Somatic hypermutation and gene conversion are two closely related processes that increase the diversity of the primary Ig repertoire. Both processes are initiated by the activation-induced cytidine deaminase that converts cytosine residues to uracils in a transcription-dependent manner; these lesions are subsequently fixed in the genome by direct replication and error-prone DNA repair. Two alternative mechanisms were proposed to explain why this mutagenic activity is targeted almost exclusively to Ig loci: 1) specific cis-acting DNA sequences; or 2) very high levels of Ig gene transcription. In this study we now identify a novel 3′ regulatory region in the chicken Ig light chain gene containing not only a classical transcriptional enhancer but also cis-acting DNA elements essential for targeting activation-induced cytidine deaminase-mediated sequence diversification to this locus.

During B cell development, functional Ig genes are assembled from individual gene segments by V(D)J recombination. Subsequently, B cells use distinct cellular programs, in particular somatic hypermutation (SHM)3 and gene conversion (GCV), to actively mutate the DNA of their Ig genes to improve Ag recognition. Both processes, as well as the related class switch recombination, share an essential initiating factor, the activation-induced cytidine deaminase (AID) (1, 2, 3, 4). This protein acts on individual cytosine residues within ssDNA that is transiently generated during transcription of the Ig genes (reviewed in Refs. 5 and 6). The resulting U:G mismatch gets resolved by direct replication and DNA repair mechanisms involving uracil DNA glycosylase, general mismatch repair factors, and error-prone DNA polymerases (reviewed in Refs. 6 and 7). While SHM creates individual “nontemplated” point mutations and occurs in all jawed vertebrates, GCV introduces single or multiple “templated” nucleotide changes and has only been observed in selected vertebrates, including chicken and rabbits. Although the end products differ, it is thought that the key molecular principles that govern their AID-dependent initiation phase are the same (reviewed in Refs. 5 and 6).

SHM and GCV are highly limited to Ig genes, but at very low frequencies non-Ig genes are also targeted and such events, as in the case of Bcl6, can contribute to B cell lymphoma formation (reviewed in Ref. 8). The molecular principles enforcing this tight restriction are unknown. Transgenic mice harboring murine Igκ mini loci suggested that the regulatory DNA elements controlling Igκ transcription, the κ intronic enhancer (iEκ) and the 3′κ enhancer (3′Eκ), serve a role in the targeting of SHM (9, 10, 11). However, neither of them was found to be essential for SHM when they were deleted individually from the endogenous Igκ locus (12). Hence, these and other approaches to identify cis-acting DNA elements (also referred to as targeting elements) recruiting AID-dependent mutation activities to Ig loci have yet to yield definitive results (reviewed in Ref. 13). Early studies in mouse models indicated that any active promoter is able to support SHM of Ig genes (9, 14, 15). The observation that AID was able to mutate highly transcribed non-Ig transgenes in AID-expressing cell lines contributed to the alternative hypothesis that the high levels of transcription themselves are the major determinants for rendering a gene to be subjected to SHM (reviewed in Ref. 13). Recent findings in chicken DT40 cells, however, indicated that not only the high activity of promoters driving Ig gene transcription but also the very nature of such promoters themselves contribute significantly to the rate of GCV/SHM (16). To resolve the controversy between the two models, we sought to identify cis-acting DNA elements that target GCV to the Ig light chain (IgL) locus in chicken DT40 cells. Using a systematic deletion strategy, we identified a novel regulatory region downstream of the only known IgL enhancer. Importantly, this region not only promoted high levels of transcription but harbors a cis-acting DNA element that targets GCV and SHM to the IgL locus.

All primer sequences will be provided upon request.

The IgL locus components of all targeting constructs were generated by PCR amplification of respective genomic DNA fragments using Phusion polymerase (New England Biolabs). Briefly, left arm (upstream) and right arm (downstream) fragments were cloned into the NotI and XhoI sites of pBluescriptSK+ (Stratagene) with a central BamHI site between the arms. The puromycin cassette of pLoxPuro (17) was inserted as a BamHI fragment to obtain the final targeting constructs. For the ΔMES construct (see Fig. 1 and legend), the SV40 enhancer was amplified from pECFP-1 and ligated as a BglII/BamHI fragment into the central BamHI site of ΔME6K before insertion of the puromycin selection cassette.

FIGURE 1.

