determine whether HKIR E4+ B cells respond to autoantigens in this way in vivo, we measured the cell cycle status of splenic HKIR E4+ B cells in naive mice. Elevation in the percentages of cycling (S-G2-M) or apoptotic cells was not detected as compared with HKI65 E4+ and nontransgenic IgMb+ control B cells (Table I). Most of the cells were in the G0-G1 phase, as expected of conventional follicular B cells. These data are consistent with our previous observation that HKIR E4+ splenic B cells do not incorporate detectable amounts of BrdU during a 1-h pulse in vivo (34).
Table I. Cell cycle analysis of B cells in vivoa
The R55 mutation does not generically alter the expression of knockin locus-derived mRNA or protein
Since HKIR B cells do not express substantially reduced levels of sBCR early in their primary development, we previously suggested that the increased levels of autoreactivity conferred to canonical BCRs by the R55 mutation was responsible for down-regulation of sBCR levels on mature E4+HKIR follicular B cells (34). However, it remained formally possible that this mutation perturbed the expression of the HKIR knockin locus due to direct effects on transcription, mRNA stability, or H chain translation, processing, or stability.
To further address this issue, MACS-purified HKIR E4+, HKI65 E4+, and littermate splenic B cells were stimulated with LPS to induce polyclonal activation. After 3 days, cells were assayed for amounts of Cμ mRNA by real-time RT-PCR. Fig. 2A shows that these levels were not significantly different in the three types of activated B cells. To evaluate H chain protein expression, supernatants from the LPS cultures were assayed for concentrations of total Ig and E4+ Ab by ELISAs. Fig. 2B shows that concentrations of secreted Ig in these cultures were indistinguishable, as were levels of E4+ Abs in the HKIR and HKI65 supernatants. These data strongly suggest that the presence of the R55 mutation does not intrinsically alter expression of the HKIR knockin locus, supporting our previous suggestion that the reduced levels of sBCR on follicular HKIR B cells results from the influence of this mutation on BCR specificity.
FIGURE 2. The R55 mutation does not directly alter expression of the HKIR knockin locus. Splenic E4+ and IgMb+ B cells were purified from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice by MACS as described in Materials and Methods. These B cells were stimulated with 2.5 μg/ml LPS at 37°C for 72 h. Cellular RNA was then extracted and real-time RT-PCR was performed on the RNA preparations as described in Materials and Methods. The results of quantification of Ig μ RNA are shown in A. The levels of secreted total Ig and E4+ Ig in the 72-h cultures were measured by sandwich ELISA (B, left panel) and competition ELISA (B, are right panel), respectively. These data are representative of two independent were experiments of pooled samples from four mice of each genotype.
HKIR E4+B cells are competent in BCR proximal signaling steps following sIgM cross-linking, but at reduced levels
Defects in BCR-mediated signaling have been suggested to be responsible for the suboptimal proliferation and Ab secretion of anergic B cells in response to anti-BCR stimulation in vitro (8, 13, 18). The molecular mechanism underlying this unresponsiveness was suggested to be Ag receptor complex destabilization, preventing Ig- and Syk phosphorylation (46). We compared the capacity of purified splenic HKIR E4+, HKI65 E4+, and nontransgenic littermate IgMb+ B cells to undergo protein tyrosine phosphorylation upon sIgM cross-linking. Syk and Ig- phosphorylation could be induced in HKIR E4+ B cells, although at reduced levels (Fig. 3).
FIGURE 3. Tyrosine phosphorylation of Syk and Ig- following anti-IgM treatment of splenic B cells. Splenic E4+ and IgMb+ B cells were purified from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice as described in Materials and Methods. B cells were stimulated with 10 μg/ml F(ab')2 goat anti-mouse IgM at 37°C for 3 min. Cells were then lysed, and 106 cell equivalents of whole lysate were resolved by SDS-PAGE and immunoblotted with an anti-phosphotyrosine mAb. The lower panels show a longer exposure of the region of the blot containing Ig- and a -actin loading control. The data are representative of three independent experiments.
