The immunoglobulin κ light chain intronic enhancer (iEκ) activates κ rearrangement and is required to maintain the earlier or more efficient rearrangement of κ versus lambda (λ). To understand the mechanism of how iEκ regulates κ rearrangement, we employed homologous recombination to mutate individual functional motifs within iEκ in the endogenous κ locus, including the NF-κB binding site (κB), as well as κE1, κE2, and κE3 E boxes. Analysis of the impacts of these mutations revealed that κE2 and to a lesser extent κE1, but not κE3, were important for activating κ rearrangement. Surprisingly, mutation of the κB site had no apparent effect on κ rearrangement. Comparable to the deletion of the entire iEκ, simultaneous mutation of κE1 and κE2 reduces the efficiency of κ rearrangement much more dramatically than either κE1 or κE2 mutation alone. Because E2A family proteins are the only known factors that bind to these E boxes, these findings provide unambiguous evidence that E2A is a key regulator of κ rearrangement.
Each B lymphocyte generates a unique set of immunoglobulin heavy (IgH) and light (IgL) chain genes through the somatic rearrangement of V, D, and J gene segments. To ensure the monospecificity of each B cell, V(D)J recombination is regulated in such a lineage- and stage-specific manner that IgH and IgL rearrangement occur at distinct stages of development (1). Because V(D)J rearrangement of all Ig loci involves the same recombination machinery, cis elements within these antigen receptor loci must play decisive roles in regulating the accessibility of each locus to V(D)J recombination machinery (2). Within the κ locus, several cis-acting elements were initially identified by their ability to activate transcription of κ reporter constructs in B cell lines, including two enhancers, one within the Jκ-Cκ intron (iEκ) and one 3′ of Cκ (3′Eκ) (3–8). More recently, a putative third enhancer 8 kb downstream of 3′Eκ (Ed) was also discovered (9). The deletion of the intronic enhancer and matrix attachment region (MiEκ) led to decreased κ rearrangement and a lower κ:λ ratio. More importantly, MiEκ is required for the earlier or more efficient rearrangement of κ versus λ loci. (10, 11). The matrix attachment region (MAR) within MiEκ does not contribute to these activities since the deletion of the κ MAR alone does not have an inhibitory effect on the overall level of κ rearrangement (12). The deletion of the 3′ enhancer (3′Eκ) results in a similar, though less dramatic decrease in the ratio of κ:λ B cells and κ rearrangement (13). In addition, 3′Eκ appears to play an important role in activating κ transcription in mature B cells (11, 13). Although the loss of either enhancer alone does not eliminate κ rearrangement, deletion of both enhancers from endogenous κ loci results in a complete block of κ rearrangement (11). Therefore, MiEκ and 3′Eκ together are the necessary elements for κ rearrangement.
Several functional motifs within the intronic enhancer were identified through a battery of biochemical and cell line transfection studies. One such functional motif is the NF-κB binding site, denoted κB (14). The potential role of the κB site in κ rearrangement was suggested by the finding that LPS could induce κ germline transcription and rearrangement through an NF-κB–dependent pathway (15–18). iEκ also contains a class of protein-binding motifs referred to as E boxes, which are also identified in enhancers of other antigen receptor genes (19, 20). iEκ contains three E boxes, labeled κE1, κE2, and κE3. κE2 was found to bind the E2A gene products E12 and E47, which are required for B cell development (21–23). E2A gene products can also bind to κE1 but not to κE3 (Murre, C., personal communication). E2A can induce germline transcription and κ rearrangement when cotransfected with the RAG genes in a nonlymphoid cell line (24). However, the role of E2A in the regulation of κ rearrangement in the physiological context remains unclear.
To delineate the mechanism through which iEκ activates κ rearrangement, we employed homologous recombination and Cre/loxP-mediated deletion to produce independently targeted deletions of four functional motifs within the endogenous κ intronic enhancer. Analysis of the effects of these mutations on κ rearrangement revealed that the E2A-binding E boxes, but not the κB site, were quantitatively important for the activation of κ rearrangement. In addition, the simultaneous mutation of both κE1 and κE2 sites impacts κ rearrangement as severely as the deletion of the entire iEκ, indicating that E2A-dependent pathways are the major mediator of iEκ's activity in activating κ rearrangement.
Materials And Methods
Generation of mκEX Embryonic Stem Cells.
