JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M507906200 on March 10, 2006

J. Biol. Chem., Vol. 281, Issue 18, 12370-12380, May 5, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
281/18/12370    most recent
M507906200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rahman, N. S.
Right arrow Articles by Rodgers, K. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rahman, N. S.
Right arrow Articles by Rodgers, K. K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

DNA Cleavage of a Cryptic Recombination Signal Sequence by RAG1 and RAG2

IMPLICATIONS FOR PARTIAL VH GENE REPLACEMENT*Formula

Negar S. Rahman{ddagger}, LeAnn J. Godderz{ddagger}1, Stephen J. Stray§2, J. Donald Capra§, and Karla K. Rodgers{ddagger}3

From the {ddagger}Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma 73190, §Molecular Immunogenetics Research Program, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma 73104

Received for publication, July 20, 2005 , and in revised form, March 8, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Antibody and T cell receptor genes are assembled from gene segments by V(D)J recombination to produce an almost infinitely diverse repertoire of antigen specificities. Recombination is initiated by cleavage of conserved recombination signal sequences (RSS) by RAG1 and RAG2 during lymphocyte development. Recent evidence demonstrates that recombination can occur at noncanonical RSS sites within Ig genes or at other loci, outside the context of normal lymphocyte receptor gene rearrangement. We have characterized the ability of the RAG proteins to bind and cleave a cryptic RSS (cRSS) located within an Ig VH gene segment. The RAG proteins bound with sequence specificity to either the consensus RSS or the cRSS. The RAG proteins nick the cRSS on both the top and bottom strands, thereby bypassing the formation of the DNA hairpin intermediate observed in RAG cleavage of canonical RSS substrates. We propose that the RAG proteins may utilize an alternative mechanism for double-stranded DNA cleavage, depending on the substrate sequence. These results have implications for further diversification of the antigen receptor repertoire as well as the role of the RAG proteins in genomic instability.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The primary repertoire of antigen-binding receptors is produced by V(D)J recombination during lymphocyte development (1–4). Functional Ig and T cell receptor genes are initially generated by recombination in the bone marrow and thymus, respectively, by assembly of component V (variable), J (joining), and sometimes D (diversity) gene segments (5). In addition to V(D)J recombination, the Ig repertoire is further expanded by somatic hypermutation in germinal centers (GC)4 of peripheral lymphoid organs, where Ig class switching and B cell maturation to plasma and memory cells also occurs (6). Together, these mechanisms produce the almost infinite variety of Ig specificities required to recognize all possible antigens the organism may encounter.

Site-specific DNA cleavage in the V(D)J process is catalyzed by the recombination activating proteins, RAG1 and RAG2 (3, 4). The site specificity of the DNA cleavage reaction is governed by the recombination signal sequence (RSS) that flanks each of the V, D, and J gene segments. The RSS contains conserved heptamer and nonamer sequences separated by 12 ± 1 or 23 ± 1 poorly conserved base pairs, referred to as the 12-RSS and 23-RSS, respectively. After DNA cleavage, the gene segments are joined by factors that mediate DNA repair via the nonhomologous end-joining (NHEJ) pathway. Efficient V(D)J recombination between two gene segments requires that each be flanked by RSS of differing spacer length, a restriction referred to as the 12/23 rule (3, 4).

The RAG proteins together produce DNA double-stranded breaks (DSB) between the RSS heptamer and the flanking coding sequence via two concerted steps (7). First, the RAG proteins form a single-stranded nick 5' to the RSS heptamer. In the second step, the RAG proteins mediate transesterification of the newly formed 3'-OH with the phosphate group directly opposite on the complementary strand. Overall, RAG-catalyzed DNA DSB yields coding sequences terminated by covalently sealed hairpins (coding ends) and blunt-ended breaks terminated at RSS heptamers (signal ends). The 12/23 rule is enforced at formation of the coding end hairpins, since transesterification preferentially occurs in the context of the synaptic complex, which consists of the RAG proteins bound to both a 12-RSS and a 23-RSS (8, 9). In contrast, the nicking reaction may occur with RAG proteins bound to a single 12- or 23-RSS. Enforcement of the 12/23 rule at the hairpin formation step ensures that production of DSB occurs at correct sequences, thereby reducing the possibility of cleavage at incorrect sites.

In addition to their role in formation of the primary repertoire of Ig genes, the RAG proteins also mediate secondary recombination events on Ig gene loci that have been previously rearranged. Immature B cells that have defective or self-reactive B cell receptors must be removed from the repertoire. This may occur by deletion of the cell or by a secondary recombination event, referred to as receptor editing (10, 11). In this case, the RAG proteins generate a de novo rearrangement of Ig genes (generally the light chain genes), which yields a modification of the Ig receptor specificity, and thus mediates tolerance.

Secondary rearrangements at the Ig heavy chain (IgH) locus were believed to be unlikely due to the deletion of the remaining D segments after the primary recombination event. However, several reports indicate that secondary rearrangements can occur at this locus and are thought to require the use of cRSS elements embedded within the VH genes. For example, a cRSS that is conserved at the 3' end of all VH gene segments has been proposed to govern RAG-mediated receptor editing, in which an entire VH gene segment replaces a portion of the rearranged VH gene, resulting in the addition of several nucleotides at the V-DJ border (12). To accomplish complete VH gene replacement, it has been proposed that the RAG proteins create DSB at both the cRSS and the conventional RSS flanking the replacing VH gene segment. Consistent with this proposal, recent studies demonstrated that in in vitro DNA cleavage assays, the RAG proteins create DSB at the cRSS combined with either a 12- or 23-RSS via the standard nick-hairpin mechanism (13). Thus, the RAG proteins can presumably bind to both the cRSS and the conventional 12- or 23-RSS in a synaptic complex.

Additional secondary rearrangement events have been proposed to occur in mature, peripheral B cells in a process referred to as receptor revision (14). Evidence for novel receptor revision events at the IgH locus have been reported in peripheral human B cells, which include tonsillar B cells (15), in the synovium of rheumatoid arthritis patients (16), in NOD/SCID mice reconstituted with human cord blood mononuclear cells (17), and in certain B cell lymphomas (18, 19). In these studies, it was found that the 5' portion of a rearranged VH gene was replaced with the homologous portion of another VH gene segment to form a hybrid antibody gene. In human tonsillar B cells, somatically hypermutated and rearranged VH 4–34 gene segments apparently recombined with other members of the VH 4 family such that the original CDR3, formed by primary V(D)J rearrangement, was conserved (15). The most frequently identified "donor" sequences were rearranged VH 4–34 alleles, with recipient sequences derived from VH 4–39 and VH 4–61 gene segments. Due to the high degree of sequence similarity between VH gene segments, particularly those of the same family, it is impossible to determine the exact break point at which hybrid formation occurred. However, inspection of the sequences revealed that each hybrid had been formed in close proximity to a cRSS, leading to the hypothesis that these secondary recombination events were mediated by RAG1 and RAG2 (15). Despite the compelling evidence for RAG-mediated receptor revision, the re-expression of the RAG proteins in GC B cells in humans remains controversial (20, 21), and alternative mechanisms based on the induction of homologous recombination following cleavage by activation-induced cytidine deaminase have been proposed (22). However, a recent report identified hybrid IgH genes in B cells of human B cell reconstituted NOD/SCID mice, indicating that such events can occur independently from the GC (17); thus, an exclusive role for activation-induced cytidine deaminase is unlikely.

