The basic helix-loop-helix transcription factor E2A is an essential regulator of B lymphocyte lineage commitment and is required to activate the expression of numerous B lineage-specific genes. Studies involving ectopic expression of Id proteins, which inhibit E2A as well as other basic helix-loop-helix proteins such as HEB, suggest additional roles of E2A at later stages of B cell development. We use E2A-deficient and E2A and HEB double-deficient pre-B cell lines to directly assess the function of E2A and HEB in B cell development after lineage commitment. We show that, in contrast to the established role of E2A in lineage commitment, elimination of E2A and HEB in pre-B cell lines has only a modest negative impact on B lineage gene expression. However, E2A single and E2A and HEB double-deficient but not HEB single-deficient cell lines show dramatically enhanced apoptosis upon growth arrest. To address the possible role of E2A in the regulation of B cell survival in vivo, we crossed IFN-inducible Cre-transgenic mice to E2A conditional mice. Cre-mediated E2A deletion resulted in a block in bone marrow B cell development and a significant reduction in the proportion and total number of splenic B cells in these mice. We show that Cre-mediated deletion of E2A in adoptively transferred mature B cells results in the rapid depletion of the transferred population within 24 h of Cre induction. These results reveal that E2A is not required to maintain B cell fate but is essential in promoting pre-B and B cell survival.
Blymphocyte development in mice is one of the best characterized models to study mammalian cell lineage commitment and differentiation (for reviews, see Refs. 1 and 2). Commitment of lymphoid progenitors to the B cell lineage in the adult bone marrow is guided by the coordinated activities of transcription factors and cytokine signals. The transcription factors Pu.1, E2A, early B cell factor (EBF),3 and Pax-5 play key roles in regulating the transition from lymphoid precursor to lineage-committed B cell progenitor (3, 4, 5, 6, 7). Progression through B cell development is characterized by the sequential somatic recombination of the Ig H (IgH) and L chain (IgL) Ag receptor genes which ultimately leads to the expression of a functional BCR on the surface of mature B lymphocytes. B lineage-specific gene expression is regulated in part by the combined activities of E2A, EBF, and Pax-5. In addition, it has been shown that E2A and Pax-5 may also play a role in the suppression of non-B lineage gene expression (8, 9).
The E2A gene encodes the basic helix-loop-helix (bHLH) transcription factors E12 and E47 and is absolutely required for commitment of lymphoid progenitors to the B cell lineage (4, 5, 10). In addition, the bHLH transcription factors HEB and E2-2 have also been implicated in the regulation of B cell development (11). E12, E47, HEB, and E2-2 are classified in a transcription factor family known as E proteins based on structural and functional similarities. Mice transheterozygous for E2A/HEB, E2A/E2-2, or HEB/E2-2 exhibit reductions in progenitor B cell number that are greater than a mouse heterozygous for any single one of these genes (11). Therefore, the combined expression of these three E proteins is required for an optimal level of B cell development. E2A has also been directly implicated in the transcriptional regulation of many B lineage-specific genes and has been shown to be essential for Ig H and L chain recombination (5, 12, 13, 14). Recent studies using E2A-deficient hemopoietic progenitors and pre-B cell lines demonstrate that E2A promotes, as well as suppresses, expression of a broad array of transcripts in B cells (8, 15). E2A is required to initiate expression of many B lineage-specific genes such as EBF, mb-1, and B29. However, analysis of E2A−/− pre-B cells revealed that E2A is not required to maintain the expression of these genes. These results raise the possibility that other E proteins, such as HEB, compensate for the loss of E2A and maintain B cell-specific gene expression.
E proteins have also been implicated in the regulation of lymphocyte survival. Multiple studies have demonstrated that increased expression of Id proteins, which are specific E protein inhibitors, correlates with enhanced apoptosis in developing T and B cells (16, 17, 18). Specifically, in B cells, induction of Id3 expression in progenitor B cells has been shown to induce apoptosis through a caspase-2-dependent mechanism (19). Although these studies suggest that E proteins play a role in promoting lymphocyte survival, they do not identify the specific E protein that may be mediating these effects. Additionally, Id proteins have been shown to directly inhibit the activity of members of the ternary complex factor subfamily of ETS transcription factors as well as members of the Pax transcription factor family (20, 21). Thus, it is possible that the observed phenotype in Id overexpression studies is a result of the combined inhibition of multiple transcription factors. Id3 expression also severely perturbs progenitor B cell proliferation raising the possibility that apoptosis may be a secondary consequence of catastrophic growth arrest.
