c-Myb Coordinates Survival and the Expression of Genes That Are Critical for the Pre-BCR Checkpoint

B cell development, like the development of each hematopoietic lineage, initiates from a multipotent self-renewing hematopoietic stem cell and is defined by the sequential expression of cell surface markers and V(D)J recombination events at the Ig H and L chain loci (1). Hematopoietic stem cells give rise to progenitor cells that gradually lose alternative lineage fate potential and gain B-lineage potential as they differentiate to the CD19⁺ pro-B cell stage, which is the first B-lineage–committed progenitor. During the pro-B cell stage, productive rearrangement at the Igh locus results in expression of an Ig μ-H chain protein, which pairs with the surrogate L chain and signaling components Igα and Igβ to form the pre-BCR. These cells differentiate into the large pre-B cell stage and undergo a limited proliferative burst, exit the cell cycle, differentiate to the small pre-B cell stage, and initiate V(D)J rearrangement at the Igκ L chain locus (2, 3). Upon productive V(D)J rearrangement at one of the Ig L chain loci, L chain protein can pair with the μ-H chain to form membrane IgM and initiate differentiation to the immature B cell stage.

Control of survival during the pro-B cell stage is crucial, because cells must have sufficient time to complete successful V(D)J rearrangements at the H chain locus but not so much time that pro-B cells with failed V(D)J recombination accumulate or for potentially oncogenic chromosome translocations to occur (4). The balance of proapoptotic and anti-apoptotic Bcl-2 family members mediates the intrinsic survival pathway during the pro-B cell stage (5). Oligomerization of the proapoptotic proteins Bak and Bax at the mitochondrial membrane leads to release of cytochrome c and initiation of apoptosis (6). The oligomerization of Bak and Bax is inhibited by interaction with a group of antiapoptotic proteins that includes Bcl-2 and Mcl-1 and gain- and loss-of-function mouse models have demonstrated that these proteins are important for survival at different stages of B cell development (7–9). The antiapoptotic proteins are opposed by a group of proapoptotic BH3-only proteins that includes Bim, Bad, Bid, and Bmf, which act as molecular sensors of cellular stress and interfere with the interaction of Bax and Bak with the antiapoptotic Bcl-2 family members (10). In general, the BH3-only proteins are highly redundant, and only Bim-deficient mice are reported to display a phenotype that is characterized by more pro-B cells (11). Although the Bim-deficient mouse model demonstrates a role for Bim in pro-B cell survival, the absolute number of pro-B cells in these mice is less than that observed in a Bcl-2–transgenic mouse model, suggesting that additional BH3-only proapoptotic proteins contribute to the survival of CD19⁺ pro-B cells.

The IL-7 signaling pathway is the major mediator of survival during the CD19⁺ pro-B cell stage and mediates survival by transcriptional and posttranslational regulation of the proapoptotic and antiapoptotic Bcl-2 family members (12). Signaling through IL-7R activates the Jak/STAT and PI3K/Akt signaling pathways (13, 14). Stat5 mediates survival during the transition from the common lymphoid progenitor stage to the pro-B cell stage by regulating expression of Mcl-1 and is important for proliferation of pro-B cells (8). PI3K/Akt signaling is also crucial for B cell development. However, simultaneous deletion of p110α and p110δ or p85α alone results in a block to B cell development at the large pre-B cell stage and does not result in decreased accumulation or proliferation of pro-B cells (15–17). Thus, the PI3K/Akt signaling pathway appears to be largely important for proliferation of large pre-B cells rather than survival or proliferation of pro-B cells (2, 3, 18). The IL-7 signaling pathway is negatively regulated by the suppressor of cytokine signaling (SOCS) and cytokine-inducible SH2-containing protein (CISH) proteins, which bind to the IL-7Rα, Jak1, and Jak3 proteins downstream of IL-7 signaling, preventing their interaction and targeting Jak1 and Jak3 proteins for proteasomal degradation (12, 19, 20). IL-7R signaling continues to play a crucial role beyond the pro-B cell stage through the pre-BCR checkpoint. The presence of the pre-BCR on the pro-B cell surface lowers the threshold for IL-7 signaling and allows for selective proliferation of pre-BCR–expressing cells in the limiting quantities of IL-7 present in the bone marrow (20). IL-7R and pre-BCR signaling induce expression of cyclin D3 and c-Myc, which are crucial for proliferation during the large pre-B cell stage, after which large pre-B cells cease proliferation to transition to the small pre-B cell stage (21–23).

The Myb (myeloblastosis oncogene) locus encodes the c-Myb transcription factor, which is a DNA-binding protein that functions as both a transcription activator and repressor (24). Myb proto-oncogene protein (c-Myb) is abundantly expressed during the immature stages of hematopoiesis and becomes downregulated as progenitor cells undergo terminal differentiation (25). A crucial role for c-Myb during hematopoiesis was first demonstrated by the embryonic lethality of the c-Myb–null mutation (26). c-Myb–deficient embryos die at approximately embryonic day 14.5 due to an inability to perform adult erythropoiesis in the fetal liver, and the embryonic lethality of the c-Myb–null mutation has been an impediment to understanding the role of c-Myb during lineage-specific differentiation. To circumvent embryonic lethality of Myb-null mutations, we have previously used conditional mutagenesis to demonstrate that c-Myb is absolutely required for B cell development and is important during several stages of B lymphopoiesis (27, 28). Myb^f/f CD19-cre mice, which initiate deletion at the Myb locus during the pro-B cell stage, display a partial block at the pro-B to pre-B cell transition, as well as defects in mature B cell homeostasis (27, 29). Ablation of c-Myb prior to B-lineage commitment in Myb^f/f Mb1-cre mice identified a crucial role for c-Myb in the prepro-B to pro-B cell transition, as well as the survival of pro-B cells (28, 30). In addition, we reported that c-Myb is required for the proper expression of IL-7Rα and Ebf1 during the pro-B cell stage. However, exogenously supplied IL-7Rα or Ebf1 alone was not sufficient to rescue the survival of c-Myb–deficient pro-B cells, suggesting that c-Myb regulates additional genes that mediate survival during the pro-B cell stage (28).