To ascertain if the reported VH gene replacements are RAG-mediated, we tested the activity of the RAG proteins on a cRSS (located 3' to CDR2 of VH 4–34), which is closely similar to the consensus 12-RSS. Whereas the RAG proteins could efficiently perform double-stranded cleavage at the cRSS, it did not occur via a mechanism consistent with V(D)J recombination. Rather, the cleavage occurred by a process that would bypass the requirement for the 12/23 rule. In addition, the resulting cleavage products would be viable substrates for either NHEJ or homologous recombination repair (HRR), which particularly in the latter case could lead to the hybrid IgH genes that have been observed in peripheral human B cells.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Protein Expression and Purification—Purified RAG1 and RAG2 proteins consisted of the core regions, which are the minimal regions required for DNA recombination activity (3, 4). MBP-core RAG1 (residues 384–1008 of the full-length RAG1 protein fused to maltose-binding protein) was expressed in Escherichia coli and purified as previously described (23). Briefly, the cells were disrupted by sonication, and protein was purified by affinity chromatography using amylose resin, followed by ion exchange chromatography through a Q-Sepharose fast flow column. In the final purification step, the fusion protein was chromatographed through a Superdex 200 gel filtration column. The protein was judged to be >95% pure by Coomassie Blue staining of SDS-polyacrylamide gels. The concentration of the purified protein was determined by UV absorbance at 280 nm using an extinction coefficient of 129.5 mM–1 cm–1.

GST-core RAG2 consisted of residues 1–387 of the full-length RAG2 protein fused to glutathione S-transferase. GST-core RAG2, either alone or co-expressed with MBP-core RAG1, was transiently expressed and purified from 293T cells as previously described (24). Briefly, after cell lysis, the fusion proteins were bound to glutathione-linked agarose resin, followed by extensive washing, and then released from the resin with purification buffer containing 20 mM reduced glutathione. Mammalian expressed proteins were judged to be >90% pure by Coomassie Blue staining of SDS-polyacrylamide gels. Protein concentrations were determined by Western blotting using monoclonal anti-GST or anti-MBP antibodies where appropriate. Standard curves were constructed using known concentrations of GST or MBP proteins, which were subjected to SDS-PAGE on the same gel as the respective GST- or MBP-RAG fusion protein. Purified HMGB1 was obtained from Sigma.

Construction of Oligonucleotide Duplexes—Oligonucleotides were commercially synthesized and purified by polyacrylamide gel electrophoresis (Integrated DNA Technologies). The 12-RSS, 23-RSS, and VH 4–34 cRSS substrates were prepared by annealing complementary oligonucleotides. The sequences of the top strand oligonucleotides were as follows: 12-RSS, d(GATATGGCTCGTCTTACACAGTGATATAGACCTTAACAAAAACCTCCAATCGA); 23-RSS, d(GGCTCGTCTTACACAGTGATGGAAGCTCTATCGGATCTCCGACAAAAACCTCGAGCGGAG); and VH 4–34 cRSS, d(CGTCCCTCAAGAGTCGAGTCACCATATCAGTAGACACGTCCAAGAACCAGTCT). The mutant heptamer and nonamer (MHMN) 12-RSS oligonucleotide sequences are identical to that of consensus 12-RSS, except that the heptamer sequence (CACAGTG) has been replaced by the sequence GAGAAGC, and the nonamer sequence (ACAAAAACC) has been replaced by the sequence AGGCTCTTG. Prior to annealing, either the top or bottom strand of each duplex was labeled at the 5' end with [{gamma}-32P]ATP using T4 polynucleotide kinase. To generate nicked substrates, two separate oligonucleotides were synthesized that corresponded to the sequence 5' and 3' to the nick site, respectively. The oligonucleotides were annealed to the full-length complement. For example, to construct the duplex containing the nick 5' to the heptamer in the consensus 12-RSS, two oligonucleotides, one containing bases 1–16 and the other bases 17–53 of the top strand, were annealed with the 12-RSS bottom strand oligonucleotide. In each experiment, the oligonucleotide 3' to the nick was phosphorylated at the 5' end using nonradiolabeled ATP. In addition, prior to annealing, either the oligonucleotide 5' to the nick or full-length complement were labeled at the 5' end with [{gamma}-32P]ATP using T4 polynucleotide kinase. To anneal substrate duplexes, the appropriate oligonucleotides were combined at equimolar ratios and heated to 95 °C, and the mixture was slowly cooled to ambient temperature. The nicked VH 4–34 con-hep cRSS substrate was identical to the wild type VH 4–34 cRSS sequence, except that the cryptic heptamer (CACCATA) was replaced with the consensus heptamer (CACAGTG). Likewise, in the nicked VH 4–34 con-non cRSS sequence, the cryptic nonamer (CCAAGAACC) was replaced with the consensus nonamer (ACAAAAACC).

Single RSS Substrate DNA Cleavage Assays Analyzed by Denaturing Polyacrylamide Gels—Single RSS (or cRSS) DNA cleavage assays were performed with buffer containing 5 mM MnCl2, 5% glycerol, 50 mM NaCl, 0.1 mg/ml bovine serum albumin, and 20 mM Na-HEPES, pH 7.5. MBP-core RAG1 and GST-core RAG2 were mixed with 1 nM 32P-labeled duplex substrate and incubated for the time indicated at 37 °C. In the experiments with the fusion proteins expressed separately, MBP-core RAG1 was incubated with GST-core RAG2 for 15 min at 4 °C prior to the addition to DNA substrate. Reactions were stopped by the addition of an equal volume of formamide gel loading buffer and heating at 95 °C for 5 min. The cleavage products were separated on 8% denaturing polyacrylamide gels. The gels were analyzed by autoradiography and phosphorimaging using an Amersham Biosciences SI PhosphorImager. Where indicated, bands were quantitated using the program ImageJ.


Figure 1
View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 1.
Overview of partial VH gene replacement. A, schematic of rearrangement events that lead to partial VH gene replacement. The separate V, D, and J gene segments are labeled. The 12- and 23-RSS elements are represented as black and gray triangles, respectively. Inverted black triangles indicate the location of the conserved VH cRSS that is 3' to CDR2 (see supplemental Fig. 1). The first recombination reaction represents primary rearrangement events that yield the combined V-D-J segments. The second reaction represents secondary rearrangement events utilizing the cRSS, which results in partial VH gene replacement. Although not represented in this schematic, it should be noted that in the human germ line sequence not all VH gene segments are in the same orientation. B, comparison of the heptamer and nonamer elements in the consensus 12-RSS versus the VH 4–34 cRSS. The spacers between the two conserved sequences are 12 and 13 base pairs for the 12-RSS and cRSS, respectively. The boxes enclose the bases that are identical in the cRSS with those in the consensus 12-RSS.

 
DNA Cleavage Assays Analyzed by Nondenaturing Polyacrylamide Gels—DNA cleavage assays analyzed by nondenaturing polyacrylamide gels were performed as described above with the following exceptions. The RAG proteins were incubated with 1 nM radiolabeled oligonucleotide duplex for 2 h at 37 °C. Subsequently, EDTA and SDS were added to each sample to final concentrations of 7.5 mM and 0.4%, respectively. The cleavage products were separated on 6% nondenaturing polyacrylamide gels.