We created pre-B cell lines deficient in E2A, HEB, or both E2A and HEB to evaluate the contribution of these E proteins to B cell gene expression, proliferation, and survival. In addition, we created mice in which E2A deletion could be induced to directly examine the consequence of E2A deletion beyond B lineage commitment. We found B lineage identity is maintained in E2A−/−HEB−/− pre-B cell lines and that E2A deficiency correlated with enhanced apoptosis when growth arrest was induced by treatment with STI-571 (Gleevec). We also show that disruption of E2A in mature B cells results in rapid cell loss. Together, these data directly implicate E2A in the regulation of progenitor and mature B lymphocyte survival.
Materials and Methods
Mice and cell lines
E2Af/fMx-Cretg mice were created by crossing mice carrying the type I IFN-inducible Cre transgene (Mx-Cre) (22) with E2A conditional mice (23). Mx-Cre expression was induced by i.p. injection of 400 μg of polyinosinic-polycytidylic acid (pIpC) (gamma-irradiated; Sigma-Aldrich) dissolved in 1× PBS. Wild-type CD45.1+CD45.2+ donor cells used in adoptive transfer experiments were obtained from the F1 progeny of CD45.1 and CD45.2 congenic strains on a B6 background. E2A/HEB double conditional pre-B cell lines (E2Af/f HEBf/f) were derived from Abelson murine leukemia virus (Ab-MuLV) transformed bone marrow pro-B cells of an E2Af/f HEBf/f mouse (J. Wojciehowski and Y. Zhuang, unpublished data). Cre recombinase was expressed in the cell lines by transduction with MIGR1-Cre retrovirus. Floxed and deleted E2A alleles were detected by genomic PCR with the flowing primers: E2Aflox for (CTGCACTCCGAATTGTGCCTG), PGKneo for (GCCCATTCGACCACCAAGCG), and YZ198 (GATCCTCGTCTTCATTGGTACTG). Floxed and deleted HEB alleles were detected by genomic PCR with the following primers: JW1 (CTGGGACAGAAGTTCAGCACTTAGTAC) and JW2 (CATTCCTATACATCAGCTTCTTGGACG). All pre-B cell lines were maintained at 37°C in RPMI 1640 medium (Invitrogen Life Technologies) supplemented with 10% heat-inactivated FBS (HyClone), 100 U/ml penicillin (Invitrogen Life Technologies), 100 μg/ml streptomycin (Invitrogen Life Technologies), and 50 μM 2-ME (J. T. Baker).
STI-571 (Gleevec; Novartis) was prepared as a 10-mM stock solution dissolved in H2O plus 10 mM HCl, sterile filtered, and stored at −20°C until use.
Quantitative real-time PCR gene expression analysis
Total RNA was prepared using TRIzol (Invitrogen Life Technologies) followed by isopropanol precipitation. For all samples, 2 μg of purified total RNA was treated with RNase free DNase I (Sigma-Aldrich) and random-primed cDNA was made using Moloney murine leukemia virus reverse transcriptase (Invitrogen Life Technologies). Quantitative RT-PCR was performed with a Roche LightCycler and the FastStart DNA master SYBR green I kit (Roche) as per the manufacturer’s instructions. All of the samples were normalized to the expression of GAPDH. The primers for RT-PCR are as follows: EBF, EBF no. 2 forward (for) (CATGTCCTGGCAGTCTCTGA) and EBF no. 2 reverse (rev) (CAACTCACTCCAGACCAGCA); Pax-5, Pax-5 for (CCGCCAAAGGATAGTGGAACTTG) and Pax-5 rev (CACAGTGTCATTGTCACAGACTCGC); E2-2, E2-2 for (GGTTCCGGTCCCACAACTTC) and E2-2 rev (CGCTCAGCCTTCTGCTCTGG); IgH Iμ, Imu for (GAGAGCCCCCTGTCTGATAAGAATC) and Imu rev(CGGTTTTGGAGTGAAGTTCGTG); mb-1 (Igα), mb-1 for (CCTCTCCTCCTCTTCTTGTCATACG) and mb-1 rev(CCCCTGTGTTCTTGTTTACTTCGG); B29 (Igβ), B29 for (TGTTCCTGCTGCTGCTCTTCTC) and B29 rev (TCGGTGACATTATGGTTGGCG); λ 5, λ 5 for (CAGATCATCCCACGGGGAGC) and λ 5 rev (TGAGTGACAGGGACCCCATC); VpreB, VpreB no. 2 for (CGTCTGTCCTGCTCATGCT) and VpreB no. 2 rev (ACGGCACAGTAATACACAGCC); GAPDH, GAPDH for (CCTGGAGAAACCTGCCAAGTATG) and GAPDH rev (AGAGTGGGAGTTGCTGTTGAAGTC).