The 740-bp κ intronic enhancer was cloned into pBluescript. The four sites were deleted and replaced with diagnostic EcoRI sites independently by site-directed mutagenesis. The primers used to replace each site are as follows: mκB, 5′-TGGCATCTCAACGAATTCAGAGCCATCTGG-3′; mκE1, 5′-GACTTTCCGAGAGAATTCAGTTGCTTAAGA-3′; mκE2, 5′-CAGTTCCTCCGAATTCGATTACAG-3′; and mκE3, 5′-TGGCTAAAAATTGA-ATTCAAACCATTAGAC-3′.
The mκE1/2 double mutations were generated by PCR-mediated mutagenesis. The κE1 and κE2 sequences are as follows, with the underlined C nucleotides mutated to A: κE1, CATCTGGC; κE2, CAGGTGGC. The MiEκ deletion (mκD) was generated as described (10). All mκEX clones were inserted into the mκD targeting construct. 20 μg of each targeting construct was linearized with PvuI and transfected into embryonic stem (ES) cells. Transfectants were selected with G418 (300 μg/ml) and gancyclovir (1 μM). Homologous recombination was screened by Southern blot, and positive clones were then transiently transfected with 10–20 μg of a plasmid containing the Cre gene to remove the PGK-neor gene as described (10). PGK-neor–deleted ES clones were subcloned, confirmed by Southern blot, and analyzed by RAG-2–deficient blastocyst complementation as described (10). mκE1/2 ES cells were microinjected into blastocysts before Cre transfection and then bred to Cre-expressing mice to generate mκE1/2 knock-in mice.
Quantitative Analysis of Vκ Jκ Rearrangements.
To detect Vκ Jκ rearrangements, genomic DNA from mature B cells (CD19+/IgM+) was purified and amplified by PCR using the degenerate Vκ primer (VκD) (15) and a primer within the kappa intronic enhancer (K1). PCR reactions were conducted under the following conditions: 95°C for 5 min, followed by 26 cycles of 94°C for 1 min, 60°C for 1 min, and 72°C for 3.5 min, followed by an additional 10 min at 72°C. To control for the total amount of B cell DNA used in each PCR reaction, VDJH rearrangement was also analyzed using a primer for the J558 family (J558) and a primer that can bind to all four JH gene segments (JH(1+4)). The intensity of each PCR product was quantified with a Storm phosphorimager. Primers used are as follows: VκD, 5′-GGCTGCAGSTTCAGTGGCAGTGGRTCWGGRAC-3′; K1, 5′-CCAT-GACTTTTGCTGGCTGTAGATTTTACCTC-3′; J558, 5′-CTTCAGTGAAGCTGTCCTGCAAGGCTT-3′; and JH(1+4), 5′-CAGCTTACCTGAGGAGACGGTGA-3′. Hybridomas were generated and analyzed as described (10). The University of California, San Diego Animal Subject Committee approved all experiments that involved mice.
Quantitative Analysis of κ Transcription by Real-Time PCR.
B cells from the spleens of WT and homozygous mκE1/2 mutant mice were purified by magnetic-activated cell sorting using αCD43 microbeads according to the manufacturer's protocol (Miltenyi Biotec). In mκE1/2 mutants, λ+ B cells were removed by costaining with 0.5 μg/million cells biotin-conjugated α-mouse Igλ antibodies (clone R26-46; BD Biosciences) and then, after washing, staining with 0.5 μg/million cells α-biotin microbeads (Miltenyi Biotec). RNA from 1 million sorted cells was purified using the RNAeasy kit (QIAGEN) combined with on-column DNase digestion (QIAGEN) and converted into cDNA using the Superscript First Strand System (Invitrogen) according to the manufacturers' protocols. Primers for amplifying the constant regions of μ (Cμ1 and Cμ2) and κ (Cκ1 and Cκ2) were designed using Primer Express software (Applied Biosystems) and used at concentrations of 200 and 400 nM, respectively. Real-time PCR reactions were performed using the SYBR Green PCR Master Mix (Applied Biosystems) in an ABI Prism 7000 Sequence Detection System according to the manufacturer's protocols. Relative transcription levels were calculated using ABI Prism 7000 SDS software using the standard curve method. Primer sequences are as follows: Cμ1, 5′-ACACCTGCCGTGTGGATCA-3′; Cμ2, 5′-GAGGAAGATGTCGGCAAAGG-3′; Cκ1, 5′-CAACTGTATCCATCTTCCCACCA-3′; and Cκ2, 5′-GGCACCTCCAGATGTTAACTGCT-3′.
Results And Discussion
Generation of mκEX ES Cells.