We demonstrate that two proapoptotic Bcl-2 family members, Bmf and Bim, are repressed by c-Myb and control the intrinsic survival (survival in the absence of IL-7) of CD19⁺ pro-B cells. In addition, we demonstrate that levels of Bmf and Bim mRNA are further repressed by IL-7 signaling. Bypassing IL-7 signaling with constitutively active (CA) Stat5b (CA STAT5) suppressed Bmf and Bim mRNA expression and induced expression of Mcl-1 mRNA in c-Myb–deficient pro-B cells, whereas CA AKT only partially suppressed Bmf and did not induce expression of Mcl-1. c-Myb regulates IL-7Rα expression, but forced expression of IL-7Rα on the surface of c-Myb–deficient pro-B cells is not sufficient for repression of Bmf and Bim expression, suggesting that c-Myb controls the expression of other components of the IL-7R signaling pathway. We further demonstrate that c-Myb represses the expression of SOCS3, a negative regulator of IL-7 signaling. Overexpression of SOCS3 in pro-B cells inhibits their ability to accumulate in response to IL-7, suggesting that upregulation of SOCS3 in the absence of c-Myb can impede proper IL-7 signaling in pro-B cells. Overexpression of Bcl-2 is able to rescue the survival of c-Myb–deficient pro-B cells but is not sufficient to rescue transition from the pro-B to pre-B cell stage of differentiation in the absence of c-Myb, and c-Myb–deficient large pre-B cells are hypoproliferative. We demonstrate several proteins, in addition to IL-7Rα, that are crucial during the pro-B to pre-B cell transition, including λ5, cyclin D3 and CXCR4 are downregulated in the absence of c-Myb, and the promoter of Igll1 (λ5) is a direct c-Myb target. Thus, c-Myb coordinates the survival of pro-B cells with the expression of genes that are necessary for proliferation and differentiation into the pre-B cell compartment.

Myb ^f/f, Bmf ^−/−, CD19-cre, and Bcl2Tg mice have been described previously (29, 31–33). Bmf ^−/− mice were a generous gift from Dr. R. Davis (University of Massachusetts Medical School). Rag2^−/− (Taconic Farms), Myb ^f/f CD19-cre Bcl2Tg, Myb ^f/f Rag2^−/− Bcl2Tg, Myb ^f/f Rag2^−/− Bmf ^−/−, and Rag2^−/− Bmf ^−/− mice were bred at the University of Virginia. Mice were 6–10 wk old when used for experiments and were housed in a barrier facility at the University of Virginia. These studies were reviewed and approved by the Institutional Animal Care and Use Committee at the University of Virginia.

Pro-B cells from Rag2^−/− and Rag2^−/− Bmf^−/− mice were positively selected from bone marrow using anti-CD19–labeled magnetic beads (Miltenyi Biotec) and were cultured for 24 h in Opti-MEM supplemented with 15% (v/v) FBS (Life Technologies), 100 U/ml penicillin-streptomycin, 2 mM l-glutamine, 50 μM 2-ME, and 5 ng/ml IL-7 (PeproTech). Cells were then washed, plated in 96-well plates in Opti-MEM supplemented with 15% (v/v) FBS (Life Technologies), 100 U/ml penicillin-streptomycin, 2 mM l-glutamine, and 50 μM 2-ME, and total CD19⁺ cells per well were analyzed by flow cytometry. For transduction of pro-B cells from Myb^f/f Rag2^−/−, Myb^f/f Rag2^−/− Bcl2Tg, and Myb^f/f Rag2^−/− Bmf^−/− mice, cells were positively selected from bone marrow using anti-CD19–labeled magnetic beads (Miltenyi Biotec), cultured for 24 h in Opti-MEM supplemented with 15% (v/v) FBS (Life Technologies), 100 U/ml penicillin-streptomycin, 2 mM l-glutamine, 50 μM 2-ME, and 5 ng/ml IL-7 (PeproTech), and transduced with retroviral vectors, as previously described (28, 34). Following transduction, pro-B cells were cultured in Opti-MEM supplemented with 15% (v/v) FBS (Life Technologies), 100 U/ml penicillin-streptomycin, 2 mM l-glutamine, and 50 μM 2-ME and were analyzed 24, 48, and 72 h later by flow cytometry. The pan-caspase inhibitor Q-VD-OPH (SM Biochemicals) was used at 100 μM.

Single-cell suspensions from bone marrow were prepared from 6–10-wk-old mice, and 2 × 10⁶ cells were stained with optimal amounts of fluorochrome-conjugated Abs, as previously described (27). Cells were subsequently analyzed on a FACSCalibur, FACSCanto (BD Immunocytometry Systems), or Cytoflex (Beckman Coulter Life Sciences). Total cells were determined using AccuCount Blank Particles, 5.27 μm (Spherotech). Flow cytometric data were analyzed using FlowJo software (TreeStar). Cell sorting was performed on a FACSVantage SE Turbo Sorter with DIVA option (BD Immunocytometry Systems). Abs and reagents were purchased from eBioscience (anti-B220 PE-Cy7 [RA3-6B2], anti-CD19 PE [6D5], anti-CD19 PerCP-Cy5.5 [6D5], anti-CD25 PE [PC61.5], anti-CD117 allophycocyanin [2B8], anti-CD127 PE [A7R34], anti-IgM FITC [R6-60.2]), BioLegend (anti-NGFR allophycocyanin [ME20.4]), Molecular Probes (7-aminoactinomycin D), and Sigma-Aldrich (DAPI). For DRAQ5 staining, cells were stained for expression of surface markers, followed by incubation with 50 μM DRAQ5 (eBioscience) for 20 min at 37°C prior to analysis.

The retrovirus vectors pMIG-R1, pMSCV-IRES-tNGFR, ptNGFR-Cre, pMIG-c-Myb, pMIG-Bcl2, pMIG-Cre, pMIG-CA-STAT5B, pMIG-CA-AKT, and pMIG-IL-7Rα have been described previously (28, 35–40). pMIG-Bcl2 was provided by Dr. M. Kondo (Duke University Medical Center). pMIG-CA-STAT5B and pMIG-CA-AKT were provided by Dr. M. Clark (University of Chicago). To generate pMIG-Bmf, Bmf encoding cDNA was isolated from pBabe-3XFLAG-mouse Bmf (41) and cloned into the BglII/EcoRI site of pMIG-R1. pBabe-3XFLAG-mouse Bmf was provided Dr. J. Brugge (plasmid #17243; Addgene). To generate pMIG-BimEL, BimEL cDNA was isolated from pCMV-Tag2b-Flag-BimEL (42) and cloned into the BglII/EcoRI site of pMIG-R1. pCMV-Tag2b-Flag-BimEL was provided by Dr. R. David (plasmid #23090; Addgene). To generate pMIG-SOCS3, SOCS3 cDNA was isolated from pCMV-SOCS3 (43) and cloned into the EcoRI site of pMIG-R1. pCMV-SOCS3 was provided by Dr. R. Kahn (plasmid #11486; Addgene). To generate pSUPER-Puro-IRES-GFP-shLuc and pSUPER-Puro-IRES-GFP-shBim, IRES-GFP from pMIG-R1 was cloned into the NsiI site of pSUPER-shLuc and pSUPER-shBim (44). pSUPER-shLuc and pSUPER-shBim were gifts of Dr. E. Cheng (Memorial Sloan-Kettering Cancer Center). Retrovirus supernatants were generated by transient transfection of HEK-293T cells and titered on NIH-3T3 cells by flow cytometry, as previously described (28).