Quantitative PCR analysis of E2A deletion
A standard curve of deleted E2A alleles was created by mixing predetermined numbers of E2A−/−HEB−/− (del) and E2Af/fHEBf/f (flox) pre-B cells together. A total of 1 × 105 pre-B cells were mixed in the following ratios: 90% del: 10% flox; 75% del: 25% flox; 50% del: 50% flox; 25% del: 75% flox; 10% del: 90% flox. Cells were lysed in Triton lysis buffer (Tris-EDTA (pH 7.5) + 0.2% Triton X-100 + 0.2 mg/ml proteinase K) by incubation at 55°C for 30 min followed by 95°C for 10 min then quickly cooled and stored at 4°C. Quantitative PCR was performed on these samples to generate a standard curve of E2A deletion. Quantitative PCR was performed on a Roche LightCycler using the FastStart DNA master SYBR green I kit (Roche) as per the manufacturer’s instructions. Each sample was first normalized by quantitative PCR amplification of the CD14 genomic locus and then the quantitative value of E2A deletion for each sample was plotted. A linear trend line was fitted to the data points using Microsoft Excel. The equation of the trend line was used to calculate the percent of E2A deletion in unknown samples which had been normalized to CD14. The primers E2Aflox for and YZ198 are used to amplify the deleted E2A allele and CD14 is amplified with CD14 for (GCTCAAACTTTCAGAATCTACCGAC) and CD14 rev (AGTCAGTTCGTGGAGGCCGGAAATC).
Adoptive B cell transfer experiments
Single-cell suspensions of donor (age: 2–3 mo) splenocytes were prepared in FACS buffer (1× PBS (pH 7.4) + 5% bovine calf serum). RBC were lysed using 0.16 M NH4Cl and total splenocytes were labeled on ice with CD43 (BD Pharmingen), CD93 (eBioscience), CD4 (BD Pharmingen), CD8β (BD Pharmingen), Mac-1 (Caltag Laboratories), Ter119 (BD Pharmingen), and pan-NK (BD Pharmingen) PE-conjugated Abs and Gr-1 (Caltag Laboratories) Cy5-PE (Cy5PE) conjugated Ab in FACS buffer. Immature, transitional and non-B splenocytes were depleted by AutoMACS (Miltenyi Biotec) using anti-PE Ab labeled microbeads (Miltenyi Biotec) as per the manufacturer’s instructions. Enriched B cells were washed and suspended in 1× PBS (pH 7.4) and injected via the tail vein into sublethally irradiated (300 rad) recipients (B6 CD45.1 congenic mice, age 2–3 mo). All animal procedures were approved by the Duke University Institutional Animal Care and Use Committee.
Cell cycle analysis and annexin V staining
For cell cycle analysis, pre-B cells were fixed in ice-cold ethanol and stained with propidium iodide in the presence of RNase A (Sigma-Aldrich). For annexin V staining, cells were first stained with appropriate Abs, washed in 1× annexin V binding buffer (BD Pharmingen) and then labeled with Annexin VPE conjugate (BD Pharmingen) at room temperature in 1× annexin V binding buffer as per the manufacturer’s instructions.
Bone marrow was prepared in FACS buffer and stained with CD43 PE, B220 allophycocyanin-conjugated Abs (BD Pharmingen), and 7-actinomycin-D (7AAD; BD Pharmingen). Splenocytes were prepared in FACS buffer and stained with IgM FITC, CD19 PE, IgD biotin, and B220 allophycocyanin-conjugated Abs followed by streptavidin-Cy5 PE staining. Mice analyzed were 2–3 mo old. All staining was done on ice. Samples were run on a FACSCalibur (BD Biosciences) and analyzed with FlowJo software (Tree Star).
