Spi-C and PU.1 Counterregulate Rag1 and Igκ Transcription to Effect Vκ-Jκ Recombination in Small Pre-B Cells

Bcell development requires the ordered rearrangement of Ig genes encoding H and L chain proteins that assemble into BCRs or Abs capable of recognizing specific Ags. Ig rearrangement proceeds in a developmental sequence in which IgH V(D)J recombination precedes Igκ or Igλ V-J recombination. Igκ/λ V-J recombination occurs in small pre-B cells that have previously rearranged a functional IgH allele, have ceased proliferation in response to IL-7, and have turned on high expression of recombinase-activating gene (RAG) proteins RAG1 and RAG2 (1, 2). Igκ rearrangement is promoted by chromatin accessibility, regulated through transcription, and regulated by relative abundance of RAG1/2 proteins (3).

Igκ accessibility and transcription are controlled by two locus-specific enhancers: the intronic enhancer and the 3′ enhancer (4). Igκ chromatin accessibility correlates with germline transcription initiating upstream of the Jκ1 segment; this transcript is alternatively named k⁰ or germline κ-1 (Glk1) (5, 6). The 3′ enhancer contains an E26 transformation-specific (ETS) transcription factor binding site that regulates B cell versus T cell specificity of Vκ-Jκ joining (7). Igκ transcription and accessibility are repressed by STAT5 interaction with the Igκ intronic enhancer downstream of IL-7 receptor signaling in pro-B cells and large pre-B cells (8, 9).

Regulation of RAG protein abundance is an important second mechanism for regulation of Igκ recombination. RAG1 and RAG2 together form the recombinase that catalyzes V(D)J recombination by recognition and cleavage of recombination signal sequences (10). Appropriate developmental regulation of Rag1/Rag2 is essential to prevent off-target mutations leading to leukemia or lymphoma (11). RAG1/2 protein levels are regulated by specific degradation linked to the cell cycle (12). The closely linked Rag1 and Rag2 genes are also regulated transcriptionally (13). Several regulatory regions or enhancers of the Rag locus have been identified, including Erag (14), R-Ten, and R1B/R2B (15). These regulatory regions interact with lineage-specific transcription factors to activate or repress Rag transcription at specific developmental stages (15). RAG-induced DNA double-stranded breaks (DSBs) promote allelic exclusion by signaling through ataxia telangiectasia mutated (ATM) kinase and the Nemo (NF-κB essential modulator) protein to repress Rag and Igκ transcription (16).

One of the genes activated by DSBs in small pre-B cells through the ATM/NF-κB signaling pathway is Spic encoding the ETS transcription factor Spi-C. Spi-C is a lineage-instructive transcription factor that is important for the generation of myeloid and lymphoid cells (17). Induction of Spi-C in pre-B cells functions to repress cell cycle progression and Igκ recombination by interacting with and repressing the Syk, Blnk, and Igκ genes (18, 19). Spi-C is required for the development of splenic red pulp macrophages (20, 21). Spi-C is dynamically regulated in response to growth factor signals to regulate gene expression in B cells (22, 23).

Spi-C is closely related to the ETS family transcription factors PU.1 and Spi-B that function as complementary transcriptional activators of BCR signaling and Igκ rearrangement (24). Combined deletion of Spi1 and Spib genes encoding PU.1 and Spi-B in mice results in a block to B cell development at the small pre-B cell stage followed by development of precursor B cell acute lymphoblastic leukemia (25, 26). PU.1 and Spi-B function as transcriptional activators, whereas Spi-C functions as a transcriptional repressor by competing for PU.1 binding sites (23, 27, 28). In addition, DSBs induced by RAG activate Spi-C to displace PU.1 binding from the Igκ 3′ enhancer to repress transcription (19). Therefore, PU.1/Spi-B and Spi-C may function to oppose one another’s function to regulate BCR signaling and Igκ rearrangement during B cell development.

In this study, we investigated the hypothesis that Spi-C and PU.1 function as regulators of Igκ recombination by regulation of Rag transcription as well as Igκ transcription. To do this, we constructed a doxycycline-inducible Spi-C expression system in 38B9 pre-B cells. Induction of Spi-C inhibited Igκ recombination and inhibited Igκ and Rag mRNA transcription in 38B9 cells. To determine if Spi-C is required to repress Igκ and Rag mRNA transcription during B cell development, we generated Spic^−/− mice by intercrossing Spic^+/− mice onto a mixed C57BL/6 and 129S/v background. Igκ and Rag mRNA transcripts were upregulated in small pre-B cells enriched from Spic^−/− mice relative to wild-type pre-B cells. Igκ and Rag1 transcript levels were activated by PU.1. Finally, we used chromatin immunoprecipitation (ChIP) to identify a regulatory binding site for PU.1 and Spi-B upstream of the Rag1 gene. These results indicate that Spi-C and PU.1 counterregulate Rag and Igκ transcription during B cell development.