Retrovirus-transduced CD19⁺ pro-B cells were electronically sorted based on the expression of NGFR and GFP, and total cellular RNA was isolated using TRIzol Reagent (Invitrogen), according to the manufacturer’s protocol. Contaminating genomic DNA was removed by treatment with RNase-free DNase I (Invitrogen), and cDNA was prepared with a Superscript III First-Strand Synthesis System (Invitrogen). Quantitative RT-PCR (qRT-PCR) was performed on cDNA with Titanium Taq Polymerase (BD Clontech) with 1× SYBR Green (Molecular Probes) and 0.4 μM the primer set of interest in 25-μl reaction mixtures in a MyiQ Single-Color Real-Time PCR Detection System (Bio-Rad). Conditions for qRT-PCR were as follows: 95°C for 3 min, 40 cycles of 95°C for 40 s, 66°C for 20 s and 72°C for 30 s, followed by an extension at 72°C for 1 min. Melting curve analysis was performed to ensure equivalent and appropriate melting temperatures. Each sample was normalized to the expression of Hprt. Primers used are listed in Supplemental Table I.

Pro-B cells were harvested from cell culture and lysed in 20 mM Tris (pH 7.4), 100 mM NaCl, 10 mM EDTA, 1 mM EGTA, 1% Triton X-100 (45) containing EDTA-free protease inhibitor mixture (Roche), and 1 mM PMSF (Sigma-Aldrich). Ten micrograms of protein was loaded on 15% SDS–polyacrylamide gels and transferred to Protran nitrocellulose transfer membranes (Whatman). Membranes were blocked in PBS plus 0.05% Tween-20 (PBS-T) with 5% nonfat dry milk for 1 h and then incubated with the appropriate primary Ab at 4°C overnight. Membranes were washed three times with PBS-T and probed with anti-rat HRP- or anti-rabbit HRP-conjugated Abs (GE Healthcare Bioscience) in PBS-T for 1 h at room temperature. After washing the membranes three times with PBS-T, the proteins were detected by ECL (Amersham). Protein expression was quantified using ImageQuant TL 2005 software. The following primary Abs were used for Western blotting: anti–c-Myb (EP769Y; Epitomics), anti-Bim (2819; Cell Signaling Technology), anti-Bmf (17A9; Enzo Life Sciences), and anti–β-actin (AC-15; Sigma-Aldrich).

CD19⁺ pro-B cells from Rag2^−/− mice were harvested and cultured for 72 h in Opti-MEM supplemented with 15% (v/v) FBS (Life Technologies), 100 U/ml penicillin-streptomycin, 2 mM l-glutamine, 50 μM 2-ME, and 5 ng/ml IL-7 (PeproTech). Protein was cross-linked to chromatin by adding 1% formaldehyde to each culture dish at room temperature for 10 min. The reaction was stopped by addition of 125 mM glycine, cells were incubated at room temperature for 5 min while rocking. Cells were harvested, pelleted, and washed with cold PBS. Cells were resuspended at 10⁷ cells per milliliter in cold cytoplasmic lysis buffer (20 mM Tris-HCl [pH 8], 85 mM KCl, 0.5% Nonidet P-40, 1 mM PMSF, and EDTA-free protease inhibitor mixture; Roche) and incubated on ice for 10 min. Nuclei were centrifuged, resuspended at 10⁷ cells per milliliter in cold sonication buffer (10 mM Tris-HCl [pH 8], 0.1 mM EDTA, 1% Nonidet P-40, 0.01% SDS, 1 mM PMSF, and EDTA-free protease inhibitor mixture), and sonicated using a W-375 cell disruptor (Ultrasonics) to generate a chromatin fragment. Debris was cleared by centrifugation, and chromatin was supplemented with 5% glycerol and 127 mM NaCl. Chromatin was precleared with protein A/G PLUS-Agarose (Santa Cruz Biotechnology) for 1 h and immunoprecipitated overnight with 5 μg of anti–c-Myb clone 1-1 (Millipore), anti–c-Myb clone EP769Y (Epitomics), normal mouse IgG (Santa Cruz Biotechnology), or normal rabbit IgG (Santa Cruz Biotechnology), with rotation, at 4°C. Immune complexes were collected on protein A/G PLUS-Agarose for 1 h, with rotation, at 4°C. Bound immune complexes were washed for 3 min on ice with low-salt wash buffer (10 mM Tris-HCl [pH 8], 2 mM EDTA, 150 mM NaCl, 0.1% SDS, 1% Triton X-100), high-salt wash buffer (10 mM Tris-HCl [pH 8], 2 mM EDTA, 500 mM NaCl, 0.1% SDS, 1% Triton X-100), LiCl wash buffer (10 mM Tris-HCl [pH 8], 1 mM EDTA, 250 mM LiCl, 1% sodium deoxycholate, 1% Nonidet P-40), and TE wash buffer (10 mM Tris-HCl [pH 8], 1 mM EDTA). All wash buffers were supplemented with PMSF and EDTA-free protease inhibitor mixture. Bound immune complexes were eluted from agarose in elution buffer (0.1 M NaHCO₃, 1% SDS) for 30 min, with rotation, at room temperature. Formaldehyde cross-linking was reversed in the presence of 200 mM NaCl at 65°C overnight. Chromatin was treated with RNase A and proteinase K, and DNA was isolated by phenol/chloroform extraction and analyzed by qRT-PCR. Primers are listed in Supplemental Table I.

Differences between data sets were analyzed using the two-tailed Student t test and at a confidence level of 95% for all experiments; error bars represent SEM. Data sets were analyzed and figures were prepared with Prism v5.01 and v7.0 (GraphPad Software).

We previously reported that c-Myb is absolutely required for B cell development and the survival of CD19⁺ pro-B cells (28). During the initial analysis of c-Myb–deficient pro-B cells, we did not detect changes in the expression of Bcl-2, Mcl-1, or Bim mRNA, which are known to be involved in the control of c-Myb–deficient pro-B cell survival (28). However, additional Bcl-2 family members, such as Bak, Bax, Bid, Bad, and Bmf, are expressed during B cell development (4). To examine the expression of these Bcl-2 family members in the absence of c-Myb, Myb^f/f Rag2^−/− pro-B cells were cultured in the presence of IL-7 for 24 h, transduced with a retrovirus that encodes Cre and GFP (MIG-Cre) to inactivate the Myb locus, and cultured in the absence of IL-7 for 24 h (Fig. 1A). Deletion efficiency by MIG-Cre is >95% in these cultures (Supplemental Fig. 1A). GFP⁺ cells were electronically sorted 24 h posttransduction, and expression of mRNAs encoding Bcl-2 family members was quantified by qRT-PCR (Fig. 1B). As previously reported (28), we did not detect decreased expression of Bcl-2, Bcl-x_L, Mcl-1, or Bim mRNA. However, the amount of steady-state mRNA encoding the proapoptotic BH3-only family member Bmf was increased 5-fold in the absence of c-Myb, suggesting that Bmf expression is repressed by c-Myb in CD19⁺ pro-B cells.

