Deletion of immunoglobulin β in developing B cells leads to cell death

Abstract

Inducible gene-targeting experiments have shown that Igμ expression is essential for maintaining survival of mature B cells, but the role of Igμ expression in immature B cell survival has not been determined. To assess whether continued B cell receptor (BCR) expression is required for bone marrow B cell precursor development and survival, we developed a method for conditional gene deletion in these cells. Recombination-activating gene regulatory elements were used to express Igβ cDNA as a transgene to complement Igβ^−/− mice. Transgenic Igβ expression was found in proB and small preB cells and was extinguished in large preB and immature B cells. Igβ deletion from large preB cells and immature B cells resulted in cell death that could be rescued by transgenic bcl-2 expression. However, transgenic bcl-2 expression was unable to restore B cell development in the absence of Igβ. We conclude that Igβ expression is essential to maintain preB cell and immature B cell survival and to mediate B cell differentiation. In addition, complementation of null mutations with cDNAs under the control of heterologous bacterial artificial chromosomes is a useful method for cell-type-specific and developmentally regulated gene ablation in vivo.

B cell development is regulated by membrane IgM (mIgμ) through Igα and Igβ, two membrane-anchored Ig superfamily proteins that are noncovalently associated with mIgμ in preB or B cell receptors (BCR) (1–3). A requirement for BCR expression in B cell survival has been demonstrated by conditional gene-targeting experiments. Loss of Igμ expression in peripheral B cells induces programmed cell death (4). It was proposed that cell death in response to Igμ ablation is an important mechanism for B lymphocyte quality control in the periphery leading, for example, to deletion of B cells that accumulate deleterious mutations in their Ig genes during the germinal center reaction (4). In contrast, T cells lacking T cell antigen receptor (TCR) do not die precipitously suggesting that requirement of BCR and TCR expression was different in B and T lymphocyte survival, respectively (5, 6).

Immature B cells differ from mature B cells in their responses to BCR crosslinking. Whereas BCR stimulation in mature B cells leads to clonal expansion, the same stimulus leads to tolerance in immature B cells. In the bone marrow, a large fraction of developing B cells undergo receptor editing, and an unknown number of these cells lose BCR expression by out-of-frame or RS V(D)J recombination (7–9). Therefore, programmed cell death in response to receptor deletion would serve the same quality-control purpose in developing B cells. However, the role of Igμ and BCR expression in immature B cell survival could not be examined by conventional conditional gene targeting because Igμ deletion by inducible Cre ablates normal B cell development (4).

To determine whether BCR expression in preB cells and immature B cells is required to induce their further differentiation and to maintain their survival, we developed a method for terminating Igβ expression in developing B cells in the bone marrow.

Materials and Methods

Transgenic Mice.

A bacterial artificial chromosome (BAC) containing RAG1 and RAG2 genes and all of the regulatory elements required to direct RAG expression in vivo was used to generate an Igβ transgene [N-BAC (10)]. Mouse Igβ cDNA was inserted at the RAG2 start codon in N-BAC by homologous recombination (11) and used to produce transgenic mice (Igβtg). Igβtg-Igβ^−/−, αHEL-Igβtg-Igβ^−/−, bcl-2-αHEL-Igβtg-Igβ^−/−, and bcl-2-Igβtg-Igβ^−/− mice (12–14) were produced by breeding. All mice were maintained under specific pathogen-free conditions.

Flow Cytometry and Cell Sorting.

Bone marrow and spleen B cells from mutant or wild-type mice were enriched by positive selection by using MACS CD19 microbeads (Miltenyi Biotec, Auburn, CA) and stained with FITC, phycoerythrin, allophycocyanin, and biotin-conjugated monoclonal antibodies that were visualized with Strepavidin red 613 (GIBCO/BRL). Monoclonal antibodies were anti-CD43, anti-IgM, anti-B220, anti-CD25, anti-IgD, anti-HSA, anti-CD19, and anti-CD22 (PharMingen). Data were collected on a FACScalibur and analyzed with CELLQUEST software (Becton Dickinson).

