Continued differentiation during B lymphopoiesis ... - Semantic Scholar

8 downloads 0 Views 1MB Size Report
Jun 18, 1997 - 14 Schuler, W., Weiler, I. J., Schuler, A., Phillips, R. A., Rosenberg, N., ... 15 Pennycook, J. L., Chang, Y., Celler, J., Phillips, R. A. and Wu,.
International Immunology, Vol. 9, No. 10, pp. 1481–1494

© 1997 Oxford University Press

Continued differentiation during B lymphopoiesis requires signals in addition to cell survival David M. Tarlinton, Lynn M. Corcoran and Andreas Strasser The Walter and Eliza Hall Institute of Medical Research, Post Office Royal Melbourne Hospital, Victoria 3050, Australia Keywords: apoptosis, B cell differentiation, immunoglobulin, knockout mice

Abstract During B lymphopoiesis, cells undergo successive rounds of division and growth arrest coupled to intermittent selection on the basis of Ig expression. It is unresolved whether differentiation requires specific signaling or is merely the consequence of sustained cell survival. Transgenic expression of the cell death antagonist, Bcl-2, promoted accumulation of B lymphoid cells in mice deficient in antigen receptor rearrangement (scid or rag-1–/–) and in mice lacking the IgM transmembrane domain (µMT). Continued differentiation occurred, however, only in the bcl-2/scid and bcl-2/µMT mice. The appearance of B lineage cells expressing CD21, CD22 and CD23 was associated with DHJH rearrangements which encode a truncated Cµ-containing protein called Dµ in bcl-2/scid mice and with expression of Ig heavy chain classes other than IgM in the bcl-2/µMT mice. In neither case, however, were proliferating cells observed in the more mature B lineage compartments in the bone marrow. Thus, continued B cell development requires signaling via Ig heavy chain-containing receptors and is not simply a consequence of blocking apoptosis. Introduction It remains an unresolved question whether development of hematopoietic cells requires specific signaling pathways for differentiation or whether the differentiation program is predetermined and merely requires signals for cell survival for its expression (1). B lymphopoiesis with the wealth of mouse strains with gain or loss of function mutations of key regulators of development serves as an ideal model system to resolve this issue. B cell development occurs in the bone marrow by the orderly progression of precursor cells through a series of genetically and phenotypically defined stages of differentiation (Fig. 1; reviewed in 2). The earliest committed B cell precursors can be identified by surface expression of the pan-B lymphocyte marker CD45R (B220) together with CD43 (S7) and c-Kit. Cells within this pro-B/pre-BI compartment (for nomenclature see 2) either possess unrearranged Ig heavy chain gene loci or DH–JH joinings only (3,4). Upon completion of a productive VHDHJH rearrangement pro-B/pre-BI cells progress to the early pre-BII stage, lose expression of CD43 and c-Kit, and the so-called pre-B receptor appears on the surface. This receptor complex is composed of an Ig µ heavy chain with the λ5 and Vpre-B surrogate light chains in

association with mb-1 (Igα) and B29 (Igβ) co-receptors (5). Large, cycling early pre-BII cells then become small, quiescent late pre-BII cells that commence Ig light chain gene (Igκ, Igλ) rearrangement. Virgin B cells, defined by expression of surface IgM (sIgM), are generated by productive rearrangement at one of the light chain gene loci. Those bearing highavidity self-antigen-specific receptors are eliminated in the bone marrow by a process termed activation-induced cell death (6,7). The remainder enter the periphery where a small fraction are selected by an unknown mechanism to become long-lived recirculating mature B cells, characterized by surface expression of IgM, IgD, CD21, CD22 and CD23 (8–10). Mice bearing mutations that preclude generation of productive Ig gene rearrangements show stage-specific blockages in B cell development. For example, the bone marrow of animals lacking either of the two recombinase activating genes, rag-1 or rag-2, contain only pro-B/pre-BI cells that lack any Ig heavy or light chain gene rearrangement (11,12). Mice homozygous for the scid mutation also are blocked at the pro-B/pre-BI stage. They do undergo very limited Ig gene rearrangement, albeit abnormally (13–15). B lymphocytes lacking either the membrane spanning region

Correspondence to: D. Tarlinton Transmitting editor: A. Kelso

Received 17 April 1997, accepted 18 June 1997

1482 Cell survival and differentiation are distinct

Fig. 1. Schematic representation of B cell development in adult mouse bone marrow. The presently identified stages of B cell development are indicated along with the compartment size per whole mouse bone marrow and growth factor requirements which characterize each stage. The points at which blockages in B cell development are known to occur as a result of specific mutations are also indicated.

of the Ig µ heavy chain (the so-called µMT mutation) or the Igβ co-receptor (B29) also fail to develop beyond the pro-B/ pre-BI stage, although a proportion of the B2201 cells in µMT mice contains productive IgH gene rearrangements (16–18). Analysis of such mutant mice provided compelling evidence that the pro-B/pre-BI to pre-BII transition is regulated by the rearrangement status of the IgH locus and requires surface expression of the pre-B cell receptor. Additional evidence for this notion comes from sequence analysis of the IgH gene loci of single cells from the pro-B/pre-BI and pre-BII stages, and from analysis of their expression of Ig µ heavy chains. These studies have shown that all cells in the pre-BII compartment contain cytoplasmic Ig µ protein; it derives from a single productive VHDHJH rearrangement, while the other IgH allele bears either a non-productive VHDHJH or an incomplete DHJH rearrangement (4). Certain normal rearrangements of Ig gene segments may also block B cell development. During IgH gene rearrangement DH elements can be joined to JH elements in any of the three possible translational reading frames. Among mouse B lymphocytes from the early, cycling pre-BII stage onwards, one DH reading frame (arbitrarily called rf2), is significantly under-represented in both DHJH and VHDHJH rearrangements (19–21). This DH reading frame is the only one that has the potential to encode a truncated Ig µ-like protein, termed Dµ, comprising the DH and JH segments fused to Cµ (22). It is thought that Dµ surface expression interferes with B cell development at the pro-B/pre-BI to pre-BII transition, probably by inhibiting VH to DHJH gene rearrangement by a mechanism analogous to that used to impose allelic exclusion at the IgH locus (4,19,23). Clonal expansion of precursor B cells occurs at distinct stages of differentiation. Pro-B/pre-BI cell proliferation requires IL-7, stem cell factor and possibly other signals from stromal cells (24–26). Cell division at the early pre-BII stage appears to be mediated by pre-B receptors (5,27,28) and presumably serves to increase the number of cells with productive IgH gene rearrangements which after growth arrest will undergo Ig light chain gene rearrangement. There is little proliferation