Lymphocyte progenitor cells very rapidly undergo apoptotic cell death in the absence of c-Myb (28, 46), and we have found that it is often necessary to rescue survival of c-Myb–deficient lymphocyte progenitor cells to accurately determine changes in gene expression, as well as to identify potential c-Myb–mediated activities, in addition to maintaining survival (46). To rescue survival in c-Myb–deficient pro-B cells, Myb^f/f Rag2^−/− pro-B cells were cotransduced with a retrovirus that encodes Cre recombinase and a truncated nerve growth factor receptor that serves as a marker of transduction (tNGFR-Cre), as well as a retrovirus that encodes Bcl-2 and GFP (MIG-Bcl2). Forced expression of Bcl-2 was able to rescue the survival of c-Myb–deficient pro-B cells (Fig. 1C), and comparable results were obtained using pro-B cells isolated from Myb^f/f Rag2^−/− Bcl2Tg mice, in which Bcl-2 transgene expression is driven by an SV40 promoter and Eμ IgH enhancer that is constitutively expressed in B- and T-lineage progenitor cells (47) (Supplemental Fig. 1B). Survival of c-Myb–deficient Rag2^−/− pro-B cells could also be rescued using the pharmacological caspase inhibitor Q-VD-OPH (Fig. 1D). To determine whether additional differences in the expression of mRNA encoding Bcl-2 family members could be detected in c-Myb–deficient Rag2^−/− pro-B cells when the underlying survival defect was overcome, Myb^f/f Rag2^−/− Bcl2Tg CD19⁺ pro-B cells were transduced with MIG-R1 or MIG-Cre and cultured for 48 h in the absence of IL-7. GFP⁺ cells were electronically sorted, and the expression of mRNA encoding Bcl-2 family members was determined by qRT-PCR (Fig. 1E). Similar to c-Myb–deficient pro-B cells that lack a Bcl-2 transgene, the expression of Bmf mRNA was increased ∼5-fold in the absence of c-Myb. Furthermore, we also detected a 5-fold increase in the amount of Bim(EL) mRNA in the absence of c-Myb. Comparable results were obtained in c-Myb–deficient pro-B cells rescued by Q-VD-OPH (data not shown). Thus, the steady-state level of mRNAs that encode the proapoptotic Bcl-2 family proteins Bmf and Bim is repressed by c-Myb in CD19⁺ pro-B cells, demonstrating that c-Myb is important for setting the baseline levels of Bmf and Bim in pro-B cells.

Rag2^−/− CD19⁺ pro-B cells rapidly died after transduction with retroviruses encoding Bmf or Bim, demonstrating that either proapoptotic protein can induce apoptosis when overexpressed in pro-B cells (Supplemental Fig. 2A). To determine whether the increased expression of Bmf contributes to poor survival in c-Myb–deficient pro-B cells, we bred Myb^f/f Rag2^−/− Bmf^−/− mice. The Rag2 and Bmf loci are 17 Mb apart on chromosome 2 (31), and a cross-over event was required to produce Myb^f/f Rag2^−/− Bmf^−/− mice. CD19⁺ pro-B cells from Myb^f/f Rag2^−/− Bmf^−/− mice were transduced with MIG-Cre and cultured for 72 h in the absence of IL-7. The Bmf loss-of-function mutation provided a 3-fold increase in the survival of c-Myb–deficient CD19⁺ pro-B cells in culture 48 h posttransduction and a 10-fold increase in survival 72 h posttransduction with MIG-Cre compared with Myb^f/f Rag2^−/− pro-B cells (Fig. 2A). Thus, Bmf contributes to the intrinsic survival of pro-B cells.

The Bim locus is also located on chromosome 2, 9 Mb away from the Bmf locus, 26 Mb away from the Rag2 locus and a cross over event would have been required to produce Myb^f/f Rag2^−/− Bim^−/− mice. As an alternative, we used a short hairpin RNA (shRNA)-mediated knockdown approach to determine whether increased Bim expression in c-Myb–deficient pro-B cells contributes to poor survival (44). Myb^f/f Rag2^−/− CD19⁺ pro-B cells were transduced with a retrovirus encoding an shRNA against Bim and cultured for 72 h in the presence of IL-7 to allow for knockdown of Bim expression. Subsequently, Myb^f/f Rag2^−/− pro-B cells were transduced with tNGFR-Cre to inactivate the Myb locus and cultured for 72 h in the absence of IL-7. Knockdown of Bim provided a 1.5-fold increase in the survival of c-Myb–deficient pro-B cells 48 h posttransduction and a 3-fold rescue 72 h posttransduction compared with Myb^f/f Rag2^−/− pro-B cells cotransduced with tNGFR-Cre and shRNA against Luc (Fig. 2B). Bim may contribute more to the intrinsic survival of CD19⁺ pro-B cells than is apparent from these results due to the poor knockdown efficiency of Bim that was achieved in c-Myb–deficient pro-B cells (see 17Discussion). To determine whether there was an additive effect of Bmf and Bim in c-Myb–regulated pro-B cell survival, Myb^f/f Rag2^−/− Bmf^−/− pro-B cells were transduced with the retrovirus encoding an shRNA against Bim and cultured for 72 h in the presence of IL-7. Myb^f/f Rag2^−/− Bmf^−/− pro-B cells were subsequently transduced with tNGFR-Cre and cultured for an additional 72 h in the absence of exogenous IL-7. The rescue of survival provided by knockdown of Bim in c-Myb–deficient Rag2^−/− Bmf^−/− pro-B cells was greater than that provided by deficiency of Bim or Bmf alone (Fig. 2C), demonstrating that suppression of Bmf and Bim expression by c-Myb is important for the intrinsic survival (survival in the absence of IL-7) of CD19⁺ pro-B cells.