RNA Preparation and Reverse Transcription (RT)-PCR.

Total RNA was extracted from 10⁵ purified B cells by using TRIzol Reagent (GIBCO/BRL) and reverse transcribed in 20 μl with Superscript II (GIBCO/BRL). For RT-PCR reactions, 1 μl of cDNA was amplified for either16 cycles (actin) or 30 cycles of 30 s at 94°C, 30 s at 60°C, and 30 s at 72°C with a final 10-min extension at 72°C with Hotstar TaqDNA polymerase (Qiagen, Chatsworth, CA) and the following primers: Igα sense 5′-CGATGCCAGGGGGTCTAGAAG-3′, antisense 5′-GAAGAACAGCTGGCCTGTGGTA-3′; Igβ sense and antisense as Igβtg primers above (see Transgenic Mice). RAG1-, RAG2-, and β-actin-specific primers have been described (15). Transgenic Igβ expression in wild-type background was detected by RT-PCR by using 5′ sense RAG2 and Igβ antisense primers. RT-PCR products were analyzed on 2% agarose gels and Southern blots were hybridized with cognate cDNAs. Phosphorimager analysis was performed on PCR samples amplified in the linear range.

B Cell Cultures.

B cells from bone marrow from mutant or wild-type mice were enriched by positive selection by using MACS CD19 microbeads (Miltenyi Biotec); stained with Fab′ anti-IgM, and monoclonal anti-B220, -CD25, -CD43, or anti-B220, -CD25, -IgD; and sorted into B220⁺CD43⁺CD25⁻IgM⁻ proB, B220⁺CD43^-CD25⁺IgM⁻ preB, and B220⁺CD43^-IgM^lowIgD⁻ immature B cells. Purified B cells were cocultured at 10⁶/ml with irradiated S17 stromal cells in DMEM supplemented with 10% FCS and 10 ng/ml IL-7 (PharMingen). B cell viability was assessed by flow cytometry after phycoerythrin-Annexin V (PharMingen) and propidium iodide staining on day 0 and after a 36-h culture period.

Results

Igβ Expression Under RAG2 Regulation.

To direct Igβ expression in early developing B cells but not in mature B cells, we inserted an Igβ cDNA into RAG2 in a RAG locus containing bacterial artificial chromosome (11). Two Igβ transgenic lines were obtained and showed identical patterns of transgenic Igβ expression (Igβtg mice, Fig. 1A).

Fig 1.

Igβ expression in Igβtg mice. (A) Diagrammatic representation of RAG locus and the Igβ containing RAG BAC (10). (B) RT-PCR of endogenous Igβ and Igβtg mRNA in B cell subpopulations purified from bone marrow and spleen. Bone marrow B cells were sorted into: (i) proB cells, (ii) preB cells that were further separated on the basis of size by forward and side scatter parameters, (iii) immature B cells, (iv) mature recirculating B cells. Spleen B cells were sorted into immature B cells and mature B cells. Serial 5-fold dilutions of cDNA are shown. PCR-negative control without cDNA (−) is included.

To determine whether Igβtg expression resembles the documented pattern of endogenous RAG2 expression, we measured transgenic Igβ mRNA by RT-PCR (Fig. 1B). Like endogenous RAG2, high levels of the transgenic mRNA were detected in proB cells, expression was down-regulated in proliferating (large) preB cells, reinduced in resting (small) preB cells (Fig. 1B) (16), and gradually extinguished in the immature B cell compartment. Transgenic Igβ mRNA was not detectable in mature B cells (Fig. 1B) (10, 17). By comparison, endogenous Igβ mRNA was expressed at similar levels in all B cell fractions (Fig. 1B). We conclude that the RAG regulatory elements in the Igβtg direct Igβ cDNA expression in a pattern similar to that of endogenous RAG2.

B Cell Development Depends on Igβ Expression.