Spic^+/− mice were maintained as previously described on a C57BL/6 background (29). To generate Spic^−/− mice, Spic^+/− mice were mated to 129Sv mice (Charles River Laboratories, Pointe-Claire, QC, Canada) to generate F1 C57BL/6/129Sv Spic^+/− mice. F1 mice were intercrossed to generate F2 Spic^−/− mice. All experiments with F2 Spic^−/− mice used F2 Spic^+/+ or Spic^+/− littermate mice as controls. All mice were housed in a specific pathogen-free animal facility and monitored under an approved animal use protocol approved by the Western University Animal Care Committee.

Bone marrow (BM) cells were prepared from euthanized mice, and erythrocytes were removed using ammonium-chloride-potassium lysis. Erythrocyte-depleted BM cells were stained with Abs including PE-conjugated anti-CD19 (6D5; BioLegend, San Diego, CA), allophycocyanin-conjugated anti-B220 (RA3-6B2; eBioscience, San Diego, CA), Brilliant Violet 421–conjugated anti-B220 (RA3-6B2; BioLegend), allophycocyanin-conjugated anti-IgM (II/41; BD Biosciences, Franklin Lakes, NJ), FITC-conjugated anti-CD24 (M1/69; BioLegend), PE-conjugated anti-BP-1 (BP-1; BD Biosciences), biotin-conjugated anti-CD43 (S7; BD Biosciences), and PE/cyanine 7–conjugated streptavidin (BioLegend). Infected cultured cells were sorted on the basis of GFP expression. Dead cells were distinguished using SYTOX Blue Dead Cell Stain (Thermo Fisher Scientific, Waltham MA) or 7-aminoactinomycin D viability stain (Thermo Fisher). Cell sorting was performed using a FACSAria II instrument. Flow cytometric analysis was performed using a FACSCanto instrument (BD Immunocytometry Systems, San Jose, CA). Flow data were analyzed using FlowJo 9.1 (BD Biosciences, Ashland, OR). The purity of sorted cells was determined to be ≥98%.

The 38B9 pre-B cell line (30) was cultured in IMDM (Wisent, St.-Jean-Baptiste, QC, Canada) supplemented with 10% FBS (Wisent), 1× penicillin/streptomycin/l-glutamine (Wisent), and 5 × 10⁻⁵ M 2-ME (Sigma-Aldrich, St. Louis, MO). Cell lines were maintained in a 5% CO₂ atmosphere at 37°C. To induce differentiation and Igκ V-J rearrangement, 38B9 cells were treated with 0.5 μM imatinib (Sigma-Aldrich) for 48 h (31). PU.1-inducible i660BM cells were cultured as previously described (32) in complete IMDM supplemented with 5% supernatant from the J558L-IL7 cell line (33).

Murine 3xFLAG-tagged Spi-C cDNA (34) was PCR amplified and ligated into the pRetro-Tre3G (Tet-regulatable element third generation) plasmid (Thermo Fisher). MIG-Tet3G was previously described (24). pRetro-Tre3G-Spi-C and MIG-Tet3G retrovirus was produced by transient transfection of Plat-E retroviral packaging cells (35). 38B9 cells were infected by “spinoculation” followed by enrichment for green fluorescent cells using cell sorting. Sorted cells were grown continuously in media supplied with 1 μg/ml puromycin (Bio Basic, Markham, ON, Canada) for 2 wk for selection of resistant cells. The resulting cell line was named “inducible 38B9” (i38B9). To induce Spi-C overexpression, i38B9 cells were treated with doxycycline (Bio Basic) at 1.5 μg/ml.