The IL-7 signaling pathway is the major survival pathway in CD19⁺ pro-B cells and regulates the expression of Bcl-2 family members at the level of transcription and posttranslational modification (12). Bim is expressed at each stage of B cell development (48), and Bmf expression has been reported in pre-B, immature B, and mature B cells (49). However, we also detect Bmf protein in freshly isolated Rag2^−/− CD19⁺ pro-B cells (Supplemental Fig. 2B). Suppression of Bim expression is important for the maintenance of the pro-B cell compartment downstream of IL-7 (11, 48). In contrast, little is known about the function of Bmf during B cell development, although it plays a role in the survival of large pre-B cells (50). To determine whether suppression of Bmf expression is important for the maintenance of pro-B cell survival downstream of IL-7, Rag2^−/− and Rag2^−/− Bmf^−/− pro-B cells were cultured for 24 h in the presence of IL-7, followed by withdrawal of IL-7 for 24–72 h. Throughout the time course of the experiment, a greater proportion of Rag2^−/− Bmf^−/− pro-B cells survived in the absence of IL-7 compared with Rag2^−/− pro-B cells (Fig. 3A), demonstrating that suppression of Bmf expression, like Bim expression, is an important regulator of pro-B cell survival downstream of IL-7.

To determine whether expression of Bmf and Bim mRNA is regulated downstream of IL-7 signaling in pro-B cells, Rag2^−/− pro-B cells were cultured in the presence or absence of IL-7 for 24 and 48 h, and expression of Bmf and Bim mRNA was determined by qRT-PCR. The amount of Bmf mRNA detected in Rag2^−/− CD19⁺ pro-B cells increased 20-fold after 24 h and 40-fold after 48 h in culture in the absence of IL-7 compared with Rag2^−/− pro-B cells cultured in the presence of IL-7 (Fig. 3B). In addition, the amount of Bim(EL) mRNA detected in c-Myb–deficient pro-B cells increased 7-fold after 24 h and 10-fold after 48 h in culture in the absence of IL-7, whereas Bim(L) mRNA was upregulated 3-fold after 48 h in culture in the absence of IL-7 compared with Rag2^−/− CD19⁺ pro-B cells cultured in the presence of IL-7.

Pro-B cells die quickly in culture in the absence of IL-7, and we were unable to measure Bmf and Bim protein levels in pro-B cells by Western blot after removal of IL-7 from the growth medium (data not shown). To circumvent this problem, we examined Bmf and Bim mRNA and protein expression in Rag2^−/− Bcl2Tg pro-B cells. Similar to Rag2^−/− pro-B cells, the amount of Bmf, Bim(EL), and Bim(L) mRNA increased in Rag2^−/− Bcl2Tg CD19⁺ pro-B cells cultured in the absence of IL-7 compared with Rag2^−/− Bcl2Tg pro-B cells cultured in the presence of IL-7 (Fig. 3C). The amount of Bmf protein increased 3-fold and Bim protein increased 5-fold after 48 h in culture in the absence of IL-7 compared with Rag2^−/− Bcl2Tg pro-B cells cultured in the presence of IL-7 (Fig. 3D). It is important to note that IL-7 signaling does not regulate c-Myb expression in pro-B cells. c-Myb mRNA is not regulated downstream of IL-7 signaling in B-lineage progenitors (51). Similarly, c-Myb protein expression was not altered in Rag2^−/− Bcl2Tg pro-B cells cultured for 48 h in the presence or absence of IL-7 (Supplemental Fig. 2C). Thus, IL-7 signaling represses the expression of Bmf and Bim mRNA and protein in CD19⁺ pro-B cells.

Signaling through IL-7R activates the Jak/STAT and the PI3K/Akt signaling pathways in pro-B cells. These pathways regulate survival and proliferation during the pro-B cell stage, as well as inhibit differentiation into the pre-B cell compartment (2, 12). To determine whether either of these pathways contributes to the regulation of Bmf and Bim mRNA expression, Rag2^−/− pro-B cells were transduced with retroviruses encoding a CA Stat5b protein and GFP (MIG-CA-STAT5) or a CA AKT protein and GFP (MIG-CA-AKT) (40) and were cultured in the absence of IL-7 for 48 h. Transduced pro-B cells were isolated by electronic cell sorting based on GFP expression, and Bmf and Bim(EL) mRNA expression was determined by qRT-PCR. The amount of Bmf mRNA was decreased 50% by CA STAT5 and 87% by CA AKT compared with MIG-R1–transduced pro-B cells. In contrast, expression of Bim(EL) mRNA was decreased 70% by CA STAT5, but CA AKT did not appear to alter expression of Bim(EL) mRNA in CD19⁺ pro-B cells (Fig. 3E), consistent with the finding that Foxo1 represses the expression of Bim (34). Similarly, we determined whether CA STAT5 or CA AKT could suppress expression of Bmf and Bim mRNA in c-Myb–deficient pro-B cells. For this purpose, we cotransduced Myb^f/f Rag2^−/− pro-B cells with retroviruses encoding NGFR-Cre and MIG- CA-STAT5 or NGFR-Cre and MIG-CA-AKT (Fig. 3F). As we observed upon IL-7 withdrawal in c-Myb–sufficient Rag2^−/− pro-B cells, repression of Bmf mRNA was restored by CA STAT5 and CA AKT, whereas repression of Bim mRNA was restored by CA STAT5, but not CA AKT, in c-Myb–deficient Rag2^−/− pro-B cells. In addition, IL-7 signaling through Stat5 is the major regulator of Mcl-1 expression in B-lineage progenitor cells (8) and, as expected, we found that Mcl-1 mRNA was decreased in Myb^f/f Rag2^−/− pro-B cells 48 h after transduction with NGFR-Cre and that CA STAT5, but not CA AKT, was able to restore expression of Mcl-1 mRNA after cotransduction with NGFR-Cre (Fig. 3F). Thus, steady-state levels of Bmf and Bim are suppressed by IL-7. Furthermore, CA STAT5 is able to suppress expression of Bmf and Bim, as well as induce expression of Mcl-1 mRNA. In contrast, CA AKT can only suppress Bmf mRNA and does not induce expression of Mcl-1.

c-Myb regulates the expression of IL-7Rα during the pro-B cell stage, but forced expression of IL-7Rα is not sufficient to rescue survival in c-Myb–deficient pro-B cells in the presence or absence of IL-7 (28, 30). Thus, we sought to determine whether c-Myb–deficient pro-B cells provided with an exogenous source of IL-7Rα could repress expression of Bmf and Bim mRNA when cultured in the presence of IL-7. Myb^f/f Rag2^−/− Bcl2Tg pro-B cells were transduced with MIG-IL-7Rα and cultured for an additional 24 h in the presence of IL-7 to allow for expression of IL-7Rα. These pro-B cells were subsequently transduced with tNGFR-Cre and cultured for an additional 48 h in the presence of IL-7, after which truncated nerve growth factor receptor (tNGFR)⁺ GFP⁺ cotransduced pro-B cells were electronically sorted, and the expression of Bmf and Bim(EL) mRNA was measured by qRT-PCR. Despite surface expression of IL-7Rα and the presence of IL-7 in the culture medium, c-Myb–deficient pro-B cells still failed to repress expression of Bmf and Bim(EL) (Fig. 4A), suggesting that c-Myb controls the expression of additional components of the IL-7 signaling pathway. The transcription factors STAT5 and Foxo1 are regulated by IL-7 signaling in pro-B cells and are crucial for survival in the pro-B cell compartment (8, 34). However, the expression of STAT5a, STAT5b, and Foxo1 mRNA was not significantly altered in the absence of c-Myb (Fig. 4B), suggesting that c-Myb does not regulate expression of STAT5 or Foxo1 in pro-B cells and, instead, regulates other components of the IL-7 signaling pathway.