To determine whether Igβtg can reconstitute B cell development in Igβ^−/− mice, we crossed these two strains of mice to produce Igβtg-Igβ^−/− mice. In contrast to arrested proB cell development in Igβ^−/− mice, Igβtg-Igβ^−/− B cell development proceeded to the preB stage (Fig. 2). However, preB cells did not accumulate in Igβtg-Igβ^−/− mice, and these cells did not differentiate into IgM⁺ immature B cells (Fig. 2). Failure to proceed beyond the preB cell stage in Igβtg-Igβ^−/− mice coincided with decreased transgenic Igβ expression at this stage in B cell development (Fig. 1).

Fig 2.

Bone marrow B cell development in Igβtg-Igβ^−/− mice. Bone marrow B cells from wild-type (WT), Igβ^−/−, Igβtg-Igβ^−/−, αHEL-Igβ^−/−, and αHEL-Igβtg-Igβ^−/− mice were preenriched by using CD19 MACS beads. Dot plots show staining with anti-CD43 or anti-IgM, and anti-CD25. Numbers indicate percentages of B220⁺ cells in each quadrant.

To bypass the preB cell stage block in Igβtg-Igβ^−/− mice, we introduced the αHel Ig heavy- and light-chain transgene (αHEL) into the Igβtg-Igβ^−/− background [αHEL-Igβtg-Igβ^−/− mice (13)]. In contrast to the arrested B cell development in control Igβtg-Igβ^−/− mice and αHEL-Igβ^−/− mice, B cells in αHEL-Igβtg-Igβ^−/− mice develop to the CD25^lowIgM^lowIgD⁻CD21⁻CD23⁻immature B cell stage (Fig. 2 and not shown). However, αHEL-Igβtg-Igβ^−/− B cells fail to develop beyond the IgM^low immature B cell stage, and no mature recirculating CD25^-IgM^high B cells are in these mice (Fig. 2). The few B cells that were found in the spleen resembled their immature bone marrow counterparts in that they expressed low levels of surface IgM, IgD, and CD25 (Fig. 3 A and B) and low levels of endogenous RAG1 and RAG2 genes (Fig. 3C). As expected the level of transgenic Igβ mRNA in these cells was low and similar to the level of endogenous RAG2, whereas Igα and actin expression was similar to control (Fig. 3C; refs. 10 and 17). Failure to progress beyond the immature B cell stage was associated with decreased Igβ expression in immature B cells in Igβtg mice (Figs. 1B, 2, and 3).

Fig 3.

Splenic B cells analysis. (A) Relative absence of spleen B cells in Igβtg transgenic mice. CD19 vs. CD22 staining. Numbers indicate percentages of B lymphocytes. (B) Dot plots show staining with anti-IgD vs. anti-IgM, and anti-CD25 vs. anti-CD43 on CD19 MACS-enriched B cells from wild-type (WT) and αHEL-Igβtg-Igβ^−/− mice. Numbers indicate percentages of CD19⁺ cells in each quadrant. (C) RAG1, RAG2, Igα, and Igβ mRNA expression in peripheral B cells. RNA from CD19-enriched peripheral B cells from wild-type and αHEL-Igβtg-Igβ^−/− mice was analyzed by RT-PCR and visualized with ³²P-labeled probes. Actin RT-PCR was used as mRNA loading control.

PreB and Immature B Cell Survival Requires Igβ Expression.

BCR ablation in mature B cells results in cell death by apoptosis (4). To determine whether absence of precursor B cell accumulation in Igβtg-Igβ^−/− and αHEL-Igβtg-Igβ^−/− mice is associated with cell death, we measured precursor B cell survival in in vitro bone marrow cultures (18). ProB cells from Igβ^−/−, preB cells from Igβtg-Igβ^−/−, and immature B cells from αHEL-Igβtg-Igβ^−/− mice and controls were purified and cell death measured by staining with annexin V and propidium iodide at the initiation of culture and after 36 h (19). Annexin V staining varies during B cell development and is therefore unreliable when comparing B cells in different stages (19). However, annexin is a reliable marker for apoptosis when comparing cells at similar stages in development.