Rapid elevation in intracellular Ca2+ concentration is one of the earliest markers of B cell activation (47). In many anergic B cells, stimulation with anti-Ig results in only a transient Ca2+ flux (17, 18, 46). Therefore, we next compared the ability of these B cells to mobilize intracellular calcium after stimulation by anti-IgM. HKIR E4+ B cells were able to flux Ca2+ rapidly upon anti-IgM stimulation and sustain this response for the same duration as control B cells (Fig. 4A). However, the amplitude of this response was reduced. Analogous results were obtained from HKIR and HKI65 E4+ B cells purified by E4-MACS (Fig. 4A, left panel) or present in bulk cultures that were stained with E4 at the time of anti-IgM addition (Fig. 4A, right panel). This suggests that the use of E4-MACS to purify these cells does not alter subsequent responses to BCR stimuli.
FIGURE 4. Intracellular calcium flux in HKIR E4+, HKI65 E4+, and littermate IgMb+ B cells in response to anti-IgM. A, Splenic E4+ or IgMb+ B cells isolated by MACS purification from HKIR/JHD, HKI65/JHD, or nontransgenic littermates mice (see Materials and Methods) were loaded with fluo-3-AM. The basal level of intracellular Ca2+ was monitored for the first 30 s before 25 μg/ml anti-IgM (Fab')2 was added (arrow). Alternatively, total splenocytes were used and E4+ or IgMb+ cells were elaborated by staining at 4°C just before the assay, followed by analysis via flow cytometry. Basal Ca2+ levels varied subtly in both cases from experiment to experiment but no reproducible trends were noted. The results are representative of four independent experiments; cells purified from two mice for each sample were used. The data are consistent with results obtained from transgene hemizygous HKIR and HKI65 mice in two other independent experiments. B, Experiments were conducted as described above on flow cytometric gated B cells from HKIR/μMT and HKI65/μMT mice and littermates, except using the indicated doses of anti-IgM (Fab')2. C, Experiments were conducted as described for A and B, but using the indicated B cell subpopulations that had been purified by FACS from HKIR/JHD and HKI65/JHD mice as described in Materials and Methods.
To determine whether the quantitatively reduced Ca2+ flux responses displayed by HKIR E4+ B cells might be a direct effect of their low levels of sIgM, flow cytometry gated HKIR E4+, HKI65 E4+, and littermate splenic B cells were stimulated with different concentrations of anti-IgM and mobilization of intracellular Ca2+ was monitored. Fig. 4B shows that as the dose of anti-IgM was increased, the differences in the magnitude of Ca2+ flux in the three types of B cells decreased. This idea was further tested by comparing the Ca2+ responses of FACS-purified HKIR E4+, HKI65 E4highsIgDlow, and HKI65 E4lowsIgDhigh splenic B cells, since the latter two subpopulations express similar amounts of sIgM (Fig. 1C). Fig. 4C shows that at a low dose of anti-IgM both HKI65 subpopulations fluxed Ca2+ to similar extents and in amounts higher than observed in HKIR E4+ B cells. In contrast, raising the dose of anti-IgM 3-fold resulted in indistinguishable Ca2+ flux responses among the three types of B cells.
