|Home | About | Journals | Submit | Contact Us | Français|
Signaling from the BCR and B cell activating factor receptor (BAFF-R or BR3) differentially regulates apoptosis within early transitional (T1) and late transitional (T2; CD21int-T2) B cells during selection processes to generate mature B lymphocytes. However, molecular mechanisms underlying the differential sensitivity of transitional B cells to apoptosis remain unclear. In this study, we demonstrate that BCR signaling induced more long-term c-Rel activation in T2 and mature than in T1 B cells leading to increased expression of anti-apoptotic genes as well as prosurvival BAFF-R and its downstream substrate p100 (NF-κB2). Sustained c-Rel activation required de novo c-Rel gene transcription and translation via Btk-dependent mechanisms. Like T1 cells, mature B cells from Btk- and c-Rel-deficient mice also failed to activate these genes. These findings suggest that the gain of survival potential within transitional B cells is dependent on the ability to produce a long-term c-Rel response, which plays a critical role in T2 B cell survival and differentiation in vivo by inducing anti-apoptotic genes, BAFF-R and NF-κB2, an essential component for BAFF-R survival signaling. Thus, acquisition of resistance to apoptosis during transitional B cell maturation is achieved by integration of BCR and BAFF-R signals.
Development of B lymphocytes in the bone marrow yields surface IgM+ immature B cells that immigrate to the spleen as early transitional (T1) B cells. Selection and maturation of transitional B cells continues in the periphery leading to the generation of at least mature follicular (Fo B)6 and marginal zone (MZ) B cells likely via late transitional T2-follicular (CD21int-T2) precursor stage (1–5). Peripheral differentiation is accompanied by increasing resistance to passive or BCR-induced apoptosis (1–5). In this regard, T1 cells are the most sensitive and have been shown to die in response to BCR cross-linking (1, 4, 6, 7). They are therefore the likely targets of negative selection, which weeds out self-reactive B cells (8–10). Prior studies also indicate that positive selection of transitional B cells contributes to generation of the mature B cell repertoire (2, 11, 12). Because this process involves selection of BCR specificities, it is believed to occur in late transitional or T2 cells that have reduced sensitivity to death in response to BCR cross-linking (4).
B cells gain a second form of survival advantage during positive selection by becoming responsive to B cell activating factor (BAFF) (4, 13–15). Loss of BAFF or BAFF-receptor 3 (BAFF-R or BR3) results in a severe B cell deficiency, particularly affecting T2 and mature B cells (16–18). Recent studies show that CD21int-T2 B cells express higher levels of BAFF-R and display enhanced responsiveness to BAFF relative to T1 B cells (4). BAFF-R activates several signaling pathways, including the canonical and noncanonical NF-κB pathways that increase B cell survival (13, 15, 19–21). Thus, differential survival competence in response to both BCR and BAFF-R engagement is a hallmark of the T1- to T2 transition. However, the mechanisms that underlie this “gain of survival potential” remain unclear.
To account for the hypersensitivity of T1 cells to BCR cross-linking, we and others have compared the signaling characteristics and biological responses of T1, T2, and mature Fo B cells (1, 3, 4, 6, 7, 15). The unifying feature of these studies is that T1 cells respond poorly to BCR cross-linking compared with T2 cells in many biochemical and biological assays. However, it remains unclear how this hypersensitivity is implemented to cause enhanced T1 cell death. In this regard, studies with B cells deficient in BCR signal transducer, Bruton’s tyrosine kinase (Btk) show that, like T1 cells, Btk-deficient B cells undergo apoptosis in response to BCR cross-linking (6, 7, 22). This similarity may be significant as, like wild-type (WT) T1 B cells, Btk-deficient B cells fail to induce Bcl-xL and A1 anti-apoptotic genes of the Bcl-2 family in response to BCR signaling (6, 7, 22). Impaired activation of these genes in T1 cells likely contributes to their extreme sensitivity to BCR-induced cell death, as evidenced by the relative resistance of B cells from Bim-deficient or Bcl-xL transgenic mice to apoptosis (23, 24). In this regard, the NF-κB family member c-Rel has been previously shown to regulate Bcl-xL and A1 gene expression in B cells (25). Consistent with an anti-apoptotic function of c-Rel, mature B cells deficient in c-Rel also display extreme sensitivity to BCR-induced cell death (25–28). Thus, B cells deficient for Btk and c-Rel behave similarly to T1 cells in their response to BCR cross-linking. We therefore considered the possibility that differential c-Rel activity is under Btk control and may at least in part explain the extreme sensitivity of T1 cells to BCR-induced death.