Freshly isolated Igβ^−/− proB cells showed a 2-fold increase in annexin V and propidium iodide staining compared with wild-type controls (Fig. 4 A and B). During a 36-h culture period, Igβ^−/− proB cells failed to develop, whereas control proB cells progressed to the immature B cell stage (Fig. 4 A and B). Freshly isolated Igβtg-Igβ^−/− preB cells showed a 5-fold increase in annexin V staining compared with controls (Fig. 4 C and D). During a 36-h culture period, Igβtg-Igβ^−/− preB cells remained CD25⁺CD43^lowIgM⁻ and a 3- to 4-fold increase occurred in propidium iodide staining. In contrast, about 50% of wild-type preB cells developed into IgM⁺ immature B cells and of those that remained as preB cells, only a small fraction became propidium iodide positive (Fig. 4 C and D). Freshly isolated immature B cells differed from pro- and preB cells in that both mutants and controls showed high levels of annexin staining (Fig. 4 E and F). After a 36-h culture period immature B cells from wild-type mice developed into mature IgM^highIgD⁺ B cells, whereas αHEL-Igβtg-Igβ^−/− immature B cells did not develop beyond the B220⁺CD43⁻IgM^lowIgD⁻ immature B cell stage (Fig. 4E). Annexin V and propidium iodide staining after 36 h revealed an increase in apoptotic and dead cells in αHEL-Igβtg-Igβ^−/− immature B cell cultures compared with wild-type controls (Fig. 4F). We conclude that Igβ deletion in preB or immature B cells results in both arrested B cell development and cell death.

Fig 4.

B cell development and death in bone marrow cell cultures. Bone marrow proB, preB, immature B cells from wild-type (WT) and Igβ^−/−, Igβtg-Igβ^−/−, αHEL-Igβ^−/−, αHEL-Igβtg-Igβ^−/− mice were cocultured with S17 stromal cells. (A) Phenotype of proB cells from wild-type and Igβ^−/− mice. Staining with anti-CD25 and anti-IgM at the initiation of culture and after 36 h. (B) Survival of proB cells from wild-type and Igβ^−/− mice. Bar graphs show the percentage of proB cells that were Annexin V- or propidium iodide-positive at the initiation of culture and after 36 h. (C) Phenotype of preB cells from Igβtg-Igβ^−/− and wild-type mice. PreB cells from wild-type or Igβtg-Igβ^−/− mice were stained with anti-CD25 and anti-IgM at the initiation of culture and after 36 h. (D) PreB cells from Igβtg-Igβ^−/− mice show impaired survival. Bar graphs show the percentage of preB cells that were Annexin V- or propidium iodide-positive at the initiation of culture and after 36 h. (E) Phenotype of immature B cells from αHEL-Igβtg-Igβ^−/− and wild-type mice. Immature B cells from wild-type or αHEL-Igβtg-Igβ^−/− mice were stained with anti-IgD and anti-IgM at the initiation of culture and after 36 h. (F) Immature B cells from αHEL-Igβtg-Igβ^−/−mice show impaired survival. Bar graphs show the percentage of immature B cells that were Annexin V- or propidium iodide-positive at the initiation of culture and after 36 h. Results are the average of duplicate cultures performed on two independent mice. The variation between samples and mice was less than 5%. * indicates statistically significant difference (t test). For all fluorescence-activated cell sorter plots; numbers indicate percentages of B220⁺ cells in each quadrant.

Bcl-2-Dependent Rescue of Receptorless B Cell Precursors.