among late pre-BII cells, virgin B cells and recirculating, mature B cells in the bone marrow (28). The lifespan of B cell progenitors is limited to ~3 days in the absence of productive Ig gene rearrangement or surface Ig expression (29). The size of the pro-B/pre-BI cell compartment of rag-1–/–, rag-2–/–, scid and µMT mice is similar to that of normal mice (11–13). When an IgH transgene is introduced into rearrangement-deficient mice (rag-1–/–, rag-2–/– and scid), B cell development is arrested at the late pre-BII cell stage with no enlargement of that compartment compared to controls (30–32). In all cases of developmental arrest described above, very few B cell precursors appear in peripheral lymphoid organs. Thus survival, differentiation and emigration of B cells from the bone marrow depend on surface Ig expression. In a previous study we reported that transgenic expression of the inhibitor of apoptosis Bcl-2 facilitated the appearance and accumulation of near normal numbers of sIg– B lineage cells in the peripheral blood and spleen of mice homozygous for the scid mutation (33). Surprisingly, a significant fraction (~20–50%) of these B cells was found to express markers normally restricted to recirculating, mature B cells, such as CD21, CD22, CD23 and high levels of Ia (33). One interpretation of these data is that the role of Ig molecules in B cell development is to provide a survival signal and that subsequent differentiation occurs spontaneously. By comparing maturation of B cells in bcl-2 transgenic mice homozygous for either the scid, rag-1–/– or µMT mutations we have explored further the role of Ig expression in B cell differentiation, and attempted to separate this role from effects on cell survival and proliferation. Our results demonstrate that while Bcl-2 promotes survival of B cells in rag-1–/–, µMT and scid mutant mice, their differentiation requires expression of an IgH chain or a Dµ protein. These results suggest a role for Dµ protein in this differentiation and, consequently, may explain the fate of Dµ-expressing cells in normal mice. Our data also demonstrate that IgH classes other than IgM can promote differentiation of B cells. These results are discussed in the context of the relationship between Ig gene rearrange-

Cell survival and differentiation are distinct 1483 ment, differentiation, proliferation and cell survival during B lymphopoiesis.

on the supernatants using purified IgM and IgG1 myeloma proteins (Sigma, St Louis, MO) as standards. Cell cycle measurements

Methods Mice All mice were bred and maintained in specific pathogen-free conditions within the Walter and Eliza Hall Institute. The generation and characterization of the Eµ-bcl-2-36 (34) and Eµ-bcl-2-36/scid mice (33) have been described, as has the generation of rag-1–/– mice (31) and µMT mice (16), which were kindly provided by K. Rajewsky. Eµ-bcl-2-36/rag-1–/– mice and Eµ-bcl-2-36/µMT mice were generated by serially crossing Eµ-bcl-2-36 (C57BL/6, N . 8) transgenic mice for two or more generations with rag-1–/– (C57BL/6, N . 4) mice or µMT (C57BL/6, N . 4) mice. Inheritance of the bcl-2 transgene was determined by PCR as previously described (35). Inheritance of the rag-1 knock-out allele was revealed by Southern blot analysis (31). As µMT mice lack sIgM1 B cells (16), homozygosity for this mutation was diagnosed by immunofluorescence staining (see below) of peripheral blood leukocytes with FITC-labeled anti-IgD antibody and phycoerythrin-conjugated anti-CD45R(B220) antibody (Caltag, South San Francisco, CA). Cell preparation and immunofluorescence staining Single-cell suspensions were prepared from bone marrow and spleen essentially as described (23). Red blood cells were lysed using Tris-buffered ammonium chloride. All cell suspensions were filtered through nylon mesh to remove debris and clumps. Cells were stained with fluorochrome- or biotin-conjugated antibodies as previously described (36). Antibodies used in this work were as follows: monoclonals 5.1 anti-IgM, RA3-6B2 anti-CD45R(B220), 7G6 anti-CD21, 2D6 anti-CD22, B3B4 anti-CD23, S7 anti-CD43, 187.1 antiIgκ light chain, JC5.1 anti-λ light chain, polyclonal sheep anti-mouse Ig (Caltag) and goat anti-mouse IgD (Nordic, Tilburg, The Netherlands). B cell subpopulations were isolated using a FACStar Plus cell sorter (Becton Dickinson, Mountain View, CA) after staining with fluorescence-labeled and biotinylated antibodies specific for the appropriate markers (as described in the text). In vitro assays of cell differentiation B lineage cells were purified by sorting from either the bone marrow or spleen of the indicated mouse strains. Ten thousand B cells of each phenotype were then cultured in the presence of IL-4, IL-5, FCS (10%), 2-mercaptoethanol and 5000 NIH 3T3 cells transfected with mouse CD40 ligand (37). As a negative control, B cells were co-cultured with control NIH 3T3 cells plus IL-4 and IL-5. When proliferation was to be measured, the fibroblasts were irradiated with 30 Gy prior to the addition of B cells. Proliferation was determined by [3H]thymidine incorporation during a 6 h pulse on day 3 of culture. The degree of proliferation was calculated as a stimulation index: SI 5 (c.p.m. cultured with CD40 ligand NIH 3T3/c.p.m. cultured with control NIH 3T3). Ig secretion was determined after 7 days culture by quantitative ELISA assays

The proportion of cells in the various phases of the cell cycle was determined by measuring the DNA content of nuclei isolated from sorted cells. Cells were stained for 30 min at 4°C with 50 µg/ml propidium iodide (Sigma) in 0.1% sodium acetate with 0.2% Triton X-100 (BDH Chemicals, Poole, UK) and then analyzed on a FACScan analyzer (Becton Dickinson). The CellFit program was used to calculate the percentages of cells in the various stages of the cycle. Analysis of expression of transgene encoded Bcl-2 protein and cell survival assays The intracellular levels of transgene encoded human Bcl-2 protein were determined as described (35). Bone marrow cells were fixed for 10 min with 1% paraformaldehyde, stained with the mouse anti-human Bcl-2 mAb Bcl-2-100 (38) in the presence of 0.3% saponin, which was also included in all washing steps. Bound antibody was revealed with FITCconjugated sheep anti-mouse IgG (Caltag). MOPC 21 (Sigma) was used as an irrelevant mouse IgG antibody in these stainings. Stained cells were analyzed using a FACScan (Becton Dickinson) and the data processed using the software provided. Cell survival assays were performed on sorted B2201 bone marrow cells from control mice, Eµ-bcl-2-36 transgenic mice, scid mice, Eµ-bcl-2-36/scid mice, rag-1–/– mice, Eµ-bcl-2-36/rag-1–/– mice, µMT mice and Eµ-bcl-2-36/ µMT mice. Ten thousand sorted cells were resuspended in 10 µl of the high glucose version of Dulbecco’s modified Eagle’s medium supplemented with 13 µM folic acid, 250 µM L-asparagine, 50 µM 2-mercaptoethanol and 10% fetal bovine serum, deposited into a single well of a Terasaki plate, and incubated at 37°C in 5% CO2. At regular intervals the fraction of viable cells was determined by visual inspection using phase contrast on an inverted microscope. Cloning and sequencing of DHJH junctions B2201CD231 and B2201CD23– cells were isolated by cell sorting from the bone marrow of bcl-2/scid mice as described above, and genomic DNA was prepared as previously described (23). DHJH junctions were amplified by PCR using the primers and protocol as described (23). Rearrangements were cloned into pGEM5, recombinant plasmid DNAs were purified and directly sequenced using the M13 universal primer and the ABI PRISM sequencing kit (Applied Biosystems, Foster City, CA) according to the manufacturers protocol. The sequences were analyzed using an automated sequencer ABI 360. Results bcl-2 transgene expression promotes accumulation of B lymphocytes in scid, rag-1–/– and µMT mutant mice The ability of bcl-2 transgene expression to rescue B cells in mice homozygous for scid, rag-1 and µMT mutations was determined by enumerating B2201 cells in the bone marrow and spleen. Without the transgene, B cell differentiation is