In this study, we show that sustained BCR-induced nuclear c-Rel activity depends on de novo c-Rel gene transcription and translation, and that this process occurs in T2 and mature Fo B but not in T1 cells. Btk deficiency blocks c-Rel induction in mature B cells, implicating Btk-dependent signaling as the basis for the differential regulation of c-Rel in T1 and T2 B cells. Importantly, c-Rel also regulates BCR-inducible expression of BAFF-R and substrate p100 of the noncanonical NF-κB pathway, thereby providing additional anti-apoptotic functions in T2 cells. We propose that gain of c-Rel inducibility at the T1 to T2 transition contributes to increased B cell resistance to apoptosis by two distinct mechanisms, by directly regulating the expression of anti-apoptotic genes and by up-regulating expression of BAFF-R and its substrates.
The generation of Btk-, c-Rel-deficient mice as well as B cell-specific bcl-2 transgenic mice has been previously described (29–31). C57BL/6 mice from The Jackson Laboratory were used as a source of WT B cell populations. Mice were treated humanely in accordance with federal and state government guidelines and their use was approved by the institutional animal committee.
Primary B lymphocytes from C57BL/6 mice were isolated using MiniMACS magnetic sorting by negative selection (CD43 depletion) to avoid inadvertent activation of B cells as previously described (3, 6). B cell purity was 90–95% based on flow cytometric (FCM) analysis following CD19 staining.
Splenic B cell subsets were purified based on a combination of schemes previously described (1, 3, 4, 6, 32, 33). Highly purified T1 and T2 B cells were obtained by MACS enrichment of AA4.1+ transitional B cells followed by FACS purification using Abs directed against CD23, CD24, and CD21. The purity of T1 (AA4.1+HSAhighCD21−CD23−) and T2 (AA4.1+HSAhighCD21+CD23+) B cell populations was 85–92 and 85–95%, respectively, as determined by FCM (see Fig. 2A). Using this sorting scheme, T2-preMZ (AA4.1low/−HSAhighCD21highCD23+) B cells were excluded from sorted T1 and T2 B cells. Thus, the T2 population analyzed in this study is the same as we and Meyer-Bahlburg et al. have recently reported (CD21intT2; Refs. 3 and 4). Mature Fo B cells were obtained by MACS-enrichment of the AA4.1− fraction with anti-CD43 conjugated microbeads (Miltenyi Biotec) followed by FACS purification using the same Ab combination as the T1 and T2 sorts. The purity of Fo B cells (AA4.1−HSAlowCD21+CD23+) was 95–100%, as determined by FCM. B cell subsets from Bcl-2 transgenic (Tg) mice were also purified using this scheme as shown in Fig. 3A. In experiments with Bcl-2 Tg B cell subsets (Fig. 3C), CD23+ cells were selected from AA4.1+ depleted splenocytes.
In the data presented in Fig. 4D, bcl-2 Tg T1 B cells were purified using a series of MACS depletions and positive selections, which included CD62L depletion followed by AA4.1 positive selection. The T2/M cells were obtained by MACS enrichment of CD23+ cells. The purity of T1 (CD62−AA4.1+) and T2 (CD23+) B cell populations was 65–90 and 90%, respectively, as determined by FCM.
Purified B cells were cultured in RPMI 1640 (HyClone Laboratories) supplemented with 10% FCS, 55 nM 2-ME, 2 mM L-glutamine, and 100 IU penicillin/streptomycin in a 37°C humidified incubator.
For kinetics studies of cRel and RelA, whole-cell extracts were prepared from total splenic B cells following lysis with RIPA buffer, and nuclear extracts were prepared as described previously (34). Ten micrograms of whole-cell extracts or 3 μg of nuclear extracts were separated on SDS-PAGE gel and subjected to Western blot analysis. Abs for RelA, cRel, PKCμ (Santa Cruz Biotechnology), and TATA box binding protein (Ab-cam) were used according to the manufacturer’s instructions.