RNA was prepared from cultured or sorted cells using the RNeasy kit (Qiagen, Germantown MD), and cDNA was synthesized using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules CA) following the manufacturer’s instructions. Reverse transcriptase Quantitative RT-PCR (RT-qPCR) was performed using SensiFast SYBR No-Rox Master Mix (FroggaBio, Toronto, ON, Canada) on a QuantStudio5 instrument (Applied Biosystems/Thermo Fisher Scientific). Relative frequencies of mRNA transcripts were all determined as ratios compared with mRNA levels of β-actin. Genomic DNA was extracted using the Wizard genomic DNA purification kit (Promega, Madison, WI). DNA (500 ng) from each sample was used for PCR amplification of Vκ-Jκ rearrangements (36). For amplification of Igκ Vκ-Jκ5 rearrangements from cDNA, 50-μl PCRs included magnesium chloride at 1.5 mM, Vκ primer at 3.2 μM, Jκ5 primer at 0.2 μM, and 0.5 U of Takara LA-Taq DNA polymerase (Takara Bio, Kusatsu, Japan). The PCR cycle was 95°C for 4 min, followed by 31 cycles at 95°C denaturing for 1 min, 62°C annealing for 1 min, and 72°C extension for 1 min 45 s, and a final 10-min extension at 72°C. Primer sequences are described in Supplemental Table I.

Cell lysates were prepared using Nonidet P-40 lysis buffer with moderate agitation for 30 min at 4°C and cleared by centrifugation. Protein concentrations were determined using a bicinchoninic acid protein assay kit (Thermo Fisher Scientific). Lysates were separated using 12% acrylamide gel electrophoresis and transferred to nitrocellulose using semidry transfer (Thermo Fisher Scientific). Blocking was performed using LI-COR Intercept blocking buffer (LI-COR Biosciences, Lincoln, NE). Detection was performed using anti-Flag M2 HRP Ab (Sigma-Aldrich Canada Co., Oakville, ON, Canada) to detect Spi-C or anti-β-actin (C4) Alexa Fluor 680 Ab (Santa Cruz Biotechnology, Dallas, TX).

Spi-C was induced for 48 h in i38B9 cells with 1.5 μg/ml doxycycline, followed by crosslinking using 1% formaldehyde (MilliporeSigma), and halted using glycine. Crosslinked pellets of up to 1 × 10⁷ cells were flash frozen in liquid nitrogen prior to sonication. Frozen fixed pellets were resuspended in lysis buffer supplemented with HALT protease inhibitor (Thermo Fisher) and sonicated for 25 cycles using the Bioruptor UCD-300 (Diagenode, Sparta, NJ). Immunoprecipitation of 3xFLAG-Spi-C bound chromatin was performed using anti-FLAG M2 magnetic beads (MilliporeSigma). DNA was eluted from input chromatin or immunoprecipitated chromatin by heating. Eluted DNA was purified using the ChIP DNA clean and concentrator kit (Zymo Research, Irvine CA). qPCR analysis was performed using primers described in Supplemental Table I.

Statistical analysis was performed using Prism 9.4.1 (GraphPad Software, La Jolla, CA). Data are presented as mean ± SEM unless otherwise indicated. Statistical significance was determined using tests indicated in the figure legends. All data points in the figures in this study represent independent biological replicate experiments.

To study the role of Spi-C in Igκ transcription and rearrangement, we used the Abelson (Abl) kinase-transformed pre-B cell line 38B9 (30). Cell cycle arrest, Igκ V-J rearrangement, and B cell differentiation can be induced in 38B9 cells using the Abl kinase inhibitor imatinib (also known as STI571 or Gleevec [Novartis]) (31). We constructed a two-vector doxycycline-inducible system for ectopic expression of 3xFLAG-tagged murine Spi-C in 38B9 cells (Fig. 1A). With this system, Spi-C expression could be induced in inducible 38B9 (i38B9) cells up to 10-fold at 48 h of doxycycline treatment at a concentration of 1500 ng/ml (Fig. 1B). Induction of Spi-C did not affect cell cycle progression or induce Igκ recombination in i38B9 cells (Supplemental Figs. 1 and 2).

Next, 38B9 cells or i38B9 cells were treated with imatinib to induce Igκ V-J rearrangement (Fig. 1C). Imatinib treatment induced 38B9 cell cycle arrest by 48 h, as shown by flow cytometric analysis of a shift from high forward scatter to low forward light scatter, suggesting a reduction in cell size (Supplemental Fig. 1A). Spic was expressed at low levels in untreated 38B9 or i38B9 cells and was increased substantially by 0.5 μM imatinib treatment alone (Fig. 1D, 1E). Prior to imatinib treatment, both 38B9 and i38B9 cells had detectable V_k-J_k1 and V_k-J_k2 rearrangements but undetectable Vκ-Jκ4 and Vκ-Jκ5 rearrangements (Supplemental Fig. 2, Fig. 1F). Imatinib treatment for 24 h induced Vκ-Jκ4 rearrangements in both 38B9 and i38B9 cells (Supplemental Fig. 2B).