Transgenic bcl-2 expression delays apoptotic cell death in many cell types including B cells that are autoreactive or mature B cells that have lost their receptors by Igμ deletion (4, 14, 20). To determine whether bcl-2 expression can rescue B cell development despite Igβ deletion we bred bcl-2 transgenic mice with Igβtg-Igβ^−/− or αHEL-Igβtg-Igβ^−/− mice. We found that transgenic bcl-2 did not alter B cell development, but increased the number of CD19⁺ cells in either Igβtg-Igβ^−/− (2.5 ± 0.7 × 10⁶ and 7.3 ± 1.3 × 10⁶ in Igβtg-Igβ^−/− and Igβtg-Igβ^−/−bcl-2⁺mice, respectively) or αHEL-Igβtg-Igβ^−/− mice (6.7 ± 1.9 × 10⁶ and 12.5 ± 2.4 × 10⁶ in αHEL-Igβtg-Igβ^−/− and αHEL-Igβtg-Igβ^−/−bcl-2⁺ mice, respectively). Deletion of Igβ in bcl-2 transgenic Igβ^−/− proB, Igβtg-Igβ^−/− preB, and αHEL-Igβtg-Igβ^−/− immature B cells produced a CD19⁺CD43⁻CD25⁻ B cell population that did not express detectable surface IgM (Fig. 5A). These B cells survived long enough to populate the spleen in bcl-2 transgenic αHEL-Igβtg-Igβ^−/− mice and to a lesser extent in bcl-2 transgenic Igβtg-Igβ^−/− mice but not in bcl-2 transgenic Igβ^−/− mice (Fig. 5 B and C). B cells represented 8% of splenocytes (4.1 ± 1.1 × 10⁶ CD19⁺ cells) in bcl-2 transgenic Igβtg-Igβ^−/− mice and reached 24% (10.5 ± 2.2 × 10⁶ CD19⁺ cells) in bcl-2 transgenic αHEL-Igβtg-Igβ^−/− mice (Fig. 5B). Thus, bcl-2 transgene expression rescues preB and immature B cells from cell death in response to Igβ deletion but fails to induce further B cell differentiation.

Fig 5.

Bcl-2 transgene rescues Igβtg B cell survival. (A) Bone marrow B cells from wild-type (WT), αHEL-Igβtg-Igβ^−/−, bcl-2-Igβ^−/−, bcl-2-Igβtg-Igβ^−/−, and bcl-2-αHEL-Igβtg-Igβ^−/− mice were enriched by using CD19 MACS beads. Dot plots show staining with anti-IgD vs. anti-IgM (Top) and anti-CD25 vs. anti-CD43 (Bottom). (B) CD19 vs. CD22 staining of splenic B cells from bcl-2-Igβtg mice. (C) anti-IgD vs. anti-IgM staining of CD19 MACS-enriched splenic B cell from wild-type, αHEL-Igβtg-Igβ^−/−, bcl-2-Igβtg-Igβ^−/−, and bcl-2-αHEL-Igβtg-Igβ^−/− mice.

Discussion

Gene Ablation by BAC Complementation.

Inducible gene targeting is an established method for altering specific gene expression in fully developed tissues in mice (21). The advantage of this method compared with conventional gene targeting is that it allows the target gene to contribute to normal cellular maturation before deletion. Disadvantages of conditional gene targeting are that the deletion is frequently less than 100% efficient, it requires inducing agents such as IFN that may alter normal physiology, and high levels of cre recombinase expression can damage normal cellular DNA (22).

We have developed a method for cell type- and developmental stage-specific gene ablation by using transgenic BACs. We would like to propose that this method could be used in a variety of different settings to achieve developmentally restricted gene expression. BAC transgene complementation for programmed gene extinction has many of the same advantages as inducible gene targeting. Potential disadvantages of the BAC approach include: (i) loss of expression is not inducible; (ii) the requirement of a null mutant for complementation; (iii) the potential for variegated or abnormal BAC transgene expression if the BAC does not contain all of the necessary cis elements. Advantages of the method include rapid BAC modification by homologous recombination in Escherichia coli and homogenous gene extinction or activation in a cell type-specific fashion without exogenous agents for gene deletion.

Igβ Expression Is Essential for Precursor B Cell Development.

Failure to assemble BCRs in μMT^−/−, J_H^−/−, Igβ^−/−, or RAG-deficient mice results in arrested development at the proB cell stage (reviewed in refs. 23 and 24). These traditional gene-targeting experiments differ from our experiments in that B cells never developed to the preB cell or immature B cell stage; therefore, the role of preBCR or BCR expression in promoting further precursor B cell differentiation and survival could not be assessed. Our experiments show that Igβ deletion in preB or immature B cells results in arrested B cell development and cell death by apoptosis. Bcl-2 transgene expression rescues preB and immature B cells from cell death in response to Igβ deletion but does not promote the differentiation of these precursors. Therefore, cell death is not responsible for the inability of preB and immature B cells to continue to differentiate in the absence of Igb.