The IL-7 signaling pathway is negatively regulated by SOCS family proteins, which act as E3 ubiquitin ligases and prevent the interaction between IL-7R and Jak and STAT proteins, causing Jak proteins to be targeted for degradation (19). The expression of mRNA encoding SOCS family members SOCS1 and SOCS3, as well as CISH, is induced by IL-7 signaling in pro-B and large pre-B cells and inhibits IL-7 signaling during the small pre-B cell stage, which allows for initiation of recombination at the κ L chain locus (12, 20). The amount of SOCS1 and CISH mRNA in CD19⁺ pro-B cells modestly increased in the absence of c-Myb, but the increase did not reach statistical significance. However, the amount of SOCS3 mRNA increased 4-fold in the absence of c-Myb (Fig. 4C), suggesting that c-Myb represses expression of SOCS3. To determine whether increased expression of SOCS3 could inhibit the pro-B cell response to IL-7, we transduced Rag2^−/− pro-B cells with a SOCS3-expressing retrovirus (MIG-SOCS3) and cultured these cells for 72 h in the presence of IL-7. The number of Rag2^−/− pro-B cells transduced with MIG-R1 increased 4-fold 48 h posttransduction and 7-fold 72 h posttransduction (Fig. 4D). In contrast, Rag2^−/− pro-B cells transduced with MIG-SOCS3 did not increase in number over the time course of the experiment (Fig. 4D), demonstrating that overexpression of SOCS3 during the pro-B cell stage can inhibit the pro-B cell response to IL-7. Thus, c-Myb controls the expression of two critical components of the IL-7 signaling pathway: it regulates the expression of the IL-7Rα-chain, which pairs with the γ_c chain to form IL-7R, and it represses the expression of SOCS3, a negative regulator of IL-7 signaling.

Because circumventing IL-7 signaling using CA STAT5 and CA AKT restored repression of Bim and Bmf mRNA and resulted in increased expression of Mcl-1 mRNA, we determined whether this was sufficient to rescue survival of c-Myb–deficient Rag2^−/− pro-B cells. For this purpose, Myb^f/f Rag2^−/− pro-B cells were isolated and cotransduced with retroviruses that produce NGFR-Cre and MIG-CA-STAT5 or NGFR-Cre and MIG-CA-AKT, followed by culture in the absence of IL-7. CA STAT5 was able to increase the relative cell recovery of c-Myb–deficient pro-B cells 6-fold after the loss of c-Myb, whereas constitutive activation of Akt provided an ∼3-fold increase in the relative recovery of c-Myb–deficient pro-B cells (Fig. 5A). The decreased amount of rescue by CA AKT compared with CA STAT5 is consistent with the finding that CA AKT repressed expression of Bmf RNA, but not Bim mRNA, and did not result in increased expression of Mcl-1 mRNA in c-Myb–deficient pro-B cells (Fig. 3F). However, neither CA STAT5 nor CA AKT completely rescued the relative recovery of c-Myb–deficient pro-B cells. Because Rag2^−/− pro-B cells proliferate in response to IL-7, it was also possible that complete rescue was impeded by a lack of proliferation in c-Myb–deficient pro-B cells. To determine whether c-Myb is important for the proliferation of pro-B cells in addition to survival, we cotransduced Myb^f/f Rag2^−/− pro-B cells with NGFR-Cre and MIG-CA-STAT5 or NGFR-Cre and MIG-CA-AKT, as described above, cultured them in the absence of IL-7, and stained them with DRAQ5 48 h later to measure DNA content by flow cytometry. In c-Myb–sufficient RAG2^−/− pro-B cells, CA STAT5 resulted in 17.3% of pro-B cells with >2N DNA content after withdrawal of IL-7, whereas CA AKT had little ability to stimulate proliferation in the absence of IL-7 (Fig. 5B). In contrast, CA STAT5 induced a much smaller proportion of cells (2.9%) with a >2N DNA content in c-Myb–deficient Rag2^−/− pro-B cells, a >80% decrease in the proportion of cells with a >2N DNA content detected in c-Myb–sufficient Rag2^−/− pro-B cells. Thus c-Myb is important for the survival and proliferation of pro-B cells in response to IL-7.

To determine whether c-Myb is important for the proliferative expansion of large pre-B cells during the pro-B to pre-B cell transition, in addition to regulating survival, we crossed a Bcl-2 transgene onto the Myb^f/f CD19-cre background and analyzed the pro-B, large pre-B, and small pre-B cell compartments from Myb^f/f CD19-cre Bcl2Tg mice and control mice using flow cytometry. Myb^f/f CD19-cre Bcl2Tg mice exhibited an increase in the proportion of pro-B cells and a decrease in the proportion of pre-B cells in bone marrow compared with Myb^f/f Bcl2Tg control mice (Fig. 6A), but the absolute number of pro-B cells in Myb^f/f CD19-cre Bcl2Tg mice was equivalent to that in control mice (Fig. 6B). However, the absolute numbers of large and small pre-B cells detected in Myb^f/f CD19-cre Bcl2Tg mice were reduced 40 and 50%, respectively, compared with the numbers detected in Myb^f/f Bcl2Tg control mice. These results suggest that, in addition to regulating survival, c-Myb controls proliferation and/or differentiation during the pre-BCR checkpoint. To determine whether c-Myb is important for the proliferation of large pre-B cells, we examined DNA content in freshly isolated large pre-B cells by DRAQ5 staining directly ex vivo. Although ∼50% of large pre-B cells from Myb^f/f and Myb^f/f Bcl2Tg mice had a >2N DNA content, <20% of large pre-B cells from Myb^f/f CD19-cre mice and <5% of large pre-B cells from Myb^f/f CD19-cre Bcl2Tg mice had >2N DNA content (Fig. 6C). Thus, c-Myb is important for proliferation during the large pre-B cell stage.