For studies performed with btk−/−, c-rel−/− or Bcl-2 Tg mice, cells were stimulated with anti-IgM (Jackson ImmunoResearch Laboratories) or anti-CD40 from (BD Pharmigen). After culture, btk−/− live cells were obtained using Lympholite M (Cedarlane Laboratories). Whole cell extracts for these experiments were resolved as described in Ref. 35 and immunoblotted with anti-cRel (sc-70; see Fig. 6B) or (sc-71; see Figs. 1F and and3C),3C), anti-p100/p52 (sc-7386) p38 (sc-535) from Santa Cruz Biotechnology and anti-β-tubulin (T0198) was purchased from Sigma-Aldrich. For quantitative analysis of c-Rel protein by immunoblotting, c-Rel and p38 (loading control) were visualized by infrared imaging (Odyssey, Licor) using anti-Ig Alexa 670 (Invitrogen).
For the Bcl-2 family protein analyses, cellular extracts were immunoblotted with anti-A1 (R&D Systems) anti-Bcl-xL (BD Transduction Laboratory), anti-Bcl-2 (Santa Cruz Biotechnology), anti-MCl-1 from (Rockland Immunochemicals), anti-p38, and β-actin (sc-1616) purchased from Santa Cruz Biotechnology.
Equal amounts of nuclear extracts (1.0 μg) were prepared with the Nuclear and Cytoplasmic Extraction Reagent kit (NE-PER; Pierce), and were pre-incubated for 20 min at room temperature in the presence or absence of polyclonal Abs specific for p50(sc-1192X), c-Rel(sc-70X) from Santa Cruz Biotechnology. Subsequently, [γ-32P]ATP radio-labeled probe derived from κB enhancer sequences in the IL-2R promoter (36) was added and incubated on ice for 15 min. DNA binding reactions were performed as previously described (37). DNA protein complexes were resolved by electrophoresis on 4% native polyacrylamide gels and exposed to x-ray film. Total NF-κB nuclear activity was quantified and the composition of the NF-κB complexes identified using a phosphoimager (FujiFilm Medical Systems).
RNA was extracted from T1, T2, M, or total B cells using the RNeasy Mini Kit (Qiagen) and used to synthesize cDNA. For real-time PCR, Taq-man Universal Master Mix (Applied Biosystems) and Stratagene Max 3000p Detection Systems were used. Primers and FAM-labeled probes were obtained from Applied Biosystems (TaqMan Assay on Demand): NF-κB2 (Mm00479807), c-Rel (Mm00485657_m1), RelB (Mm00485672), Bcl-xL (Mm00437783_m1), and BAFF-R (Mm00840578_g1). The sequence of the primers and probe used for A1 were also obtained from (AB): A1 forward, CAGGAGAATGGATACGGCAGA; A1 reverse, CAGATCTGTCCTGTCATCTGCAG; (MGB): TCTTCCCAACCTCCA (38). RNA was quantified before use in RT-PCR for experiments that used Bcl-2 Tg mice and for kinetics with WT, btk−/− and crel−/− B cells. Data are expressed as the relative RNA concentrations relative to 18S or GAPDH expression and calibrated to nonstimulated cells.
MACS-enriched B cells were stimulated with 10 μg/ml goat anti-mouse IgM F(ab′)2 Abs for 18 h, stained for CD19, CD23, CD24, and CD21 for identification of T1 (CD19+CD23−CD24highCD21−), T2 (CD19+CD23+ CD24highCD21int), and mature Fo B cells (CD19+CD23+CD24low CD21+). After staining, the cells were fixed with 2% paraformaldehyde, permeabilized with 0.3% Triton X-100, and stained with AlexaFluor 647 labeled anti-c-Rel (SC-71). Cold competition was preformed by staining first with unlabeled anti-c-Rel, followed by staining with AlexaFluor 647 anti-c-Rel. Data were acquired on BD LSR II flow cytometer.
Data were compared with Student’s t test. All data are represented as mean ± SEM where indicated. Values of *, p ≤ 0.05 were considered statistically significant.