The loss of E2A and HEB perturbs but does not abolish B lineage-specific gene expression in pre-B cells
We have previously established E2A-deficient Ab-MuLV transformed progenitor B cell lines to study E2A regulation of pre-B lymphocyte gene expression and differentiation (13, 15). E2A−/− pre-B cells exhibit a block in Ig L chain recombination but retain expression of many B lineage-specific genes including known E2A targets such as EBF, Ig H chain, and mb-1 (13, 15). These findings raise the possibility that other E proteins may be able to compensate for the loss of E2A. To test this hypothesis, we created Ab-MuLV transformed pre-B cell lines that are deficient in both E2A and HEB. Conditional E2A (E2Af), previously referred to as E2Aloxp, mice were crossed to mice carrying a conditional HEB allele (HEBf) to produce E2Af/fHEBf/f offspring. The conditional E2A and HEB alleles were created by flanking the exons that encode the E47 and E12 bHLH domains of E2A and the bHLH domain of HEB with loxP recombination sites (Ref. 23 ; J. Wojciehowski and Y. Zhuang, unpublished data). Whole bone marrow from an adult E2Af/f HEBf/f mouse was transformed with Ab-MuLV and the resulting stable pre-B cell line was infected with a retrovirus expressing Cre recombinase to simultaneously delete E2A and HEB (Fig. 1,A). Clonal cell lines were derived by limiting dilution plating of retrovirally infected cells and genomic PCR screening was performed to confirm the deletion of E2A and HEB. Both E2A and HEB alleles were deleted in the majority of Cre-transduced clones screened (Fig. 1 B).
Once E2A−/−HEB−/− clones were established, we examined the expression of known E2A target genes by quantitative RT-PCR (qRT-PCR). Gene expression in E2A−/−HEB−/− clones was compared with that of E2A−/− clones and the parental E2Af/fHEBf/f cell line. E2A is required to properly induce EBF expression upon commitment of lymphoid progenitors to the B cell lineage (12, 24). Therefore, we measured the expression of EBF and Pax-5 in E2Af/f HEBf/f, E2A−/−, and E2A−/− HEB−/− pre-B cells. We have previously shown that EBF and Pax-5 expression is unchanged in E2A−/− pre-B cells as determined by semiquantitative RT-PCR analysis (15). Quantitative RT-PCR analysis revealed a small decrease in the expression of both EBF and Pax-5 in E2A−/− and E2A−/−HEB−/− pre-B cells (Fig. 1 C). These data demonstrate that the expression of EBF and Pax-5 can be maintained despite the loss of both E2A and HEB.
E2A is known to regulate the expression of Ig H chain (IgH) (25). Therefore, we examined IgH expression by measuring transcription occurring through the μ H chain constant exon (Cμ). Cμ expression was reduced 2-fold in E2A−/− cells and 4-fold in E2A/HEB double-deficient pre-B cells relative to wild type indicating that IgH expression is influenced by E protein gene dose (Fig. 1 C). We also examined expression of known E2A targets CD79a (Igα), CD79b (Igβ), λ5, and VpreB. Expression of Igα and Igβ was decreased 2-fold in E2A−/−HEB−/− pre-B cells relative to the control pre-B cell line. The expression of surrogate L chain genes λ5 and VpreB was reduced 2-fold in both E2A−/− and E2A−/−HEB−/− pre-B cell lines when compared with E2Af/fHEBf/f pre-B cells. We confirmed that expression of the E protein E2-2 remained unchanged by E2A and HEB gene status thus ruling out the possibility that increased E2-2 expression was compensating for the deficiency in E2A and HEB expression. These data indicate that B lineage-specific gene expression is maintained in the absence of both E2A and HEB and show that HEB is involved in regulating the expression of a subset of E2A target genes in B cells.
E2A-deficient pre-B cells undergo rapid cell death upon induction of growth arrest and differentiation
Previous studies have shown that overexpression of Id3 proteins, which block E protein activity, can inhibit progenitor B cell proliferation and induce cell death (19). Because Id proteins are nonspecific inhibitors of E protein activity, it is not clear whether Id-mediated growth inhibition and cell death is due to the direct inhibition of E2A or due to the inhibition of another target. Therefore, we investigated whether E2A−/− and/or HEB−/− pre-B cells exhibit altered proliferation.
We examined total cellular DNA content to determine the cell cycle kinetic profiles of the E2Af/f and E2A−/− transformed pre-B cell lines that we previously established (15). We observed that slightly more of the E2Af/f pre-B cells were actively cycling (57% cycling cells) versus E2A−/− pre-B cells (46% cycling cells) (Fig. 2,A). Next, we examined the cell cycle kinetics of pre-B cell lines deficient in both E2A and HEB. We measured total cellular DNA content of the parental E2Af/fHEBf/f cell line, a Cre-transduced clone which had lost HEB but retained both floxed copies of E2A (E2Af/fHEB−/−) and three E2A−/−HEB−/− clones. Cell cycle analysis of three independent E2A−/−HEB−/− clones revealed variable differences in the proportion of cycling cells compared with E2Af/fHEBf/f or E2Af/fHEB−/− cells (Fig. 2 B). We observed that in E2A−/−HEB−/− clones 7, 9, and 11, ∼37, 45, and 47% of the cells were actively cycling, respectively. Approximately 51% of the E2Af/fHEBf/f or E2Af/fHEB−/− cells were actively cycling. Analysis of population growth kinetics confirmed that E2A−/−HEB−/− clones 7 and 9 proliferate slower than either of the E2Af/fHEBf/f or E2Af/fHEB−/− cell lines (data not shown).