1484 Cell survival and differentiation are distinct

Fig. 2. bcl-2 transgene expression promotes accumulation of surface IgM– B lineage cells in mutant scid, rag-1–/– and µMT Mice. (A) The proportion of B2201sIgM– cells in the bone marrow of wild-type, scid, rag-1–/– and µMT mice was determined by flow cytometric analysis of cells doubly stained with fluorochrome-conjugated antibodies specific for CD45R(B220) and IgM. Staining profiles from control littermates are shown in the upper row and from bcl-2 transgenic mice in the lower row. (B) As for (A), but presenting analysis of spleen cells. The mean number of nucleated cells recovered from each femur or spleen is indicated. At least three animals of each genotype were examined, with a representative staining profile being shown. A summary of these data is presented in Table 1.

arrested at the pro-B/pre-BI stage in these mutant animals, and few B2201 cells can be found in the bone marrow, peripheral blood and spleen (Fig. 2, Table 1 and data not shown). As previously reported (33), bcl-2 transgene expres-

sion resulted in the accumulation of near normal numbers of B2201 cells in bone marrow, spleen and blood of bcl-2/scid mice (Fig. 2 and Table 1). Similar accumulation of B lineage cells was observed in rag-1–/– and µMT mice expressing the

Cell survival and differentiation are distinct 1485 Table 1. The effect of bcl-2 transgene expression on accumulation of B2201 cells in wild-type and rearrangement-deficient mice Bone marrow (3106)a

Strain

Wild-type bcl-2 scid bcl-2/scid rag-1–/– bcl-2/rag-1–/– µMT bcl-2/µMT aNumber bNumber

Spleen (3106)

B2201 totalb

B2201CD231

6.4 6 0.7 18.4 6 0.9 1.1 6 0.3 7.8 6 1.1 1.4 6 0.6 10.8 6 3.3 0.9 6 0.1 4.7 6 0.6

1.5 7.7 ,0.2 1.6 ,0.2 ,0.2 ,0.2 0.6

6 0.2 6 1.0 6 0.4

6 0.2

B2201 total

B2201CD231

62 6 7 331 6 7 1.0 6 0.6 22.3 6 8.7 1.3 6 0.8 22.8 6 9.1 1.3 6 0.5 15 6 1.4

52 6 7 267 6 13 ,0.5 11.0 6 6 ,0.5 ,0.5 ,0.5 4.5 6 1.1

of cells per femur or per spleen, calculated as an average 6 SD of four or five mice per group of cells of each phenotype calculated from the percentages determined by immunofluorsecent staining and flow cytometry.

same bcl-2 transgene (Fig. 2 and Table 1). Cell cycle and cell turnover analysis of these B2201sIg– B lineage cells revealed that essentially all are non-dividing and as longlived as recirculating, mature sIgD1 B lymphocytes in normal mice (L. O’Reilly et al., manuscript in preparation). Differentiation of B lymphoid cells occurs in bcl-2/scid and bcl-2/µMT but not in bcl-2/rag-1–/– mice The B lymphoid cells that had accumulated within the bone marrow, peripheral blood and spleen of bcl-2 transgenic scid, rag-1–/– and µMT mice expressing the bcl-2 transgene were examined for differentiation beyond the pro-B/pre-BI stage by measuring the proportion of B2201 cells that were CD43–, CD211, CD221 and CD231. In scid, µMT and rag1–/– mice all bone marrow and splenic B2201 cells expressed CD43 and none expressed CD21, CD22 or CD23 (Fig. 3 and data not shown), consistent with previous reports (11,12,16,39). Expression of a bcl-2 transgene promoted maturation of some B2201 cells to the CD211CD221CD231 stage in both scid and µMT mice but notably not in rag-1–/– mice. The distribution of CD23 on bone marrow and spleen cells from control and transgenic mice has been used to exemplify this point (Fig. 3). The proportion of mature cells was greater in bcl-2/scid mice than in bcl-2/µMT animals (20– 40 versus 5–15% of all B2201 cells). Differentiation was not due to rare cells generating productive IgM and IgL gene rearrangements since all B2201 cells remained sIgM– as determined by immunofluorescence staining and flow cytometric analysis (Fig. 2). The ability of B lineage cells from the various mutant mice to respond to mitogenic stimulation in vitro correlated with their cell surface marker profile. B cells from rag-1–/– and bcl2/rag-1–/– mice did not proliferate and, as expected, did not secrete Ig upon co-culture with CD40 ligand-transfected NIH 3T3 fibroblasts (Table 2). Similarly, immature B2201CD23– B lineage cells from scid, bcl-2/scid, µMT and bcl-2/µMT mice also failed to respond (Table 2). In contrast, the more differentiated B2201CD231 B cells from bcl-2/scid and bcl-2/µMT mice proliferated and, in the case of the bcl-2/µMT cultures, they matured into antibody-secreting cells upon CD40 stimulation (Table 2). Although some secreted Ig was detected in the supernatants of B2201CD23– cultures from bcl-2 transgenic µMT mice, the level was ~5-fold less than in the CD231

cultures. No IgM or IgG1 was secreted by either scid or rag1–/– B cells, irrespective of the presence of the bcl-2 transgene. An additional indicator of the differentiation potential of B lineage cells in the various mutant strains was obtained by measuring serum Ig levels. As expected, no Ig of any isotype was detectable in serum from rag-1–/– mice, irrespective of the presence or absence of the bcl-2 transgene (Fig. 4). In contrast, two of the six scid samples had low but detectable levels of IgA, IgG3, IgG2b and IgG2a, and all scid mice showed low levels of IgM. These Igs presumably arise as a result of the known ‘leakiness’ of the scid mutation (40), although it did not extend to IgG1. The bcl-2 transgene did not significantly alter the frequency of serum Ig1 scid mice or the levels of Ig in the serum except for IgG2b which was present in all bcl-2/scid samples. The findings with the µMT samples were strikingly different: in the absence of a bcl-2 transgene, no Ig of any class (save occasional traces of IgM) was detectable whereas all sera from bcl-2/µMT mice contained IgA, IgG2b, IgG2a, IgG1 and IgM. The levels of each Ig isotype were consistently higher than in any of the other strains and approached those seen in normal mice. IgG3 was, however, undetectable in three bcl-2/µMT serum samples and was low in the other two (Fig. 4). In sum, these data demonstrate that the ability of Bcl-2 to promote further B cell differentiation was restricted to the two mutant strains in which rearrangement at the IgH locus can be initiated. Maturation of B cells in the bone marrow of bcl-2/scid mice is not associated with extensive proliferation Having observed differentiation beyond the pro-B/pre-BI stage in bcl-2 transgenic mutant scid and µMT mice, we wished to determine whether this was associated with proliferation, as would occur at the early pre-BII stage in normal animals. To this end we purified pro-B/pre-BI cells (B2201CD431) and the more mature B lymphoid cells (B2201CD43–) from bone marrow of BALB/c mice, scid mice (pro-B/pre-BI cells only) and bcl-2/scid mice, and determined their cell cycle distribution by flow cytometry. The results (Table 3) showed that the more mature B lineage cells in bcl2/scid mice were not proliferating to any measurable extent. This was in contrast to the equivalent B cell fraction in control animals, where 20% of cells were in the S, G2 or M phase of