Prior studies have shown that several anti-apoptotic genes are regulated by c-Rel in mature B cells after BCR stimulation (26, 27). Because T1 and T2 B cells display distinct survival properties in response to BCR cross-linking, we explored the possibility that c-Rel was differentially regulated in T1 and T2 B cell populations. We first examined the c-Rel response of mature splenic B cells to BCR stimulation. CD43-depleted naive splenic B cells were activated with anti-IgM for various times followed by analysis of nuclear or whole cell extracts by immunoblotting. We found that BCR stimulation resulted in long-term c-Rel, but not p65/RelA, nuclear expression (Fig. 1A). Extended c-Rel induction coincided with increased total cellular c-Rel protein levels (Fig. 1B) that was likely mediated by increased c-Rel gene transcription (Fig. 1C). Total p65/RelA levels did not change over this time course (data not shown). These results are consistent with and extend our previous observation in T cells (39). We conclude that BCR stimulation activates two distinct phases of nuclear c-Rel. To test whether anti-apoptotic gene expression coincides with c-Rel induction, we performed kinetic studies of A1 and Bcl-xL in WT and c-rel−/− B cells. The kinetics of A1 and Bcl-xL appear to largely coincide with the biphasic nature of BCR-induced c-Rel (Fig. 1, D–F). The correspondence of c-Rel induction to A1 and Bcl-xL gene expression suggested that inducible transcription of the c-Rel gene may be a key factor in its effectiveness as an anti-apoptotic transcription factor.
After determining that c-Rel was induced upon BCR stimulation, we compared c-Rel gene transcription in transitional B cell subsets. T1, T2, and mature Fo B cells were purified by flow cytometry (Fig. 2A) as previously described (3, 4), activated with anti-IgM for 4 h before c-Rel mRNA analyses. We observed increased c-Rel transcripts in T2 and mature B cells, but not in T1 cells (Fig. 2B). Consistent with this, BCR-induced expression of anti-apoptotic genes A1 and Bcl-xL was also more increased in T2 and mature B cells relative to T1 B cells (Fig. 2C). As a control, activation of T1 cells via CD40 efficiently induced these anti-apoptotic genes excluding the possibility that T1 B cells are generally nonresponsive to stimulation (data not shown). As expected, this difference between T1 and T2 cells also reflected in increased NF-κB DNA binding activity in T2 relative to T1 cells (Fig. 2, D and E). The reduced NF-κB DNA binding activity that we observe in T1 cells are consistent with the findings previously reported using immature transitional B cells (25). Our results extend these findings by analyzing distinct populations within transitional B cells. Based on our findings in the mature splenic B cells (Fig. 1) and an increase at early time point (Fig. 2, B and C) we asked whether T1, T2, and mature B cells display differences in the anti-apoptotic gene expression after long-term BCR stimulation. Levels of several anti-apoptotic proteins were analyzed in FACS-sorted T1, T2, and mature Fo B cells after 16 h of culture with or without anti-IgM. Like mature splenic B cells a BCR signaling-dependent increase was observed in T2 and mature B cells (Fig. 2F). In contrast, T1 B cells which did not show an increase in the transcripts for c-Rel or A1 and Bcl-xL also did not display an increase in the protein levels encoded by these anti-apoptotic genes. These results are consistent with a higher caspase activity and sensitivity to apoptosis of T1 cells after long-term BCR stimulation and may reflect a relatively higher rate of cell death in T1 B cells. Together, these data suggest that unlike T1 and like naive splenic mature B cells, T2 B cells are capable of inducing anti-apoptotic genes and their protein products.
Given the close relationship between the long-term induction of c-Rel and anti-apoptotic genes (Fig. 1), we analyzed the effects of long-term BCR stimulation on c-Rel expression in transitional B cell subsets. For these studies, we FACS-sorted transitional B cell subsets from apoptosis resistant Bcl-2 transgenic mice (as in Fig. 2A) to avoid potential effects of differential sensitivity of T1 and T2 cells to BCR-induced cell death (Fig. 3A) (31). c-Rel transcripts were not induced in T1 cells even after 16 h of BCR activation, whereas this response was evident in T2 B cells (Fig. 3B). Accumulation of c-Rel protein also significantly increased in T2 and mature B cells following BCR stimulation when compared with T1 B cells (Fig. 3, C–E). These results suggested that one key feature that distinguishes T2 from T1 cells was their ability to up-regulate c-Rel expression following BCR stimulation. Taken together, these results demonstrate that T2 cells induce the c-Rel gene transcription and protein production that is required for sustained nuclear c-Rel activity and anti-apoptotic gene expression, whereas T1 cells fail to do so rendering them sensitive to apoptosis.