The structure of the IgL locus in DT40 cells and in derivatives thereof. The leader (L), VJ, and constant (C) region exons are depicted as open boxes, the putative MAR (M) and enhancer (E) as red ovals, and the SV40 enhancer (SV40E) as a purple oval. The position and size of each of these elements are drawn to scale. The polyadenylation site (Poly(A)) is represented by a green circle, and the loxP sites that are retained in the genome after Cre-mediated deletion of the puromycin selection cassette are depicted as filled triangles. The areas that are deleted in the DT40 derivatives are shown in lighter colors (gray and pink, respectively). The location of the newly discovered 3′ regulatory region, 3′RR, is indicated with a set of brackets, and its sequence corresponds to nt113822–117952 in GenBank entry AC171016. WT, Wild type.

FIGURE 1.

The structure of the IgL locus in DT40 cells and in derivatives thereof. The leader (L), VJ, and constant (C) region exons are depicted as open boxes, the putative MAR (M) and enhancer (E) as red ovals, and the SV40 enhancer (SV40E) as a purple oval. The position and size of each of these elements are drawn to scale. The polyadenylation site (Poly(A)) is represented by a green circle, and the loxP sites that are retained in the genome after Cre-mediated deletion of the puromycin selection cassette are depicted as filled triangles. The areas that are deleted in the DT40 derivatives are shown in lighter colors (gray and pink, respectively). The location of the newly discovered 3′ regulatory region, 3′RR, is indicated with a set of brackets, and its sequence corresponds to nt113822–117952 in GenBank entry AC171016. WT, Wild type.

Close modal

All DT40 Cl18 clones were grown at 41°C and 5% CO2 in RPMI 1640 (Mediatech) supplemented with 10% FBS (Invitrogen Life Technologies), 1% chicken serum (Invitrogen Life Technologies), 10 mM HEPES, 2 mM l-glutamine, and penicillin/streptomycin. Transfections were performed using 30 μg of linearized plasmid DNA with a Bio-Rad Gene Pulser at 580 volts, 25 microfarads, and infinite resistance. Stable transfectants were selected with 0.5 μg/ml puromycin, and the genotypes of clones with targeted integrations were confirmed by Southern blot analysis. The probes were amplified by PCR from genomic DNA, gel purified, and body labeled with [α-32P]dCTP using the NEB blot kit (New England Biolabs). Cre-mediated deletion of puromycin resistance cassettes was done using a cell-permeable Tat-NLS-Cre fusion protein (where NLS is “nuclear localization signal”) as described previously (16) and confirmed by Southern blotting.

Gene conversion and mutation analysis was performed as described previously (16). Briefly, single cell clones were obtained by limiting dilution and cultured continuously for 4 wk (corresponding to ∼84 generations). Surface IgM reversion was determined by staining cell aliquots with PE-labeled anti-IgM Ab (Southern Biotechnology), and VJ region sequences were amplified and sequence from the genomic DNA of unsorted cultures.

Total RNA was isolated using RNABee (Tel-Test) according to the manufacturer’s protocol, and transcript levels were determined by Northern blotting using probes for the constant region of IgL and GAPDH. Signals were detected using a Storm PhosphorImager (Amersham Biosciences) and quantified using ImageQuant (Amersham Biosciences).

To identify and locate cis-acting elements that recruit AID-mediated sequence diversification to the IgL locus in chicken DT40 cells, we systematically removed noncoding DNA segments from this locus by gene targeting (Fig. 1). All targeted manipulations of the IgL locus, the initial gene targeting, and the subsequent Cre-mediated removal of the puromycin-resistance cassette were confirmed by Southern blotting (data not shown). The effect of each deletion on GCV was determined by tracking the appearance of IgM+ cells from individual IgM single cell clones over 4 wk of culture, and mutation event frequencies were determined by sequencing the IgL genes at the end of this period. To minimize experimental errors, we adhered to the following five precautions for each of our IgL deletion mutants. First, to rule out any effects caused by clone to clone variations, all experiments were performed using two clones that had undergone the desired gene targeting event obtained from independent transfections. Second, puromycin-resistant clones with a random integration of the targeting constructs (random integrants) were cultured in parallel to targeted clones to ensure that any observed phenomenon is indeed caused by the specific modification of the IgL locus. Third, the absence of additional randomly integrated copies of the targeting vectors was confirmed for each target-integrated clone by Southern blotting, as such additional exogenous sequences incorporated elsewhere in the genome might interfere with the GCV/SHM of the endogenous IgL locus. Fourth, we relied on two independent methods, surface IgM reversion (3) and IgL gene sequencing, to determine GCV and SHM activity in each clone. Fifth, to rule out the possibility that the Cre-mediated deletion would lead to a transient susceptibility to GCV/SHM of an otherwise protected IgL locus (18), all clones were grown for at least 10 days before the 4-wk culture experiments.