1486 Cell survival and differentiation are distinct

Fig. 3. bcl-2 transgene expression promotes differentiation of B2201 cells in scid and µMT mice, but not in rag-1–/– mice. The degree of B lymphoid differentiation was determined by measuring the proportion of cells simultaneously expressing B220, CD21, CD22 and CD23 (shown in this figure). The upper row in each part of the figure is an example of the staining found in non-transgenic animals, while the lower panel is from bcl-2 transgenic animals. (A) Bone marrow. (B) Spleen. Details of the number of animals analyzed and the number of cells in each phenotypic compartment are given in Table 1.

the cycle. Similar observations were made in analyzing B lineage cells from bcl-2/µMT and control µMT animals (data not shown). These results demonstrate that differentiation beyond the pro-B/pre-BI stage in bcl-2/scid and bcl-2/µMT mice is not accompanied by continued proliferation.

The bcl-2 transgene is expressed functionally at equal levels in B lymphoid cells of bcl-2/rag-1–/–, bcl-2/µMT and bcl-2/ scid mice A possible explanation for the lack of B cell differentiation in bcl-2/rag-1–/– mice is that the timing or level of expression of

Cell survival and differentiation are distinct 1487

Fig. 4. bcl-2 transgene expression promotes development of antibody-secreting cells in mutant µMT mice. Serum samples from the indicated number of adult scid, bcl-2/scid, rag-1–/–, bcl-2/rag-1–/–, µMT and bcl-2/µMT mice were titrated against defined standards in specific ELISA assays to determine the concentration of each isotype. The level of each Ig isotype in samples from recombination proficient animals on the same genetic background (C57BL/6 and BALB/c, lacking a bcl-2 transgene) housed in the same facility as the experimental animals is indicated by a grey line in each panel. Similarly, the detection limit for each Ig class is indicated by the dashed line. bcl-2 transgenic animals on a wild-type background have serum Ig levels that are ~2-fold higher than littermate controls (34).

Table 2. Differentiation potential of B lymphocyte subsets Strain

Wild-type bcl-2 scid bcl-2/scid rag-1–/– bcl-2/rag-1–/– µMT bcl-2/µMT

Proliferation (SI 6 SD)

IgG1 secretion (pg/ml)

B2201CD23–

B2201CD231

B2201CD23–

B2201CD231

3.7 6 0.2 11.0 6 1.0 1.0 6 0.1 1.1 6 0.1 1.0 6 0.1 1.0 6 0.1 1.0 6 0.1 1.0 6 0.1

9.4 6 3.5 12.0 6 1.0

3700 3800 ND ND ND ND ND 250

3600 2100

10.5 6 2.0

2.9 6 0.3

ND

1300

Sorted cells were stimulated with CD40 ligand in the presence of IL-4 and IL-5. Proliferation was measured on day 3 by [3H]thymidine incorporation. IgG1 secretion was measured in day 7 culture supernatants. IgM secretion, although not shown, was also measured and mirrored IgG1. Sorted cells cultured in the absence of CD40 ligand did not secrete detectable levels of IgG1. ND: not above the detection limit of 30 pg/ml. Where a cell type does not exist, no value is entered.

1488 Cell survival and differentiation are distinct

Fig. 5. Expression of the Eµ-bcl-2-36 transgene is not influenced by genetic background. (A) Intracellular levels of human Bcl-2 were determined by flow cytometric analysis of permeabilized bone marrow cells from bcl-2 transgenic (solid line) and littermate control (dashed line) mice on a wild-type, scid, rag-1–/– and µMT background. (B) B2201 B lineage cells were purified from bcl-2 transgenic and littermate control mice on a wild-type, scid, rag-1–/– and µMT background, and suspended in simple tissue culture medium. Their viability was determined at the indicated times. Values shown are arithmetic means 6 SD from three mice of each genotype.

Table 3. bcl-2 transgene expression promotes B lymphocyte differentiation but not proliferation in scid mice Strain

Cell phenotype

BALB/c

B2201CD431Ig– B2201CD43–

scid bcl-2/scid

B2201CD431 B2201CD43–

Percent of sorted cells in G0/G1

S

G2 1 M

70 80 71 100

21 17 20 0

9 3 9 0

the transgenic Bcl-2 protein differs from that in bcl-2/scid and bcl-2/µMT mice. The fact that all strains accumulated similar numbers of B2201 cells in bone marrow, spleen and peripheral blood (Fig. 2 and Table 1) suggested that this was not in fact the case. Immunofluorescence staining of fixed and permeabilized cells revealed that B2201 cells from bcl-2 transgenic rag-1–/–, scid and µMT mice all expressed the same level of transgene encoded human Bcl-2 protein (Fig. 5). Furthermore bcl-2 transgene expression enhanced the in vitro survival of sorted B2201 cells from bcl-2/rag-1–/–, bcl-2/scid and bcl-2/µMT mice to a similar extent (Fig. 5). Collectively these results rule out the possibility that the observed differences in differentiation of B2201 cells reflect differences in transgene expression. Expression of surface Ig classes other than IgM promotes B cell differentiation in bcl-2/µMT mice The inability of B2201 cells from bcl-2/rag-1–/– animals to undergo differentiation when compared to either bcl-2/scid or bcl-2/µMT mice suggested that the capacity to initiate Ig gene rearrangement may be involved. Although B cells from µMT mice cannot express surface µ IgH chain, they can still generate functional IgH gene rearrangements (17). Since a