HKIR E4+ B cells display somewhat reduced cell cycle progression in response to anti-IgM but not other mitogenic stimuli
We previously found that HKIR E4+ B cells proliferate suboptimally upon anti-IgM stimulation, as indicated by a [3H]TdR incorporation assay. We suggested that this was attributable to their low levels of sBCR, since these B cells responded to non-BCR stimuli in this assay as well as HKI65 E4+ and nontransgenic IgMb+ B cells. To evaluate the possibility that the reduced proliferation of these B cells was due to activation-induced cell death or proliferation of only a subpopulation of B cells in the cultures, we evaluated the cell cycle status of HKIR E4+ B cells via flow cytometry after stimulation with anti-IgM for 48 h. The percentage of apoptotic and necrotic cells in HKIR E4+ cultures was found to be comparable to that of control cultures, but the percentage of cycling HKIR E4+ B cells was somewhat lower than that in the control cultures (Table II). This suggests that HKIR E4+ B cells are indeed stimulated to undergo cell cycle progression to a slightly lesser extent than control B cells, but this is not due to increased levels of activation-induced cell death. As expected from our previous results, when HKIR E4+ B cells were stimulated with LPS or anti-CD40 plus IL-4, these B cells displayed levels of cell cycle progression comparable to those of control B cells.
Table II. Cell cycle analysis of B cells after in vitro stimulationa
HKIR E4+ B cells differentiate efficiently to IgM- and IgG-producing Ab-forming cells (AFCs)
Previous studies have demonstrated that most anergic B cells inefficiently differentiate to secretory phenotype following in vitro stimulation (48, 49, 50, 51, 52, 53). However, the data presented in Fig. 2 suggested this might not be the case for HKIR E4+ B cells, and therefore to further investigate this question we stimulated purified HKIR E4+, HKI65 E4+, and nontransgenic IgMb+ B cells with LPS, anti-IgM, and anti-IgM plus anti-CD40 and IL-4. IgM- or IgG-secreting cells in the cultures were enumerated 72 h later. As shown in Fig. 5, HKIR E4+ B cells did not differ from control B cells in their ability to produce IgM or IgG AFCs in response to LPS. All three types of B cells failed to give rise to many AFCs in response to anti-IgM alone (data not shown). However, all cells activated with anti-IgM in the presence of mitogenic anti-CD40 and IL-4 produced a robust AFC response, and HKIR E4+ B cells produced IgM- and IgG-secreting AFCs in numbers comparable to HKI65 E4+ and nontransgenic IgMb+ B cells.
FIGURE 5. In vitro differentiation of B cells to AFCs. AFC formation by splenic HKIR E4+ (), HKI65 E4+(), and nontransgenic littermate B cells () purified from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice were evaluated by ELISPOT assay. B cells were incubated in vitro in medium alone or with either LPS (2.5 μg/ml) or 25 μg/ml F(ab')2 anti-IgM plus 15 μg/ml anti-CD40 and 50 ng/ml IL-4 for 72 h at 37°C. Total IgM (A)- and IgG (B)-secreting B cells were detected by ELISPOT. Triplicate samples from each B cell population were analyzed. The data obtained are representative of three separate experiments, and pools of cells purified from three mice of each genotype for each experiment were used. These data are also consistent with results obtained from transgene hemizygous HKIR and HKI65 in three other independent experiments.
Up-regulation of costimulatory molecules on HKIR E4+, HKI65 E4+, and nontransgenic littermate IgMb+ B cells
B cells can act as potent APCs, in part, because of the ability of the BCR to mediate rapid and specific Ag uptake (54, 55, 56). However, productive T cell activation requires signals delivered by costimulatory molecules, which naive B cells appear to lack (57, 58, 59). Previous studies have shown that anergic B cells are deficient in their ability to up-regulate costimulatory molecules after anti-BCR stimulation (18, 60, 61, 62, 63). To examine the effect of BCR engagement on the expression of the CD28 counterreceptors CD80 (B7.1) and CD86 (B7.2) and other activation molecules, purified HKIR E4+, HKI65 E4+, and nontransgenic IgMb+ B cells were stimulated with anti-IgM for 16 h and surface levels of expression of these molecules were assayed by flow cytometry. Parallel cultures were also stimulated with LPS or anti-CD40 plus IL-4. All of these stimuli resulted in increased surface expression of both CD80 and CD86 as well as the induction of MHC-II and CD69 on HKIR E4+ B cells. Furthermore, the levels of surface expression of these molecules were comparable to nonautoreactive B cells both before and after stimulation (Fig. 6).