The pre-BCR checkpoint requires that pro-B cells express components of signaling pathways that mediate selection into the pre-B cell compartment and allow for proliferative expansion of large pre-B cells (52). Myb^f/f Rag2^−/− Bcl2Tg pro-B cells were transduced with MIG-R1 or MIG-Cre, GFP⁺ cells were electronically sorted 48 h posttransduction, and we used qRT-PCR to analyze the expression of mRNAs that encode proteins known to play critical roles during the pre-BCR checkpoint, including the proliferation factor cyclin D3, the cytokine/chemokine receptors IL-7Rα and CXCR4, and components of the pre-BCR (λ5, VpreB, mb1, B29) (Fig. 7A). Of these genes, mRNAs encoding cyclin D3, IL-7Rα, CXCR4, and λ5 were downregulated in the absence of c-Myb, demonstrating that c-Myb is important for the proper expression of key molecules that are required for proliferation and differentiation during the pre-BCR checkpoint.

c-Myb functions as a transcriptional activator and repressor by directly binding to chromatin (53–55). Using chromatin immunoprecipitation (ChIP), c-Myb has been reported to directly bind to the Cxcr4 promoter in MCF-7 human breast carcinoma cells and the Bcl2l11 (Bim) promoter in PC12 neuronal cells (56, 57). The Igll1 (λ5) promoter has been suggested to be a c-Myb target in murine pre-B cell lines based on EMSAs and luciferase reporter assays, but direct binding of c-Myb to the Igll1 promoter in vivo was not reported (58). In addition, the promoters of Il7ra and Ccnd3, as well as intron 2 of the Bmf gene, contain potential c-Myb binding sites, although direct binding of c-Myb to these sites has not been detected (28, 30, 59, 60). To determine whether putative c-Myb binding sites in the Bcl2l11, Bmf, Ccnd3, Igll1, Il7ra, or Cxcr4 genes are direct c-Myb target genes during the pro-B cell stage, we performed ChIP assays using chromatin from Rag2^−/− CD19⁺ pro-B cells cultured for 72 h in the presence of IL-7. Abs directed against the N-terminal end (EP769Y) and C-terminal end (1-1) of c-Myb were used for these experiments, and immunoprecipitated chromatin was analyzed by qRT-PCR for enrichment of c-Myb binding at the potential c-Myb binding sites in each gene. Of the promoters tested, enrichment for c-Myb binding was only detected on the Igll1 promoter compared with normal mouse/rabbit IgG and the negative control Ccng2 (cyclin G2) promoter (Fig. 7B). Thus, c-Myb controls the expression of critical signaling pathway components that are important for the pre-BCR checkpoint, and at least one of these genes, Igll1, is a direct c-Myb target. Taken together, our results demonstrate that c-Myb coordinates survival with the expression of genes that are important to drive proliferation and differentiation across the pre-BCR checkpoint.

The c-Myb transcription factor is crucial for the regulation of survival, proliferation, and differentiation of hematopoietic progenitor cells (24, 25). We previously reported an absolute requirement for c-Myb during B cell differentiation and the control of survival in CD19⁺ pro-B cells during B cell development (28), but downstream mechanisms and factors that mediate c-Myb–regulated survival are poorly understood. We have now demonstrated that c-Myb regulates the intrinsic survival (survival in the absence of IL-7) of pro-B cells by repression of the proapoptotic proteins Bmf and Bim. Thus, c-Myb is crucial to set the basal level of Bmf and Bim expression in pro-B cells. IL-7 signaling via Stat5 is thought to be the major determinant of pro-B cell survival, at least in part, by controlling expression of the antiapoptotic factor Mcl-1 (8). We have now demonstrated that IL-7 signaling further represses the expression of Bmf and Bim in pro-B cells largely via Stat5 in c-Myb–sufficient and -deficient pro-B cells, although CA AKT partially suppressed expression of Bmf. The lack of suppression of Bim mRNA and modest decrease in Bmf mRNA expression mediated by CA AKT are consistent with reports that loss-of-function mutations in components of PI3K, as well as Akt1/2, have little effect on pro-B cell development (15–17). Thus, in addition to increasing the expression of Mcl-1 (8), IL-7 signaling acts on the baseline level of Bmf and Bim expression that is determined by c-Myb in pro-B cells, creating a balance of pro- and antiapoptotic factors that favors survival.

We note that the rescue of intrinsic survival in c-Myb–deficient pro-B cells by the loss of Bmf and Bim was not to the level of control in Rag2^−/− pro-B cells, which could be a consequence of incomplete knockdown of Bim (data not shown). We were able to achieve essentially complete knockdown of Bim expression in several Abelson virus–transformed pro-B and pre-B cell lines, but we only achieved very inefficient knockdown of Bim protein expression in c-Myb–deficient CD19⁺ pro-B cells, likely due to the large amount of Bim encoding mRNA and protein that accumulates in pro-B cells in the absence of c-Myb. It remains possible that additional Bcl2 family members are involved in controlling the intrinsic survival of pro-B cells. We detected a 50% decrease in the expression of Bcl-x_L in the absence of c-Myb, and Bcl-x_L has been described as a mediator of c-Myb–regulated survival during the double-positive stage of thymopoiesis (46). However, lymphoid-specific deletion of Bcl2l1 suggests that it is important for the survival of pre-B cells but dispensable prior to the pre-B cell stage (8). Bim has previously been described as a direct c-Myb target in neuronal cells during nerve growth factor withdrawal (57). In neuronal cells, c-Myb is reported to activate expression of Bim, as opposed to the repression of Bim expression that we observe in pro-B cells, suggesting that c-Myb regulates Bim expression in a cell- and stage-specific manner. In addition to the c-Myb binding site identified within the Bcl2l11 promoter in neuronal cells, analysis of anti-Myb ChIP sequencing data performed in the MCF-7 human breast adenocarcinoma cell line revealed a potential c-Myb binding site within intron 2 of the Bmf locus (56). However, using ChIP, we did not detect direct binding of c-Myb to the Bcl2l11 promoter or the potential c-Myb site in Bmf intron 2 in Rag2^−/− CD19⁺ pro-B cells (data not shown). It remains possible that c-Myb suppresses Bim and Bmf expression during the pro-B cell stage directly through unidentified regulatory regions, by association with other proteins that tether c-Myb to the Bmf or Bcl2l1 regulatory regions, or through an indirect mechanism.