BAFF-R plays a critical role in B cell survival beyond the T1 stage (17). The findings that B cells become further resistant to apoptosis during the T1 to T2 transition by gaining responsiveness to BAFF further support a unique role for BAFF beginning at the T2 stage (4). This is, in part, due to increased expression of BAFF-R, perhaps as a consequence of BCR signaling (15). Consistent with an enhanced responsiveness of T2 cells to BAFF (4), we found that BCR-induced BAFF-R expression increased in T2 and mature but not in T1 B cells (Fig. 4A). Because BAFF-dependent survival is mediated by both canonical and the noncanonical NF-κB pathways (20, 21, 35), effective anti-apoptotic signaling by BAFF requires not only increased BAFF-R expression, but also sustained availability of the noncanonical pathway substrate p100. We therefore examined whether substrate availability was also differentially regulated in transitional B cell subsets.
BCR stimulation increased p100 mRNA in T2, but not in T1 cells after 4 h of BCR stimulation (Fig. 4B). To establish a link between long-term c-Rel induction, we activated FACS-sorted T1, T2, and mature B cells from Bcl-2 Tg mice and analyzed p100 mRNA after 16 h of culture with or without anti-IgM. We found elevated levels of p100 in T2 and further enhanced in mature B cells, however, this increase was minor in T1 B cells (Fig. 4C), which reflects severely reduced levels of BCR-induced p100 protein in T1 cells in even after long-term stimulation via the BCR (Fig. 4D). The levels of p100 protein in T2 + M B cells were increased significantly, conceivably by p100 production by both T2 and mature B cells (Fig. 4D). These results extend previous findings with BAFF-R to the regulation of its downstream substrates p100 (15). Together, these data suggest that increased expression of BAFF-R and noncanonical NF-κB pathway mediator p100 contribute to the higher survival response of T2 and mature Fo B cells compared with T1 cells in response to BAFF. A recent report showing a critical role for the noncanonical NF-κB pathway in B cell survival further supports this hypothesis (20).
NF-κB2 has been shown to be an NF-κB target gene (35, 40). The coordinate gain of BAFF-R, p100, and c-Rel inducibility in T2 cells prompted us to examine the c-Rel dependence of genes involved in BAFF-mediated signaling. WT and c-Rel-deficient B cells were activated with anti-IgM for varying times followed by RNA and protein analyses. We observed a time-dependent increase of BAFF-R, and p100 mRNA in WT but not c-Rel-deficient B cells (Fig. 5, A and B). In concordance with the RNA data, p100 protein was robustly up-regulated in normal B cells activated via the BCR, but not in c-Rel-deficient B cells (Fig. 5C). Further, the increase in BAFF-R and p100 mRNA and p100 protein occurred more robustly at later time points, which is consistent with their induction by long-term c-Rel induction by BCR signaling. These results indicate that c-Rel plays a key role in the inducible activation of BAFF-R and p100 genes. This role is in addition to the well-established function of c-Rel in the regulation of anti-apoptotic genes under canonical NF-κB control, including Bcl-xL and A1 (Fig. 2). We note, however, that basal expression of all genes analyzed was comparable in WT and c-Rel-deficient B cells. This may be due to compensatory NF-κB activity in c-Rel-deficient B cells or it may be mediated by other constitutively active transcription factors. Together, these data demonstrate that c-Rel pathway is particularly important not only for expression of anti-apoptotic genes of the Bcl-2 family but also for long-term expression of BAFF-R, and p100 in response to BCR stimulation.
We and others have previously shown that like T1 cells, Btk-deficient B cells fail to induce anti-apoptotic genes and undergo apoptosis in response to BCR engagement (6, 7). If the failure to induce c-Rel contributed causally to the enhanced death of T1 cells, we expect impaired c-Rel expression in Btk-deficient B cells after BCR stimulation. We tested this prediction by activating btk−/− B cells with anti-IgM and assaying for c-Rel mRNA and protein. BCR stimulation resulted in c-Rel transcriptional induction and protein accumulation in WT, but not btk−/−, B cells (Fig. 6, A and B). CD40 signaling, which is known to promote survival of WT and btk−/− B cells resulted in comparable c-Rel induction (Fig. 6B). Lack of BCR-induced c-Rel expression also affected its nuclear DNA binding in btk−/− B cells (Fig. 6C). Consistent with a role for Btk in the regulation of c-Rel, Btk-deficient B cells also failed to increase p100 protein in response to BCR signals (Fig. 6D). Taken together, results in Figs. 5 and and66 show that btk−/− and c-rel−/− B cells are defective in mechanisms that promote survival in a manner similar to T1 B cells and suggest that Btk/c-Rel signaling axis is critical in the regulation of B cell survival.