FIGURE 6. Up-regulation of costimulatory and activation markers on HKIR E4+ B cells. MACS-purified B cells from HKIR/JHD, HKI65/JHD, or nontransgenic littermate mice were activated in vitro for 16 h using the indicated stimuli and assayed for surface levels of the indicated activation or costimulatory molecules by flow cytometry. The data obtained from experiments in which polymixin B was not added to the medium are representative of those obtained in three independent experiments, and pools of cells purified from three mice per sample per experiment were used. These data are also consistent with results obtained from transgene hemizygous HKIR and HKI65 mice in one other independent experiment.
Since HKIR E4+ B cells respond to LPS in a manner indistinguishable from HKI65 E4+ and littermate B cells, contamination of the in vitro cultures by endotoxin might mask potential qualitative differences in responsiveness of HKIR E4+ and the control B cell populations to stimulation via the BCR. To address this possibility, the anti-IgM stimulations were repeated in the presence of polymixin B, a compound that blocks the effects of the lipid A component of LPS on B cells (64). The lower panels in Fig. 6 illustrate that the relative levels of induction of costimulatory molecules on HKIR E4+ B cells were unaltered by the presence of this agent.
HKIR E4+ B cells efficiently present Ag to a T cell hybridoma taken up nonspecifically, but not via the BCR
To further evaluate whether the reduced levels of sBCR characteristic of HKIR E4+ B cells might alter their ability to undergo cognate interaction with T cells, we measured their efficiency of presentation of OVA to the T cell hybridoma B17.10 (specific for OVA323–329 in the context of I-Ab). Purified HKIR E4+, HKI65 E4+, and littermate IgMb+ splenic B cells were incubated with B17.10 cells and varying concentrations of OVA for 20–24 h, and levels of IL-2 in culture supernatants were quantified via ELISA. HKIR E4+ B cells showed an Ag dose-dependent ability to induce IL-2 production from the T cell hybridoma comparable to that of control B cells. This was true either without (Fig. 7A) or with (Fig. 7B) addition of LPS to the cultures. Similar data were obtained using another Ag (HEL) and T cell hybridoma (BO4, specific for the epitope HEL74–78 in the context of I-Ab) (data not shown). This suggests that HKIR E4+ B cells have no intrinsic defect in their ability to process and present protein Ags.
FIGURE 7. B cell presentation of Ag to a T cell hybridoma. HKIR E4+, HKI65 E4+, and littermate IgMb+ splenic B cells were purified from HKIR/μMT, HKI65/μMT mice or nontransgenic littermates and incubated with varying concentrations of the indicated Ag and 5 x 104 BO17.10 T cell hybridoma cells for 20–24 h before quantifying the levels of IL-2 in culture supernatants (see Materials and Methods). Background levels of IL-2 secretion were evaluated by incubation of nontransgenic IgMb+ B cells with BO17.10 cells and various concentrations of BSA. The assay was performed in the absence (A and C) and presence (B and D) of 2.5 μg/ml LPS. The data are representative of three independent experiments. Pooled B cells purified from three mice of each genotype were used for each experiment.
To assess the potential of HKIR E4+ B cells to process and present Ag taken up via the BCR, we made the use of the fact that canonical BCRs containing and lacking the R55 mutation are "dual reactive" for both the hapten p-azophenylarsonate (Ars) and nuclear autoantigens. At low doses of an Ars-OVA conjugate, HKIR E4+ B cells gave rise to reduced levels of activation of the B17.10 hybridoma as compared with HKIR65 E4+ B cells (Fig. 6C). As expected, IgMb+ B cells were also inefficient in presenting this Ag at low concentration. When B cells were activated with LPS, the deficiency of HKIR E4+ and IgMb+ B cells in presentation was somewhat diminished (Fig. 6D).