Why do preB cells fail to differentiate into immature B cells in the absence of Igβ expression? In Igβtg-Igβ^−/− mice, transgenic Igβ expression allows the assembly and selection of Igμ-containing preBCRs that induce the differentiation of proB cells into large preB cells. In the absence of Igβ, normal levels of preBCR and the BCR are not expressed on the cell surface. PreBCR expression in these cells results in RAG and transgenic Igβ down-regulation and therefore preBCR expression induces its own deletion (16). In the absence of the preBCR large preB cells fail to progress to the small preB cell stage where RAG and light-chain (IgL) gene recombination are induced (reviewed in ref. 25). The observation that bcl-2 expression fails to promote immature B cell development in bcl-2 transgenic Igβtg-Igβ^−/− mice suggests that in the absence of Igβ preB cell survival alone is not sufficient to induce further differentiation. Consistent with this interpretation, introduction of a pre-rearranged αHEL transgene rescues B cell development in Igβtg-Igβ^−/− mice up to the immature B cell stage.

Igβ Expression Is Essential for Precursor B Cell Survival.

When a lox-flanked Igμ gene was deleted from mature B cells by inducible Cre expression the cells underwent programmed cell death and it was proposed that apoptosis is a physiologic response to disruption of the BCR (4). Igμ deletion also resulted in loss of preB and immature B cells (4). However, the possible role of BCR in preB and immature B cell survival could not be assessed because Igμ deletion by Cre induction abrogated B cell development after the proB cell stage and no further preB or immature B cells could be produced (4). Our method of BAC complementation allowed us to determine that preBCR and BCR expression was essential for preB and immature B cell survival, respectively, and that these B cell precursors fail to be maintained in the absence of Igβ.

Our in vitro experiments suggested that B cell precursors that lose BCR expression die by apoptosis. The finding that transgenic bcl-2 expression rescued receptorless B cells from apoptosis in Igβtg-Igβ^−/− and αHEL-Igβtg-Igβ^−/− mice supports this idea (26, 27). In contrast, constitutive bcl-2 expression does not lead to accumulation of receptorless B cells in Igβ^−/−and Rag2^−/− mice in which B cell development does not proceed beyond the proB cell stage (27). PreBCR expression is therefore required to produce cells that can be maintained by constitutive bcl-2 expression.

How does the BCR produce a survival signal? Three mechanisms have been considered to explain how BCR expression might protect B cells from programmed cell death. First, receptor assembly on the cell surface might be sufficient to produce a constitutive survival signal (28). Alternatively the tonic signal might be produced by low-affinity interaction between mIgμ and self-antigens (29). A third possibility is that loss of BCR expression results in the up-regulation of Fas and in the down-regulation of MHC class I molecules, and these receptorless B cells are therefore recognized and killed by cytotoxic T or NK cells (4). Our experiments do not distinguish between receptor aggregation and receptor engagement by self-antigen as the mechanisms for producing the survival signal. However, we can rule out a requirement for NK or cytotoxic T cells in receptorless B cell deletion, because purified B cell precursors died upon receptor deletion in vitro and therefore this response is cell autonomous.

Whereas BCR crosslinking by antigen in mature B cells leads to proliferative responses and clonal expansion, BCR triggering in immature B cells induces tolerance responses including receptor editing, anergy and deletion by apoptosis (reviewed in ref. 25). It has been proposed that these distinct biological responses are caused by differences in the signaling cascades activated by the Igα and Igβ components of the BCR at different stages in development (30). Our results show that Igβ and BCR expression are essential for preB and immature B cell survival and therefore a BCR signal that maintains B cell survival is a universal B cell function common to all stages of B cell development.

Abbreviations

• BCR, B cell receptor

• RAG, recombination activating gene

• BAC, bacterial artificial chromosome

• RT, reverse transcription