DNA Coupled Cleavage Assays—DNA coupled cleavage assays were performed as described under "Single RSS Substrate DNA Cleavage Assays Analyzed by Denaturing Polyacrylamide Gels" with the following exceptions. The reactions were performed in 5 mM MgCl2, 5% glycerol, 0.1 M NaCl, and 10 mM Tris-HCl, pH 8.0. MBP-core RAG1 and GST-core RAG2 were incubated with unlabeled consensus 12-RSS or VH 4–34 cRSS for 30 min at 4 °C, followed by incubation with 10 nM 32P-labeled 23-RSS duplex substrate for 1 h at 37 °C.

Measurement of Protein-DNA Interactions—Electrophoretic mobility shift assays (EMSA) demonstrating binding of the RAG proteins to consensus 12-RSS and VH 4–34 cRSS substrates were performed as previously described (25). EMSA samples were subjected to electrophoresis on discontinuous, nondenaturing gels of 3.5/8% polyacrylamide and analyzed by autoradiography and phosphorimaging using a Molecular Dynamics SI phosphorimager. Bands were quantitated using the program ImageJ.

Specificity of Protein-DNA Interactions—The buffer used in the competition assays was the same as for the protein-DNA interaction assays (25). Competition assays were accomplished by adding MBP-core RAG1 and GST-core RAG2 proteins to a mixture of 1 nM 32P-labeled VH 4–34 cRSS duplex with increasing concentrations of unlabeled competitor DNA (5–10 nM), which was either the VH 4–34 cRSS or MHMN 12-RSS oligonucleotide duplex.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Several cRSS have been implicated in partial VH gene replacement (supplemental Fig. 1). In these cases, the replacement resulted in a portion of the VH gene replacing the homologous region in the rearranged V-DJ gene to create a hybrid IgH gene (Fig. 1A). Since one of the more commonly observed hybrid joints occurred at a cRSS with a sequence closely similar to the consensus 12-RSS (Fig. 1B), this event was proposed to be RAG-mediated (15). To test this proposal, we have examined the action of the RAG proteins on this cRSS, which we refer to as VH 4–34 cRSS.

Binding of the RAG Proteins to VH 4–34 cRSS Versus Consensus 12-RSS—To confirm that the RAG proteins could bind the VH 4–34 cRSS, EMSA were used (Fig. 2). RAG1-RAG2 complexes bound to both the consensus 12-RSS (Fig. 2A, lanes 4 and 5) and the VH 4–34 cRSS (lanes 9 and 10) substrates. At the protein concentrations used, the RAG proteins shifted ~50% less of the latter substrate, corresponding to a decreased binding affinity of 2–3-fold to the VH 4–34 cRSS compared with the consensus 12-RSS substrate. Protein-DNA complexes were formed inefficiently with RAG1 alone (lanes 2 and 7), requiring prolonged exposure for detection (data not shown). As expected, in the absence of RAG1, there is no detectable complex formed between RAG2 and either DNA substrate (lanes 3 and 8) even with longer exposure times (data not shown). The addition of increasing concentrations of RAG2 produced substantial interaction of the RAG proteins together to both the consensus 12-RSS (lanes 4 and 5) and the VH 4–34 cRSS (lanes 9 and 10). These results are consistent with previous reports that RAG2 enhances the affinity of RAG1 for the DNA by up to 20-fold (25).

The RAG proteins bound with detectable sequence specificity to the VH 4–34 cRSS as determined by competition assays (Fig. 2B). Each sample contained the combined RAG proteins with radiolabeled VH 4–34 cRSS. To each sample, increasing concentrations (5–10 nM) of unlabeled VH 4–34 cRSS duplex (lanes 2 and 3) or mutated 12-RSS (lanes 4 and 5) were added. The mutated 12-RSS duplex (referred to as MHMN 12-RSS) was identical to consensus 12-RSS, except that the heptamer and nonamer were replaced with unrelated sequences. With the higher concentration of VH 4–34 cRSS competitor, <4% of radiolabeled protein-DNA complexes were detected (lane 1 versus lane 3). In contrast, the MHMN 12-RSS duplex was a weaker competitor, since 19% of the radiolabeled complex remained at the higher competitor concentration (lane 1 versus lane 5).


Figure 2
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 2.
Interaction of RAG1 and RAG2 with VH 4–34 cRSS. A, EMSA of 32P-radiolabeled consensus 12-RSS (lanes 1–5) and VH 4–34 cRSS (lanes 6–10) incubated with the RAG proteins as indicated. When indicated in the lanes, MBP-core RAG1 was present at a concentration of 0.025 µM. Samples loaded into lanes 4 and 9 contained 0.020 µM, and lanes 2, 5, 7, and 10 contained 0.025 µM GST-core RAG2. Following incubation for 30 min at 25 °C, the reactions were subjected to electrophoresis on 3.5/8% nondenaturing polyacrylamide gels. B, the RAG proteins bound with sequence specificity to the VH 4–34 cRSS substrate as determined by EMSA competition assays. Each lane contained 32P-labeled VH 4–34 cRSS duplex combined with MBP-core RAG1 and GST-core RAG2 (each at 0.025 µM). Unlabeled VH 4–34 cRSS competitor DNA (lanes 2 and 3) and unlabeled MHMN 12-RSS competitor DNA (lanes 4 and 5) were added at increasing concentrations of 5 and 10 nM. Bands corresponding to VH 4–34 cRSS unbound and complexed to the RAG proteins are labeled.

 


Figure 3
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 3.
RAG proteins create DSB on both consensus 12-RSS and VH 4–34 cRSS substrates. Shown are consensus 12-RSS (A) and VH 4–34 cRSS (B)5'-labeled with 32P on the top strand and incubated with the RAG proteins (separately expressed) as indicated. The concentrations of the RAG proteins were MBP-core RAG1 at 0.2 µM and GST-core RAG2 at 0.4 µM. The samples were subjected to electrophoresis on 6% nondenaturing polyacrylamide gels.

 
The RAG Proteins Catalyze Cleavage of the VH 4–34 cRSS—We tested the ability of RAG1-RAG2 complexes to cleave the VH 4–34 cRSS substrate using an in vitro DNA cleavage assay. The RAG proteins were incubated with the consensus RSS or cRSS substrates in Mn2+-containing buffers and analyzed on nondenaturing polyacrylamide gels (Fig. 3). Under these conditions, the DNA strands remain annealed to each other; thus, the appearance of smaller products demonstrated the creation of a DSB. Upon the addition of both RAG proteins to the radiolabeled consensus 12-RSS duplex, a product of faster mobility was formed, presumably corresponding to the DNA hairpin product (Fig. 3A). Similar results were observed with the VH 4–34 cRSS duplex in that a cleavage product was produced upon the addition of both RAG proteins (Fig. 3B), demonstrating that the RAG proteins are capable of creating DSB on the cRSS substrate.