Abelson virus-transformed pre-B cells can be induced to undergo growth arrest and differentiation by inhibiting the activity of the v-Abl oncogene with the small molecule inhibitor STI-571 (26). Therefore, we asked whether E proteins are required for cell survival of pre-B cell lines that are reverted to an untransformed state with STI-571. We treated E2Af/fHEBf/f, E2Af/fHEB−/−, E2A−/− or E2A−/−HEB−/− pre-B cell lines with 1 μM STI-571 for 12 h and monitored cell viability by annexin V and 7AAD staining. We observed a large proportion of apoptotic cells (7AAD−annexin V+) in E2A−/− and E2A−/−HEB−/− pre-B cell cultures but not in E2Af/fHEBf/f or E2Af/fHEB−/− cultures after 12 h of STI-571 treatment (Fig. 2,C). The E2Af/fHEB−/− pre-B cell line showed only a small increase in 7AAD−annexin V+ cells after 12 h of STI-571 treatment. The increase in annexin V+ cells observed in E2A−/− and E2A−/−HEB−/− clones after STI-571 treatment coincided with a rapid decline in the number of viable cells over the 12 h of treatment (Fig. 2 D). STI-571-treated E2Af/fHEBf/f and E2Af/fHEB−/− pre-B cells showed little decline in cell viability over this same time period. Clones deficient in both E2A and HEB did not exhibit a further decrease in cell viability when compared with E2A single-deficient clones, indicating that the loss of E2A but not HEB is sufficient to sensitize pre-B cell lines to apoptosis. These data indicate that E2A protects pre-B cell lines from apoptotic cell death when the cells are induced to undergo growth arrest and differentiation. These data raise the possibility that E2A may play a broader role in promoting B cell survival.
Conditional disruption of E2A in vivo perturbs B lymphocyte development in the bone marrow
Results from E2A-deficient pre-B cell lines indicate that E2A regulates the expression of multiple B lineage-specific genes and may play an important role in promoting B lymphocyte survival. Because E2A knockout mice lack B cells, we used a conditional gene disruption approach to investigate the in vivo function of E2A in developing and mature B cells. We crossed type I IFN inducible Cre (Mx-Cre) transgenic mice with our E2A conditional mice (E2Af) to generate E2Af/fMx-Cretg mice (22). E2A deletion was induced by injecting these mice with pIpC. E2Af/f or E2Af/fMx-Cretg mice were given three i.p. injections of 400 μg of pIpC every other day and sacrificed 2 days after the last injection. Analysis of pIpC-treated Cre+ mice revealed a near complete loss of pro-B cells (B220+CD43high) and a significant reduction in pre-B cells (B220+ CD43low) in the bone marrow (Fig. 3 A). Cre− mice treated with pIpC also had significantly smaller proportion of pre-B cells than untreated controls. This side effect of pIpC treatment is thought to be due to the ability of type I IFN to antagonize IL-7-dependent growth and development of progenitor B lymphocytes (27). We also found that the proportion of pre-B cells in untreated Cre+ mice was reduced as well, possibly as a consequence of leaky Cre expression.
To determine the efficiency of E2A deletion, we developed a quantitative PCR (qPCR) based method. We adopted this method because it can be used to assay small populations of cells and it is faster than traditional Southern analysis. The qPCR assay was validated using genomic DNA obtained from populations of E2Af/fHEBf/f and E2A−/−HEB−/− pre-B cells that were mixed together in known ratios (Fig. 3,B). qPCR analysis revealed that ∼60% of the E2A alleles had been deleted in the remaining bone marrow cells of pIpC-treated Cre+ mice (Fig. 3 C). However, because B cell development depends on E2A, it is likely that this value underestimates the actual level of Cre-mediated E2A deletion in the bone marrow. A small percentage (<10%) of deleted E2A alleles was also detected in the untreated Cre+ mice indicating that there is leaky expression of the Cre transgene. No E2A deletion was detected in the untreated or pIpC-treated E2Af/fCre− mice by qPCR. These results indicate that remaining bone marrow B cells still retain at least one copy of the floxed E2A allele after pIpC treatment. The significant reduction of the pro-B cell population observed in pIpC-treated E2Af/fMx-Cretg mice could be due to the partial or complete loss of E2A.