HKIR E4+ B cells are efficient in presenting Ag to and activating primary CD4 T cells, whether or not the Ag is taken up via the BCR
The activation of T cell hybridomas by APCs does not require nor is enhanced by costimulation (60). Therefore, we next tested the ability of HKIR E4+ B cells to present Ags to primary CD4 T cells ex vivo. For this purpose, we used splenic CD4 T cells isolated from OVA-specific, class II-restricted TCR-transgenic mice (OT-II mice). Purified HKIR E4+ B cells presenting both OVA and Ars-OVA activated OT-II T cells to levels comparable to control B cells (Fig. 8A). As expected, control IgMb+ B cells presented OVA with comparable efficiency, but were far less capable in presentation of Ars-OVA. We next investigated the levels of surface expression of CD80 (B7.1), CD86 (B7.2), and MHC-II molecules on HKIR E4+ and control B cells after their incubation with OT-II T cells and Ag. Induction of all of these molecules on the surface of HKIR E4+ B cells reached levels similar to those displayed by control B cells (Table III and data not shown). Thus, in a situation mimicking cognate T cell-B cell interaction in vivo, HKIR E4+ B cells act as APCs as efficiently as nonautoreactive B cells.
FIGURE 8. B cell Ag presentation to primary T cells. A, The indicated types of MACS-purified splenic B cells from HKIR/μMT, HKI65/μMT, and littermate mice were incubated with OVA or Ars-OVA and purified splenic T cells from OT-II TCR-transgenic mice, and IL-2 secretion was assayed as described in Materials and Methods. The data are representative of three independent experiments. B, The same assay was performed as illustrated in A, with the exception that polymixin B (10 μg/ml) was used to inhibit endotoxin activity. Pooled CD4 T cells from two OT-II mice and pooled B cells purified from three mice of each genotype were used for each experiment.
Table III. B cell surface expression of CD80 (B7.1), CD86 (B7.2), and MHC-II molecules after incubation with Ag and Ag-specific primary T cellsa
As discussed above, contamination of cultures with endotoxin might mask differences in the ability of HKIR E4+ B cells to be functionally activated via the BCR as compared with control B cells. Since sBCR levels on HKIR E4+ and HKI65 E4+ follicular B cells (Fig. 1) correlate with degree of sBCR-mediated Ag processing and presentation to T cell hybridomas (Fig. 7), but not primary T cells (Fig. 8A), we addressed the issue of whether endotoxin contamination might account for this difference using polymixin B. Fig. 8B illustrates that although the addition of this compound appeared to have a generic inhibitory effect on IL-2 production in the cultures, the levels of IL-2 induced by HKIR E4+ and HKI65 E4+ B cells presenting either OVA or Ars-OVA to OT-II T cells were indistinguishable.
Discussion
Induction of B cell tolerance is a multifaceted process, involving a variety of "checkpoint" mechanisms operative at numerous stages of primary and Ag-driven B cell development (4, 6, 8, 10, 14). Among these, receptor editing and anergy have been documented to be associated with down-modulation of sBCR, particularly sIgM, levels (9, 50, 65, 66, 67). Unfortunately, past studies have not revealed whether there are functional consequences of this down-regulation or whether it simply represents an epiphenomenon of autoantigen engagement. Nonetheless, a variety of observations have suggested that the fate of developing B cells is critically dependent on both the amount and specificity of BCR that is expressed (4, 68). Indeed, peripheral survival requires the expression of a signaling-competent BCR complex (69), immature B cells expressing suboptimal levels of sBCR fail to progress in development (70, 71), and positive selection based on BCR specificity appears to play an important role in shaping the composition of the peripheral clonotype repertoire (72, 73, 74, 75, 76). Moreover, a body of anecdotal evidence has indicated that developing B cells are capable of altering levels of sBCR expression in response to changes in the signaling capacity of the BCR complex (68, 77, 78, 79, 80, 81, 82).