Next, the combined effects of inducing Spi-C using doxycycline and inducing Igκ rearrangement with imatinib were investigated. Spic mRNA levels were induced over 50-fold by combined doxycycline and imatinib treatment (Fig. 1D, right side). Strong induction of Spi-C protein by doxycycline and imatinib treatment was confirmed by anti-FLAG immunoblotting (Fig. 1E). Induction of Spi-C in i38B9 cells inhibited imatinib-induced Ig Vκ-Jκ rearrangement at the 48-h time point (Fig. 1F, right panels). This result suggests that ectopic expression of Spi-C negatively regulates Igκ rearrangement in pre-B cells.

Igκ rearrangement might be inhibited by Spi-C through repression of Igκ transcription, repression of RAG1/2 expression, or a combination of both. To determine the mechanism of inhibition of Igκ rearrangement by Spi-C, gene expression was examined using RT-qPCR. Vcam1 encodes vascular cell adhesion protein-1 and was previously shown to be directly activated by Spi-C (20). Rag1 and Rag2 encode the recombinase-activating 1 and 2 proteins, whereas Glk1 encodes an Igκ germline “sterile” transcript that correlates with locus accessibility (Fig. 2A) (6). Vcam1 was activated following imatinib treatment in the presence or absence of Dox (Fig. 2B). Glk1 but not Rag1 or Rag2 transcripts were inhibited by Spi-C at the 24-h time point (Fig. 2C–2E). However, Rag1, Rag2, and Glk1 mRNA transcript levels were all reduced at 48 h after imatinib treatment in doxycycline-induced i38B9 cells (Fig. 2C–2E). Therefore, ectopic expression of Spi-C inhibits Igκ germline transcription, Rag1, and Rag2 transcription in 38B9 cells.

We set out to determine the effect of Spic loss of function on gene expression in developing pre-B cells in mice. Spic^−/− mice are born at low frequency due to preimplantation lethality of undetermined cause (20, 37). We found that Spic^−/− mice on a C57BL/6 strain background were rarely generated, either at 3 wk of age or fetal 14.5 d postimplantation (Table I). However, by intercrossing F1 progeny of C57BL/6 Spic^+/− × 129Sv mice, we were able to obtain Spic^−/− mice with a frequency of 14.8% by 3 wk of age (Table I, Fig. 3A). This outcome suggests that Spic^−/− mice on a mixed C57BL/6 and 129.Sv strain background have reduced preimplantation lethality. This procedure allowed generation of Spic^−/− mice for further analysis.

There are substantial sex differences between frequencies of lymphocytes in various mouse strains (38). We found that female B6/129 F2 mice had increased frequencies of CD19⁺ B220⁺ B cells in the BM compared with male mice (Fig. 3B, 3C). There were no significant differences in total BM cell numbers between Spic^−/− and control mice (Fig. 3D, 3F). However, Spic^−/− mice had higher total numbers of CD19⁺ B220⁺ B cells in female mice compared with control animals (Fig. 3E). There were no significant differences between Spic^−/− and control mouse B cell frequencies in male mice (Fig. 3G). This result suggests a sex-specific effect of Spi-C on B cell development.

The frequency of BM B cell development stages was determined using the Hardy scheme in which B cell development is separated into fractions A–F (39). No significant differences were detected between Spic^−/− female mice and Spic^+/+ or Spic^+/− female control mice (Fig. 4A, 4B) or between male mice of the same genotypes (data not shown). Next, Igκ V-J rearrangements were determined. RT-PCR analysis of RNA prepared from enriched fraction D BM pre-B cells showed similar frequencies of Vκ-Jκ rearrangements in Spic^−/− mice and control mice (Fig. 4C). To perform quantitative analysis, Vκ-Jκ5 PCR products were sequenced and used to design qPCR primers (Fig. 4D). RT-qPCR analysis of Vκ-Jκ5 rearrangements revealed that mRNA levels were higher in fraction D pre-B cells than in fraction C, E, or F (Fig. 4E). However, no difference was found between Spic^−/− mice and control mice (Fig. 4E). In summary, these results suggest that loss of Spi-C function does not impair early B cell development.