The 740-bp HindIII–AflII region spanning the κ intronic enhancer and associated matrix attachment region (MiEκ) was cloned into pBluescript. The NF-κB, κE1, κE2, or κE3 sites were each individually replaced with a diagnostic EcoRI restriction site by site–directed mutagenesis (Fig. 1
F). These iEκ mutants, collectively referred to as mκEX and individually as mκB, mκE1, mκE2, and mκE3, were independently inserted into the targeting construct used previously to delete MiEκ, so that homologous recombination between the targeting vector and the endogenous locus would replace the endogenous MiEκ with mκEX (10) (Fig. 1 B). Homologous recombination events were screened by Southern blotting with EcoRI digestion and hybridization to probe A (Fig. 1, A and C; not depicted). The PGK-neor gene was excised from the targeted allele by transient expression of the Cre gene in positive ES clones (Fig. 1 D). ES clones with the PGK-neor deleted were subcloned and confirmed by Southern blot with EcoRI digestion and hybridization to probe A (Fig. 1 E, lanes 2–5).
Analysis of Vκ Jκ Recombination in mκEX B Cells.
The effects of mκEX mutations on κ rearrangement in B cells were assayed by recombination activating gene-2–deficient (RAG-2−/−) blastocyst complementation as described (10). Since heterozygous mutant ES cells were used, this enabled us to compare the rearrangement frequency between the WT and mutant κ alleles. B cells were sorted from mκEX–RAG-2−/− mice and analyzed for Vκ Jκ recombination using a quantitative PCR assay as previously described (11, 12). This assay could detect rearrangement to all four functional Jκ gene segments and distinguish PCR products derived from the WT allele from those from the mutant allele due to an additional 150 bp within the mutant allele (Fig. 2
A). Therefore, the efficiency of Vκ Jκ1-5 rearrangement of the mutant κ allele could be compared with that of the WT allele in the same PCR reaction (Fig. 2 C). Our analysis indicated that neither the NF-κB (mκB) nor κE3 (mκE3) site mutations had any apparent effect on κ rearrangement (Fig. 2 B). However, mutation of the κE1 (mκE1) or κE2 (mκE2) site reduced the rearrangement efficiency of the mutant κ allele. The mutation of the κE2 site had a more inhibitory effect on κ rearrangement (Fig. 2 B, lane 4). In this context, the ratio of the rearrangement frequency of Vκ to Jκ1, Jκ2, Jκ4, and Jκ5 gene segments of the mutant allele versus that of the WT allele was 52.1, 31.9, 26.2, and 27.9%, respectively (Fig. 2 C). In mκE1 B cells, the rearrangement efficiency of the mutant allele was approximately twofold reduced for each of the four possible Vκ Jκ rearrangements compared with that of the WT allele (Fig. 2 B, lane 3, and C).
Analysis of Vκ Jκ Recombination in mκEX κ+ Hybridomas.
To further confirm the impact of mκEX mutations on κ rearrangement efficiency and analyze the rearrangement frequency of mutant and WT alleles in individual B cells, we generated hybridomas from the spleen cells of all four mκEX–RAG-2−/− mice as described (10). Genomic DNA derived from κ+ hybridomas was analyzed by Southern blotting to detect Vκ Jκ rearrangements at the WT and mutant alleles. Consistent with the data from the PCR analysis, we observed the most dramatic decrease in the rearrangement frequency of the mutant allele in mκE2 hybridomas (Fig. 3
B). In this context, of the 67 κ+ hybridomas analyzed 48 had rearrangements on only the WT allele, 10 had rearrangements on only the mutant allele, and 9 had rearrangements on both alleles (Fig. 3 C). The ratio of the rearrangement frequency of the WT allele versus that of the mutant allele was ∼3:1. A less dramatic decrease was observed in mκE1 hybridomas (Fig. 3 C). Consistent with the data derived from the PCR analysis, little difference was detected in the rearrangement frequency of the WT and mutant alleles in hybridomas derived from the mκB and mκE3 B cells (Fig. 3, A and C).
To determine the contribution of the κE1 and κE2 sites to the full activity of iEκ, hybridomas were also generated from B cells in which the entire intronic enhancer was deleted from one allele (mκD). Only 3 of 43 κ+ mκD hybridomas analyzed harbored a Vκ Jκ rearrangement on the mutant allele, whereas the WT allele was rearranged in all 43 hybridomas (Fig. 3 C). Therefore, the reduction of the rearrangement efficiency caused by the deletion of the entire iEκ was much more dramatic than that caused by κE1 or κE2 mutations alone. In addition, these data indicate that the rearrangement of the mutant allele occurs only after the rearrangement of the WT allele in mκD B cells.