Our previous data on the fate of developing canonical R55 B cells in HKIR mice provided no evidence for their clonal deletion or developmental arrest, and these cells were observed to acquire a mature, follicular phenotype (34). Taken together with the results discussed above, this led us to propose that active down-regulation of sIgM on these cells upon autoantigen engagement in the bone marrow resulted in an adjustment of sBCR expression to levels resulting in signaling sufficient to promote developmental progression and survival, but inadequate to induce deletion, development arrest, or receptor editing. Subsequently, we imagined that levels of sIgD were also regulated in this fashion to ensure peripheral stability under conditions of chronic autoantigen engagement (34). However, these past studies did not carefully examine the question of whether follicular canonical R55 B cells harbor deficiencies in responsiveness to BCR or other stimuli, a distinct possibility given the previous results of others on the fate of peripheral, autoreactive clonotypes (8, 9, 10, 19, 24).
The data presented herein strongly suggest that the only limitation on the responsiveness of canonical R55 B cells in vitro results from their low levels of sBCR. Stimulation of HKIR E4+ splenic B cells with anti-IgM resulted in reduced levels of Syk and Ig- phosphorylation and Ca2+ responses that were blunted in magnitude as compared with HKI65 E4+ and nontransgenic control B cells. As expected from these reductions in proximal BCR signaling, levels of ensuing proliferation were somewhat reduced. Nonetheless, such stimulation did induce an intracellular Ca2+ response that was sustained and resulted in induction of costimulatory and activation markers in amounts comparable to those of controls. Moreover, increasing the dose of anti-IgM resulted in minimization of the amplitude differences in Ca2+ flux observed in canonical HKIR and control B cell populations.
Whether the quantitative differences in BCR responsiveness noted for HKIR E4+ B cells in vitro translates into alterations in in vivo responsiveness during an Ag-driven immune response remain to be thoroughly investigated. Treatment with anti-CD40 and IL-4 resulted in functional outcomes indistinguishable from control B cells. Analogous results were obtained with LPS stimulation. It might be argued that such conditions are not an accurate representation of the in vivo situation, as B cells with reduced sBCR levels might fail to compete effectively for the limiting amounts of T or other accessory cell help most probably available during the initial stages of the immune response to a TD Ag (83, 84, 85, 86). Indeed, we observed that HKIR E4+ B cells were less efficient than HKI65 E4+ B cells at presenting an Ars-protein conjugate to T cell hybridomas. However, this difference was not detected in Ag presentation assays where CD4 T cells from TCR-transgenic mice were used, suggesting that under conditions of cognate T-B interaction in which Ag-specific T cells are not limiting, the impact of this functional deficiency is mollified. Nonetheless, it is unlikely that the in vitro conditions we used for our studies truly mimic the events that result in recruitment of B cells into a T cell-dependent immune response in vivo.
Another important issue for future studies is whether the low levels of sBCR characteristic of canonical R55 B cells is reflective of a state of reduced "competitiveness" with other B cells for peripheral trophic factors required for normal behavior during immune responses in vivo (21, 22, 87). Our previous studies on HKIR mice containing a functional, endogenous Igh locus showed that 50% of peripheral B cells exclusively expressed this locus, apparently as a consequence of inactivation of the knockin locus due to Rag-mediated DNA rearrangements (34). Even when the development of such B cells is precluded by crosses to mice with inactivated Igh loci, the resulting B cell compartment is "semidiverse," due to expression of endogenous L chain V genes. Despite the presence of large "competitor" populations in both cases, canonical R55 B cells stably inhabit follicles in substantial numbers (35). Moreover, most of the experiments reported here were performed with canonical B cells obtained from both transgene hemizygous mice and mice in which the transgene was present in the context of an inactivated endogenous Igh locus. In all such experiments, the B cells from the different sources yielded analogous results (see figure legends). This suggests that potential alterations in trophic factor-driven processes in vivo do not overtly influence the ability of canonical R55 B cells to mount vigorous proliferative and differentiative responses in vitro.