We first deleted the known enhancer (ΔE) and the VJ-C intron (ΔM) containing a predicted matrix attachment region (MAR) (Fig. 1). The absence of the entire VJ-C intron (ΔM), including the MAR, resulted in a 3-fold increase of the steady-state IgL transcripts levels (Fig. 2,A), similar to the result of an earlier deletion of only the MAR (19). In contrast, deletion of the 467-bp core IgL enhancer showed a modest reduction in transcription (Fig. 2,A). Although one earlier report concluded that the enhancer is essential for IgL transcription (19), our observation is in better agreement with a more recent report showing that this enhancer plays only a minor role in IgL transcription in DT40 cells (16). The frequency of GCV and SHM was not dramatically altered in the ΔM and ΔE cell lines, consistent with previous reports, and small differences in the magnitude of effects are likely to be caused by differences between the targeting construct used (16, 19) (Figs. 2,B and 3A).

FIGURE 2.

IgL transcription and mutation frequencies in mutant DT40 cell with deletions in the IgL gene locus. A, Two Northern blots with independently derived RNA samples from each cell clone were hybridized with probes for the IgL gene and GAPDH as a loading control. The signals were quantified, and all values were first normalized to the GAPDH signal in each lane and subsequently normalized to the IgL level in wild-type (WT) DT40 cells. The bars represents the average of four independent experiments (two per genotype) except for ΔMES1.1 and ΔMES34.2, where the average of two values per clone is shown. Sorted IgM cells were used for the RNA preparation to avoid issues of differential RNA stability due to nonsense-mediated decay. B, The frequency of mutation events (nontemplated and templated combined) was determined by sequencing the VJ region of the IgL gene from two individual clones of each genotype after 4 wk of continuous culture. A minimum of 77 sequences was obtained for each genotype. The dashed line indicates the background level of 5 × 10−5 events/bp for our experimental design as determined by sequencing the IgL gene in AID-deficient DT40 clones.

FIGURE 2.

IgL transcription and mutation frequencies in mutant DT40 cell with deletions in the IgL gene locus. A, Two Northern blots with independently derived RNA samples from each cell clone were hybridized with probes for the IgL gene and GAPDH as a loading control. The signals were quantified, and all values were first normalized to the GAPDH signal in each lane and subsequently normalized to the IgL level in wild-type (WT) DT40 cells. The bars represents the average of four independent experiments (two per genotype) except for ΔMES1.1 and ΔMES34.2, where the average of two values per clone is shown. Sorted IgM cells were used for the RNA preparation to avoid issues of differential RNA stability due to nonsense-mediated decay. B, The frequency of mutation events (nontemplated and templated combined) was determined by sequencing the VJ region of the IgL gene from two individual clones of each genotype after 4 wk of continuous culture. A minimum of 77 sequences was obtained for each genotype. The dashed line indicates the background level of 5 × 10−5 events/bp for our experimental design as determined by sequencing the IgL gene in AID-deficient DT40 clones.

Close modal
FIGURE 3.

Fluctuation analysis of IgM reversion in DT40 IgL deletion mutants. Eight to 12 single cell subclones of each genotype were obtained by limiting dilution and continuously culturing for 4 wk. Each cell clone starts as surface IgM due to a stop codon in the IgL gene. After 4 wk the surface IgM status is determined by FACS and the appearance of IgM+ cells (reversion) is indicative of GCV events that reverted this stop codon. The percentage of IgM+ cells for each subclone is indicated by a diamond, and the average for each genotype is shown and indicated as a horizontal line. Because this assay only scores a subset of GCV events even in clones with no apparent IgM reversion, sequencing can still reveal ongoing sequence diversification. Data in A and B were from one experiment (hence the ΔM3.5 data points are identical), whereas C and D represent two additional independent experiment in which the indicated lines were tested. Random integrants (RI) serve as controls for each set of targeted integrants. WT, Wild type.

FIGURE 3.