small fraction of B cells are known to undergo Igκ gene rearrangement in the absence of productive IgH gene rearrangements (15,17,41,42), µMT mice might contain a population of B cell precursors with both functional IgH and IgL gene rearrangements. Heavy chain isotype switching would then allow expression of membrane-bound Ig of another class. Accordingly, inhibiting apoptosis by over-expression of Bcl-2 would be expected to increase the probability of these three events occurring within the same cell. Three arguments suggest that the alternative isotype most likely to replace the function of µ in bcl-2/µMT mice is δ. First, transcription of the δ locus occurs at developmental stages earlier than those in which IgD is expressed on the cell surface (43). Second, a B cell does not have to undergo isotype switch recombination, but only alternative IgH mRNA splicing to express IgD (44). Third, experiments in transgenic mice have demonstrated that Cδ can replace Cµ in B cell differentiation (45,46). To determine if sδ1sµ– cells were present in bcl-2/µMT mice, spleen cells were examined for the presence of B2201 cells expressing IgD on their surface. Such cells were indeed found and furthermore all were shown to express CD23 (Fig. 6). No cells of this phenotype were found in control µMT mice. Thus, in bcl-2/µMT mice, a significant fraction of the CD231 subset of B2201 cells expressed surface IgD (Fig. 6); the remaining CD231 cells presumably expressed more distal IgH isotypes. These results suggested that signaling via sIg receptors containing heavy chains other than µ can promote differentiation of pro-B/pre-BI cells. Expression of Dµ protein may promote B cell differentiation in bcl-2/scid mice Previous analysis of the IgH gene loci of scid B cells has shown that the majority of recognizable rearrangement junctions are DHJH, rather than VHDHJH (15,33). Since we were unable to detect surface IgM on CD231B2201 B lineage cells in bcl-2/

Cell survival and differentiation are distinct 1489

Fig. 6. B2201 cells in bcl-2/µMT mice express sIg. Spleen cells from bcl-2/µMT mice and control µMT littermates were stained with combinations of antibodies specific for CD45R(B220), IgD and CD23. Ten thousand events were collected for each sample. All sIgD1 cells were shown to be simultaneously CD231. The staining profiles shown are representative of those obtained from analysis of at least five mice of each genotype.

scid mice (Fig. 2 and 33), we considered the possibility that their differentiation might have resulted from the presence of Dµ-encoding DHJH rearrangements. A fraction of DHJH rearrangements in bcl-2/scid B2201 cells might be joined in rf2 with sufficient coding sequence to generate a Dµ polypeptide. To examine this possibility, B2201CD231 and B2201CD23– cells were sorted from the bone marrow of bcl2/scid mice, and DHJH rearrangements amplified by PCR from their genomic DNA. These junctions were recovered and individual clones sequenced (Fig. 7 and Table 4). From the B2201CD23– fraction we recovered DHJH junctions containing substantial deletions, often encompassing both the DH and JH elements. Where both the DH and JH elements were identifiable, they invariably encoded rf1 and rf3 junctions as has been observed for pre-BII cells, virgin B cells and mature B cells in normal animals (19,20). In striking contrast, 50% of the junctions recovered from the B2201CD231 cells of bcl-2/ scid mice could potentially encode a Dµ protein; the remainder were in rf1, in rf3 or could not be assigned a reading frame due to extensive deletion of the DH or JH elements (Fig. 7 and Table 4). Although the DH coding segment itself was sometimes deleted, the sequences remaining still contained the potential promoter, initiation codon, leader sequence and the JH splice donor necessary for proper RNA processing. The distribution of DHJH reading frames requires some comment. If B cell differentiation in bcl-2/scid mice is mediated by Dµ expression, then we would expect every B2201CD231 cell to contain a rf2 DHJH junction while no B2201CD23– cell should contain such a junction. If expression of the Dµ protein

triggered both differentiation and the cessation of further gene rearrangement at the IgH loci (23), then one would expect 56% of DHJH rearrangements in such a population to be rf2: three of nine from the first IgH locus to rearrange plus two of nine from the second. Our observations closely matched these expectations with 50% of DHJH junctions recovered from B2201CD231 cells being rf2, while none of nine such junctions from B2201CD23– cells were rf2 (Table 3). Using currently available detection systems such as flow cytometry and immunoprecipitation we have been unable to reproducibly demonstrate a Cµ-containing protein on the surface of the bone marrow B cell precursors in bcl-2/scid mice. This failure may reflect either the inherent difficulty in detecting Dµ protein on the surface of pre-B cells ex vivo and/or the transient nature of its expression in bcl-2/scid mice. Although differentiation out of the pro-B/pre-BI compartment requires an Ig-mediated signal, persistence of the more mature phenotype may be independent of such a signal in the presence of transgenic BCL-2. Despite this, the genetic data provides evidence that differentiation of bcl-2/scid B cells is regulated by a receptor complex containing a Dµ protein. Discussion B cell differentiation proceeds through a succession of genotypically and phenotypically identifiable stages, and is regulated by the interaction of several processes (47). These include proliferation of clones with productive IgH gene rearrangements (28), growth arrest at the point of IgL gene

1490 Cell survival and differentiation are distinct

Cell survival and differentiation are distinct 1491 Table 4. DH reading frame usage in B2201 cells from bcl-2/ scid mice Cell phenotype B2201CD23– B2201CD231

No. of clones in DHJH reading frame 1

2

3

Undefined

Total

3 7

0 10

1 0

5 4

9 21

rearrangement, apoptosis of cells lacking sIg and positive selection of those with potentially useful antigen receptor specificities into the pool of long-lived mature, recirculating B lymphocytes (Fig. 1). We wanted to determine whether differentiation requires active signaling via surface receptors containing IgH chains or is merely a consequence of sustained B cell survival. This was achieved by constructing novel strains of mice bearing mutations that block Ig gene rearrangement (scid, rag-1–/–) or plasma membrane deposition of sIgM (µMT) and that express in their B cells a bcl-2 transgene to inhibit apoptosis. We found that over-expression of Bcl-2 was sufficient to allow a large fraction of B lymphoid cells to survive, emigrate from the bone marrow, and accumulate in peripheral lymphoid organs of bcl-2/rag-1–/–, bcl-2/µMT and bcl-2/scid mice. The degree of rescue was similar in each of these strains (Fig. 2 and Table 1). These strains differed, however, in whether or not the persisting cells underwent further differentiation in vivo as manifested by expression of markers characteristic of mature B cells. Such markers appeared on a subset of the B2201 cells in the bone marrow and spleen of bcl-2/scid and bcl-2/µMT mice, but on none of the B2201 cells in bcl-2/rag-1–/– mice (Fig. 3 and Table 1). Expression of the more differentiated cell surface marker phenotype correlated with the capacity to respond to mitogenic stimulation in vitro (Table 2). Clearly the developmental potential of pro-B/pre-BI cells in bcl-2/scid and bcl-2/µMT mice differs fundamentally from those in bcl-2/rag-1–/– mice. Detailed phenotypic and genotypic examination of the mature B cells in bcl-2/scid and bcl-2/µMT mice revealed that their differentiation resulted from expression of receptors containing either Dµ protein or IgH chains of an isotype other than µ (Figs 6 and 7). Thus, continued B cell differentiation requires signaling via an Ig or Ig-like surface receptor and cannot be simply a consequence of blocking physiological cell death. Implications for the normal function of the pre-B receptor Despite the differentiation of B cell precursors beyond the pro-B/pre-BI stage in bcl-2/scid and bcl-2/µMT mice, no