Given the many previously characterized mechanisms contributing to B cell tolerance, what might be the utility of an additional pathway allowing the normal follicular development of certain autoreactive B cell clonotypes via sBCR down-regulation? Given the arguments above, we suggest that follicular B cells expressing very low levels of sBCR will not be efficiently recruited into the early AFC response because of reduced sBCR cross-linking and their inability to garner large amounts of T cell help as a consequence of low levels of Ag capture, processing, and presentation (88). Nonetheless, the data we present here suggest that such B cells are fully competent to undergo productive cognate interaction with CD4 T cells, a process that could result in their recruitment into the germinal center (GC) response. Once in the GC microenvironment, V gene hypermutation, combined with stringent clonal selection processes, might result in the emergence of mutant derivatives of the founding, autoreactive clones that had lost affinity for autoantigen, resulting in expression of high levels of sBCR, proficient Ag capture, processing, and presentation to GC T cells, and induction of memory cells and secondary AFCs. Such a process of "specificity maturation" (89, 90) might also facilitate the nucleation of the GC response by ensuring that not all B cells are driven to terminal differentiation during the initial stages of a TD immune response. We are presently testing these ideas by determining whether canonical R55 B cells exclusively enter the GC response after Ars immunization in vivo, and whether hypermutated progeny of this clonotype can contribute to the anamnestic AFC response and give rise to memory B cells expressing high levels of sBCR.
Disclosures
The authors have no financial conflict of interest.
Acknowledgments
We thank Scot Fenn for technical assistance, Bice Perussia for polymixin B, Kishore Kunapuli, Dr. G. J. Hammerling for the T cell hybridomas, Connie Krawczyk Jones, Dr. H. Shen for the OT-II mice, and all members of the Manser laboratory for their indirect contributions.
Footnotes
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 study was supported by National Institutes of Health Grant AI 38965 (to T.M.).
2 Address correspondence and reprint requests to Dr. Tim Manser, Department of Microbiology and Immunology and Kimmel Cancer Center, Jefferson Medical College, Bluemle Life Sciences Building, 708, 233 South 10th Street, Philadelphia, PA 19017-5541. E-mail address: manser{at}mail.jci.tju.edu
3 Abbreviations used in this paper: s, surface; AFC: Ab-forming cell; BAFF (BLys): B Cell-activating factor belonging to the TNF family (B lymphocyte stimulator); GC, germinal center; HEL, hen egg lysozyme; Ars, p-azophenylarsonate; fluo-3-AM, fluo-3-acetoxymethyl ester; MFI, mean fluorescence intensity; MHC-II, MHC class II; WT, wild type.
Received for publication July 26, 2004. Accepted for publication January 31, 2005.
References
Monroe, J. G.. 1996. Tolerance sensitivity of immature-stage B cells: can developmentally regulated B cell antigen receptor (BCR) signal transduction play a role?. J. Immunol. 156:2657
Nossal, G. J.. 1992. A lifetime‘s flirtation with repertoire purging. Res. Immunol. 143:268.
Klinman, N. R.. 1996. The "clonal selection hypothesis" and current concepts of B cell tolerance. Immunity 5:189.
Goodnow, C. C., J. G. Cyster, S. B. Hartley, S. E. Bell, M. P. Cooke, J. I. Healy, S. Akkaraju, J. C. Rathmell, S. L. Pogue, K. P. Shokat. 1995. Self-tolerance checkpoints in B lymphocyte development. Adv. Immunol. 59:279.
Tiegs, S. L., D. M. Russell, D. Nemazee. 1993. Receptor editing in self-reactive bone marrow B cells. J. Exp. Med. 177:1009.
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.
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.
Santulli-Marotto, S., M. W. Retter, R. Gee, M. J. Mamula, S. H. Clarke. 1998. Autoreactive B cell regulation: peripheral induction of developmental arrest by lupus-associated autoantigens. Immunity 8:209.