Establishing tolerance is an essential aspect of B cell development, whereby cells that recognize self-Ags are functionally discarded. Positive and negative selection of clonotypes in the periphery is modulated by the differential responses of transitional B cells to BCR engagement. In this study, we provide evidence that reduced c-Rel response in T1 cells controls their hypersensitivity to BCR cross-linking. As a consequence, anti-apoptotic genes targeted by c-Rel, such as Bcl-xL and A1, are poorly induced in T1 B cells. In contrast, T2 B cells resist BCR-induced apoptosis due their ability to activate and sustain signal-induced expression of c-Rel, which leads to increased anti-apoptotic gene expression relative to T1 cells. Our observations also explain the enhanced sensitivity of Btk-deficient B cells to BCR-induced apoptosis by placing c-Rel transcriptional induction downstream of Btk as an essential mediator of survival signals.
We previously showed that T1 cells fail to produce diacylglycerol and activate PKC in response to BCR engagement (3). This suggested that T1 cells would not activate NF-κB leading to reduced expression of NF-κB-dependent anti-apoptotic genes. We and others tested this hypothesis and were surprised to find that IKK was phosphorylated, IκBα was degraded, NF-κB was translocated to the nucleus and bound DNA comparably in T1, T2, and mature B cell populations (Ref. 41 and data not shown). Yet, NF-κB-dependent anti-apoptotic and c-Rel genes are differentially regulated in T1, T2, and mature B cells (Figs. 1, D and E and 2, B and C). One potential explanation for this unexpected finding may be that unlike T2 and mature, T1 B cells are in a state of nuclear nonresponsiveness as shown by BCR-specific defect in the assembly of active transcriptional machinery in T1 cells (41). This view is supported by our data showing that the first phase of A1 and Bcl-xL transcription is also defective in T1 B cells, whereas these genes are clearly up-regulated within the first 1– 4 h following BCR stimulation in T2 and mature B cells (Figs. 1, D and E and and2C2C).
Our data show that in addition to the early phase, T2 and mature B cells produce a second and long-term phase of c-Rel induction. Importantly, T1 B cells are defective in this response as well. The second and late phase of c-Rel induction and consequent A1 and Bcl-xL gene transcription in T2 and mature Fo B cells may be critical for an effective resistance to apoptosis in response to BCR engagement because the later phase coincides with a substantial increase in the corresponding anti-apoptotic proteins. It is unclear at present whether a reduced long-term c-Rel induction in T1 B cells in response to BCR stimulation is also associated with their nuclear nonresponsiveness. However, two distinct mechanisms operate in T2 and mature Fo B cells that allow early as well as sustained expression of c-Rel and its downstream target genes. In contrast, neither of the two phases of c-Rel induction operates in T1 B cells, resulting in an inability to produce anti-apoptotic factors and rendering them susceptible to BCR-induced cell death. Although the mechanism of long-term c-Rel activation remains unknown, it may be largely independent of the regulation by classical NF-κB pathway as newly synthesized c-Rel may not be sequestered in the cytoplasm by IκB proteins, as previously suggested (42). Regardless of the mechanism, our data suggest de novo synthesis of c-Rel in T2 cells as a mechanism to limit the expression of the anti-apoptotic genes to this apoptosis-resistant transitional B cell population. Together, our results suggest that an ability to produce a c-Rel activation response following BCR engagement is associated with resistance to apoptosis in T2 and mature Fo B cells.
In addition to BCR-initiated signals, B lymphocyte survival and selection requires activation of the BAFF/BAFF-R pathway (17, 20). We show in this study that the Btk/c-Rel pathway also up-regulates expression of BAFF-R and its substrates p100 in T2 cells. Thus, proapoptotic BCR-mediated signaling in T1 cells is converted to an anti-apoptotic response during the T1 to T2 transition by two c-Rel-regulated mechanisms: First, c-Rel directly activates anti-apoptotic gene expression and second, c-Rel enhances the levels of both BAFF-R and its substrate, p100, which likely increases BAFF responsiveness of T2 relative to T1 cells. This hypothesis is consistent with the recent findings showing that T2 cells express higher levels of cell surface BAFF-R and respond more robustly to BAFF than T1 B cells (4). In addition, prior studies are consistent with a critical role for c-Rel, anti-apoptotic genes, and BAFF-R in B cell survival. Although mature B cells develop in c-Rel-deficient mice, these B cells respond similarly to normal immature B cells in their sensitivity to BCR-induced apoptosis, whereas induced expression of Bcl-xL and cyclin E can rescue immature B cells from this form of cell death (25, 28, 29, 43, 44). Likewise, B cells deficient for Btk show similar but more pronounced defects in the survival pathways discussed above in the context of c-Rel deficiency. A vital function for BAFF-R in B cell survival beyond the T2 stage has also been well documented. Thus, loss of either BAFF or BAFF-R results in a severe deficiency of T2 and mature B cells (17, 18, 45, 46). In addition, reduced sensitivity of T2 B cells to BCR-induced apoptosis is further accentuated by developmental alterations in the BAFF/BAFF-R pathway of survival signaling (4, 13, 15, 18, 47, 48).