The conditional deletion of E2A results in the loss of splenic B lymphocytes
The E2Af/fMx-Cretg mouse is a particularly useful model to examine the effect of E2A gene disruption on mature B lymphocytes. Cre+ and Cre− mice were given three injections of pIpC every other day and splenic B cells were examined 2 days after the final injection. We found that the total number and percent of splenic B cells was significantly reduced in pIpC-treated Mx-Cre+ mice when compared with pIpC-treated Cre− mice (Fig. 4,A). The mean total number of B220+CD19+IgM+ B cells was reduced from 1.0 × 107± 1.8 × 106 in Cre− mice (n = 3) to 3.8 × 106± 1.2 × 106 in Mx-Cre+ mice (n = 4) and the mean proportion of splenic B cells was 56 ± 5.7% (n = 3) and 37 ± 2.9% (n = 8) in Cre− and Mx-Cre+ mice, respectively. Because E2A regulates the expression of the Ig genes as well as Igα and Igβ which are key signaling molecules required for BCR expression, we asked whether E2A disruption could influence the expression level of surface IgM and IgD on splenic B cells. However, we found that the expression of IgM and IgD does not differ significantly between pIpC-treated Cre− and Cre+ mice (Fig. 4,B). Quantitative analysis of E2A deletion in pIpC-treated mice revealed that the floxed E2A allele was significantly enriched (60% floxed) in the CD19+-enriched fraction of E2Af/fMx-Cretg splenocytes versus the CD19− splenocyte fraction (20% floxed) (Fig. 4,C). We performed an additional experiment to determine whether enrichment of the floxed E2A allele in pIpC-treated CD19+ E2Af/fMx-Cretg splenocytes could be due to inefficient E2A deletion in this population. E2A+/fMx-Cretg and E2Af/fMx-Cretg mice were given two pIpC treatments (on days 0 and 2) and sacrificed for analysis 48 h after the last treatment. The relative proportion of wild-type/floxed E2A to deleted E2A in whole spleen and CD19+ enriched splenocytes was determined by Southern analysis. We found that enrichment of the floxed allele in E2Af/f Mx-Cretg mice is not due to inefficient E2A deletion in splenic B cells because CD19+ enriched splenocytes from the pIpC-treated E2A+/fMx-Cretg mouse show approximately equal representation of wild-type and deleted E2A alleles. (Fig. 4 D). We also observed that whole spleen from the E2Af/fMx-Cretg mouse exhibited ∼50% floxed and 50% deleted E2A while only the floxed E2A allele was detectable in the CD19+-enriched splenocyte population. Thus, we consistently observe an overrepresentation of floxed allele in the CD19+ splenocyte fraction of E2Af/fMx-Cretg mice treated with pIpC. It is unlikely that the loss of E2A results in a down-regulation of CD19 surface expression on mature B cells because CD19 expression is normal on E2A−/− pre-B cell lines (15). These results indicate that E2A-deficient B cells are lost from the spleens of pIpC-treated E2Af/fMx-Cretg mice.
Adoptively transferred E2Af/fMx-Cretgmature B lymphocytes are lost upon induction of E2A deletion
The underrepresentation of E2A-deficient splenic B cells observed may result from a block in B cell development in the bone marrow or may be directly due to the deletion of E2A in mature B cells. To distinguish these possibilities, we isolated mature B cells from E2A+/f, E2A+/fMx-Cretg, or E2Af/fMx-Cretg mice and transferred these cells to sublethally irradiated wild-type recipients. Recipient mice then received a total of three pIpC injections given every other day and were sacrificed 2 days after the final treatment (Fig. 5,A). Approximately 2 × 106 mature B cells of each genotype were transferred. Transferred B cells were distinguished from host cells by CD45.2 surface expression. E2A+/f and E2A+/fMx-Cretg B cells were readily detectable in spleens of the recipient mice after pIpC treatment, however, very few E2Af/fMx-Cretg splenic B cells could be detected after the pIpC treatments (Fig. 5 B). These results indicate that the deletion of E2A in mature B cells results in the loss of these cells in vivo.