proliferation was apparent among the more mature cells (Table 2). In contrast, normal B cells undergo an estimated four to six divisions at the early pre-BII stage (28,29). This discrepancy suggests that proliferation of B cell precursors at the pre-BII stage probably is dependent on surface expression of a complete VHDHJHCµ molecule and cannot be replaced by a Dµ protein. The stimulus for pre-BII cells to proliferate is mediated by the so-called pre-B cell receptor, comprising a full length µ IgH chain associated with the surrogate light chain complex, λ5 and Vpre-B, associated with mb-1 (Igα) and B29 (Igβ) co-receptor molecules (5,48). While surface expression of the pre-B receptor has been demonstrated (5), it is still unclear whether a specific ligand exists or whether receptor assembly alone is sufficient to trigger proliferation. In either case, it appears that the forms of pre-B receptor expressed in bcl-2/scid and bcl-2/µMT mice (presumably Dµ, λ5, Vpre-B and δ, λ5, Vpre-B respectively) can only induce differentiation; they are insufficient to trigger proliferation. Pre-B receptor-mediated differentiation and proliferation can thus be dissociated, providing evidence that they are regulated by distinct signaling routes. Implications for the fate of Dµ protein-expressing B cell precursors in normal mice Our results provide evidence that continued B cell differentiation in bcl-2/scid mice is due to expression of a Dµ protein receptor (Fig. 7). We have also shown that the more mature bcl-2/scid B cells are quiescent (Table 3). Taken together, these results provide insight into the fate of Dµ-expressing B cell precursors in normal mice and may explain why cells bearing rf2 DHJH rearrangements are selectively under represented among B lymphocytes beyond the pro-B/pre-BI stage (4,19,20). A normal pro-B/pre-BI cell bearing a rf2 DHJH rearrangement would receive a differentiation-inducing signal via a Dµ-containing pre-B receptor similar to that transduced by a full-size µ-containing pre-B receptor. This signal would block further rearrangement at the IgH locus and the cell would enter the pre-BII stage. The normal fate of a pre-BII cell would be a period of rapid clonal expansion (an estimated four to six divisions) followed by growth arrest and initiation of Ig light chain gene rearrangement. If Dµ-expressing B cells in wild-type animals behaved like the putative Dµ-expressing cells in bcl-2/scid mice and failed to proliferate after entry into the pre-BII compartment, then they would simply be lost from the B cell repertoire by dilution. Dµ-encoding rearrangements presumably also arise in scid pro-B cells but do not result in the appearance of a more mature B cell population. The appearance of such a population would therefore appear to depend on the survival function of the bcl-2 transgene,

Fig. 7. The distribution of DHJH reading frames in B cell subsets from the bone marrow of bcl-2/scid mice. The nucleotide sequences and corresponding reading frames of DHJH rearrangements amplified and cloned from (A) B2201CD231 and (B) B2201CD23– cells are shown. The first sequence in each panel is that of germline DSP2.Y and JH4 elements, indicated by a G. Identity between the test sequences and this reference are indicated by a dash. Sequence differences that do not correspond to known germline differences between DH elements (19) are indicated by lower case letters. The putative Dµ leader sequence is indicated, as are N regions and the D elements themselves. In some cases the entire JH coding sequence was deleted which is indicated by a blank line. DH element reading frame was determined from the upstream ATG. The putative leader sequence of Dµ, the DH element coding sequence, N region and JH coding sequences are indicated. DFl16.1-containing sequences are aligned for maximum homology with the DSP2.Y, thereby placing the Dµ ATG 45 bp upstream of the sequence displayed (54). The heptamer and nonamer recombination signal sequences are underlined in the reference sequence.

1492 Cell survival and differentiation are distinct suggesting that the Dµ protein is unable to support the continued survival of B cell precursors because of either its signaling properties or its transient expression. This observation reinforces the role of continued signaling through the B cell receptor in B cell maturation (49). Implications for the regulation of IgD expression and IgD function The continued differentiation of a fraction of the B cell precursors in bcl-2/µMT mice appears to be mediated by surface expression of IgD. This IgD presumably arises from splicing of the rearranged VHDHJH gene segment to the Cδ exons contained on a mRNA precursor initiated at the normal V segment promoter. The VHDHJH–Cδ transcript is very likely to be generated by alternate splicing since in IgH allotype heterozygous µMT1/– mice, the targeted µ locus is not subject to allelic exclusion and the allotype of the ‘included’ IgD is the same as that of the targeted µ locus, suggesting a cisacting phenomenon (17 and our unpublished data). Irrespective of the mechanism by which the productive VHDHJH gene segment becomes attached to Cδ, the resultant protein can mediate the differentiation of the B cell precursor. Previous work has established that expression of an IgD transgene can ‘allelically exclude’ the endogenous IgH loci, can trigger IgL gene rearrangement and can allow apparently normal phenotypic development to a state of antigen responsiveness (45,46). Data presented here confirm the ability of IgD to promote B cell differentiation, and furthermore imply that IgD can enable the remaining downstream steps of B cell activation such as Ig isotype switching and antibody secretion (Figs 4 and 6). A remarkable feature of the bcl-2/µMT mice is the regular production of all classes of circulating Ig except IgM and the T cell-independent isotype IgG3 (Fig. 4). Moreover, sIgD1 B cells sorted from bcl-2/µMT mice undergo isotype switching in vitro upon stimulation with CD40 ligand and IL-4 (data not shown). Together these results suggest that the serum Ig in bcl-2/µMT mice is the result of T cell-dependent immune responses involving the sIg1 B cells. Other studies have shown that the T cell compartment in µMT mice is functional (50) and should thus be able to provide help to antigen-activated B cells. The low level of the T cell-independent isotype IgG3 in bcl-2/µMT mice has interesting implications for the ability of the Ig expressed on these cells to mediate all types of B cell activation. While the relatively low density of Ig molecules on the surface of bcl-2/µMT B cells may be responsible for their inability to mediate T cellindependent activation, our results may provide evidence that sIgM is essential for this process. Finally, the fact that a fraction of bcl-2/µMT B cells expresses sIgD implies that such expression is not dependent on prior surface expression of IgM.