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.
Mandik-Nayak, L., A. Bui, H. Noorchashm, A. Eaton, J. Erikson. 1997. Regulation of anti-double-stranded DNA B cells in nonautoimmune mice: localization to the T-B interface of the splenic follicle. J. Exp. Med. 186:1257.
Lang, J., D. Nemazee. 2000. B cell clonal elimination induced by membrane-bound self-antigen may require repeated antigen encounter or cell competition. Eur. J. Immunol. 30:689.
Aplin, B. D., C. L. Keech, A. L. de Kauwe, T. P. Gordon, D. Cavill, J. McCluskey. 2003. Tolerance through indifference: autoreactive B cells to the nuclear antigen La show no evidence of tolerance in a transgenic model. J. Immunol. 171:5890.
Jenkins, M. K., R. H. Schwartz. 1987. Antigen presentation by chemically modified splenocytes induces antigen-specific T cell unresponsiveness in vitro and in vivo. J. Exp. Med. 165:302.
Li, Y., H. Li, M. Weigert. 2002. Autoreactive B cells in the marginal zone that express dual receptors. J. Exp. Med. 195:181.
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.
Cooke, M. P., A. W. Heath, K. M. Shokat, Y. Zeng, F. D. Finkelman, P. S. Linsley, M. Howard, C. C. Goodnow. 1994. Immunoglobulin signal transduction guides the specificity of B cell-T cell interactions and is blocked in tolerant self-reactive B cells. J. Exp. Med. 179:425.
Adams, E., A. Basten, C. C. Goodnow. 1990. Intrinsic B-cell hyporesponsiveness accounts for self-tolerance in lysozyme/anti-lysozyme double-transgenic mice. Proc. Natl. Acad. Sci. USA 87:5687
Thien, M., T. G. Phan, S. Gardam, M. Amesbury, A. Basten, F. Mackay, R. Brink. 2004. Excess BAFF rescues self-reactive B cells from peripheral deletion and allows them to enter forbidden follicular and marginal zone niches. Immunity 20:785.
Engel, P., L. J. Zhou, D. C. Ord, S. Sato, B. Koller, T. F. Tedder. 1995. Abnormal B lymphocyte development, activation, and differentiation in mice that lack or overexpress the CD19 signal transduction molecule. Immunity 3:39.
Kraus, M., L. I. Pao, A. Reichlin, Y. Hu, B. Canono, J. C. Cambier, M. C. Nussenzweig, K. Rajewsky. 2001. Interference with immunoglobulin (Ig) immunoreceptor tyrosine-based activation motif (ITAM) phosphorylation modulates or blocks B cell development, depending on the availability of an Ig cytoplasmic tail. J. Exp. Med. 194:455.
Otipoby, K. L., K. B. Andersson, K. E. Draves, S. J. Klaus, A. G. Farr, J. D. Kerner, R. M. Perlmutter, C. L. Law, E. A. Clark. 1996. CD22 regulates thymus-independent responses and the lifespan of B cells. Nature 384:634.
Wang, L. D., J. Lopes, A. B. Cooper, M. Dang-Lawson, L. Matsuuchi, M. R. Clark. 2004. Selection of B lymphocytes in the periphery is determined by the functional capacity of the B cell antigen receptor. Proc. Natl. Acad. Sci. USA 101:1027.
Manser, T., K. M. Tumas-Brundage, L. P. Casson, A. M. Giusti, S. Hande, E. Notidis, K. A. Vora. 1998. The roles of antibody variable region hypermutation and selection in the development of the memory B-cell compartment. Immunol. Rev. 162:183.
Hande, S., E. Notidis, T. Manser. 1998. Bcl-2 obstructs negative selection of autoreactive, hypermutated antibody V regions during memory B cell development. Immunity 8:189.(Xiaohe Liu and Tim Manser)