Fluctuation analysis of IgM reversion in DT40 IgL deletion mutants. Eight to 12 single cell subclones of each genotype were obtained by limiting dilution and continuously culturing for 4 wk. Each cell clone starts as surface IgM due to a stop codon in the IgL gene. After 4 wk the surface IgM status is determined by FACS and the appearance of IgM+ cells (reversion) is indicative of GCV events that reverted this stop codon. The percentage of IgM+ cells for each subclone is indicated by a diamond, and the average for each genotype is shown and indicated as a horizontal line. Because this assay only scores a subset of GCV events even in clones with no apparent IgM reversion, sequencing can still reveal ongoing sequence diversification. Data in A and B were from one experiment (hence the ΔM3.5 data points are identical), whereas C and D represent two additional independent experiment in which the indicated lines were tested. Random integrants (RI) serve as controls for each set of targeted integrants. WT, Wild type.

Close modal

The robust levels of AID-mediated IgL diversification observed in the individual VJ-C intron and enhancer deletion mutants led us to wonder whether these DNA control elements might function redundantly. There is precedence for such redundancy between the κ intronic enhancer iEκ and the 3′κ enhancer 3′Eκ in mice with respect to their roles in the targeting of SHM to an Igκ transgene itself (10). We created a large deletion (ΔME) that lacks the VJ-C intron, the enhancer, and an additional 2-kb of noncoding DNA located between the polyadenylation signal and the enhancer (Fig. 1). Northern blot analysis revealed that the steady-state level of IgL transcripts was slightly elevated (1.35-fold) compared with that of wild-type DT40 cells (Fig. 2,A). Interestingly, the surface IgM reversion assay and the sequencing analysis after 4 wk of continuous culture revealed a modest 2-fold decrease of GCV/SHM rates in both of the ΔME clones (Fig. 2, B and B), suggesting that the deleted sequences might play a minor role in targeting. We concluded that a putative essential targeting sequence did not reside in these gene proximal sequences and that neither the VJ-C intron nor the enhancer is essential for IgL transcription in DT40 cells.

To search for other distal control elements within the chicken IgL locus, we deleted additional noncoding DNA regions downstream of the enhancer. Deletion of an additional 4 kb downstream of the enhancer (ΔME6K; Fig. 1) resulted in the complete absence of IgL transcripts as determined by Northern blotting and RT-PCR (Fig. 2,A and data not shown). Because GCV and SHM are strictly dependent on active transcription of the Ig genes, it was not surprising that AID-mediated sequence diversification of IgL was absent in the ΔME6K clones (Figs. 2,B and 3 C). This was not caused by a loss of AID activity, as the mutation frequencies in the unaltered Ig heavy chain (IgH) locus of ΔM (6.22 × 10−4events/bp) and ΔME6K (4.79 × 10−4events/bp) were found to be comparable. We conclude that this 4 kb region (hereafter referred to as the 3′-regulatory region, 3′RR) contains a transcriptional control element that is essential for high levels of IgL gene transcription in DT40 cells.

To determine whether the 3′RR also contained cis-acting DNA elements that target GCV/SHM to this locus, we decided to rescue IgL transcription by replacing the 3′RR with a single copy of the strong SV40 enhancer in the context of the ΔME6K genotype (ΔMES; Fig. 1). Two such cell clones, ΔMES1.1 and ΔMES34.2, were chosen for further analysis. As expected, Northern blot analysis showed that the SV40 enhancer strongly increased the level of steady-state IgL transcripts close to that found in wild-type, ΔE, and ΔME cells (Fig. 2,A), all of which showed readily detectable GCV and SHM activity. Strikingly, none of the 20 subclones of ΔMES1.1 and ΔMES34.2 showed any significant increase in the percentage of IgM+ cells (Figs. 4 and 3,D). This complete loss of GCV/SHM in the IgL gene was further confirmed by sequence analysis (Fig. 2 B). The mutation frequency (0.977 × 10−5 events/bp) is dramatically lower (20-fold) than that of ΔME cells and is again below the background level of our experiments. Importantly, mutation frequencies in the unaltered IgH locus in ΔMES (4.26 × 10−4 events/bp) remained comparable to that of the parental ΔM lines (6.22 × 10−4 events/bp). These observations demonstrate that the 3′RR harbors DNA sequences that are essential for targeting AID-mediated sequence diversification to the chicken IgL locus.

FIGURE 4.

Ig gene conversion in DT40 ΔMES cell clones. Eight to 12 individual subclones of the parental ΔM3.5 cell line, a random integrant (RI) of the MES targeting vector, and two ΔMES lines were grown continuously for 4 wk. After 2, 3, and 4 wk the percentages of cells that reverted from IgM to IgM+ were determined by FACS analysis. The average percentage of surface IgM+ cells from each cell clone is shown. Note that these data were obtained from the same identical cultures used in Fig. 3 D.