The low frequency of BM B cells in male B6/129 F2 mice Spic^−/− mice made enrichment of B cells from male mice by cell sorting technically unfeasible. We therefore enriched total BM CD19⁺ B220⁺ B cells from female B6/129 F2 Spic^−/− mice, as well as female littermate Spic^+/+ or Spic^+/− mice as controls. RNA was prepared from enriched B cells for RT-qPCR analysis of gene expression. Spic mRNA transcripts were undetectable in Spic^−/− BM B cells, as expected (Fig. 5A). Rag1 and Glk1 were expressed at significantly higher levels in B cells from Spic^−/− female mice than those from Spic^+/+ or Spic^+/− female mice (Fig. 5B, 5D). Rag2 trended toward higher levels of expression in Spic^−/− female mice than in Spic^+/+ or Spic^+/− female mice (Fig. 5C). These results suggested that Spi-C controls levels of Rag1 and Glk1 transcripts in BM B cells.

Next, fractions C, D, E, and F cells were enriched by cell sorting from Spic^−/− female mice and Spic^+/+ or Spic^+/− female mice, and RNA was prepared to determine gene expression using RT-qPCR. Rag1 was expressed at higher levels in Spic^−/− fraction D cells and fraction F cells than in controls (Fig. 5E). Rag2 and Glk1 were expressed at higher levels in Spic^−/− fraction D cells than in control fraction D cells. The majority of fraction D cells represent pre-B cells that have stopped proliferating in response to IL-7 and have initiated rearrangement of Igκ but do not yet express a BCR (39). In agreement with this observation, we found that fraction D cells expressed higher levels of Rag and Glk1 mRNA transcripts than fraction C, E, or F cells (Fig. 6). For Rag1, there was a larger difference between fractions C and D in Spic^−/− than in wild type (9-fold compared with 3-fold; Fig. 6A and 6D) or between fractions E/F to D (82-fold compared with 70-fold; Fig. 6A and 6D). In summary, our results suggest that Spi-C represses Rag and Glk1 transcription starting at the pre-B (fraction D) cell stage.

PU.1 and Spi-B are highly related to Spi-C in the ETS transcription factor family. We previously showed that PU.1 is an activator of Igκ transcription and rearrangement during B cell development (24). To further explore this observation, we made use of the i660BM pre-B cell line in which PU.1 is doxycycline inducible (32). Doxycycline induction of PU.1 in i660BM cells for 48 h (Fig. 7A) increased Rag1 and Rag2 mRNA transcript levels (Fig 7B, 7C). This result suggests that PU.1 is an activator of Rag transcription. Next, we looked at mice in which PU.1 is inducibly deleted under control of the Mb1 (Cd79a) gene starting at the pro-B stage (Mb1-CreΔPB mice) (24). These mice are also germline null for the Spib gene. We found that Rag1 and Glk1 mRNA transcripts were decreased in fraction D cells enriched from Mb1-CreΔPB mice compared with Spib^−/− control mice (Fig. 7D–7G). These data suggest that PU.1 functions as an activator of Igκ and Rag1 transcription and Igκ rearrangement. Taken together with the previous results, these data suggest that Spi-C and PU.1 act in an opposite manner to regulate Igκ and Rag1 transcription and Igκ rearrangement starting at the small pre-B cell stage.

To determine if Spi-C regulates Rag transcription directly by interaction with a binding site in a regulatory region, we performed anti-Spi-C ChIP. 3xFLAG-Spi-C was induced by doxycycline in i38B9 cells for 48 h, after which cells were formaldehyde fixed and chromatin prepared for immunoprecipitation. ChIP was performed using anti-FLAG Ab as described in Materials and Methods. First, qPCR analysis was performed on a previously identified Spi-C binding site (23) in the first intron of the Bach2 gene (Fig. 8A). Using qPCR analysis, Spi-C binding at Bach2 region of interest 1 was increased 40-fold compared with binding at a negative control region (Fig. 8B). Next, qPCR analysis was performed to determine Spi-C binding at a previously identified PU.1 binding site within the Rag locus (24) (Fig. 8C). Four independent biological replicate experiments found 8–45-fold increased binding of Spi-C at the Rag peak compared with a negative control region in the same locus (Fig. 8D–8G). In summary, Spi-C and PU.1 interact with a binding site located directly upstream of the Rag1 gene, suggesting that Spi-C and PU.1 regulate Rag transcription directly.

In this study, we investigated the ability of the ETS transcription factor Spi-C to negatively regulate Igκ L chain recombination. Using an inducible expression system in a pre-B cell line, we found that Spi-C negatively regulated Igκ rearrangement, Igκ transcript levels, and Rag1 transcript levels. By analyzing BM-derived B cells from Spic^−/− mice, we found that Igκ and Rag1 transcript levels were negatively regulated by Spi-C in small pre-B cells. In contrast, we found that Igκ and Rag1 transcript levels were activated by PU.1 in small pre-B cells. Using ChIP analysis, we identified a binding site for PU.1 and Spi-C located 5′ of the Rag1 promoter. We conclude that Spi-C and PU.1 counterregulate Igκ L chain recombination and Rag1 transcription in small pre-B cells.