To determine where DNA cleavage occurred in the VH 4–34 cRSS substrate, the in vitro DNA cleavage assays were analyzed on denaturing polyacrylamide gels (Fig. 4). As expected, the reaction containing consensus 12-RSS duplex with the top strand 5'-labeled yielded the nicked product (at the expected location 5' to the heptamer) and the hairpin product (Fig. 4B, lane 4) in the presence of both RAG proteins. Moreover, when the bottom strand of the consensus 12-RSS duplex was 5'-labeled, treatment with both RAG proteins produced the expected byproduct of hairpin formation (Fig. 4B, lane 8). In comparison, using the VH 4–34 cRSS duplex as substrate, the RAG proteins together mediated cleavage on the top strand (Fig. 4D, lane 4) as well as on the complementary bottom strand (Fig. 4E, lane 4). However, the major product formed from the top strand was 21 nucleotides (nt) in length (Fig. 4D, lane 4). This fragment resulted from cleavage not 5' of the CAC, as expected, but between the A and C (at CA{downarrow}C). A minor cleavage product of 34 nt was evident on overexposure of the gel, which was probably due to cleavage at a second CAC sequence in the VH 4–34 cRSS duplex. On the bottom strand of the VH 4–34 cRSS duplex, two major pairs of cleavage products at 15 and 16 nt and 33 and 34 nt in length were produced upon incubation with both RAG proteins (Fig. 4E, lane 4). We observed identical cleavage products on the cRSS substrate using RAG proteins (MBP-core RAG1 and GST-core RAG2) that were co-expressed in 293T cells (Fig. 4, D and E, lanes 6), showing that the activity is independent of the expression system for RAG1. Moreover, no cleavage products from the MHMN 12-RSS substrate were observed upon incubation with RAG1 and RAG2, indicating that cleavage is specific to RSS or cRSS sequences (data not shown). Thus, the RAG proteins are able to perform double-stranded DNA cleavage on the VH 4–34 cRSS substrate, although the formation of hairpin products was not observed.

Kinetic analyses of the DNA cleavage reactions with the cRSS substrate were also inconsistent with a nick-hairpin mechanism (Fig. 5). In the normal nick-hairpin mechanism, the top strand is nicked by the RAG1-RAG2 complex, and then hairpins are formed by attack of the bottom strand by the 3'-hydroxyl; thus, formation of the hairpin product lags behind that of the nicked product, as observed for products from the consensus 12-RSS substrate (Fig. 5A, lanes 2–5). Likewise, the appearance of the cleavage product from the bottom strand matched that of hairpin formation (Fig. 5A, lanes 7–10). However, with the VH 4–34 cRSS substrate, the sequential appearance of two cleavage products of n and 2n nt in length, along with formation of a corresponding product on the opposite strand, were not apparent (Fig. 5B). Instead, cleavage on the top strand occurred within the cryptic heptamer, yielding the 21-nt product with no additional products of ~42 nt formed. With the 32P-labeled bottom strand substrate, products of 15 and 16 nt and 33 and 34 nt in length were produced, which could correspond to hairpin formation. However, significant levels of products on the top strand of ~38–39 nt in length were not observed. RAG1 and RAG2 proteins expressed separately showed similar results in the kinetics of DNA cleavage as compared with the co-expressed proteins (data not shown). However, the 15- and 16-nt products on the bottom strand were formed more efficiently by the co-expressed proteins.


Figure 4
View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 4.
RAG-mediated cleavage on duplex substrates analyzed with denaturing polyacrylamide gels. A, the sequence of the consensus 12-RSS duplex with the heptamer and nonamer sequences shaded gray. Below the sequence of the 12-RSS are the cleavage products that would be observed if the top or bottom strands were radiolabeled at the 5' end. The cleavage products observed with the radiolabel on the 5' end of the top strand are referred to as N and HP for nick and hairpin, respectively. The product observed with the bottom strand 5'-labeled is referred to as 5'b. B, RAG-mediated cleavage of the consensus 12-RSS with the top strand (left) and the bottom strand (right)5'-labeled. The concentrations of the RAG proteins (separately expressed) in the indicated lanes are MBP-core RAG1 at 0.1 µM and GST-core RAG2 at 0.2 µM. The proteins were incubated with the DNA substrate for 1 h at 37 °C. The samples were subjected to electrophoresis on 8% denaturing polyacrylamide gels. Bands corresponding to the cleavage products are labeled accordingtothe nomenclature in A. C, the sequence of the VH 4–34 cRSS duplex with the cryptic heptamer and nonamer sequences shaded gray. The cleavage products are represented as solid bars with arrows indicating the nicking site. With the top strand 5'-radiolabeled, one cleavage product referred to as 21T is observed. With the bottom strand 5'-radiolabeled, two pairs of cleavage products referred to as 15,16B and 33,34B are observed. D, RAG-mediated cleavage of the VH 4–34 cRSS duplex with the top strand 5'-labeled with 32P. The left panel shows the RAG proteins (separately expressed) with MBP-core RAG1 at 0.1 µM and GST-core RAG2 at 0.2 µM. The right panel shows the co-expressed RAG proteins with MBP-core RAG1 at 0.1 µM and GST-core RAG2 at 0.2 µM. In all samples, the proteins were incubated with the substrate at 37 °C for 1 h. Bands corresponding to cleavage products are labeled according to the nomenclature in C. Lanes containing standards are labeled M. The standards are labeled according to their length in bases followed by st. E, same as in D, except that the substrate was the VH 4–34 cRSS duplex radiolabeled at the 5' end of the bottom strand. The cleavage products are labeled according to C. The standards are labeled as in D.

 


Figure 5
View larger version (32K):
[in this window]
[in a new window]
  		FIGURE 5.
Time course analysis of RAG-mediated DNA cleavage of consensus 12-RSS substrate (A) and VH 4–34 cRSS substrate (B). Co-expressed MBP-core RAG1 (at 0.1 µM) and GST-core RAG2 (at 0.2 µM) were incubated with the radiolabeled substrate in each sample for the indicated time intervals. After incubation, the samples were subjected to electrophoresis on 8% denaturing polyacrylamide gels. The cleavage products are labeled as in Fig. 4. The asterisk represents a product probably due to nicking of the hairpin product 3' to the hairpin tip (28, 29). The sequences of the standards in the left panel of A (labeled nick st and hairpin st) correspond to the expected nick and hairpin products from the consensus 12-RSS substrate. The remaining standards are labeled as in Fig. 4.

 
The RAG Proteins Do Not Cleave the 4–34 Cryptic RSS via a Hairpin Intermediate—To further show that the RAG proteins did not efficiently catalyze hairpin formation on the VH 4–34 cRSS substrate, nicked substrates were used in the in vitro DNA cleavage assays. To test if the double-stranded cleavage of the VH 4–34 cRSS can occur via the nick-hairpin mechanism, we used the nicked VH 4–34 cRSS substrate, in which the nick occurred between bases 21 and 22 of the cRSS duplex (Fig. 6A, lane 1). With the addition of both RAG proteins, the input substrate, along with minor levels of the 42-nt hairpin, were observed (Fig. 6A, lane 2). The formation of the hairpin product was very inefficient and only occurred after long incubation time with the RAG proteins. In contrast, the nicked VH 4–34 cRSS substrate with the bottom strand labeled at the 5'-end (Fig. 6A, lane 3), yielded efficient formation of the 33- and 34-nt products after incubation with the RAG proteins (Fig. 6A, lane 4). Significantly, the levels of the bottom strand 33- and 34-nt products at 30% of total DNA (lane 4) versus the 42-nt product from the 5'-labeled top strand at <2% of total DNA (lane 2) were not consistent with the same DNA cleavage event, demonstrating that the 33- and 34-nt fragments produced from the bottom strand were not due to the hairpin formation but rather to an independent nicking reaction. Moreover, a nicked substrate with the nick placed between bases 15 and 16 from the 5' end of the bottom strand (corresponding to the nicked product observed in Fig. 4E, lanes 4 and 6) also did not yield formation of a hairpin product upon the addition of the RAG proteins (data not shown).