fail to undergo positive selection (51,52). In contrast to these conclusions, data from other experimental systems have been interpreted as supporting the hypothesis that control of differentiation of hematopoietic precursors is intrinsic to the cells and that continued differentiation is therefore a natural consequence of sustained cell survival (53). In that model, signaling by external factors such as cytokines or colony stimulating factors functions merely to prevent apoptosis of precursor cells. This conclusion was based on the observation that over-expression of Bcl-2 facilitated both survival and differentiation in growth factor deprived cells of the FDCP Mix cell line (53). The exact relationship between this immortalized hematopoietic progenitor line and its normal counterpart, however, is unclear. While our results derive from the lymphoid rather than the myeloid lineage, they directly address the interdependence of cell survival and differentiation within normal cells in vivo. Our data from bcl-2/rag-1–/– mice strongly support the view that cell survival alone is insufficient to promote differentiation of precursors. B cell differentiation requires specific signals which may well have an effect on cell survival, but this is distinct to their effect on cell maturation. In the case of B lymphocytes, the differentiationinducing signal requires surface expression of an Ig or Iglike protein. We speculate that, analogous to the function of antigen receptors in lymphocytes, receptors for colony stimulating factors could also trigger distinct signaling pathways for differentiation, proliferation and survival of myeloid and erythroid cells. Acknowledgements We thank Drs A. Harris, S. Cory, D. Mason, T. Kinoshita, E. Clark, H. Karasuyama, A. Rolink, F. Melchers and K. Rajewsky for their generous gifts of mutant mice, cell lines, antibodies and cDNA clones, and Drs S. Cory, J. Adams, A. Harris and G. Nossal for discussions and comments on the manuscript. We are grateful to M. Stanley, L. Barlough and A. Light for expert technical assistance, J. Scott and K. Patane for animal husbandry, and Dr F. Battye for help with flow cytometry. D. T. was the recipient of a fellowship from the National Health and Medical Research Council (Canberra), L. C. was a Cancer Research Institute investigator, while A. S. was supported by the Leukemia Society of America, the Swiss National Science Foundation and by a Clinical Investigator Award of the Cancer Research Institute. This work was supported by the National Health and Medical Research Council (Canberra), the US National Cancer Institute (CA43540), and by a grant from the Human Frontier Science Program (Principal Investigator Professor D. Mathis).

Abbreviations Cδ Cµ rf sIgM

IgD heavy chain constant region IgM heavy chain constant region translational reading frame of the DH element in IgH rearrangements surface IgM

General implications for cell differentiation Our data demonstrate that inhibiting apoptosis is not sufficient to promote differentiation of B lineage cells, providing evidence that differentiation and cell survival are subject to distinct control. This interpretation is compatible with previous observations that Bcl-2 over-expression enhances survival but does not promote differentiation of T cell precursors that

References 1 Metcalf, D. 1989. The molecular control of cell division, differentiation commitment and maturation in haemopoietic cells. Nature 339:27. 2 Rolink, A., Karasuyama, H., Haasner, D., Grawunder, U., Martensson, I. L., Kudo, A. and Melchers, F. 1994. Two pathways of B-lymphocyte development in mouse bone marrow and the

Cell survival and differentiation are distinct 1493

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20 21 22

23

24

roles of surrogate L chain in this development. Immunol. Rev. 137:185. Hardy, R. R., Carmack, C. E., Shinton, S. A., Kemp, J. D. and Hayakawa, K. 1991. Resolution and characterization of pro-B and pre-pro-B cell stages in normal mouse bone marrow. J. Exp. Med. 173:1213. Ehlich, A., Martin, V., Muller, W. and Rajewsky, K. 1994. Analysis of the B-cell progenitor compartment at the level of single cells. Curr. Biol. 4:573. Karasuyama, H. and Rolink, A. 1994. The expression of Vpre-B/λ5 surrogate light chain in early bone marrow precursor B cells of normal and B cell-deficient mutant mice. Cell 77:133. Nemazee, D. A. and Burki, K. 1989. Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody gene. Nature 337:562. Hartley, S. B., Crosbie, J., Brink, R., Kantor, A. B., Basten, A. and Goodnow, C. C. 1991. Elimination from peripheral lymphoid tissues of self-reactive B lymphocytes recognizing membranebound antigens. Nature 353:765. Gu, H., Tarlinton, D., Muller, W., Rajewsky, K. and Forster, I. 1991. Most peripheral B cells in mice are ligand selected. J. Exp. Med. 173:1357. Cyster, J. G., Hartley, S. B. and Goodnow, C. C. 1994. Competition for follicular niches excludes self-reactive cells from the recirculating B-cell repertoire. Nature 371:389. Torres, R. M., Flaswinkel, H., Reth, M. and Rajewsky, K. 1996. Aberrant B cell development and immune response in mice with a compromised BCR complex. Science 272:1804. Mombaerts, P., Iacomini, J., Johnson, R. S., Herrup, K., Tonegawa, S. and Papaioannou, V. E. 1992. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68:869. Shinkai, Y., Rathbun, G., Lam, K.-P., Oltz, E. M., Stewart, V., Mendelsohn, M., Charron, J., Datta, M., Young, F., Stall, A. M. and Alt, F. W. 1992. RAG-2- deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangements. Cell 68:855. Bosma, G. C., Custer, R. P. and Bosma, M. J. 1983. A severe combined immunodeficiency mutation in the mouse. Nature 301:527. Schuler, W., Weiler, I. J., Schuler, A., Phillips, R. A., Rosenberg, N., Mak, T. W., Kearney, J. F., Perry, R. P. and Bosma, M. J. 1986. Rearrangement of antigen receptor genes is defective in mice with severe combined immune deficiency. Cell 46:963. Pennycook, J. L., Chang, Y., Celler, J., Phillips, R. A. and Wu, G. E. 1993. High frequency of normal DJH joints in B cell progenitors in severe combined immunodeficiency mice. J. Exp. Med. 178:1007. Kitamura, D., Roes, J., Kuhn, R. and Rajewsky, K. 1991. A B celldeficient mouse by targeted disruption of the membrane exon of the immunoglobulin µ chain gene. Nature 350:423. Kitamura, D. and Rajewsky, K. 1992. Targeted disruption of µ chain membrane exon causes loss of heavy-chain allelic exclusion. Nature 356:154. Gong, S. C. and Nussenzweig, M. C. 1996. Regulation of an early developmental checkpoint in the B cell pathway by Igb. Science 272:411. Gu, H., Kitamura, D. and Rajewsky, K. 1991. B cell development regulated by gene rearrangement: arrest of maturation by membrane-bound Dµ protein and selection of DH element reading frames. Cell 65:47. Meek, K. 1990. Analysis of junctional diversity during B lymphocyte development. Science 250:820. Kaartinen, M. and Makel, O. 1985. Reading of D genes in variable frames as a source of antibody diversity. Immunol. Today 6:324. Reth, M. G. and Alt, F. W. 1984. Novel immunoglobulin heavy chains are produced from DJH gene segment rearrangements in lymphoid cells. Nature 312:418. Tarlinton, D., Strasser, A., McLean, M. and Basten, A. 1995. DH element reading frame selection is influenced by an Ig heavy chain transgene, but not by bcl-2. J. Immunol. 154:3341. Rolink, A., Kudo, A., Karasuyama, H., Kikuchi, Y. and Melchers, F. 1991. Long-term proliferating early pre B cell lines and clones