BAFF/BAFF-R interaction activates the alternative NF-κB pathway by IKKα-mediated phosphorylation of p100 (NF-κB2) that results in its proteolytic processing to p52 (19, 40, 49). The active transcription factor consists of the p52/RelB heterodimer, which activates anti-apoptotic genes such as Pim-2 (20). These and earlier studies show that constitutive, as well as BCR-inducible, expression of BAFF-R is higher in T2 B cells compared with T1 B cells (4, 15). For increased surface expression of BAFF-R to be effective, however, components of the alternative NF-κB pathway must be readily available. We find that p100 is induced preferentially in T2 B cells after BCR cross-linking; like c-Rel, this induction requires the presence of Btk. Re-enforcement of the BAFF-R signaling pathway by the BCR thus confines survival advantage among transitional B cells to the T2 population potentially facilitating their positive selection into mature B cells.
Taken together with earlier studies of signaling differences between T1 and T2 B cells, our observations lead us to the following model for the gain of apoptosis resistance to BCR signaling during the T1 to T2 transition; in contrast to T2, T1 B cells do not induce sustained c-Rel function that is mediated by Btk. This long-term c-Rel activation in T2 B cells results in stable induction of anti-apoptotic genes, which in turn rescues T2 B cells from BCR-induced cell death. Our observations account for the enhanced sensitivity of Btk-deficient B cells to apoptosis in response to the BCR by placing c-Rel induction downstream of Btk (6, 7, 22). Moreover, we demonstrate a unique role for Btk and c-Rel in the production of p100 in response to BCR, thus providing additional support for the re-enforcement of BAFF-R induced activation of the alternate NF-κB pathway by the BCR. The requirement for Btk in the up-regulation of p100 strongly implicates the long-term c-Rel activation in cross-talk between the canonical and alternate NF-κB pathways in productive B cell activation. Thus, the ability of transitional B cells to maintain anti-apoptotic c-Rel and BAFF-R activity following BCR engagement will determine whether these cells will survive and undergo positive selection or be negatively selected by apoptosis. Thus, our findings suggest that the ability to produce sustained c-Rel responses contributes to the gain of survival potential in T2 B cells, thus facilitating their survival and differentiation in vivo.
We thank M. Pia Arrate for expert technical assistance. We thank Dr. Emily Clark for assistance in establishing the intracellular staining protocol for c-Rel in primary splenic B cells. We acknowledge the skilled assistance of Kevin P. Weller, David K. Flaherty, and Jim Higginbotham of Vanderbilt University Medical Center Flow Cytometry and Dr. Oliver Umland of University of Miami Diabetes Research Institute Flow Cytometry Core. We thank Dr. Hsiou-Chi Liou for permission to use and Dr. Mark Boothby for providing c-Rel-deficient mice.
6Abbreviations used in this paper: Fo B, follicular B cell; MZ, marginal zone; BAFF, B cell activating factor; Btk, Bruton’s tyrosine kinase; WT, wild type; FCM, flow cytometric; Tg, transgenic; qRT-PCR, quantitative RT-PCR.
The authors have no financial conflict of interest.
1This study is supported in part by National Institutes of Health AI060729 (to W.N.K.), AI041054 and AI057463 (to R.T.W), AI043534 (to R.M.G), P30DK32520 (to University of Massachusetts Medical Flow Cytometry Core) and DK32520 (to the University of Miami Miller School of Medicine Diabetes and Endocrinology Center). K.L.H and N.P.S. are supported by NIH T32 HL69715–0, I.C. by NIH T32 CA09385–20 and J.A.W. by NIH 5 T32 HL069765. B.D. and R.S. are supported by intramural research program of the National Institute of Health on Aging, Baltimore, MD.