Treatment of the nicked VH 4–34 cRSS substrate with the coexpressed RAG complex also showed robust formation of the 33- and 34-nt bottom strand products that were not consistent with hairpin formation, given that the hairpin product comprised >4% and the 33- and 34-nt products comprised 40% of the total DNA in the respective samples (Fig. 6B, lane 2 versus lane 4). As described above with the kinetic experiments, the co-expressed RAG proteins more efficiently formed the bottom strand 15- and 16-nt products at 40% of total DNA (Fig. 6B, lane 4) versus the separately expressed proteins at <1% of total DNA (Fig. 6A, lane 4). Moreover, the addition of HMGB1 to the co-expressed RAG proteins did not alter the ratio of hairpin to 33- and 34-nt products in the hairpin formation assay (supplemental Fig. 2). A relatively high ratio of HMGB1 to the RAG complex (~6:1 HMGB1 to RAG2) did yield a significant enhancement of the nicked products (33 and 34 nt) on the bottom strand, with only minute amounts of hairpin product (supplemental Fig. 2). Thus, whereas HMGB1 has been reported to facilitate hairpin formation on canonical RSS substrates (26), this is not the case with the VH 4–34 cRSS substrate.


Figure 6
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 6.
Hairpin formation assay. Nicked VH 4–34 cRSS substrate (A and B) and nicked consensus 12-RSS substrate (C) were labeled with 32P and incubated with the RAG proteins. The DNA substrates were prepared as described under "Experimental Procedures," and schematics of the substrates are shown at the top of each panel. The concentration of the RAG proteins were as in Figs. 4 and 5. After incubation at 37 °C for 1 h, the samples were subjected to electrophoresis on 10% denaturing polyacrylamide gels. In each panel the top strand 5' to the nick (lanes 1 and 2) and the bottom strand (lanes 3 and 4) were 5'-labeled. The standards and cleavage products are labeled according to the nomenclature in Fig. 4. The bands labeled with asterisks in C indicate cleavage products resulting from production of a hairpin (HP) of larger than usual size (left panel) and its corresponding fragment on the bottom strand (right panel) as discussed under "Results."

 


Figure 7
View larger version (19K):
[in this window]
[in a new window]
  		FIGURE 7.
Coupled cleavage assay. The RAG proteins (separately expressed) with MBP-core RAG1 at 0.05 µM and GST-core RAG2 at 0.1 µM were incubated with HMGB1 (at 0.3 µM) and 10 nM 23-RSS duplex substrate radiolabeled at the 5' end of the top strand. Each sample was incubated for 1 h at 37 °C in Mg2+-containing buffer. Increasing concentrations (1, 2.5, and 5 nM) of unlabeled consensus 12-RSS (A) or VH 4–34 cRSS (B) duplexes were added in lanes 3–5. The sequences of the standards labeled nick st and hairpin st correspond to the expected products from the top strand 5'-radiolabeled 23-RSS duplex. The nick and hairpin cleavage products from the 23-RSS are labeled 23N and 23HP, respectively.

 
As a control for the hairpin formation assay, the nicked consensus 12-RSS duplex with the nick placed 5' to the consensus heptamer sequence was tested (Fig. 6C). The 16-nt coding flank on the consensus 12-RSS top strand was 5'-labeled in the assay (Fig. 6C, lane 1). After incubation with the RAG proteins, the resulting products were separated on a denaturing polyacrylamide gel. In addition to the 16-nt coding flank, a product of 32 nt was present, consistent with hairpin formation (Fig. 6C, lane 2). The presence of a minor product of larger size than the expected 32-nt hairpin may be due to transesterification at the 3' side of the heptamer resulting in a larger hairpin product, as has been previously observed (27). Likewise, minor products of a size less that the 32-nt hairpin may be due to hairpin nicking, previously shown to be RAG-mediated in in vitro DNA cleavage assays (28, 29). Standard hairpin formation of the consensus 12-RSS duplex was further confirmed using the nicked consensus 12-RSS duplex with the bottom strand 5'-labeled (Fig. 6C, lane 3). Upon the addition of both RAG proteins, a 43-nt product was formed (Fig. 6C, lane 4) at levels that coincided with that of the 32-nt hairpin, consistent with the introduction of a DSB by conversion of the nick to a hairpin.

The VH 4–34 Cryptic RSS Substrate Does Not Support the 12/23 Rule—Given that cleavage on the VH 4–34 cRSS substrate seems to occur via a nick-nick mechanism rather than a coordinated nick-hairpin mechanism, we asked if the cRSS substrate could still promote double-stranded DNA cleavage on a 23-RSS in compliance with the 12/23 rule. In Mg2+-containing buffers, the RAG proteins efficiently nicked a radiolabeled consensus 23-RSS but formed only minute amounts of hairpin product (Fig. 7, A and B, lanes 2). The addition of unlabeled consensus 12-RSS substantially enhanced the formation of a hairpin product from the 23-RSS substrate, as expected for the 12/23 rule (Fig. 7A, lanes 3–5). Hairpin formation occurred over a 5-fold range of 12-RSS to 23-RSS ratios. In contrast, the addition of unlabeled VH 4–34 cRSS at the same concentrations as the consensus 12-RSS not only did not enhance hairpin formation but actually inhibited formation of the nicked product from the 23-RSS substrate (Fig. 7B, lanes 3–5). It is apparent that the cRSS cannot functionally replace the consensus 12-RSS in a synaptic complex. Rather, the results are consistent with the cRSS acting as an efficient competitor of the 23-RSS, which led to reduced nicking activity of the 23-RSS substrate as the concentration of cRSS/23-RSS approached equimolar ratios. The VH 4–34 cRSS also could not replace the 23-RSS in enhancing hairpin formation on the consensus 12-RSS (data not shown).


Figure 8
View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 8.
Hairpin formation assays using variant VH 4–34 cRSS substrates. The experiment was performed as in Fig. 6. Each sample contained 1nM nicked substrate that was 5'-radiolabeled on either the top strand (A) or the bottom strand (B) with a schematic of the substrate used shown at the top of each panel. The solid circle in the schematics represents the 32P label. Where indicated, the RAG proteins (separately expressed) were at concentrations of MBP-core RAG1 at 0.1 µM and GST-core RAG2 at 0.2 µM. In both A and B, lanes 1 and 2 contain the wild type (WT) VH 4–34 cRSS (sequence shown in Fig. 4C), lanes 3 and 4 contain the VH 4–34 cRSS (con-hep) substrate (cryptic heptamer replaced with the consensus heptamer sequence), and lanes 5 and 6 contain the VH 4–34 cRSS (con-non) substrate (cryptic nonamer replaced with the consensus nonamer sequence). An extra cleavage product from the VH 4–34 cRSS (con-non) substrate in B, which is either due to nicking or hairpin formation, is marked with an asterisk. C, quantitation of the relative amounts of each product shown in A and B. The percentage of each cleavage product relative to the total amount of DNA in its respective sample is shown for each substrate. The heptamer and nonamer sequences of each substrate are given below the graph.