FIGURE 4.

Ig gene conversion in DT40 ΔMES cell clones. Eight to 12 individual subclones of the parental ΔM3.5 cell line, a random integrant (RI) of the MES targeting vector, and two ΔMES lines were grown continuously for 4 wk. After 2, 3, and 4 wk the percentages of cells that reverted from IgM to IgM+ were determined by FACS analysis. The average percentage of surface IgM+ cells from each cell clone is shown. Note that these data were obtained from the same identical cultures used in Fig. 3 D.

Close modal

The 3′RR constitutes one of the first cis-acting targeting element within an endogenous Ig locus. While previous studies suggested that high levels of transcription and the promoters themselves are important determinants in licensing Ig genes for AID-mediated sequence diversification (20, 21), our data show that distinct DNA sequences within the Ig loci mark the genes as targets for localized mutagenesis. Future analysis will reveal whether the same elements within the 3′RR are responsible for both enhancing transcription and GCV/SHM or whether these functions are conferred by different sets of sequences. One popular model of how targeting elements work invokes the binding of transcription factors to such DNA elements and the recruitment of AID via protein-protein interactions (reviewed in Ref. 13). Evidence is emerging that E2A could be one such factor, as the deletion/overexpression of E2A in DT40 cells leads to a corresponding decrease/increase of GCV and SHM without significantly affecting Ig transcription (19, 22). Similarly, additional E2A binding sites in murine Igκ transgenes also increase SHM frequencies (23). The 3′RR described here contains five predicted E2A binding sites (data not shown), and additional experiments will reveal whether any of these are important for targeting. It is worth noting, however, that none of the six putative E2A sites within the IgL enhancer is essential for GCV (16). To narrow our search to smaller regions within the 3′RR, we compared this DNA segment with IgL and IgH loci of other vertebrate species, but no highly conserved sequence stretches indicative of conserved elements were found (data not shown). This is not surprising, as control elements in Ig loci are rapidly evolving at the sequence level while still maintaining their regulatory function (24). Because AID-mediated Ig gene diversification is an evolutionary conserved pathway to generate Ag receptor diversity, we predict that cis-acting targeting elements are present in the Ig loci of all vertebrates, including humans and mice, and that functional complementation assays represent a direct approach to reveal their nature and identity. The identification of the 3′RR now sets the stage toward understanding the molecular mechanism and the protein factors involved that restrict these unique mutagenic processes to distinct genomic loci. It is tempting to speculate that the 3′RR is the member of a novel category of cis-acting elements that are able to control genomic (in)stability on a local scale, e.g., in the context of antigenic variation of pathogens and recombination hot spots.

We are grateful to Dr. Frank Edenhofer for proving the HTNC Cre-protein expression construct, and we thank Dr. Ranjan Sen, Dr. Patricia Gearhart, and Dr. Shu Yuan Yang for comments on this 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.

1

This work was supported by the Intramural Research Program of the National Institute on Aging/National Institutes of Health.

3

Abbreviations used in this paper: SHM, somatic hypermutation; AID, activation induced cytidine deaminase; GCV, gene conversion; IgH, Ig heavy chain; IgL, Ig light chain; 3′RR, 3′ regulatory region; MAR, matrix attachment region.