The nearly complete lack of E2Af/fMx-Cretg B cells found in recipient spleens after the three pIpC treatments suggests that the deletion of both copies of E2A results in the rapid loss of mature B cells. To better understand the kinetics by which transferred E2Af/fMx-Cretg B cell are lost, we mixed purified mature B cells from wild-type mice with an equal number of B cells from E2Af/f or E2Af/fMx-Cretg mice and transferred these mixed populations to sublethally irradiated wild-type recipients. Differential surface expression of the markers CD45.1 and CD45.2 was used to distinguish donor wild-type (CD45.1+CD45.2+), donor E2Af/f or E2Af/fMx-Cretg (CD45.1−CD45.2+), and recipient (CD45.1+CD45.2−) B cells. A total of 2 × 106 wild-type + 2 × 106 E2Af/f or E2Af/fMx-Cretg B cells were transferred to each recipient and 400 μg of pIpC was administered to the recipient mice 4 h after cell transfer (Fig. 6,A). Recipient mice were sacrificed at 4 and 24 h after the pIpC treatment and the ratio of donor-derived wild-type to E2Af/f or E2Af/fMx-Cretg splenic B cells was determined by FACS. We found that the proportion of both Cre− and Cre+ B cells to wild-type B cells had decreased 4 h after pIpC treatment (Fig. 6, B and C). These results suggest that both E2Af/f and E2Af/fMx-Cretg B cells engraft less efficiently than wild-type cells. By 24 h post-pIpC treatment, the ratio of donor E2Af/fMx-Cretg to wild-type B cells decreased an additional 30%, whereas the ratio of E2Af/f to wild-type B cells increased by ∼20%. We examined the change in the ratio of transferred E2Af/fMx-Cretg to wild-type B cells over time in two independent experiments and observed an equivalent decline in the proportion of Cre+ B cells between 4 and 24 h after pIpC treatment (Fig. 6 D). These data indicate that a single pIpC treatment is sufficient to eliminate a significant proportion of the transferred E2Af/fMx-Cretg B cells within 24 h of Cre induction. We were able to detect both floxed and deleted E2A alleles in splenic genomic DNA from recipients that received E2Af/fMx-Cretg B cells at 24 h, indicating that a single round of Cre induction does not completely delete all floxed E2A alleles (data not shown). These data support the conclusion that mature B cells are lost upon deletion of E2A.
E2A-deficient pre-B cells lines express many B lineage-specific genes, including known E2A targets, raising the possibility that other E proteins can regulate B lineage-specific gene expression in the absence of E2A. As expected, disruption of E2A alone reduced the transcript levels of known E2A targets. However, the deletion of both E2A and HEB in pre-B cells negatively impacted the expression of only a subset of E2A target genes. Of the genes we assayed, IgH Cμ expression was most dependent on expression of E2A and HEB. These data indicate IgH expression is particularly sensitive to E protein gene dose and suggest that the expression level of IgH may be regulated by the overall level of E protein activity within the B cell. Interestingly, the disruption of both E2A and HEB had only a minimal effect on the expression of EBF. E2A directly associates with the 5′ region of the EBF locus and is required to induce EBF expression at the earliest stages of B cell development (12, 14). Based on these results it seems plausible that E2A may play a role in regulating the expression level of EBF in B cells. However, we find that EBF expression is maintained at near normal levels in E2A−/−HEB−/− pre-B cells, suggesting that at later stages of B cell development, EBF expression is regulated independently of E protein activity. These data suggest that EBF expression switches from an E2A-dependent mechanism to a mechanism less dependent on E proteins once lymphoid precursors commit to the B cell lineage. It is possible that this switch in control of EBF expression may serve as one method to enforce commitment to the B cell lineage. Our results also demonstrate that while disruption of E2A and HEB negatively impacts the expression of multiple B lineage-specific genes, overall B cell lineage identity is retained, presumably due to E2A- and HEB-independent maintenance of EBF and Pax-5 expression.
E2A been has been shown to both promote or inhibit cell proliferation under different experimental conditions (28, 29, 30). Studies using EBF-reconstituted E2A−/− progenitor B cells revealed a role for E proteins in the optimal expression of the pro-proliferative genes N-myc and c-myc and in the promotion of IL-7-dependent proliferation (24). However, we observe only a minor defect in proliferation of E2A−/− or E2A−/−HEB−/− pre-B cell lines suggesting that there may be other mechanisms that drive proliferation of these cells independently of E2A and HEB. It has been shown that v-Abl can promote c-myc expression, thus v-Abl signaling may be sufficient to maintain c-myc in the absence of E2A and HEB to promote proliferation (31). Alternately, it is possible that the regulation of N-myc and c-myc expression in our pre-B cell lines is mediated through a mechanism that is independent of E2A and HEB. Future studies should help to resolve these questions.