 
Sequence Determinants within the VH 4–34 Cryptic RSS Inhibit Hairpin Intermediate—We hypothesized that sequence determinants in the cRSS were prompting altered cleavage activity by the RAG proteins. To test this hypothesis, we used nicked cRSS substrates with either the cryptic heptamer or nonamer sequences replaced with the corresponding consensus RSS sequences (Fig. 8). Replacement of the cryptic nonamer with the consensus sequence restored the ability of the RAG proteins to form a DNA hairpin on this substrate (Fig. 8, A and C). This result demonstrates how alterations in the RSS, although distal from the cleavage site, can result in an alternative cleavage mechanism. In contrast, substituting in the consensus heptamer had no effect on the level of hairpin formation compared with the nonaltered cRSS (Fig. 8, A and C). It may be that the consensus heptamer cannot rescue hairpin formation, since the nick in the DNA substrate is located within the heptamer rather than at the 5' side. In addition, the consensus heptamer apparently suppressed nicking on the bottom strand, since diminished levels of the 33- and 34-nt bottom strand products were formed from the chimeric cRSS containing the consensus heptamer versus the wild type VH 4–34 cRSS (Fig. 8, B and C).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Receptor revision of the IgH gene, in which a portion of the VH gene in the combined V-D-J sequence is replaced with the corresponding region of another VH gene, has been demonstrated in peripheral B cells isolated from human tonsils (15), arthritic synovium (16), and NOD/SCID mice colonized with human cord blood mononuclear cells (17). The DNA sequence adjacent to the break point of the hybrid VH gene bears striking homology to the conventional RSS. This is particularly the case for the cRSS located 3' to CDR2 in the VH 4–34 family (Fig. 1B and supplemental Fig. 1). This cRSS contains both RSS-like heptamer and nonamer sequences separated by 13 base pairs and is thus comparable with the 12-RSS. Given that the sequence of the VH 4–34 cRSS nearly qualifies as a conventional RSS, we tested the ability of the RAG proteins to cleave at this site using standard in vitro DNA cleavage assays. In summary, we found that RAG1 and RAG2 together created double-stranded breaks near the 5' side of the cryptic heptamer. The RAG proteins produced DSB on the cRSS via a nick-nick mechanism, in which both the top and bottom strands are nicked, rather than the standard nick-hairpin mechanism that has been well characterized for the conventional RSS. The nick-nick reaction occurred on the cRSS even in the presence of Mn2+, which in the conventional reaction allows abundant hairpin formation on a single RSS. In addition, the nick-nick mechanism was preferred even with nicked substrates, which would normally facilitate hairpin formation (30, 31). Moreover, the cRSS did not enhance DSB formation at a 23-RSS and thus does not adhere to the 12/23 rule. Overall, a nick-nick mechanism would have fewer restrictions than the 12/23-restricted nick-hairpin mechanism in the creation of DSB.

The RAG proteins have been found to perform multiple DNA cleavage activities. Besides creation of DSB via the nick-hairpin mechanism, these include formation of hybrid and open and shut joints (32) as well as in vitro demonstrations of transposition (33–35), hairpin nicking (28, 29), and structure-dependent cleavage at single-stranded/double-stranded junctions (36, 37). Indeed, the RAG proteins in vitro can create DSB on a final product of V(D)J recombination, the signal joint, via a nick-nick mechanism (38). However, it is difficult to draw comparisons between the VH 4–34 cRSS substrate used in the present study and the signal joint, since the latter consists of a 12/23 RSS pair fused heptamer-to-heptamer.

RAG-mediated cleavage of the VH 4–34 cRSS differs from cleavage of conventional RSS in three significant ways. First, with the cRSS substrate, nicking occurred two bases 3' to the expected site on the top strand (at CA{downarrow}C rather than {downarrow}CAC). Second, the formation of the hairpin was very inefficient, and third, nicks were produced at the GTG sequence on the complementary strand. Comparison of the VH 4–34 cRSS sequence with that of the consensus 12-RSS has yielded few clues into the difference in cleavage mechanisms on the two DNA substrates. The heptamer of the VH 4–34 cRSS (CACCATA) contains four of the seven conserved bases in the consensus heptamer (CACAGTG). Significantly, the first three bases are CAC, the most highly conserved bases in the conventional RSS. The fourth base (C-4) of the VH 4–34 cRSS heptamer is rarely observed at this position in standard intergenic RSS (see the international ImMunoGeneTics data base on the World Wide Web at imgt.cines.fr (Initiator and coordinator Marie-Paule Lefranc)); however, a C at this position does not prevent normal V(D)J recombination reactions (40). Strikingly, the VH 4–34 cRSS nonamer contains seven of the nine bases present in the consensus RSS nonamer. However, one of the differences is at position 5, which is G in VH 4–34 cRSS versus a highly conserved A in the consensus RSS. Despite this difference, previous studies demonstrated that mutation of the entire nonamer in a consensus 12-RSS could still be cleaved by the RAG proteins via the nick-hairpin mechanism, with only a 5–10-fold decrease in product formation (30, 31). In addition to the heptamer and nonamer sequences, it is necessary to inspect the intervening spacer sequence as well as that 5' to the heptamer of the VH 4–34 cRSS. The 13-base pair spacer between the heptamer and nonamer sequences of the VH 4–34 cRSS falls within the acceptable length range for a 12-RSS (12 ± 1 base pairs). Although poorly conserved, the spacer sequences of conventional RSS have recently been found to contain sequence elements that contribute significantly to RAG-mediated cleavage activity (41, 42). The VH 4–34 cRSS spacer did not contain sequences that would diminish RAG-mediated cleavage or any unusual characteristics that could result in an altered cleavage mechanism. The sequence 5' to conventional RSS heptamer corresponds to the V, D, or J gene segment and is referred to as the coding flank. Previous studies have demonstrated that the sequence of the coding flank can reduce or eliminate DSB (43–45). However, the sequence 5' to the VH 4–34 cRSS bears greater resemblance to "good" coding flank sequences, which allow abundant hairpin formation.

Even with the similarities in sequence between VH 4–34 cRSS and the consensus 12-RSS, the RAG proteins processed the two substrates differently. The difference in nicking site at the heptamers, with nicking occurring two bases into the cRSS at CA{downarrow}C rather than 5' to the heptamer can be accounted for from previous reports. Extensive studies with mutated RSS substrates have shown that the consensus heptamer sequence is not critical to the nicking reaction (30, 31). Substrates containing the conserved nonamer sequence, but no heptamer site, still yielded RAG-mediated nicking ~17–19 bases 5' to the nonamer, where the heptamer would normally be positioned in a 12-RSS substrate. It is possible that with the VH 4–34 cRSS, the CACCATA heptamer may not be optimal for directing cleavage 5' to the CAC. Alternatively, the 13-base pair spacer and/or the cryptic nonamer sequence may result in misalignment of the RAG proteins relative to the CAC sequence.