IL-7 signaling provides the major survival signals in pro-B cells (2, 12) and, in addition to controlling the basal level of Bmf and Bim in the absence of IL-7, we have demonstrated that c-Myb controls the expression of two key components of the IL-7 signaling pathway. c-Myb is crucial for expression of the IL-7Ra chain (25, 28). However, we have found that, in the absence of c-Myb, exogenously supplied IL-7Ra is not sufficient to repress expression of Bmf and Bim and does not rescue survival. Thus, we examined the expression of downstream components of the IL-7 signaling pathway and further demonstrate that c-Myb represses expression of SOCS3, a negative regulator of IL-7 signaling (12). The SOCS family of proteins inhibit signaling downstream of cytokine receptors by binding to the Jak tyrosine kinases and targeting them for proteosomal degradation and, in some cases, interfering with the interaction of Jak kinases with Stat proteins (19). The expression of SOCS1, SOCS3, and CISH mRNA is induced downstream of IL-7 signaling in pro-B and large pre-B cells (12, 20, 61), and upregulation of SOCS and CISH family members inhibits IL-7 signaling during the small pre-B cell stage, which may allow for the initiation of recombination at the Ig κ L chain loci by relieving Stat5-mediated repression (8, 12, 62). We detected increased expression of SOCS3 mRNA in c-Myb–deficient pro-B cells and demonstrate that overexpression of SOCS3 during the pro-B cell stage was able to inhibit accumulation of pro-B cells in response to IL-7. Thus, c-Myb regulates IL-7 signaling in CD19⁺ pro-B cells through at least two mechanisms: it is required for expression of the IL-7Rα-chain, which associates with the γ_c chain to form the functional IL-7R, and it represses SOCS3, which functions to inhibit signaling through IL-7R. Therefore, in addition to regulating the baseline levels of Bmf and Bim in pro-B cells, c-Myb is required for proper expression of the IL-7Rα component of IL-7R and the negative regulator of IL-7R signaling, SOCS3, which provides a mechanism to further modulate the expression of Bmf and Bim and control the lifespan of CD19⁺ pro-B cells (Supplemental Fig. 3).

In addition to restoring repression of Bim and Bmf, we found that forced activation of IL-7 signaling pathways through Stat5 (CA STAT5) and, to a lesser extent, Akt (CA AKT), led to significantly increased survival of c-Myb–deficient pro-B cells. However, the relative cell recovery of c-Myb–deficient pro-B cells was not completely rescued by exogenous CA STAT5 or CA AKT expression in c-Myb–deficient pro-B cells following IL-7 withdrawal, suggesting that processes downstream of IL-7 signaling other than survival were impacted by the loss of c-Myb. Signals transmitted through IL-7R are important for the proliferation of pro-B and large pre-B cells (2, 12), and we found that CA STAT5, despite inducing expression of cyclin D3 in c-Myb–deficient Rag2^−/− pro-B cells, and CA AKT are not able to rescue proliferation in c-Myb–deficient pro-B cells. The inability of CA AKT to drive proliferation in pro-B cells is consistent with reports demonstrating that activation of PI3K is not important for the proliferation of pro-B cells (15–17, 63). We conclude that control of pro-B cell survival by c-Myb is primarily mediated by IL-7–driven activation of Stat5. c-Myb controls the expression of cyclin D3, which is crucial for the proliferation of pro-B cells, via IL-7–mediated activation of Stat5. However, the control of pro-B cell proliferation by c-Myb is also mediated by genes that are regulated by c-Myb, independently of IL-7 signaling via Stat5 or PI3K/Akt.

We previously identified a critical role for c-Myb during the pro-B to pre-B cell transition in Myb^f/f CD19-cre mice (27). CD19-cre is produced late during the pro-B cell stage, and we did not detect decreased production or turnover of pro-B cells in Myb^f/f CD19-cre mice. However, the number of pre-B cells was reduced in Myb^f/f C19-cre mice, and we detected counterselection of the deleted Myb allele in pre-B cells compared with pro-B cells, suggesting that c-Myb was important for the proliferation, survival, or differentiation of large pre-B cells (27, 28). In principle, decreased survival of large and small pre-B cells could explain the decreased number of pre-B cells in these mice. However, when we crossed a Bcl-2–producing transgene onto the Myb^f/f CD19-cre background, which rescued survival of c-Myb–deficient pro-B cells, we found that it failed to rescue the number of large and small pre-B cells, suggesting that c-Myb is important beyond controlling survival during the transition from the pro-B cell to pre-B cell compartment. Indeed, we found that c-Myb–deficient large pre-B cells are hypoproliferative compared with c-Myb–sufficient large pre-B cells, demonstrating that c-Myb is crucial for proliferation of large pre-B cells during the pro-B to small pre-B cell transition.

To gain insight into the changes in gene expression that underlie the failure of c-Myb–deficient large pre-B cells to proliferate, we examined the expression of critical genes that are required for proliferation during the large pre-B cell stage. We found that mRNAs encoding IL-7Rα (Il7r), lambda-5 (Igll1), Cxcr4 (Cxcr4), and cyclin D3 (Ccnd3) were downregulated in the absence of c-Myb. Importantly, lambda-5, which is a component of the pre-BCR surrogate L chain, and cyclin D3 are crucial for the development and proliferation of large pre-B cells and progress across the pre-BCR checkpoint (21, 64–66). In addition, mRNA encoding CXCR4 (Cxcr4), which is thought to guide the migration of pre-B cells within the bone marrow microenvironment (67, 68), was also downregulated in the absence of c-Myb, suggesting that c-Myb may play a crucial role in guiding the migration of pre-B cells away from IL-7–producing stromal cells, with the consequence of effectively dampening IL-7 signaling, which is important for differentiation to the small pre-B cell stage (2, 3). Of these genes, Igll1 and Cxcr4 have previously been implicated as direct c-Myb targets (56, 58), although direct binding of c-Myb to the Igll1 promoter has not previously been reported. Direct c-Myb binding to the Cxcr4 promoter region was reported in MCF-7 breast carcinoma cells (56). ChIP of c-Myb in Rag2^−/− pro-B cells revealed that the Igll1 promoter is a direct c-Myb target during the pro-B cell stage; however, we were unable to detect direct interaction of c-Myb with the CXCR4, cyclin D3, or IL-7Rα promoters in pro-B cells. It remains possible that c-Myb directly regulates expression of these genes in pro-B cells through unknown regulatory elements or indirect mechanisms. Our results make clear that c-Myb is important beyond the maintenance of survival during B cell development and coordinates survival with the expression of genes that are important for differentiation to the next developmental stage. In addition, these findings suggest that c-Myb is important at stages of B cell development after the pro-B cell stage.

We thank Dr. Ulrike Lorenz, Dr. Kodi Ravichandran, and Dr. Loren Erickson for advice and valuable discussions. We are grateful to the Flow Cytometry Core Facility (University of Virginia) and, in particular, we thank Joanne Lannigan, Michael Solga, Claude Chew, and Sebastien Coquery for expert help and advice.

The authors have no financial conflicts of interest.