Our studies in E2A-deficient pre-B cell lines and in E2A-conditional mice strongly suggest that E2A regulates an important intrinsic survival pathway in B cells. We found that the loss of E2A correlates with dramatically enhanced cell death in pre-B cell lines. Interestingly, this phenotype was only observed when we treated E2A−/− or E2A−/−HEB−/− pre-B cell lines with STI-571. These data suggest that v-Abl provides survival signals which can complement E2A deficiency. Mature B cell survival is dependent in part on signaling through a functional BCR on the cell surface. Ablation of the Ig H chain or the BCR signaling molecule Igα on mature B cells has been shown to block this necessary BCR survival signal ultimately resulting in cell death (32, 33). Because E2A regulates the expression of multiple genes which are required to produce the functional BCR, we investigated whether insufficient BCR expression could explain the reduction of B cells observed in E2Af/fMx-Cretg mice after Cre induction. However, we did not observe a down-regulation of surface IgM or IgD on B220+ splenic B cells in pIpC-treated E2Af/fMx-Cretg mice. Adoptive transfer experiments revealed that E2Af/fMx-Cretg B cells were lost within 24 h after Cre induction. These data show that B cells die very rapidly upon E2A deletion. It is possible that BCR expression is quickly down-regulated upon E2A deletion and we fail to observe a population of B cells with low BCR expression because these cells do not survive long enough to establish a significant, detectable population.
However, we do not believe that ablation of BCR expression is the primary cause of B cell death in our experimental system for the following reasons. First, qRT-PCR analysis of our E-protein-deficient pre-B cell lines suggests that the expression of Igα, Igβ, or IgH is not absolutely dependent on E2A and may be regulated by HEB and/or E2-2. Second, IgMlowB220+ B cells can be detected in the spleen of mice after Mx-Cre mediated disruption of Ig H chain on mature B cells (33). These data suggest that the pre-existing BCR on the cell surface is gradually down-regulated when the synthesis of new receptors is blocked. In our studies, we do not observe an analogous IgMlowB220+ B cell upon E2A disruption indicating that E2A-deficient B cells die before a reduction in BCR expression can be observed. Finally, E2A protects transformed pre-B cells from apoptosis in a manner which is not dependent on BCR expression as these cells are at a developmental stage in which they have not fully rearranged their Ig genes. These lines of evidence strongly suggest that E2A promotes B cell survival through a novel mechanism which is independent of its role in regulating B cell Ag receptor expression.
We also observed that E2Af/f B cells failed to engraft as efficiently as wild-type B cells in a competitive adoptive transfer. The E2Af allele is hypomorphic due to the presence of the PGKneo cassette 3′ of the E2A gene. Therefore, E2Af/f mice exhibit a phenotype that is intermediate between wild-type and E2A+/−. It is possible that decreased E2A expression may contribute to differences in initial B cell engraftment. Additionally, differences in the genetic backgrounds of E2Af/f donor mice, which are mixed between B6, 129, and SJL, and wild-type donor mice, which are B6 CD45 congenic, may also contribute to the observed differences in B cell engraftment efficiency.
The specific mechanism through which E2A may regulate B cell survival is currently unclear. It has been shown that Id3 overexpression in progenitor B cells leads to growth arrest and death via a caspase-2-dependent pathway (19). It was demonstrated that Id3-mediated cell death occurred in a p53-independent manner and could not be rescued by a Bcl-2 transgene. Currently, none of the genes that are known to be directly regulated by E2A can fully account for the apparent antiapoptotic activity of E2A in our pre-B cell lines. However, E2A−/− pre-B cell lines will serve as a useful tool to dissect the molecular pathways through which E2A regulates B cell survival.
The finding that E2A is required for the survival of mature B cells is notable. We propose that E2A promotes B cell survival by a common mechanism in progenitor and mature B cells. The level of E2A protein declines significantly as developing B lymphocytes mature and enter the resting peripheral B cell pool but increases rapidly upon B cell activation (34, 35, 36). It is well-established that pIpC can transiently activate resting B cells in T cell-independent manner; therefore, it is possible E2A may promote the survival of recently activated B cells. Thus, it will be particularly interesting to determine whether E2A is involved in promoting the survival of resting versus activated peripheral B cells.
We thank the laboratory of Dr. Mike Krangel for technical assistance and advice, the laboratory of Dr. Motonari Kondo for assistance with the B cell enrichments, Dr. Garnett Kelsoe for helpful discussions, and Mary Elizabeth Jones and Jingquan Jai for review and critical discussions of the manuscript.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grant CA072433 (to Y.Z.).
Abbreviations used in this paper: EBF, early B cell factor; bHLH, basic helix-loop-helix; pIpC, polyinosinic-polycytidylic acid; Ab-MuLV, Abelson murine leukemia virus; for, forward; rev, reverse; 7AAD, 7-actinomycin-D; qRT-PCR, quantitative RT-PCR; qPCR, quantitative PCR; Cμ, μ H chain constant exon.