The reason for the lack of hairpin formation with the VH 4–34 cRSS substrate is less apparent, especially in light of the finding that other cRSS do apparently participate in standard nick-hairpin V(D)J recombination with conventional RSS, leading to the formation of "expanded" VH-VH hybrids (13, 46). Previous studies have indicated that the CAC sequence directly 3' to the nicked site is critical for hairpin formation (30). However, nicked VH 4–34 cRSS substrates with the nick placed at {downarrow}CAC were only poorly converted to hairpins by the RAG proteins (data not shown). This is consistent with the possibility that the RAG proteins are not aligned at the CAC sequence of the cryptic heptamer, as they would be in the conventional RSS.

Last, altered binding of the RAG proteins to the cRSS may explain how this substrate is nicked on both strands. Although the RAG proteins are positioned two bases into the cryptic heptamer, they may still induce base unpairing of the cRSS. RAG-induced base unpairing of the conventional RSS has been previously proposed based on results from RAG-mediated DNA cleavage of altered substrates (30, 31), from the occurrence of hypersensitive bands in DNA footprinting experiments (47, 48), and from the preferential interaction of a RAG1 domain with single-stranded over double-stranded RSS (49–51). Thus, sequence restrictions of the VH 4–34 cRSS may prevent hairpin formation; however, RAG-induced base unpairing could allow nicking on the opposite strand. This would not be unexpected, given that the RAG proteins can cleave at double-stranded/single-stranded junctions with no apparent sequence specificity (36, 37). Overall, we propose that there may be altered conformations or orientation of the RAG subunits on the VH 4–34 cRSS, which allows nicking reactions on both DNA strands to be substantially preferred over the nick-hairpin reaction.

Establishment of a nick-nick mechanism in RAG-mediated cleavage of the VH 4–34 cRSS introduces new possibilities for the formation of hybrid IgH genes. It was previously hypothesized that VH gene replacement occurred through a hybrid joint recombination reaction (15). In hybrid joint formation, the RAG proteins bind to both a 12- and 23-RSS in a synaptic complex, as in normal V(D)J recombination (32). However, after RAG-mediated cleavage of both RSS via the standard nick-hairpin mechanism, hybrid joints are formed in which a blunt signal end from one RSS is joined to a hairpin-containing coding end of the partner RSS. To achieve VH gene replacement that was observed in human tonsillar B cells, such hybrid joint formation would require a synaptic complex of RAG proteins with two cRSS located on different VH genes. Whereas synaptic complexes may occur infrequently between cRSS and a 12 or 23-RSS, the likelihood of synapsis between two VH cRSS is unlikely. Our data also show that the VH 4–34 cRSS apparently inhibits normal synaptic complex function, so a V(D)J-like recombination mechanism seems unlikely.

Here, we propose that double-stranded cleavage of the VH 4–34 cRSS via a nick-nick mechanism would allow joining to occur without the stringent requirements for synaptic complex formation. With the nick-nick mechanism, DSB on two cRSS could be brought together and repaired by NHEJ. However, NHEJ would require that DSB occur nearly simultaneously on cRSS on two different VH genes. An attractive alternative is that repair of the DSB at the cRSS could occur by HRR, which would necessitate that only one of the cRSS be cleaved by the RAG proteins. Given the high sequence similarity among VH genes of the same family and between different families, HRR could occur between misselected homologous sequences of different VH genes rather than strictly between the two alleles of the same gene. Indeed, if repair occurs after V(D)J recombination in G1 or S phase, there is no complementary allele to the rearranged VH gene. This model has several advantages over previous models invoking V(D)J-like recombination events. Using HRR, hybrids can be formed with both upstream and downstream VH genes, since repair can be directed in trans; V(D)J recombination occurs in cis, and thus hybrids can only be formed with upstream VH genes, since the downstream genes are deleted in the initial V to DJ rearrangement. Indeed, the partial VH gene replacement events observed in human tonsillar B cells showed that the donor VH gene segment could be either upstream or downstream of the target VH gene segment in the germ line sequence (15). This observation indicates that HRR, instead of NHEJ, would be the likely repair pathway that completes such receptor revision events. Moreover, there is precedent for the repair of some RAG-mediated cleavage events by the HRR pathway. It has recently been shown that RAG-mediated DNA breaks may be repaired by the HRR pathway if the RAG-RSS complex is not an optimal interaction (52) or if NHEJ factors are deficient (53). Since we have proposed here that the RAG activity on the VH 4–34 cRSS is due to an altered interaction, access of HRR factors to the DNA ends may be allowed that would not normally occur in a complex containing conventional RSS.

Further studies are necessary to identify determinants that lead to the preference for nick-nick over "normal" nick-hairpin RAG-mediated DNA cleavage. This nick-nick mechanism suggests an alternate pathway by which the RAG proteins could influence the antigen receptor repertoire, such as by mediating the receptor revision events previously reported (15, 17). Furthermore, putative cRSS sites have been identified throughout the genome (54). Inappropriate V(D)J recombination between these cRSS and conventional RSS may lead to deleterious chromosomal translocations (39). If the RAG proteins cleave other cRSS via the nick-nick mechanism, then this would add another dimension to the oncogenic potential of RAG activity in initiating events that lead to lymphomagenesis.


   FOOTNOTES
 
* This work was supported by Oklahoma Center for Advancement in Science and Technology awards for project numbers HR02-008 and HR05-101 (to K. K. R.) and National Institutes of Health Grant AI-12127 (to J. D. C.). 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. Back

Formula The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. 1 and 2. Back

1 Supported by a National Science Foundation Graduate Research Fellowship. Back

2 Present address: Dept. of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73190. Back

3 To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, 940 S. L. Young Blvd., Oklahoma City, OK 73190. Tel.: 405-271-2227 (ext. 1248); Fax: 405-271-3139; E-mail: karla-rodgers{at}ouhsc.edu.

4 The abbreviations used are: GC, germinal center(s); CDR, complementarity-determining region; RSS, recombination signal sequence; cRSS, cryptic recombination signal sequence; DSB, double-stranded breaks; EMSA, electrophoretic mobility shift assay; GST, glutathione S-transferase; HMGB1, high mobility group protein B1; HRR, homologous recombination repair; MBP, maltose-binding protein; NHEJ, nonhomologous end-joining; nt, nucleotide; IgH, Ig heavy chain; MHMN, mutant heptamer and nonamer. Back