Critical Roles of κE1 and κE2 in Mediating iEκ's Function in Activating κ Rearrangement.
Although the reduction of κ rearrangement caused by the κE1 or κE2 mutations alone was significantly less than that caused by the deletion of the entire enhancer, it is possible that these two sites have redundant functions since both sites can be bound by E2A family proteins. To test this hypothesis, we mutated both κE1 and κE2 sites simultaneously through homologous recombination and assayed for effects on κ rearrangement. The mutation introduced at both sites was a single nucleotide (C to A) mutation that destroyed the canonical basic helix-loop-helix binding site (CANNTG) (Fig. 4
A). To determine the effects of the κE1/2 mutation on κ rearrangement, we used the same PCR assay described in Fig. 2 to analyze κ rearrangement of WT and mutant alleles in κ+ mκE1/2 B cells purified from the spleens of RAG−/− chimeric mice or heterozygous mutant mice. The reduction in κ rearrangement caused by the mκE1/2 mutation was much more severe than that caused by either κE1 or κE2 mutations alone (Fig. 4, B and C). Analysis of κ+ hybridomas derived from mκE1/2 B cells revealed a more than 10-fold reduction in the rearrangement efficiency of the mutant allele, comparable to that caused by the deletion of the entire enhancer (Fig. 4 D). In addition, similar to findings in mκD hybridomas, the mκE1/2 allele only rearranged after the WT allele had already rearranged.
To rule out the possibility that the greatly reduced rearrangement efficiency of mκE1/2 alleles observed in splenic B cells is due to the impaired expression of rearranged κ at mκE1/2 alleles during the transition from pre–B cells to immature B cells, we examined the rearrangement frequency of WT and mutant alleles in pre–B cells (B220+/CD43−/IgM−) sorted from the BM of mκE1/2 mice (Fig. 4 E). Similar to that observed in splenic B cells, the rearrangement frequency of the mκE1/2 allele was significantly reduced in pre–B cells compared with the WT allele. To further determine whether the mκE1/2 mutation affects κ expression in B cells, we analyzed κ expression in κ1 B cells purified from the spleens of WT mice and homozygous mκE1/2 mutant mice. Similar levels of κ mRNA were detected in WT and mκE1/2 κ1 B cells, indicating that the mκE1/2 mutation had no apparent effect on κ transcription (Fig. 4 F). Therefore, κE1 and κE2 play synergistic roles in iEκ's function in activating κ rearrangement.
Sequential rearrangement of the IgH and IgL chain genes is critical to ensure B cell monospecificity. The κ intronic enhancer plays an important role in this process (10, 11). By analyzing the functional motifs within iEκ in vivo, we demonstrated that the κB site mutation had no apparent impact on κ rearrangement. The κB site is bound by NF-κB family members and was found to play an important role in activating κ rearrangement of recombination substrates in cell line transfection and transgenic studies (16, 25). However, the findings that the impacts of κE1 and κE2 double mutation on κ rearrangement are very similar to the deletion of the entire iEκ rule out any important roles of other protein-binding motifs within iEκ, such as the NF-κB binding site, in activating κ rearrangement. Based on the findings that iEκ and 3′Eκ are the essential elements activating κ rearrangement and that there is no NF-κB binding site within 3′Eκ, it remains unclear how NF-κB family transcription factors might activate κ rearrangement. Based on the findings that NF-κB is activated during B cell activation induced by cross-linking of the antigen receptor (for review see reference 26), the potential roles of the κB site in receptor editing and somatic hypermutation of κ should be examined.
Our findings provide unambiguous evidence that the κE1 and κE2 sites play critical roles in activating κ rearrangement. E2A family transcription factors are the primary transcription factors binding to the κE1/2 sites, and when coexpressed with the RAG proteins, are able to activate κ rearrangement in a non–B cell line (24). Therefore, our findings provide a mechanism for how E2A family transcription factors might function in activating κ rearrangement. Based on the findings that E2A can interact with histone acetyltransferases such as p300/CBP and the SAGA complex (27–29), κE1/2 might activate κ rearrangement by increasing histone acetylation of the κ locus.
The authors thank David Baltimore, Cornelis Murre, and Ann Feeney for helpful discussion and critical reading of the manuscript.
This work was supported by a National Institutes of Health grant (AI44838) to Y. Xu.
The authors have no conflicting financial interests.