25

26

27 28

29 30

31

32

33 34

35 36 37 38

39

40 41

42

43

with the potential to develop to surface Ig-positive, mitogen reactive B cells in vitro and in vivo. EMBO J. 10:327. Peschon, J. J., Morrissey, P. J., Grabstein, K. H., Ramsdell, F. J., Maraskovsky, E., Gliniak, B. C., Park, L. S., Ziegler, S. F., Williams, D. E., Ware, C. B., et al. 1994. Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deficient mice. J. Exp. Med. 180:1955. Kodama, H., Nose, M., Hamaguchi, Y., Tsunoda, J.-i., Suda, T., Nishikawa, N. and Nishikawa, S.-i. 1992. In vitro proliferation of primitive hemopoietic stem cells supported by stromal cells: evidence for the presence of a mechanism(s) other than that involving c-kit receptor and its ligand. J. Exp. Med. 176:351. Kitamura, D., Kudo, A., Schaal, S., Muller, W., Melchers, F. and Rajewsky, K. 1992. A critical role of λ5 protein in B cell development. Cell 69:823. Grawunder, U., Leu, T. M. J., Schatz, D. G., Werner, A., Rolink, A. G., Melchers, F. and Winkler, T. H. 1995. Down-regulation of RAG1 and RAG2 gene expression in preB cells after functional immunoglobulin heavy chain rearrangement. Immunity 3:601. Osmond, D. G. 1993. The turnover of B-cell populations. Immunol Today 14:34. Reichman-Fried, M., Hardy, R. R. and Bosma, M. J. 1990. Development of B-lineage cells in the bone marrow of scid/ scid mice following the introduction of functionally rearranged immunoglobulin transgenes. Proc. Natl Acad. Sci. USA 87:2730. Spanopoulou, E., Roman, C. A. J., Corcoran, L. M., Schlissel, M. S., Silver, D. P., Nemazee, D., Nussenzweig, M. C., Shinton, S. A., Hardy, R. A. and Baltimore, D. 1994. Functional immunoglobulin transgenes guide ordered B-cell differentiation in Rag-1-deficient mice. Genes Dev. 8:1030. Young, F., Ardman, B., Shinkai, Y., Lansford, R., Blackwell, T. K., Mendelshon, M., Rolink, A., Melchers, F. and Alt, F. W. 1994. Influence of immunoglobulin heavy- and light-chain expression on B-cell differentiation. Genes Dev. 8:1043. Strasser, A., Harris, A. W., Corcoran, L. M. and Cory, S. 1994. bcl-2 expression promotes B but not T lymphoid development in scid mice. Nature 368:457. Strasser, A., Whittingham, S., Vaux, D. L., Bath, M. L., Adams, J. M., Cory, S. and Harris, A. W. 1991. Enforced BCL2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease. Proc. Natl Acad. Sci. USA 88:8661. Strasser, A., Harris, A. W., Huang, D. C. S., Krammer, P. H. and Cory, S. 1995. Bcl-2 and Fas/APO-1 regulate distinct pathways to lymphocyte apoptosis. EMBO J. 14:6136. Strasser, A., Harris, A. W. and Cory, S. 1991. Bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell 67:889. Tarlinton, D., Mclean, M. and Nossal, G. J. V. 1995. B1 and B2 cells differ in their potential to switch immunoglobulin isotype. Eur. J. Immunol. 25:3388. Pezzella, F., Tse, A. G. D., Cordell, J. L., Pulford, K. A. F., Gatter, K. C. and Mason, D. Y. 1990. Expression of the bcl-2 oncogene protein is not specific for the 14;18 chromosomal translocation. Am. J. Pathol. 137:225. Hardy, R. R., Kemp, J. D. and Hayakawa, K. 1989. Analysis of lymphoid population in scid mice; detection of a potential B lymphocyte progenitor population present at normal levels in scid mice by three color flow cytometry with B220 and S7. Curr. Top. Microbiol. Immunol. 152:19. Bosma, M. J. and Carroll, A. M. 1991. The scid mouse mutant: definition, characterization, and potential uses. Annu. Rev. Immunol. 9:323. Kubagawa, H., Cooper, M. D., Carroll, A. J. and Burrows, P. D. 1989. Light-chain gene expression before heavy-chain gene rearrangement in pre-B cells transformed by Epstein–Barr virus. Proc. Natl Acad. Sci. USA 86:2356. Ehlich, A., Schaal, S., Gu, H., Kitamura, D., Muller, W. and Rajewsky, K. 1993. Immunoglobulin heavy and light chain genes rearrange independently at early stages of B cell development. Cell 72:695. Yuan, D. and Tucker, P. W. 1983. Transcriptional regulation of the µ-δ heavy chain locus in normal murine B lymphocytes. J. Exp. Med. 160:564.

1494 Cell survival and differentiation are distinct 44 Maki, R., Roeder, W., Traunecker, A., Sidman, C., Wabl, M., Raschke, W. and Tonegawa, S. 1981. The role of DNA rearrangement and alternative RNA processing in the expression of immunoglobulin δ genes. Cell 24:353. 45 Iglesias, A., Lamers, M. and Kohler, G. 1987. Expression of immunoglobulin δ chain causes allelic exclusion in transgenic mice. Nature 330:482. 46 Brink, R., Goodnow, C. C., Crosbie, J., Adams, E., Eris, J., Mason, D. Y., Hartley, S. B. and Basten, A. 1992. Immunoglobulin M and D antigen receptors are both capable of mediating B lymphocyte activation, deletion, or anergy after interaction with specific antigen. J. Exp. Med. 176:991. 47 Rolink, A. and Melchers, F. 1991. Molecular and cellular origins of B lymphocyte diversity. Cell 66:1081. 48 Lassoued, K., Illges, H., Benlagha, K. and Cooper, M. D. 1996. Fate of surrogate light chains in B lineage cells. J. Exp. Med. 183:421. 49 Cyster, J. G., Healy, J. I., Kishihara, K., Mak, T. W., Thomas, M. L., Goodnow, C. C. 1996. Regulation of B-lymphocyte negative and positive selection by tyrosine phosphatase CD45. Nature 381:325.

50 Epstein, M. M., Di Rosa, F., Jankovic, D., Sher, A. and Matzinger, P. 1995. Successful T cell priming in B cell-deficient mice. J. Exp. Med. 182:915. 51 Linette, G. P., Grusby, M. J., Hedrick, S. M., Hansen, T. H., Glimcher, L. H. and Korsmeyer, S. J. 1994. Bcl-2 is upregulated at the CD41CD81 stage during positive selection and promotes thymocyte differentiation at several control points. Immunity 1:197. 52 Strasser, A., Harris, A. W., Von Boehmer, H. and Cory, S. 1994. Positive and negative selection of T cells in T cell receptor transgenic mice expressing a bcl-2 transgene. Proc. Natl Acad. Sci. USA 91:1376. 53 Fairbairn, L. J., Cowling, G. J., Reipert, B. M. and Dexter, T. M. 1993. Suppression of apoptosis allows differentiation and development of a multipotent hemopoietic cell line in the absence of added growth factors. Cell 74:823. 54 Ichihara, Y., Hayashida, H., Miyazawa, S. and Kurosawa, Y. 1989. Only DFL16: DSP2 and DQ52 gene families exist in mouse immunoglobulin heavy chain diversity gene loci, of which DFL16 and DSP2 originate from the same primordial DH gene. Eur. J. Immunol. 19:1849.