   ACKNOWLEDGMENTS
 
We thank David Schatz and Patrick Swanson for plasmids used in the transient expression in 293T cells of GST-core RAG2 and MBP-core RAG1, respectively. We thank Brian Sauer for helpful discussions.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Lewis, S. M. (1994) Adv. Immunol. 56, 27–150[Medline] [Order article via Infotrieve]
  2. Bassing, C., Swat, W., and Alt, F. (2002) Cell 109, (suppl.) S45–S55[CrossRef][Medline] [Order article via Infotrieve]
  3. Gellert, M. (2002) Annu. Rev. Biochem. 71, 101–132[CrossRef][Medline] [Order article via Infotrieve]
  4. Fugmann, S. D., Lee, A. I., Shockett, P. E., Villey, I. J., and Schatz, D. G. (2000) Annu. Rev. Immunol. 18, 495–527[CrossRef][Medline] [Order article via Infotrieve]
  5. Hesslein, D., and Schatz, D. (2001) Adv. Immunol. 78, 169–232[Medline] [Order article via Infotrieve]
  6. de Villartay, J. P., Fischer, A., and Durandy, A. (2003) Nat. Rev. Immunol. 3, 962–972[CrossRef][Medline] [Order article via Infotrieve]
  7. McBlane, J. F., van Gent, D. C., Ramsden, D. A., Romeo, C., Cuomo, C. A., Gellert, M., and Oettinger, M. A. (1995) Cell 83, 387–395[CrossRef][Medline] [Order article via Infotrieve]
  8. Eastman, Q. M., and Schatz, D. G. (1997) Nucleic Acids Res. 25, 4370–4378[Abstract/Free Full Text]
  9. Yu, K., and Lieber, M. R. (2000) Mol. Cell. Biol. 20, 7914–7921[Abstract/Free Full Text]
  10. Casellas, R., Shih, T., Kleinewietfeld, M., Rakonjac, J., Nemazee, D., Rajewsky, K., and Nussenzweig, M. (2001) Science 291, 1503–1505[Abstract/Free Full Text]
  11. Nemazee, D., and Hogquist, K. (2003) Curr. Opin. Immunol. 15, 182–189[CrossRef][Medline] [Order article via Infotrieve]
  12. Radic, M., and Zouali, M. (1996) Immunity 5, 505–511[CrossRef][Medline] [Order article via Infotrieve]
  13. Zhang, Z., Zemlin, M., Wang, Y. H., Munfus, D., Huye, L. E., Findley, H. W., Bridges, L. S., Roth, D. B., Burrows, P. D., and Cooper, M. D. (2003) Immunity 19, 21–31[CrossRef][Medline] [Order article via Infotrieve]
  14. Kouskoff, V., and Nemazee, D. (2001) Life Sci. 69, 1105–1113[CrossRef][Medline] [Order article via Infotrieve]
  15. Wilson, P. C., Wilson, K., Liu, Y. J., Banchereau, J., Pascual, V., and Capra, J. D. (2000) J. Exp. Med. 191, 1881–1894[Abstract/Free Full Text]
  16. Itoh, K., Meffre, E., Albesiano, E., Farber, A., Dines, D., Stein, P., Asnis, S., Furie, R., Jain, R., and Chiorazzi, N. (2000) J. Exp. Med. 192, 1151–1164[Abstract/Free Full Text]
  17. Kolar, G. R., and Capra, J. D. (2004) Scand. J. Immunol. 60, 108–111[CrossRef][Medline] [Order article via Infotrieve]
  18. Lenze, D., Greiner, A., Knorr, C., Anagnostopoulos, I., Stein, H., and Hummel, M. (2003) Mol. Pathol. 56, 249–255[Abstract/Free Full Text]
  19. Bellan, C., Lazzi, S., Zazzi, M., Lalinga, A. V., Palummo, N., Galieni, P., Marafioti, T., Tonini, T., Cinti, C., Leoncini, L., Pileri, S. A., and Tosi, P. (2002) Diagn. Mol. Pathol. 11, 2–8[CrossRef][Medline] [Order article via Infotrieve]
  20. Jankovic, M., Casellas, R., Yannoutsos, N., Wardemann, H., and Nussenzweig, M. (2004) Annu. Rev. Immunol. 22, 485–501[CrossRef][Medline] [Order article via Infotrieve]
  21. Hillion, S., Rochas, C., Youinou, P., and Jamin, C. (2005) Ann. N. Y. Acad. Sci. 1050, 10–18[Abstract/Free Full Text]
  22. Darlow, J. M., and Stott, D. I. (2005) Immunology 114, 155–165[CrossRef][Medline] [Order article via Infotrieve]
  23. Arbuckle, J. L., Fauss, L. J., Simpson, R., Ptaszek, L. M., and Rodgers, K. K. (2001) J. Biol. Chem. 276, 37093–37101[Abstract/Free Full Text]
  24. Spanopoulou, E., Zaitseva, F., Wang, F.-H., Santagata, S., Baltimore, D., and Panayotou, G. (1996) Cell 87, 263–276[CrossRef][Medline] [Order article via Infotrieve]
  25. Godderz, L. J., Rahman, N. S., Risinger, G. M., Arbuckle, J. L., and Rodgers, K. K. (2003) Nucleic Acids Res. 31, 2014–2023[Abstract/Free Full Text]
  26. Swanson, P. C. (2002) Mol. Cell. Biol. 22, 1340–1351[Abstract/Free Full Text]
  27. Santagata, S., Aidinis, V., and Spanopoulou, E. (1998) J. Biol. Chem. 273, 16325–16331[Abstract/Free Full Text]
  28. Besmer, E., Mansilla-Soto, J., Cassard, S., Sawchuk, D. J., Brown, G., Sadofsky, M., Lewis, S. M., Nussenzweig, M. C., and Cortes, P. (1998) Mol. Cell 2, 817–828[CrossRef][Medline] [Order article via Infotrieve]
  29. Shockett, P. E., and Schatz, D. G. (1999) Mol. Cell. Biol. 19, 4159–4166[Abstract/Free Full Text]
  30. Ramsden, D. A., McBlane, J. F., van Gent, D. C., and Gellert, M. (1996) EMBO J. 15, 3197–3206[Medline] [Order article via Infotrieve]
  31. Cuomo, C. A., Mundy, C. L., and Oettinger, M. A. (1996) Mol. Cell. Biol. 16, 5683–5690[Abstract]
  32. Melek, M., Gellert, M., and van Gent, D. C. (1998) Science 280, 301–303[Abstract/Free Full Text]
  33. Agrawal, A., Eastman, Q. M., and Schatz, D. G. (1998) Nature 394, 744–751[CrossRef][Medline] [Order article via Infotrieve]
  34. Hiom, K., Melek, M., and Gellert, M. (1998) Cell 94, 463–470[CrossRef][Medline] [Order article via Infotrieve]
  35. Shih, I.-H., Melek, M., Jayaratne, N. D., and Gellert, M. (2002) EMBO J. 21, 6625–6633[CrossRef][Medline] [Order article via Infotrieve]
  36. Santagata, S., Besmer, E., Villa, A., Bozzi, F., Allingham, J. S., Sobacchi, C., Haniford, D. B., Vezzoni, P., Nussenzweig, M. C., Pan, Z.-Q., and Cortes, P. (1999) Mol. Cell 4, 935–947[CrossRef][Medline] [Order article via Infotrieve]
  37. Raghavan, S. C., Swanson, P. C., Ma, Y., and Lieber, M. R. (2005) Mol. Cell. Biol. 25, 5904–5919[Abstract/Free Full Text]
  38. Neiditch, M. B., Lee, G. S., Huye, L. E., Brandt, V. L., and Roth, D. B. (2002) Mol. Cell 9, 871–878[CrossRef][Medline] [Order article via Infotrieve]
  39. Greaves, M. F., and Wiemels, J. (2003) Nat. Rev. Cancer 3, 639–649[CrossRef][Medline] [Order article via Infotrieve]
  40. Cowell, L. G., Davila, M., Yang, K., Kepler, T. B., and Kelsoe, G. (2003) J. Exp. Med. 197, 207–220[Abstract/Free Full Text]
  41. Lee, A. I., Fugmann, S. D., Cowell, L. G., Ptaszek, L. M., Kelsoe, G., and Schatz, D. G. (2003) PLoS Biol. 1, 56–69