DNA DSBs induced at the Igκ L chain locus in small pre-B cells have been demonstrated to negatively feed back on Rag and Igκ transcription to promote allelic exclusion at the Igκ locus (16). The signaling pathway induced by DSBs to promote allelic exclusion requires ATM kinase and the Nemo protein (16). This signaling pathway activates several genes, including Spic (18). Spi-C is highly related to PU.1 and Spi-B and can interact with and/or compete for similar binding sites in the genome of B cells and macrophages (23, 29, 40). Spi-C can repress Igκ recombination and cell division by interacting with BCLAF1 to displace PU.1, resulting in repression of the Syk, Blnk, and Igκ genes (18, 19) Although Rag1 transcription was shown to be negatively regulated by DSBs, it has not been clear what factors mediate this transcriptional repression (41). Our study now demonstrates that Spi-C is a direct negative regulator of Rag1 transcription in pre-B cells.

PU.1 has been shown in multiple studies to function primarily as an activator of transcription (42). In addition, PU.1 functions as a pioneer protein, having the ability to interact with binding sites in closed chromatin and to recruit histone acetyltransferases to open chromatin sites (43). In our previous studies and in the study by Soodgupta et al. (19), Spi-C appeared to function as a negative regulator of gene transcription primarily by displacing PU.1 from binding sites. ChIP studies suggest that most Spi-C binding sites are also PU.1 binding sites, whereas there are fewer unique Spi-C binding sites (23). It remains to be determined whether Spi-C interaction with target sites requires prior PU.1 binding or pioneer activity.

The Igκ 3′ enhancer contains an important ETS transcription factor binding site that regulates B cell versus T cell specificity of Vκ-Jκ joining (7). PU.1 was shown to interact with this site to promote Igκ transcription and accessibility (44, 45). Because of functional redundancy and complementarity with PU.1, Spi-B very likely also regulates Igκ transcription through this site (24). Spi-C was shown to displace PU.1 interaction with binding sites in the Igκ locus (19). Therefore, Spi-C likely negatively regulates the Glk1 mRNA transcript by direct interaction with PU.1 interaction sites in the Igκ locus.

In all experiments performed in this study, Rag1 was affected to a greater degree than Rag2 by Spi-C ectopic expression or deletion. We speculate that this is because of the proximity of the Rag1 transcription start site to the binding site that we identified for PU.1/Spi-C. Mechanisms of transcriptional regulation of the Rag locus have not been well studied, with only the identification of the ERag enhancer reported for many years (14). However, a more detailed study of Rag transcriptional regulation was reported in 2020 by Miyazaki et al. using assay for transposase-accessible chromatin with sequencing to identify additional regulatory regions named R-TEn, R2B, R1pro, and R1B (15). R1pro represents the Rag1 promoter region and was shown to interact with multiple transcription factors, including E2A, Pax5, ETS1, Ikaros, and IRF4. We have now shown that Spi-C and PU.1 also interact with the R1pro region, adding to the evidence that this is an important regulatory region for the Rag1 gene.

We speculate that in small B cells, PU.1 activates Igκ and Rag transcription using its pioneer activity. PU.1 is a strong transcriptional activator of Igκ (24) and, as we show here, Rag1 to promote Igκ recombination. Following RAG1/2-mediated DNA cleavage at one Igκ allele, The Spic gene is activated, resulting in Spi-C interaction with BCLAF1 to displace PU.1 from target genes in small pre-B cells. Spi-C expression may persist into the transitional 1 stage to modulate BCR signaling. In this manner, PU.1 and Spi-C are proposed to function as counterregulators of Igκ allelic exclusion and are potentially involved in regulation of BCR signaling and receptor editing. In summary, our experiments support the idea that Spi-C and PU.1 counterregulate Igκ transcription and Rag1 transcription to effect Igκ recombination in small pre-B cells.

The authors have no financial conflicts of interest.

We thank Dr. Malay Haldar (Perelman School of Medicine, University of Pennsylvania) for advice regarding breeding and genotyping of Spic^−/− mice. We thank Kristin Chadwick and the London Regional Flow Cytometry Facility for assistance with cell sorting.