1
Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo.
2000
. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme.
Cell
102
:
553
-563.
2
Revy, P., T. Muto, Y. Levy, F. Geissmann, A. Plebani, O. Sanal, N. Catalan, M. Forveille, R. Dufourcq-Labelouse, A. Gennery, et al
2000
. Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the hyper-IgM syndrome (HIGM2).
Cell
102
:
565
-575.
3
Arakawa, H., J. Hauschild, J. M. Buerstedde.
2002
. Requirement of the activation-induced deaminase (AID) gene for immunoglobulin gene conversion.
Science
295
:
1301
-1306.
4
Harris, R. S., J. E. Sale, S. K. Petersen-Mahrt, M. S. Neuberger.
2002
. AID is essential for immunoglobulin V gene conversion in a cultured B cell line.
Curr. Biol.
12
:
435
-438.
5
Di Noia, J. M., M. S. Neuberger.
2007
. Molecular mechanisms of antibody somatic hypermutation.
Annu. Rev. Biochem.
76
:
1
-22.
6
Maizels, N..
2005
. Immunoglobulin gene diversification.
Annu. Rev. Genet.
39
:
23
-46.
7
Martomo, S. A., P. J. Gearhart.
2006
. Somatic hypermutation: subverted DNA repair.
Curr. Opin. Immunol.
18
:
243
-248.
8
Storb, U., H. M. Shen, N. Michael, N. Kim.
2001
. Somatic hypermutation of immunoglobulin and non-immunoglobulin genes.
Philos. Trans. R. Soc. Lond. B
356
:
13
-19.
9
Betz, A. G., C. Milstein, A. Gonzalez-Fernandez, R. Pannell, T. Larson, M. S. Neuberger.
1994
. Elements regulating somatic hypermutation of an immunoglobulin κ gene: critical role for the intron enhancer/matrix attachment region.
Cell
77
:
239
-248.
10
Goyenechea, B., N. Klix, J. Yelamos, G. T. Williams, A. Riddell, M. S. Neuberger, C. Milstein.
1997
. Cells strongly expressing Igκ transgenes show clonal recruitment of hypermutation: a role for both MAR and the enhancers.
EMBO J.
16
:
3987
-3994.
11
Klix, N., C. J. Jolly, S. L. Davies, M. Bruggemann, G. T. Williams, M. S. Neuberger.
1998
. Multiple sequences from downstream of the J κ cluster can combine to recruit somatic hypermutation to a heterologous, upstream mutation domain.
Eur. J. Immunol.
28
:
317
-326.
12
Inlay, M. A., H. H. Gao, V. H. Odegard, T. Lin, D. G. Schatz, Y. Xu.
2006
. Roles of the Ig κ light chain intronic and 3′ enhancers in Igk somatic hypermutation.
J. Immunol.
177
:
1146
-1151.
13
Odegard, V. H., D. G. Schatz.
2006
. Targeting of somatic hypermutation.
Nat. Rev. Immunol.
6
:
573
-583.
14
Fukita, Y., H. Jacobs, K. Rajewsky.
1998
. Somatic hypermutation in the heavy chain locus correlates with transcription.
Immunity
9
:
105
-114.
15
Tumas-Brundage, K., T. Manser.
1997
. The transcriptional promoter regulates hypermutation of the antibody heavy chain locus.
J. Exp. Med.
185
:
239
-250.
16
Yang, S. Y., S. D. Fugmann, D. G. Schatz.
2006
. Control of gene conversion and somatic hypermutation by immunoglobulin promoter and enhancer sequences.
J. Exp. Med.
203
:
2919
-2928.
17
Arakawa, H., D. Lodygin, J. M. Buerstedde.
2001
. Mutant loxP vectors for selectable marker recycle and conditional knock-outs.
BMC Biotechnol.
1
:
7
18
Yang, S. Y., S. D. Fugmann, H. S. Gramlich, D. G. Schatz.
2007
. Activation-induced Cytidine Deaminase-mediated Sequence Diversification Is Transiently Targeted to Newly Integrated DNA Substrates.
J. Biol. Chem.
282
:
25308
-25313.
19
Conlon, T. M., K. B. Meyer.
2006
. The chicken Ig light chain 3′-enhancer is essential for gene expression and regulates gene conversion via the transcription factor E2A.
Eur. J. Immunol.
36
:
139
-148.
20
Yoshikawa, K., I. M. Okazaki, T. Eto, K. Kinoshita, M. Muramatsu, H. Nagaoka, T. Honjo.
2002
. AID enzyme-induced hypermutation in an actively transcribed gene in fibroblasts.
Science
296
:
2033
-2036.
21
Martin, A., M. D. Scharff.
2002
. Somatic hypermutation of the AID transgene in B and non-B cells.
Proc. Natl. Acad. Sci. USA
99
:
12304
-12308.
22
Schoetz, U., M. Cervelli, Y. D. Wang, P. Fiedler, J. M. Buerstedde.
2006
. E2A expression stimulates Ig hypermutation.
J. Immunol.
177
:
395
-400.
23
Michael, N., H. M. Shen, S. Longerich, N. Kim, A. Longacre, U. Storb.
2003
. The E box motif CAGGTG enhances somatic hypermutation without enhancing transcription.
Immunity
19
:
235
-242.
24
Hikima, J., M. L. Lennard, M. R. Wilson, N. W. Miller, L. W. Clem, G. W. Warr.
2006
. Conservation and divergence of the Emicro3′ enhancer in the IGH locus of teleosts.
Immunogenetics
58
:
226
-234.